Methods in

Molecular Biology 1672

Marco Muzi-Falconi
Grant W. Brown Editors

Genome
Instability
Methods and Protocols
METHODS IN MOLECULAR BIOLOGY

Series Editor
John M. Walker
School of Life and Medical Sciences
University of Hertfordshire
Hatfield, Hertfordshire, AL10 9AB, UK

For further volumes:

http://www.springer.com/series/7651
 Genome Instability

Methods and Protocols

Edited by

Marco Muzi-Falconi
Dipartimento di Bioscienze, Università degli Studi di Milano, Milano, Milano, Italy

Grant W. Brown
Donnelly Centre and Department of Biochemistry, University of Toronto, Toronto, ON, Canada
Editors
Marco Muzi-Falconi Grant W. Brown
Dipartimento di Bioscienze Donnelly Centre and Department of Biochemistry
Università degli Studi di Milano University of Toronto
Milano, Milano, Italy Toronto, ON, Canada

ISSN 1064-3745 ISSN 1940-6029 (electronic)

Methods in Molecular Biology
ISBN 978-1-4939-7305-7 ISBN 978-1-4939-7306-4 (eBook)
DOI 10.1007/978-1-4939-7306-4
Library of Congress Control Number: 2017953940

© Springer Science+Business Media LLC 2018

This work is subject to copyright. All rights are reserved by the Publisher, whether the whole or part of the material is
concerned, specifically the rights of translation, reprinting, reuse of illustrations, recitation, broadcasting, reproduction
on microfilms or in any other physical way, and transmission or information storage and retrieval, electronic adaptation,
computer software, or by similar or dissimilar methodology now known or hereafter developed.
The use of general descriptive names, registered names, trademarks, service marks, etc. in this publication does not imply,
even in the absence of a specific statement, that such names are exempt from the relevant protective laws and regulations
and therefore free for general use.
The publisher, the authors and the editors are safe to assume that the advice and information in this book are believed to
be true and accurate at the date of publication. Neither the publisher nor the authors or the editors give a warranty,
express or implied, with respect to the material contained herein or for any errors or omissions that may have been made.
The publisher remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Printed on acid-free paper

This Humana Press imprint is published by Springer Nature

The registered company is Springer Science+Business Media, LLC
The registered company address is: 233 Spring Street, New York, NY 10013, U.S.A.
Preface

The integrity of the genome is a fundamental determinant of cellular identity, cellular fitness,
and interactions between a cell and its environment. The study of genome integrity is now a
mature field, but one marked by continuous innovations in techniques, technology, and
systems, both in vitro and in vivo. We present 42 methods and protocols to analyze diverse
aspects of genome instability.
Beginning in the realm of mutagenesis and repair, we present classic genetic assays to
detect chromosome loss, mutation, and genome rearrangements, whole genome
approaches to mapping base modifications and repair events, and a method for analyzing
engineered base lesions in the genomic context. Methods to quantify and analyze the
properties of DNA double-strand breaks include traditional and single-molecule approaches
to measure double-strand break resection, and modern methods to map double-strand
breaks at high resolution and high sensitivity. Given the importance of DNA replication
errors as a source of genome instability, we include methods to profile replication, to probe
replication and replication proteins strand specifically, to analyze replication intermediates at
high resolution, and to specifically perturb DNA synthesis at specific sites. The increasing
interest in the role of ribonucleotides and RNA–DNA hybrids in genome instability is
reflected in methods to detect and map ribonucleotides and RNA-DNA hybrids.
Techniques to study genome instability at specialized regions, in particular the telomeres
and triplet nucleotide repeats, are presented and include molecular biological, genetic, and
imaging-based methods. The application of imaging techniques to study genome instability
has become common in the field. We present fluorescence microscopic techniques to detect
and analyze genome instability, including single-molecule and single-cell analysis, as well as
high-resolution methods to probe DNA structural properties. Finally, the contributions of
genomic and proteomic approaches to identifying and defining genome instability pathways
and networks are reflected in procedures for measuring cell fitness, protein interactions,
gene and protein expression, protein-DNA interactions, and protein modifications, on a
genome/proteome scale.
Together, the methods and protocols here form a comprehensive resource for the
discovery and analysis of the proteins and pathways that are critical for stable maintenance
of the genome.

Milano, Italy Marco Muzi-Falconi

Toronto, Canada Grant W. Brown

v
Contents

Preface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . v
Contributors. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xi
1 The A-Like Faker Assay for Measuring Yeast Chromosome III Stability. . . . . . . . 1
Carolina A. Novoa, J. Sidney Ang, and Peter C. Stirling
2 The Chromosome Transmission Fidelity Assay for Measuring
Chromosome Loss in Yeast. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
Supipi Duffy and Philip Hieter
3 Measuring Mutation Rates Using the Luria-Delbr€ uck Fluctuation Assay . . . . . . . 21
Gregory I. Lang
4 Molecular Genetic Characterization of Mutagenesis Using
a Highly Sensitive Single-Stranded DNA Reporter System
in Budding Yeast . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33
Kin Chan
5 Analyzing Genome Rearrangements in Saccharomyces cerevisiae . . . . . . . . . . . . . . . 43
Anjana Srivatsan, Christopher D. Putnam, and Richard D. Kolodner
6 High-Resolution Mapping of Modified DNA Nucleobases
Using Excision Repair Enzymes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63
Monica Ransom, D. Suzi Bryan, and Jay R. Hesselberth
7 Integrated Microarray-based Tools for Detection of Genomic
DNA Damage and Repair Mechanisms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 77
Patrick van Eijk, Yumin Teng, Mark R. Bennet, Katie E. Evans,
James R. Powell, Richard M. Webster, and Simon H. Reed
8 Study of UV-induced DNA Repair Factor Recruitment: Kinetics
and Dynamics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 101
Sarah Sertic, Stefania Roma, Paolo Plevani, Federico Lazzaro,
and Marco Muzi-Falconi
9 Inserting Site-Specific DNA Lesions into Whole Genomes . . . . . . . . . . . . . . . . . . . 107
Vincent Pagès and Robert P. Fuchs
10 A qPCR-Based Protocol to Quantify DSB Resection . . . . . . . . . . . . . . . . . . . . . . . . 119
Matteo Ferrari, Shyam Twayana, Federica Marini,
and Achille Pellicioli
11 Alkaline Denaturing Southern Blot Analysis to Monitor Double-Strand
Break Processing. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 131
Chiara Vittoria Colombo, Luca Menin, and Michela Clerici
12 Single Molecule Analysis of Resection Tracks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 147
Pablo Huertas and Andrés Cruz-Garcı́a
13 Mapping DNA Breaks by Next-Generation Sequencing . . . . . . . . . . . . . . . . . . . . . 155
Laura Baranello, Fedor Kouzine, Damian Wojtowicz, Kairong Cui,
Keji Zhao, Teresa M. Przytycka, Giovanni Capranico, and David Levens

vii
viii Contents

14 Genome-Wide Profiling of DNA Double-Strand Breaks

by the BLESS and BLISS Methods. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 167
Reza Mirzazadeh, Tomasz Kallas, Magda Bienko,
and Nicola Crosetto
15 DNA Replication Profiling Using Deep Sequencing. . . . . . . . . . . . . . . . . . . . . . . . . 195
Xanita Saayman, Cristina Ramos-Pérez, and Grant W. Brown
16 Quantitative Bromodeoxyuridine Immunoprecipitation Analyzed
by High-Throughput Sequencing (qBrdU-Seq or QBU) . . . . . . . . . . . . . . . . . . . . 209
Joanna E. Haye-Bertolozzi and Oscar M. Aparicio
17 Strand-Specific Analysis of DNA Synthesis and Proteins Association
with DNA Replication Forks in Budding Yeast . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 227
Chuanhe Yu, Haiyun Gan, and Zhiguo Zhang
18 Analysis of Replicative Polymerase Usage by Ribonucleotide
Incorporation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 239
Andrea Keszthelyi, Izumi Miyabe, Katie Ptasińska, Yasukazu Daigaku,
Karel Naiman, and Antony M. Carr
19 Dynamic Architecture of Eukaryotic DNA Replication Forks
In Vivo, Visualized by Electron Microscopy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 261
Ralph Zellweger and Massimo Lopes
20 A Molecular Toolbox to Engineer Site-Specific DNA Replication
Perturbation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 295
Nicolai B. Larsen, Ian D. Hickson, and Hocine W. Mankouri
21 Single Cell Gel Electrophoresis for the Detection of Genomic
Ribonucleotides . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 311
Barbara Kind, Christine Wolf, Kerstin Engel, Alexander Rapp,
M. Cristina Cardoso, and Min Ae Lee-Kirsch
22 Measuring the Levels of Ribonucleotides Embedded in Genomic DNA . . . . . . . 319
Alice Meroni, Giulia M. Nava, Sarah Sertic, Paolo Plevani,
Marco Muzi-Falconi, and Federico Lazzaro
23 Mapping Ribonucleotides Incorporated into DNA by Hydrolytic
End-Sequencing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 329
Clinton D. Orebaugh, Scott A. Lujan, Adam B. Burkholder,
Anders R. Clausen, and Thomas A. Kunkel
24 Detection of DNA-RNA Hybrids In Vivo . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 347
Marı́a Garcı́a-Rubio, Sonia I. Barroso, and Andrés Aguilera
25 Analysis of De Novo Telomere Addition by Southern Blot . . . . . . . . . . . . . . . . . . . 363
Diego Bonetti and Maria Pia Longhese
26 Assays to Study Repair of Inducible DNA Double-Strand Breaks
at Telomeres . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 375
Roxanne Oshidari and Karim Mekhail
 Contents ix

27 Telomerase RNA Imaging in Budding Yeast and Human Cells

by Fluorescent In Situ Hybridization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 387
David Guérit, Maxime Lalonde, and Pascal Chartrand
28 Methods to Study Repeat Fragility and Instability
in Saccharomyces cerevisiae. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 403
Erica J. Polleys and Catherine H. Freudenreich
29 Quantitative Analysis of the Rates for Repeat-Mediated Genome
Instability in a Yeast Experimental System. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 421
Elina A. Radchenko, Ryan J. McGinty, Anna Y. Aksenova,
Alexander J. Neil, and Sergei M. Mirkin
30 Measuring Dynamic Behavior of Trinucleotide Repeat Tracts
In Vivo in Saccharomyces cerevisiae . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 439
Gregory M. Williams and Jennifer A. Surtees
31 The Detection and Analysis of Chromosome Fragile Sites. . . . . . . . . . . . . . . . . . . . 471
Victoria A. Bjerregaard, Özg€ u n Özer, Ian D. Hickson,
and Ying Liu
32 Imaging of DNA Ultrafine Bridges in Budding Yeast . . . . . . . . . . . . . . . . . . . . . . . . 483
Oliver Quevedo and Michael Lisby
33 Detection of Ultrafine Anaphase Bridges. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 495
Anna H. Bizard, Christian F. Nielsen, and Ian D. Hickson
34 A Chromatin Fiber Analysis Pipeline to Model DNA Synthesis
and Structures in Fission Yeast. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 509
Sarah A. Sabatinos and Marc D. Green
35 Long-Term Imaging of DNA Damage and Cell Cycle Progression
in Budding Yeast Using Spinning Disk Confocal Microscopy. . . . . . . . . . . . . . . . . 527
Riccardo Montecchi and Etienne Schwob
36 The CellClamper: A Convenient Microfluidic Device for Time-Lapse
Imaging of Yeast . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 537
Gregor W. Schmidt, Olivier Frey, and Fabian Rudolf
37 Characterization of Structural and Configurational Properties
of DNA by Atomic Force Microscopy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 557
Alice Meroni, Federico Lazzaro, Marco Muzi-Falconi,
and Alessandro Podestà
38 Genome-Wide Quantitative Fitness Analysis (QFA) of Yeast Cultures . . . . . . . . . 575
Eva-Maria Holstein, Conor Lawless, Peter Banks, and David Lydall
39 Rewiring the Budding Yeast Proteome using Synthetic
Physical Interactions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 599
Guðjón Ólafsson and Peter H. Thorpe
40 Reporter-Based Synthetic Genetic Array Analysis: A Functional
Genomics Approach for Investigating Transcript or Protein
Abundance Using Fluorescent Proteins in Saccharomyces cerevisiae . . . . . . . . . . . . 613
Hendrikje Göttert, Mojca Mattiazzi Usaj, Adam P. Rosebrock,
and Brenda J. Andrews
x Contents

41 Statistical Analysis and Quality Assessment of ChIP-seq Data

with DROMPA. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 631
Ryuichiro Nakato and Katsuhiko Shirahige
42 Quantitative Analysis of DNA Damage Signaling Responses
to Chemical and Genetic Perturbations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 645
Francisco M. Bastos de Oliveira, Dongsung Kim, Michael Lanz,
and Marcus B. Smolka

Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 661
Contributors

ANDRÉS AGUILERA  Centro Andaluz de Biologı́a Molecular y Medicina Regenerativa,

CABIMER, Universidad de Sevilla, Seville, Spain
ANNA Y. AKSENOVA  Amyloid Biology, Saint-Petersburg University, Saint Petersburg, Russia
BRENDA J. ANDREWS  The Donnelly Centre, University of Toronto, Toronto, ON, Canada;
Department of Molecular Genetics, University of Toronto, Toronto, ON, Canada
J. SIDNEY ANG  Michael Smith Laboratories, University of British Columbia, Vancouver,
Canada
OSCAR M. APARICIO  Molecular and Computational Biology Program, University of
Southern California, Los Angeles, CA, USA
PETER BANKS  Institute for Cell and Molecular Biosciences, Newcastle University, Medical
School, Newcastle upon Tyne, UK
LAURA BARANELLO  Laboratory of Pathology, NCI/NIH, Bethesda, MD, USA; Department
of Cellular and Molecular Biology, Karolinska Institutet, Stockholm, Sweden
SONIA I. BARROSO  Centro Andaluz de Biologı́a Molecular y Medicina Regenerativa,
CABIMER, Universidad de Sevilla, Seville, Spain
MARK R. BENNET  Division of Cancer and Genetics, School of Medicine, Cardiff University,
Cardiff, UK
MAGDA BIENKO  Science for Life Laboratory, Department of Medical Biochemistry and
Biophysics, Karolinska Institutet, Stockholm, Sweden
ANNA H. BIZARD  Department of Cellular and Molecular Medicine, Center for Chromosome
Stability, University of Copenhagen, Panum Institute, Copenhagen N, Denmark
VICTORIA A. BJERREGAARD  Department of Cellular and Molecular Medicine, Center for
Chromosome Stability, University of Copenhagen, Copenhagen N, Denmark
DIEGO BONETTI  Institute for Molecular Biology (IMB) gGMBH, Mainz, Germany
GRANT W. BROWN  Donnelly Centre and Department of Biochemistry, University of Toronto,
Toronto, ON, Canada
D. SUZI BRYAN  Department of Biochemistry and Molecular Genetics, Program in Molecular
Biology, University of Colorado School of Medicine, Aurora, CO, USA
ADAM B. BURKHOLDER  Integrative Bioinformatics, National Institute for Environmental
Health Sciences, National Institute of Health (NIH), Research Triangle Park, NC, USA
GIOVANNI CAPRANICO  Department of Pharmacy and Biotechnology, University of Bologna,
Bologna, Italy
M. CRISTINA CARDOSO  Department of Biology, Technische Universit€ a t Darmstadt,
Darmstadt, Germany
ANTONY M. CARR  University of Sussex, Brighton, UK
KIN CHAN  Department of Biochemistry, Microbiology, and Immunology, Faculty of
Medicine, University of Ottawa, Ottawa, Canada
PASCAL CHARTRAND  Department of Biochemistry and Molecular Medicine, Université de
Montréal, Montréal, QC, Canada
ANDERS R. CLAUSEN  Department of Medical Biochemistry and Cell Biology, University of
Gothenburg, Gothenburg, Sweden
MICHELA CLERICI  Dipartimento di Biotecnologie e Bioscienze, Università di Milano-
Bicocca, Milano, Italy

xi
xii Contributors

CHIARA VITTORIA COLOMBO  Dipartimento di Biotecnologie e Bioscienze, Università di

Milano-Bicocca, Milano, Italy
NICOLA CROSETTO  Science for Life Laboratory, Department of Medical Biochemistry and
Biophysics, Karolinska Institutet, Stockholm, Sweden
ANDRÉS CRUZ-GARCÍA  Centro Andaluz de Biologı́a Molecular y Medicina Regenerativa-
CABIMER, Universidad de Sevilla-CSIC-Universidad Pablo de Olavide, Sevilla, Spain
KAIRONG CUI  Systems Biology Center, NCI/NIH, Bethesda, MD, USA
YASUKAZU DAIGAKU  Frontier Research Institute for Interdisciplinary Sciences, Tohoku
University, Sendai, Japan; Graduate School of Life Sciences, Tohoku University, Sendai,
Japan
SUPIPI DUFFY  Michael Smith Laboratories, University of British Columbia, Vancouver, BC,
Canada
PATRICK VAN EIJK  Division of Cancer and Genetics, School of Medicine, Cardiff University,
Cardiff, UK
KERSTIN ENGEL  Department of Pediatrics, Medizinische Fakult€ at Carl Gustav Carus,
Technische Universit€a t Dresden, Dresden, Germany
KATIE E. EVANS  Division of Cancer and Genetics, School of Medicine, Cardiff University,
Cardiff, UK
MATTEO FERRARI  Dipartimento di Bioscienze, Università degli Studi di Milano, Milano,
Italy
CATHERINE H. FREUDENREICH  Department of Biology, Tufts University, Medford, MA,
USA
OLIVIER FREY  Department of Biosystems Science and Engineering, ETH Zurich, Basel,
Switzerland
ROBERT P. FUCHS  Cancer Research Center of Marseille, Team DNA Damage Tolerance,
CNRS, UMR7258, Marseille, France; Inserm, U1068, Marseille, France; Institut Paoli-
Calmettes, Marseille, France; Aix Marseille Univ., UM 105, Marseille, France
HAIYUN GAN  Institute for Cancer Genetics, Columbia University, New York, NY, USA;
Department of Pediatrics, Columbia University, New York, NY, USA; Department of
Genetics and Development, Columbia University, New York, NY, USA; Irving Cancer
Research CenterNew York, NY, USA
MARÍA GARCÍA-RUBIO  Centro Andaluz de Biologı́a Molecular y Medicina Regenerativa,
CABIMER, Universidad de Sevilla, Seville, Spain
HENDRIKJE GÖTTERT  The Donnelly Centre, University of Toronto, Toronto, ON, Canada;
Department of Molecular Genetics, University of Toronto, Toronto, ON, Canada
MARC D. GREEN  Royal Ontario Museum, Toronto, ON, Canada
DAVID GUÉRIT  Department of Biochemistry and Molecular Medicine, Université de
Montréal, Montréal, QC, Canada
JOANNA E. HAYE-BERTOLOZZI  Molecular and Computational Biology Program, University
of Southern California, Los Angeles, CA, USA
JAY R. HESSELBERTH  Department of Biochemistry and Molecular Genetics, Program in
Molecular Biology, University of Colorado School of Medicine, Aurora, CO, USA
IAN D. HICKSON  Center for Chromosome Stability, Department of Cellular and Molecular
Medicine, University of Copenhagen, Panum Institute, Copenhagen N, Denmark
PHILIP HIETER  Michael Smith Laboratories, University of British Columbia, Vancouver, BC,
Canada
EVA-MARIA HOLSTEIN  Institute for Cell and Molecular Biosciences, Newcastle University,
Medical School, Newcastle upon Tyne, UK
 Contributors xiii

PABLO HUERTAS  Centro Andaluz de Biologı́a Molecular y Medicina Regenerativa-

CABIMER, Universidad de Sevilla-CSIC-Universidad Pablo de Olavide, Sevilla, Spain
TOMASZ KALLAS  Science for Life Laboratory, Department of Medical Biochemistry and
Biophysics, Karolinska Institutet, Stockholm, Sweden
ANDREA KESZTHELYI  University of Sussex, Brighton, UK
DONGSUNG KIM  Department of Molecular Biology and Genetics, Weill Institute for Cell and
Molecular Biology, Cornell University, Ithaca, NY, USA
BARBARA KIND  Department of Pediatrics, Medizinische Fakult€ at Carl Gustav Carus,
Technische Universit€a t Dresden, Dresden, Germany
RICHARD D. KOLODNER  Ludwig Institute for Cancer Research, University of California
School of Medicine, San Diego, La Jolla, CA, USA; Department of Cellular and Molecular
Medicine, University of California School of Medicine, San Diego, La Jolla, CA, USA;
Moores-UCSD Cancer Center, University of California School of Medicine, San Diego, La
Jolla, CA, USA; Institute of Genomic Medicine, University of California School of
Medicine, San Diego, La Jolla, CA, USA
FEDOR KOUZINE  Laboratory of Pathology, NCI/NIH, Bethesda, MD, USA
THOMAS A. KUNKEL  Genome Integrity and Structural Biology Laboratory, National
Institute for Environmental Health Sciences, National Institute of Health (NIH),
Research Triangle Park, NC, USA
MAXIME LALONDE  Department of Biochemistry and Molecular Medicine, Université de
Montréal, Montréal, QC, Canada
GREGORY I. LANG  Department of Biological Sciences, Lehigh University, Bethlehem, PA,
USA
MICHAEL LANZ  Department of Molecular Biology and Genetics, Weill Institute for Cell and
Molecular Biology, Cornell University, Ithaca, NY, USA
NICOLAI B. LARSEN  The Novo Nordisk Foundation Center for Protein Research, Faculty of
Health and Medical Sciences, University of Copenhagen, Copenhagen, Denmark
CONOR LAWLESS  Institute for Cell and Molecular Biosciences, Newcastle University, Medical
School, Newcastle upon Tyne, UK
FEDERICO LAZZARO  Dipartimento di Bioscienze, Università degli Studi di Milano, Milano,
Italy
MIN AE LEE-KIRSCH  Department of Pediatrics, Medizinische Fakult€ a t Carl Gustav Carus,
Technische Universit€a t Dresden, Dresden, Germany
DAVID LEVENS  Laboratory of Pathology, NCI/NIH, Bethesda, MD, USA
MICHAEL LISBY  Department of Biology, University of Copenhagen, Copenhagen N,
Denmark; Department of Cellular and Molecular Medicine, Center for Chromosome
Stability, University of Copenhagen, Copenhagen N, Denmark
YING LIU  Department of Cellular and Molecular Medicine, Center for Chromosome
Stability, University of Copenhagen, Copenhagen N, Denmark
MARIA PIA LONGHESE  Dipartimento di Biotecnologie e Bioscienze, Università di Milano-
Bicocca, Milano, Italy
MASSIMO LOPES  Institute of Molecular Cancer Research, University of Zurich, Zurich,
Switzerland
SCOTT A. LUJAN  Genome Integrity and Structural Biology Laboratory, National Institute
for Environmental Health Sciences, National Institute of Health (NIH), Research
Triangle Park, NC, USA
DAVID LYDALL  Institute for Cell and Molecular Biosciences, Newcastle University, Medical
School, Newcastle upon Tyne, UK
xiv Contributors

HOCINE W. MANKOURI  The Novo Nordisk Foundation Center for Protein Research, Faculty
of Health and Medical Sciences, University of Copenhagen, Copenhagen, Denmark
FEDERICA MARINI  Dipartimento di Bioscienze, Università degli Studi di Milano, Milano,
Italy
RYAN J. MCGINTY  Department of Biology, Tufts University, Medford, MA, USA
KARIM MEKHAIL  Department of Laboratory Medicine and Pathobiology, Faculty of
Medicine, University of Toronto, Toronto, ON, Canada
LUCA MENIN  Dipartimento di Biotecnologie e Bioscienze, Università di Milano-Bicocca,
Milano, Italy
ALICE MERONI  Dipartimento di Bioscienze, Università degli Studi di Milano, Milano, Italy
SERGEI M. MIRKIN  Department of Biology, Tufts University, Medford, MA, USA
REZA MIRZAZADEH  Science for Life Laboratory, Department of Medical Biochemistry and
Biophysics, Karolinska Institutet, Stockholm, Sweden
IZUMI MIYABE  University of Sussex, Brighton, UK
RICCARDO MONTECCHI  IGMM, CNRS, University of Montpellier, Montpellier, France
MARCO MUZI-FALCONI  Dipartimento di Bioscienze, Università degli Studi di Milano,
Milano, Italy
KAREL NAIMAN  University of Sussex, Brighton, UK
RYUICHIRO NAKATO  Institute of Molecular and Cellular Biosciences, The University of
Tokyo, Tokyo, Japan
GIULIA M. NAVA  Dipartimento di Bioscienze, Università degli Studi di Milano, Milano,
Italy
ALEXANDER J. NEIL  Department of Biology, Tufts University, Medford, MA, USA
CHRISTIAN F. NIELSEN  Department of Cellular and Molecular Medicine, Center for
Chromosome Stability, University of Copenhagen, Panum Institute, Copenhagen N,
Denmark
CAROLINA A. NOVOA  Terry Fox Laboratory, BC Cancer Agency, Vancouver, Canada
GUðJÓN ÓLAFSSON  Mitotic Control Laboratory, The Francis Crick Institute, London, UK
FRANCISCO M. BASTOS DE OLIVEIRA  Instituto de Biofı́sica Carlos Chagas Filho,
Universidade Federal do Rio de Janeiro, Cidade Universitária, Rio de Janeiro, Brazil
CLINTON D. OREBAUGH  Genome Integrity and Structural Biology Laboratory, National
Institute for Environmental Health Sciences, National Institute of Health (NIH),
Research Triangle Park, NC, USA
ROXANNE OSHIDARI  Department of Laboratory Medicine and Pathobiology, Faculty of
Medicine, University of Toronto, Toronto, ON, Canada
ÖZG€uN ÖZER  Department of Cellular and Molecular Medicine, Center for Chromosome
Stability, University of Copenhagen, Copenhagen N, Denmark
VINCENT PAGÈS  Cancer Research Center of Marseille, Team DNA Damage Tolerance,
CNRS, UMR7258, Marseille, France; Inserm, U1068, Marseille, France; Institut Paoli-
Calmettes, Marseille, France; Aix Marseille Univ., UM 105, Marseille, France
ACHILLE PELLICIOLI  Dipartimento di Bioscienze, Università degli Studi di Milano, Milano,
Italy
PAOLO PLEVANI  Dipartimento di Bioscienze, Università degli Studi di Milano, Milano,
Italy
ALESSANDRO PODESTÀ  Dipartimento di Fisica and C.I.Ma.I.Na, Università degli Studi di
Milano, Milano, Italy
ERICA J. POLLEYS  Department of Biology, Tufts University, Medford, MA, USA
 Contributors xv

JAMES R. POWELL  Division of Cancer and Genetics, School of Medicine, Cardiff University,
Cardiff, UK
TERESA M. PRZYTYCKA  Computational Biology Branch, NCI/NIH, Bethesda, MD, USA
KATIE PTASIŃSKA  University of Sussex, Brighton, UK
CHRISTOPHER D. PUTNAM  Ludwig Institute for Cancer Research, University of California
School of Medicine, San Diego, La Jolla, CA, USA; Department of Medicine, University of
California School of Medicine, San Diego, La Jolla, CA, USA
OLIVER QUEVEDO  Department of Biology, University of Copenhagen, Copenhagen N,
Denmark; Department of Cellular and Molecular Medicine, Center for Chromosome
Stability, University of Copenhagen, Copenhagen N, Denmark
ELINA A. RADCHENKO  Department of Biology, Tufts University, Medford, MA, USA
CRISTINA RAMOS-PÉREZ  Donnelly Centre and Department of Biochemistry, University of
Toronto, Toronto, ON, Canada
MONICA RANSOM  Department of Biochemistry and Molecular Genetics, Program in
Molecular Biology, University of Colorado School of Medicine, Aurora, CO, USA
ALEXANDER RAPP  Department of Biology, Technische Universit€ at Darmstadt, Darmstadt,
Germany
SIMON H. REED  Division of Cancer and Genetics, School of Medicine, Cardiff University,
Cardiff, UK
STEFANIA ROMA  Dipartimento di Bioscienze, Università degli Studi di Milano, Milano,
Italy
ADAM P. ROSEBROCK  The Donnelly Centre, University of Toronto, Toronto, ON, Canada;
Department of Molecular Genetics, University of Toronto, Toronto, ON, Canada;
Department of Pathology and University Cancer Center, Stony Brook Medicine, Stony
Brook, NY, USA
FABIAN RUDOLF  Department of Biosystems Science and Engineering, ETH Zurich, Basel,
Switzerland
XANITA SAAYMAN  Donnelly Centre and Department of Biochemistry, University of Toronto,
Toronto, ON, Canada
SARAH A. SABATINOS  Ryerson University, Toronto, ON, Canada
GREGOR W. SCHMIDT  Department of Biosystems Science and Engineering, ETH Zurich,
Basel, Switzerland
ETIENNE SCHWOB  IGMM, CNRS, University of Montpellier, Montpellier, France
SARAH SERTIC  Dipartimento di Bioscienze, Università degli Studi di Milano, Milano, Italy
KATSUHIKO SHIRAHIGE  Institute of Molecular and Cellular Biosciences, The University of
Tokyo, Tokyo, Japan
MARCUS B. SMOLKA  Department of Molecular Biology and Genetics, Weill Institute for Cell
and Molecular Biology, Cornell University, Ithaca, NY, USA
ANJANA SRIVATSAN  Ludwig Institute for Cancer Research, University of California School of
Medicine, San Diego, La Jolla, CA, USA
PETER C. STIRLING  Terry Fox Laboratory, BC Cancer Agency, Vancouver, Canada;
Department of Medical Genetics, University of British Columbia, Vancouver, Canada
JENNIFER A. SURTEES  Department of Biochemistry, School of Medicine and Biomedical
Sciences, State University of New York at Buffalo, Buffalo, NY, USA; Genetics, Genomics
and Bioinformatics Program, Jacobs School of Medicine and Biomedical Sciences, State
University of New York at Buffalo, Buffalo, NY, USA
YUMIN TENG  Division of Cancer and Genetics, School of Medicine, Cardiff University,
Cardiff, UK
xvi Contributors

PETER H. THORPE  Mitotic Control Laboratory, The Francis Crick Institute, London, UK
SHYAM TWAYANA  Dipartimento di Bioscienze, Università degli Studi di Milano, Milano,
Italy
MOJCA MATTIAZZI USAJ  The Donnelly Centre, University of Toronto, Toronto, ON, Canada
RICHARD M. WEBSTER  Division of Cancer and Genetics, School of Medicine, Cardiff
University, Cardiff, UK
GREGORY M. WILLIAMS  Department of Biochemistry, School of Medicine and Biomedical
Sciences, State University of New York at Buffalo, Buffalo, NY, USA
DAMIAN WOJTOWICZ  Laboratory of Pathology, NCI/NIH, Bethesda, MD, USA
CHRISTINE WOLF  Department of Pediatrics, Medizinische Fakult€ at Carl Gustav Carus,
Technische Universit€
a t Dresden, Dresden, Germany
CHUANHE YU  Department of Biochemistry and Molecular Biology, Mayo Clinic, Rochester,
MN, USA
RALPH ZELLWEGER  Institute of Molecular Cancer Research, University of Zurich, Zurich,
Switzerland
ZHIGUO ZHANG  Institute for Cancer Genetics, Columbia University, New York, NY, USA;
Department of Pediatrics, Columbia University, New York, NY, USA; Department of
Genetics and Development, Columbia University, New York, NY, USA; Irving Cancer
Research Center, New York, NY, USA
KEJI ZHAO  Systems Biology Center, Bethesda, MD, USA
 Chapter 1

The A-Like Faker Assay for Measuring Yeast Chromosome III

Stability
Carolina A. Novoa, J. Sidney Ang, and Peter C. Stirling

Abstract
The ability to rapidly assess chromosome instability (CIN) has enabled profiling of most yeast genes for
potential effects on genome stability. The A-like faker (ALF) assay is one of several qualitative and
quantitative marker loss assays that indirectly measure loss or conversion of genetic material using a
counterselection step. The ALF assay relies on the ability to count spurious mating events that occur
upon loss of the MATα locus of haploid Saccharomyces cerevisiae strains. Here, we describe the deployment
of the ALF assay for both rapid and simple qualitative, and more in-depth quantitative analysis allowing
determination of absolute ALF frequencies.

Key words Chromosome instability, Genome instability, Aneuploidy, Gene conversion, Gross chro-
mosomal rearrangement, A-like faker, Marker loss

1 Introduction

Baker’s Yeast (Saccharomyces cerevisiae) can divide stably as haploid

MATa or MATα cells. This phenotype is determined by the com-
position of the active mating locus MAT which carries one of two
mating type alleles. Silent copies of MATa and MATα also exist at
distal positions on the right and left arms of chromosome III and
are called HMR and HML, respectively [1]. In 1981, it was recog-
nized that the transcriptional program driven by MATa is constitu-
tive, such that deletions causing loss of the MATα locus lead to a
MATa phenotype [2]. While the mutant cells were able to mate,
they usually did not sporulate normally and thus were not truly
switched to normal MATa cells. As a result, these MATα mutants
were called A-Like Fakers (ALF). This model helped explain how
deletion alleles could lead to a mating type switch, including the
first of such observations, the so-called Hawthorne Deletion [3].
Subsequently, it was realized that the frequency of ALF in a cell
population reflected instability in yeast Chromosome III [4–6].
The outcomes of the ALF phenotype were finally characterized

Marco Muzi-Falconi and Grant W. Brown (eds.), Genome Instability: Methods and Protocols, Methods in Molecular Biology,
vol. 1672, DOI 10.1007/978-1-4939-7306-4_1, © Springer Science+Business Media LLC 2018

1
2 Carolina A. Novoa et al.

and applied in large-scale screening less than a decade ago [7].

Events leading to ALF cells include chromosome loss, MAT allele
disruption by deletions or unbalanced translocations, and gene
conversion from the silent mating type locus HMRa. The latter is
likely to be a product of inappropriate homologous recombination
between MATα and silent locus HMRa, and the subsequent gener-
ation of an active MATa locus [7].
The measurement of relative ALF frequency can be applied to
large collections of mutants using a qualitative patch-based assay
[7–9], and absolute frequencies can also be quantified using a more
demanding assay. The presence of ALF cells is detected by a mating
test involving selection of prototrophic mated products after expo-
sure to a MATα tester strain (YPH316, See Subheading 2.3). In the
protocol described here, MATα locus loss can be followed because
MATα strains are deficient in HIS3 (and carry other auxotrophies)
while the MATα tester strain is deficient in HIS1; therefore, only
diploid cells can synthesize histidine and grow normally on media
lacking amino acids (Fig. 1). Virtually any auxotrophic MATα strain
can be used for ALF, so long as it contains a functional HIS1 gene
to complement the mating tester. As the mating tester strain is a
constant across all control and experimental conditions, any change
in the frequency of ALF can be ascribed to the experimental strains
under study and compared with an otherwise isogenic control.
While here we primarily focus on assessing genomically integrated
mutant alleles, ALF can also be conducted on strains bearing plas-
mids or in the presence of alternative media which we highlight in
our notes (e.g., to induce expression, such as with galactose).

2 Materials

2.1 Lab Ware 1. Disposable 10 cm plastic petri dishes.

2. Autoclaved toothpicks or sticks.
3. Sterile 1.5 mL microcentrifuge tubes.
4. Sterile velvet squares (~15
15 cm).
5. Replica plating block.
6. 10 cm round filter paper discs.

2.2 Media 1. Yeast Peptone Dextrose (YPD) agar plates: 10 g/L yeast extract,
Preparation 20 g/L peptone, 0.33 g/L tryptophan, 20 g/L dextrose, 20 g/L
agar. All solutions dissolved in distilled H2O (see Notes 1 and 2).
(a) In separate autoclavable flasks, prepare yeast extract-pep-
tone-tryptophan broth, 4% agar, and 20% dextrose as
indicated above. Make sure to use at least one flask with
enough volume carrying capacity to hold the desired final
volume.
 The A-Like Faker Assay for Measuring Yeast Chromosome III Stability 3

Fig. 1 Schematic of the ALF assay genetics. A MATα query strain deleted for any gene (yfgΔ ¼ your favorite
gene) is grown and at some frequency loses the MATα locus (black arrow at left). A dotted box expands on
possible mechanisms of this event which are, from left to right, focal deletion or GCRs that disrupt MAT;
inappropriate recombination with HMRa, leading to gene conversion, and whole chromosome loss. That MAT
[null] strain is rescued by mating with a tester strain lacking only the HIS1 gene. Fully prototrophic diploids are
selected on minimal media and scored

(b) Autoclave at 121  C for at least 20 min, then cool to

~60  C.
(c) While stirring the autoclaved mixture contained in the
largest flask, mix in the other two solutions.
(d) While still warm, pour 20–25 mL of media into 10 cm
plastic petri dishes.
(e) Allow agar mixture to solidify and dry overnight on the
bench top before using.
(f) Unused YPD plates can be stored at 4  C for several
months.
2. YPD liquid media: 10 g/L yeast extract, 20 g/L peptone,
0.32 g/L tryptophan, 20 g/L dextrose. All the solutions dis-
solved in distilled H2O.
(a) As for agar plates, prepare and autoclave dextrose solution
separately from yeast extract-peptone-tryptophan broth.
(b) Combine sterilized mixtures and stir well.
(c) Cool prior to use.
4 Carolina A. Novoa et al.

3. Synthetic Dextrose agar plates (SD): 1.7 g/L yeast nitrogen

base, 5 g/L ammonium sulfate, 20 g/L dextrose, 20 g/L agar.
All the solutions dissolved in distilled H2O.
(a) In separate autoclavable flasks prepare SD broth, 4% agar,
and 20% dextrose as indicated above. Make sure to use at
least one flask with enough volume carrying capacity to
hold the desired final volume.
(b) Autoclave each mixture.
(c) While stirring the autoclaved mixture contained in the
largest flask, mix in the other solutions.
(d) While still warm, pour media into 10 cm diameter plastic
petri dishes.
(e) Allow agar mixture to cool and solidify before using.
(f) Unused SD plates can be stored at 4  C for several
months.

2.3 Yeast Strains 1. Mating tester strain YPH316 which is a MATα strain bearing an
inactivating mutation in HIS1 but otherwise prototrophic.
2. Positive control for high levels of ALF. There are many possible
controls among the hundreds of known ALF-inducing mutants
[7, 9]. A BY4742-derived haploid deletion of BIM1 from the
yeast knockout collection [10] serves well as a strong ALF mutant.
3. Negative control for ALF. A wild-type strain representing your
strain background of study is the best control. Here, we use
BY4742 (MATα his3Δ1 leu2Δ0 lys2Δ0 ura3Δ0) as our wild-
type control.
4. Positive control for mating. A MATa strain can provide an
additional control for mating. Here, we use BY4741 (MATa
his3Δ1 leu2Δ0 met15Δ0 ura3Δ0).

3 Methods

3.1 Qualitative ALF 1. Day 1. Suspend 1
106–1
107 α-tester strain cells in water,
Assay spread yeast suspension on a YPD plate, and incubate at 30  C
overnight. The plate should be fully covered with a lawn of α-
tester cells the following day.
2. Day 1. On a second YPD plate, using a toothpick or stick, patch
MATα haploid strains in an array of approximately 1 cm2
patches. At least 15 patches can be comfortably fit onto a
10 cm petri dish (Fig. 2). The following controls can be included
on each plate: wild-type MATα (BY4742), wild-type MATa
(BY4741), and the MATα bim1Δ::kanMX from the MATα
deletion collection (or another ALF phenotype positive control).
Incubate at 30  C degrees for 24 h (see Notes 3 and 4).
 The A-Like Faker Assay for Measuring Yeast Chromosome III Stability 5

Fig. 2 Examples of qualitative and quantitative ALF data. (a) Sample plate setup for ALF strain patches in
qualitative assay. At least 15 1
1 cm2 patches can be fit on a 10 cm petri dish. At right a sample output plate
showing the BY4742 and bim1Δ controls on top, along with various positive and negative test strains and the
BY4741 control at the bottom. This panel is an excerpt reproduced from Figure S7 in ref. [7]. (b) Output SD
plates from the quantitative ALF assay. A known number of wild-type BY4742 (left) or MATα sgs1Δ (right)
cells were mixed with an excess of MATα tester cells and co-cultured on SD media. The sgs1Δ strain exhibits
an increased frequency of ALF [7] which can be quantified as described in the text, by comparing to the
viability counts from the associated nonselection YPD plate (not shown)

3. Day 2. Mate the α-tester strain with the arrayed patches by

replica plating the α-tester lawn, and the mutant patches, in
that order on a fresh YPD plate. Incubate overnight at 30  C
(see Note 5).
6 Carolina A. Novoa et al.

4. Day 3. Replica plate the lawn of mated strains onto an SD plate.

Incubate at 30  C.
5. Day 3–5. Growth of diploid cells in colonies should appear
within 48–72 h.
6. Day 5. Score the number of colonies on the SD plate for each
strain. Comparing the average number of colonies in mutant
patches to the number of colonies in the wild-type MATα
control patches gives a relative indication of chromosome
instability.
7. Optional: retest all probable ALF phenotype hits as described
above, but using 4–12 independent colonies to create patches
for each mutant. Median values for colony number on the SD
plates that are twofold higher than WT are usually considered
ALF mutants [7, 9].

3.2 Quantitative ALF 1. Day 1. Inoculate 9–12 independent colonies of each MATα
Assay strain to be tested in 2–5 mL of liquid YPD broth, and incubate
overnight at 30  C (see Note 5).
2. Day 1. In parallel, inoculate YPD broth with 4 mL of α tester
cells per MATα strain of interest, and incubate overnight at
30  C (see Note 6).
3. Day 2. Separate overnight cultures by MATα strain and follow
steps 4–11 for one strain at a time, then repeat for the rest of
the mutants.
4. Day 3. Label one YPD and one SD plate with strain name.
5. Day 3. Make 1:100,000 dilutions of the overnight cultures in
water (see Note 7).
6. Day 3. Dry YPD plate by placing a piece of ~10 cm diameter
filter paper on the top of a replica plating block and pressing the
surface of the plate against the filter paper. Allow water to wick
into the filter paper until it is visibly wet. If necessary repeat
with a new filter paper to remove more liquid (see Note 8).
7. Day 3. Plate 100 μL of each independent colony dilution on
the dried YPD plate to determine cell viability counts. Record
plating order and orientation on the plate. See Fig. 2 as refer-
ence for spotting position (see Notes 9 and 10).
8. Day 3. Dry the SD plate as described in step 4.
9. Day 3. Mix 100 μL of each undiluted overnight culture with
300 μL of overnight culture of the MATα tester in a 1.5 mL
microcentrifuge tube.
10. Day 3. Spin at 3000
g for 5 min and resuspend each pellet in
100 μL of H2O.
11. Day 3. Plate the mixtures from step 9 on the dried SD plate
(step 8) in the same order and orientation as in the YPD plate.
 The A-Like Faker Assay for Measuring Yeast Chromosome III Stability 7

12. Day 3. Allow all the plates to completely absorb the liquid on
the bench top, then flip and incubate at 30  C for 24 h or 72 h
for YPD and SD plates, respectively. After 2 days in culture,
colonies of mated yeast should become visible on the SD plate
inside the thick white circles left behind by the excess unmated
cells (see Note 11).
13. Day 4–6. Calculate the frequency of ALF by finding the ratio
between mated and plated cells on the SD plates. The number of
mated cells is equal to the number of colonies grown on
each spot on the SD plate. The number of cells plated on
SD can be calculated from the number of viable cells in each
spot on the YPD plate, multiplied by the dilution factor
(see Notes 12 and 13).

4 Notes

1. For all solid media, autoclave at 121 ˚C for at least 20 min on a

liquid cycle. Cool agar mixture to approximately 55–60 ˚C, and
add other solutions, mix well before pouring 20–25 mL in each
petri dish.
2. If a drug needs to be included to maintain a plasmid, ammo-
nium sulfate should be substituted with 1 g/L monosodium
glutamate.
3. Patches can grow at any desired temperature if one is looking
for the effects of temperature on ALF. Standard conditions are
listed for most experiments but can generally be varied.
4. It is advisable to initially fit only 15 patches on one plate to
ensure even patch sizes and avoid sample cross-contamination.
It can also be helpful to mark the petri dish bottom carefully
with a grid and labels to ensure accurate patch size and unam-
biguous identification of strains to patches at the end of the
experiment.
5. The density of cells in each patch should be approximately
equal after the growth period. If some strains appear signifi-
cantly less dense than WT, make a note of it as it will reduce the
cell population transferred and thus the perceived rate of ALF
colonies. In cases where many strains are sick, an extra 24 h of
growth at 30  C can be useful in allowing saturation of all
patches.
6. If many MATα mutants will be tested in one day, a larger
volume of α tester stain can be cultured overnight in the same
flask. ~4 mL are required per strain.
7. Making at least two serial dilutions to get to a final 1:100,000
dilution of the overnight culture is recommended to achieve
accurate results due to variability in pipetting small volumes.
8 Carolina A. Novoa et al.

8. The dryness of the plate will influence the spreading of the

100 μL spots. If the plate is too dry, the liquid will not spread
enough, making counting of the colonies more difficult. If the
plate is not dry enough samples will run into each other or the
plates will take too long to absorb in the samples. Approxi-
mately 1 g of water can be removed by the filter paper, and
overdried plates should be used the same day.
9. It is recommended to fit only nine spots on the plates to avoid
samples running into each other. Also, avoid making spots too
close to the edges of the plate, as those areas are the hardest to
dry with a replica plating tool and might be missed.
10. Using a grid template underneath the YPD and SD plates will
facilitate arraying of liquid spots.
11. Some deletion mutants might take longer than 48 h to appear
on the SD plate. Colony counts are made after 72 normally but
can be extended to 96 h or longer for slow growers.
12. We think of ALF in terms of frequency not rate, as the ALF
assay differs from mutation rate fluctuation analyses. This is
because each mutation of the MATα locus causing ALF is
captured only once and fixed by mating with the mating tester.
Thus, there is little possibility of jackpots. A single ALF event
usually leads to a single colony and should be expressed as a
frequency of marker loss.
13. ALF with other media and plasmids: The ALF assay can also be
used to test the effect of plasmids or plasmid borne genes on
any MATα strain. The conditions and specific media changes
will vary depending on the plasmid system. In our experience,
it is important to: (a) maintain selection of the plasmid
throughout patch growth, overnight culturing and mating,
but it is not necessary on the final SD plate. (b) For galactose
inductions the time of galactose exposure and the necessity of
including other nonrepressive sugars like raffinose needs to be
determined empirically.

Acknowledgments

We thank Philip Hieter for helpful discussions. PCS is a Canadian

Institutes of Health Research (CIHR) New Investigator and a
Michael Smith Foundation for Health Research Scholar. We
acknowledge operating support from CIHR (MOP-136982) and
this research is funded by the Canadian Cancer Society (grant
703263).
 The A-Like Faker Assay for Measuring Yeast Chromosome III Stability 9

References
1. Herskowitz I (1988) Life cycle of the budding 8. Ben-Aroya S, Coombes C, Kwok T, O’Donnell
yeast Saccharomyces cerevisiae. Microbiol Rev KA, Boeke JD, Hieter P (2008) Toward a com-
52(4):536–553 prehensive temperature-sensitive mutant
2. Strathern J, Hicks J, Herskowitz I (1981) Con- repository of the essential genes of Saccharo-
trol of cell type in yeast by the mating type myces cerevisiae. Mol Cell 30(2):248–258.
locus. The alpha 1-alpha 2 hypothesis. J Mol doi:10.1016/j.molcel.2008.02.021
Biol 147(3):357–372 9. Stirling PC, Bloom MS, Solanki-Patil T, Smith
3. Herskowitz I (1988) The Hawthorne deletion S, Sipahimalani P, Li Z, Kofoed M, Ben-Aroya
twenty-five years later. Genetics 120(4):857–861 S, Myung K, Hieter P (2011) The complete
4. Liras P, McCusker J, Mascioli S, Haber JE spectrum of yeast chromosome instability
(1978) Characterization of a mutation in genes identifies candidate CIN cancer genes
yeast causing nonrandom chromosome loss and functional roles for ASTRA complex com-
during mitosis. Genetics 88(4 Pt 1):651–671 ponents. PLoS Genet 7(4):e1002057. doi:10.
1371/journal.pgen.1002057
5. Warren CD, Eckley DM, Lee MS, Hanna JS,
Hughes A, Peyser B, Jie C, Irizarry R, Spencer 10. Winzeler EA, Shoemaker DD, Astromoff A,
FA (2004) S-phase checkpoint genes safeguard Liang H, Anderson K, Andre B, Bangham R,
high-fidelity sister chromatid cohesion. Mol Benito R, Boeke JD, Bussey H, Chu AM, Con-
Biol Cell 15(4):1724–1735. doi:10.1091/ nelly C, Davis K, Dietrich F, Dow SW, El
mbc.E03-09-0637 Bakkoury M, Foury F, Friend SH, Gentalen
E, Giaever G, Hegemann JH, Jones T, Laub
6. Lemoine FJ, Degtyareva NP, Lobachev K, M, Liao H, Liebundguth N, Lockhart DJ,
Petes TD (2005) Chromosomal translocations Lucau-Danila A, Lussier M, M’Rabet N,
in yeast induced by low levels of DNA polymer- Menard P, Mittmann M, Pai C, Rebischung
ase a model for chromosome fragile sites. Cell C, Revuelta JL, Riles L, Roberts CJ, Ross-
120(5):587–598. doi:10.1016/j.cell.2004.12. MacDonald P, Scherens B, Snyder M,
039 Sookhai-Mahadeo S, Storms RK, Veronneau
7. Yuen KW, Warren CD, Chen O, Kwok T, S, Voet M, Volckaert G, Ward TR, Wysocki R,
Hieter P, Spencer FA (2007) Systematic Yen GS, Yu K, Zimmermann K, Philippsen P,
genome instability screens in yeast and their Johnston M, Davis RW (1999) Functional
potential relevance to cancer. Proc Natl Acad characterization of the S. cerevisiae genome
Sci U S A 104(10):3925–3930. doi:10.1073/ by gene deletion and parallel analysis. Science
pnas.0610642104 285(5429):901–906
 Chapter 2

The Chromosome Transmission Fidelity Assay

for Measuring Chromosome Loss in Yeast
Supipi Duffy and Philip Hieter

Abstract
The budding yeast Saccharomyces cerevisiae has served as an excellent model system for studying highly
conserved biological pathways including pathways involved in genome transmission and maintenance. The
Chromosome Transmission Fidelity (CTF) colony color assay was developed to assess chromosome
instability (CIN) in yeast, by monitoring the loss or gain during cell division of an artificial chromosome
fragment carrying a visual marker. The CTF assay monitors changes in chromosome number, allowing the
detection of mutants that exhibit increased rates of chromosome nondisjunction or chromosome loss. In
this article, we describe the SUP11-marker-based CTF assay system, and the methodologies for both
qualitative analysis of mutants affecting chromosome transmission, and quantitative analysis for determin-
ing the types and rates of errors in chromosome transmission using half-sector analysis.

Key words Chromosome instability, Genome instability, Aneuploidy, Whole chromosomal loss,
Chromosome transmission fidelity, Marker loss

1 Introduction

Chromosome replication and segregation during the mitotic cell

cycle relies on the correct execution of a complex series of events,
with functional determinants that act in cis (DNA sequence
domains) and in trans (gene products). As these processes are
essential to cell viability, functional characterization of mutants
that are defective in chromosome transmission in higher eukaryotes
has been challenging. In the budding yeast Saccharomyces cerevisiae,
DNA replication and chromosome segregation occur with
extremely high fidelity, as errors in chromosome transmission hap-
pen very infrequently with rates of chromosome mis-segregation,
for example, occurring on the order of once per 105 cell divisions
[1, 2]. However, as this rate of chromosome transmission fidelity in
wild-type strains is much greater than the fidelity necessary for cell
viability, it is possible to perform genetic analysis with mutants that

Marco Muzi-Falconi and Grant W. Brown (eds.), Genome Instability: Methods and Protocols, Methods in Molecular Biology,
vol. 1672, DOI 10.1007/978-1-4939-7306-4_2, © Springer Science+Business Media LLC 2018

11
12 Supipi Duffy and Philip Hieter

reduce chromosome transmission fidelity over a several order of

magnitude range [3, 4].
Taking advantage of the findings that yeast mutants with
altered chromosome transmission fidelity remain viable [5] and
can tolerate aneuploidy for many individual chromosomes [6],
assays were developed for monitoring changes in chromosome
ploidy [7, 8]. The two visual chromosome transmission fidelity
(CTF) assays, ade3-2p and SUP11, can be used to detect and
analyze mutants involved in chromosome transmission. This chap-
ter will only describe the use of the SUP11 system for qualitative
and quantitative analyses of CTF. The general principles, however,
apply directly to the ade3-2p system.
Both CTF assays make use of mutations in the purine biosyn-
thesis pathway [9] to facilitate the colony color assay that detects
chromosome loss. Wild-type yeast cells are colorless and produce
white colonies. Mutations in either the ADE1 or the ADE2 genes
required for purine biosynthesis cause the accumulation of an
intermediate that generates a red hue. The degree of redness in
yeast colonies harboring these mutations depends on accumulation
of biosynthetic intermediates, which leads to the formation of a red
pigment [3]. In the SUP11 system cells contain an ochre mutation
in the ADE2 gene (ade2oc) and are red. This mutation is sup-
pressed by the introduction of a chromosome fragment (CF) that
carries the ochre-suppressing form of a tRNA gene, SUP11
[10–12].
The standard CF used is a 125-kilobase fragment with the short
arm containing a selectable marker that can be used during initial
transformation and a colony color marker (SUP11), which is used
in the visual colony assay. The structure and genotype of the CF
makes it ideal for monitoring CTF [12, 13]. As the color marker is
embedded in pBR322 that bears no homology to yeast DNA on
the CF short arm, it will not be lost by mitotic recombination
between the CF and endogenous chromosomes [4]. The presence
of the CF results in disomy for less than 1% of the yeast genome
with no effects on either cell growth or endogenous chromosome
fidelity; in wild-type cells, the loss rate of the CF is ~2 in 104 cell
divisions [12, 13].
In the CTF assay, a homozygous ade2-ochre diploid cell with
zero copies of the CF produces no functional Ade2 and thus
accumulates sufficient pigment to generate red colonies. A cell
with only one copy of the CF will produce small amounts of
Ade2, and will only suppress the ochre mutation partially, thus
will generate pink colonies. A cell containing two or more copies
of the CF that will be colorless will generate white colonies, because
enough ADE2 gene product will be made to prevent the accumu-
lation of the intermediates that produce the red color. Thus, diploid
cells with zero, one, and two or more copies of CF will generate
red, pink, and white cell lineages respectively and the degree of red
 The Chromosome Transmission Fidelity Assay for Measuring. . . 13

CEN
ARS1
A Mating B

CFVII(RAD2.d)::URA
)::URA
URA3
yfg K anMX
KanMX
X ade2-101::NatMX
MAT
MAT

yfg deletion
M
MA
MATTa
1:0
YPH1725
Diploid selection (2X)
SD-Ura+G418
Transformation of Replacing the WT copy of
the linearized CF Sporulation the mutant in the CF strain
1:1
Haploid selection for
ade2-101, CF and yfgg deletionn
ade2-101,

2:0 2:1
CFVII(RAD2.d)::URA
ade2-101::NatMX
yfg KanMX

Streak for singles on media selective for CF

Streak or plate on media non-selective for CF

C
SC with 20% ade 2:0
Loss of CF

ade2-101::NatMX
yfg KanMX

1:0

Wild type Mild sectoring Severe

sectoring

Fig. 1 The qualitative CTF assay and the half-sectoring assay. (a) Detailed steps involved in generating a CTF
strain. For more details on generating a strain by transforming the CF or by introducing the mutant allele to the
CF strain, refer to [4]. A starter strain (YPH1725) containing the ade2-101 mutation and the SUP11 (CFVII
(RAD2.d)::URA3) is mated to a deletion strain of YFG and diploids are selected by plating onto appropriate
selective media. Following sporulation, a haploid strain is generated where yfgΔKAN is combined with the
ade2-101::NAT mutation and the CF. Cells are then streaked onto media with the selection for the CF and after
48 h, are either streaked or plated on low adenine media plates. After growing 7–10 days at 25
C, plates are
placed at 4
C for 2–3 days before scoring. Representative images from qualitative CTF phenotypes are shown
at the bottom. (b) Segregation properties of the CF adapted from [3]. The CF includes an ARS element, a CEN
element, and a selectable marker in yeast (URA). When the CTF strains are placed on nonselective media the
pattern of sectoring in each colony reflects the inheritance of the CF. In about 99% of cell divisions, the CF
replicates once and partitions equally to daughter cells (1! 1:1 segregation). Aberrant CF transmission events
(1!1:0, 1!2:0, and 1!2:1) are depicted here. Sample images are also included. (c) Samples from an
output plate that shows red, pink, and white diploid colonies with 2:0 and 1:0 patterns of CF segregation

sectoring in colonies reflects the frequency of mitotic chromosome

loss. In haploid cells, one copy of the CF will cause cells to appear
white and a haploid cell that has lost the CF will appear red (Fig. 1a
and c). The CTF assay can therefore distinguish between chromo-
some loss (1:0 segregation) and nondisjunction (2:0 segregation)
14 Supipi Duffy and Philip Hieter

in diploid cells (Fig. 1b), which makes it an appropriate method for

determining the types of aberrant transmissions that lead to cells
with altered chromosome ploidy [7].
The qualitative CTF assay can be easily applied to large collec-
tions of loss-of-function (nonessential genes), reduction-of-func-
tion (essential genes) mutants and overexpression (essential and
nonessential) mutants [14–16], whereas the half-sector analysis
for determining the types and rates of chromosome aberrations is
more labor intensive. It is also possible to assess CTF rates using
fluctuation analysis [13]; however, this chapter will be limited to
qualitative CTF and the half-sectoring assay.
To test the effects of a specific genetic perturbation, the muta-
tion will need to be introduced into the CTF background using one
of the following methods. Mutations to be tested can be intro-
duced directly into CTF strain backgrounds (YPH1725 and
YPH1726, see Subheading 2.3) using standard techniques. Alter-
natively, the CF can be generated de novo by transforming mutant
strains of interest with the appropriate linearized plasmid [4].
Finally, strains can also be generated using mating and sporulation
(Fig. 1a). Following strain generation using one of the methods
above, cells are placed on low adenine containing agar media to
allow accumulation of intermediates that give rise to red pigment
(Fig. 1a). As wild-type cells will generate white colonies, changes in
colony sectoring will reflect changes in the rates of chromosome
loss and nondisjunction. The protocols described here primarily
focus on genomically integrated loss-of-function or reduction-of-
function alleles; however, the CTF assay can be conducted in strains
with plasmids in the presence of alternate media for inducible
expression.

2 Materials

2.1 Lab Ware 1. Disposable 10 cm plastic petri dishes.

2. Autoclaved toothpicks or sticks.
3. Sterile 1.5 mL microcentrifuge tubes.

2.2 Media 1. Synthetic Complete Plates with Limiting Adenine (SC+20%

Preparation ade): 1.7 g/L yeast nitrogen base, 5 g/L ammonium sulfate,
20 g/L dextrose, 20 g/L agar (see Notes 1 and 2).
2. Prepare SC broth, 4% agar, and 20% dextrose in three separate
autoclavable flasks/bottles as indicated above. Use at least one
flask that will allow the desired final volume of media. Also
place a stir bar in this flask as it will aid in the mixing process.
3. Autoclave each mixture.
 The Chromosome Transmission Fidelity Assay for Measuring. . . 15

4. Place the largest container on a stir plate and add the agar and
the dextrose.
5. To this add 100 mL/3 L of solution containing the following
supplements: adenine, 0.015%; uracil, 0.06%; L-lysine, 0.06%;
L-histidine, 0.09%; L-trypsine, 0.09%; L-leucine, 0.06%.

6. Pour ~25 mL of media into 10 cm diameter plastic petri dishes

after the media has cooled to a temperature of approximately
55 ˚C.
7. Allow agar mixture to cool and solidify before using.
8. Unused SC+1/5ade plates can be stored at 4
C for several
months.
9. Synthetic Complete agar plates (SC-URA): 1.7 g/L yeast nitro-
gen base, 5 g/L ammonium sulfate, 2 g/L of—Uracil dropout
mix, 20 g/L dextrose, 20 g/L agar (see Notes 3 and 4).
(a) Prepare SC broth, 4% agar, and 20% dextrose in three
separate autoclavable flasks/bottles as indicated above.
Use at least one flask that will allow the desired final
volume of media. Also place a stir bar in this flask, as it
will aid in the mixing process.
(b) Autoclave each mixture.
(c) Place the largest container on a stir plate and add the agar
and the dextrose.
(d) Pour ~25 mL of media into 10 cm diameter plastic petri
dishes after the media has cooled to a temperature of
approximately 55
C.
(e) Allow agar mixture to cool and solidify before using.
(f) Unused SC-URA plates can be stored at 4
C for several
months.

2.3 Yeast Strains 1. The CTF starter strains are YPH1725 or YPH1726, which are
MATa and MATalpha strains with the ade2-101 ochre muta-
tion marked with NatMX and the CF marked with URA3 [17].
The NatMX marker allows you to select for the presence of the
ade2-101 marker; therefore, the strains can be used in high-
throughput screens. These strains can be used to generate final
strains with genes of interest (Fig. 1a and [4]).
2. Positive control for high levels of CTF. These can be found in
previously published work from the Hieter lab [12, 16–18].
The CTF19 mutant strain used in [18] serves well as a control
for the half-sectoring assay.
3. Negative control for CTF. Either YPH1725 or YPH1726
without perturbation serves as a good control.
16 Supipi Duffy and Philip Hieter

3 Methods

3.1 Qualitative CTF 1. Day 1. On a SC-URA plate, using a toothpick or stick, streak
Assay the CTF strains with YFG perturbation to achieve single colo-
nies. Include the CTF starter strain and the CTF19 mutant
strains as controls (see Note 5).
2. Day 3. Pick two single colonies for each strain with a wooden
stick and dilute in 1 mL of H2O. Dilute the cells 1:1000 in a
second microcentrifuge tube and plate 100 μL into SC plates
with 20% of the standard adenine concentration to obtain
100–250 colonies per plate (see Note 6).
Alternatively, two single colonies for each strain can be
streaked onto SC plates with 20% of the standard adenine
concentration to achieve single colonies (see Note 7).
3. Day 3. Incubate plates at 25
C for 6–7 days (see Note 8).
4. Day 9–10. Incubate plates at 4
C for additional 5–7 days to
enhance the development of red pigment (see Note 9).
5. Day 14–17. Score the number of sectored colonies on the
plates. The CTF starter strain should have almost no sectored
colonies, whereas the CTF19 mutant will have many sectored
colonies giving a relative indication of chromosome instability
(see Note 10).

3.2 Half-Sector 1. Day 1. On a SC-URA plate, using a toothpick or stick, streak

Assay out the CTF strains with YFG perturbation to achieve single
colonies. Include the CTF starter strain and the CTF19 mutant
strains as controls (see Note 5).
2. Day 3. Pick 9–12 single colonies for each strain with a wooden
stick and dilute in 1 mL of H2O. Dilute the cells 1:1000 in a
second microcentrifuge tube and plate 100 μL into SC plates
with 20% of the standard adenine concentration to obtain
100–250 colonies per plate. Plate 100 μL into SC plates with
20% of the standard adenine concentration to obtain 100–250
colonies per plate (see Note 7).
3. Day 3. Incubate plates at 25
C for 6–7 days (see Note 8).
4. Day 9–10. Incubate plates at 4
C for additional 5–7 days to
enhance the development of red pigment (see Note 9).
5. Day 14–17. Score the number of “half-sectored” colonies on
the plates to calculate the rate of chromosome loss relative to
wild-type cells. Chromosome loss (1:0 segregation, Fig. 1b)
will result in pink/red half-sectored colonies, whereas nondis-
junction (2:0 segregation) will result in white/red half-
sectored colonies (Fig. 1b). Note that colonies with greater
than 1/2 red:white sectoring (e.g., ¾ red and ¼ white) are
scored as half-sectored colonies; this rule was empirically
 The Chromosome Transmission Fidelity Assay for Measuring. . . 17

determined by comparisons to rates determined using fluctua-

tion analysis.
6. The CTF rate is a ratio of “half-sectored” colonies to total CF
containing colonies. Colonies that are completely red should
be discounted, as these represent cells that did not have the CF
when they were plated. For the CTF19 mutant strain, the rate
of chromosome loss should be ~100 times greater than wild-
type and the rate of nondisjunction should be 60-fold higher
than wild-type [18] (see Notes 10 and 11).

4 Notes

1. For all solid media (>1 L), autoclave at 121 ˚C for at least
30 min on a liquid cycle. Mix well after adding the solutions
together.
2. Standard protocol includes 1/5 the concentration of adenine;
however, it may be necessary to adjust this concentration as the
red colony phenotype may vary when using different adenine
stock solutions or strain backgrounds.
3. If a drug needs to be included to maintain a plasmid, ammo-
nium sulfate should be substituted with 1 g/L monosodium
glutamate.
4. To make the—Uracil dropout mix amino acids as follows. 6 g
of each of serine, arginine, glycine, glutamic acid, alanine,
histidine, glutamine, threonine, asparagine, phenylalanine,
methionine, valine, isoleucine, proline, tryptophan, tyrosine,
aspartic acid, lysine and cysteine, 12 g of leucine and 1.5 g of
adenine hemisulfate.
5. It is important to maintain selection (URA) for the CF until
you are ready to begin the CTF experiment, as the loss rate of
the CF is higher than for native chromosomes.
6. It is assumed a single colony contains approximately 106 cells;
therefore, 100 μL from the diluted micocentrifuge tube repre-
sents approximately 100 cells. For slow growing strains it may
be necessary to plate more than 100 μL to obtain 100–250
cells. In the first pass it is advisable to plate at least two different
volumes to get a final cell number of 100–250 cells/plate.
7. While streaking for singles is acceptable it is much harder to
control for cell numbers using this method. It is advisable to
score at least 100 single colonies during the first pass of a CTF
experiment.
8. CTF starter strains can grow at any desired temperature; how-
ever, we have observed that slower growth at 25
C enhances
the accumulation of red pigment compared to growth at 30
C.
18 Supipi Duffy and Philip Hieter

9. This extended incubation at 4
C is especially beneficial for

detecting mild CTF phenotypes as it further enhances the red
pigment development.
10. With a mis-segregation of 104 approximately
30,000–100,000 single colonies will need to be counted to
generate the rate of chromosome mis-segregation events for
wild-type cells. Depending on the degree of the CTF pheno-
type it may be necessary to score at least 3000 colonies per gene
tested. A rule of thumb is to plate sufficient numbers of colo-
nies so that you score >10 independent half-sectored colonies.
Colonies with greater than 1/2 red:white sectoring (e.g., ¾ red
and ¼ white are scored as half-sectored) are designated as half
sectored colonies.
11. CTF with other media and plasmids: The CTF assay can also be
used to test the effect of plasmids or plasmid borne genes. The
conditions and specific media changes will vary depending on
the plasmid system. In our experience it is important to: (a)
maintain selection of the plasmid throughout growth on
plates, but it is not necessary on the final SC+1/5 ade plate.
(b) For galactose inductions in our experience it is better to
incorporate two rounds of inductions each lasting 48 h of
growth.

Acknowledgments

This work was supported by grants from the Canadian Institute of

Health Research (MOP 38096) and the National Institute of
Health (RO1CA158162) to P.H. S.D. was supported by a Michael
Smith Foundation for Health Research Trainee Award. P.H. is a
Senior Fellow in the Genetic Networks program at the Canadian
Institute for Advanced Research.

References

1. Esposito MS, Maleas DT, Bjornstad KA et al 4. Shero JH, Koval M, Spencer F et al (1991)
(1982) Simultaneous detection of changes in Analysis of chromosome segregation in Saccha-
chromosome number, gene conversion and romyces cerevisiae. Methods Enzymol
intergenic recombination during mitosis of 194:749–773
Saccharomyces cerevisiae: spontaneous and 5. Maine GT, Sinha P, Tye BK (1984) Mutants of
ultraviolet light induced events. Curr Genet 6 S. cerevisiae defective in the maintenance of
(1):5–11. doi:10.1007/BF00397633 minichromosomes. Genetics 106(3):365–385
2. Hartwell LH, Smith D (1985) Altered fidelity 6. Parry EM, Cox BS (1970) The tolerance of
of mitotic chromosome transmission in cell aneuploidy in yeast. Genet Res 16(3):333–340
cycle mutants of S. cerevisiae. Genetics 110 7. Hieter P, Mann C, Snyder M et al (1985)
(3):381–395 Mitotic stability of yeast chromosomes: a col-
3. Koshland D, Hieter P (1987) Visual assay for ony color assay that measures nondisjunction
chromosome ploidy. Methods Enzymol and chromosome loss. Cell 40(2):381–392
155:351–372
 The Chromosome Transmission Fidelity Assay for Measuring. . . 19

8. Koshland D, Kent JC, Hartwell LH (1985) (36):9967–9976. doi:10.1073/pnas.

Genetic analysis of the mitotic transmission of 1611839113
minichromosomes. Cell 40(2):393–403 15. Stirling PC, Bloom MS, Solanki-Patil T et al
9. Roman H (1956) Studies of gene mutation in (2011) The complete spectrum of yeast chro-
Saccharomyces. Cold Spring Harb Symp mosome instability genes identifies candidate
Quant Biol 21:175–185 CIN cancer genes and functional roles for
10. Hawthorne DC, Mortimer RK (1968) Genetic ASTRA complex components. PLoS Genet 7
mapping of nonsense suppressors in yeast. (4):e1002057. doi:10.1371/journal.pgen.
Genetics 60(4):735–742 1002057
11. Manney TR (1964) Action of a super- 16. Yuen KW, Warren CD, Chen O et al (2007)
suppressor in yeast in relation to allelic Systematic genome instability screens in yeast
mapping and complementation. Genetics and their potential relevance to cancer. Proc
50:109–121 Natl Acad Sci U S A 104(10):3925–3930.
12. Spencer F, Gerring SL, Connelly C et al (1990) doi:10.1073/pnas.0610642104
Mitotic chromosome transmission fidelity 17. Measday V, Baetz K, Guzzo J et al (2005)
mutants in Saccharomyces cerevisiae. Genetics Systematic yeast synthetic lethal and synthetic
124(2):237–249 dosage lethal screens identify genes required
13. Hegemann JH, Shero JH, Cottarel G et al for chromosome segregation. Proc Natl Acad
(1988) Mutational analysis of centromere Sci U S A 102(39):13956–13961. doi:10.
DNA from chromosome VI of Saccharomyces 1073/pnas.0503504102
cerevisiae. Mol Cell Biol 8(6):2523–2535 18. Hyland KM, Kingsbury J, Koshland D et al
14. Duffy S, Fam HK, Wang YK et al (2016) Over- (1999) Ctf19p: a novel kinetochore protein in
expression screens identify conserved dosage Saccharomyces cerevisiae and a potential link
chromosome instability genes in yeast and between the kinetochore and mitotic spindle.
human cancer. Proc Natl Acad Sci U S A 113 J Cell Biol 145(1):15–28
 Chapter 3

€ck
Measuring Mutation Rates Using the Luria-Delbru
Fluctuation Assay
Gregory I. Lang

Abstract
The Luria-Delbr€ uck fluctuation assay is one of the most commonly used methods for measuring the
mutation rate in microorganisms. Specifically, it is used to measure the mutation rate at a particular locus
or loci at which mutations give rise to a selectable phenotype. Here, I outline the essential features of
performing Luria-Delbr€ uck fluctuation assays as well as common missteps and tips for improving the
accuracy of mutation rate estimates. In addition, I provide tools for analyzing data from fluctuation assays.
This 96-well plate protocol has been optimized for use in yeast but should perform equally well for a range
of microorganisms using standard microbiological methods.

Key words Mutation rate, Fluctuation test, Poisson distribution

1 Introduction

1.1 Principle of the The principle of the fluctuation assay (first introduced by Salvador
Fluctuation Assay Luria and Max Delbr€ uck in their classic 1943 Genetics paper [1]) is
simple, and if implemented properly, provides a powerful way to
measure phenotypic mutation rates. The key to understanding the
logic of the Luria-Delbr€ uck experiment is to grasp the distinction
between the distribution of the number of mutation events per
culture and the distribution of the number of mutant cells per
culture. Fig. 1 shows three hypothetical cultures that each start
from a single cell and proceed through four generations of growth
to produce sixteen cells. In each of these cultures exactly two
mutation events occur, but the number of mutant cells differs
depending upon when during the growth of the culture those
mutations arose.
The distribution of the number of mutation events per culture
follows the Poisson distribution and the distribution of the number

Electronic Supplementary Material: The online version of this chapter (doi: 10.1007/978-1-4939-7306-4_3)
contains supplementary material, which is available to authorized users.

Marco Muzi-Falconi and Grant W. Brown (eds.), Genome Instability: Methods and Protocols, Methods in Molecular Biology,
vol. 1672, DOI 10.1007/978-1-4939-7306-4_3, © Springer Science+Business Media LLC 2018

21
22 Gregory I. Lang

Culture 1 Culture 2 Culture 3

1 2 3 4 Generations

Fig. 1 Cartoon illustrating the principle of the Luria-Delbr€uck fluctuation assay.

During the growth of a culture the number of mutation events will follow the
Poisson distribution; however, the number of mutants per culture will have a
larger variance because early arising mutations produce more mutant cells. For
example, each of the three cultures had two mutation events during growth;
however, the number of mutant cells differs depending upon when during growth
the mutations arose

of mutant cells per culture follows the Luria-Delbr€ uck distribution

(Fig. 2). Both these distributions are described by a single parame-
ter m, the expected number of mutation events per culture. Notice
that for all values of m, the zero class (p0) is the same for both the
Poisson and the Luria-Delbr€ uck distributions. This is because a
culture will have zero mutant cells if, and only if, zero mutation
events occurred. A single mutation event could produce just one
mutant cell (if it arose in the last generation) but could also lead to a
“jackpot” of mutant cells if it arose early in the growth of the
culture. These “jackpots” are rare. Their rarity can be explained
by looking retrospectively at the growth of a culture, recognizing
that half of the mutation events occurred in the last generation, one
quarter in the generation previous to that, etc. The “jackpot”
principle was the key observation in the Luria Delbr€ uck 1943
Genetics paper that proved that bacterial resistance to bacteriophage
occurred through genetic mutation and not to acquired immunity
following exposure to the bacteriophage at the time of plating.

1.2 Assumptions of The formulation of the Luria-Delbr€ uck distribution in Fig. 2 was
€
the Luria-Delbruck derived by Ma, Sandri, and Sarkar [2]. In order for the number of
Distribution mutant cells per culture to follow this distribution, several
biological assumptions must be met. When performing fluctuation
assays it is important to be aware of these assumptions because
deviations could affect the accuracy of mutation rate estimates.
 €ck Fluctuation Assay
Measuring Mutation Rates Using the Luria-Delbru 23

a
0.4
mn -m

Frequency of cultures
pn = e
0.3 m=1 n!

0.2 m=4

0.1 m=8

0
0 4 8 12 16 20
Mutation events (n) per culture
b
0.4
Frequency of cultures r-1
0.3 p0 = e-m , pr =
m
r Σ p /(r-i+1)
i=0
i

0.2 m=1

0.1 m=4
m=8
0
0 4 8 12 16 20
Mutant cells (r) per culture

Fig. 2 Distributions for the number of mutation events per culture and the
number of mutant cells per culture for three values of the parameter m,
the expected number of mutation events per culture. (a) The distribution of the
number of mutation events per culture follows the Poisson distribution with an
average number of mutation events per culture equal to m. (b) The distribution of
mutant cells per culture follows the Luria-Delbr€uck distribution. Note that p0 is
the same for both distributions; in other words, a culture will have no mutant
cells if, and only if, there were no mutation events

I have listed these assumptions in the order in which they are likely
to be violated in a typical fluctuation assay.
1. Growth of each culture starts with a single cell. This assumption
will almost always be violated, but as long as the initial inocu-
lum size is negligible relative to the final number of cells per
culture, this is not a concern. This is because the number of cell
divisions is equal to the number final cell number minus the
inoculum size.
2. Mutant and non-mutant cells have the same growth rate. This
will skew the distribution of the number of mutant cells that
result from each mutation event. This will have less of an effect
on mutation rate estimates using the p0 method than the MSS-
maximum-likelihood method (see Subheading 4), since differ-
ential growth will not affect the zero class.
3. No post-plating mutations. In the case where the selection does
not effectively kill or arrest non-mutant cells, additional cell
divisions will produce additional mutant cells. These post-
plating mutants will be Poisson-distributed.
24 Gregory I. Lang

4. All mutants are detected. If plating efficiency is less than 100%

or if there is protein perdurance (for instance in the case of
counter-selectable markers such as CAN1 or URA3), the num-
ber of true mutant cells will be undercounted.
5. No change in mutation rate during growth. Changes in muta-
tion rate in response to changes in growth conditions will skew
the distribution of mutant cells per culture. An elevated muta-
tion rate early would lead to more “jackpot” cultures; elevated
mutation rate late in growth would increase the proportion of
the lower classes of the distribution.
6. Same final number of cells in replicate populations. This is
typically a safe assumption, though complications with the
experimental setup (such as uneven evaporation across the
96-well plate) could result in this assumption being violated.
7. No reverse mutations. Under most circumstances, this is a safe
assumption. Though it is possible, for example if a large class of
suppressor mutations exists for the phenotype being assayed,
that this assumption will not always hold.

1.3 Common From my experience performing fluctuation assays and advising

Missteps in others on this method, I have come to recognize several common
Performing Fluctuation missteps that should be avoided.
Assays 1. Not plating the whole culture. The principle of the Luria-
Delbr€ uck fluctuation is based on accurately measuring the
distribution of the numbers of mutant cells per culture. This
is best determined by plating the entirety of the culture. For
this reason, culture volumes should typically be kept small. If
the entire volume of the culture is not plated, it is necessary to
correct for this by adjusting for the dilution factor before
calculating m from the data. When you have a large number
of mutant cells per culture and subsampling is required, it
becomes more difficult to determine how well (or how poorly)
the data fit to the Luria-Delbr€
uck distribution, which flattens
out at high m.
2. Not having a zero class. There are two reasons why I think it is
important to have a zero class. First, it allows the p0 method to
be used to calculate mutation rate in addition to the MSS-
maximum-likelihood (see Subheading 4). Second, the maxi-
mum likelihood fitting method is most sensitive to the lower
end of the distribution of the data.
3. Using too few populations. There is no specific number of popu-
lations required, but the accuracy of the fluctuation assays (if
performed properly) increases as the square root of the number
of cultures [3]. Fig. 1 in [4] can be used as a guide for selecting
an appropriate number. I routinely use 72 cultures to estimate
m, though I have used as little as 36 and as many as 720.
 €ck Fluctuation Assay
Measuring Mutation Rates Using the Luria-Delbru 25

2 Materials

1. 96-well plates.
2. Aluminum plate seals.
3. Whatman filter paper circles (Grade 3, Cat. No. 1003-090).
4. Replica-plating block.
5. Replica-plating velvets, sterile.
6. Liquid nonselective growth medium (see Note 1).
7. Agar plates containing selective growth medium (see Note 2).
8. Beckman Coulter particle counter (see Note 3).
9. Vials for Beckman Coulter particle counter.
10. ISOTON II Diluent for Beckman Coulter particle counter.
11. Sonicator.

3 Performing Fluctuation Assays

3.1 Overdrying 1. Set up a replica-plating block with a sterile velvet.

Plates 2. Using sterilized forceps, place a sterile 90 mm Whatman filter
paper circle on the velvet (see Note 4).
3. Press plate onto the replica-plating block to transfer the filter to
the plate (see Note 5).
4. Allow the filter to sit on the plate for at least 30 min, up to
overnight, to allow the filter to pull water out of the plate.
5. Repeat for each plate.

3.2 Set Up 96 Parallel 1. Grow a single overnight culture of the strain to be tested (see
Cultures Note 6).
2. Dilute culture 1:10,000 into 40 ml of nonselective (see Note 7).
3. Add the appropriate volume of culture to each well of a 96-well
plate (see Note 8).
4. Cover the plate with a plate seal to avoid evaporation.
5. Incubate the 96-well plate at 30
C (or other appropriate
temperature) until cultures reach saturation (see Note 9).

3.3 Determining the 1. For each 96-well plate, pool 24 of the cultures, and set the
Average Number of remaining 72 cultures aside to be plated on selective medium
Cells Per Culture (N) (see Note 10).
2. If necessary, sonicate the pooled cultures to break up cell
clumps for 1–2 min in a bath sonicator or for 10 half-second
pulses using a probe sonicator.
3. Make three independent dilutions of each culture into filtered
ISOTON II Diluent (see Note 11).
26 Gregory I. Lang

4. Count each vial three times (see Note 12).

5. Calculate the average number of cells per culture (N) as the
median-of-the-median cell counts, correcting for the initial
dilution and culture volume (see Note 13).
Example data:

Count 1 Count 2 Count 3 Median

Dilution 1 33,284 32,912 33,518 33,284
Dilution 2 33,268 33,126 32,766 33,126
Dilution 3 33,972 33,578 33,440 33,578
Median 33,284

3.4 Determining the 1. For the remaining 72 cultures, bring the volume up to 100 μl
Number of Mutant by adding water to each well using a multichannel pipette.
Cells Per Culture 2. Plate the entire volume of each of the 72 cultures onto the
overdried plates by spot-plating nine cultures per plate (see
Fig. 3 and Note 14).
3. Allow the plates to sit at room temperature for at least a half
hour or until all of the liquid is absorbed (see Note 15).
4. Incubate plates for 1–2 days at 30
C until colonies are large
enough to count.
5. Count the number of mutant cells per culture (see Note 16).

Fig. 3 Spot plating cultures onto overdryed plates. Example of a fluctuation assay on 10 canavanine medium
from ref. [7]. Seventy-two 100 μl cultures were spot-plated onto eight canavanine plates. Colonies were
counted after 2 days of growth. For larger culture volumes (~200 μl), only six cultures are spot plated onto
each plate. For small volumes, cultures should be brought up to 100 μl.
 €ck Fluctuation Assay
Measuring Mutation Rates Using the Luria-Delbru 27

4 Analyzing Data from Fluctuation Assays

Many methods exist to calculate m (the expected number of muta-

tion events per culture) from fluctuation data (reviewed in [4, 5]).
In practice, only two methods should be used: the p0 method and
the MSS-maximum likelihood method.

4.1 Calculate the 1. Determine the fraction of cultures with zero mutant cells. This
Mutation Rate Using value is p0.
the p0 Method 2. Calculate the expected number of mutation events per culture
(m):
m ¼  ln(p0) (see Note 17).
3. Calculate mutation rate (μ):
μ¼N m
.

4.2 Calculate the In 1992, Ma, Sandri, and Sarkar provided a solution to the Luria-
Mutation Rate Using Delbr€uck distribution for the single parameter m [2]. This made it
the MSS-Maximum possible to calculate the most likely value of m based on fluctuation
Likelihood Method assay data [6]. Stewart [3] shows that this method, known as the
MSS-maximum-likelihood method, is the most accurate method
for determining m, and provides a formula for calculating 95%
confidence intervals for estimates of m calculated using this
method.
1. Calculate the expected number of mutation events per culture
(m) using the provided supplemental Matlab script
“findMLm” (see Note 18) or Excel spreadsheet “fluctuatio-
nAssay_TEMPLATE.xls” (see Note 19).
2. Calculate 95% confidence intervals on m using the following
equations:
m95 ¼ ln(m)  1.96σ(e1.96σ)0.315, and
m95+ ¼ ln(m) + 1.96σ(e1.96σ )0.315,
0:315
pffiffiffi
where σ ¼ 1:255m , and C is the number of cultures used (see
C
Note 20).
3. Calculate mutation rate (μ):
μ¼N
m
, where N is the average number of cells per culture.
4. Calculate 95% confidence intervals for the mutation rate:
μ95 ¼ mN95 , and
μ95 ¼ mN95 .

5 Notes

1. This is typically a rich medium or a synthetic complete medium,

for example YPD or SC for yeast.
28 Gregory I. Lang

2. The exact medium used will depend on the organism and the
particular locus or loci used for the fluctuation assay. For exam-
ple, in yeast, 5FOA (1 mg/ml) and canavanine (0.6 mg/ml) is
used to measure mutation rates at URA3 and CAN1, respec-
tively. In the case of canavanine, it is necessary to use higher
drug concentrations than is necessary for counter-selection to
prevent post-plating growth, typically 60 μg/ml [7].
3. Cells counts can also be performed using other methods such
as a hemocytometer or dilution plating.
4. It is possible to sterilize the Whatman filters, but I have not
found this to be necessary.
5. Press evenly so that the entire surface of the plate is in contact
with the filter. The 90 mm filter circles will be slightly larger on
the surface of the agar, which has diameter of ~85 mm for
standard petri dishes.
6. This can be done in a nonselective medium or in a medium that
selects against the mutants. For example, when selecting for
5FOA resistance, which occurs primarily through mutations at
the URA3 locus, I perform this overnight growth in medium
lacking uracil.
7. Before beginning a set of experiments it is useful to pilot the
experiment to identify the optimal culture conditions. To ana-
lyze the data using the p0 method, ~20–80% of the cultures
should have zero mutation events, thus zero mutants. An
estimate of the mutation rate will help to select a culture
volume to pilot. For example, when using 5FOA resistance
(μ ~ 5  108 per generation) as a selection, I typically use
200 μl cultures with 2% glucose; when using αF resistance
(μ ~ 5  106 per generation) as a selection, I typically use
10 μl cultures with only 0.2% glucose.
8. It saves time to use a multichannel pipette for this step.
9. It is not necessary to shake the plates. In fact, shaking could
increase evaporation or introduce inconsistencies between
wells. Side-by-side measurements of mutation rate without
shaking and with shaking on a Titramax 1000 orbital shaker
were indistinguishable.
10. I pool 24 wells along the diagonals, rather than two rows, to
sample wells along the edge and in the center to control for
possible edge effects such as uneven evaporation. This can be
done without changing pipette tips. It is important that cells be
resuspended well either by pipetting up and down several times
or using a plate vortex such as the Titramax 1000 orbital
shaker.
11. Filter ISOTON II Diluent with a 0.45 μm filter to remove
particles. Use an automatic dispenser for the solution to
 €ck Fluctuation Assay
Measuring Mutation Rates Using the Luria-Delbru 29

maintain uniformity in volume. After dispensing the diluent,

let the vials sit for a few minutes before counting to avoid
counting bubbles. I aim to dilute such that I count in the
range of 30,000–90,000 events (For example, I dilute a 107
culture 1:2000, 10 μl into 20 ml).
12. Before beginning rinse and flush until the background is below
500 (ideally below 100, but I tolerate up to 1000. Sometimes
this takes a while). Record the background before and after
each run. It is not necessary to rinse and flush in between
samples. If anything lodges in the aperture, rinse and flush,
noting when this occurs. If the counts spike, stop and restart
the count.
13. I do not subtract background. I use medians to ensure that
these estimates are robust to non-normally distributed data.
14. If the volume is greater than 100 μl, it may not be possible to
spot-plate nine cultures per plate. For yeast fluctuation assays
on 5FOA, I use 200 μl cultures and spot-plate six cultures per
plate. It is important that cells be resuspended well prior to
plating. This can be accomplished by pipetting up and down
several times. I have also used a Titramax 1000 orbital shaker to
resuspend cells prior to plating.
15. Be careful not to disturb the plates. I usually allow an hour, but
I have allowed plates to dry overnight before transferring them
to the incubator.
16. It is possible to count unaided by eye; however, I prefer to use a
dissection scope (10 magnification) and a gooseneck light
source. I use threshold counting. For example, for fluctuation
assays on canavanine, colonies smaller than 1 mm at 10
magnification are presumed to result from mutations that had
occurred after the cells were plated and were counted
separately.
17. This equation is derived from the Poisson distribution. Specifi-
cally, the probability of observing zero mutant cells (and thus,
zero mutation events) is p0 ¼ em.
18. The most likely value for m given fluctuation assay data can be
calculated by executing the following Matlab script: findMLm
(data), where data is a column vector containing counts of
mutant cells per culture. This program requires the following
Matlab scripts to run: scoreData and generateLD. scoreData
is a Matlab script that takes as its input data from a fluctuation
assay and a value for m. This program outputs the -log proba-
bility of observing the data given m. This program requires the
Matlab script generateLD. generateLD takes as its input a
value of m and a maximum number of mutant cells per culture
(max) for which to calculate the probability distribution. It
outputs the Luria-Delbr€ uck distribution from 0 to max with
30 Gregory I. Lang

parameter m. All three scripts are provided as supplemental

information and were originally published in ref. [7]. Alterna-
tively, several web-based tools exist for analyzing fluctuation
assays [8, 9].
19. Open the “fluctuationAssay_TEMPLATE.xls” spreadsheet,
provided as supplemental information. Remove the example
data in blue and enter your data here. These data will be in the
form of the number of cultures with a given number of mutant
cells. Check that the number “# of cult.” value is the same as
the number you expect. Select a value for m. The objective is to
find the value for m that maximizes the value for “-log P(data|
m). Using the workspace provided pick several values of m: I
typically start with values of m such as 0.1, 1, 2, 3, 4. Find the
minimum value. Then determine if the actual minimum is
greater or less than this value. For instance if 2 is the best
value of m out of 0.1, 1, 2, and 3, I next check if 1.9 or 2.1
improves the fit. If m of 2.1 is better than 2.0, I would then test
all values between 2 and 3 in increments of 0.1. I repeat this
process until I find the best most likely value of m given the
data to three significant digits.
20. The expected error of fluctuation assays was determined by
Stewart [3] and assumes that the data truly fit the Luria-
Delbr€uck distribution. If the quality of the data is poor, calcu-
lated values for m and the 95% confidence intervals on this
value will be unreliable. For this reason it is useful to compare
values for the mutation rate as determined by the P0 method
and the MSS-maximum likelihood method. When performing
fluctuation assays one needs to be cognizant of the assumptions
that underlie Luria-Delbr€ uck distribution and how violations
of these assumptions could affect estimates of mutation rates.
Some efforts have been made to account for deviations such as
post-plating growth [7], plating efficiency [10], and differ-
ences in the growth rate of mutant and non-mutant cells
[8, 11, 12].

Acknowledgments

I thank Sean Buskirk, Katie Fisher, and Dan Marad for comments
on this manuscript. This work was supported by a New Investigator
grant from the Charles E. Kaufman Foundation of The Pittsburgh
Foundation.

References
1. Luria S, Delbr€
uck M (1943) Mutations of bac- 2. Ma WT, Sandri GH, Sarkar S (1992) Analysis
teria from virus sensitivity to virus resistance. of the Luria-Delbr€
uck distribution using dis-
Genetics 28:491–511 crete convolution powers. J App Prob
29:255–267
 €ck Fluctuation Assay
Measuring Mutation Rates Using the Luria-Delbru 31

3. Stewart FM (1994) Fluctuation tests: how reli- mutation rates from fluctuation analysis. G3
able are the estimates of mutation rates? Genet- (Bethesda) 5(11):2323–2327
ics 137(4):1139–1146 9. Hall BM et al (2009) Fluctuation analysis Cal-
4. Rosche WA, Foster PL (2000) Determining culatOR: a web tool for the determination of
mutation rates in bacterial populations. Meth- mutation rate using Luria-Delbr€ uck fluctua-
ods 20(1):4–17 tion analysis. Bioinformatics 25
5. Foster PL (2006) Methods for determining (12):1564–1565
spontaneous mutation rates. Methods Enzy- 10. Stewart FM (1991) Fluctuation analysis: the
mol 409:195–213 effect of plating efficiency. Genetica 84
6. Sarkar S, Ma WT, Sandri GH (1992) On fluc- (1):51–55
tuation analysis: a new, simple and efficient 11. Stewart FM, Gordon DM, Levin BR (1990)
method for computing the expected number Fluctuation analysis: the probability distribu-
of mutants. Genetica 85(2):173–179 tion of the number of mutants under different
7. Lang GI, Murray AW (2008) Estimating the conditions. Genetics 124(1):175–185
per-base-pair mutation rate in the yeast Saccha- 12. Zheng Q (2005) New algorithms for Luria-
romyces cerevisiae. Genetics 178(1):67–82 Delbr€uck fluctuation analysis. Math Biosci
8. Gillet-Markowska A, Louvel G, Fischer G 196(2):198–214
(2015) bz-rates: a web tool to estimate
 Chapter 4

Molecular Genetic Characterization of Mutagenesis

Using a Highly Sensitive Single-Stranded DNA Reporter
System in Budding Yeast
Kin Chan

Abstract
Mutations are permanent alterations to the coding content of DNA. They are starting material for the
Darwinian evolution of species by natural selection, which has yielded an amazing diversity of life on Earth.
Mutations can also be the fundamental basis of serious human maladies, most notably cancers. In this
chapter, I describe a highly sensitive reporter system for the molecular genetic analysis of mutagenesis,
featuring controlled generation of long stretches of single-stranded DNA in budding yeast cells. This
system is ~100- to ~1000-fold more susceptible to mutation than conventional double-stranded DNA
reporters, and is well suited for generating large mutational datasets to investigate the properties of
mutagens.

Key words Mutagenesis, Genomic instability, DNA damage, Mutation clusters, Localized hypermu-
tation, Translesion DNA synthesis

1 Introduction

Mutations are an important form of genetic diversity within popu-

lations of biological entities [1]. Mutations can be the end result of
biochemical or physical processes that damage the nitrogenous
bases in DNA. Such processes can be endogenous to cells, includ-
ing spontaneous hydrolytic, oxidative, or alkylation damage [2].
Base damage can also be induced by exogenous mutagens, includ-
ing chemical species [3] and high-energy radiation [4]. Model
systems have been used extensively to investigate spontaneous and
induced mutagenesis, yielding important insights into these pro-
cesses ([5] and references therein).
Conventional systems for studying mutagenesis typically utilize
a reporter gene, such that mutational inactivation or reversion of
the reporter can be detected using selection media [5]. In these
systems, the reporter gene can be assumed to exist in the canonical
double-stranded DNA (dsDNA) state most of the time, as the

Marco Muzi-Falconi and Grant W. Brown (eds.), Genome Instability: Methods and Protocols, Methods in Molecular Biology,
vol. 1672, DOI 10.1007/978-1-4939-7306-4_4, © Springer Science+Business Media LLC 2018

33
34 Kin Chan

DNA should be transiently single stranded (i.e., existing as ssDNA)

only during transcription, replication, or repair. dsDNA is less
prone to damage than ssDNA, since the nitrogenous bases are less
solvent accessible in the former state [2]. Additionally, dsDNA is
less susceptible to mutation even if base damage does occur, since
an undamaged strand can serve as the template for the repair of the
damaged strand [6]. As a result, conventional dsDNA reporter
systems can be used to elucidate the mutagenic properties of rela-
tively strong mutagens, but can yield ambiguous results for weakly
acting mutagens (see also Discussion in [7]).
Curiously, a number of weakly acting mutagens are also sus-
pected carcinogens [8]. But the weak mutagenic signal induced by
these agents is difficult to distinguish above the background of
spontaneous mutagenesis using conventional dsDNA systems. To
obtain robust data on the mutagenic properties of weakly acting
agents, a much more sensitive reporter system created by the Gor-
denin group at the U.S. National Institute of Environmental
Health Sciences has been used to good effect ([7] and see Fig. 1).

Cdc13-1
ADE2 CAN1 3'
A
URA3 5'
telomere uncapping,
shift to 37oC
5' → 3' resection

3'
B
5'

keep at 37oC mutagen treatment

3'
C
5'

TLS creates mutations

restore to 23oC
opposite base damage
3'
D
5'

Fig. 1 (a) Three reporter genes (CAN1, URA3, and ADE2) are embedded near the
left telomere of Chromosome V in a cdc13-1 haploid. (b) Shifting to 37
C causes
telomere uncapping, and ensuing 50 to 30 resection generates a long single-
stranded overhang encompassing the reporter cassette. (c) Treatment with a
mutagen creates base damage in the 30 overhang. (d) Restoration to 23
C
enables resynthesis of the resected strand. Error-prone translesion DNA
synthesis (TLS) creates mutations opposite the base damage
 Single-Stranded DNA Mutagenesis Reporter in Yeast 35

Three reporter genes (CAN1, URA3, and ADE2) were deleted

from their native loci and reintroduced near a de novo telomere
on the left arm of Chromosome V (see Fig. 1a). Due to the cdc13-
1 temperature sensitive mutation, the proteinaceous capping
complex at the telomeres dissociates when cells are shifted to
37
C (see Fig. 1b). The exposed telomere end is then enzymati-
cally resected in the 50 to 30 direction [9]. This resection generates
a long 30 overhang that encompasses the three reporter genes.
The exposed ssDNA triggers the DNA damage checkpoint,
arresting cells in the G2 phase of the cell cycle [10]. When these
cells are treated with a mutagen, base damage is induced within
the 30 ssDNA overhang (see Fig. 1c). Upon restoration to permis-
sive temperature (23
C), the DNA is restored to a double-
stranded state but in the process, specialized error-prone transle-
sion DNA synthesis (TLS) polymerases create mutations opposite
to the sites of base damage (see Fig. 1d). By plating the cells on
selection media, it is possible to identify clusters of closely spaced
mutations which inactivate multiple reporter genes. These muta-
tion clusters are a form of localized hypermutation and are remi-
niscent of similar strand-biased clusters of mutations observed,
e.g., in cancer genomes [11, 12]. The current generation reporter
system is a considerable improvement over earlier, less sophisti-
cated versions, which had been used to study clustered mutagen-
esis [13, 14].
The ssDNA system has been used in multiple molecular genetic
studies of mutagenesis by chemical mutagens [7, 15] and cytidine
deaminases [7, 16, 17]. ssDNA is some 100- to 1,000-fold more
susceptible to mutation than dsDNA. As a result, the ssDNA
reporter system is sufficiently sensitive to detect mutation clusters
induced by weak mutagens that are essentially inert toward dsDNA
[7]. Furthermore, the genetic requirements for mutagenic TLS
bypass opposite to abasic sites in the ssDNA system are consistent
with the consensus view in the field [16]. Since abasic sites are well-
studied mutagenic lesions, this validation lends additional confi-
dence that the ssDNA system can be used to study mutagenic TLS
polymerase-mediated bypass of other kinds of base damage as well.
A high frequency of selectable mutation clusters is a reliable proxy
for high mutation counts elsewhere in the genome, especially
within unselected mutation clusters near other telomeres. For exam-
ple, overexpression of the human APOBEC3B cytidine deaminase
generated an average of >400 unique mutations per sequenced
yeast genome of ~12 megabases [17]. Also note that due to the
known strandedness of the 30 ssDNA overhang, the exact identity
of base substitutions can be assigned. This is an additional advan-
tage over conventional dsDNA reporter systems, which cannot
distinguish the specific substitution type that is actually induced
by a mutagen (e.g., distinguishing C to A from G to T).
36 Kin Chan

2 Materials

2.1 Day 1: Yeast 1. ySR127 ssDNA reporter yeast. This is isogenic to CG379 [18],
Cultures with the following genotype: MATα his7-2 leu2-3,112 trp1-289
cdc13-1. The reporter gene cassette (lys2::CAN1-URA3-
ADE2) is situated ~3 kilobases from the left telomere of Chro-
mosome V.
2. yKC023 dsDNA reporter control yeast (optional). The
reporter gene cassette is in Chromosome II, 345 kilobases
from the right telomere and 232 kilobases from the
centromere.
3. YPDA liquid media: For 1 L, autoclave the following mixture
for 30 min at 121
C: 20 g D-glucose, 20 g peptone, 10 g yeast
extract, and deionized water to 1 L total volume. Supplement
with 2 mL of 0.5% adenine sulfate solution (filter sterilized).
4. 75 mL tissue culture flasks. Microbiological culture tubes or
50 mL Falcon tubes will do as well.
5. Refrigerated incubating shaker.

2.2 Day 4: Mutagen 1. 10 cm petri dishes, each filled with 30–35 mL of media.
Treatment 2. NaOH solution (5 N).
3. Synthetic Complete solid media: For 1 L, autoclave the follow-
ing mixture for 30 min at 121
C: 20 g agar, 20 g D-glucose,
5 g ammonium sulfate, 1.7 g yeast nitrogen base without
amino acids or ammonium sulfate, 60 mg adenine sulfate,
50 mg L-arginine HCl, 75 mg L-aspartic acid, 100 mg L-glu-
tamic acid, 20 mg L-histidine HCl, 50 mg L-isoleucine, 100 mg
L-leucine, 120 mg L-lysine HCl, 20 mg L-methionine, 50 mg L-
phenylalanine, 375 mg L-serine, 100 mg L-threonine, 50 mg L-
tryptophan, 50 mg L-tyrosine, 150 mg L-valine, 60 mg uracil,
and deionized water to 1 L total volume. Adjust pH to 5.8
before autoclaving. Dispense into 10 cm petri dishes.
4. Canavanine, Low Adenine solid media (see Note 1): For 1 L,
autoclave the following mixture for 30 min at 121
C: 20 g
agar, 20 g D-glucose, 5 g ammonium sulfate, 1.7 g yeast
nitrogen base without amino acids or ammonium sulfate,
20 mg adenine sulfate, 75 mg L-aspartic acid, 100 mg L-glu-
tamic acid, 20 mg L-histidine HCl, 50 mg L-isoleucine, 100 mg
L-leucine, 120 mg L-lysine HCl, 20 mg L-methionine, 50 mg L-
phenylalanine, 375 mg L-serine, 100 mg L-threonine, 50 mg L-
tryptophan, 50 mg L-tyrosine, 150 mg L-valine, 60 mg uracil,
and deionized water to 1 L total volume. After cooling media
to 60
C, add 6 mL of 1% canavanine sulfate solution (filter
sterilized), stir thoroughly to mix, then pour plates. Adjust pH
to 5.8 before autoclaving. Dispense into 10 cm petri dishes.
 Single-Stranded DNA Mutagenesis Reporter in Yeast 37

5. Sterile water.
6. Cell spreader, metal or glass.
7. 95% ethanol.
8. Bunsen burner.
9. Turntable for spreading cells (optional).
10. Sterile glass beads (optional).
11. Incubator set to 23
C.
12. Hemocytometer (optional).
13. Microscope (optional).

2.3 Day 9: Colony 1. Colony counter, automated or manual.

Counting and 2. YPG solid media: For 1 L, autoclave the following mixture for
Phenotype Verification 30 min at 121
C: 20 g agar, 20 g glycerol, 20 g peptone, 10 g
yeast extract, and deionized water to 1 L total volume. Dis-
pense into 10 cm petri dishes.
3. Uracil Dropout solid media: For 1 L, autoclave the following
mixture for 30 min at 121
C: 20 g agar, 20 g D-glucose, 5 g
ammonium sulfate, 1.7 g yeast nitrogen base without amino
acids or ammonium sulfate, 60 mg adenine sulfate, 50 mg L-
arginine HCl, 75 mg L-aspartic acid, 100 mg L-glutamic acid,
20 mg L-histidine HCl, 50 mg L-isoleucine, 100 mg L-leucine,
120 mg L-lysine HCl, 20 mg L-methionine, 50 mg L-phenylal-
anine, 375 mg L-serine, 100 mg L-threonine, 50 mg L-trypto-
phan, 50 mg L-tyrosine, 150 mg L-valine, and deionized water
to 1 L total volume. Adjust pH to 5.8 before autoclaving.
Dispense into 10 cm petri dishes.
4. YPDA solid media: For 1 L, autoclave the following mixture for
30 min at 121
C: 20 g agar, 20 g D-glucose, 20 g peptone,
10 g yeast extract, and deionized water to 1 L total volume.
Supplement with 2 mL of 0.5% adenine sulfate solution (filter
sterilized). Dispense into 10 cm petri dishes.
5. Adenine Dropout solid media: For 1 L, autoclave the following
mixture for 30 min at 121
C: 20 g agar, 20 g D-glucose, 5 g
ammonium sulfate, 1.7 g yeast nitrogen base without amino
acids or ammonium sulfate, 50 mg L-arginine HCl, 75 mg L-
aspartic acid, 100 mg L-glutamic acid, 20 mg L-histidine HCl,
50 mg L-isoleucine, 100 mg L-leucine, 120 mg L-lysine HCl,
20 mg L-methionine, 50 mg L-phenylalanine, 375 mg L-serine,
100 mg L-threonine, 50 mg L-tryptophan, 50 mg L-tyrosine,
150 mg L-valine, 60 mg uracil, and deionized water to 1 L total
volume. Adjust pH to 5.8 before autoclaving. Dispense into
10 cm petri dishes.
6. Velveteen squares.
7. Replica plating block.
38 Kin Chan

3 Methods

3.1 Day 1: Yeast 1. For each replicate culture, inoculate ySR127 yeast from a single
Cultures colony into 5 mL of YPDA in each flask.
2. Shake at 23
C and 250 revolutions per minute (RPM) for
3 days, with flasks in a vertical orientation (see Note 2).
3. (Optional) A control yeast strain is available (yKC023), where
the triple reporter gene cassette is situated at an internal locus
in Chromosome II [7]. Telomere uncapping and resection also
occur at 37
C, but the reporter DNA remains in a double-
stranded state. This control can be included to assess the muta-
genicity of a given agent toward dsDNA.

3.2 Day 4: Mutagen 1. Dilute cultures from Day 1 by tenfold using fresh YPDA (e.g.,
Exposure dilute 500 μL of culture with 4.5 mL of YPDA) in fresh flasks.
2. Shake at 37
C for 3–6 h at 250 RPM (see Note 3).
3. Add exogenous mutagen and maintain at 37
C, with shaking,
for desired duration of treatment. Optimal duration of treat-
ment and mutagen concentration will have to be worked out by
trial and error. Set up mock treated controls in parallel (see
Note 4).
4. Wash cells from Day 1 cultures in water three times. Determine
cell titer using microscope and hemocytometer (see Note 5).
5. Spread appropriate dilutions from Day 1 cultures on Synthetic
Complete and on Canavanine, Low Adenine plates to assess
baseline survival and mutant frequency, respectively (see Note
6). This can be done using a cell spreader (flame sterilized with
ethanol between platings) with or without a turntable, or by
shaking sterile glass beads to spread cells around a plate (see
Note 7). All platings should be done in triplicate.
6. When mutagen treatment is completed, wash treated cells and
mock treated controls in water three times. Determine cell
titers using microscope and hemocytometer. Spread appropri-
ate dilutions on Synthetic Complete and on Canavanine, Low
Adenine (see Note 8).
7. Incubate plates at 23
C for 5 days (see Note 9).

3.3 Day 9: Colony 1. Count the number of colonies on Synthetic Complete plates.
Counting and Divide this number by the number of cells plated to obtain the
Phenotype Verification fraction of viable, colony forming cells within a sample. This is
the plating efficiency.
2. Count the total number of colonies on Canavanine, Low Ade-
nine plates. This is the total number of colonies with CAN1
inactivation (see Note 10).
 Single-Stranded DNA Mutagenesis Reporter in Yeast 39

3. Count the number of red (and pink) colonies on Canavanine,

Low Adenine plates. This is an estimate of the number of
colonies with both CAN1 and ADE2 inactivated (see Note 11).
4. Replica the Canavanine, Low Adenine plates onto: (fresh)
Canavanine, Low Adenine; YPG; and Uracil Dropout media,
in that order (see Note 12).
5. After incubation at 23
C, compare these replicas to their
source plates. Colonies that fail to grow on YPG have lost
mitochondrial respiratory function, while those that fail to
grow on Uracil Dropout are inactivated for both CAN1 and
URA3 (see Note 13).
6. Streak from colonies of interest (i.e., Canr Ade, Canr Ura, or
Canr Ade Ura) onto YPDA and grow at 23
C.
7. Patch a single colony from each streak onto YPDA and grow at
23
C. ~2 cm-long patches will suffice.
8. Replica the patches onto the following media to verify pheno-
types: Canavanine, Low Adenine; YPG; Adenine Dropout; and
Uracil Dropout, in that order. Patches with confirmed pheno-
types of interest are genetically homogeneous and suitable for
downstream analyses, e.g., sequencing to identify mutations
(see Note 14).

4 Notes

1. Be sure to omit arginine from these media to promote efficient

uptake of its toxic analog, canavanine, into cells via functional
Can1 protein. This is critical, as it is the basis for the selection of
canavanine resistant (i.e., Canr) mutants.
2. Do not exceed 23
C, as doing so risks sporadic telomere
uncapping. Growth for 3 days effectively synchronizes the
vast majority of cells in the G1 phase, since the cultures are
well past logarithmic growth. Growth for longer is also fine.
Using cultures grown for only 2 days is possible, but at the cost
of somewhat less uniform arrest at 37
C. Growth in tubes will
be slower than in flasks and cell titers likely will not be as high.
3. Vigorous shaking in flasks promotes synchronous arrest. A 6-
h incubation at 37
C yields a more uniformly arrested cell
population than 3-h incubation, as the former presumably
allows more cells enough time to reach the G2 arresting point.
4. It is sensible to check (under the microscope) that the cultures
are arrested properly prior to adding mutagen. >90% of cells
should be arrested as large dumbbells, where the buds are
roughly the same size as mother cells.
40 Kin Chan

5. It is possible to skip cell counting and simply plate sensible

dilutions of cells. This would be adequate for determining
mutant frequency, but the fraction of viable cells within the
sample would be unknown. This latter quantity, the plating
efficiency, is important for assessing whether a treatment is
cytotoxic.
6. As starting points, I recommend spreading 1,000 cells per
Synthetic Complete and 108 per Canavanine, Low Adenine
plate. Spreading more than 108 per selection plate is inadvis-
able, as pseudo-resistant papillae can grow when cells are plated
at too high of a density. 3-day old starter cultures can have
>109 cells per mL. Thus if skipping cell counting, one can try
spreading a 106- and/or 107-fold dilution on Synthetic Com-
plete and a 10- and/or 100-fold dilution on Canavanine, Low
Adenine. A culture with an unusually high spontaneous Canr
frequency should be dropped from further consideration, as
pre-existing Canr mutations in the starting culture will con-
found the analysis of induced mutagenesis. The spontaneous,
background Canr frequency is typically ~106 per viable cell.
7. The glass bead method is quicker, as one can add cells to
multiple plates and swirl the plates altogether as a stack. The
drawback is that a non-negligible fraction of colonies would
grow up near the rim of the plates and would not be included
when replica plating to check for respiratory competency and
inactivation of URA3.
8. Depending on the mutagen treatment, as few as 104 and as
many as 106 cells should be spread per Canavanine, Low Ade-
nine plate. A typical pilot experiment should try 1,000 cells per
Synthetic Complete, and 105 and/or 106 cells per Canavanine,
Low Adenine plate. Mock-treated controls should be tested
initially with 106 cells per Canavanine, Low Adenine plate.
Adjustments to platings can be made in follow-on experiments.
If plating cells without counting titers, choose order of magni-
tude dilutions that should yield (approximately) the desired
number of cells per plate as mentioned here.
9. 5 days’ incubation at 23
C is a bare minimum, but identifica-
tion of red colonies can be unreliable because there may not be
enough time for red pigments to build up in some mutant
colonies. For the most reliable quantification of red colonies,
7 days at 23
C followed by 7 days at 4
C are recommended.
Low temperature incubation markedly intensifies red
pigmentation.
10. To estimate the number of viable cells plated on Canavanine,
Low Adenine, multiply the total number of plated cells by the
plating efficiency from the Synthetic Complete plates. To cal-
culate mutant frequency, divide the number of Canr colonies
 Single-Stranded DNA Mutagenesis Reporter in Yeast 41

by the number of viable cells plated. If plating dilutions with

undetermined titers, the number of viable cells plated on
Canavanine, Low Adenine can still be estimated. For example,
suppose 500 colonies are observed on Synthetic Complete
when plated with a 106-fold dilution. If a tenfold dilution is
plated on selection media, then one would expect
500  105 ¼ 5  107 viable cells.
11. This only gives an estimate for Canr Ade mutation cluster
frequency, as the entire reporter cassette can be lost at a fre-
quency of ~106 per viable cell plated. Presumably, this is due
to strand breakage within the long 30 ssDNA overhang. Cas-
sette loss events can be recognized easily, as the red pigmenta-
tion emerges early upon incubation and is always a uniformly
deep hue; these colonies tend to be small; and they are always
Ura.
12. Take care to use a light touch when replica plating, so that
excessive numbers of cells are not transferred onto the replicas.
This is especially important when replica plating onto Uracil
Dropout media, as too heavy of an imprint will result in a high
background that can make Ura colonies hard to distinguish
reliably.
13. Colonies that are respiratory deficient (i.e., ρ or rho) should
be dropped from further consideration, as these colonies have
profoundly altered metabolism. Also note that respiratory defi-
ciency can result in reddish pigmentation, so diagnostic replicas
are necessary to distinguish between Ade- and ρ.
14. Partial loss-of-function Ade- patches will grow slowly on Ade-
nine Dropout. These can exhibit a range of reddish hues, so it is
important to compare with YPG to verify that the redness
cannot be attributed to respiratory deficiency.

Acknowledgments

KC thanks Dr. Dmitry A. Gordenin for an historical perspective on

the development and usage of previous generation ssDNA reporter
systems. This work was supported by NSERC Discovery Grant
147465 and start-up funding from the University of Ottawa.

References

1. Ayala FJ, Fitch WM (1997) Genetics and the 3. Irigaray P, Belpomme D (2010) Basic proper-
origin of species: an introduction. Proc Natl ties and molecular mechanisms of exogenous
Acad Sci 94:7691–7697 chemical carcinogens. Carcinogenesis
2. Lindahl T (1993) Instability and decay of the 31:135–148. doi:10.1093/carcin/bgp252
primary structure of DNA. Nature 4. Ikehata H, Ono T (2011) The mechanisms of
362:709–715. doi:10.1038/362709a0 UV mutagenesis. J Radiat Res 52:115–125.
doi:10.1269/jrr.10175
42 Kin Chan

5. Rogozin IB, Pavlov YI (2003) Theoretical McLaren S, Butler AP, Teague JW, Jönsson G,
analysis of mutation hotspots and their DNA Garber JE, Silver D, Miron P, Fatima A, Boy-
sequence context specificity. Mutat Res ault S, Langerød A, Tutt A, Martens JWM,
544:65–85. doi:10.1016/S1383-5742(03) Aparicio SAJR, Borg Å, Salomon AV, Thomas
00032-2 G, Børresen-Dale A-L, Richardson AL, Neu-
6. Boiteux S, Jinks-Robertson S (2013) DNA berger MS, Futreal PA, Campbell PJ, Stratton
repair mechanisms and the bypass of DNA MR (2012) Mutational processes molding the
damage in saccharomyces cerevisiae. Genetics genomes of 21 breast cancers. Cell
193:1025. doi:10.1534/genetics.112.145219 149:979–993. doi:10.1016/j.cell.2012.04.
7. Chan K, Sterling JF, Roberts SA, Bhagwat AS, 024
Resnick MA, Gordenin DA (2012) Base dam- 13. Yang Y, Sterling J, Storici F, Resnick MA, Gor-
age within single-strand DNA underlies in vivo denin DA (2008) Hypermutability of damaged
hypermutability induced by a ubiquitous envi- single-strand DNA formed at double-strand
ronmental agent. PLoS Genet 8:e1003149. breaks and uncapped telomeres in yeast saccha-
doi:10.1371/journal.pgen.1003149 romyces cerevisiae. PLoS Genet 4:e1000264.
8. International Agency for Research on Cancer doi:10.1371/journal.pgen.1000264
(2010) Identification of research needs to 14. Burch LH, Yang Y, Sterling JF, Roberts SA,
resolve the carcinogenicity of high-priority Chao FG, Xu H, Zhang L, Walsh J, Resnick
IARC carcinogens MA, Mieczkowski PA, Gordenin DA (2011)
9. Booth C, Griffith E, Brady G, Lydall D (2001) Damage-induced localized hypermutability.
Quantitative amplification of single-stranded Cell Cycle 10:1073–1085. doi:10.4161/cc.
DNA (QAOS) demonstrates that cdc13-1 10.7.15319
mutants generate ssDNA in a telomere to cen- 15. Degtyareva NP, Heyburn L, Sterling J, Resnick
tromere direction. Nucleic Acids Res MA, Gordenin DA, Doetsch PW (2013) Oxi-
29:4414–4422. doi:10.1093/nar/29.21. dative stress-induced mutagenesis in single-
4414 strand DNA occurs primarily at cytosines and
10. Garvik B, Carson M, Hartwell L (1995) Single- is DNA polymerase zeta-dependent only for
stranded DNA arising at telomeres in cdc13 adenines and guanines. Nucleic Acids Res
mutants may constitute a specific signal for 41:8995–9005. doi:10.1093/nar/gkt671
the RAD9 checkpoint. Mol Cell Biol 16. Chan K, Resnick MA, Gordenin DA (2013)
15:6128–6138. doi:10.1128/MCB.15.11. The choice of nucleotide inserted opposite aba-
6128 sic sites formed within chromosomal DNA
11. Roberts SA, Sterling J, Thompson C, Harris S, reveals the polymerase activities participating
Mav D, Shah R, Klimczak LJ, Kryukov GV, in translesion DNA synthesis. DNA Repair
Malc E, Mieczkowski PA, Resnick MA, Gorde- 12:878–889. doi:10.1016/j.dnarep.2013.07.
nin DA (2012) Clustered mutations in yeast 008
and in human cancers can arise from damaged 17. Chan K, Roberts SA, Klimczak LJ, Sterling JF,
long single-strand DNA regions. Mol Cell Saini N, Malc EP, Kim J, Kwiatkowski DJ,
46:424–435. doi:10.1016/j.molcel.2012.03. Fargo DC, Mieczkowski PA, Getz G, Gordenin
030 DA (2015) An APOBEC3A hypermutation
12. Nik-Zainal S, Alexandrov LB, Wedge DC, Van signature is distinguishable from the signature
Loo P, Greenman CD, Raine K, Jones D, Hin- of background mutagenesis by APOBEC3B in
ton J, Marshall J, Stebbings LA, Menzies A, human cancers. Nat Genet 47:1067–1072
Martin S, Leung K, Chen L, Leroy C, Ramak- 18. Morrison A, Bell JB, Kunkel TA, Sugino A
rishna M, Rance R, Lau KW, Mudie LJ, Varela (1991) Eukaryotic DNA polymerase amino
I, McBride DJ, Bignell GR, Cooke SL, Shlien acid sequence required for 30 ! 50 exonuclease
A, Gamble J, Whitmore I, Maddison M, Tarpey activity. Proc Natl Acad Sci 88:9473–9477.
PS, Davies HR, Papaemmanuil E, Stephens PJ, doi:10.1073/pnas.88.21.9473
 Chapter 5

Analyzing Genome Rearrangements in Saccharomyces

cerevisiae
Anjana Srivatsan, Christopher D. Putnam, and Richard D. Kolodner

Abstract
Genome rearrangements underlie different human diseases including many cancers. Determining the rates
at which genome rearrangements arise and isolating unique, independent genome rearrangements is critical
to understanding the genes and pathways that prevent or promote genome rearrangements. Here, we
describe quantitative S. cerevisiae genetic assays for measuring the rates of accumulating genome rearrange-
ments including deletions, translocations, and broken chromosomes healed by de novo telomere addition
that result in the deletion of two counter-selectable genes, CAN1 and URA3, placed in the nonessential
regions of the S. cerevisiae genome. The assays also allow for the isolation of individual genome rearrange-
ments for structural studies, and a method for analyzing genome rearrangements by next-generation DNA
sequencing is provided.

Key words Genome instability, Deletion, Monocentric translocation, Dicentric translocation,

De novo telomere addition, Genetics, GCR rates, Whole-genome sequencing

1 Introduction

Maintaining the stability of the genome during cell division is

critical for cell survival and normal cell growth. Understanding
the pathways and mechanisms by which cells prevent genome rear-
rangements (herein called Gross Chromosomal Rearrangements or
GCRs) has been greatly facilitated by the development of simple
genetic methods for measuring the rate at which GCRs occur in
different wild-type and mutant S. cerevisiae strains [1–3]. Such
methods have made it possible to identify genes in which defects
cause increased or decreased GCR rates, perform epistasis analysis
of these mutations, and isolate independent GCRs for structural
analysis (for example see refs. 4, 5). These types of studies have
provided a growing picture of the cellular pathways that act to
prevent GCRs as well as the pathways that produce GCRs.
The key observation that led to the development of a broad
series of assays for measuring the rate of accumulation of GCRs was

Marco Muzi-Falconi and Grant W. Brown (eds.), Genome Instability: Methods and Protocols, Methods in Molecular Biology,
vol. 1672, DOI 10.1007/978-1-4939-7306-4_5, © Springer Science+Business Media LLC 2018

43
44 Anjana Srivatsan et al.

A. “Classical” GCR assay

URA3
CEN5
DSF1 HXT13 CAN1 PCM1

breakpoint region

B. Observed Canr 5FOAr GCRs

CEN5
Interstitial deletion PCM1

CEN5
De novo telomere addition PCM1

CEN5
Translocation other chromosome PCM1

C. uGCR assay
URA3 CAN1
CEN5
DSF1 HXT13 PCM1

breakpoint region

D. dGCR assay
URA3 CAN1
CEN5
DSF1 HXT13 PCM1

breakpoint region

Fig. 1 GCR assays. (a) The “classical” GCR assay is constructed by inserting URA3 into HXT13. The breakpoint
region (dashed line) is between the most telomeric essential gene, PCM1, and the most centromeric marker
gene, CAN1. (b) A variety of GCR products have been observed, including interstitial deletions, terminal
deletions healed by de novo telomere additions, and translocations. The types of translocations are dictated by
the orientation, presence, or absence of a centromere and source of the sequence joined to the broken
chromosome V. (c) The unique sequence or uGCR assay contains a CAN1 URA3 cassette centromeric to the
DSF1-HXT13 segmental duplication (gray). (d) The duplication-mediated or dGCR assay contains a CAN1 URA3
cassette telomeric to the DSF1-HXT13 segmental duplication (gray)

that the CAN1-containing portion of the left arm of chromosome

V from the telomere to PCM1 was nonessential [6]. This allowed
the development of the first GCR assay (referred to as the “classi-
cal” GCR assay) in which the URA3 gene was inserted into HXT13
between CAN1 and the telomere (Fig. 1; [2]). During normal
mitotic growth of haploid cells containing this genetically marked
chromosome, progeny arise that are resistant to both canavanine
(Can) and 5-fluoroorotic acid (5-FOA). These double-drug-resis-
tant progeny contain a rearranged chromosome V in which URA3
and CAN1 have been deleted by the formation of GCRs including:
 Genome Instability in S. Cerevisiae 45

(1) interstitial deletions, (2) terminal deletions of chromosome V L

that are healed by the addition of a new telomere, and (3) terminal
deletions of chromosome V L healed by fusion to a fragment of
another chromosome, generating a translocation, which can either
be stable monocentric translocations or dicentric translocations
that are unstable and undergo additional rounds of rearrangement
(Fig. 1) [2, 6–8]. By measuring the rate of occurrence of Carr 5-
FOAr progeny, GCR rates can be determined and by determining
the structure of the GCRs present in individual progeny, the rates of
occurrence of specific types of GCRs can be determined.
To increase the utility of GCR assays, CAN1 and URA3 were
inserted into a cassette. This made it possible to move the two
counter-selectable genes to different chromosomal locations in
haploid strains containing mutations at the CAN1 and URA3
chromosomal loci to evaluate the effect of chromosomal context
on GCR rates. In two such assays (Fig. 1), the CAN1 URA3
cassette was placed either centromeric (the yel068c::CAN1/URA3
or uGCR assay) or telomeric (the yel072w::CAN1/URA3 or dGCR
assay) to the HXT13-DSF1 segmental duplication region on chro-
mosome V L [3]. These assays revealed strong influences of having
a sequence in the breakpoint region with divergent homology with
other regions of the genome on GCR rates, on the pathways that
suppress GCRs and the mechanisms by which GCRs are formed
[3]. In addition to these GCR assays, a number of other haploid
strain-based GCR assays utilizing counter-selectable markers in
nonessential regions of the genome have been described (for exam-
ple see refs. 9, 10). In addition, GCR assays in which amplification
of genetic markers is selected for in haploid strains have been
described, as have a limited number of diploid strain-based GCR
assays (for example see refs. 11–13).
GCR structures can be analyzed using a variety of methods
such as PCR-based mapping and amplification of breakpoint junc-
tions, multiplex ligation-dependent probe amplification (MLPA),
pulsed-field gel electrophoresis (PFGE) combined with Southern
blotting, and microarray-based comparative genome hybridization
(aCGH), and usually, determining the complete structure of an
individual GCR requires multiple methods and approaches [2, 3,
7, 14, 15]. A method that has been recently adapted for the analysis
of GCR structures is next-generation whole-genome sequencing
(NGS) (Fig. 3; [16, 17]). Analysis of GCRs using paired-end
sequencing, which provides sequence from both ends of genomic
DNA fragments, is particularly useful for the analysis of GCRs and
should be the method used when performing NGS on GCR-
containing strains. Compared to prior methods, NGS is less
labor-intensive and more rapidly provides genome-wide data on
copy number, point mutations, and genome rearrangements. Fur-
thermore, the small size of the S. cerevisiae genome allows multiple
46 Anjana Srivatsan et al.

strains to be sequenced simultaneously (multiplexed), greatly

reducing the cost of analysis.
In this article, we describe the use of GCR assays based on
selection against the presence of URA3 and CAN1 to both mea-
sure GCR rates and isolate progeny containing unique GCRs for
structural analysis. We also describe a method for preparing 600 bp
multiplexed libraries from S. cerevisiae strains for paired-end whole-
genome sequencing using an Illumina HiSeq 2500 for characteriz-
ing GCR structures (see Note 1). The general approach described is
applicable to many other types of GCR assays.

2 Materials

2.1 Determining GCR 1. YPD (Yeast extract Peptone Dextrose) liquid medium: For 1 l
Rates of medium, add 10 g Bacto-yeast extract and 20 g Bacto-
peptone to 950 ml deionized water, and sterilize by autoclaving
at 121
C for 20 min. Cool and add 50 ml of a 40% (w/v)
sterile glucose (D-dextrose) solution.
2. YPD plates: For 1 l of medium (40 plates), add 10 g Bacto-
yeast extract, 20 g Bacto-peptone, and 21 g of Bacto-agar to
950 ml deionized water in a 2 l flask, place a magnetic stir bar in
the flask and sterilize by autoclaving at 121
C for 20 min. Cool
to 60
C and add 50 ml of a 40% (w/v) sterile glucose (D-
dextrose) solution. Mix well by stirring and pour 25 ml per
plate in 100  15 mm round petri dishes. Allow the plates to
dry for 1–2 days at room temperature.
3. 40% (w/v) glucose solution: Dissolve 400 g glucose (D-dex-
trose) in 600 ml warm deionized water with continuous stir-
ring, adjust the volume to 1000 ml, and sterilize by autoclaving
at 121
C for 20 min.
4. Can 5-FOA plates: Add 23–25 g Bacto-agar to 750 ml deio-
nized water in a 2 l flask and place a magnetic stir bar in the
flask. Sterilize the agar mixture by autoclaving at 121
C for
20 min and then stir on a magnetic stir plate to allow the
mixture to cool to ~65
C (see Note 2). While the agar mixture
cools, prepare the dropout mix as follows. Dissolve 60 mg L-
canavanine sulfate (Can) and 1 g 5-fluoroorotic acid (5-FOA)
in 200 ml warm deionized water. Add 2 g of dropout powder
minus arginine (US Biological, D9518), 6.7 g of yeast nitrogen
base without amino acids (with ammonium sulfate), and 50 ml
of 40% (w/v) sterile glucose solution, and stir well to dissolve.
The dropout solution can be filter sterilized, although this is
generally not necessary because the combination of Can and 5-
FOA is capable of killing most potential contaminants. Keep
the dropout mix at 55–60
C until the autoclaved agar mixture
has cooled sufficiently. Carefully transfer the dropout mix to
 Genome Instability in S. Cerevisiae 47

the autoclaved agar mixture, avoiding the formation of air

bubbles. Pour approximately 70 ml per plate in
150  15 mm round petri dishes or 25 ml per plate in
100  15 mm round petri dishes as needed. Allow the plates
to dry for 1–2 days at room temperature.
5. Sterile deionized water.
6. Sterile plastic culture tubes (15 ml round-bottomed tubes,
50 ml conical tubes).
7. Sterile glass 250 ml culture flasks for larger cultures.
8. Sterile toothpicks to streak out, patch, and inoculate colonies.
9. Surgical scalpel.
10. Sterile microcentrifuge tubes—1.5 ml.
11. Sterile 4 mm glass beads.
12. Benchtop microcentrifuge.
13. Centrifuge with swinging bucket rotor.
14. Incubator set to 30
C.
15. Shaker set to 30
C.
16. Sterile velvets.
17. Replica-plating block.

2.2 For Preparing 1. YPD liquid medium.

Libraries for NGS to 2. YPD plates.
Analyze GCR
3. 40% (w/v) sterile glucose solution.
Structures
4. Sterile deionized water.
5. Gentra Puregene Yeast/Bact. kit (Qiagen) or preferred geno-
mic DNA extraction kit.
6. 4 mg/ml RNase A (Qiagen or equivalent).
7. Covaris microTUBEs: microTUBE-15AFA beads screw-cap
vials and M220 holder XTU insert microTUBE 15 μl. For
50 μl sample volumes, use microTUBE-50AFA fiber screw-
cap vials and M220 holder XTU insert microTUBE 50 μl.
8. Covaris Focused Ultrasonicator M220.
9. FlashGel DNA Cassette 1.2%, 12 + 1, single tier (Lonza).
10. FlashGel Loading Dye 5  1 ml, 5 concentrate (Lonza).
11. FlashGel DNA Marker 100 bp to 4 kb (Lonza).
12. FlashGel Dock System (Lonza).
13. End-It DNA end-repair kit (Epicentre Technologies).
14. Klenow DNA polymerase (30 –50 exo) (5 U/μl) (New Eng-
land Biolabs (NEB) or equivalent).
15. dATP.
48 Anjana Srivatsan et al.

16. Quick ligation kit (NEB or equivalent).

17. Library amplification readymix (KAPA Biosystems).
18. TruSeq PCR-Free LT DNA sample prep kit set A (Illumina).
19. MinElute PCR purification kit (Qiagen or equivalent).
20. MinElute gel extraction kit (Qiagen or equivalent).
21. Certified low-range ultra agarose (Bio-Rad or equivalent).
22. 50 TAE Buffer, pH 8.0.
23. 10 Orange Loading Buffer (NEB or equivalent).
24. 100 bp DNA ladder (Bioline or similar).
25. SYBR Safe DNA gel stain (Thermo Fisher Scientific).
26. Benchtop microcentrifuge.
27. Microcentrifuge tubes—0.5 ml (clear, thin-walled; Thermo
Fisher Scientific, or equivalent) and 1.5 ml.
28. Centrifuge with swinging bucket rotor.
29. Sterile toothpicks to streak out and inoculate strains.
30. Sterile culture tubes (15 ml round-bottomed tubes, 50 ml
conical tubes).
31. Razor blades and forceps.
32. Nanodrop 2000 UV-Vis spectrophotometer.
33. Dark Reader transilluminator (Iso BioExpress) or other UV
transilluminator.
34. Electrophoresis power supply, slab gel trays, and electrophore-
sis tank.
35. Thermal cycler.
36. Qubit fluorometer.
37. Qubit dsDNA HS assay kit.

3 Methods

3.1 Determining GCR 1. Streak out the S. cerevisiae strains of interest containing the
Rates desired GCR assay (see Note 3) on YPD plates (or an appropri-
ate selective medium if plasmid selection is required), and
incubate the plates for 2–3 days at 30
C to obtain well-
separated single colonies. Streak out at least two independent
biological isolates for each strain.
2. Using a sterile surgical scalpel (see Note 4), excise and inoculate
at least seven colonies from each biological isolate into individ-
ual sterile tubes or flasks containing an appropriate volume (see
Note 5) of YPD liquid medium (or appropriate selective
medium, if plasmid selection is required). Ensure that the
 Genome Instability in S. Cerevisiae 49

entire colony is used for inoculation. Grow the cultures to

saturation (16–18 h; see Note 6) in a 30
C shaker with
vigorous shaking (250–300 rpm).
3. To obtain the viable cell count, plate an appropriate dilution of
each culture on YPD (or appropriate selective medium) plates
as follows. For each culture, prepare five microcentrifuge tubes
containing 90 μl sterile deionized water. Transfer 10 μl of the
saturated culture into the 1st tube and vortex vigorously. Then
transfer 10 μl from the 1st tube into the 2nd tube, and so on,
until the 5th tube, vortexing vigorously before each serial
dilution, to obtain a tenfold dilution series. Transfer the entire
0.1 ml of the 105 dilution to a YPD plate and spread well
using sterile glass beads.
4. To select for cells that contain a GCR, plate an appropriate
volume (see Note 7) of each culture on Can 5-FOA plates. Use
100  15 mm and 150  15 mm petri dishes to plate up to
1 ml (108 cells) and 10 ml (109 cells) of the culture, respec-
tively. Use multiple plates if larger culture volumes must be
plated; for instance, while plating RDKY3615 (the classical
GCR assay), which requires ~50 ml cultures, use five
150  15 mm Can 5-FOA plates per culture, i.e., 10 ml culture
(109 cells) per plate as follows. After plating the appropriate
dilution on YPD plates (step 3), centrifuge the remaining
culture at 3000 rpm (~2000  g) for 10 min at room tempera-
ture and discard the supernatant. For up to 1 ml of culture,
resuspend the cells in 0.1 ml of sterile deionized water and plate
on a 100  15 mm Can 5-FOA plate. For larger culture
volumes, resuspend the cells in 0.2 ml of sterile deionized
water per 10 ml of original culture, and plate 0.2 ml per
150  15 mm Can 5-FOA plate.
5. Incubate the YPD and Can 5-FOA plates at 30
C for 2–3 days
and 3–5 days, respectively.
6. Count and record the number of colonies on each YPD plate
(see Note 8) and the corresponding Can 5-FOA plates (see
Note 9).
7. Calculate the GCR rate using the Lea-Coulson method of the
median [18]. The spreadsheet illustrated in Fig. 2 provides a
method to calculate the median and 95% confidence interval of
the median based on the total volume of the cultures and the
count of the colonies on the nonselective and selective plates. A
crucial step in the calculation is the use of the number of
mutants per culture, r, to estimate m, which is the most likely
number of mutations per culture. The nonlinear relationship
between m and r requires, in each spreadsheet, the CalcM()
macro, which calculates the Lea-Coulson median estimator
using the Newton-Raphson method (Fig. 2).
 A. A B C D E F G H
1 Culture Volume (mL) 15
2
3 Culture Quantitation
4 Culture Dilution Plated Number Cells
5 Number Factor Volume (mL) Colonies per mL =E6/D6*C6
6 1 1.00E+05 0.1 112 1.12E+08
7 2 1.00E+05 0.1 158 1.58E+08
8 3 1.00E+05 0.1 184 1.84E+08 =E18/D18*C18
9 4 1.00E+05 0.1 103 1.03E+08
10 5 1.00E+05 0.1 98 9.80E+07
11 6 1.00E+05 0.1 101 1.01E+08
12 7 1.00E+05 0.1 119 1.19E+08 =CalcM(F18*
13 8 1.00E+05 0.1 135 1.35E+08
9 1.00E+05 0.1 137 1.37E+08
$B$1)/(F6*$B$1)
14
15 Rate Determination
16 Culture Dilution Plated Number Mutants
17 Number Factor Volume (mL) Colonies per mL Rate
18 1 1 14.99 70 4.67 1.02E-08
19 2 1 14.99 67 4.47 6.99E-09
20 3 1 14.99 55 3.67 5.13E-09
21 4 1 14.99 98 6.54 1.46E-08
22 5 1 14.99 112 7.47 1.71E-08
23 6 1 14.99 108 7.2 1.61E-08
24 7 1 14.99 81 5.4 1.08E-08
25 8 1 14.99 132 8.81 1.42E-08
26 9 1 14.99 30 2 4.28E-09
27 Summary
28 Rate Number Lower Pos Upper Pos Median 95% CI 95% CI
29 Location Rates 95% CI 95% CI Rate Lower Upper
30 G18:G26 9 2 8 1.08E-08 5.13E-09 1.61E-08

=COUNT(INDIRECT(B30))
=FLOOR((C30+1)/2-0.9789*SQRT(C30),1)
=C30+1-D30
=MEDIAN(INDIRECT(B30))
=SMALL(INDIRECT(B30),D30)
=SMALL(INDIRECT(B30),E30)
B. ‘ Visual Basic Macro to Calculate m
‘ Defines the function CalcM() for use
‘ in spreadsheet programs
‘
Function CalcM(r As Double) As Double
Dim fm As Double, dfm As Double, m As Double
If (r <= 0#) Then
m=0
Else
m = Exp ( Log(r) * 0.5 )
fm = m * (1.24 + Log(m)) – r
dfm = 2.24 + Log(m)
While (Abs(fm) > 0.01)
m = m – fm / dfm
fm = m * (1.24 + Log(m)) – r
dfm = 2.24 + Log(m)
Wend
EndIf
CalcM = m
End Function

Fig. 2 Spreadsheet for calculating the median GCR rate of a series of cultures and 95% confidence intervals
for the median GCR rate. (a) The spreadsheet is displayed with both sample data and the underlying formulas.
Callout boxes indicate cells that should contain formulas; the required formulas are indicated by the callout
 Genome Instability in S. Cerevisiae 51

3.2 Preparation of S. 1. Select strains for the analysis by NGS. Multiplexed libraries
cerevisiae Genomic containing DNA from up to 12 strains can be sequenced on
DNA Libraries for an Illumina HiSeq 2500 to obtain sufficient read depth (>20-
Multiplexed Paired- fold) to reliably identify rearrangements, point mutations, and
End NGS copy number changes. GCR-containing strains can be obtained
from Canr 5FOAr colonies on plates from fluctuation tests or
from patches replica plated onto Can 5-FOA medium when
strains have sufficiently high GCR rates (see Note 7). It is
critical, however, to ensure that all GCR-containing strains
analyzed are independently isolated so that multiple descen-
dants of a single GCR-containing cell are not analyzed. Hence,
only one colony per culture used in fluctuation analysis or only
one colony per patch should be analyzed. When sequencing
genomic DNA from GCR-containing strains, it is preferable to
also sequence genomic DNA from any parental strains from
which the GCR-containing strains were derived so that pre-
existing rearrangements and point mutations can be identified.
A typical sample of multiplexed libraries from a GCR experi-
ment will contain libraries from 1 parental strain and 11 inde-
pendently isolated GCR-containing strains; however, the
method can be scaled up or down in regard to the number of
strains analyzed so long as sufficient coverage for each genome
can be achieved.
2. Streak selected GCR-containing strains onto Can 5-FOA plates
for single colonies to purify the isolate. Incubate the freshly
streaked Can 5-FOA plates for 3–5 days at 30
C. Inoculate a
single colony for each GCR-containing strain into 5 ml of YPD
liquid and grow in a 30
C shaker for 12–16 h or until satura-
tion is reached. In parallel, inoculate the parent strain(s) from
which the GCR-containing isolates were derived. To generate a
glycerol stock for each overnight culture of a GCR-containing
strain, take 0.5 ml of the culture, mix with 0.5 ml of 40%
glycerol, and freeze at 80
C.
3. Prepare genomic DNA from the overnight cultures as follows
(see Note 10). Harvest the cells from 1.5 ml of each culture by
centrifugation at 14,000 rpm (~21,000  g) for 30 s to 1 min
in a benchtop microcentrifuge. Carefully remove the superna-
tant by pipetting and discard. Suspend the cells in 1 ml of sterile
ä

Fig. 2 (continued) boxes. For cells F6 to F14, F18 to F26, and G18 to G26, the row specifiers in the formula
should be incremented in each row (for example, F6 should have the formula ¼ E6/D6*C6 and F7 should have
the formula ¼ E7/D7*C7). Most modern spreadsheets have fill functions that will automatically increment
these values. (b) To calculate the rate in cells G18 to G26, the CalcM() Visual Basic macro must be defined for
the spreadsheet. CalcM() initializes the test value for m to the square root of r and uses the iterative Newton-
Raphson method to solve the transcendental equation r/m  ln(m)  1.24 ¼ 0. This macro functions in both
Microsoft Excel and in LibreOffice Calc
52 Anjana Srivatsan et al.

deionized water, harvest the cells again by centrifugation, and

discard the supernatant. The genomic DNA purification
method described here is based on the Gentra PureGene
Yeast/Bact. kit with some modifications. Resuspend the cell
pellet in 300 μl Cell Suspension Solution and pipet to mix well.
Add 1.5 μl Lytic Enzyme Solution, mix well, and incubate the
cells for 1 h at 37
C. Pellet the cells by centrifugation at
14,000 rpm (~21,000  g) for 1 min and carefully remove
the supernatant by pipetting and discard. Resuspend the cells in
300 μl Cell Lysis solution and pipet to mix well. Add 100 μl
Protein Precipitation Solution and vortex at high speed
for 20–30 s. Centrifuge the samples at 14,000 rpm
(~21,000  g) for 3 min to pellet the proteins. Transfer the
supernatant to a fresh microcentrifuge tube, add 1.5 μl of
4 mg/ml RNAse A to the supernatant, mix well and incubate
at 37
C for 1 h. Then add 300 μl isopropanol, mix well, and
centrifuge the samples at 14,000 rpm for 1 min to precipitate
the DNA (the pellet may not always be visible). Discard the
supernatant carefully by pouring. Wash the pellet with 300 μl
70% ethanol, and centrifuge the samples at 14,000 rpm
(~21,000  g) for 1 min. Remove the ethanol carefully and
air-dry the pellet for 5 min. Add 50–100 μl DNA Hydration
Solution and dissolve the DNA by incubating the tubes at
65
C for 1 h. Measure the DNA concentration using a Nano-
drop spectrophotometer (see Note 11). If needed, the genomic
DNA samples can be stored at 20
C overnight or longer.
4. Prepare sonicated DNA samples as follows. Reserve 1 μl of each
genomic DNA sample as a “pre-sonicated” control. Sonicate
1–5 μg of the remainder of the genomic DNA to fragments
with an average size of 600 bp using the Covaris M220 or
preferred sonicator (see Note 12). To verify fragment size
range, load 1 μl of the pre-sonicated and sonicated DNA
samples on a 1.2% FlashGel (use 2 μl Flash Gel Dye +2 μl sterile
deionized water +1 μl DNA) with 2 μl Flash Gel Ladder, and
run the gel at 200 V for 5 min. Sonication should result in a
smear of fragments ranging from 200 to 800 bp. The fragment
size range can be adjusted by changing the sonication time. If
needed, the sonicated DNA samples can be stored at 20
C
overnight or longer.
5. Prepare blunt-ended 50 -phosphorylated DNA fragments using
T4 DNA polymerase and T4 polynucleotide kinase. If using the
End-It DNA End-Repair kit, treat up to 5 μg DNA in a 50 μl
reaction prepared as follows:
1–34 μl DNA (see Note 13)
5 μl 10 End-Repair buffer
 Genome Instability in S. Cerevisiae 53

5 μl 2.5 mM dNTP mix

5 μl 10 mM ATP
1 μl End-Repair enzyme mix
Sterile water (for a final volume of 50 μl)

Mix the reaction well and incubate the reactions at room tempera-
ture for 45 min. Purify the DNA using a Qiagen MinElute PCR
purification kit or equivalent. If the Qiagen MinElute PCR
purification kit is used, follow the manufacturer’s protocol
with the modification that two elutions of the columns should
be performed: the first elution should use 20 μl of elution
buffer, and the second elution should use 12 μl of elution
buffer. Combine the two elutions to obtain 32 μl of eluate.
The two elution steps are required to elute most of the end-
repaired DNA from the column. At this stage, all the samples
should have approximately equal concentrations of DNA. If
needed, the end-repaired DNA samples can be stored at
20
C overnight or longer.
6. Adenylate the 30 ends of the DNA fragments using Klenow
DNA polymerase as follows:
32 μl end-repaired DNA
5 μl 10 NEBuffer 2
10 μl 1 mM dATP
3 μl Klenow DNA polymerase (exo, 5 U/μl)

Mix well and incubate the reactions in a thermal cycler at 37
C for

30 min. Purify the DNA using a Qiagen MinElute PCR purifi-
cation kit, eluting in 15 μl of elution buffer. If needed, the A-
tailed DNA samples can be stored at 20
C overnight or
longer.
7. Ligate Illumina adapters to the 50 and 30 ends of the A-tailed
DNA fragments using Quick DNA ligase. Each adapter mix
contains both a universal adapter and an indexed adaptor that
tags the fragments so that multiplexed sequences can be
assigned to specific samples after sequencing. Ensure that a
different indexed adapter mix is used for each sample within
one multiplexed library. Set up ligation reactions as follows:
15 μl A-tailed DNA
20 μl 2 DNA ligase buffer
2 μl 1:10 diluted Illumina indexed adapter mix
3 μl Quick DNA ligase

Mix well and incubate the reactions at 20
C for 15 min in a thermal

cycler. Purify the DNA using Qiagen MinElute PCR
54 Anjana Srivatsan et al.

purification kit, eluting in 17 μl elution buffer. If needed, the

samples can be stored at 20
C overnight or longer.
8. Purify 600–800 bp ligated samples by gel electrophoresis and
gel extraction as follows. Prepare 200 ml of 2% high-resolution
agarose in 1 TAE (diluted from 50 TAE in sterile deionized
water), add 20 μl Sybr Safe, and cast 2 gels (100 ml each) using
15-well combs. Load 10 μl HyperLadder 100 bp in the first
lane of the gel. Add 5 μl Orange gel loading dye to each DNA
sample from the previous step and load the first six samples on
the first gel, leaving a gap of one lane after each sample to
prevent cross-contamination. Load 10 μl HyperLadder
100 bp on the final lane of the gel. Flanking the samples with
markers facilitates excision of the desired fragment sizes. Simi-
larly, load the next set of six samples and 100 bp ladder on the
second gel. Run the gels at 120 V for 60 min (6–10 V/cm).
View the fragments using a Dark Reader transilluminator. At
this stage, it is normal to observe very weak or no DNA signal
using a transilluminator. Using a fresh razor blade, excise a
piece of the gel containing fragments ranging from 600 to
800 bp from each sample lane. Use a fresh blade for each
sample or clean the blade thoroughly by washing in water and
ethanol between samples to avoid cross contamination. Trans-
fer the gel piece to a microcentrifuge tube using a forceps,
cleaning the forceps thoroughly between samples. Purify the
DNA from the excised gel pieces using the Qiagen MinElute
gel extraction kit, eluting in 11.5 μl elution buffer. The purified
samples can be stored at 20
C overnight or longer.
9. Enrich adapter-ligated DNA fragments by PCR amplification
(see Note 14) using the following reaction mix:
11.5 μl DNA from the previous step
12.5 μl 2 KAPA master mix
1 μl 25 μM PCR primer mix

The sequences of the primers in the PCR primer mix are 50 -AAT-
GATACGGCGACCACCGAGATCTACAC-30 and 50 -CAAG-
CAGAAGACGGCATACGAGAT-30 , and are designed to
hybridize to the proprietary Illumina adapter sequences.
Use the following PCR conditions:
Denaturation: 98
C for 45 s
18 cycles (see Note 15): denaturation (98
C for 15 s), anneal-
ing (65
C for 30 s), and extension (72
C for 30 s)
Final extension: 72
C for 1 min
Hold: 4
C as long as desired
 Genome Instability in S. Cerevisiae 55

10. Purify the final libraries by running the samples on 2% agarose

gels as described in Subheading 3.2, step 8. Excise
600–800 bp fragments as described above, and purify the
fragments from the gel pieces using the Qiagen MinElute
gel extraction kit, eluting the DNA in 15 μl elution buffer.
The samples can be stored at 20
C overnight or longer.
11. Measure the DNA concentrations using an assay that is highly
specific to double-stranded DNA, such as the Qubit dsDNA
HS assay kit, which uses a double-stranded DNA-specific
fluorescent dye. Standard spectrophotometric measurements
based on the absorbance at 260 nm are often influenced by
the presence of RNA and other contaminants in the sample
and hence are not suitable for obtaining accurate DNA con-
centrations for NGS libraries. To prepare samples for the
Qubit fluorometer, add 2 μl of each sample to 198 μl of
Qubit working solution (prepared by diluting the Qubit
dsDNA HS reagent 1:200 in the Qubit dsDNA HS buffer)
in thin-walled, clear 0.5 ml microcentrifuge tubes. To prepare
the standards, add 190 μl of Qubit working solution to 10 μl
of each Qubit standard and mix well. Calibrate the Qubit
fluorometer before measuring the sample concentrations.
Calculate the molar concentration (nM) of each library
based on the Qubit reading (ng/μl) assuming that the average
fragment length is 600 bp, and hence the average molecular
weight is 600 bp  660 g/mol/bp ¼ 396,000 g/mol. Dilute
all the samples to the same concentration (5 or 10 nM but not
less than 2 nM). Pool 2 μl of each sample for multiplexing and
mix well. Use 5 μl of the pooled libraries for paired-end
sequencing using the Illumina HiSeq 2500.
12. Sequencing of NGS libraries is now a regularly available ser-
vice at both commercial companies and sequencing facilities
at universities and research institutes. The analyses performed
and details such as the file formats of the data returned are
very facility-dependent. This protocol assumes that the data
returned will be unmapped reads, separated by sample/index,
in either a FASTQ or BAM format (which can be converted to
a FASTQ format by samtools [19]).
13. Analysis of the sequencing data first requires mapping the
sequencing reads to the reference sequence of the S. cerevisiae
genome (http://www.yeastgenome.org) using programs such
as Bowtie2 [20]. When mapping paired-end data, many read-
mapping programs filter out reads that are not concordant
with the reference genome, which eliminates reads that span
novel junctions present in GCRs. To preserve these reads,
provide the paired-end data as two sets of “unpaired” data
to the mapping program (for example using the “-U” flag to
Bowtie2) and add back the read identifier using programs
56 Anjana Srivatsan et al.

such as graft in the Pyrus suite ([16], http://www.

sourceforge.net/p/pyrus-seq). The Pyrus suite of programs
can then be used to sort the mapped reads and perform copy
number analysis, sequence variant identification, and identifi-
cation and sequencing of genome rearrangements by identifi-
cation of anomalous paired-end alignments. Many other
programs for identifying structural variants that use a variety
of strategies have been developed and can also be used to
perform these types of analyses [21]. The combination of
copy number analysis and identification of rearrangements
targeting single copy sequences in the genome can provide
great insight into the structure of individual GCRs (Fig. 3),
particularly if the junction sequence can be identified using
programs such as comice in the Pyrus suite [16]. Many GCRs,
however, are mediated by non-allelic homologous recombi-
nation between repetitive elements (such as Ty elements) that
are larger than 600 bp. These rearrangements cannot be
spanned by the NGS libraries described here and will not be
identifiable by anomalous paired-end alignment analysis. In
these cases, copy number analysis can provide hints as to the
structures of these rearrangements, but proof of these struc-
tures requires the use of complementary techniques such as
PCR and PFGE [7, 9].

4 Notes

1. The method for generating NGS libraries was modified from

the Illumina library preparation guide (Illumina). It should be
noted that there are continual improvements in the methods
and reagents for NGS library generation that should be eval-
uated from time to time.
2. Do not allow the autoclaved agar mixture to cool too much
because it will rapidly cool once the dropout mix is added and
can solidify before it is poured.
3. Strains containing three widely used haploid strain-based assays
include RDKY3615 (classical GCR assay; MATa ura3-52
leu2Δ1 trp1Δ63 his3Δ200 lys2ΔBgl hom3-10 ade2Δ1 ade8
hxt13::URA3), RDKY6677 (uGCR assay; MATa ura3-52
leu2Δ1 trp1Δ63 his3Δ200 lys2ΔBgl hom3-10 ade2Δ1 ade8
can1::hisG yel068c::CAN1-URA3 iYEL072W::hph) and
RDKY6678 (dGCR assay; MATa ura3-52 leu2Δ1 trp1Δ63
his3Δ200 lys2ΔBgl hom3-10 ade2Δ1 ade8 can1::hisG yel072w::
CAN1-URA3 iYEL072W::hph) [2, 3]. Mutations of interest
can be introduced into these strains by methods such as PCR-
based gene disruption using the following selectable markers:
HIS3, TRP1, G418-resistance (e.g., kanMX4), and
 Genome Instability in S. Cerevisiae 57

A. NGS copy number analysis B. NGS sequenced rearrangement

parental strain
Read depth

1n

0 50 100 150 200 250

deletion + de novo telomere addition

Read depth

1n

0 50 100 150 200 250

inverted duplication
Read depth

2n

1n

0 50 100 150 200 250

Chromosome V Coordinate (kbp)

Fig. 3 Examples of GCR structures deciphered through analysis of NGS data. (a) Copy number analysis for the
left arm of chromosome V is plotted for a parental strain (top), a de novo telomere addition GCR (middle), and a
GCR involving an inverted duplication (bottom). Note that both rearranged chromosomes have deleted the
region containing the CAN1 URA3 cassette (around coordinate 25,700 in this GCR assay strain). The inverted
duplication has additionally duplicated the region from around coordinate 44,700 to the Ty-related sequence
YELWdelta6 (around coordinate 138,300). (b) Sequence of the novel junctions involved in the observed GCR. A
de novo telomere (sequence prior to the colon) healed the terminal deletion at chromosome V 34,899 for the
strain displayed in the middle of the panel. For the strain displayed at the bottom of the panel, non-concordant
read pairs in the NGS data revealed the presence of an inverted duplication, which was also confirmed by the
increase in copy number (panel a) and through the sequence of the reads that are associated with the novel
junction and hence did not map to the reference sequence. For this inverted duplication, there is a six-base
identity (sequence between the colons) at the junction between the source and target of the inversion. The
inversion was not a perfect hairpin, however, as ~4000 bp separate the two fused sequences

nourseothricin-resistance (e.g., natNT2). Alternatively, the

classical GCR assay can be introduced into any CAN1 ura3
strain of interest by replacing HXT13 with URA3 (hxt13::
URA3) [2]. The uGCR or dGCR assay can be introduced
into any can1 ura3 strain of interest by inserting the desired
GCR assay cassette (yel068c::CAN1-URA3 for uGCR and
yel072w::CAN1-URA3 for dGCR), which can be amplified
from the plasmids pRDK1379 and pRDK1378, respectively
[3], at the appropriate chromosomal site. To facilitate system-
atic GCR studies using the BY4741 MATa yeast deletion
58 Anjana Srivatsan et al.

collection, MATalpha query strains that contain the appropri-

ate markers as well as the uGCR or dGCR assay have been
constructed [22]: RDKY8625 (uGCR; MATalpha hom3-10
ura3Δ0 leu2Δ0 trp1Δ63 his3Δ200 lyp1::TRP1 cyh2-Q38K
iYFR016::PMFA1-LEU2 can1::PLEU2-NAT yel068c::CAN1-
URA3) and RDKY7635 (dGCR; MATalpha hom3-10 ura3Δ0
leu2Δ0 trp1Δ63 his3Δ200 lyp1::TRP1 cyh2-Q38K iYFR016::
PMFA1-LEU2 can1::PLEU2-NAT yel072w::CAN1-URA3).
Additional mutations can be introduced into these query
strains by PCR-based gene disruption using the HIS3 and
hygromycin-resistance (e.g., hphNT1) markers.
4. Sterilize a scalpel/surgical blade by dipping in ethanol and
flaming. Before excising colonies, ensure that the blade is suffi-
ciently cooled by pressing into a region of agar on the edge of
the plate, away from any colonies.
5. The culture volume depends on the assay used to monitor the
formation of GCRs. The wild-type rates for the classical,
uGCR, and dGCR assays are 3.5  1010, 2.27  109, and
1.97  108, respectively [2, 3]. For strains with estimated
GCR rates that are similar to or greater than wild-type, use
50 ml, 10 ml, and 4 ml cultures, respectively, for these three
assays. For mutant strains that have estimated GCR rates that
are lower than wild-type, larger culture volumes can be used if
very few or no colonies are observed in patch tests (see Note 7)
on Can 5-FOA plates. To allow sufficient aeration, the cells
must be cultured in tubes or flasks with a capacity of at least
four to five times the desired culture volume.
6. For strains with a severe growth defect, use longer incubation
times and/or larger culture volumes.
7. The volume of culture plated depends on the estimated GCR
rate, which is influenced by the sequences within the break-
point region (i.e., the GCR assay used) and the mutations
present in the strain of interest. Patch tests can be performed
to make a preliminary assessment of the effect of any mutation
of interest on the GCR rate before performing fluctuation
tests. The dGCR assay is most amenable to this type of test
because the rate of accumulating GCRs in the dGCR assay is
sufficiently high. Patch tests can be performed as follows. The
strains of interest (including a wild-type control) are streaked
on YPD plates and incubated for 2–3 days at 30
C. Using
sterile toothpicks, three independent colonies of each strain are
spread into individual ~1 cm2 squares on YPD plates (including
the wild-type strain as a control). The plates are incubated for
2–3 days at 30
C, and the patches are then replica-plated onto
GCR plates using sterile velvets. The number of papillae in each
patch is counted after 3–5 days of growth at 30
C. The wild-
 Genome Instability in S. Cerevisiae 59

type dGCR strain generally yields 1–5 papillae per patch, and
different mutations can cause either increased or decreased
papillation. The number of papillae can be used to make an
initial estimate of the volume of saturated culture to be plated
on GCR plates for fluctuation tests as follows: 1–5 papillae,
0.75–1 ml; 6–15 papillae, 0.3–0.4 ml; ~16–150 papillae,
0.1–0.25 ml; papillae that are too many or too close together
to count, 0.1–0.2 ml of a 1:10 dilution of the saturated culture.
If plating these volumes yields too few or too many colonies on
the Can 5-FOA plates, then increase or decrease the volume
plated accordingly. It is important to note that the patch test is
significantly influenced by the growth rate of the strain(s) of
interest, and therefore patch tests are not reliable predictors of
GCR rates for strains with growth defects. Additionally, certain
mutations can have assay-specific effects, for instance by
increasing the GCR rate in the uGCR but not the dGCR
assay (as observed when genes like EXO1, RAD10, or RAD6
are deleted [3]). In such cases, performing patch tests using
only the dGCR assay can lead to the potentially inaccurate
conclusion that the mutation does not impact the GCR rate.
Therefore, while patch tests are extremely useful in rapidly
screening large numbers of mutations, it is crucial to perform
fluctuation tests especially when the mutation of interest causes
a growth defect or potentially has an assay-specific effect on the
GCR rate.
8. S. cerevisiae strains yield on average 108 cells per ml of saturated
culture in YPD; therefore, 0.1 ml of the 105 dilution plated on
YPD will yield approximately 100 colonies.
9. If all Can 5-FOA plates contain very few or no colonies, plate
larger culture volumes.
10. The genomic DNA extraction method described here uses the
Gentra Puregene Yeast/Bact. kit (Qiagen), but other genomic
DNA extraction methods can be used at this step, with the
modification that RNase treatment should be performed prior
to DNA precipitation and resuspension.
11. If the DNA yield from 1.5 ml of the overnight culture is low,
the culture volume used for the DNA isolation should be
increased. For example, a 10 ml overnight culture can be split
into multiple separate 1.5 ml scale DNA isolations. When
multiple DNA isolations per culture are performed, the DNA
can be combined in the final step by resuspending the DNA
from all the tubes with the same 50 μl of DNA Hydration
Solution.
12. The Covaris microTUBE-15 and microTUBE-50 vials have a
DNA input limit of 1 μg and 5 μg, respectively; therefore, use 1
or 2 μg genomic DNA in a final volume of volume of 15–20
60 Anjana Srivatsan et al.

() 1 μl and 50 () 3 μl, respectively, depending on the DNA

concentration. Use the manufacturer settings to obtain 600 bp
fragments in each case.
13. If the DNA concentration is low, use the entire 50 μl of
sonicated DNA sample and scale up the end-repair reaction
to 125 μl with 50 μl DNA, 12.5 μl each of the 10 End-repair
buffer, 2.5 mM dNTP mix, and 10 mM ATP, 2.5 μl of End-It
enzyme mix, and sterile water to a final volume of 125 μl.
14. There are alternative PCR-free methods to enrich adapter-
ligated fragments, as described at www.illumina.com.
15. If insufficient amplification is observed, use real-time PCR to
determine the optimal number of cycles.

Acknowledgments

This work was supported by NIH grant GM26017 and the Ludwig
Institute for Cancer Research.

References
1. Chan JE, Kolodner RD (2011) A genetic and 8. Putnam CD, Pennaneach V, Kolodner RD
structural study of genome rearrangements (2004) Chromosome healing through terminal
mediated by high copy repeat Ty1 elements. deletions generated by de novo telomere addi-
PLoS Genet 7:e1002089 tions in Saccharomyces cerevisiae. Proc Natl
2. Chen C, Kolodner RD (1999) Gross chromo- Acad Sci U S A 101:13262–13267
somal rearrangements in Saccharomyces cere- 9. Narayanan V, Mieczkowski PA, Kim HM, Petes
visiae replication and recombination defective TD, Lobachev KS (2006) The pattern of gene
mutants. Nat Genet 23:81–85 amplification is determined by the chromo-
3. Putnam CD, Hayes TK, Kolodner RD (2009) somal location of hairpin-capped breaks. Cell
Specific pathways prevent duplication- 125:1283–1296
mediated genome rearrangements. Nature 10. Kanellis P, Gagliardi M, Banath JP, Szilard RK,
460:984–989 Nakada S, Galicia S et al (2007) A screen for
4. Myung K, Chen C, Kolodner RD (2001) Mul- suppressors of gross chromosomal rearrange-
tiple pathways cooperate in the suppression of ments identifies a conserved role for PLP in
genome instability in Saccharomyces cerevisiae. preventing DNA lesions. PLoS Genet 3:e134
Nature 411:1073–1076 11. Koszul R, Caburet S, Dujon B, Fischer G
5. Pennaneach V, Kolodner RD (2004) Recombi- (2004) Eucaryotic genome evolution through
nation and the Tel1 and Mec1 checkpoints the spontaneous duplication of large chromo-
differentially effect genome rearrangements somal segments. EMBO J 23:234–243
driven by telomere dysfunction in yeast. Nat 12. Umezu K, Hiraoka M, Mori M, Maki H
Genet 36:612–617 (2002) Structural analysis of aberrant chromo-
6. Chen C, Umezu K, Kolodner RD (1998) somes that occur spontaneously in diploid Sac-
Chromosomal rearrangements occur in S. cer- charomyces cerevisiae: retrotransposon Ty1
evisiae rfa1 mutator mutants due to mutagenic plays a crucial role in chromosomal rearrange-
lesions processed by double-strand-break ments. Genetics 160:97–110
repair. Mol Cell 2:9–22 13. Zhang Y, Saini N, Sheng Z, Lobachev KS
7. Pennaneach V, Kolodner RD (2009) Stabiliza- (2013) Genome-wide screen reveals replication
tion of dicentric translocations through sec- pathway for quasi-palindrome fragility depen-
ondary rearrangements mediated by multiple dent on homologous recombination. PLoS
mechanisms in S. cerevisiae. PLoS One 4: Genet 9:e1003979
e6389
 Genome Instability in S. Cerevisiae 61

14. Lemoine FJ, Degtyareva NP, Lobachev K, 18. Lea DE, Coulson CA (1949) The distribution
Petes TD (2005) Chromosomal translocations of the numbers of mutants in bacterial popula-
in yeast induced by low levels of DNA polymer- tions. J Genet 49:264–285
ase a model for chromosome fragile sites. Cell 19. Li H, Handsaker B, Wysoker A, Fennell T,
120:587–598 Ruan J, Homer N et al (2009) The Sequence
15. Chan JE, Kolodner RD (2012) Rapid analysis Alignment/Map format and SAMtools. Bioin-
of Saccharomyces cerevisiae genome rearrange- formatics 25:2078–2079
ments by multiplex ligation-dependent probe 20. Langmead B, Salzberg SL (2012) Fast gapped-
amplification. PLoS Genet 8:e1002539 read alignment with Bowtie 2. Nat Methods
16. Putnam CD, Pallis K, Hayes TK, Kolodner RD 9:357–359
(2014) DNA repair pathway selection caused 21. Guan P, Sung WK (2016) Structural variation
by defects in TEL1, SAE2, and de novo telo- detection using next-generation sequencing
mere addition generates specific chromosomal data: a comparative technical review. Methods
rearrangement signatures. PLoS Genet 10: 102:36–49
e1004277 22. Putnam CD, Srivatsan A, Nene RV, Martinez
17. Serero A, Jubin C, Loeillet S, Legoix-Ne P, SL, Clotfelter SP, Bell SN et al (2016) A
Nicolas AG (2014) Mutational landscape of genetic network that suppresses genome rear-
yeast mutator strains. Proc Natl Acad Sci U S A rangements in Saccharomyces cerevisiae and
111:1897–1902 contains defects in cancers. Nat Commun
7:11256
 Chapter 6

High-Resolution Mapping of Modified DNA Nucleobases

Using Excision Repair Enzymes
Monica Ransom, D. Suzi Bryan, and Jay R. Hesselberth

Abstract
Modification of DNA nucleobases has a profound effect on genome function. We developed a method that
maps the positions of the modified DNA nucleobases throughout genomic DNA. This method couples
in vitro nucleobase excision with massively parallel DNA sequencing to determine the location of modified
DNA nucleobases with single base precision. This protocol was used to map uracil incorporation and UV
photodimers in DNA, and a modification of the protocol has been used to map sparse modification events
in cells. The Excision-seq protocol is broadly applicable to a variety of base modifications for which an
excision enzyme is available.

Key words Uracil incorporation, Base excision repair, UV photodimers, Circular ligation, High-
throughput sequencing

1 Introduction

Methods for the detection of modified DNA bases include chro-

matographic techniques, PCR-based assays, comet assays, mass
spectrometry, electrochemistry, radioactive labeling, and immuno-
chemical methods [1]. These methods are powerful for studying
single or small groups of loci but less effective for unknown targets
or on a larger scale. Direct sequencing methods such as SMRT
sequencing that rely on polymerase dynamics as it crosses a modifi-
cation are powerful tools for detecting single types of DNA modi-
fication but do not scale well to large genomes [2].
We developed a method that uses base excision repair enzymes
to create nicks or gaps in DNA, which are then recovered in a
sequencing library for their identification in genomic DNA. Base
excision repair is a conserved repair system that recognizes and
removes DNA lesions to maintain genomic stability [3]. Repair of
these lesions is initiated by the recognition and cleavage of the
modification by a modification-specific DNA glycosylase yielding
in an apurinic site [3]. Apurinic sites are resolved to single-stranded

Marco Muzi-Falconi and Grant W. Brown (eds.), Genome Instability: Methods and Protocols, Methods in Molecular Biology,
vol. 1672, DOI 10.1007/978-1-4939-7306-4_6, © Springer Science+Business Media LLC 2018

63
64 Monica Ransom et al.

breaks and are repaired by additional systems involving DNA poly-

merase and ligase [4].
We developed a general method to map modified nucleobases
in genomic DNA that couples base excision with next-generation
DNA sequencing. We have successfully implemented several varia-
tions of the protocol to map different nucleobases or to produce
different signals at sites of nucleobase modification 1. In pre-
digestion Excision-seq, samples containing modified nucleobases
are first digested with a DNA glycosylase, yielding breaks at the sites
of modification (Fig. 1a). Fragments released from DNA digestion
are converted into Illumina libraries with standard protocols such
that the 50 -end of each sequencing read corresponds to the site of
DNA modification. This protocol is applicable when high levels of
DNA modification (~1% or more) are expected. In addition, we
developed a complementary approach termed “post-digestion” in
which libraries are first created from sheared genomic DNA fol-
lowed by adaptor ligation (Fig. 1b). Prior to PCR, the double-
stranded DNA fragments are digested with a base excision repair
enzyme, destroying fragments containing modified bases and ren-
dering them incompetent for PCR (Fig. 2). Analysis of reads from
these libraries creates a signal opposite to that of “pre-digestion”
libraries. Again, this protocol is applicable to DNA with high
levels of DNA modification. We used these protocols to map
uracil throughout the genomes of E. coli and S. cerevisiae cells in
which dUTP biosynthesis was genetically or pharmacologically
perturbed [5].
To map DNA photodimers we modified the pre-digestion
protocol to include a photolyase repair step following glycosylase
cleavage to remove the photo dimer, which inhibits subsequent
Illumina library steps. We used photolyase enzymes specific for
cyclobutane pyrimidine dimers or 6-4 photodimers [6] and used
this repaired DNA to generate libraries containing sites of UV
damage (Fig. 1c). Again, these libraries require high modification
levels because the distribution of modification determines the frag-
ment size upon cleavage, and Illumina sequencing requires frag-
ments <800 bp in size.
To address the limitations of the “pre-digestion” and “post-
digestion” Excision-seq methods and create libraries using lower
levels of genomic damage, we developed an alternative method
involving DNA circularization. To generate these libraries, geno-
mic DNA was sheared, polished, a-tailed and ligated to adapters
that are appropriate for circularization. DNA was then cleaved with
a DNA glycosylase generating a free 30 -OH terminus. These mole-
cules are made single stranded and circularized with circ-ligase.
Linear DNA is then degraded with T5 exonuclease and PCR is
performed on intact circles (Fig. 2a). We used this protocol to
generate libraries from low dose UV damaged cells (Fig. 2b). We
were able to see libraries with less damage than with the basic
 Genomewide Mapping of Modified Nucleobases 65

A B C
5´ U A A AU 3´
3´ A U T UA 5´
5´ T T 3´
UDG / Endo IV 5´ U A A AU 3´ 3´ T T 5´
U U 3´ A T T TA 5´
A A A UVDE
Polishing &
T adaptor ligation
T T A A
U U 5´ U A A AU 3´ A A T T
Chew-back 3´ A T T TA 5´
Photolyase
5´ A 3´
UDG / Endo IV
3´ T 5´ 5´ TT AA 3´
Fill-in Polishing & 3´ A T T TA 5´ 3´ AA TT 5´
adaptor ligation Adaptor ligation
A PCR
T TT
T AA
AA TT

Fig. 1 Excision-seq methods for mapping modified nucleobases in genomic DNA. (a) In “pre-digestion” UDG
Excision-seq, DNA is sheared with UDG glycosylase and Endo IV leaving double-strand breaks. These
fragments are then treated with the standard Illumina polishing and ligation reactions followed by PCR to
generate Illumina libraries. Sequences derived from these libraries identify the position of base modifications.
(b) In “post-digestion” UDG Excision-seq DNA is sheared mechanically and made into Illumina libraries with
the standard polishing and ligation protocols. Libraries containing uracil are then cleaved with UDG and Endo
IV and the remaining non-uracil containing library is PCR amplified. (c) In “pre-digestion” UV Excision-seq, UV
photodimers are cleaved by the UVDE enzyme followed by the removal of the 50 dipyrimidine base with either
CPD or 6-4 photolyases. These single pyrimidine ends are then compatible with standard Illumina polishing,
ligation, and PCR

A B UVDE
Circligase
- + + + +
+ - + + +
C
T5 exo + + - + +
5´ T T 3´
3´ 5´ Template + + + + - 4 1000 J/m2
20 J/m2
Fold enrichment

5´-P T T 3´-NH2 3
3´-NH2 5´-P
UVDE
Nick
5´-P T T 3´-NH2 2
3´-NH2 5´-P
Denature
Gel 1
5´-P
purified
Circularize
Degrade linear DNA 0
with T5 exonuclease
C
A
C

A
T
TC

TA
G

G
A
C
TT

T
TG
G
C

G
C
A
A

G
G
A

PCR

1 2 3 4 5 Inferred dinuleotide downstream of nick

Illumina sequencing M

Fig. 2 Excision-seq methods for mapping individual modified nucleobases in genomic DNA. (a) In circular UV
Excision-seq, DNA is mechanically sheared and ligated to modified circular adapters with standard polishing
and ligation reactions. Double-stranded DNAs are then treated with UVDE to generate nicks upstream of the
dipyrimidine, yielding a free 30 OH. The DNA is denatured with heat and the strand with a free 50 P and 30 OH is
circularized with CircLigase II (Epicentre). The remaining linear DNA is destroyed with T5 exonuclease and
circle PCR generates libraries with standard Illumina sequences. (b) Representative circ-ligase preparation
with controls including no UVDE (lane 1), no CircLigase II (lane 2), no T5 exonuclease (lane 3), and no template
(lane 5). Doubled-stranded from lane 4 was gel purified and submitted for Illumina sequencing. (c) Analysis of
low UV dosage sequencing libraries treated with 1000 J/m2 (black) or 20 J/m2 (gray) shows bias of
dipyrimidine ends relative to genomic dinucleotide content
66 Monica Ransom et al.

pre-digestion protocol but lost the specificity of the photodimer

type as shown by the prevalence of pyrimidine dimers directly
upstream of our sequencing libraries (Fig. 2c).

2 Materials

2.1 General 1. Micro centrifuge.

Equipment 2. Size Select E Gel apparatus (Invitrogen).
3. E.Z.N.A. Cycle-Pure DNA kit (Omega Bio-tek).
4. E.Z.N.A. MicroElute Cycle-Pure Kit (Omega Bio-tek).
5. Qubit Fluorimeter.
6. Diagenode bioruptor (or similar DNA shearing apparatus).

2.2 Enzymes 1. End Repair kit(NEB).

2. Klenow Exo-.
3. Rapid T4 DNA ligase.
4. Zymolase 20T.
5. Uracil DNA glycosylase.
6. Endonuclease IV.
7. Proteinase K 20 mg/ml.
8. RNAse A 10 mg/ml.
9. CircLigase (Epicenter).
10. Platinum Taq.
11. Phusion DNA polymerase (or similar high fidelity polymerase).

2.3 Primers See Tables 1 and 2.

2.4 Cell Growth 1. LB for bacterial cultures: 10 g/l Bacto-tryptone, 5 g/l yeast
extract, 10 g/l NaCL.
2. YPD for yeast cultures: Bacto peptone 20 g/l, yeast extract
10 g/l, glucose 2% (w/v).

2.5 Isolation of DNA 1. Buffer Z: 1 M sorbitol, 50 mM Tris pH 8.0, 10 mM BME.

and Shearing 2. Lysis buffer: 1 M sorbitol, 1% SDS, 20 mM EDTA, 10 μg
Proteinase K.
3. PBS-S: PBS + 1 M sorbitol.
4. Stop solution: 5% SDS, 100 mM EDTA.
5. Shear buffer: 10% glycerol, TE.
6. Spectrophotometer.
7. Stratalinker with UVC or UVB bulbs (or similar UV
crosslinker).
 Genomewide Mapping of Modified Nucleobases 67

Table 1
Oligonucleotide sequences

JH805 ACA CTC TTT CCC TAC ACG ACG CTC TTC CGA TCT GAT C*T
BC1
JH806 /5Phos/GAT CAG ATC GGA AGA GCG GTT CAG CAG GAA TGC CGA G
BC1
JH807 ACA CTC TTT CCC TAC ACG ACG CTC TTC CGA TCT ACT G*T
BC2
JH808 /5Phos/CAG TAG ATC GGA AGA GCG GTT CAG CAG GAA TGC CGA G
BC2
JH813 ACA CTC TTT CCC TAC ACG ACG CTC TTC CGA TCT GTC A*T
BC5
JH814 /5Phos/TGA CAG ATC GGA AGA GCG GTT CAG CAG GAA TGC CGA G
BC5
JH815 ACA CTC TTT CCC TAC ACG ACG CTC TTC CGA TCT ATC G*T
BC6
JH816 /5Phos/CGA TAG ATC GGA AGA GCG GTT CAG CAG GAA TGC CGA G
BC6
JH1139 /5Phos/GAT CGG AAG AGC ACA CGT CT
Universal
JH0804 ACA CTC TTT CCC TAC ACG ACG CTC TTC CGA TC*T
Universal
JH1323 /5Phos/NNN NNN NNA GAT CGG AAG AGC GTC GTG TAG GGA AAG AGG GAG
Circle TTC AGA CGT GTG CTC TTC CGA TCT AGC CAG CGC AGA CCG TGA GGT
JH1324 /5Phos/CCT CAC GGT CTG CGC TGG CT/3AmMO/
Circle
Adapter sequences used for this study. A “*” denotes a phosphothioate bond. A “3AmMO” is a 30 amino modifier group
from IDT used to block this end

8. Radiometer (UVX or similar).

9. 365 nM black light.
10. E. coli lysis buffer: 0.006% SDS, 12 μg/ml Proteinase K, 10 μg/
ml RNAse A.

2.6 Photolyase 1. Vibrio CPD buffer: 50 mM Tris–HCl pH 7.0, 50 mM NaCl,

Cleavage 5 mM DTT, 1 mM EDTA pH 8.0, 20% glycerol [6].
2. Xenopus 6-4 buffer: 50 mM Tris–HCl pH 7.5, 100 mM NaCl,
10 mM DTT, 10% (v/v) glycerol [7].
3. E. coli CPD buffer: 50 mM Tris–HCl pH 7.0, 50 mM NaCl,
5 mM DTT, 1 mM EDTA pH 8.0, 20% glycerol [8].
4. A. thaliana 6-4 photolyase buffer: 50 mM Tris–HCl pH 7.5,
100 mM NaCl, 10 mM DTT, 10% (v/v) glycerol [9].
68 Monica Ransom et al.

Table 2
Primer sequences used for this study

JH801 CAA GCA GAA GAC GGC ATA CGA GCT CTT CCG ATC T
Universal
JH802 AAT GAT ACG GCG ACC ACC GAG ATC TAC ACT CTT TCC CTA CAC GAC GCT CTT
Universal CCG ATC T
JH1141 CAA GCA GAA GAC GGC ATA CGA GAT CGG TAT CAC GGT GAC TGG AGT TCA
GAC GTG TGC TCT TCC GAT C
JH1147 CAA GCA GAA GAC GGC ATA CGA GAT CGG TCG ATG TGT GAC TGG AGT TCA
GAC GTG TGC TCT TCC GAT C
JH1151 CAA GCA GAA GAC GGC ATA CGA GAT CGG TGC CAA TGT GAC TGG AGT TCA
GAC GTG TGC TCT TCC GAT C
Bold bases are barcodes for demultiplexing

2.7 DNA Purification 1. Ampure XP beads.

2. Magnetic separator.
3. EBT: EB + 0.05% Tween.
4. DNA cycle pure kits from zymogen.
5. Invitrogen size select gel system.
6. Agilent tape station (optional) with D1000 tapes.
7. Qubit Fluorimeter and HS DNA kit.
8. Illumina HiSeq or MiSeq.

3 Methods

3.1 Uracil Pre- 1. CJ1036 ung-1, dut-1 E. coli were grown to log phase at 37  C
digestion Excision-seq (~2.5  1010 cells) in Luria Broth (LB) and collected by
for E. coli centrifugation.
2. DNA was harvested following lysis in E. coli lysis buffer,
extracted with phenol chloroform, and DNA precipitated.
3. High molecular weight DNA was digested with 5 units of
UDG (1 μl) and 10 (1 μl) units of T4 Endonuclease IV for
2 h at 37  C (see Note 1).
4. DNA cleavage was confirmed on a 1% agarose TBE gel looking
for appropriate cleavage fragments from 100 to 800 bp. Uracil
should make up 1–10% of the genome so the majority of DNA
following cleavage should be low molecular weight.
5. Fragments were polished using T4 DNA polymerase
(150 units), T4 PNK (50 units) and 1 mM dNTP for 30 min
at 20  C (see Note 2).
 Genomewide Mapping of Modified Nucleobases 69

6. Fragments were cleaned up with Cycle-Pure kits and eluted in

30 μl of elution buffer.
7. Fragments were A-tailed with Klenow Exo- fragment
(15 units) and 10 mM dATP for 30 min at 37  C.
8. Fragments were cleaned up with MicroElute Cycle-Pure kits
and eluted in 10 μl of elution buffer.
9. Fragments were ligated with pre-annealed universal adapters
for 15 min at 16  C with 3000 units of T4 Rapid DNA ligase.
10. Libraries were size selected on 2% size select E gels and frag-
ments between 200 and 400 bp were collected in ddH2O.
11. Libraries were PCR amplified using Platinum Taq polymerase
(see Note 3) and primers containing Illumina 6 bp barcode
sequences.
12. Ampure XP beads were used to purify libraries at 1.8 concen-
tration and eluted in 30 μl EBT.
13. Libraries were quantified on a qubit fluorimeter (see Note 4)
and submitted for sequencing at 10 nM.

3.2 Uracil Pre- 1. 50 ml of YJH516 dut1–1 ung1Δ or other yeast strains

digestion Excision-seq (~6  108 cells) were grown to mid-log phase at 30  C in
for Yeast YPD and collected by centrifugation.
2. Cells were resuspended in 5 ml of Buffer Z and spheroplasted
with 0.35 mg of zymolyase (20T) for 30 min at 30  C.
3. Spheroplasts were centrifuged and resuspended in 400 μl of
lysis buffer and incubated O/N at 55  C.
4. High molecular weight DNA was isolated by phenol/chloro-
form extraction followed by DNA precipitation (see Note 5).
5. 5–10 μg of high molecular weight DNA was digested with
5 units of UDG(1 μl) and 10 units of T4 Endonuclease IV
(1 μl) for 2 h at 37  C.
6. DNA cleavage was confirmed on a 1% agarose TBE gel looking
for appropriate cleavage fragments from 100 to 800 bp.
7. Fragments were polished using T4 DNA polymerase
(150 units), T4 PNK (50 units), and 1 mM dNTP for 30 min
at 20  C (5 μl of end repair kit).
8. Fragments were cleaned up with Cycle-Pure kits and eluted in
30 μl of elution buffer.
9. Fragments were A-tailed with Klenow Exo- fragment
(15 units/3 μl) and 10 mM dATP for 30 min at 37  C.
10. Fragments were cleaned up with MicroElute Cycle-Pure kits
and eluted in 10 μl of elution buffer.
70 Monica Ransom et al.

11. Fragments were ligated with pre-annealed universal adapters

for 15 min at 16  C with 3000 units of T4 Rapid DNA ligase
(5 μl).
12. Libraries were size selected on 2% size select E gels and frag-
ments between 200 and 400 bp were collected in ddH2O (see
Note 6).
13. Libraries were PCR amplified using Platinum Taq polymerase
and primers containing Illumina 6 bp barcode sequences.
14. Ampure XP beads were used to purify libraries at 1.8 concen-
tration and eluted in 30 μl EBT.
15. Libraries were quantified on a qubit fluorimeter and submitted
for sequencing at 10 nM.

3.3 Uracil Post- 1. 50 ml of YJH516 dut1–1 ung1Δ or other yeast strains

digestion Excision-seq (~6  108 cells) were grown to mid-log phase at 30  C in
YPD and collected by centrifugation.
2. Cells were resuspended in 5 ml of Buffer Z and spheroplasted
with 0.35 mg of zymolyase(20T) for 30 min at 30  C.
3. Spheroplasts were centrifuged and resuspended in 400 μl of
lysis buffer and incubated O/N at 55  C.
4. High molecular weight DNA was isolated by phenol/chloro-
form extraction followed by DNA precipitation.
5. DNA was mechanically sheared in shearing buffer using a
Diagenode bioruptor for 12 min (30 s on/off at maximum
intensity) until DNA was mainly sheared to 200–600 bp (see
Note 7).
6. DNA shearing was confirmed with agarose gel electrophoresis.
7. Fragments were polished using T4 DNA polymerase
(150 units), T4 PNK(50 units), and 1 mM dNTP (5 μl of
NEB end repair) for 30 min at 20  C (see Note 2).
8. Fragments were cleaned up with Cycle-Pure kits and eluted in
30 μl of elution buffer.
9. Fragments were A-tailed with Klenow Exo- fragment
(15 units/3 μl) and 10 mM dATP for 30 min at 37  C.
10. Fragments were cleaned up with MicroElute Cycle-Pure kits
and eluted in 10 μl of elution buffer.
11. Fragments were ligated with pre-annealed universal adapters
for 15 min at 16  C with 3000 units of T4 Rapid DNA ligase
(5 μl).
12. Libraries were size selected on 2% size select E gels and frag-
ments between 200 and 400 bp were collected in ddH2O.
13. Libraries were PCR amplified using Platinum Taq polymerase
and primers containing Illumina 6 bp barcode sequences.
 Genomewide Mapping of Modified Nucleobases 71

14. PCR libraries containing Uracil were sheared with 5 units of

UDG (1 μl) and 10 units of T4 Endonuclease IV (1 μl) for 2 h
at 37  C.
15. Ampure XP beads were used to purify libraries at 1.8 concen-
tration and eluted in 30 μl EBT.
16. Libraries were quantified on a qubit fluorimeter and submitted
for sequencing at 10 nM.

3.4 UV Pre-digestion 1. 50 μl of YJH302 wt yeast were grown to an OD600 of ~0.4

Excision-seq (~4  108 cells) and arrested with alpha factor.
2. Cells were resuspended in 5 ml of Buffer Z in dark tubes (see
Note 8).
3. Cells were irradiated in a 100 mm Petri dish until a specified
UV dosage was obtained as measured by a dosimeter (see Note
9).
4. 0.35 mg of zymolyase (20T) was added to the cells in the dark
tubes for 30 min at 30  C.
5. Spheroplasts were centrifuged and resuspended in 400 μl of
lysis buffer and incubated O/N at 55  C.
6. High molecular weight DNA was isolated by phenol/chloro-
form extraction followed by DNA precipitation.
7. UV damaged DNA was digested with 10 μg of UVDE for 2 h
at 30  C.
8. DNA cleavage was confirmed with agarose gel electrophoresis
increased UV dosage leads to increased shearing on the gel.
9. Samples were treated with either CPD or 6-4 photolyase in the
appropriate buffer for 2 h under illumination of a 365 nm light
(~280 kJ/m2) (see Note 10).
10. Fragments were polished using T4 DNA polymerase
(150 units), T4 PNK (50 units), and 1 mM dNTP (5 μl of
NEB end repair kit) for 30 min at 20  C.
11. Fragments were cleaned up with Cycle-Pure kits and eluted in
30 μl of elution buffer.
12. Fragments were A-tailed with Klenow Exo- fragment
(15 units/3 μl) and 10 mM dATP for 30 min at 37  C.
13. Fragments were cleaned up with MicroElute Cycle-Pure kits
and eluted in 10 μl of elution buffer.
14. Fragments were ligated with pre-annealed universal adapters
for 15 min at 16  C with 3000 units of T4 Rapid DNA ligase
(5 μl).
15. Libraries were size selected on 2% size select E gels and frag-
ments between 300 and 500 bp were collected in ddH2O (see
Note 11).
72 Monica Ransom et al.

16. Libraries were PCR amplified using Phusion DNA polymerase

(see Note 12).
17. Ampure XP beads were used to purify libraries at 1.8 concen-
tration and eluted in 30 μl EBT.
18. Libraries were run on D1000 lane on a Tapestation to validate
library size and determine the percentage of library that was
adapter dimer (see Note 13).
19. Libraries were quantified on a qubit fluorimeter and submitted
for sequencing at 10 nM.

3.5 Circular Ligation 1. Anneal circle adapters (see Note 14).

of Low Dose UV 2. UV dose and extract DNA as in Subheading 3.4 (steps 1–6).
Fragments
3. Shear for 15 min on Diagenode bioruptor (30 s on/off at
max).
4. DNA cleavage was confirmed with agarose gel electrophoresis
expecting a low molecular weight shear from 100 to 1000 base
pairs.
5. Fragments were polished using T4 DNA polymerase
(150 units), T4 PNK (50 units), and 1 mM dNTP (5 μl of
NEB end repair kit) for 30 min at 20  C.
6. Fragments were cleaned up with Cycle-Pure kits and eluted in
30 μl of elution buffer.
7. Fragments were A-tailed with Klenow Exo- fragment
(15 units/3 μl) and 10 mM dATP for 30 min at 37  C.
8. Fragments were cleaned up with MicroElute Cycle-Pure kits
and eluted in 12 μl of elution buffer.
9. Fragments were ligated to pre-annealed circular adaptor
15 min at 16  C with 3000 units of T4 Rapid DNA ligase
(5 μl).
10. Ampure XP beads were used to purify libraries at 1.8 concen-
tration and eluted in 30 μl EBT.
11. DNA was digested with 10 μg of UVDE and 10 units of
EndoIV (1 μl) for 2 h at 30  C.
12. Heat denature 95  C for 15 min (see Note 15) and then keep
on ice until circularization.
13. Denatured DNA was circularized with 1 mM dATP, 50 mM
MnCl2, and 100 units of CircLigase (1 μl) at 60  C for 1 h.
14. Ampure XP beads were used to purify libraries at 1.8 concen-
tration and eluted in 30 μl EBT.
15. T5 Exonuclease was used to shear noncircular DNA with
50 units of T5 exonuclease (5 μl) in NEB buffer 4 for 2 h at
37  C.
 Genomewide Mapping of Modified Nucleobases 73

16. Ampure XP beads were used to purify libraries at 1.8 concen-

tration and eluted in 30 μl EBT.
17. Libraries were PCR amplified using Phusion polymerase (see
Note 16).
18. Ampure XP beads were used to purify libraries at 1.8 concen-
tration and eluted in 30 μl EBT.
19. Libraries were quantified on a qubit fluorimeter and submitted
for sequencing at 10 nM.

3.6 Data Analysis for 1. Sequences were analyzed by alignment to a reference genome
Uracil Modification (sacCer1) using Bowtie 2 [10] and SAMtools [11].
2. Alignments were processed to bedGraph format using BED-
Tools [12], and visualized in the UCSC Genome Browser [13].
3. Coverage at each position was normalized by the number of
reads aligned in the library (i.e., reads per million, RPM).
Using this method, the level of coverage at a specific site or
region in the genome represents the relative quantity of uracil
at that position.
4. Raw and processed sequencing data (FASTQ and bedGraph
formats) from step 2 from this study have been submitted to
the NCBI Gene Expression Omnibus (GEO; http://www.
ncbi.nlm.nih.gov/geo/) [15] under accession number
GSE51361.

3.7 Data Analysis for 1. Sequences were aligned to the S. cerevisiae genome using Bow-
UV Modification tie 2 [10] and SAMtools [11].
2. For the circular libraries the UMI tag was used to remove PCR
duplicates from the library using UMItools [14].
3. Aligned reads were separated by strand, and dinucleotide
counts for the 50 ends of the reads were determined.
4. The percentages of each dinucleotide combination were nor-
malized to the mononucleotide and dinucleotide frequencies
found in S. cerevisiae genomic DNA to account for A:T bias in
the genome.

4 Notes

1. Uracil digestion of 10 μg of yeast DNA with 5 units of UDG

and 10 unit Endo IV for 2 h at 37  C was sufficient to yield
small molecular weight DNA in strains mutant for repair
enzymes as reported in [5]. Increased enzyme or time did not
yield additional cleavage indicating that the majority of uracil
was cleaved. For other cell types or mutant backgrounds the
74 Monica Ransom et al.

cleavage conditions may need to be optimized. Small molecular

weight DNA <600 bp is needed for efficient library generation.
2. The end repair module from NEB is used for polishing DNA
for the most consistent results. The three enzymes can also be
purchased individually.
3. Pfu DNA polymerases and their derivatives cannot be used to
create Ex-seq libraries for uracil, because these polymerases are
inhibited by uracil in templates.
4. Quantification of libraries for Illumina sequencing in our hands
has shown the best results with the Qubit fluorimeter. We get
the most accurate clustering using this method. qPCR and
other methods are available for quantification and many
sequencing companies offer this service.
5. Vortexing should not be done at any stage in this protocol as it
shears high molecular weight DNA. All mixing should be done
by hand optionally with a cut pipette tip.
6. We prefer using size select gels to purify adapter ligated libraries
prior to PCR. This allows for the removal of excess adapters
which will PCR downstream, as well as allowing for the isola-
tion of a variety of fragment sizes. We use the different frag-
ments to find the optimal library in terms of background bands
and size. Also the ability to see a size selected ladder of DNA vs.
a randomly sized smear can help determine real libraries from
artifacts.
7. Biorupting is very cell specific with different mutants and cell
types requiring different amounts of biorupting time. Prior to
performing library preparation, genomic DNA as prepared
above should be sheared for a variety of time points and
about 10% should be analyzed via agarose gel. Smears between
~200 bp and ~1000 bp are optimal and shear time can be
optimized to achieve that. Hard to shear samples can also be
sheared in TE with no glycerol or in special Eppendorf tubes
that improve shearing.
8. For UV experiments all the steps following UV irradiation are
performed in dark tubes due to the photolyases present in
wild-type cells that can repair UV photodimers upon exposure
to light. Once DNA has been obtained this is no longer
required.
9. UV exposure was measured with a dosimeter within the Stra-
talinker to confirm the exposure we saw slight variation
between the Stratalinker dosage and the actual dosage given.
10. UV libraries were repaired by photolyase prior to PCR because
nothing else tried allowed PCR through the adduct on the 50
end of our molecules. UVA light was used to activate these
enzymes by placing the tubes within an inch of a black light.
 Genomewide Mapping of Modified Nucleobases 75

The UVA dosage was measured by a dosimeter as >200 kJ and

should be sufficient for enzyme activity in vitro.
11. UV libraries were isolated at a higher molecular weight from
size select gels due to less efficient shearing by UVDE. PCR
from higher molecular weight fractions contained less dimer
and more specific libraries. Also performing a laddering PCR
with a series of fractions was helpful to determine real libraries
from artifacts (which all PCR at the same size).
12. Phusion Polymerase was used to PCR UV libraries because
both high fidelity and high activity were needed to amplify
these low copy libraries.
13. Tapestation analysis helped us to confirm that these libraries
contained mainly >400 bp DNA fragments. Since the input is
low in these libraries there is a high chance for Illumina adapter
contamination to occur. We see this in our analysis so try to
submit samples with low levels of adapter contamination.
14. Circular adapters have long overhanging ends and as such we
perform a slower annealing process and make them relatively
fresh to maintain their annealing. We have also tried adding a
spacer to improve the adaptors to prevent rolling circle PCR.
15. Heat for 15 min or use 0.3 M final concentration of KOH
followed by neutralization and precipitation can be used to
denature the DNA.
16. To determine the yield of a circle library a variety of controls are
useful. A no T5 Exonuclease step gives a strong smear follow-
ing PCR as an indication of what a robust library amplification
would look like. No UV and no CircLigase negative controls
are useful for troubleshooting nonspecific amplification.

Acknowledgments

We thank A. Sancar and D. Zhang for graciously providing purified

photolyase enzymes during the development of the Excision-seq
protocol. This work was supported in part by a March of Dimes
Basil O’Connor Research Grant (J.H.), a Damon Runyon-Rachleff
Innovation award (J.H.), a Research Scholar Grant from the Amer-
ican Cancer Society (RSG-13-216-01-DMC), and a Department of
Defense Visionary Postdoc Award (to M.R., W81XWH-12-1-
0333).
76 Monica Ransom et al.

References
1. Kumari S, Rastogi RP et al (2008) DNA dam- (6-4) photoproduct by photolyase. Nature
age: detection strategies. EXCLI J 7:44–62 466:887–890
2. Clark TA, Spittle KE, Turner SW et al (2011) 10. Langmead B, Trapnell C, Pop M et al (2009)
Direct detection and sequencing of damaged Ultrafast and memory-efficient alignment of
DNA bases. Genome Integr 2:10 short DNA sequences to the human genome.
3. Baute J, Depicker A (2008) Base excision repair Genome Biol 10:R25
and its role in maintaining genome stability. 11. Li H, Handsaker B, Wysoker A et al (2009)
Crit Rev Biochem Mol Biol 43:239–276 The Sequence Alignment/Map format and
4. Hegde ML, Hazra TK, Mitra S (2008) Early SAMtools. Bioinformatics 25:2078–2079
steps in the DNA base excision/single-strand 12. Quinlan AR, Hall IM (2010) BEDTools: a
interruption repair pathway in mammalian flexible suite of utilities for comparing genomic
cells. Cell Res 18:27–47 features. Bioinformatics 26:841–842
5. Bryan DS, Ransom M, Adane B et al (2014) 13. Karolchik D, Hinrichs AS, Kent WJ (2001)
High resolution mapping of modified DNA The UCSC Genome Browser. Wiley, Hobo-
nucleobases using excision repair enzymes. ken, NJ
Genome Res 24:1534–1542 14. Smith TS, Heger A, Sudbery I (2016) UMI-
6. Sancar GB, Sancar A (2006) Purification and tools: modelling sequencing errors in Unique
characterization of DNA photolyases. Methods Molecular Identifiers to improve quantification
Enzymol 408:121–156 accuracy. bioRxiv 051755
7. Ryoji M, Katayama H, Fusamae H et al (1996) 15. Barrett T, Wilhite SE, Ledoux P, Evangelista C,
Repair of DNA damage in a mitochondrial Kim IF, Tomashevsky M, Marshall KA, Phil-
lysate of Xenopus laevis oocytes. Nucleic lippy KH, Sherman PM, Holko M, Yefanov A,
Acids Res 24:4057–4062 Lee H, Zhang N, Robertson CL, Serova N,
8. Liu Z, Tan C, Guo X et al (2011) Dynamics Davis S, Soboleva A (2012) NCBI GEO:
and mechanism of cyclobutane pyrimidine archive for functional genomics data sets–up-
dimer repair by DNA photolyase. Proc Natl date. Nucleic Acids Research 41 (D1):
Acad Sci U S A 108:14831–14836 D991–D995
9. Li J, Liu Z, Tan C et al (2010) Dynamics and
mechanism of repair of ultraviolet-induced
 Chapter 7

Integrated Microarray-based Tools for Detection

of Genomic DNA Damage and Repair Mechanisms
Patrick van Eijk, Yumin Teng, Mark R. Bennet, Katie E. Evans,
James R. Powell, Richard M. Webster, and Simon H. Reed

Abstract
The genetic information contained within the DNA molecule is highly susceptible to chemical and physical
insult, caused by both endogenous and exogenous sources that can generate in the order of thousands of
lesions a day in each of our cells (Lindahl, Nature 362(6422):709–715, 1993). DNA damages interfere
with DNA metabolic processes such as transcription and replication and can be potent inhibitors of cell
division and gene expression. To combat these regular threats to genome stability, a host of DNA repair
mechanisms have evolved. When DNA lesions are left unrepaired due to defects in the repair pathway,
mutations can arise that may alter the genetic information of the cell. DNA repair is thus fundamental to
genome stability and defects in all the major repair pathways can lead to cancer predisposition. Therefore,
the ability to accurately measure DNA damage at a genomic scale and determine the level, position, and
rates of removal by DNA repair can contribute greatly to our understanding of how DNA repair in
chromatin is organized throughout the genome. For this reason, we developed the 3D-DIP-Chip protocol
described in this chapter. Conducting such measurements has potential applications in a variety of other
fields, such as genotoxicity testing and cancer treatment using DNA damage inducing chemotherapy. Being
able to detect and measure genomic DNA damage and repair patterns in individuals following treatment
with chemotherapy could enable personalized medicine by predicting response to therapy.

Key words Genome stability, Nucleotide excision repair, Microarrays, 3D-DIP-Chip, DNA damage,
DNA repair, Chemotherapy, Genotoxicity

1 Introduction

We have developed a method that we refer to as 3D-DIP-Chip

(DNA Damage Detection by DNA Immuno-Precipitation (DIP)
on microarray chips, [2]) to detect genomic DNA damage and
measure lesion removal on a genome-wide scale. The method is
based on the standard ChIP-chip workflow but relies on the immu-
noprecipitation of DNA lesions within naked DNA. We will provide
details for the immunoprecipitation of three types of DNA damage
(UV, cisplatin, and oxaliplatin), but the protocol can be adapted to
any type of aberrant DNA base or DNA damage granted a specific

Marco Muzi-Falconi and Grant W. Brown (eds.), Genome Instability: Methods and Protocols, Methods in Molecular Biology,
vol. 1672, DOI 10.1007/978-1-4939-7306-4_7, © Springer Science+Business Media LLC 2018

77
78 Patrick van Eijk et al.

antibody against that epitope is available. Similarly, procedures for

the use of chromatin and/or DNA from both budding yeast,
Saccharomyces cerevisiae, and cultured human cell lines will be
provided.
In order to analyze multiple inter- and intra-experimental data-
sets of this new type of repair and indeed standard ChIP-chip data,
we had to develop software tools to enable this. Generally, routine
ChIP-chip experiments provide genome-wide chromatin protein
occupancy information through peak detection of the measured
signal. This traditionally allows for binary comparison of data sets
and methods are available for comparative analysis of this type of
data. However, biological systems naturally contain much more
information beyond that which can be described using a binary
concept of presence or absence. Similarly, the nature of the abun-
dant and heterogeneous distribution of DNA lesions such as UV-
induced Cyclobutane Pyrimidine Dimers (CPDs) is not suitable for
this classical approach to data analysis. This potential for more
advanced analysis we have tried to fulfil by developing the Sandcas-
tle R software package—Software for the Analysis and Normaliza-
tion of Data from ChIP-chip AssayS of Two or more Linked
Experiments [3]—that is an integral part of the microarray-based
genomic toolset described here. This software can be used for
normalization, analysis, and visual description of data in 3D-DIP-
Chip and standard ChIP-chip datasets. The Sandcastle normaliza-
tion procedure, described in more detail below, uniquely facilitates
the comparative analysis of datasets generated under different
experimental conditions, allowing the extraction of latent, biologi-
cally relevant, information.
We have used this extended workflow consisting of 3D-DIP-
Chip, ChIP-chip, and Sandcastle’s data analysis algorithms, to
study the mechanisms and the organization of DNA repair and
DNA damage-induced chromatin remodeling during Global
Genome Nucleotide Excision Repair (GG-NER) in yeast [4].
This was achieved by measuring: (1) repair using 3D-DIP-Chip
and UV-induced changes in chromatin occupancy of (2) DNA
repair proteins, (3) chromatin remodelers, and (4) chromatin mod-
ifications [4]. The introduction of specific mutations or deletion of
relevant factors can provide important mechanistic insights into the
change in genome-wide chromatin occupancy, redistribution of
proteins and protein complexes, and changes in relative DNA repair
rates in response to DNA damage. The comparison of datasets
derived from different mutant strains or data from different treated
versus untreated cells can now be compared using Sandcastle to
derive mechanisms of DNA repair on a genomic scale.
The ability to detect and measure DNA damage and genomic
repair spectra is becoming increasingly important as recent
advances in DNA sequencing technologies now permit whole
genome mutation analyses [5, 6]. These sequencing projects have
 Genomic DNA Damage and Repair Detection Methods 79

revealed the presence of mutational signatures captured in individ-

ual cancer genomes derived from different types of cancer cells [7].
Efforts to determine the relationship between the distribution of
genomic DNA repair rates with the mutational end-points detected
in cancer genomes are emerging [8]. Progress in this field would
enable the use of genomic DNA damage and repair measurements
as early indicators and predictive biomarkers, including markers
that predict predisposition to, and the progression of, disease, as
well as the ability to determine the response of individuals to drug
treatments of these diseases.

2 Materials

2.1 Yeast Cell 1. Glassware, relevant equipment, media, and other reagents,
Culture and Media where necessary, were sterilized by autoclaving at 121
C for
15 min.
2. Yeast Extract Peptone Dextrose (YPD): 2% glucose, 2% pep-
tone; 1% yeast extract.
3. Glass beads, acid-washed (425–600 μm).

2.2 Human Cell 1. Dulbecco’s Modified Eagle Medium (DMEM).

Culture and Media 2. Fetal calf serum.
3. 1% L-Glutamine.
4. 1% penicillin streptomycin solution (e.g., Sigma-Aldrich).
5. Trypsin-EDTA solution (e.g., Gibco by Life Technologies,
Invitrogen).

2.3 DNA Purification 1. Sorbitol-TE: 0.9 M Sorbitol, 0.1 M Tris–HCl, pH 8.0, 0.1 M
Buffers and Solutions EDTA.
2. Zymolyase 20 T, 20,500 U/g, Amsbio or similar (10 mg/ml in
sorbitol-TE solution).
3. β-mercaptoethanol.
4. RNAse A (10 mg/ml stock).
5. 10% SDS.
6. DNA lysis buffer: 4 M Urea, 200 mM NaCl, 100 mM
Tris–HCl pH 8.0, 10 mM EDTA, 0.5% SDS.
7. Pronase (20 mg/ml in TE buffer) (e.g., Roche).
8. Phenol/Chloroform 1:1 (v/v).
9. Prechilled 20
C 100% Ethanol.
10. 100% Isopropanol.
11. TE: 10 mM Tris–HCl pH 8.0, 1 mM EDTA.
80 Patrick van Eijk et al.

2.4 ChIP/IP Buffers, 1. Rat anti-cisplatin modified DNA, CP9/19, antibody

Solutions, (ab103261, Abcam).
and Reagents 2. Mouse Anti-CPD antibody (CAC-NM-DND-001, COSMO
Bioscience).
3. Dynabeads (Invitrogen): Sheep anti-Rat IgG, Goat anti-Mouse
and ProtG dynabeads.
4. PBS-BSA: solutions of 0.1% BSA (1 mg/ml) and 1% BSA
(10 mg/ml) in PBS.
5. PMSF (200 mM in Isopropanol).
6. FA/SDS: 50 mM HEPES–KOH pH 7.5, 150 mM NaCl,
1 mM EDTA, 1% Triton-X, 0.1% Na deoxycholate, 0.1% SDS
(1 mM PMSF added fresh before use, complete protease inhib-
itor cocktails are optional).
7. FA/SDS 0.5 M NaCl.
8. FA/SDS 1.0 M NaCl.
9. LiCl IP buffer: 10 mM Tris–HCl pH 8.0, 250 mM LiCl, 1 mM
EDTA, 0.5% NP40, 0.5% Na deoxycholate.
10. Pronase buffer: 125 mM Tris–HCl pH 7.5, 25 mM EDTA,
2.5% SDS.
11. ChIP lysis buffer: 10 mM Tris–HCl pH 8, 10 mM NaCl, 0.5%
NP-40.
12. ChIP shearing buffer/SDS lysis buffer: 50 mM Tris–HCl
pH 8, 10 mM EDTA pH 8, 1% SDS.
13. Protease Inhibitor cocktail (Roche, product code 11697498001).
14. ddH2O.

2.5 Reagents 1. PreCR DNA repair kit (New England Biolabs).

2. T4 DNA polymerase 3 U/μl (New England Biolabs).
3. T4 DNA ligase 400 U/μl (New England Biolabs).
4. 10 NEBuffer 2 (T4 DNA Pol buffer).
5. Q5 DNA Polymerase 2 U/μl (New England Biolabs).
6. dNTPs (10 mM).
7. PureLink TM Quick PCR Purification Kit (Invitrogen).
8. iTaq™ Universal SYBR® Green Supermix (Bio-Rad).
9. Hard-Shell® 96-well semi-skirted PCR plate (Bio-Rad).
10. BioPrime Total Genome Labelling System (Invitrogen).
11. GenomePlex Complete Whole Genome Amplification kit
(WGA2, Sigma-Aldrich).
12. Human Cot-1 DNA (1.0 mg/ml) (Agilent).
13. 10 Agilent Blocking Agent.
14. 2 Agilent Hybridization Buffer.
 Genomic DNA Damage and Repair Detection Methods 81

2.6 Equipment 1. Cell counting chamber (e.g., Neubauer by Hawksley).

2. Shaking incubator for basic microbiology cell culture (e.g.,
Infors HT multitron).
3. Diagenode BioRuptor (Diagenode).
4. Water bath (65
C) and circulating cooled water bath (i.e.,
NESLAB Digital Plus RTE7, ThermoScientific).
5. Refrigerated mid-speed centrifuge and rotor for large volumes
of 100–500 ml (e.g., Beckman-Coulter centrifuge with JA-10
and JA-20 rotor).
6. Refrigerated microcentrifuge (e.g., Beckman-Coulter Micro-
fuge 22R or Eppendorf Microfuge 5424R).
7. Refrigerated Bench-top Eppendorf centrifuge 5810R.
8. Bench-top Eppendorf centrifuge 5804R and A-2-W-ADP
rotor for 96-well plates.
9. Thermomixer (Eppendorf).
10. UV light source, VL-21G mineralight lamp (UV products, San
Gabriel, CA, USA).
11. NanoDrop-1000, NanoDrop Technologies, ThermoScientific.
12. Quantitative real-time PCR equipment (e.g., CFX Connect™
Real-Time PCR detection system, Bio-Rad).
13. CFX Manager™ Software v3.1 (Bio-Rad).
14. Agilent 2200 TapeStation (or Bioanalyzer) and accompanying
D1000 screentapes.

3 Methods

3.1 3D-DIP-Chip The 3D-DIP-Chip procedure follows the general steps of a tradi-
Procedure Overview tional ChIP-chip workflow (Fig. 1). However, the immunoprecipi-
tation (IP) is performed using naked DNA instead of chromatin.
Hence, the initial purification of DNA (not chromatin) and shear-
ing of the DNA are adapted for the use of naked DNA (see Sub-
heading 3 for more details). In short, DNA is purified from cells
treated with a DNA damaging agent and fragmented by sonication.
Antibodies raised against the DNA lesion of interest are used to IP
the DNA. At this stage input samples are also included for compar-
ison, as is common practice in any ChIP-chip workflow. Next, the
DNA is purified and the DNA damages are reversed so as not to
inhibit amplification. Purified input and IP DNA is amplified and
differentially labeled using two dyes and applied to the appropriate
DNA microarray. Following the hybridization, the arrays are
washed and the fluorescent signal from the dye is read out using
an optical scanner. The ratio of the red and green fluorescence at
each feature of the array represents the enrichment for the DNA
82 Patrick van Eijk et al.

Fig. 1 Representation of the ChIP-chip procedure. Proteins are crosslinked to

chromatin (a) which is extracted, sonicated, and split into two samples. IP is
carried out on one sample to separate out the chromatin bound to the factor of
interest (b). Both the samples are purified to DNA, amplified by PCR and
differentially labeled (c). They are allowed to hybridize to the microarray
probes and the resulting intensity values from the scanned image (d) are
converted to numerical values that can be plotted (e) and processed as
required by the investigation. Figure adapted from [3]

damage detected and is converted into numerical values during a

process referred to as feature extraction. The data as such are now
ready for data analysis and are loaded into the R statistical program-
ming environment for further investigation using a bespoke soft-
ware package that we developed called Sandcastle [3, 9]. We
describe the typical workflow for our 3D-DIP-Chip method for
the detection of UV-induced CPD lesions as well as platinum
adducts, resulting from cisplatin and oxaliplatin treatment of both
yeast and human cells. Isolation of DNA and chromatin is described
for yeast and human cells. These are followed by procedures for
shearing the DNA, immunoprecipitation, and quantification by
real-time PCR and microarray analysis. Finally, we present our
 Genomic DNA Damage and Repair Detection Methods 83

data analysis workflow. Our workflow is not limited to 3D-DIP-

Chip but can also be applied to conventional ChIP-chip, provided
that a chromatin preparation is used.

3.2 Yeast Cell Work described here uses the haploid BY4742 strain from Euro-
Culture, Growth, Scarf [10]. All work described in this protocol refers to this strain
Treatment, and and mutants are derived from this background, unless stated other-
Harvesting wise. Yeast cell growth and handling were all performed under
sterile conditions and cells were grown on Yeast Extract Peptone
Dextrose (YPD) agar plates at 30
C and stored at 4
C for day-to-
day use and short-term storage. Yeast grown on a plate was used for
inoculating liquid precultures that were incubated in an Infors HT
multitron standard incubator shaking at 180 rpm. Cell counting
was performed on a Neubauer cell counting chamber (Hawksley)
and optical density (OD600) was measured using a Jenway 6300
spectrophotometer (Bibby Scientific Ltd.) (see Note 1 for details).
Centrifugation of large liquid yeast cultures and cell suspensions
was performed using a Beckman-Coulter centrifuge and JA-10
rotor at 4
C unless stated otherwise.
1. Inoculate a 10 ml preculture of the relevant strain and incubate
overnight. When the cells reach stationary phase the next day
they can be stored at 4
C.
2. A calculated amount of this preculture can now be used to
inoculate a large volume culture. Account for 100 ml of culture
per sample and multiply this by the total number of samples
needed to establish the final volume of culture required. Follow
the growth by calculating the cell density or measuring optical
density (OD600) regularly, until the cells reach logarithmic-phase
and a cell density of 2  107 cells/ml (or OD600 ¼ 0.6–1.0).
3. Collect the cells by centrifugation at 4000  g for 5 min at
4
C and resuspend them in prechilled phosphate-buffered
saline (PBS) at a cell density of 2  107 cells/ml. Continue to
step 4 for platinum treatment or step 5 for UV treatment.
4. Treatment of cells with platinum: Cells were treated with
2.5 mM cisplatin or oxaliplatin for 4 h and following treatment
resuspended in YPD for the required repair time.
5. Treatment of cells with UV irradiation was carried out using a
254 nm UVC light (UV products, San Gabriel, CA, USA) at a
dose of 100 J/m2. Set aside 100 ml of cell suspension before
UV irradiation as the nonirradiated (UV) control. Fifty milli-
liters of cell suspension were irradiated in a Pyrex dish
(Φ ¼ 14 cm). This process was repeated for the remainder of
the cell suspension. Keep the irradiated cells in the dark in a
sterile flask (to avoid photoreactivation of CPDs) and set aside a
single volume of 100 ml cell suspension in PBS as the zero
repair-time sample. Collect the remainder of the cells by
84 Patrick van Eijk et al.

centrifugation (4000  g for 5 min) and resuspend them in

fresh YPD at 2  107 cells/ml and incubate them at 30
C for
the desired repair time (e.g., 2 h).
6. For 3D-DIP-chip samples, collect the cells after the repair time
by centrifugation at 4000  g for 5 min at 4
C and resuspend
them in ice-cold PBS and wash the cells twice. After the final
wash step, the cell pellet can now be snap-frozen using liquid
nitrogen and stored at 80
C (see Note 2).
7. For standard ChIP-chip samples:
(a) Crosslink cells for 10 min at room temperature with con-
tinuous shaking using 3.3 ml of 37% formaldehyde for
100 ml culture. The reaction can be quenched using
5.5 ml of 2.5 M glycine for 5 min at room temperature
with continued shaking. Cells are collected by centrifuga-
tion at 4000  g for 5 min at 4
C and washed twice in
40 ml of ice-cold PBS.
(b) After the second wash step resuspend the pellet in 1 ml
FA/SDS (1 mM PMSF) and transfer the suspension to a
2 ml microcentrifuge tube.
(c) Spin down the cells by centrifugation for 5 min at 13,500
rpm(13,500  g) at 4
C (Beckman-Coulter Microfuge
22R). Remove the supernatant and snap-freeze the pellet
using liquid nitrogen and store at 80
C (see Note 2).

3.3 Human Cell Human cell culture was performed under standard sterile condi-
Culture tions and all equipment was autoclaved at 121
C for 15 min, where
applicable. We have successfully applied our 3D-DIP-Chip method
using immortalized human dermal fibroblasts (AG16409, Coriell
Cell Repository, Camden, NJ, USA).
Fibroblasts were grown in DMEM containing 10% foetal calf
serum, 1% L-Glutamine, and 1% penicillin-streptomycin (Sigma-
Aldrich) as monolayers and incubated at 37
C in a humidified
incubator under 5% CO2. We used T75 cm2 or T25 cm2 cell culture
flasks with 10 ml or 5 ml of growth media, respectively, to grow the
cells. Fresh growth media is replaced approximately every 3 days
and cells are reseeded every 4 weeks. The fibroblasts were passaged
close to confluence (~80–90%) every 7–10 days, after which they
were trypsinized and 2  105 to 4  105 cells were reseeded.
1. Remove growth media from the flask and wash the cells with
2 ml of pre-warmed PBS.
2. Add 2 ml of 1 trypsin-EDTA and incubate at 37
C for
15–30 min.
3. Resuspend the cells in the required amount of growth media
for reseeding.
 Genomic DNA Damage and Repair Detection Methods 85

4. Count the cells using an improved Neubauer hemocytometer

(Hawksley) using 10 μl of cell suspension. For these cells the
population doubling (PD) was ~0.4 PD/day and this remained
constant throughout these experiments.
5. Seed 2  105 to 4  105 fibroblast cells (AG16409) on a tissue
culture dish (100  20 mm) and grow for 7–10 days to obtain
~2  106 cells. Alternatively, use a larger tissue culture dish
(150  25 mm) to yield 4–5  106 cells after 10 days of
incubation.
6. Grow cells to confluence before treatment.
7. Remove growth media and wash the cells twice with 2 ml of
warm PBS (5 ml for larger dishes). The cells are now ready for
treatment with platinum-based chemotherapeutics (see step 8)
or UV irradiation (see step 10).

3.4 Platinum 1. Treat cells with an end-concentration of 2.5 mM cisplatin and

Treatment of Human 2.5 mM oxaliplatin in serum-free media. Add the serum-free
Cells media containing the drugs and incubate the dishes at 37
C
for 2 h.
2. In addition to the treated cells, incubate the untreated cells in
serum-free media for 2 h at 37
C as well.

3.5 UV Treatment 1. Add 5 ml PBS to the culture dish and set aside a 5 ml PBS cell
of Human Cells suspension as untreated control.
2. Irradiate cells with 50 J/m2 using a 254 nm UVC lamp (UV
products, San Gabriel).

3.6 Human Cell 1. Harvest the cells after treatment using 1 trypsin-EDTA and
Harvesting and DNA add PBS to the cells suspension to transfer the cells to a 15 ml
Isolation Falcon tube.
2. Pellet the cells by centrifugation at 1000 rpm (200  g) for
5 min at room temperature (Eppendorf centrifuge 5810R).
3. Remove the supernatant and continue immediately with DNA
extraction using DNeasy Blood & Tissue Kit (Qiagen No.
69504). The DNA is now ready for sonication (see
Subheading 3.7).

3.7 Human When chromatin from human cell culture is required, the following
Chromatin Preparation protocol describes the cross-linking and cell lysis necessary to pre-
from Nonadherent pare chromatin for a human ChIP-chip workflow. We have success-
Cells fully used this approach for TK6 cells and HT1080 cells. We won’t
go into detail on the maintenance and propagation of these cells as
this is outside the scope of this method. The procedure outlined
here was optimized for TK6 cells and uses four to eight T175 flasks
grown to a density of ~5  106 cells/ml.
86 Patrick van Eijk et al.

1. Spin down the cells at 1000  g for 5 min at room temperature.

2. Resuspend the pellet in PBS maintaining the original cell den-
sity of ~5  106 cells/ml and transfer the suspension to a 2 l
Erlenmeyer flask.
3. Treat the cells with 1% formaldehyde for cross-linking on an
orbital shaker at room temperature for 8 min.
4. Quench the cross-linking reaction for 5 min at room tempera-
ture by adding 2.5 M glycine to an end concentration of
125 mM (see Note 3).
5. Cool the cells on ice and transfer the cell suspension to 50 ml
tubes.
6. Centrifuge the cells at 1000  g for 5 min at 4
C. Keep the
cells on ice during the following steps.
7. Wash the cells three times in cold PBS, pellet the cells each time
by centrifugation at 1000  g for 5 min at 4
C.
8. Efficient shearing of TK6 cell-derived chromatin requires two
lysis steps: resuspend the cells in 1 ml (~1  107 cells/ml) ChIP
lysis buffer, transfer the cell suspension to a 1.5 ml microcen-
trifuge tube, and lyse the cells by incubating them on ice for
10 min (see Note 4).
9. Spin down the “nuclear pellet” by centrifugation at 1000  g
for 5 min at 4
C and remove the supernatant.
10. Resuspend pellet in 330 μl ChIP shearing buffer. The suspen-
sion is now ready for sonication (see Subheading 3.11).
11. Load six tubes of chromatin for shearing and sonicate for
7 cycles of 30s on/off using “High” settings.
12. Spin down the samples at 15,000  g for 10 min at 4
C to
remove any cell debris.
13. Chromatin preps from similar treatments can now be pooled
and aliquoted for storage.
14. Snap freeze the chromatin preps using liquid nitrogen and
store the samples at 80
C.

3.8 Human 1. Grow cells for chromatin extraction on 10 cm dishes and use
Chromatin Preparation three plates with cells grown to 80% confluence for a single
from Adherent Cells large scale ChIP (yields ~200 μg of chromatin).
2. Cross-link the cells by formaldehyde treatment. Add 1% final
concentration of FA and incubate the dishes for 3–5 min at
room temperature on an orbital shaker.
3. Quench the reaction by adding 2.5 M Glycine to a final con-
centration of 125 mM and incubate the cell suspension for
5–10 min at room temperature.
 Genomic DNA Damage and Repair Detection Methods 87

4. Discard the media (HT1080 cells are adherent) after quench-

ing and cross-linking and wash the cells on plate by adding
10 ml of ice-cold PBS to the corner of the dish and swirl. Avoid
applying large volumes of liquid directly onto the plate to
prevent the cells from peeling off.
5. Scrap the cells from the plate into 10–30 ml PBS containing a
protease inhibitor cocktail.
6. Spin down the cells in tabletop centrifuge at 1200 rpm
(290  g) for 5 min at 4
C.
7. Discard the supernatant and take care to remove any residual
PBS because it will interfere with the lysis buffer.
8. Resuspend the pellet in SDS lysis buffer supplemented with
protease inhibitor cocktail to 1–1.5  107 cells/ml and aliquot
in microcentrifuge tubes in volumes of 300 μl.
9. Lyse the cells by incubating the suspension on ice for
20–30 min.
10. Shear the chromatin by sonication using a Diagenode BioR-
uptor. Sonicate for 5 min using ten cycles at the “high” setting
cycling for 30 s on 20 s off. Place the samples on ice for a couple
of minutes (see Subheading 3.11).
11. Spin down the sonicated chromatin prep for 10 min at
13,000 rpm (16,000  g) at 4
C.
12. Chromatin preps from similar treatments can now be pooled
and aliquoted for storage.
13. Snap freeze the chromatin preps using liquid nitrogen and
store the samples at 80
C.

3.9 Yeast DNA 1. Resuspend the untreated, platinum-treated, or UV-treated cell

Preparation pellet in 5 ml of sorbitol-TE solution containing 5 mg Zymo-
lyase and 10 μl β-mercaptoethanol to the samples and mix by
shaking.
2. Spheroplasts are generated by incubating the cells at 4
C
overnight or at 37
C for 30–60 min with intermittent shaking.
Spheroplast production can be monitored under a microscope
(see Note 5).
3. Gentle centrifugation at 3000 rpm (800  g) for 5 min at 4
C
will pellet the spheroplasts, which can then be washed once in
sorbitol-TE and resuspended in 5 ml of DNA lysis buffer/PBS
1:1 (v/v).
4. Add 200 μl RNAse A and mix by vortexing. Incubate the
samples at 37
C for 1 h with occasional shaking.
5. Add 200 μl pronase (20 mg/ml) and incubate the samples at
37
C for 1 h, follow by 1 h incubation at 65
C with occasional
shaking.
88 Patrick van Eijk et al.

6. Perform two phenol/chloroform extractions to guarantee

complete deproteinization. Add 6 ml phenol/chloroform to
the cell lysate (6 ml) and vortexing vigorously until
homogeneous.
7. Spin down the sample at 10,000 rpm (12,000  g) (Beckman-
Coulter centrifuge, JA-20 rotor) for 10 min and transfer of the
upper aqueous phase to a 15 ml Falcon tube.
8. Repeat steps 7 and 8 for the second phenol/chloroform
extraction.
9. Precipitate the DNA by adding two volumes (12 ml) of cold
100% ethanol (20
C) with gentle mixing and incubate over-
night at 20
C (or 30 min at 80
C).
10. Recover the DNA precipitate by centrifuging for 15 min at
4000 rpm (3200  g) at 4
C and resuspend the pellet in 1 ml
TE. Allow the DNA to dissolve completely. This may take time.
11. Precipitate the DNA again using isopropanol. Add 1 ml of
prechilled 100% isopropanol (20
C) and shake gently (geno-
mic DNA is sensitive to shearing). Incubate at room tempera-
ture for 30 min with occasional, gentle mixing. Collect the
DNA precipitate carefully using a pipette tip and resuspend in
1 ml TE.
12. Check the DNA for purity by agarose gel electrophoresis and
quantify the DNA (i.e., using a NanoDrop UV spectropho-
tometer). The DNA is now ready for sonication (see Subhead-
ing 3.11).

3.10 Yeast 1. Thaw the cell pellet on ice and resuspended in 0.5 ml FA/SDS
Chromatin Preparation (1 mM PMSF) in a 2 ml microcentrifuge tube.
2. Lyse the cells by bead beating. Add 500 μl of glass beads and
vortex for 10 min at 4
C.
3. Retrieve the cell lysate by puncturing a hole in the bottom of
the 2 ml tube by using a hot needle.
4. Place the punctured tube in a 15 ml Falcon tube and centrifuge
at 2000 rpm (300  g) for 2 min at 4
C (Eppendorf centrifuge
5810R).
5. Wash the glass beads with 500 μl of FA/SDS (1 mM PMSF)
and centrifuge again at 2000 rpm (300  g) for 2 min at 4
C.
6. Transfer the lysate to a fresh 2 ml tube and centrifuge at
4000 rpm (3200  g) for 20 min at 4
C (Beckman-Coulter
Microfuge 22R or Eppendorf). This removes any non-
crosslinked or soluble proteins.
7. Discard the supernatant and resuspend the pellet in 1 ml FA/
SDS (1 mM PMSF). Split the chromatin prep over two 2 ml
tubes containing 500 μl each.
 Genomic DNA Damage and Repair Detection Methods 89

8. Sonicate the lysate using a BioRuptor sonicator (Diagenode)

on “high” power settings for eight cycles of 30 s on and 30 s off
at 4
C.
9. Spin down the sonicated chromatin at 13,000 rpm
(16,000  g) for 10 min at 4
C to remove cell debris.
10. Recover the supernatant in a fresh microtube and centrifuge
again at max speed for 10 min at 4
C (Beckman-Coulter
Microfuge 22R). Collect the supernatant and use it to prepare
an input sample (see Subheading 3.12, step 12).
11. Check the DNA for purity by agarose gel electrophoresis and
quantify the DNA (i.e., using a NanoDrop UV
spectrophotometer).
12. Snap freeze the chromatin prep in liquid nitrogen to prepare it
for storage at 80
C. The chromatin is now ready for immu-
noprecipitation, see Subheading 3.8.

3.11 Sonication We use a Diagenode BioRuptor for sonication of both chromatin

of Chromatin and DNA and naked DNA. The settings described in this protocol are stan-
dardized for four 2.0 ml microcentrifuge tubes containing 200 μl of
chromatin or DNA leaving two slots open in the sonicators tube
holder, unless stated otherwise (see below and individual sections
for more details).
1. Human DNA is to be diluted to 100 ng/μl for optimal shear-
ing, yeast DNA is sonicated using a 200 ng/μl concentration.
2. Samples are sonicated at “high” power settings at 4
C using a
circulating cooled water bath.
3. The duration and number of on-and-off cycles were calibrated
based on the source material (DNA versus chromatin). The
following numbers are a guideline (see Note 6):

Yeast DNA 20 cycles 300 on 300 off ~300–500 bp

Yeast chromatin 8 cycles 300 on 300 off ~500 bp
Human DNA 25 cycles 300 on 300 off ~200–300 bp
Human 7 cycles 300 on 300 off ~500 bp (TK6)
chromatin
Human 10 cycles 300 on 200 off ~500 bp (HT1080)
chromatin

4. The size-distribution of the sonicated DNA was assessed by gel

electrophoresis on a 1.2% agarose gel using the FastRuler™
low range DNA ladder (Fermentas).
5. Alternatively, size-distribution can be checked using a TapeSta-
tion (Agilent Technologies) using D1000 screentapes.
90 Patrick van Eijk et al.

3.12 Both proteins (ChIP-chip) and DNA lesions (3D-DIP-Chip) can

Immunoprecipitation now be immunoprecipitated according to the following protocol.
The most important distinction is the use of naked DNA for DIP
and chromatin for ChIP. For DIP, we use the anti-cisplatin mod-
ified DNA, CP9/19, antibody (ab103261, Abcam), or Mouse
Anti-CPD antibody (CAC-NM-DND-001, COSMO Bioscience).
Use 0.5 mg of CP9/19 antibody 50 μl of Dynabeads for the
detection of cisplatin-induced DNA damage or 1.5 μg of antibody
for the detection of oxaliplatin-induced damage (see Note 7).
1. Take 50 μl Dynabeads per sample and wash three times using
500 μl PBS-BSA (1 mg/ml).
2. Resuspend the beads in 50 μl PBS-BSA (1 mg/ml) and add the
indicated amount of antibody.
3. Incubate the antibody with the washed Dynabeads for 30 min
at 30
C in a thermomixer (Eppendorf) shaking at 1300 rpm
(optional: incubate overnight at 4
C on a rotator).
4. Wash the Dynabeads again for three rounds using 500 μl
1 mg/ml PBS-BSA and resuspend the beads in 50 μl PBS
after the final washing step.
5. Add 30 μl PBS-BSA (10 mg/ml), 6–10 μg sonicated DNA or
3 μg sheared chromatin and add 1 PBS to an end-volume of
300 μl.
6. Incubate the IP/ChIP reaction at 21
C for 3 h shaking at
1300 rpm using a thermomixer.
7. Wash the DynaBeads. All buffers are prechilled at 4
C and kept
on ice. Resuspend the beads in 500 μl FA/SDS, incubate on a
rotator for ~1 min, precipitate the beads on a magnet.
8. Remove the supernatant and repeat this process twice using
FA/SDS 0.5 M NaCl (for yeast) or FA/SDS 1.0 M NaCl (for
human).
9. Next, wash the beads using LiCl solution and 1 TE.
10. Elute the DNA from the beads by incubating them in 125 μl
pronase buffer at 65
C shaking at 900 rpm for 30 min in a
thermomixer.
11. Add 6.25 μl pronase (20 mg/ml) to the IP samples
12. Prepare an input sample containing ~1/10 of IPed DNA by
diluting DNA into 100 μl using TE. Add 25 μl 5 pronase
buffer and 6.25 μl pronase to the input DNA and incubate the
samples together with the IP samples, overnight at 65
C to
remove all proteins and reverse the cross-linking.
13. For platinum-treated DNA:
(a) Treat the DNA with NaCN to chemically remove the
platinum-induced adducts. Incubate both the IN and IP
 Genomic DNA Damage and Repair Detection Methods 91

DNA samples with 0.2 M NaCN for 2 h at 65
C. Add

5 μl RNAse (10 mg/ml) for the last 45 min of the
incubation.
(b) Purify the Input and IP DNA using PureLink™ Quick
PCR Purification Kit (Invitrogen) following the supplier’s
instructions, eluting in 50 μl elution buffer.
(c) The DNA is now ready for qRT-PCR validation and
microarray sample preparation.
14. For UV treated DNA:
(a) Add 5 μl of RNAse A (10 mg/ml) and incubate at 37
C
for 1 h.
(b) For yeast: Purify the Input and IP DNA using PureLink™
Quick PCR Purification Kit (Invitrogen) following the
supplier’s instructions, eluting in 50 μl elution buffer.
(c) For human: extract the IP and Input DNA by phenol-
chloroform extraction.
(d) Treat 40 μl sample (IN and IP) with the PreCR DNA
repair kit (New England Biolabs) according to the suppli-
er’s instructions.
(e) Purify the Input and IP DNA using PureLink™ Quick
PCR Purification Kit (Invitrogen) following the supplier’s
instructions, eluting in 50 μl elution buffer.
(f) The DNA is now ready for qRT-PCR validation and
microarray sample preparation.

3.13 Quantitative To assess the successful enrichment of the IP compared to the input
Real-time PCR DNA before continuing with the microarray, perform qRT-PCR
using your favorite locus and RT-PCR system and reagents. Here,
we provide details for the use of the iTaq™ Universal SYBR Green
Supermix and CFX Connect™ Real-Time PCR detection system
(Bio-Rad). We used the MFA2 locus for yeast 3D-DIP-Chip exper-
iment for the detection of DNA damages. Work on human cells
made use of the 28S ribosomal RNA gene to verify successful DIP.
The volumes of samples and dilutions listed here are enough for
enrichment detection at 3 loci, with samples run in triplicate.
1. Dilute the input and IP samples 10 by adding 5 μl sample to
45 μl ddH2O.
2. Create an internal dilution series of input DNA ranging from
1/10 to 1/100,000 (5 μl in 45 μl ddH2O).
3. Create a master mix by adding forward and reverse primers to
the SYBR Green Supermix to an end concentration of 1 μM
diluting from 100 μM stock.
4. Aliquot 5 μl of the SYBR Green Supermix and primer master
mix in a 96-well PCR plate (Bio-Rad). Account for samples
92 Patrick van Eijk et al.

(input and IP) to be run in triplicate and include a set of three

non-template controls (NTC). Add 5 μl of sample to each well.
5. Vortex the 96-well PCR plate and spin down for 2 min at
300 rpm at room temperature to collect the sample at the
bottom of the wells.
6. Load the 96-well PCR plate into the CFX Connect™ Real-
Time PCR detection system (Bio-Rad).
7. Generate melt-curves to check for nonspecific amplicons.
8. Analyze the data using the CFX Manager™ Software v3.1
(Bio-Rad). Eliminate data points that deviated more than
0.5 units of the threshold (Ct).

3.14 Yeast DNA Yeast DIP and ChIP and accompanying input DNA samples can
Labeling now be processed per the Agilent Technologies Yeast ChIP on chip
and Microarray protocol (Agilent Technologies Yeast ChIP-on-chip Analysis Pro-
Hybridization tocol, version 9.2). A short summary of the steps is described
below. The method for each of these steps has been described in
detail previously [11] and is described extensively in the Agilent
Yeast ChIP on chip protocol.
1. Blunt-end the DNA using T4 DNA polymerase.
2. Perform a phenol/chloroform extraction and precipitate the
DNA using ice-cold 100% ethanol.
3. Ligate the linker hybrid to the input and IP DNA using T4
DNA ligase.
4. Amplify the DNA using two rounds of ligation-mediated PCR
(LM-PCR).
5. Measure the DNA concentration in a NanoDrop spectropho-
tometer (ThermoScientific) and normalize the concentration
of all samples to 150 ng/μl using ddH2O.
6. Differentially label 10.5 μl the IP and IN DNA with the Cy5
and Cy3 fluorophores, respectively, using the BioPrime Total
Genome Labelling System (Invitrogen) for 2 h at 37
C (see
Note 8).
7. Purify the DNA using the PureLink PCR Purification Kit
(Invitrogen) and elute in 55 μl elution buffer E1.
8. Apply 5 μl of labeled sample on a NanoDrop spectrophotome-
ter (ThermoScientific) using the MicroArray Measurement
Software Module to establish the labeling efficiency.
9. Combine the labeled IP and IN samples (100 μl) and purify the
DNA using ethanol precipitation (see Agilent Yeast ChIP on
chip protocol for details).
10. Resuspend the DNA in 39 μl ddH2O and add the hybridiza-
tion mixture (end volume is 110 μl).
 Genomic DNA Damage and Repair Detection Methods 93

11. Apply 100 μl of the hybridization mixture to each Agilent yeast

4  44 k microarray (see Agilent Yeast ChIP on chip protocol
for details).
12. Hybridize for 24 h at 65
C at 20 rpm. For larger format arrays,
e.g., 4x44K the hybridization time is 40 h.

3.15 Human DNA Successful amplification of human samples was performed using the
Labeling GenomePlex Complete Whole Genome Amplification kit (WGA2,
and Microarray Sigma-Aldrich). This method was introduced for these samples
Hybridization because the LM-PCR-based method relies on two rounds of ampli-
fication, which is a potential source of amplification bias particularly
when working with larger genomes [12]. This step of the 3D-DIP-
Chip workflow is therefore slightly different for yeast and human-
derived DNA samples and details are described below. The manu-
facturer’s instructions were followed for amplification of human
DNA using the WGA2 kit, with the following alterations:
1. No fragmentation was necessary as our DNA is fragmented by
sonication already (see Subheading 3.11).
2. Use the entire DIP sample and add 2 μl of 1 library prepara-
tion buffer and 1 μl of library stabilization buffer.
3. Incubate the mixture at 95
C for 2 min and directly after, cool
the sample on ice.
4. Add 1 μl of library preparation enzyme and incubate in a
thermal cycler as follows; 20 min at 16
C, 20 min at 24
C,
20 min at 37
C, and 5 min at 75
C.
5. Next, add 10 amplification master mix, 47.5 μl nuclease-free
H2O, and 5 μl WGA DNA polymerase to each sample. Mix and
incubate in a thermal cycler for amplification using: 3 min at
95
C, 14–16 cycles of 15 s at 95
C, and 5 min at 65
C.
6. Purify the amplified DNA using the PureLink PCR Purification
Kit (Invitrogen) and elute in 15 μl nuclease-free H2O.
7. Quantify the DNA using a NanoDrop spectrophotometer
(ThermoScientific). A yield between 1 and 5 μg is to be
expected and is sufficient for microarray analysis.
8. Label the library DNA as described previously using the Bio-
Prime Total Genomic Labelling System (Invitrogen, Subhead-
ing 3.14, steps 6–9) but now incubate the reaction for 3 h at
37
C.
9. Combine the IP and IN samples and apply 100 μl of the
hybridization mixture to a custom-designed Agilent Technol-
ogies human 1  244 k microarray.
10. Hybridize for 40 h at 65
C at 20 rpm.
94 Patrick van Eijk et al.

3.16 Microarray 1. Both human and yeast arrays are washed twice using SSPE
Washing, Scanning, washing buffers as described in detail in the Agilent Technolo-
and Feature Extraction gies Yeast ChIP on chip protocol (Agilent Technologies Yeast
ChIP-on-chip Analysis Protocol, version 9.2).
2. A washed and dried microarray slide is now ready for scanning.
3. The scanned image of the array records the fluorescence of
both the Cy3 and Cy5 dyes in a TIFF image, the Agilent
Feature Extraction software retrieves the raw data for analysis.
4. The raw data files can now be loaded into the R statistical
programming language environment [9] by using the built-in
functions of Sandcastle developed in our lab [3].
Details of data loading, normalization, visualization, and peak
detection are discussed in the next section.

3.17 Overview The following section describes a typical data analysis workflow
of Data Analysis Using using Sandcastle. The Sandcastle software package is written in
Sandcastle the statistical programming language of R [9] and provides a
novel normalization procedure, enrichment detection, and graphi-
cal tools for plotting data. For full details on how to use each
command, we refer to the Sandcastle instruction document
(http://reedlab.cardiff.ac.uk/sandcastle/sandcastleinstructions.
pdf) and the Sandcastle documentation and vignettes that can be
accessed from within R.
In short, the normalization procedure performed by Sandcastle
consists of four stages (Fig. 2) and works only when the back-
ground and enrichment signal can be clearly differentiated (see
Fig. 2). The first step of preprocessing (1) removes unused or
unusable data points, ensuring that all datasets contain the same
amount of useable data points before moving forward. It should be
noted that this step only removes a small set of the data points. (2)
During the second stage of the normalization process, within-
condition quantile normalization is applied to reduce variation
among replicate datasets derived from the same experimental con-
dition. The third and fourth stages of normalization introduce the
unique between-condition normalization by (3) shifting and (4)
scaling the data and can only be applied to data of which the
background population can be identified. Shifting the datasets is
achieved by centering the background subpopulation to the y-axis
(x ¼ 0) for each individual dataset. This generates the same mean of
zero for each of the background subpopulation of data points for all
datasets. Next, to be able to scale the data, the left-hand, negative
portion of the background data is mirrored along the y-axis to
create a symmetrical dataset that acts as a simulated background
distribution (Fig. 2). The background population can now be
scaled to a standard normal distribution with a mean of zero and
 Genomic DNA Damage and Repair Detection Methods 95

Fig. 2 Representation of the normalization procedure. (a) Raw density profiles of datasets from two experi-
mental conditions (red and blue), each with three replicates. Differences in the shapes of the profiles indicate
96 Patrick van Eijk et al.

SD of 1. This transforms each dataset to have the same theoretically

consistent background subpopulation against which it is now pos-
sible to assess the enriched, biological data in relative terms.

3.18 Loading Data Sandcastle can use raw ChIP-chip data derived from Agilent’s
and Compatible Data Feature Extraction software or tab-delimited files from other
Formats sources. When the data are loaded by Sandcastle the arrayData
object that is created contains the log2 IP/IN ratios for each
probe or feature of the microarray, including information on the
genomic coordinates and probe annotation.
1. The findArrayFiles function will identify files suitable for load-
ing into R in your R working directory, giving each file a
numbered identifier that can be used when loading the data.
2. First, a quality control can be performed on the raw data. Using
checkData, files found by the findArrayFiles function will be put
through QC analysis. A “check.pdf” file will be generated in
your working directory that can be printed out for logging the
experiment. This file provides fields to write down the experi-
mental conditions, as well as pseudo images of the red and
green channel derived from the scanned array for visual inspec-
tion. The distribution of signal in the dataset is presented as a
box-plot, density profile and scatter plot.
3. Using the loadArrayFiles function, data will be loaded into an
arrayData object. Making use of the numbers assigned to each

Fig. 2 (continued) experimentally induced biologically relevant changes, but these cannot be compared in
their raw state. (b) Quantile normalizing all datasets together (1) removes much of the experimentally induced,
biologically relevant differences between them. This is not desirable, as these differences cannot then be
investigated. Sandcastle quantile normalizes the datasets from each experimental condition separately, to
maintain these biological differences. Quantile normalization makes each of the datasets follow the same
distribution, meaning all density profiles from each experimental condition overlap each other (2). This
reduces intra-condition—but not inter-condition—technical variations. (c) Each dataset consists of two
overlapping subpopulations (dashed lines), background (BG), and enriched (EN). These cannot be fully
discerned in the data and only the overall population (solid lines) is known. Sandcastle performs inter-
condition normalization based on estimated background subpopulations. This requires the central (modal)
point of the background subpopulations to be identifiable (marked with triangles). If this central point cannot
be discerned (for example, if the background subpopulation is too small), then the Sandcastle normalization
cannot be applied. (d) Data are first shifted to center the modal point of the estimated background
subpopulation on zero (indicated by arrows). (e) To estimate the properties of the whole background
subpopulation all negative values (the left-hand side of the estimated background subpopulation following
the shift step) are mirrored into the positive (indicated by arrow; dashed lines show mirrored data). This allows
the standard deviation of the estimated background subpopulation to be calculated. (f) Data are scaled to
make the calculated standard deviation of the estimated background subpopulation 1 (indicated by arrows).
(g) The resulting fully normalized datasets have estimated background subpopulations with the same mean (0)
and standard deviation (1). Comparisons of data between conditions can now be made relative to this common
background. For clarity axis labels are only shown in (a)—all other x- and y-axes are ratio and density values
respectively. Vertical gray lines indicate 0, which are only labeled in (f)
 Genomic DNA Damage and Repair Detection Methods 97

dataset by findArrayFiles the order and grouping of the data-

sets to be loaded can be specified here.
4. Next, the user needs to load the annotation data from the yeast
ensemble database to an annotation data object using loadAn-
notation. The annotation data contains information about all
annotated genes used for genomic analysis and graphical repre-
sentation of genomic data. The default data loaded is that of
yeast but human annotation data can also be loaded (see the
Sandcastle documentation for more details).

3.19 Normalization 1. Data loaded as an arrayData object can now be normalized by

calling the normalize function.
2. The “batches” argument can be used with the normalization
function to combine the replicate datasets that need to be
quantile normalized to eliminate the “within” condition varia-
tion. After this step the group(s) of quantile normalized data
will be subjected to the scaling and shifting algorithms to allow
the “between” condition comparison.
3. A graphical representation of the individual steps of the nor-
malization procedure can be generated using the “plot” argu-
ment, allowing the user to follow exactly each step of the
normalization process.
4. Knowledge of the dataset and nature of the signal detected in
the 3D-DIPChip or ChIP-chip experiment are important in
determining whether the dataset is amenable to Sandcastle’s
full normalization procedure (Fig. 2). By default, all the four
steps of the normalization procedure, described earlier (Sub-
heading 3.17), are performed. However, for 3D-DIP-Chip
data that does not contain a discernible background signal,
the scaling and shifting steps must be omitted by setting the
“shift” and “scale” arguments to FALSE.
5. After applying the appropriate normalization to the data, the
normalisationAssumptions function provides a means of vali-
dating whether two datasets that were normalized in different
batches (e.g., treated and untreated) are suited to the normali-
zation procedure (Fig. 2).

3.20 Enrichment After following the previous steps enrichment detection and peak
Detection detection can be performed by calling the enrichmentDetection func-
tion. Enrichment detection describes the probe enrichment, whereas
peak detection finds local maxima and generates a sandcastlePeaks list
object, identifying potential binding regions of the protein of interest
in the case of ChIP-chip data analysis (see Note 9). 3D-DIP-Chip
data describing UV-induced CPDs and their removal is not amenable
98 Patrick van Eijk et al.

of peak detection since CPD distribution is random and dispersed and

is not expected to generate defined local peaks at significant
resolution.

3.21 Plotting Sandcastle provides means to plot data in a linear, genomic fashion
as well as aggregating data into composite plots around genomic
features such as ORFs, TSSs, TESs but also peaks or other genomic
coordinates supplied by the user.
1. genomePlot plots the data for the first chromosome, but can
also be instructed to display user-defined sections of a chromo-
some or export a whole-genome plot to a PDF file. This
function uses the annotation data loaded through loadAnnota-
tion (see Subheading 3.18, step 4) to plot the location and
direction of genes. Using standard R-commands, data can be
group together and averaged to plot results in different
context.
2. profilePlot allows users to generate so-called composite plots that
aggregate genomic data from a user-defined selection of geno-
mic positions or structures and compiles them into a single
plot. These genomic features can, for instance, be protein
binding sites, peaks detected using Sandcastle’s enrichmentDe-
tection (Subheading 3.20) or ORF-features extracted from the
ensemble annotation data.

4 Notes

1. Following yeast cell growth using OD600 can vary based on the
yeast strain grown and spectrophotometer used. Cell-counting
is more reliable but also more time consuming. Yeast cell
doubling time can vary from 1.4 to close to 2.0 h depending
on the genetic background or mutant used.
2. Snap-freezing and storing yeast cell pellets at 80
C after
treatment improves the efficiency of cell lysis.
3. When performing cross-linking in medium (i.e., for the
HT1080 cells), quenching the formaldehyde cross-linking
reaction using glycine changes the color of the medium due
to the pH of the glycine stock (~pH 2.5) if a pH indicator is
present.
4. In yeast only use PMSF during lysis because of high protease
activity in yeast extracts. Use of protease inhibitor cocktails is
possible but not required. Use of protease inhibitor cocktails
(i.e., from Roche) is recommended for human cell lysis and
chromatin preparation.
 Genomic DNA Damage and Repair Detection Methods 99

5. Monitoring spheroplast production:

(a) Add 100 μl 10% SDS to 100 μl cell suspension.
(b) Deposit 5–10 μl of this mixture together with 5–10 μl of
cell suspension on an object slide.
(c) The absence of the cell wall should be clearly discernible
comparing the cell suspension with or without SDS. In
the presence of SDS the cells are very faintly visible com-
pared to the untreated suspension.
(d) Spheroplast production is complete if most cells are poorly
visible in the presence of SDS.
6. Processing sonication human DNA samples for microarray
analysis revealed that a shorter fragment length was required
to improve microarray resolution. For yeast studies, an average
fragment length of 300–500 bp was used, while for human
studies a shorter fragment length of 200–300 bp was required.
7. To increase the overall DNA yield in the human DIP samples,
DIP samples can be processed in parallel and two samples
pooled together prior to the qRT-PCR stage. Multiple labeled
IP DNA samples can also be combined prior to microarray
hybridization as a means of boosting the microarray signal.
8. You can store the IP and IN DNA after labeling for any length
of time prior to purification if 5 μl of 0.5 M EDTA is added to
the reaction mixture to quench the reaction.
9. Enrichment detection power improves by providing replicate
datasets, but detection of enrichment or peaks in individual
datasets can be performed as well.

References

1. Lindahl T (1993) Instability and decay of the 7. Alexandrov LB et al (2013) Signatures of

primary structure of DNA. Nature 362 mutational processes in human cancer. Nature
(6422):709–715 500(7463):415–421
2. Powell JR et al (2015) 3D-DIP-Chip: a 8. Perera D et al (2016) Differential DNA repair
microarray-based method to measure genomic underlies mutation hotspots at active promo-
DNA damage. Sci Rep 5:7975 ters in cancer genomes. Nature 532
3. Bennett M et al (2015) Sandcastle: software for (7598):259–263
revealing latent information in multiple experi- 9. R Core Team R (2015) A language and envi-
mental ChIP-chip datasets via a novel normal- ronment for statistical computing
isation procedure. Sci Rep 5:13395 10. Brachmann CB et al (1998) Designer deletion
4. Yu S et al (2016) Global genome nucleotide strains derived from Saccharomyces cerevisiae
excision repair is organized into domains that S288C: a useful set of strains and plasmids for
promote efficient DNA repair in chromatin. PCR-mediated gene disruption and other
Genome Res 26(10):1376–1387 applications. Yeast 14(2):115–132
5. Nik-Zainal S et al (2016) Landscape of somatic 11. Teng Y et al (2011) A novel method for the
mutations in 560 breast cancer whole-genome genome-wide high resolution analysis of DNA
sequences. Nature 534(7605):47–54 damage. Nucleic Acids Res 39(2):e10
6. Morganella S et al (2016) The topography of 12. Powell JR (2014) Thesis, Cardiff University
mutational processes in breast cancer genomes. ORCA portal. http://orca.cf.ac.uk/id/
Nat Commun 7:11383 eprint/65976
 Chapter 8

Study of UV-induced DNA Repair Factor Recruitment:

Kinetics and Dynamics
Sarah Sertic, Stefania Roma, Paolo Plevani, Federico Lazzaro,
and Marco Muzi-Falconi

Abstract
The local UV irradiation technique enables detection, kinetic measurements of recruitment, and quantifi-
cation of DNA Damage Response (DDR) proteins at the site of UV-induced DNA damage.
Using Isopore filters with high density pores of a broad range of sizes, it is possible to UV irradiate and
damage only a very small portion of the nucleus of a cell by letting UV light pass only through the pores.
Immunofluorescent analyses of modified DNA nucleotides, proteins, or fluorescently tagged versions of
target factors can be used as markers to label and study UV-induced lesions and their repair.

Key words UV-induced DNA damage, Immunofluorescence, Protein recruitment, Sequential assem-
bly, NER

1 Introduction

Nucleotide Excision Repair (NER) is involved in the removal of

bulky lesions on DNA, and in human cells is the only process that
repairs UV-induced DNA injuries. Several genetic and biochemical
studies during the last few decades have defined details of para-
mount importance for the mechanistic characterization of NER
[1]. It is a multistep process involving many enzymes starting
from recognition of DNA damage, unwinding of double helix,
endonucleolytic cleavage 50 and 30 to the lesion, gap-refilling and
ligation.
However, characterization of the sequential assembly, dynamic
and kinetics of complexes involved in the different steps of DNA
lesion removal and repair has been possible thanks to the (r)evolu-
tion of sophisticated fluorescence microscopy techniques. In par-
ticular, different laboratories [2–4] developed a very powerful
technique, based on the induction of Local UV Damage (LUD).
The added value of this technique resides in its simplicity.

Marco Muzi-Falconi and Grant W. Brown (eds.), Genome Instability: Methods and Protocols, Methods in Molecular Biology,
vol. 1672, DOI 10.1007/978-1-4939-7306-4_8, © Springer Science+Business Media LLC 2018

101
102 Sarah Sertic et al.

UV-light produces bulky lesions on DNA, which are very

dangerous if not properly repaired as they can be led to double-
strand breaks and block both DNA replication and transcription.
Mutations in human genes coding for NER factors are responsible
for the onset of severe pathologies, among which are xeroderma
pigmentosum (XP), Cockayne syndrome (CS), and Tricothiody-
strophy (TTD). The principal and common symptom to all these
conditions is extreme sensitivity to UV irradiation, while other
manifestations can vary. XP mutations have been grouped into
eight complementation groups (from XP-A to XP-G and XP-V).
The LUD technique has been very important to define who is first
recruited to the UV lesion and the sequential assembly of all the
factors [4]. For example, it has been used to demonstrate that the
incisions at the 50 side of the lesions by XPF and 30 side of the lesion
by XPG are sequential and they occur in this order [5]. Using XP
cells from patients, recruitment at LUDs has been used to identify
the genetic dependencies for the recruitment at UV lesions of
different new factors, such as XRCC1-Lig3, MutSβ, MDC1-
53BP1, Pol κ, and EXO1 [6–10].
Here, we describe the technical procedure to generate LUDs
that can be used to define details for the recruitment and involve-
ment of proteins in UV-induced DNA damage processing and
repair.

2 Materials

2.1 Cell Cultures 1. Cell dishes, multiwell plates.

2. PBS sterile for cell culture.
3. Specific cell culture medium complemented with Pen/Strep,
Glutammine, Fetal Bovine Serum (at different concentrations).
4. Glass coverslips 20  20 mm.

2.2 Local UV 1. Thin tweezers.

Irradiation 2. Isopore filters 5 μm (Merck-Millipore TMTP04700).
3. UV dosimeter.
4. Timer.

2.3 Immunostaining 1. PBS.

2. 4% Paraformaldehyde (2).
3. Formaldehyde 37% (10).
4. Permeabilization Buffer: 0.5% Triton-X100 in PBS (4  C).
5. Blocking Buffer: 10% BSA in PBS (4  C).
6. Antibodies incubation buffer: 0.1% Tween-20 in PBS.
7. ProlongGold with DAPI.
 Repair of UV Lesions 103

2.4 Detection and 1. Widefield Fluorescence Microscope.

Analysis 2. High-resolution Spectral Confocal Microscope.

3 Methods

1. Day 1: cell seeding (see Note 1) on coverslip, one for each

treatment, UV dose or/and different kinetics.
2. Day 2: transfect cells or treat them with any chemical that needs
24 h incubation.
3. Day 3: Local UV irradiation and immunofluorescence.
4. Day 4: Fluorescence microscopy analyses.

3.1 Local UV 1. Place UV-Lamp 254 nm on a proper support to ensure appro-

Irradiation priate distance between lamp and sample, so that cells receive a
UV dose of 0.5–1 J/m2/s (see Note 2). Switch on lamp.
2. Control UV dose with a dosimeter after a lamp warm-up time
of at least 15–20 min.
3. Prepare petri dish lid covered with parafilm.
4. Wash cells with PBS.
5. Lift the coverslip with tweezers and position it on the parafilm
covered lid, with cells on top.
6. Wash the Isopore filter with PBS in a petri dish.
7. Place the filter on the top of the coverslip (Fig. 1).
8. Set an alarm for the seconds of irradiation needed to reach the
desired UV dose (e.g., 40 J/m2, corresponding approximately
to a total irradiation of 2 J/m2, when using 5 μm pores).
9. Place the petri dish lid holding the coverslip and filter under the
warmed UV lamp.
10. After the appropriate time, when desired UV dose is reached,
remove it and lift Isopore filter with tweezers with one hand,
while adding PBS with a hand pipette using the other hand (see
Note 3).
11. Lift coverslip and put it back in a multiwell plate.

Glass coverslip

Isopore Filter

Plastic support, i.e.

petri dishes lid

Fig. 1 Local UV damage casting setup

104 Sarah Sertic et al.

Fig. 2 Example of two different NER factors at LUDs by immunofluorescence. MRC5VI cells were seeded and
irradiated with 40 J/m2 through Isopore filter of 5 μm dimension and incubated for 30 min. Scale bar ¼ 5 μm

12. Incubate for the desired period of time (e.g., 1 h) in a 37  C

humidified incubator.
13. Wash cells twice with PBS.
14. Fix cells with 2% paraformaldehyde for 20 min at room tem-
perature or 3.7% formaldehyde for 15 min.
15. Wash cells twice with PBS (see Note 4).
16. Add ice-cold permeabilization buffer for 5 min.
17. Add blocking buffer for 30–60 min at room temperature.
18. Incubate with primary antibody (diluted in PBST) at the
appropriate dilution for 1.5–2 h at room temperature in a
humidified chamber.
19. Wash three times with PBST.
20. Incubate with fluorescent secondary antibody (AlexaFluor®,
Molecular Probes) for 1 h at room temperature (diluted in
PBST) protected from light.
21. Wash three times with PBST.
22. Mount coverslip with ProLong Gold mounting solution with
DAPI.
23. Let coverslip dry overnight at room temperature (see Note 5).

3.2 Microscopic Samples can be analyzed under a widefield microscope or confocal

Analysis microscope depending on the need. A 63 or 100 immersion oil
objective can be used for the analysis. Example images are pre-
sented in Figs. 2 and 3.

4 Notes

1. Be careful of seeding cells at a maximum confluence of 70% the

day of Local UV damage protocol. Higher confluence would
give rise to a bad LUD accumulation and immunofluorescence,
while lower confluence would favor cell loss when removing
Isopore filters, leaving few cells for the analyses.
 Repair of UV Lesions 105

Fig. 3 Application of LUD technique. MRC5VI cells were seeded and transfected with EXO1-mCherry tagged
protein. The following day, cells were Local UV irradiated at 100 J/m2 in the presence or absence of AraC.
Representative images, acquired under a confocal microscope (Leica SPE) for each condition, are shown.
Scale bar ¼ 5 μm

2. An appropriate UV dose (J/m2/s) is suggested. If the dose is

too high reproducibility can be affected, while too low doses
will need long incubation times of living cells under the lamp in
an almost dry condition. We generally use a 6 W 254 nm lamp
(VL-6.C) placed 35 cm away from the cells.
3. Be careful when lifting the filter, pipetting PBS simultaneously
is essential to avoid scraping cells off the coverslip.
4. If needed, fixed cells can be stored at 4  C for up to 1 week.
5. Do not place coverslip under an immersion oil objective before
drying is completed, since wrong movement of the sample
might result in scratching of cells and morphological changes.

References
1. Gillet LCJ, Sch€arer OD (2006) Molecular nucleotide excision repair requires XRCC1
mechanisms of mammalian global genome and DNA ligase III alpha in a cell-cycle-specific
nucleotide excision repair. Chem Rev manner. Mol Cell 27:311–323
106:253–276 7. Hong Z, Jiang J, Hashiguchi K et al (2008)
2. Katsumi S, Kobayashi N, Imoto K et al (2001) Recruitment of mismatch repair proteins to the
In situ visualization of ultraviolet-light-induced site of DNA damage in human cells. J Cell Sci
DNA damage repair in locally irradiated human 121:3146–3154
fibroblasts. J Invest Dermatol 117:1156–1161 8. Ogi T, Limsirichaikul S, Overmeer RM et al
3. Moné MJ, Volker M, Nikaido O et al (2001) (2010) Three DNA polymerases, recruited by
Local UV-induced DNA damage in cell nuclei different mechanisms, carry out NER repair
results in local transcription inhibition. EMBO synthesis in human cells. Mol Cell 37:714–727
Rep 2:1013–1017 9. Sertic S, Pizzi S, Cloney R et al (2011) Human
4. Volker M, Moné MJ, Karmakar P et al (2001) exonuclease 1 connects nucleotide excision
Sequential assembly of the nucleotide excision repair (NER) processing with checkpoint acti-
repair factors in vivo. Mol Cell 8:213–224 vation in response to UV irradiation. Proc Natl
5. Staresincic L, Fagbemi AF, Enzlin JH et al Acad Sci U S A 108:13647–13652
(2009) Coordination of dual incision and 10. Marteijn JA, Bekker-Jensen S, Mailand N et al
repair synthesis in human nucleotide excision (2009) Nucleotide excision repair-induced
repair. EMBO J 28:1111–1120 H2A ubiquitination is dependent on MDC1
6. Moser J, Kool H, Giakzidis I et al (2007) Seal- and RNF8 and reveals a universal DNA damage
ing of chromosomal DNA nicks during response. J Cell Biol 186:835–847
 Chapter 9

Inserting Site-Specific DNA Lesions into Whole Genomes

Vincent Pagès and Robert P. Fuchs

Abstract
Here, we describe a methodology that allows the insertion of site-specific DNA lesions into genomes in
living cells. The technique involves the integration of a plasmid containing a site-specific lesion engineered
in vitro into a precise location in the genome via the site-specific recombination reaction from phage
lambda. The notion of DNA lesion is not restricted to chemically modified nucleotides but also refers to
unusual DNA structures. This method will be instrumental to study qualitatively and quantitatively the
genetic consequences of site-specific lesions in vivo; moreover, it does also allow analyzing the molecular
structure of stalled replication forks at well-defined locations.

Key words DNA adducts, Unusual DNA structures, DNA damage tolerance, Site-specific recombi-
nation, Translesion synthesis, Damage avoidance, Homology-directed recombination

1 Introduction

In cells, replication-blocking lesions are tolerated via two major

pathways, mutagenic Translesion synthesis (TLS) and error-free
Damage Avoidance (DA). Currently, most molecular and genetic
studies of TLS and mutagenesis are achieved by using plasmids
probes carrying site-specific single adducts. However, it became
clear that plasmids are inappropriate tools for the study of DA
mechanisms [1, 2]. Indeed, replication fork uncoupling [3, 4]
that occurs locally at the site of a transiently blocked replication
fork leads to complete unwinding of the parental strands in plas-
mids, thus preventing DA pathways to be functional. To overcome
this difficulty we developed a method to insert a construct contain-
ing a single lesion into a whole genome by harnessing the site-
specific recombination machinery of phage lambda. The method-
ology, described in the present chapter for the insertion of a single
lesion into the E. coli genome, can be transposed to yeast or
mammalian cells. The methodology can be applied to the study of
chemical lesions (UV induced, lesions, adducts formed by chemical
carcinogens) or to sequences potentially forming unusual DNA

Marco Muzi-Falconi and Grant W. Brown (eds.), Genome Instability: Methods and Protocols, Methods in Molecular Biology,
vol. 1672, DOI 10.1007/978-1-4939-7306-4_9, © Springer Science+Business Media LLC 2018

107
108 Vincent Pagès and Robert P. Fuchs

structures (G4 quadruplex, H-DNA, cruciforms, etc.) or to short

ribonucleotides stretches embedded in DNA. This approach not
only allows the genetic consequences induced by a site-specific
lesion in a genome to be analyzed, it also allows the structure and
the dynamics of a single well-defined stalled replication fork to be
dissected at the molecular level. The genetic and molecular con-
sequences as a function of the site of stalling (close to replication
origins, to replication termini, etc.) can be studied.

2 Material

1. E. coli strain EC100D pir-116 (from Epicentre

Biotechnologies—cat# EC6P0950H).
2. LB liquid media.
3. LB agar plates.
4. SOC (0.5% yeast extract, 2% tryptone, 10 mM NaCl, 2.5 mM
KCl, 10 mM MgCl2, 10 mM MgSO4, 20 mM glucose).
5. IPTG (Isopropyl β-D-1-thiogalactopyranoside). Stock at
200 mM in H2O.
6. X-Gal (5-bromo-4-chloro-3-indolyl-β-D-galactopyranoside).
Stock at 80 mg/ml in N,N-Diméthylformamide.
7. Ampicillin. Stock at 100 mg/ml in H2O.
8. Kanamycin. Stock at 50 mg/ml in H2O.
9. Chloramphenicol. Stock at 30 mg/ml in ethanol.
10. EtOH/Na-acetate solution: dissolve 21.8 g of Sodium Acetate
(MW ¼ 136) into 20 ml of water. Complete to 1 l with 100%
Ethanol.
11. 5 GM buffer:
12. 0,5 ml 1 M TrisHCl pH 7 (50 mM final).
13. 0,1 ml 5 M NaCl (50 mM final).
14. 0,1 ml 0,5 M EDTA (5 mM final).
15. H2O complete to 10 ml.
16. Filter sterilize and store at 4  C.
17. 10 Gap Duplex Ligation Buffer: 500 μl 1 M Tris–HCl pH 7,5
(100 mM final), 500 μl 1 M MgCl2 (100 mM final), 500 μl 1 M
DTT (100 mM final), 3500 μl H2O.
18. Isopropyl-alcohol/TE/CsCl saturated: Saturate TE solution
with CsCl. Then add Isopropyl-alcohol.
19. MaxiPrep kit (Qiagen).
20. 30 ml glass tubes, type Corex 30.
21. Ultracentrifugation tube 8  35 mm (Beckman # 361082).
 Single DNA Lesions in Genomes 109

22. Electroporation apparatus: BioRad Gene Pulser.

23. Ultracentrifuge type Beckman Optima™ L-90K and rotor
NVT 65.2.
24. Ultracentrifuge type Beckman Optima™ Max-XP with rotor
TLN120.
25. Automatic serial diluter and plater EasySpiral Dilute
(Interscience).
26. Scan 1200 automatic colony counter (Interscience).

3 Methods

3.1 Construction The recipient cell must contain a unique attR integration site in
of the Recipient Cell fusion with the 30 end of lacZ gene. We use MG1655 strain in
which the original λ attB site was replaced by an artificial promoter-
less operon carrying attR fused to the 30 end of lacZ upstream of the
aadA gene (conferring Spectinomycin resistance) and between
ybhC and ybhB [5]. This locates the integration site around minute
17 of the chromosome (Fig. 1a). Two parental strains were
obtained, FBG151 where the integration of the lesion occurs on
the lagging strand, and FBG152 where lacZ is inverted to locate the
integrated lesion on the leading strand (Fig. 1c).
pVP135 plasmid expresses the xis/int operon from phage
lambda under the control of a trc promoter that has been weakened
by the mutation in the 35 and the 10 regions [6] to obtain
optimal expression level of integrase. This plasmid is transformed
into this recipient strain and maintained by growing the cells with
50 μg/ml of kanamycin.

3.2 Control Plasmid To normalize integration efficiency between the damaged vector
to Measure and the non-damaged control, we use an internal standard
Transformation (pVP146) in the electroporation experiment. pVP146 is derived
Efficiency from pACYC184 plasmid where the chloramphenicol resistance
gene has been deleted by BsaAI digestion and re-ligation. This
vector, which carries only the tetracycline resistance gene, serves
as an internal control for transformation efficiency.

3.3 Construction Two plasmids are required for the construction of the damaged
of the Damaged integrating vector by the gapped duplex method. One plasmid
Integrating Vector named “parental” plasmid contains the sequence where the
by the “Gapped lesion-containing oligonucleotide will hybridize. The other plas-
Duplex” Method [7] mid named “helper” is lacking the sequence of the oligonucleotide
allowing generating a precise gap into which the oligonucleotide
3.3.1 Description of the will be incorporated.
Parental Plasmids In the following example, we will use the combination of
plasmid pVP141 (helper) and pVP142 (parental) that allows us to
110 Vincent Pagès and Robert P. Fuchs

A. Outline of the integration system

5’-lacZ
AttL
R6K

electroporation integration
Amp®

LacZ
OriC 3’-lacZ OriC Amp®

AttR AttP AttB

xis xis
int int

chromosome pVP135 chromosome pVP135

kan® kan®
ori ori

B. Detail of the phage lambda integration system

Chromosomal DNA Damaged vector

B O P’
P O B’ CCTGCTTTTTTATACTAAGTTGG
TCAGCTTTTTTATACTAACTTG
AttL
P’1 P’2
P1 P2 AttR henP’3
3’-lacZ
5’-lacZ
R6K

Amp®

Integration product
P O P’ B O B’
TCAGCTTTTTTATACTAAGTTGG CCTGCTTTTTTATACTAACTTG
AttP AttB
P1 P2 P’1 P’2 henP’3 R6K Amp®

C. Map of the strains

FBG151 attR
ybhC 3’ lacZ aadA (Spc®) ybhB

FBG152 attR
ybhC 3’ lacZ aadA (Spc®) ybhB

Plasmid integration

FBG151: lesion integrated on the lagging strand

attP R6K attB aadA (Spc®)

ybhC Amp® 5’ lacZ 3’ lacZ ybhB

FBG151: lesion integrated on the leading strand

aadA (Spc®)
ybhC 3’ lacZ ybhB

5’ lacZ Amp® R6K

attB attP

Fig. 1 (a) Outline of the integration system: The recipient strain contains a single attR integration site in fusion
with the 30 end of lacZ gene at min 17 in the E. coli chromosome. Following ectopic expression of phage
 Single DNA Lesions in Genomes 111

probe non-frameshift TLS events at the NarI locus modified by the

G-AAF lesion [1, 2].
Plasmids pVP141 and pVP142 plasmids contain the following
characteristics:
– the bla gene that confers resistance to ampicillin thus allowing
for the selection of the integration events.
– the R6K replication origin that prevents autonomous replication
of the plasmid in the recipient strain. This plasmid can be pro-
pagated in a strain that carries the pir gene [8].
– the 50 end of lacZ gene in fusion with the attL site of
phage lambda. The P30 site of attL has been mutated (AATGAT-
TAT to AATTATTAT) to avoid excision of the plasmid once
integrated [9].

3.3.2 Gap Duplex Production of the vectors:

Assembly of the Damaged Plasmids are produced in strain EC100D pir-116 (see Note 1).
Vector Plasmids are purified using the MaxiPrep kit.
To obtain plasmid of higher purity, we further purify the DNA
by cesium chloride/Ethidium bromide equilibrium centrifugation
on Beckman Beckman Optima™ L-90K ultracentrifuge with rotor
NVT 65.2 [10].
Digestion of the vectors (Fig. 2a):
Helper plasmid (pVP141) is digested with EcoRV, whereas
parental plasmid (pVP142) is digested with ScaI.

pVP141 pVP142
50 μl pVP141 at 1 μg/μl (¼50 μg) 50 μl pVP142 at 1 μg/μl (50 μg)
40 μl NEB buffer cutsmart 40 μl NEB buffer cutsmart
4 μl EcoRV-HF (20 U/μl) 4 μl ScaI-HF (20 U/μl)
306 μl H2O (final 400 μl) 306 μl H2O (final 400 μl)

1. Incubate for 3 h at 37  C.
2. Take a sample to check that the digestion is complete on
agarose gel electrophoresis.

Fig. 1 (continued) lambda int–xis, the lesion-carrying construct is introduced by electroporation. Its attL site
will recombine with the chromosomal attR, leading to the integration of the entire lesion-containing construct.
Integration events are selected on the basis of their resistance to ampicillin. Integration at nucleotide level
resolution restores a functional lacZ gene allowing these events to be monitored on X-gal indicator plates. (b)
Detail of the phage lambda integration system: The figure shows the detail of the core sequences of the site-
specific recombination sites of phage lambda: attR, attL, attB, and attP. (c) Map of the strains before and after
integration. In strain FBG151, the lesion is integrated in the lagging strand. In strain FBG151, attR-lacZ has
been inverted to integrate the lesion in the leading strand
112 Vincent Pagès and Robert P. Fuchs

5' 3' 5'

A EcoRV GATATC GATATCACCGGCT ACCACAATC 3'
AG
3' CTATAG 5' 3' CTATAGTGGCCGATCTGGTGTTAG 5'

helper parental
(pVP141) (pVP142)

ScaI

EcoRV digestion ScaI digestion

Equimolar mix / heat danaturation (99°C) / re-annealing (55°C)

5'
B GAT ATC 3'
3' CTATAGTGGCCG GGTGTTAG 5'
AT
TC

linear homoduplexes and

“Good” gapped-duplex “wrong” gapped-duplex

AAF damaged
5'
atcaccggcgccaca3' radiolabelled
oligonucleotide

Oligonucleotide hybridization and ligation

CsCl gradient purification of the closed circular species

C
AAF
5'
GAT atcaccggcgccacaATC 3'
3' CTATAGTGGCCG GGTGTTAG 5'
AT
TC

Damaged vector

Fig. 2 Gapped duplex method to produce vectors with a single lesion. (a) Helper
and parental plasmid are linearized by a restriction endonuclease generating
blunt ends. (b) Equal amounts of the two plasmids are mixed, heat-denaturated
and allowed to re-anneal to form to gapped-duplex molecules. Two molecular
 Single DNA Lesions in Genomes 113

3. Ethanol precipitate by adding 1.2 ml of EtOH/Na-acetate

solution.
4. Spin and wash the pellet with 1 ml 70% Ethanol.
5. Spin and dry the pellet under vacuum.
6. Resuspend in 50 μl of water.
Gapped-duplex formation (Fig. 2b):
All the following steps are executed in duplicate, one being
used for the non-damaged control, the other for the lesion-
containing plasmid.
1. Mix in a 30 ml glass tube (see Note 2):

pVP141 digested EcoRV (1 μg/μl) 25 μl

pVP142 digested ScaI (1 μg/μl) 25 μl
GM buffer 5 1 ml
H2O 3.9 ml

➔ freeze a 20 μl aliquot for electrophoresis checking (STEP 0).

2. Incubate for 3.5 min at 99  C with constant and vigorous
shaking.
3. Immediately after, cool off in a mix of water/crushed ice while
shaking for 2 min.
4. Leave three additional minutes without shaking.
➔ freeze a 20 μl aliquot for electrophoresis checking (STEP 1).
5. Add 40 μl of 5 M NaCl, vortex.
6. Incubate at 55  C in a water bath.
7. After 2 h incubation, shut off the water bath and allow it to
slowly cool down to room temperature overnight.
➔ freeze a 20 μl aliquot for electrophoresis checking (STEP 2).
8. Control the efficiency of denaturation and gapped-duplex for-
mation by migration of the aliquots of Steps 0, 1, and 2 on a
1.4% agarose gel electrophoresis without ethidium bromide (to
better resolve the gapped duplex from the linear forms). After
migration, stain the gel in TAE buffer, 0.5 μg/ml ethidium
bromide (Fig. 3a).

Fig. 2 (continued) forms of gapped-duplexes will be generated, one with a gap that is complementary to the
damaged oligonucleotide (denoted “good” gapped duplex) and one that has the same sequence as the
oligonucletides (denoted “wrong” gapped duplex). (c) Ligation of the control or damaged oligonucleotide in the
gapped-duplex and isolation of the closed circular species by CsCl gradient purification allows recovering high
purity damaged vector
114 Vincent Pagès and Robert P. Fuchs

A. Formation of the gapped duplexes B. Ligation of the oligonucleotides in the gapped-duplexes

STEP 3 STEP 4

le l

le l
ro

ro
n

n
0

sio

sio
nt

nt
EP

EP

EP

co

co
ST

ST

ST
gapped-duplexes
Linear vectors linear vectors

gapped-duplexes
Denaturated plasmids 3 kb
linear vectors
2 kb
ligated gapped-duplexes

1 kb

C. Gel quantification of the constructions

range pVP141 (ng)

le l
ro

n
sio
nt
co

30 40 50 60

Fig. 3 Controlling the steps of vector production on agarose gel electrophoresis. (a) Visualization of the three
steps leading to the formation of the gapped duplexes. Step 0: linear vectors after digestion. Step 1:
denaturated plasmids. Step 2: after re-annealing, formation of ~50% of linear homoduplexes and ~50% of
gapped heteroduplexes. (b) Control of the ligation of the oligonucleotides (control and damaged) into the
gapped-duplex molecule. Successful ligation events generate circular, covalently closed constructs that
migrate faster during agarose gel electrophoresis in the presence of ethidium bromide. (c) Quality control
and quantification of the construction: a range of the helper plasmid (pVP141) allows estimating the
concentration of the construction after CsCl gradient purification. (Note that the electrophoretic migration of
the circularized duplex is slightly faster than the pVP141 plasmid. Indeed, pVP141 was produced in vivo and
therefore contains negative supercoiling, whereas the constructed duplex was ligated in vitro and contains
thus no supercoils)

Ligation of the oligonucleotides (Fig. 2c).

The 15mer oligonucleotide containing the single lesion (G-
AAF in our example) and the non-damaged control oligonucleo-
tide are phosphorylated and purified on a polyacrylamide gel
electrophoresis.
1. To the 5 ml gapped-duplex reaction (still in the glass tube), add
a 1.5 to 3-fold excess of oligonucleotide (see Note 3).
2. Add 0.5 ml of 10 Gap Duplex ligation buffer.
 Single DNA Lesions in Genomes 115

3. Incubate for 2 min at 65  C.

4. Incubate for 15 min at 16  C.
5. Add 50 μl of 0.1 M ATP and 25 μl of 10 mg/ml BSA.
➔ take a 20 μl aliquot for electrophoresis checking (STEP 3).
6. Add 2 μl of T4 DNA ligase (400 U/μl).
7. Vortex briefly and incubate for 3 min at 16  C (see Note 4).
8. Stop the reaction on ice.
➔ take a 20 μl aliquot for electrophoresis checking (STEP 4).
9. Control the ligation on a 0.8% agarose gel electrophoresis with
ethidium bromide (Fig. 3b).
10. Inactivate the ligase by incubating for 10 min at 65  C.
11. Transfer the reaction to a 50 ml falcon tube.
12. Precipitate the DNA by adding 15 ml of Ethanol/NaAcetate
solution and incubate overnight at 20  C.
Purification of the ligated gapped-duplex on CsCl gradient:
1. Recover the precipitated DNA by centrifugation for 30 min at
12,000  g.
2. Resuspend in 1.2 ml TE/CsCL/Ethidium Bromide mix.
3. Transfer to an ultracentrifugation tube 8  35 mm and seal the
tube.
4. Centrifuge for 5 h at 15,000  g rpm at 20  C on Beckman
Optima™ Max-XP ultracentrifuge in rotor TLN120.
5. After centrifugation, recover the lower band under UV light
(312 nm) using a syringe and a needle. The recovered volume is
usually around 100 μl.
Removal of the ethidium bromide:
1. Add one volume of isopropyl-alcohol/TE/CsCl saturated.
2. Vortex.
3. Eliminate the upper phase where the ethidium bromide is
dissolved.
4. Repeat two additional times.
5. Add two volumes of water to the aqueous phase.
6. Precipitate by adding two volumes (of the total volume) of
100% EtOH and incubate for 1 h at 20  C.
7. Recover the precipitated DNA by centrifugation.
8. Wash the pellet with 150 μl of 100% EtOH.
9. Dry and resuspend the pellet in TE.
116 Vincent Pagès and Robert P. Fuchs

10. Quantify and adjust the concentration of the damaged and

non-damaged plasmid by quantification on agarose gel electro-
phoresis (Fig. 3c—Note 5).

3.4 Integration 1. Inoculate 3 ml of LB, 50 μg/ml kanamycin with the recipient

Protocol strain and grow O/N at 37  C with shaking.
3.4.1 Competent Cells 2. Inoculate 100 ml of LB, 50 μg/ml kanamycin, 200 μM IPTG
with 1 ml of the O/N starter culture.
3. Grow at 37  C under shaking until OD600 ¼ 0.4.
4. Spin and wash the cells with 100 ml of ice-cold water.
5. Spin and wash the cells with 50 ml of ice-cold water.
6. Spin and wash the cells with 2 ml of ice-cold 10% glycerol.
7. Spin and resuspend the cells in 200 μl of ice-cold 10% glycerol.
8. Divide the competent cells in 40 μl aliquot and freeze at
80  C.

3.4.2 Electroporation 1. Add 1 μl of plasmid mix (gap-duplex + transformation control

plasmid pVP146) at 1 ng/μl each to the 40 μl of thawed
competent cell.
2. Transfer to an electroporation cuvette (2 mm gap) and Pulse
2.5 kV, 25 μF, 200 Ω (time constant should be ~5 ms).
3. Immediately add 1 ml of SOC, 200 μM IPTG (pre-warmed at
37  C). Mix by pipetting and transfer 500 μl to 2 ml of LB,
200 μM IPTG (pre-warmed at 37  C) (see Note 6).
4. Incubate at 37  C for 40 min with shaking.
5. Dilute and plate the cells.
Integrations events: on LB-agar, 50 μg/ml Ampicillin, 80 μg/
ml X-gal.
Transformation control: on LB-agar, 30 μg/ml
Chloramphenicol.
6. Incubate O/N at 37  C.

3.4.3 Analysis of the Blue and white colonies are counted.

Plates (Fig. 4) Blue colonies represent TLS events. The relative integration
efficiencies of damaged vectors are compared to their non-damaged
homologs, and normalized by the transformation efficiency of
pVP146 plasmid in the same electroporation experiment, allow
the overall rate of tolerance of the lesion to be measured.

3.5 Modification As discussed above, different DNA lesions can be inserted into the
of the System genome of living cells by using the gapped duplex methodology. In
addition to the lesion of choice, genetic markers (i.e., mismatches)
can be introduced upstream and/or downstream from the lesion to
 Single DNA Lesions in Genomes 117

Integration of the Integration of the

non-damaged duplex damaged duplex

Fig. 4 Colonies following integration. After the integration of the non-damaged construct, all colonies exhibit
blue/white sectors; the white and blue sectors represent the bacterial progeny that stems from the replication
of the lacZ- and lacZ+ strands of the initial construct, respectively. After the integration of the single lesion-
containing vector, only cells in which a TLS event occurred give rise to a sectored blue/white colony.
Phenotypic monitoring of DA events requires a specific construct [11]

study strand-switching/recombination reactions. Care should be

taken to identify proper mismatches that will be resistant to mis-
match repair correction in vivo. For that purpose mismatch repair
defective strains should be tested. We recently implemented a mod-
ified version of the above-described system that allow us to monitor
homologous recombination events at the lesion site. This was
achieved by the introduction of an additional mismatch upstream
the lesion site, allowing us to distinguish the damaged from the
non-damaged strand, and to evidence any exchange of genetic
information between the two strands [11].

4 Notes

1. Vectors containing the R6K origin can only be replicated in a

strain that expresses the pir gene. To increase the yield of our
plasmid prep, we use the strain EC100D pir-116, in which the
pir-116 allele supports higher copy number of R6K origin
plasmids.
2. It is important to perform all the following steps in glass tubes,
as plastic containers do not allow sufficient thermal exchange to
insure proper plasmid denaturation efficiency.
3. Expected yield: the heat denaturation/renaturation cycle will
produce both the two reconstituted linear plasmid forms (50%)
and gapped-duplexes (~50%). Two molecular forms of gapped-
duplexes will be generated with gaps of complementary
sequences (denoted as “good” and “wrong” in Fig. 2b).
118 Vincent Pagès and Robert P. Fuchs

Thus, the maximal yield of gapped duplex carrying the proper

gap will be 25%. From an initial amount of 50 μg of parental
plasmids (25 μg each), the expected yield of proper gapped-
duplex will correspond to ~7 pmol of DNA (parental plasmid
size 2.5 kb). Therefore, 10–21 pmol of oligonucleotide is
added.
4. Since only three nicks have to be sealed, a 3 min ligation at
16  C is enough. Further incubation will lead to re-
circularization of the linear forms that we want to avoid.
5. There is a slight difference in migration between the construct
and the control plasmid (Fig. 3c). This difference is due to
differences in the supercoiling status, the covalently closed
circular (ccc) construct having essentially no supercoils.
6. Recovery of the cells after electroporation in SOC only or LB
only greatly reduces the integration efficiency. For this reason,
cells are resuspended in SOC after electroporation and diluted
in LB for the incubation.

References
1. Pagès V, Mazon G, Naiman K, Philippin G, the septal ring requires its membrane anchor,
Fuchs RP (2012) Monitoring bypass of single the Z ring, FtsA, FtsQ, and FtsL. J Bacteriol
replication-blocking lesions by damage avoid- 181:508–520
ance in the Escherichia coli chromosome. 7. Koehl P, Burnouf D, Fuchs RP (1989) Con-
Nucleic Acids Res 40:9036–9043 struction of plasmids containing a unique acet-
2. Naiman K, Philippin G, Fuchs RP, Pagès V ylaminofluorene adduct located within a
(2014) Chronology in lesion tolerance gives mutation hot spot: a new probe for frameshift
priority to genetic variability. Proc Natl Acad mutagenesis. J Mol Biol 207:355–364
Sci U S A 111:5526–5531 8. Inuzuka M (1985) Plasmid-encoded initiation
3. Pagès V, Fuchs RP (2003) Uncoupling of lead- protein is required for activity at all three ori-
ing- and lagging-strand DNA replication dur- gins of plasmid R6K DNA replication in vitro.
ing lesion bypass in vivo. Science FEBS Lett 181:236–240
300:1300–1303 9. Bauer CE, Hesse SD, Gumport RI, Gardner JF
4. Higuchi K, Katayama T, Iwai S, Hidaka M, (1986) Mutational analysis of integrase arm-
Horiuchi T, Maki H (2003) Fate of DNA rep- type binding sites of bacteriophage lambda.
lication fork encountering a single DNA lesion Integration and excision involve distinct inter-
during oriC plasmid DNA replication in vitro. actions of integrase with arm-type sites. J Mol
Genes Cells 8:437–449 Biol 192:513–527
5. Esnault E, Valens M, Espéli O, Boccard F 10. Heilig, J.S., Elbing, K.L. and Brent, R. (2001)
(2007) Chromosome structuring limits Large-scale preparation of plasmid DNA. Curr
genome plasticity in Escherichia coli. PLoS Protoc Mol Biol, Chapter 1, Unit1.7
Genet 3:e226 11. Laureti L, Demol J, Fuchs RP, Pagès V (2015)
6. Weiss DS, Chen JC, Ghigo JM, Boyd D, Beck- Bacterial proliferation: keep dividing and don’t
with J (1999) Localization of FtsI (PBP3) to mind the gap. PLoS Genet 11:e1005757
 Chapter 10

A qPCR-Based Protocol to Quantify DSB Resection

Matteo Ferrari, Shyam Twayana, Federica Marini, and Achille Pellicioli

Abstract
The nucleolytic degradation of the 50 -ending strand of a Double-Strand DNA break (DSB) is necessary to
initiate homologous recombination to correctly repair the break. This process is called DNA end resection
and it is finely regulated to prevent genome rearrangements. Here, we describe a protocol to quantify DSB
resection rate by qPCR, which could be applied to every organisms whenever the break site and its flanking
region sequences are known.

Key words Double-strand breaks, DNA end resection, Single-strand DNA, qPCR

1 Introduction

DSBs can be repaired thorough the alternative processes of ends

joining (NHEJ, non homologous ends joining) or homologous-
directed recombination (HR). NHEJ consists of a ligation of two
DNA ends without needing any homology, while a homologous
sequence is used to repair the break through HR. The coordinated
actions of multiple nucleases and associated factors mediate the
nucleolytic degradation of the 50 -ending strands of a DSB (also
called DSB resection). This process generates the 30 -ended single-
strand DNA (ssDNA) filament, which is the first intermediate of all
the HR-mediated DSB repair mechanisms. When a DSB is resected,
NHEJ-mediated repair is no longer possible. Therefore, all those
mechanisms that regulate DSB resection are fundamental to pro-
mote HR and to determinate the choice between NHEJ and HR
repair events. Indeed, CDK1-mediated phosphorylation of critical
factors restricts efficient DSB resection and HR in S and G2/M cell
cycle phases [1].
In S. cerevisiae it is possible to induce a DSB through the over-
expression of the HO endonuclease, which normally cuts a specific
sequence at the MAT locus. However, in the last years, a number of
strains have been generated in which the HO-cut site has been
moved to different genomic loci, thus allowing testing different

Marco Muzi-Falconi and Grant W. Brown (eds.), Genome Instability: Methods and Protocols, Methods in Molecular Biology,
vol. 1672, DOI 10.1007/978-1-4939-7306-4_10, © Springer Science+Business Media LLC 2018

119
120 Matteo Ferrari et al.

aspects of DSB processing and repair [2]. Therefore, S. cerevisiae has

been considered an ideal model organism to study DSB response and
different methods have been described to test DSB resection in this
organism. Early methods were based on Southern blotting
approaches to visualize either the formation of the 30 ssDNA filament
or, alternatively, the disappearance of dsDNA fragments in the flank-
ing region of a DSB, using specific probes [3]. These methods are
very useful to visualize DSB resection, but are not always easy and
reproducible to measure and quantify resection rate. Therefore, a
qPCR-based method has been described, which can give an accurate
quantification of DSB resection speed [4]. A comparison between
this qPCR-based method and a Southern blotting approach to study
the processing of the HO-cut at the MAT locus can be seen in our
recent works [5, 6].
Here, we illustrate the protocol to apply the qPCR-based tech-
nique to measure DSB resection in yeast; however, thanks to its
power and versatility this method has been recently applied to other
organisms. Remarkably, the same protocol can be applied, with
minor changes, to AsiSI-induced DSBs in human genome [7, 8].

2 Materials

2.1 Yeast Cells 1. S. cerevisiae JKM139 derivative strains (J. Haber, Waltham
Growing and In Vivo University, USA).
DSB Induction 2. Yeast liquid growth media (YP).
3. 30% Raffinose stock in water, filter sterilized.
4. 30% Galactose stock in water, filter sterilized.
5. 10% Sodium Azide solution in water, filter sterilized.
6. Flasks.
7. Incubator.

2.2 Genomic DNA 1. Reagents to prepare genomic DNA.

Extraction, Digestion, 2. StyI, RsaI restriction enzymes and appropriate buffers.
and Purification
3. 2-Propanol.
4. 70% EtOH solution in water.
5. TE 1, filter sterilized (10 mM Tris–HCl pH 7.5, 1 mM
EDTA-NaOH pH 8).

2.3 qPCR Reagents 1. Real Time PCR machine.

2. 96-well plates with appropriate adhesive sealing films.
3. 2 Master Mix with SYBR Green.
4. Ultrapure water (MilliQ).
5. Customized DNA primers (see qPCR reaction step).
 Measuring DSB Resection by qPCR 121

2.4 Southern Blot Although we do not describe the Southern Blotting procedure,
Reagents materials to perform this step of the protocol are needed.

3 Methods

Following the HO break induction, genomic DNA is purified and

prepared for the analysis of end resection. The technique we
describe here relies on the ability of restriction enzymes to cut
selectively dsDNA but not ssDNA; therefore, the ssDNA generated
by DSB resection is resistant to endonucleolytic cleavage and can be
used as a template for the amplification of short DNA fragments
either close or far from the break. The rationale of the procedure is
shown in Fig. 1. The genomic DNA is either cut or mock cut with
appropriate restriction enzyme (in our example in S. cerevisiae we

A unresected HO cut B resected HO cut

mock digestion mock digestion

HO
HO HO
1 1

RsaI 2 RsaI
2

RsaI digestion RsaI digestion

(NO PCR product)

C MATa locus

HO
StyI P StyI

Cut fragment (700bp)

Uncut fragment

Fig. 1 Schematic representation of the qPCR method used to monitor DSB

resection of one HO cut induced at MATa locus in S. cerevisiae. Expected
outcomes if the DSB ends are not resected (a) or resected (b). The relative
positions of the StyI and HO sites at the MATa locus are shown in a scheme (c).
“P” is the probe used to visualize the Uncut and Cut fragments by Southern
blotting
122 Matteo Ferrari et al.

used RsaI, an enzyme that digests the yeast genome frequently,

therefore is a useful tool to study resection at many locations).
Another sample of the genomic DNA is cut with a different enzyme
(in our example we used StyI, which cuts opportunely at MATa
locus), to measure the HO cut efficiency through a Southern
Blotting procedure. Indeed, HO cut efficiency can vary among
different strains and experiments; therefore, it is necessary to test
the HO cut efficiency to normalize the results.

3.1 Cell Sample 1. Inoculate JKM139 derivative strains in YP + 3% Raffinose,

Collection grow O/N at 28  C until cells reach the exponential phase.
2. The day after, synchronize the cell culture in G2/M with
Nocodazole (see Note 1).
3. Add 2% galactose to induce HO endonuclease expression (in
the indicated strain HO cuts only in the MATa locus, see
Fig. 1).
4. Collect 50 ml of cell culture sample for each time point in
Falcon tubes. Immediately, add Sodium Azide to the sample
(0.1% final) and keep the sample on ice till the end of the
collection of all the samples to block whole cell metabolisms
and freeze all the DSB processing enzymes.
5. After the collection of all the samples, centrifuge the tubes at
around 3200  g for 5 min and eliminate the supernatant.
Keep the pellets on ice at 4  C until you are ready to process
samples (see Note 2).

3.2 Extraction 1. Prepare genomic DNA from each sample by a standard proce-
and Digestion of the dure (see Note 3).
Genomic DNA 2. Divide each genomic DNA sample (50–100 ng/μl) into 2 ali-
quots of 25 μl each in 1.5 ml tube.
3. Prepare the following mix:

Mix 1 Mix 2 (Mock digestion)

25 μl genomic DNA 25 μl genomic DNA
5 μl 10 Buffer 5 μl 10 Buffer
0.5 μl RsaI enzyme 0 μl RsaI enzyme
19.5 μl MilliQ H20 20 μl MilliQ H2O

Incubate at 37  C for 6 h.
4. Precipitate DNA with an equal volume (50 μl) of 2-Propanol
and centrifuge at 16,000 RCF for 30 min at RT.
5. Eliminate the supernatant and wash the pellet with 1 ml of cold
70% EtOH.
 Measuring DSB Resection by qPCR 123

6. Centrifuge for 5 min at 16,000 RCF, eliminate the superna-

tant, and dry the pellet in a vacuum pump centrifuge.
7. Resuspend the pellets in 100 μl TE 1.

3.3 qPCR Reaction 1. Dilute the genomic DNA in MilliQ water at a working concen-
tration of  0.1 ng/μl. Normally, we have a genomic DNA
concentration of 50–100 ng/μl; therefore, we dilute it
500–1000. We also prepare three different dilutions of the
mock-digested time 0 sample to perform the standard curve
(see Note 4).
2. We suggest using primers with a similar melting temperature
and an amplicon length of 150–250 base pairs. Design primers
at the desired distance from the DSB (primers 1 and 2; see
Fig. 1). All the amplicons should have a RsaI site. Moreover,
an amplicon on a different chromosome in which neither RsaI
nor HO are cutting is essential to normalize all the PCR values.
For this analysis you should design specific primers (3 and 4, at
the PRE1 gene locus on chromosome V in our case). See Note
5 for the sequence of all the primers used in our practical
examples.
3. Prepare the following 2 PCR reaction mixes:

CTRL MIX (as a negative control) DSB MIX

12.5 μl Master mix with SYBR 12.5 μl Master mix with SYBR
green (2) green (2)
0.5 μl Primer 3 (10 μM) 0.5 μl Primer 1 (10 μM)
0.5 μl Primer 4 (10 μM) 0.5 μl Primer 2 (10 μM)
6.5 μl MilliQ water 6.5 μl MilliQ water

4. Add 5 μl of each diluted sample in the 96-well PCR plate (we

suggest analyzing all the samples in triplicate).
5. Add 20 μl of the reaction mix in each well and seal the plate (for
Master Mix details, see Note 6).
6. Perform the real-time PCR with the following program (con-
sidering the annealing temperature of our primers we use a
two-step program):
(a) 95  C for 10 3000
(b) 95  C for 1500
(c) 66  C for 4000
(d) GO TO 2 for 49 times
(e) Melt curve 60–95  C, increment 0.5  C
124 Matteo Ferrari et al.

7. Analyze the PCR results with the specific software of your

qPCR machine. Check the linearity of the PCR reaction
performing a standard curve analysis (efficiency of the reaction
between 90 and 105%) and the specificity of the DNA amplifi-
cation by melting curve analysis.

4 Data Analysis

4.1 Example A In this section, we illustrate a practical example of how to calculate

the ssDNA accumulation at 0.15 Kb from one HO cut at the MATa
locus in wild-type yeast cells. For each PCR reaction you will obtain
a Ct value, which is the cycle in which the florescence of the sample
became higher than the background signal and then detectable by
the machine (threshold cycle). Consequently, the higher is the
amount of DNA in the sample the lower is the Ct value.
To calculate the amount of ssDNA at a defined time and
position, four Ct values are required: two obtained using primers
1 and 2, at the DSB locus, and two using primers 3 and 4, at the
control locus using mock and RsaI digested DNA in the PCR
reactions (see Fig. 1).
Use the following formula:
ΔCt digested ¼ Ct digested DSB  Ct digested CTRL.
ΔCt mock ¼ Ct mock DSB  Ct mock CTRL.
Calculate the ΔΔCt of each time point and distance from the
break as:
ΔΔCt ¼ ΔCt digestedΔCt mock
 

%ssDNA ¼ 100= 1 þ 2ΔΔCt =2 =f ¼ %DSB resected
f ¼ HO cut efficiency.
There are different procedures to measure the f value (HO cut
efficiency). Here, we describe a Southern blotting-based method.
We recommend collecting two cell samples (one at time 0 and one
within 1 h after the HO cut induction) to get a reliable f value by
this procedure.
1. Prepare genomic DNA from each sample by a standard proce-
dure (see Note 3).
2. Digest 3–5 μg of genomic DNA with StyI enzyme.
3. Load an aliquot of the StyI digested DNA and run it at 50 V for
5 h.
4. Transfer the DNA on a nylon membrane using a standard
Southern blotting protocol.
5. After hybridization with the desired probe, that has to be inside
the HO locus, quantify the signal for normalization (see Fig. 1c
for a scheme of the MATa locus and probe).
 Measuring DSB Resection by qPCR 125

A 0 1 hours after galactose addition

B Example: cut efficiency (f) calculation
densitometric value f
Uncut Uncut 1h 157701.71
Cut Cut 1h 1893839.58 0.92313013

C Example: % DSB resected calculation

Ct Digested ∆Ct Digested Ct Mock ∆Ct Mock ∆∆Ct Ratio f % of DSBs resected
DSB 0h 26.89 4.68 23.94 -1.4 6.1 67.6 0 0
DSB 3h 21.52 1.13 20.45 -0.91 2 4.11 0.92 42.38
DSB 6h 20.51 -0.18 20.94 -1.25 1.1 2.1 0.92 69.90
Ctrl 0h 22.21 25.34
Ctrl 3h 20.39 21.36
Ctrl 6h 20.69 22.19

D E
analysis at 0.15 kb
100 100
3h 6h
80 80
% DSB resected
% DSB resected

60 60

40 40

20 20

0 0
3h 6h 0.15 1.4 5.0 10.0
hours after galactose distance (Kb) from the DSB end
addiction

Fig. 2 Data analysis: Example A. The Southern blotting procedure to measure the HO cut efficiency is shown in
(a), together with the calculation of the f value in (b). The densitometry of the bands is calculated using the
ImageJ software. (c) Data obtained from qPCR reactions and calculation of the % DSB resected. “DSB”
indicates PCR reactions close to the HO cut site (primers 1 and 2); “CTRL” indicates PCR reactions obtained
using primers 3 and 4. ΔCt values and the % DSB resected are calculated using the formula shown in the text.
(d) The plot shows the mean values of three different experiments in which DSB resection has been measured
at 0.15 kb from the HO cut, 3 and 6 h after galactose addition. Error bars show the Standard Error of the Mean
(SEM). (e) Measurement of the % DSB resected at different distances from DSB end, 3 and 6 h after galactose
addition. The mean of three different experiments is plotted. Error bars indicate the SEM

6. Perform a densitometric analysis and calculate the f value as

follows:
f ¼ Cut value=ðCut value þ Uncut valueÞ ¼ HO Cut efficiency
In our example (see Fig. 2), we show the data and the graph of
the % DSB resected at 0.15 kb from the HO cut. Using the same
procedure, we also calculate resection rate at different positions far
from the break (see Fig. 2e).
126 Matteo Ferrari et al.

4.2 Example B This qPCR-based protocol has been originally described to analyze
a DSB in S. cerevisiae. However, it can be applied to every organism
whenever the break site and its flanking region sequences are
known. To give an example, in this section, we apply the protocol
to one DSB induced in chromosome I in U2OS human cell line,
stably expressing the AsiSI enzyme fused to the oestrogen receptor
(AsiSI-ER). Upon tamoxifen (4OHT) treatment, AsiSI-ER enters
into the nucleus and cuts DNA at defined positions [7]. After
treating or mock treating cells with 4OHT, genomic DNA is
extracted with a standard commercial kit. Normally, for each time
point we use 90% confluent cells grown in a well of a 6-well plate
and obtain a final concentration of DNA around 100 ng/μl. Then,
15 μl of genomic DNA is digested or not with BsrGI enzyme, as
follows:

Mix 1 Mix 2 (Mock digestion)

15 μl genomic DNA 15 μl genomic DNA
9 μl 10 Buffer 9 μl 10 Buffer
2 μl BrsGI enzyme 0 μl BrsGI enzyme
64 μl MilliQ H20 66 μl MilliQ H2O

For the qPCR reaction 3 μl (approximately 50 ng of genomic

DNA) of each mix is used. Master Mix, DSB-Mix, and CTRL-Mix
are prepared as described above. See Note 5 for the sequences of the
primers used in U2OS cells.
The AsiSI-ER cut efficiency (f) is determined by PCR, using
specific primers (9 and 10) that anneal at 30 and 50 of the cut site
(Fig. 3a).
The % DSB resected is calculated taking into consideration the f
value, with the same formula describe above in example A.
It is important to underline that in the literature, people who
use the AsiSI-ER-U2OS system often show the % ssDNA obtained
with the same procedure without normalization with the f value in
their graphs. However, contrary to the HO system in yeast, the cut
efficiency of AsiSI-ER changes significantly among the experi-
ments, at different times after the enzyme induction and at differ-
ent genomic sites. Therefore, the percentage of ssDNA could
significantly differ from the percentage of DSB resected value and
the normalization with the f value is determinant for the reliability
of the results.
 Measuring DSB Resection by qPCR 127

A
AsiSI
9 5 7 Chr I

10 BsrGI 6 BsrGI 8
335bp
1618bp
B
Example: cut efficiency (f) calculation

Ct Mock Ct Mock Ct Ratio (2^- Ct) f (1-Ratio)

Across DSB 0h 25.01 0.98 0.00 1.00 0.00
Across DSB 6h 26.09 1.56 0.59 0.67 0.33
No DSB 0h 24.03
No DSB 6h 24.53

C
Example: % ssDNA and % DSB resected calculation

Ct Digested Ct Digested Ct Mock Ct Mock Ct Ratio (2^ Ct) % of ssDNA f % DSB resected
DSB 335 0h 33.32 9.58 27.33 3.21 6.37 82.79 2.39 0.00 0.00
DSB 335 6h 32.50 7.61 28.11 3.51 4.10 17.12 11.04 0.33 33.09
No DSB 0h 23.74 24.12
No DSB 6h 24.89 24.60

D 14 analysis at 335 bp
E 50 analysis at 335 bp
12
40
% DSB resected

10
%ssDNA

8 30

6 20
4
10
2

0 0
4OHT 0h 4OHT 6h 4OHT 0h 4OHT 6h

Fig. 3 Data analysis: Example B. (a) The relative positions of AsiSI and BsrGI sites analyzed by qPCR on
Chomosome I in U2OS cells are shown in a scheme. Arrows indicate the primers used for the qPCR. Primers 9
and 10 were used to determine the cut efficiency (f). (b) Data obtained from qPCR reactions to calculate the f
value. “Across DSB” indicates Ct values obtained using primers 9 and 10, while “No DSB” indicates Ct values
achieved using primers 11 and 12 (control genomic site where AsiSI does not cut) at different time points as
indicated. The template DNA was not digested with BrsGI (Mock). ΔCt is obtained by subtracting Ct of “No
DSB” from Ct of “Across DSB”. (c) Data obtained from qPCR reactions and calculation of % ssDNA and % DSB
resected. “DSB 335” indicates the Ct values obtained using primers 5 and 6, “No DSB” indicates Ct values as
in (b), with template DNA taken from either Mix1 (BrsGI digested) or Mix2 (Mock digested) at different time
points as indicated. ΔCt values, % of ssDNA, and % of DSB resected are calculated using the formula shown
in the text. The data calculated in (c) at 335 bp from the AsiSI cut, 0 h and 6 h after of 4OHT addition were
plotted as % ssDNA (d) and % DSB resected (e). Error bars indicate Standard Error of Mean (SEM)
128 Matteo Ferrari et al.

5 Notes

1. As described in the introduction, resection is cell cycle regu-

lated; therefore, it is important to analyze it in cells blocked in a
specific cell cycle phase.
2. Here one can eventually keep cells on ice overnight.
3. There are many different protocols to extract genomic DNA
from yeast cells. Some of them are hand-made protocol, while
others are based on commercial kits. Based on our experience,
the protocol used to extract genomic DNA is not critical for the
DSB resection analysis; therefore, you can perform this step
according to your lab indication. We normally use an adapted
Teeny prep protocol derived from the original protocol by Dr.
J. Boeke (NYU). We obtain DNA at a final concentration of
50–100 ng/μl.
4. Analysis of a standard curve is required to verify that the
concentration of each sample is within the linear range of the
PCR reaction. We use three different concentrations of the
mock digested sample to test the linearity of the PCR reaction.
The results of the PCR can be considered suitable for the
resection analysis only if the efficiency of the reaction is
between 90 and 105% and the Ct value of all the samples
analyzed is encompassed in the standard curve range.
5. List of primers in yeast:
1_ CCTGGTTTTGGTTTTGTAGAGTGG;
2_GAGCAAGACGATGGGGAGTTTC;
3_CCCACAAGTCCTCTGATTTACATTCG;
4_ATTCGATTGACAGGTGCTCCCTTTTC
List of primers in U2OS human cells.

5_AATCGGATGTATGCGACTGA;
6_AAAGTTATTCCAACCCGATCC;
7_TGAGGAGGTGACATTAGAACTCAGA;
8_AGGACTCACTTACACGGCCTTT;
9_GATGTGGCCAGGGATTGG;
10_CACTCAAGCCCAACCCGT;
11_ATTGGGTATCTGCGTCTAGTGAGG;
12_GACTCAATTACATCCCTGCAGCT
6. Many companies sell Master Mix for qPCR protocols. We use a
Master Mix that contains all the reagents for qPCR, including
the SYBR green. Other qPCR protocols require TaqMan
enzyme and specific primers.
 Measuring DSB Resection by qPCR 129

Acknowledgments

This work was supported by grants from Associazione Italiana per la

Ricerca sul Cancro [AIRC_IG Grant n.15488 to A.P.; CARIPLO
[2013-0790 to A.P.]; a fellowship from Fondazione “Gabriella
Dolfin Voyasidis”-Accademia Nazionale dei Lincei to M.F.

References
1. Symington LS (2016) Mechanism and regula- tethering and repair of a double-strand break.
tion of DNA end resection in eukaryotes. Crit PLoS Genet 11:e1004928
Rev Biochem Mol Biol 51:195–212 6. Dibitetto D, Ferrari M, Rawal CC, Balint A, Kim
2. Sugawara N, Haber JE (2012) Monitoring DNA T, Zhang Z, Smolka MB, Brown GW, Marini F,
recombination initiated by HO endonuclease. Pellicioli A (2015) Slx4 and Rtt107 control
Methods Mol Biol 920:349–370 checkpoint signalling and DNA resection at
3. White CI, Haber JE (1990) Intermediates of double-strand breaks. Nucleic Acids Res 44
recombination during mating type switching in (2):669–682
Saccharomyces cerevisiae. EMBO J 9:663–673 7. Iacovoni JS, Caron P, Lassadi I, Nicolas E, Mas-
4. Zierhut C, Diffley JF (2008) Break dosage, cell sip L, Trouche D, Legube G (2010) High-
cycle stage and DNA replication influence DNA resolution profiling of gammaH2AX around
double strand break response. EMBO J DNA double strand breaks in the mammalian
27:1875–1885 genome. EMBO J 29:1446–1457
5. Ferrari M, Dibitetto D, De Gregorio G, Eapen 8. Zhou Y, Caron P, Legube G, Paull TT (2014)
VV, Rawal CC, Lazzaro F, Tsabar M, Marini F, Quantitation of DNA double-strand break
Haber JE, Pellicioli A (2015) Functional inter- resection intermediates in human cells. Nucleic
play between the 53BP1-ortholog Rad9 and the Acids Res 42:e19
Mre11 complex regulates resection, end-
 Chapter 11

Alkaline Denaturing Southern Blot Analysis to Monitor

Double-Strand Break Processing
Chiara Vittoria Colombo, Luca Menin, and Michela Clerici

Abstract
Generation of 30 single-stranded DNA (ssDNA) tails at the ends of a double-strand break (DSB) is essential
to repair the break through accurate homology-mediated repair pathways. Several methods have been
developed to measure ssDNA accumulation at a DSB in the budding yeast Saccharomyces cerevisiae. Here,
we describe one of these assays, which is based on the inability of restriction enzymes to cleave ssDNA.
Digestion of genomic DNA prepared at different time points after DSB generation leads to the formation of
ssDNA fragments whose length increases as the 50 strand degradation proceeds beyond restriction sites.
After the separation by electrophoresis on alkaline denaturing agarose gel, these ssDNA fragments can be
visualized by hybridization with an RNA probe that anneals with the 30 -undegraded DSB strand. This assay
allows a direct and comprehensive visualization of DSB end processing.

Key words DNA double-strand breaks, Resection, Single-stranded DNA, HO endonuclease, MAT
locus, Southern blot, Electrophoresis, Alkaline denaturing conditions, RNA probe

1 Introduction

DNA double-strand breaks (DSBs) can be repaired either by non-

homologous end-joining (NHEJ), which directly rejoins the two
broken ends together, or by homologous recombination (HR) that
uses intact homologous duplex DNA sequences as a template for
accurate repair. HR initiates with the nucleolytic degradation of the
50 DNA strand on both the DSB sides (a process referred to as
resection) to yield 30 single-stranded DNA (ssDNA) tails. One of
these tails invades the homologous duplex and primes reparative
DNA synthesis [1, 2]. The initiation of resection is thought to
channel DSB repair to HR and is tightly regulated [3, 4].
Much of our knowledge of the DNA end resection mechanism
and regulation comes from genetic and biochemical studies in
Saccharomyces cerevisiae, where DNA end processing of a DSB
located at a specific genomic locus can be followed by both PCR-
based and electrophoretic methods. Among these methods, a

Marco Muzi-Falconi and Grant W. Brown (eds.), Genome Instability: Methods and Protocols, Methods in Molecular Biology,
vol. 1672, DOI 10.1007/978-1-4939-7306-4_11, © Springer Science+Business Media LLC 2018

131
132 Chiara Vittoria Colombo et al.

Southern blot analysis of genomic DNA that has been run on an

alkaline denaturing agarose gel [5] allows a direct and comprehen-
sive visualization of resection. When genomic DNA from a time-
course experiment is used, the kinetics of both resection initiation
and elongation can be determined in a single gel blot. This assay is
based on the inability of many restriction enzymes to cleave ssDNA.
As resection proceeds beyond restriction sites, additional, slower
migrating bands are revealed by an RNA or ssDNA probe that
anneals to the 30 -unresected strand on one side of the break
(Fig. 1).
Here, we describe the protocol to monitor DNA end proces-
sing in haploid yeast cells where a DSB is generated at the MAT
locus by the homothallic switching endonuclease HO, whose
expression is regulated by a galactose inducible promoter [6]. The
same method can be used to study ssDNA generation at DSBs
located at different genomic loci and/or generated by different
nucleases. This method was successfully employed to monitor end
resection of both an HO-induced DSB at LEU2 locus [7] and a
Spo11-induced DSB at the YCR048W meiotic recombination hot-
spot [8]. Furthermore, a modified version of this assay that used
denaturing polyacrilamide gels allowed us to measure ssDNA accu-
mulation at telomeres [9]. Resection analysis at DSBs located in
different genomic loci requires a detailed analysis of each locus to

r7 (15.8 kb)
r6 (8.9 kb)
r5 (6.5 kb)
r4 (5.9 kb)
r3 (4.7 kb)
r2 (3.5 kb)
r1 (1.7 kb)
HO-cut (1.1 kb)

uncut (0.9 kb)

RNA probe

Chr III
S SS S S S S S
HO-cut site

Fig. 1 System to detect DSB end resection at the MAT locus. Schematic
representation of the region immediately centromere-distal to the MAT HO site
(bottom), and of the DSB and 50 -to-30 resection products (top) detectable with
the indicated RNA probe after alkaline gel electrophoresis of SspI (S)-digested
DNA. The probe reveals a 1.1 kb fragment representing the uncut MAT locus.
When HO cuts this locus, a smaller HO-cut fragment is produced. 50 -to-30
nucleolytic degradation progressively eliminates SspI sites, generating larger
ssDNA SspI fragments (r1–r7) detected by the probe
 Resection Analysis by Alkaline Gel Electrophoresis 133

determine both the most appropriate restriction enzyme and the

probe sequence for Southern blot.
Extensive resection of a DSB induced at MAT locus can be
monitored in JKM139 derivative strains [6]. In these strains the
GAL-HO fusion is stably integrated in the genome and allows the
expression of the HO endonuclease when galactose is added to the
media. As transcription from the GAL promoter is repressed by
glucose, cells are grown in raffinose-containing media before galac-
tose addition. The HO endonuclease recognizes a 24-bp cleavage
sequence, which is unique in the yeast genome and is located
adjacent to the Y region at the MAT locus. HO cleavage yields 4-
bp, 30 -overhanging ends [10]. JKM139 derivative strains also carry
the deletion of both the HML and HMR cassettes, which are
homologous to the MAT locus and are normally used to repair
the HO-induced DSB by a gene conversion event that triggers
mating type switching [10]. Therefore, HO expression in these
cells generates an irreparable DSB, whose 50 end degradation pro-
ceeds for several kilobases.
With the aim of highlighting some critical steps, here we divide
the protocol into three main parts: (1) DSB induction in yeast cells;
(2) genomic DNA extraction; (3) alkaline denaturing electropho-
resis and Southern blot analysis. Indeed, we experienced that,
beside the alkaline Southern blot, the efficient and simultaneous
DSB induction in the yeast population, as well as the preparation of
a high quality genomic DNA, is crucial to obtain a clear result with
good resection kinetics.

2 Materials

Prepare all the solutions using filtered and deionized ultrapure

water (ddH2O; resistivity 18.2 MΩ·cm at 25  C) and analytical
grade reagents. Prepare and store all solutions at room temperature
(unless indicated otherwise).

2.1 Double-Strand 1. YEPD: 2% Bacto peptone, 1% yeast extract, 2% D(+)glucose

Break Induction monohydrate, 0.005% adenine hemisulfate salt. Dissolve in
in Yeast Cells ddH2O. Autoclave.
2. YEP + raffinose: dissolve 2% Bacto peptone, 1% yeast extract,
0.005% adenine hemisulfate salt in ddH2O. Autoclave. Add
sterilized raffinose from 30% solution to 2%.
3. 30% raffinose: dissolve D(+)raffinose pentahydrate in
ddH2O. Autoclave or sterilize by filtration.
4. 30% galactose: dissolve D(+)galactose 99.0% in ddH2O.
Sterilize by filtration. Do not autoclave (see Note 1).
134 Chiara Vittoria Colombo et al.

2.2 Genomic DNA 1. Spheroplasting solution: 0.9 M sorbitol, 0.1 M ethylenediami-

Extraction netetraacetic acid (EDTA), pH 7.5.
2. Zymolyase solution: dissolve 2 mg/mL Zymolyase 20T® (see
Note 2) from Arthrobacter luteus (Nacalai Tesque) in sphero-
plasting solution + 14 mM β–mercaptoethanol.
3. 1 TE: 10 mM Tris–HCl, pH 7.5, 1 mM EDTA, pH 7.4.
Autoclave.
4. Lysis solution: 2.2% sodium dodecyl sulfate (SDS), 278 mM
EDTA, pH 8.5, 445 mM Tris-base. Prepare the lysis solution
just before adding it to the samples by mixing the appropriate
amounts of 10% SDS, 0.5 M EDTA, pH 8.5, 2 M Tris-base
stock solutions in a 15 mL tube.
5. 5 M potassium acetate: dissolve 5 M potassium acetate in
ddH2O. Autoclave.
6. Ice-cold ethanol: 96% and 70%. Prepare a 70% ethanol solution
by diluting 96% ethanol in ddH2O. Store at 20  C aliquots of
both 70% and 96% ethanol in glass bottles.
7. RNase solution: dissolve 10 mg/mL RNase A, DNase-free, in
10 mL 10 mM Tris–HCl, pH 7.5, 15 mM NaCl. Heat to
100  C for 5 min and cool down at room temperature. Prepare
small aliquots and store them at 20  C.
8. 2–propanol anhydrous 99.5%.
9. 6 DNA loading buffer: 30% glycerol, 0.25% bromophenol
blue.
10. Regular agarose gel: melt 0.8% agarose in 1 TAE buffer. Cool
at approximately 60  C and add 10 mg/mL ethidium bromide
solution to a final concentration of 1 μg/mL. Pour the gel into
a gel tank and insert a comb.
11. 1 TAE (Tris-acetate-EDTA) buffer: 40 mM Tris, 20 mM
acetic acid, 1 mM EDTA. Prepare a 50 TAE buffer: for 1 L
dissolve 242 g of Tris-base in approximately 600 mL ddH2O.
Add 57.1 mL glacial acetic acid and 100 mL 0.5 M EDTA,
pH 8.0, and bring final volume to 1 L with ddH2O. Autoclave.
Before use, dilute in ddH2O to a final concentration of 1.
12. Ethidium bromide solution: prepare a stock of 10 mg/mL
ethidium bromide in ddH2O and store in light-tight containers
(see Note 3).
13. UV lamp with camera.

2.3 Alkaline 1. SspI restriction enzyme (20,000 U/mL; New England Bio-
Denaturing labs) and buffer supplied from distributor.
Electrophoresis 2. 3 M sodium acetate, pH 5.2: dissolve 3 M sodium acetate in
and Southern Blot ddH2O. Adjust the pH to 5.2 with glacial acetic acid.
2.3.1 DNA Digestion
Autoclave.
and Denaturation
 Resection Analysis by Alkaline Gel Electrophoresis 135

3. 0.5 M EDTA, pH 8.0: dissolve 0.5 M EDTA in ddH2O. Adjust

the pH to 8.0 with sodium hydroxide (NaOH). Autoclave.
4. 1 alkaline loading buffer: 50 mM NaOH, 1 mM EDTA,
pH 8.0, 2.5% ficoll (type 400) in ddH2O, 0.025% bromophe-
nol blue.

2.3.2 Alkaline Denaturing 1. Horizontal electrophoresis system with a large gel running
Gel Electrophoresis chamber (gel size 25  20 cm) and 32-tooth comb (thickness
and Transfer 1.0 mm and width of teeth 4.0 mm).
2. 1 alkaline electrophoresis buffer: 50 mM NaOH, 1 mM
EDTA, pH 8.5.
3. Glass plate that fits the gel.
4. 0.25 N hydrochloric acid (HCl): dilute HCl in ddH2O just
before use.
5. 0.5 N NaOH, 1.5 M NaCl: dissolve in ddH2O just before use.
6. Nylon hybridization transfer membrane (GeneScreen® from
Perkinelmer or equivalent).
7. 20 SSC buffer: 3 M NaCl, 300 mM sodium citrate. Adjust
the pH to 7.0 with HCl. Autoclave.
8. Whatman 3 MM paper.
9. Parafilm from Bemis NA or equivalent.
10. Paper towels.
11. Neutralization solution: 0.5 M Tris–HCl pH 7.5, 1 M NaCl.
12. UV crosslinker.

2.3.3 Probe Labeling 1. Plasmid pML514 (available upon request), carrying part of the
MAT locus downstream to the T7 bacteriophage promoter.
Plasmid pML514 was constructed by inserting in the
pGEM®–7Zf(+/) (purchased from Promega) EcoRI site a
900–bp fragment of the MAT locus, obtained by PCR using
yeast genomic DNA as a template and PRP643 (50 –CGG AAT
TCC CTG GTT TTG GTT TTG TAG AGT GG–30 ) and
PRP644 (50 –CGG AAT TCG AAA CAC CAA GGG AGA
GAA GAC–30 ) as primers.
2. BamHI restriction enzyme (20,000 U/mL; New England
Biolabs) and buffer supplied from distributor or equivalent
(see Note 4).
3. In vitro transcription system Riboprobe System–T7 (Purchased
from Promega and containing recombinant RNasin® RNase
inhibitor, 10 mM rATP, 10 mM rCTP, 10 mM rGTP, 10 mM
rUTP, 100 mM dithiothreitol (DTT), 5 transcription opti-
mized buffer, T7 RNA polymerase, RQ1 RNase-free DNase,
nuclease-free water, pGEM® Express positive control tem-
plate), or equivalent.
136 Chiara Vittoria Colombo et al.

4. rUTP–α32P. Specific activity: 800 Ci/mmol.

5. Sephadex G–50 grade chromatography-columns (GE–Health-
care) or equivalent.

2.3.4 Filter Hybridization 1. Hybridization oven and hybridization tubes.

2. Formamide hybridization buffer: 5 SSPE, 50% formamide,
6% dextran sulfate sodium salt, 4 Denhardt’s solution (dilute
50 Denhardt’s solution from Sigma-Aldrich or equivalent),
100 μg/mL deoxyribonucleic acid sodium salt from salmon
testes, 200 μg/mL yeast tRNA.
3. 20 SSPE buffer: 3 M NaCl, 0.2 M NaH2PO4, 20 mM EDTA.
Adjust the pH to 7.4 with NaOH. Autoclave.
4. 1 SSPE, 0.1% SDS.
5. 0.1 SSPE, 0.1% SDS.
6. 0.2 SSPE, 0.1% SDS.
7. Autoradiography cassette with intensifying screens.
8. Autoradiography films.

3 Methods

3.1 Double-Strand 1. Inoculate cells in 50–100 mL YEPD medium.

Break Induction 2. Grow the cell culture 6–8 h at 26  C until the exponentially
in Yeast Cells growing cells reach a density of 104–106 cells/mL.
3. Spin down the cells and wash with YEP + raffinose medium to
remove glucose. Resuspend the cells in an equal or larger
volume (depending on the number of samples to collect after
HO induction) of YEP + raffinose medium (see Notes 5 and 6).
4. Grow the cell culture overnight at 26  C.
5. When the cell culture has grown to a density between 6  106
and 107 cells/mL draw a 50 mL sample for the uninduced
control. Then add galactose from 30% solution to 3% final
concentration to the remaining cell culture (see Note 7).
6. Grow the cell culture at 26  C and take 50 mL samples at the
desired time points after galactose addition. During the experi-
ment the efficiency of DSB formation can be determined
(see Note 8).

3.2 Genomic DNA 1. Pellet the cells by spinning 3 min at 1600  g in 50 mL tubes.
Extraction 2. Wash the cells in 1 mL spheroplasting solution and transfer the
samples to 1.5 mL microcentrifuge tubes.
3. Spin 3 min at 1600  g and completely remove the supernatant
with a tip.
 Resection Analysis by Alkaline Gel Electrophoresis 137

4. Freeze and store the cell pellets at 20  C.

5. Thaw the samples at room temperature and resuspend the cell
pellets in 400 μL spheroplasting solution + 14 mM
β–mercaptoethanol.
6. Add 100 μL Zymolyase solution to each sample and invert the
tube four to six times. Incubate the samples at 37  C. After
30 min check the formation of spheroplasts under a light
microscope (see Note 9).
7. When >95% cells have become spheroplasts, spin 30 s at
15,000  g and carefully remove the supernatant with a tip.
8. Gently resuspend spheroplasts in 400 μL 1 TE (do not
vortex).
9. Add 90 μL lysis solution (prepared just before use). Immedi-
ately shake vigorously the tube (a foam should form) and
incubate the samples 30 min in a water-bath at 65  C.
10. Add 80 μL 5 M potassium acetate and mix by inverting the
tube several times. Place the tubes on ice for approximately 1 h.
11. Spin for 20 min at 15,000  g at 4  C. Transfer the supernatant
with nucleic acids to new 1.5 mL tubes. Discard the pellets.
12. Add 1 mL ice-cold 96% ethanol and mix by inverting the tube
several times. A white cloudy precipitate should form. Place
30 min at 80  C to facilitate precipitation.
13. Spin 10 min at 15,000  g at 4  C and remove the supernatant.
14. Wash the pellet with 1 mL ice-cold 70% ethanol and immedi-
ately discard the ethanol.
15. Air-dry the pellet until it appears glassy.
16. Add 500 μL 1 TE. Let tubes sit for 30 min at room tempera-
ture, then gently dissolve the pellet (do not vortex).
17. When all the pellets are completely dissolved and the solutions
appear clear and transparent add 2.5 μL RNase solution to each
sample and incubate for 1 h at 37  C.
18. Add 500 μL 2–propanol and invert the tube several times. A
white DNA “clew” should form. Place 30 min–overnight at
80  C to facilitate precipitation.
19. Spin for 15–30 min at 15,000  g at 4  C and remove the
supernatant.
20. Wash the pellet with 1 mL ice-cold 70% ethanol and immedi-
ately discard the ethanol.
21. Air-dry until the pellet appears glassy.
22. Add 30 μL 1 TE. Let tubes sit for 30 min at room tempera-
ture, then gently dissolve the DNA pellet (do not vortex).
138 Chiara Vittoria Colombo et al.

a Minutes after HO induction b Minutes after HO induction c Minutes after HO induction

120
150
180
210

120
150
180
210

120
150
180
210
30
60
90

30
60
90

30
60
90
0

0
Fig. 2 Evaluation of the DNA quality during sample DNA preparation. (a–c) JKM139 cells exponentially growing
in YEP + raffinose were transferred in YEP + raffinose + galactose to induce HO expression and DSB formation
(time zero). Genomic DNA was prepared from samples taken at the indicated time points after galactose
addition. (a) DNA extraction. 1 μL from 30 μL genomic DNA was visualized on regular agarose gel with
ethidium bromide. (b) DNA digestion. Genomic DNA was digested with SspI. After 5 h at 37  C, 2 μL of each
digestion reaction were analyzed on agarose gel with ethidium bromide. (c) DNA denaturation. SspI-digested
genomic DNAs were dissolved in 18 μL alkaline loading buffer. 1 μL of each sample was visualized on agarose
gel with ethidium bromide

23. Apply 1 μL of each genomic DNA sample (added to 10 μL of

1 DNA loading buffer) on a 0.8% regular agarose gel with
ethidium bromide and run in 1 TAE buffer. Check the qual-
ity of the extracted DNA under an UV lamp (see Note 10 and
Fig. 2a).

3.3 Alkaline 1. Digest each DNA sample (10–15 μg DNA) with 10 U of SspI
Denaturing (New England Biolabs) or other restriction enzymes that cut
Electrophoresis double-stranded DNA (dsDNA) but not ssDNA (see Notes 11
and Southern Blot and 12). Digest with 1 enzyme buffer in a total volume of
70 μL for 5–6 h at 37  C.
3.3.1 DNA Digestion
and Denaturation
2. Test 2 μL of each digestion reaction on a 0.8% agarose gel with
ethidium bromide and run in 1 TAE buffer to check that all
samples are digested (Fig. 2b).
3. Precipitate the digested DNA with 2 volumes 96% ethanol,
5 mM EDTA, pH 8.5, 0.3 M sodium acetate pH 5.2. Place
overnight at 80  C to facilitate precipitation.
4. Spin for 30 min at 15,000  g at 4  C and remove the
supernatant.
5. Wash the pellet with 1 mL ice-cold 70% ethanol and immedi-
ately discard the ethanol.
6. Air-dry until the pellet appears glassy.
7. Add 18 μL 1 alkaline loading buffer. Let tubes sit for 30 min
at room temperature, then gently dissolve the pellet (do not
vortex). Let tubes sit for additional 1.5–2 h at room tempera-
ture by gently mixing every 15–30 min. The DNA should
denature in single-stranded filaments. A 1 kb DNA ladder can
 Resection Analysis by Alkaline Gel Electrophoresis 139

also be treated with 1 alkaline loading buffer and then used as

marker for the size estimation of the DNA resection fragments.
8. Apply 1 μL denatured DNA to 10 μL alkaline loading buffer
and load on a 0.8% agarose gel with ethidium bromide. Run in
1 TAE buffer to check whether the DNA is completely
denatured (see Note 13 and Fig. 2c) and to evaluate the
amount of DNA in each digestion. Equilibrate the DNA con-
centration in the different tubes by adding an appropriate
amount of 1 alkaline loading buffer.

3.3.2 Alkaline Denaturing 1. Melt 0.8% agarose in 450 mL ddH2O, and pour into a gel tray.
Gel Electrophoresis and 2. When the gel is completely solidified, mount it in a large gel
Transfer box and submerge the gel in 1 alkaline electrophoresis buffer.
Allow the gel to equilibrate for 30 min or longer (see Note 14).
3. Load on the gel 15 μL of each sample dissolved and equili-
brated in alkaline loading buffer.
4. Carry out electrophoresis by running the gel overnight at
voltages <3 V/cm. As bromophenol blue diffuses rapidly out
of the gel into the alkaline electrophoresis buffer, place a glass
plate directly on the top of the gel after the dye has migrated
out of the loading slots.
5. After the DNA has migrated far enough (the dye has to migrate
approximately 13–14 cm from the loading slots), remove the
gel from the tank.
6. Stain the DNA with ethidium bromide by soaking the gel for
30 min–1 h in 1 TAE buffer with 0.5 μg/mL ethidium
bromide (see Note 15).
7. Wash the gel 10 min in ddH2O.
8. Check the gel under an UV lamp (Fig. 3a).
9. Soak the gel 7 min with gentle agitation in 0.25 N HCl.
10. Rinse the gel with ddH2O.
11. Soak the gel 30 min with gentle agitation in 0.5 N NaOH,
1.5 M NaCl.
12. Rinse the gel with ddH2O.
13. Blot overnight the DNA from the gel to a nylon hybridization
membrane by capillary transfer with 10 SSC buffer (see Notes
16 and 17). Fill a tray with 10 SSC buffer (approximately
1.5 L). Place a platform on the tray and create a bridge with a
Whatman 3 MM paper onto the platform. Pour 10 SSC
buffer over the bridge and remove air bubbles. Place the gel
on the Whatman 3 MM paper bridge. Wet a membrane that fits
with the size of the gel in 10 SSC and lay on the gel. Remove
air bubbles. Wet three sheets of Whatman 3 MM paper in 10
SSC and lay on the membrane. To maintain capillary flow
140 Chiara Vittoria Colombo et al.

Minutes after Minutes after

a HO induction b HO induction

120
150
180
210

120
150
180
210
30
60
90

30
60
90
0

0
r7

r6

r5
r4

r3

r2

r1

uncut

HO-cut

Fig. 3 DSB end resection at MAT locus. (a) Alkaline denaturing gel electrophore-
sis. SspI-cut genomic DNA was run overnight at 1.4 V/cm on an alkaline
denaturing agarose gel and stained 30 min in 1 TAE buffer with 0.5 μg/mL
ethidium bromide. (b) Southern blot analysis. ssDNA fragments were transferred
on a nylon membrane and hybridized with a MAT RNA probe. Before galactose
addition the probe reveals the uncut band, which is converted in the smaller HO-
cut band after DSB formation. The resection bands r1–r7 appear when resection
proceeds and eliminates SspI sites, as depicted in Fig. 1

through the gel instead of around it, place strips of Parafilm on

all the sides of the gel and create a barrier between the bridge
and the paper towels. Build up a big stack of paper towels and
add some weight on the top.
14. After transfer, soak the membrane 15 min in neutralization
solution with gentle agitation.
15. Air-dry the filter.
16. Crosslink the DNA on the membrane with an UV crosslinker
by following the instructions of the manufacturer.

3.3.3 Probe Labeling 1. Digest 5–10 μg pML514 DNA with BamHI at 37  C to

linearize the plasmid downstream the MAT insert.
 Resection Analysis by Alkaline Gel Electrophoresis 141

2. Check whether the plasmid is completely linearized by loading

2 μL of the digestion reaction and 1 μL of the uncut plasmid on
a 0.8% agarose gel with ethidium bromide (see Note 18).
3. Precipitate the linearized plasmid and suspend it in a small
volume of nuclease-free water (see Note 19).
4. Prepare the labeled RNA probe with the Promega Riboprobe
System-T7 kit or equivalent, according to the instructions of
the manufacturer. If the Promega Riboprobe System-T7 is
chosen, add the following components in the order listed.
4 μL 5 transcription optimized buffer; 2 μL 100 mM DTT,
20 U recombinant RNasin® RNase inhibitor, 1 μL 10 mM
rATP, 1 μL 10 mM rGTP, 1 μL 10 mM rCTP, 1–4 μL linearized
template DNA (0.2–1.0 μg/μL in nuclease-free water),
nuclease-free water to a 20 μL final volume, 5 μL rUTP–α32P
(100 μCi), 15–20 U T7 RNA Polymerase.
5. Incubate for 45 min at 40  C.
6. Add 1 U RQ1 RNase-free DNase.
7. Incubate for 15 min at 37  C.
8. Add 180 μL nuclease-free water.
9. Equilibrate a Sephadex G–50 grade chromatography-column
with nuclease-free water.
10. Load the in vitro transcription solution on the chromatogra-
phy-column.
11. Spin for 1 min at 850  g and recover the eluate.

3.3.4 Filter Hybridization 1. Insert the filter in a hybridization tube and soak the filter with
ddH2O.
2. Block the filter by incubating for 5 h in a hybridization oven at
42  C in 25 mL formamide hybridization buffer (pre-
hybridization).
3. Prepare the hybridization solution by adding the RNA probe
obtained by in vitro transcription (almost 200 μL; see Subhead-
ing 3.3.3) to 25 mL of fresh formamide hybridization buffer.
Replace the pre-hybridization solution with this hybridization
solution.
4. Incubate overnight at 42  C by gently rotating in the hybridi-
zation oven.
5. Remove the hybridization solution.
6. Wash the filter for 30 min at 42  C in 5 SSPE.
7. Wash for 30 min at 42  C in 1 SSPE, 0.1% SDS.
8. Wash for 30 min at 42  C in 0.1 SSPE, 0.1% SDS.
9. Wash for 15 min at 68  C in 0.2 SSPE, 0.1% SDS.
142 Chiara Vittoria Colombo et al.

10. Wash for 5 min at room temperature in 0.2 SSPE.

11. Air-dry the filter.
12. Expose the filter to an autoradiography film in an autoradi-
ography cassette with intensifying screens (see Notes 20 and 21
and Fig. 3b).

4 Notes

1. Galactose should not be autoclaved because it isomerizes at

elevated temperatures.
2. There are two preparations of Zymolyase from Nacai Tesque,
Zymolyase®–20T, and Zymolyase®–100T, having lytic activity
of 20 U/g and 100 U/g, respectively. Zymolyase–100T can be
used in place of Zymolyase–20T but the Zymolyase concentra-
tion in the spheroplasting solution and/or the time of incuba-
tion should be adjusted according to the different activity of
the enzyme.
3. Alternatives to ethidium bromide are less-toxic dyes such as
GelRed™ or GelGreen™.
4. Templates used for in vitro transcription should either be
blunt-ended or carry protruding 50 termini. As extraneous
transcripts can appear in addition to the expected transcript
when templates contain 30 overhangs [11], plasmids should not
be linearized with any enzyme that leaves a 30 overhang.
5. Be sure to transfer in YEP + raffinose medium enough cells to
reach a density of at least 6  106 cells/mL the next day.
6. HO expression and DSB formation can also be induced in a
strain carrying a plasmid. In this case, an appropriate selective
medium should be used in place of YEPD to maintain the
plasmid selection. However, cells grow poorly in selective
media with raffinose as a carbon source. Thus, cells can be
grown in glucose selective medium for approximately 8 h and
then transferred in YEP + raffinose medium for the overnight
growth to minimize the number of cell divisions in a rich
medium.
7. As resection is stimulated by the activity of the cyclin-
dependent kinases (Cdk1 in S. cerevisiae) and the kinetics of
resection change in the different cell cycle phases [12, 13],
sometimes it might be important to analyze the generation of
ssDNA at DSB ends in cells arrested at a particular stage of the
cell cycle. To monitor resection in G2/M phase, when Cdk1
activity is high, cells can be arrested in early metaphase before
galactose addition by treatment with nocodazole, a drug that
interferes with microtubule polymerization. Prepare a 100
 Resection Analysis by Alkaline Gel Electrophoresis 143

nocodazole stock by dissolving 1.5 mg/mL nocodazole in

100% dimethyl sulfoxide (DMSO). Add the nocodazole stock
to the exponentially growing cells in YEP + raffinose medium
to a final concentration of 1 (final concentration of nocoda-
zole 15 μg/mL). After 2.5–3 h at 26  C, visualize cells under a
light microscope. Add galactose to induce HO in G2/M-
arrested cells when 90–100% cells accumulate as large-budded
cells. Nocodazole-mediated arrest can be maintained for
8–10 h [7]. To monitor resection in G1 phase, when Cdk1
activity is low, MATa haploid cells can be treated with the α-
factor pheromone. As these cells recover from the α-factor
block after 1.5–2 h due to α-factor degradation by the Bar1
protease, a persistent G1 arrest with α-factor can be induced in
strains carrying the deletion of the BAR1 gene [14]. Dissolve
1 mg/mL α–factor in ddH2O and add this solution to the
exponentially growing cells in YEP + raffinose to a final α-factor
concentration of 0.5 μg/mL. After 2 h at 26  C, visualize cells
under a light microscope. Add galactose to induce HO in G1-
arrested cells when 90–100% cells are unbudded. α-factor-
mediated arrest in bar1Δ cells can be maintained for at least
8 h [15].
8. As a single DSB formation triggers the activation of the DNA
damage checkpoint and a G2/M arrest [16], the effectiveness
of DSB induction in checkpoint-proficient strains can be deter-
mined under a light microscope by evaluating the percentage of
large-budded cells after 4–6 h of galactose induction. Alterna-
tively, samples taken before HO induction and at the desired
time points after galactose addition (2 and/or 4 h) can be
diluted and 300–3000 cells can be plated on YEPD plates to
test if galactose addition has efficiently induced the DSB for-
mation. As in this strain the HO-induced DSB is irreparable
and its generation leads to cell death [6], strain viability after
HO induction can be used to determine the efficiency of DSB
formation.
9. Spheroplasts display a rounded morphology in contrast to the
ovoid shape of intact yeast cells. Furthermore, as spheroplasts
lyse in SDS solution, cells should disappear after the addition of
one drop of 10% SDS to 3 μL of cells on a glass slide. We usually
obtain >95% spheroplasts in 40–50 min.
10. If high molecular weight DNA molecules were extracted and
completely dissolved, a single band should be detectable in the
gel without smears or signals retained in the well.
11. Restriction endonucleases digest dsDNA by cleaving two phos-
phodiester bonds, one within each strand of the duplex DNA.
ssDNA generally cannot be cut by restriction enzymes. A few
restriction enzymes cleave ssDNA, although usually at low
efficiency.
144 Chiara Vittoria Colombo et al.

12. The conditions described here were set up for digestion with
SspI restriction enzyme purchased from New England Biolabs.
If a different enzyme and/or producer were chosen, consult
the manufacturer’s instructions for optimal digestion
conditions.
13. When DNA is completely denatured, a smear is detectable on
the agarose gel.
14. The agarose gel is equilibrated in alkaline electrophoresis
buffer after solidification because the addition of NaOH to a
warm agarose solution causes polysaccharide hydrolysis. Alter-
natively, the agarose can be melted in ddH2O and then cooled
down to 60  C, so that NaOH to 50 mM and EDTA pH 8.5 to
1 mM can be added just before pouring the gel.
15. Ethidium bromide is omitted from alkaline agarose gels
because it does not bind to DNA at high pH. DNA can be
stained with ethidium bromide after the electrophoresis. How-
ever, DNA will be faint because the ethidium bromide does not
bind very well to ssDNA.
16. Alternatively, a vacuum blotter can be used.
17. DNA can be blotted onto either neutral or positively charged
nylon membranes.
18. Plasmid DNA must be cleaved to completion, as trace amounts
of uncut supercoiled plasmid DNA can give rise to long tran-
scripts that include vector sequences. These transcripts may
incorporate a fraction of the radiolabeled rNTP.
19. The linearized plasmid DNA should be highly concentrated
because a small volume of template is required for the in vitro
transcription reaction.
20. The film can be also developed in a Typhoon instrument or
equivalent.
21. With freshly labeled rUTP-α32P, we usually obtain a good
signal after 4 h-overnight exposure at 80  C.

Acknowledgments

We thank J. Haber (Brandeis University) for the JKM139 yeast

strain, and M. P. Longhese and G. Lucchini for critical reading of
the manuscript.

References
1. Symington LS, Rothstein R, Lisby M (2014) 2. Mehta A, Haber JE (2014) Sources of DNA
Mechanisms and regulation of mitotic recom- double-strand breaks and models of recombi-
bination in Saccharomyces cerevisiae. Genetics national DNA repair. Cold Spring Harb Per-
198:795–835. doi:10.1534/genetics.114. spect Biol 6(9):a016428. doi:10.1101/
166140 cshperspect.a016428
 Resection Analysis by Alkaline Gel Electrophoresis 145

3. Symington LS (2014) End resection at double- 10. Lee CS, Haber JE (2015) Mating-type gene
strand breaks: mechanism and regulation. Cold switching in Saccharomyces cerevisiae. Micro-
Spring Harb Perspect Biol 6(8):a016436. biol Spectr 3(2):MDNA3-0013-2014. doi:10.
doi:10.1101/cshperspect.a016436 1128/microbiolspec.MDNA3-0013-2014
4. Villa M, Cassani C, Gobbini E, Bonetti D, 11. Schenborn ET, Mierendorf RC (1985) A novel
Longhese MP (2016) Coupling end resection transcription property of SP6 and T7 RNA
with the checkpoint response at DNA double- polymerases: dependence on template struc-
strand breaks. Cell Mol Life Sci 73 ture. Nucleic Acids Res 13:6223–6236
(19):3655–3663. (in press) 12. Aylon Y, Liefshitz B, Kupiec M (2004) The
5. White CI, Haber JE (1990) Intermediates of CDK regulates repair of double-strand breaks
recombination during mating type switching in by homologous recombination during the cell
Saccharomyces cerevisiae. EMBO J 9:663–673 cycle. EMBO J 23:4868–4875
6. Lee SE, Moore JK, Holmes A, Umezu K, 13. Ira G, Pellicioli A, Balijja A, Wang X, Fiorani S,
Kolodner RD, Haber JE (1998) Saccharomyces Carotenuto W, Liberi G, Bressan D, Wan L,
Ku70, mre11/rad50 and RPA proteins regu- Hollingsworth NM, Haber JE, Foiani M
late adaptation to G2/M arrest after DNA (2004) DNA end resection, homologous
damage. Cell 94:399–409. doi:10.1016/ recombination and DNA damage checkpoint
s0092-8674(00)81482-8 activation require CDK1. Nature
7. Clerici M, Mantiero D, Lucchini G, Longhese 431:1011–1017
MP (2005) The Saccharomyces cerevisiae Sae2 14. Chan RK, Otte CA (1982) Physiological char-
protein promotes resection and bridging of acterization of Saccharomyces cerevisiae mutants
double strand break ends. J Biol Chem supersensitive to G1 arrest by a factor and alpha
280:38631–38638 factor pheromones. Mol Cell Biol 2:21–29
8. Manfrini N, Guerini I, Citterio A, Lucchini G, 15. Trovesi C, Falcettoni M, Lucchini G, Clerici
Longhese MP (2010) Processing of meiotic M, Longhese MP (2011) Distinct Cdk1
DNA double strand breaks requires cyclin- requirements during single-strand annealing,
dependent kinase and multiple nucleases. J noncrossover, and crossover recombination.
Biol Chem 285:11628–11637. doi:10.1074/ PLoS Genet 7(8):e1002263. doi:10.1371/
jbc.M110.104083 journal.pgen.1002263
9. Bonetti D, Martina M, Clerici M, Lucchini G, 16. Pellicioli A, Lee SE, Lucca C, Foiani M, Haber
Longhese MP (2009) Multiple pathways regu- JE (2001) Regulation of Saccharomyces Rad53
late 30 overhang generation at S. cerevisiae tel- checkpoint kinase during adaptation from
omeres. Mol Cell 35:70–81. doi:10.1016/j. DNA damage-induced G2/M arrest. Mol
molcel.2009.05.015 Cell 7:293–300
 Chapter 12

Single Molecule Analysis of Resection Tracks

Pablo Huertas and Andrés Cruz-Garcı́a

Abstract
Homologous recombination is initiated by the so-called DNA end resection, the 50 –30 nucleolytic degra-
dation of a single strand of the DNA at each side of the break. The presence of resected DNA is an
obligatory step for homologous recombination. Moreover, the amount of resected DNA modulates the
prevalence of different recombination pathways. In different model organisms, there are several published
ways to visualize and measure with more or less detail the amount of DNA resected. In human cells,
however, technical constraints hampered the study of resection at high resolution. Some information might
be gathered from the study of endonuclease-created DSBs, in which the resection of breaks at known sites
can be followed by PCR or ChIP. In this chapter, we describe in detail a novel assay to study DNA end
resection in breaks located on unknown positions. Here, we use ionizing radiation to induce double-strand
breaks, but the same approach can be used to monitor resection induced by different DNA damaging
agents. By modifying the DNA-combing technique, used for high-resolution replication analyses, we can
measure resection progression at the level of individual DNA fibers. Thus, we named the method Single
Molecule Analysis of Resection Tracks (SMART). We use human cells in culture as a model system, but in
principle the same approach would be feasible to any model organism adjusting accordingly the DNA
isolation part of the protocol.

Key words DNA resection, High-resolution resection assay, SMART, DNA combing, Fiber assay

1 Introduction

The key event that controls the choice between different pathways
to repair DNA Double-Strand Break (DSB) is DNA-end resection.
This mechanism consists of a 50 -to-30 degradation of one strand at
each side of the break, yielding a long stretch of 30 -ended single-
stranded DNA (ssDNA). Whereas non-homologous end-joining
(NHEJ) is inhibited by DNA-end resection, the DNA-end once
resected is committed to being repaired by either homologous
recombination (HR) or microhomology-mediated end-joining
(MMEJ) [1]. Once the break end is resected, the ssDNA tail that
is generated is rapidly coated by RPA (replication protein A) [1], a
heterotrimeric protein that prevents the formation of secondary
structures and protects against degradation of the ssDNA [1].

Marco Muzi-Falconi and Grant W. Brown (eds.), Genome Instability: Methods and Protocols, Methods in Molecular Biology,
vol. 1672, DOI 10.1007/978-1-4939-7306-4_12, © Springer Science+Business Media LLC 2018

147
148 Pablo Huertas and Andrés Cruz-Garcı́a

DNA end resection has been studied by different methods

depending on if the location of the DSB was known. For those
DSB created randomly by DNA damaging agents such as ionizing
radiation or several drugs, resection has been measured by the
exposure of ssDNA-containing BrdU [2], by proxy using RPA
accumulation (either by immunofluorescence or FACS; [2, 3]),
by posttranslational modification of RPA [2] or modification of
the apparent size of a circular chromosome [4]. Such studies can
monitor large changes in resection, mainly in the number of breaks
that are resected, but they all share a common problem: they are
low-resolution techniques that cannot reflect changes on the length
or speed of DNA resection. An alternative is to use breaks created at
established position, like the ones created with nucleases such as
HO, I-SceI, or AsiSI [4–10]. In these cases, high-resolution tech-
niques based on chromatin IP, quantitative PCR, Southern blots,
or sequencing are available [4–10]. However, with those techni-
ques the study is restricted to one or several fixed loci.
Here, we describe a simple immunofluorescence-based tech-
nique to monitor DNA end resection of breaks created by any type
of DNA-damaging agents. A schematic of the protocol can be seen
in Fig. 1a. Briefly, the protocol relies on a long treatment with
BrdU (one cell cycle) to label one strand of all DNA molecules,

A B C
SMART

Cell Culture LONG BrdU PULSE

(20-24 Hr)
(Day 0-1)

DNA Isolation Genomic DNA Plug

(Day 1-2)
D E

Constant Speed

DNA fibers Stretching

(Day 3)

Immuno Detection and

Visualitation
(Day 4-5)

Anti-BrdU

Fig. 1 Single Analysis of resection Tracks. (a) Schematic representation of the SMART protocol. (b) DNA plug
formation with a casting mold. (c) Picture showing how to eject the DNA plugs onto the tubes. (d) To avoid
losing the plugs during the washes, use a cell scrapper to hold it into the tube. (e) A representative image of a
SMART coverslip. Red signal corresponds to ssDNA visualized by BrdU immunofluorescence
 Single Molecule Analysis of Resection Tracks 149

followed by the isolation of DNA in gentle conditions using cells

embedded in agarose plugs. Then, genomic DNA is stretched onto
silanized coverslips to create parallel individual DNA fibers. Finally,
the ssDNA is detected using an anti-BrdU antibody, taking advan-
tage of the fact that the BrdU epitope is hidden on double-stranded
but exposed in single-stranded DNA. The resected DNA can be
visualized by fluorescence microscopy as fibers of BrdU/ssDNA
strands whose length can serve as a metric to quantify resection in
cells responding to DNA damage. Although we developed the assay
to study DNA resection in mammalian cells, the same approach
might be applied successfully to other model organisms by adjust-
ing the BrdU labeling/DNA isolation part of the method using
established DNA combing protocols.

2 Materials

2.1 Equipment 1. Agitator plate orbital.

2. Fluorescent microscope (with FITC and TX2 filters and 40
objective).
3. Hemocytometer cell counter or automated cell counter (i.e.,
Beckman Coulter).
4. Oven (42  C and 65  C).
5. Fiber Comb® Molecular Combing System. (Genomic Vision).

2.1.1 Materials 1. Cell culture plates.

2. Coplin jar.
3. Cover slip (not silanized) (22  22 mm).
4. Cover slip silanized (22  22 mm) (GenomicVision).
5. Filter lab paper.
6. Gel plug mold (CHEF MAPPER PLUG MOLD, BIORAD).
7. Glass slide.
8. Masking tape.
9. Microtubes.
10. Pipettes.
11. Test tubes 12 ml round bottom.
12. Tips (white, yellow).

2.1.2 Buffers 1. Blocking Buffer: 1% BSA in PBS/T.

2. MES 10: 350 mM MES hydrate, 150 mM MES sodium salt.
3. PBS 10: 81.9 g NaCl, 2.01 g KCl, 14.2 g Na2HPO4, 2.45
KH2PO4 g per liter.
4. PBS/T: PBS 1, 0.1% Triton X-100.
150 Pablo Huertas and Andrés Cruz-Garcı́a

5. TE (1): 10 mM Tris, 1 mM EDTA, pH 8.0.

6. TE50: 10 mM Tris, 50 mM EDTA, pH 8.0.

2.1.3 Stocks 1. 10% (W/V) N-lauryl-sarcosyl.

2. 10 mM BrdU: 100 mg of BrdU in 32.56 ml of distilled water.
Aliquot in microtubes and store at 20  C.
3. 500 mM MES Hydrate: 9.762 g in 100 ml of distilled water.
4. 500 mM MES sodium salt: 10.861 g in 100 ml of distilled
water.
5. Agarose 1% in PBS: 250 mg in 250 ml of PBS 1. Aliquot in
microtubes and store at 20  C.
6. Proteinase K 2 mg/ml: 2.5 g in 100 ml of TE50 buffer, 1% l-
sarcosyl. Aliquot in microtubes and store at 20  C.
7. Yoyo-1 (Yoyo-1 iodide): 1 mM solution in DMSO (Molecular
Probes).

3 Methods

A schematic of the protocol can be found in Fig. 1a.

3.1 Culture Growth 1. Seed in 100 mm plates, 5  105 cells in 10 ml of proper cell
(Day 0) culture medium (see Notes 1 and 2).
2. Incubate cells at 37  C for 20–24 h.

3.2 Labeling of 1. Remove old medium from cell culture. Add fresh medium
Mammalian Cells with supplemented with BrdU, 10 μM final concentration.
BrdU (Day 1) 2. Incubate cells at 37  C for 20–24 h (one cell cycle).

3.3 Preparation of 1. Heat agarose solution at 65  C in a thermoblock for at least

Genomic DNA Plugs 15 min, until it is completely melted. Lower the temperature to
(Day 2) 42  C and let the agarose to temper.
2. Treat cells to generate the desired amount and type of DNA
damage (see Note 3).
3. Incubate the cells for 1 h to allow resection to take place (see
Note 4).
4. Rinse cells with cold PBS 1 (4  C), and harvest cells with
Accutase (eBiosciences) for 5 min (see Notes 5 and 6).
5. Resuspend cells gently in 5 ml of PBS 1, spin 3 min at
400  g. During centrifugation, label microtubes for each
sample (see Note 7).
6. Aspirate the PBS and resuspend in 5 ml of ice-cold PBS 1.
7. Count the number of cells and resuspend at a final concentra-
tion of 8.000,000 cells per 1 ml of PBS 1.
8. Transfer cells solution (100 μl) to a pre-labeled microtube.
 Single Molecule Analysis of Resection Tracks 151

9. Briefly pre-warm cell suspension to 42  C in a thermoblock.

10. To prepare 2 plugs add 100 μl of 1% LMT agarose (Bio-Rad) in
PBS 1 to 100 μl of cell solution.
11. Mix very gently with a P200 pipetman (cut tip). This has to be
done quickly; the melted agarose can become solid when the
temperature is lower than 42  C.
12. Transfer 100 μl of the agarose/cells mix per plug into a casting
mold (Fig. 1b) (see Note 8).
13. Let plugs solidify for 25 min at room temperature and addi-
tionally for 5 min at 4  C. While the solidification occurs, label
12 ml round-bottom tubes for each sample.
14. Eject plugs into 12 ml round-bottom tubes containing 500 μl
of proteinase K in TE50 (Fig. 1c). Plugs for the same condition
can be pooled in a single tube, scaling up the amount of buffer.
Make sure that DNA agarose plugs are completely covered by
the buffer solution (see Note 9).
15. Incubate overnight at 50  C.

3.4 Melting of 1. Gently, remove the liquid by blocking the agarose plug with a
Genomic DNA Plug and cell scraper (Fig. 1d).
YOYO-1 Staining (Day 3) 2. Add 500 μl of proteinase K in TE50 buffer per plug.
3. Incubate at 50  C for at least another 6 h (see Note 10).
4. Gently remove buffer without damaging plugs, which should
be now completely transparent.
5. Wash 4  10 min with 10 ml of TE50 at RT with gently shaking
(i.e., 300 rpm in Eppendorf thermomixer).
Optional: Samples can be stored at this point at 4  C in TE50
until use.
6. Put each plug to be stained with YOYO-1 in a new 12 ml
round-bottom tube (see Note 11).
Optional: To keep long-term any extra plugs, add 10 ml of
TE50 to the tube and keep it at 4  C.
7. Stain the plug with 100 μl of TE and add 1.5 μl of YOYO-1
solution (1 mM in DMSO). Incubate in the dark for 30 min at
RT.
8. Wash the plugs with 4  10 ml of TE for 5 min with gently
shaking.
9. Add 3 ml of MES 1 to each tube.
10. Incubate at 65  C until the plugs are completely melted
(15–20 min).
11. Let the temperature drop gradually to 42  C. Add 3 units of
beta Agarase (NEB) by adding 100 μl of MES 1 with 3 μl of
β-Agarase per plug see Note 12.
12. Incubate at 42  C overnight.
152 Pablo Huertas and Andrés Cruz-Garcı́a

3.5 DNA Fibers 1. Incubate at 65  C for 15 min to inactivate the Beta Agarase.
Stretching (Day 4) Then, cool the sample until it reaches room temperature.
2. Perform the DNA fiber stretching in silanized coverslip using
the Fiber Comb® Molecular Combing System (see Note 13).
3. Remove the silanized coverslips. Do not forget to mark the
fiber orientation.
4. Put the coverslips on filter lab paper.
5. Bake the coverslips at 60  C for at least 2 h.
6. Glue the coverslip to a microscope slide. Check the quality of
the fibers by visualizing the YOYO-1 staining under the micro-
scope. Only if good quality fibers are observed continue with
the immunodetection.

3.6 1. Put the slides in a coplin jar. Incubate for 15 min in a blocking
Immunodetection solution (PBS/T, 1% BSA). Be sure that the coverslips are
completely submerged in the buffer. Alternatively, commercial
blocking reagents can be used.
2. Drain the slides and set them horizontally in a wet chamber.
Add 18 μl of PBS/T containing anti BrdU (1:500) antibody to
the coverslips and cover them with another coverslip to spread
the solution. Incubate for 45 min at 37  C (or at RT) in the wet
chamber.
3. Remove the top coverslip. Wash five times for 2 min with PBS/
T.
4. Incubate with secondary antibody (Dilution, 1:1000) in PBS/
T for 30 min at 37  C (or RT) in the wet chamber as described
for the primary antibody.
Optional: Spin fresh aliquots of secondary antibodies for 5 min
at full speed with a microfuge to eliminate aggregates.
5. Wash for 5  2 min with PBS/T.
6. Dry slides and mount with 20 μl of Prolong Gold Antifade
(Molecular Probes) using a micropipette with a cut tip. Let
mounting reagent polymerize overnight at RT before proceed-
ing with microscopy. Mounted coverslips are stable for months
at 20  C in the dark.

3.7 Visualization 1. Image the coverslips using a fluorescent microscope. We use

40 objectives for visualization (Fig. 1e).
2. Measure the length of resected DNA using the ruler tool in the
Analysis menu of Photoshop Adobe Acrobat software. Other
software with measurement tools can be used, i.e.,
MetaMorph.
3. Represent the dispersion of the resected DNA length and
calculate the median.
 Single Molecule Analysis of Resection Tracks 153

4. Individual dispersion of different conditions can be compared

using a Mann–Whitney test or the average of the median of
three experiments for each condition can be plotted and com-
pared using a t-test.

4 Notes

1. This protocol has been optimized for U2OS and HeLa cells.
Other cell types might require a different seeding concentra-
tion or different growth conditions.
2. Generally for each plug 1 million cells are needed. The size of
your cell culture might be scaled up depending on how many
plugs will be required for the specific experiment. We recom-
mend preparing extra plugs, especially the first few times, just in
case.
3. In this specific example, we are using 10 Gy of Ionizing Radia-
tion. Treatments with NCS, campthotecin, and etoposide have
been used in the lab with good results. Note that the dose will
depend on your expected results. We recommend avoiding low
doses, as otherwise the amount of resected fiber might be too
scarce to find a significant number.
4. The incubation time might be adjusted for the desired applica-
tion. In our experience, 1 h is the best option to have tracks
long enough to be measured confidently. For kinetic studies,
several time points should be taken.
5. Warm up the Accutase at RT before use, do not add it at 37  C.
We usually use 700 μl of Accutase and 4–5 min at 37  C for
100 mm plate. The amount of Accutase and time should be
adjusted for specific cell types.
6. To harvest cells trypsin can be used instead of Accutase, but the
yield of DNA fibers is clearly lower.
7. In order to minimize the damage by manipulation of the cells, a
P1000 can be used to resuspend the cells, adding PBS gently
1 ml at a time.
8. This step has to be done sample by sample to avoid solidifica-
tion of the agarose.
9. If the ejected plug is deformed, perhaps the plug was not yet
completely solidified. Put the mold back to the refrigerator and
wait for another 15 min.
10. The order of proteinase K treatment can be reverted without
impacting the final results. Thus, it is possible to start with a 6 h
incubation at 50  C followed by a change of buffer and an
additional incubation of the samples overnight at 50  C.
154 Pablo Huertas and Andrés Cruz-Garcı́a

11. The YOYO-1 staining step is used to check the quality of the
fibers before immunostaining. Therefore, it could be
skipped, as the staining will anyway disappear during
immunofluorescence.
12. Do not add the agarase before the sample is at 42  C to avoid
denaturalization of the enzyme.
13. Manual stretching can be achieved by letting a drop of sample
to slide over a coverslip, but with low reproducibility between
samples.

References

1. Huertas P (2010) DNA resection in eukar- 6. Shroff R, Arbel-Eden A, Pilch D, Ira G, Bonner
yotes: deciding how to fix the break. Nat Struct WM, Petrini JH, Haber JE, Lichten M (2004)
Mol Biol 17(1):11–16. doi:10.1038/nsmb. Distribution and dynamics of chromatin modi-
1710. nsmb.1710 [pii] fication induced by a defined DNA double-
2. Sartori AA, Lukas C, Coates J, Mistrik M, Fu S, strand break. Curr Biol 14(19):1703–1711.
Bartek J, Baer R, Lukas J, Jackson SP (2007) doi:10.1016/j.cub.2004.09.047
Human CtIP promotes DNA end resection. 7. Sugawara N, Haber JE (2006) Repair of DNA
Nature 450(7169):509–514. doi:10.1038/ double strand breaks: in vivo biochemistry.
nature06337. nature06337 [pii] Methods Enzymol 408:416–429. doi:10.
3. Forment JV, Jackson SP (2015) A flow 1016/S0076-6879(06)08026-8
cytometry-based method to simplify the analysis 8. Westmoreland JW, Resnick MA (2013) Coin-
and quantification of protein association to cident resection at both ends of random,
chromatin in mammalian cells. Nat Protoc 10 gamma-induced double-strand breaks requires
(9):1297–1307. doi:10.1038/nprot.2015.066 MRX (MRN), Sae2 (Ctp1), and Mre11-
4. Ma W, Westmoreland J, Nakai W, Malkova A, nuclease. PLoS Genet 9(3):e1003420. doi:10.
Resnick MA (2011) Characterizing resection at 1371/journal.pgen.1003420
random and unique chromosome double- 9. Zhou Y, Paull TT (2015) Direct measurement
strand breaks and telomere ends. Methods of single-stranded DNA intermediates in mam-
Mol Biol 745:15–31. doi:10.1007/978-1- malian cells by quantitative polymerase chain
61779-129-1_2 reaction. Anal Biochem 479:48–50. doi:10.
5. Bothmer A, Robbiani DF, Feldhahn N, 1016/j.ab.2015.03.025
Gazumyan A, Nussenzweig A, Nussenzweig 10. Zierhut C, Diffley JF (2008) Break dosage, cell
MC (2010) 53BP1 regulates DNA resection cycle stage and DNA replication influence
and the choice between classical and alternative DNA double strand break response. EMBO J
end joining during class switch recombination. 27(13):1875–1885. doi:10.1038/emboj.
J Exp Med 207(4):855–865. doi:10.1084/ 2008.111
jem.20100244
 Chapter 13

Mapping DNA Breaks by Next-Generation Sequencing

Laura Baranello, Fedor Kouzine, Damian Wojtowicz, Kairong Cui,
Keji Zhao, Teresa M. Przytycka, Giovanni Capranico, and David Levens

Abstract
Here, we present two approaches to map DNA double-strand breaks (DSBs) and single-strand breaks
(SSBs) in the genome of human cells. We named these methods respectively DSB-Seq and SSB-Seq. We
tested the DSB and SSB-Seq in HCT1116, human colon cancer cells, and validated the results using the
topoisomerase 2 (Top2)-poisoning agent etoposide (ETO). These methods are powerful tools for the
direct detection of the physiological and pathological “breakome” of the DNA in human cells.

Key words DNA damage, Double-strand breaks (DSBs), Single-strand breaks (SSBs), Topoisomerase
2 (Top2), Etoposide (ETO)

1 Introduction

The cellular genome is constantly exposed to extracellular and

intracellular DNA damaging agents that compromise DNA integ-
rity and threaten genomic stability (Fig. 1). As a result, cellular
mechanisms to cope with DNA damage developed early in evolu-
tion and are conserved in current animal and plant species. Under-
standing the sensitivity of the genome to various DNA insults is
instrumental to implement effective preventive and treatment stra-
tegies of diseases. Most of the techniques developed to study DNA
breaks are based on their indirect detection, by examining the
localization of proteins involved in the repair of DNA damage
such as the phosphorylated histone variant γ-H2AX [1] or the
replication protein A (RPA), or by detecting the single-stranded
DNA that transiently accumulates at DSB sites [2]. However,
despite past work having provided comprehensive information on
the pathways involved in detecting and repairing DNA breakage,
our knowledge of how the genome breaks in response to various
agents is incomplete and would be advanced by improved technol-
ogy to detect DNA breaks.

Marco Muzi-Falconi and Grant W. Brown (eds.), Genome Instability: Methods and Protocols, Methods in Molecular Biology,
vol. 1672, DOI 10.1007/978-1-4939-7306-4_13, © Springer Science+Business Media LLC 2018

155
156 Laura Baranello et al.

Fig. 1 Types of DNA damage. Exogenous and endogenous DNA damaging agents generate various types of
lesions including SSBs and DSBs. PARP predominantly acts as a sensor of SSB [17]. RPA binds to regions of
single-stranded DNA (ssDNA) that are exposed to stalled replication forks or after DSB resection [18]. The
multifunctional MRN complex and KU detect DSBs, FANCM is required for the DNA interstrand crosslink (ICL)-
induced checkpoint response [19]. FANCM ¼ Fanconi anemia complementation group M; ICL ¼ interstrand
crosslink; MRN ¼ MRE11-RAD50-NBS1 complex; PARP ¼ poly(ADP-ribose) polymerase; RPA ¼ replication
protein A

Several assays to detect DSBs were developed during recent

years that either rely on the mapping of chromosomal translocation
by using a “bait” DSB introduced into the genome [3–5] or by
capturing the double-strand broken ends with oligodeoxynucleo-
tides or integrase-defective lentiviral vectors. The junctions
between the genomic DNA and the transfected DNA correspond
to the location of the DSBs and can successively be sequenced
[6, 7].
While these methods provide high resolution of DSB and low
background, they are limited to the detection of only translocated
breaks or integrated DNA. Consequently, they cannot be used to
study highly dynamic genome instability events. Additionally, due
to the necessity to introduce a “bait” DSB into the genome or
double-stranded DNA into the nuclei of living cells, the use of
these methods is now restricted to well-established cell lines, not
primary cells or tissues. Real-time capture of genome damaging
events is provided by recent methods based on direct in situ label-
ing of DSBs [8, 9]. However, the methods were not implemented
to identify SSBs and so they do not allow a direct comparison
between DSBs and SSBs in the same genome context [10].
Here, we present two comprehensive experimental and compu-
tational approaches to map DSBs and SSBs across the genome of
human cells [11]. These are based on the direct labeling of breaks
with two independent strategies. The approaches could be
extended to any cell lines and tissue and even on previously isolated
genomic DNA.
To detect SSBs (SSB-Seq, Fig. 2a) the high molecular weight
genomic DNA isolated form HCT116 cells is subjected to nick
translation. Nick translation is a tagging technique in which DNA
polymerase I is used to replace some of the nucleotides of a DNA
 Mapping DNA Breaks by Next-Generation Sequencing 157

Fig. 2 DNA breaks mapping workflow. (a) SSBs are labeled during nick translation using nucleotides covalently
linked to digoxigenin (blue circle). The DNA is subsequently purified, sonicated and incubated with anti-
digoxigenin antibody (anti-DIG). The immuno-precipitated DNA is sequenced. (b) 30 tails of DSBs are ligated to
biotinylated nucleotides (red circle). After sonication the labeled fragments are captured on streptavidin beads
(pink circle). Tails are removed from released fragments and DNA is sequenced

sequence with their labeled analogues—digoxigenin-modified

nucleotides—creating a tagged DNA fragment that can be
immuno-precipitated with anti-digoxigenin antibody and
sequenced [12]. To increase the resolution of mapping, we
restricted the digoxigenin-labeling to a small patch of DNA by
including in the reaction dideoxynucleotides (ddNTP), to inhibit
excessive chain elongation by DNA polymerase I. As a control for
the labeling, samples were also nick-translated without
digoxigenin-labeled nucleotides.
To map DSBs (DSB-Seq, Fig. 2b) the double-stranded DNA
ends were 30 -end tailed with terminal deoxynucleotidyl transferase
158 Laura Baranello et al.

(TdT), a DNA polymerase I that catalyzes the addition of nucleo-

tides to the 30 terminus of a DNA molecule. In the presence of TdT,
biotinylated nucleotides were added to the region of DSB and after
fragmentation, the biotinylated DNA was streptavidin-selected. In
parallel, 30 -tailing was performed in the absence of biotinylated
nucleotides, which constituted our negative control for labeling and
selection. To remove the biotinylated-tails, samples were treated with
S-1 nuclease and the resulting DNA was purified and sequenced (this
technique was modified from our previous work [13]).
Our methods provide maps of the DNA break landscape, in
various cell types and tissues, and in different experimental
conditions.

2 Materials

2.1 Reagents 1. Proteinase K (Solution in water, 20 mg/ml).

2. Phenol, Tris saturated.
3. Phenol:Chloroform:Isoamyl Alcohol 25:24:1.
4. Ethanol 100%.
5. Ammonium acetate 7.5 M.
6. RNase, DNase-free (500 μg/ml).
7. SDS 10%.
8. dATP, dGTP, dCTP, dTTP.
9. Digoxigenin-11-dUTP.
10. ddATP, ddGTP, ddCTP, ddTTP.
11. E. coli DNA polymerase I.
12. EDTA 0.5 M, (pH 8.0).
13. Anti-digoxigenin antibody.
14. Protein G-Sepharose beads.
15. Terminal transferase (TdT).
16. Biotin-16-dUTP.
17. TTP.
18. Streptavidin-coated beads.
19. S-1 nuclease.
20. Klenow (exo-).
21. T4 DNA ligase.
22. Agarose (agarose gels are prepared in TAE buffer).
23. 2% precast agarose gel (e.g., E-Gel, Invitrogen).
24. Illumina adapter (Adaptor oligo mix).
25. Illumina primers (Fw: 50 -aca ctc ttt ccc tac acg acg c-30 /Rv: 50 -
caa gca gaa gac ggc ata cga gc-30 ).
 Mapping DNA Breaks by Next-Generation Sequencing 159

2.2 Buffers 1. Lysis buffer: 10 mM Tris–HCl (pH 8.0), 100 mM EDTA

(pH 8.0), 0.5% SDS.
2. TE buffer: 10 mM Tris–HCl (pH 8.0), 1 mM EDTA (pH 8.0).
3. PBS buffer: 137 mM NaCl, 2.7 mM KCl, 10 mM Na2HPO4,
2 mM KHPO4 (pH 7.2).
4. NP-40 buffer: 20 mM Tris–HCl (pH 8.0), 137 mM NaCl, 10%
Glycerol, 1% NP-40, 2 mM EDTA (pH 8.0).
5. TdT reaction buffer: 50 mM Potassium acetate, 20 mM Tris-
acetate, 10 mM magnesium acetate (pH 7.9).
6. DSB washing buffer: 10 mM Tris–HCl (pH 7.5), 1 mM EDTA
(pH 8.0), and 2.0 M NaCl.
7. DSB elution buffer: 10 mM Tris–HCl (pH 7.5), 1 mM EDTA
(pH 8.0), 1 M NaCl, and 2 M β-mercaptoethanol.
8. TAE buffer: 40 mM Tris (pH 7.6), 20 mM acetic acid, 1 mM
EDTA (pH 8.0).
9. End repair buffer: 33 mM Tris-acetate (pH 7.0), 66 mM potas-
sium acetate, 10 mM magnesium acetate, 0.5 mM DTT.

2.3 Equipment 1. Gel electrophoresis apparatus.

2. Spectrophotometer.
3. Thermoblock (a heating and cooling unit).
4. Sonicator.
5. Rocker.
6. E-Gel Precast Agarose Electrophoresis System (Invitrogen).
7. Genome sequencer.

2.4 Kits 1. QIAquick PCR Purification Kit (Qiagen).

2. Epicentre DNA END-Repair kit.
3. MinElute gel extraction kit (Qiagen).

3 Methods

3.1 Purification of 1. Wash cells (1  108) twice with ice-cold PBS (see Note 1).
High Molecular Weight 2. Lyse cells with 10 ml of lysis buffer (see Note 2).
DNA
3. Collect lysate by scraping and transfer the suspension in a 50 ml
conical tube.
4. Digest the sample overnight with proteinase K (200 μg/ml) at
52  C.
5. Purify DNA twice with phenol (1:1 v/v) and once with phenol-
chloroform (1:1 v/v) (see Note 3).
160 Laura Baranello et al.

6. Precipitate DNA with 2 volumes of ethanol 100% in the pres-

ence of 2 M ammonium acetate (see Note 4).
7. Centrifuge 4,400  g for 30 min. Remove the supernatant,
add ethanol 70%, mix by inverting the sample, centrifuge
4,400  g for 25 min.
8. Air-dry the pellet (see Note 5).
9. Add to the pellet 500 μl of TE buffer and incubate it for 12 h at
room temperature with gentle rotation (see Note 6).
10. Incubate the sample with 5 μg of pancreatic RNase for 1 h at
37 .
11. Adjust the sample to 0.5% SDS and incubate for 1 h at 55  C
with proteinase K (200 μg/ml).
12. Bring the volume to 10 ml with TE buffer and extract the DNA
twice with phenol-chloroform (1:1 v/v) (see Note 5).
13. Precipitate the DNA with 2 volumes of ethanol 100% in the
presence of 2 M ammonium acetate (see Note 6).
14. Air-dry the pellet (see Note 5).
15. Add to the pellet 1 ml of TE buffer and incubate it for 12 h at
room temperature with gentle rotation.
16. Run 10 μl of DNA on a 0.6% agarose gel (see Note 7).
17. Determine DNA concentration with a spectrophotometer (see
Note 8).

3.2 SSB-Seq 1. In a final volume of 1.5 ml, incubate with gentle mixing 500 μg
of DNA for 40 s at 16  C with a mixture of 200 μM of dATP,
dGTP, dCTP and 20 μM of digoxigenin-11-dUTP, 117 μM of
ddATP, ddGTP, ddCTP and 1000 units of E. coli DNA poly-
merase I. As a negative control for labeling, incubate 500 μg of
DNA with the same reagents except digoxigenin-11-dUTP
that is substituted with 20 μM of dTTP (see Note 9).
2. Stop the reaction with 50 μM EDTA.
3. Extract DNA with phenol-chloroform (1:1 v/v).
4. Precipitate DNA in the presence of 2 volumes of 100% ethanol
and 2 M ammonium acetate.
5. Centrifuge 16,000  g for 30 min.
6. Air-dry pellet.
7. Add 1 ml of TE and mix by vortexing.
8. To help resuspension, incubate the sample at 45  C for 15 min,
mix by vortexing.
9. Precipitate DNA in the presence of 2 volumes of 100% ethanol
and 2 M ammonium acetate.
 Mapping DNA Breaks by Next-Generation Sequencing 161

10. Centrifuge 16,000  g for 30 min. Remove the supernatant.

Add 70% ethanol. Centrifuge 16,000  g for 15 min (see Note
10). Remove the supernatant.
11. Air-dry pellet.
12. Add to the pellet 300 μl of TE buffer.
13. Shear the DNA by sonication to an average fragment size of
250 bp (see Note 11).
14. Incubate the sample with 10 μg of anti-digoxigenin antibody at
4  C, overnight with gentle rotation.
15. To recover the immuno-complexes add 60 μl of Protein G-
Sepharose beads (see Note 12) and incubate for 4 h at 4  C.
16. Wash the beads once with PBS buffer, three times with NP-40
buffer; twice with TE buffer; and finally add to the beads 200 μl
of TE buffer (see Note 13).
17. Adjust the sample to 0.5% SDS and digest with proteinase K
(200 μg/ml) at 65  C overnight.
18. Purify DNA using QIAquick PCR Purification Kit according to
the manufacturer’s instructions and quantify the eluate (see
Note 14).

3.3 DSB-Seq 1. Incubate 500 μg of DNA in 3 ml of TdT buffer with 24,000 U

TdT, 0.5 mM dCTP and 5 mM CoCl2 at 37  C for 5 min.
2. Add 0.02 mM Biotin-16-dUTP and incubate at 37  C for
30 min. As a control of labeling, incubate 500 μg of DNA
with the same reagents substituting TTP for Biotin-16-dUTP.
3. Stop the reaction by adding EDTA to a final concentration of
20 μM.
4. Extract the sample with phenol-chloroform (1:1 v/v).
5. Precipitate DNA in the presence of 2 volumes of 100% ethanol
and 2 M ammonium acetate.
6. Centrifuge 16,000  g for 30 min.
7. Air-dry pellet.
8. Add 1 ml of TE and mix by vortexing.
9. To help resuspension, incubate the sample at 45  C for 15 min,
mix by vortexing.
10. Precipitate DNA in the presence of 2 volumes of 100% ethanol
and 2 M ammonium acetate.
11. Centrifuge 16,000  g for 30 min. Remove the supernatant.
Add ethanol 70%. Centrifuge 16,000  g for 15 min (see Note
15). Remove the supernatant.
12. Air-dry the pellet and dissolve it in 200 ml of TE buffer.
162 Laura Baranello et al.

13. Sonicate the biotinylated DNA to generate 200–400 bp DNA

fragments (see Note 11).
14. Capture the biotinylated fragments by mixing the DNA with
200 μl of streptavidin-coated beads (see Note 16).
15. Incubate for 3 h at room temperature with agitation.
16. Wash the beads four times with DSB washing buffer, incubat-
ing the sample at 50  C (see Note 17).
17. Wash the beads four times with DSB washing buffer, incubat-
ing the sample at room temperature (see Note 18).
18. To disrupt biotin-streptavidin complexes incubate the sample
in 200 μl DSB elution buffer at 75  C for 4 h (see Note 19).
19. Purify free DNA fragments with a QIAquick PCR Purification
Kit.
20. To remove the biotinylated tails from DNA, incubate the
sample with 30 U of S-1 nuclease in 110 μl of recommended
buffer for 30 min at 37  C.
21. Purify DNA with a QIAquick PCR Purification Kit.
22. Quantify the recovered DNA.

3.4 Template The DNA recovered from the immuno-precipitation/biotin-step-

Preparation for tavidin selection as well as 10 μl of the Input DNA—genomic DNA
Sequencing Analysis not subjected to immuno-precipitation—will be subjected to
library preparation and sequencing (see Note 20).
1. To generate blunt-ended DNA, incubate the DNA for 45 min
at room temperature in the 25 μl reaction with a mixture of
End repair buffer, 0.25 mM of each dNTPs, 1 mM ATP, and
1 μl End-Repair Enzyme mix (T4 DNA polymerase þ T4
PNK) (see Note 21).
2. Purify the DNA with a MinElute Reaction Cleanup Kit (see
Note 22).
3. In the 25 μl reaction, treat the blunt-ended DNA with 15 units
of Klenow(exo-) for 30 min at 37  C in the presence of 0.2 mM
dATP to generate a protruding 3’A base used for adaptor
ligation.
4. Purify the DNA with a MinElute Reaction Cleanup Kit.
5. In the 20 μl reaction, ligate Illumina adapter to the end of DNA
fragments by incubating with 0.1 μl Adaptor oligo mix and
1000 units of T4 DNA ligase at room temperature for 30 min.
6. Purify DNA using MinElute Reaction Cleanup Kit.
7. To size-select the adapter ligated DNA, run the sample
through 2% E-Gel electrophoresis.
8. Excise the gel slice, around the 200–400 bp region.
 Mapping DNA Breaks by Next-Generation Sequencing 163

9. Purify DNA from the gel using the MinElute gel extraction kit
and elute in a final volume of 12 μl elution buffer.
10. Amplify the DNA for 18 cycles using Illumina primers (Fw: 50 -
aca ctc ttt ccc tac acg acg c-30 /Rv: 50 -caa gca gaa gac ggc ata cga
gc-30 ) according to the following protocol: 98  C for 30 s;
65  C for 30 s; 72  C for 30 s.
11. Run the PCR product through 2.5% agarose gel and excise the
gel slice around 220 bps–500 bps.
12. Purify the DNA from the gel using MinElute gel extraction kit.
13. The purified DNA is used directly for cluster generation and
sequencing analysis using the Illumina Genome Analyzer fol-
lowing the manufacturer’s protocols.

3.5 Processing of 1. Process sequencing data from SSB-Seq and DSB-Seq protocols
Sequencing Data using Illumina Analysis Pipeline (image analysis and base
calling).
2. Check quality of high-throughput sequencing data with the
FastQC software (http://www.bioinformatics.babraham.ac.
uk/projects/fastqc/).
3. Align short sequencing reads of length 36 bp to the reference
human genome using the Bowtie 2 tool (version 2.2.2) with
default parameters [14].
4. Remove redundant reads from the datasets, to minimize poten-
tial PCR bias, using Samtools package (http://www.htslib.
org/doc/samtools.html).
5. Generate a read density visualization of the aligned sequencing
data that can be viewed in most genome browsers (e.g., UCSC
Genome Browser http://genome.ucsc.edu/cgi-bin/
hgGateway). For wiggle track format (wig), extend the reads
to the average length of the genomic fragments, count the
number of reads at each position in the genome, and normalize
the library size to 1 million reads [15]. For faster upload and
display of wiggle file it can be compressed to bigWig format
with the wigToBigWig tool (http://hgdownload.cse.ucsc.
edu/admin/exe/).

4 Notes

1. This protocol is designed for adherent cells. If suspension cells

are used, centrifuge cells 140  g for 4 min, at 4  C.
2. Lysis buffer should be kept at room temperature.
3. In general, during the steps preceding nick-translation DNA
should be handled very gently. After the addition of phenol or
phenol-chloroform mix the phases by inverting the tube for a
164 Laura Baranello et al.

Fig. 3 Representative example of High Molecular Weight (HMW) DNA after the
purification steps described in Subheading 3.1. In lanes 1 and 2 we run two
different markers. The numbers on the left refer to the molecular weight of the
marker in lane 2

few minutes. The aqueous phase should be purified from the

organic phase with a 25 ml pipette, sucking the liquid very
slowly. Do not vortex the sample.
4. Mix by gently inverting the sample for 3 min until a white puff
forms.
5. The sample should not be completely dry, otherwise resuspen-
sion will be difficult.
6. Do not resuspend the pellet by pipetting.
7. This step is performed to check the quality of DNA, which
should run around 23 kb or higher molecular weight. The
sample is viscous; therefore, the tip should be cut before trans-
ferring the aliquot for gel electrophoresis. See Fig. 3 as a repre-
sentative example.
8. 1  108 HCT116 cells give approximately 1 mg of DNA. The
260/280 ratio should be around 1.8 and 1.9 and the 260/230
ratio between 2.0 and 2.2.
9. It is important to control the time of nick translation. Thus, it is
suggested to incubate one sample at the time.
10. The double precipitation with ammonium acetate is necessary
to remove free digoxigenin-11-dUTP.
11. Check the DNA fragment sizes by running a 1% agarose gel. In
our procedure sonication was performed with an ultrasonic
 Mapping DNA Breaks by Next-Generation Sequencing 165

sonicator (Bioruptor, Diagenode) at medium power, by puls-

ing 30 times for 30 s and incubating on ice for 30 s between
each pulse.
12. To prepare the beads, mix the slurry, take 60 μl of beads per
immuno-precipitation, wash the beads three times with ice-
cold PBS buffer. Perform each wash by adding 900 μl of ice-
cold PBS buffer. Incubate the beads for 10 min by rocking at
4  C. Centrifuge the sample 4 min at 1500  g, 4  C. Remove
the supernatant.
13. Perform each wash by adding 900 μl of ice-cold washing
buffer. Incubate the immuno-complex for 10 min by rocking,
centrifuge the sample for 4 min at 1500  g, 4  C. Remove the
supernatant.
14. This protocol is suitable for mapping SSBs with a free 30 end.
SSBs generated during topoisomerase 1 catalytic cycle will be
labeled only after the treatment of DNA with tyrosyl-DNA
phosphodiesterase 1 (TDP1) [16].
15. The double precipitation with ammonium acetate is necessary
to remove free Biotin-16-dUTP.
16. Available from Invitrogen (Dynabeads kilobase BINDER Kit,
Dynal). To prepare the beads, use the magnet to separate beads
from the supernatant. Add to the beads 200 μl of binding
buffer (provided with the beads). Mix for 5 min. Remove the
supernatant. Repeat the wash.
17. To perform each wash, add 900 μl of washing solution. Incu-
bate at 50 for 5 min with agitation. Use the magnet to
separate the beads from the washing buffer. Add a new washing
solution.
18. To perform each wash, add 900 μl of the washing solution.
Incubate at room temperature for 5 min, with agitation. Use
the magnet to separate the beads from the washing buffer. Add
a new washing solution.
19. For a better yield, add 100 μl of DSB elution buffer, incubate at
75  C for 2 h, use the magnet to separate the supernatant from
the beads. Keep the supernatant. Add to the beads 100 μl of
new DSB elution buffer, incubate at 75  C for 2 h, use the
magnet to separate the supernatant from the beads. Pool the
supernatants together.
20. The Input DNA is sonicated.
21. The Epicentre DNA END-Repair kit is available at Epicentre
Biotechnologies.
22. This kit is available at QIAGEN.
166 Laura Baranello et al.

References
1. Iacovoni JS et al (2010) High-resolution 9. Canela A et al (2016) DNA breaks and end
profiling of gammaH2AX around DNA double resection measured genome-wide by end
strand breaks in the mammalian genome. sequencing. Mol Cell 63(5):898–911
EMBO J 29(8):1446–1457 10. Aguilera A, Garcia-Muse T (2013) Causes of
2. Blitzblau HG, Hochwagen A (2011) Genome- genome instability. Annu Rev Genet 47:1–32
wide detection of meiotic DNA double-strand 11. Baranello L et al (2014) DNA break mapping
break hotspots using single-stranded DNA. reveals topoisomerase II activity genome-wide.
Methods Mol Biol 745:47–63 Int J Mol Sci 15(7):13111–13122
3. Hu J et al (2016) Detecting DNA double- 12. Rigby PW et al (1977) Labeling deoxyribonu-
stranded breaks in mammalian genomes by lin- cleic acid to high specific activity in vitro by
ear amplification-mediated high-throughput nick translation with DNA polymerase I. J
genome-wide translocation sequencing. Nat Mol Biol 113(1):237–251
Protoc 11(5):853–871 13. Kouzine F et al (2013) Global regulation of
4. Klein IA et al (2011) Translocation-capture promoter melting in naive lymphocytes. Cell
sequencing reveals the extent and nature of 153(5):988–999
chromosomal rearrangements in B lympho- 14. Langmead B, Salzberg SL (2012) Fast gapped-
cytes. Cell 147(1):95–106 read alignment with bowtie 2. Nat Methods 9
5. Chiarle R et al (2011) Genome-wide transloca- (4):357–359
tion sequencing reveals mechanisms of chro- 15. Bardet AF et al (2012) A computational pipe-
mosome breaks and rearrangements in B cells. line for comparative ChIP-seq analyses. Nat
Cell 147(1):107–119 Protoc 7(1):45–61
6. Tsai SQ et al (2015) GUIDE-seq enables 16. Caldecott KW (2008) Single-strand break
genome-wide profiling of off-target cleavage repair and genetic disease. Nat Rev Genet 9
by CRISPR-Cas nucleases. Nat Biotechnol 33 (8):619–631
(2):187–197
17. Barnes DE, Lindahl T (2004) Repair and
7. Wang XL et al (2015) Unbiased detection of genetic consequences of endogenous DNA
off-target cleavage by CRISPR-Cas9 and base damage in mammalian cells. Annu Rev
TALENs using integrase-defective lentiviral Genet 38:445–476
vectors. Nat Biotechnol 33(2):175–178
18. Sartori AA et al (2007) Human CtIP promotes
8. Crosetto N et al (2013) Nucleotide-resolution DNA end resection. Nature 450(7169):509–514
DNA double-strand break mapping by next-
generation sequencing. Nat Methods 10 19. Symington LS, Gautier J (2011) Double-
(4):361–365 strand break end resection and repair pathway
choice. Annu Rev Genet 45:247–271
 Chapter 14

Genome-Wide Profiling of DNA Double-Strand Breaks

by the BLESS and BLISS Methods
Reza Mirzazadeh, Tomasz Kallas, Magda Bienko, and Nicola Crosetto

Abstract
DNA double-strand breaks (DSBs) are major DNA lesions that are constantly formed during physiological
processes such as DNA replication, transcription, and recombination, or as a result of exogenous agents
such as ionizing radiation, radiomimetic drugs, and genome editing nucleases. Unrepaired DSBs threaten
genomic stability by leading to the formation of potentially oncogenic rearrangements such as transloca-
tions. In past few years, several methods based on next-generation sequencing (NGS) have been developed
to study the genome-wide distribution of DSBs or their conversion to translocation events. We developed
Breaks Labeling, Enrichment on Streptavidin, and Sequencing (BLESS), which was the first method for
direct labeling of DSBs in situ followed by their genome-wide mapping at nucleotide resolution (Crosetto
et al., Nat Methods 10:361–365, 2013). Recently, we have further expanded the quantitative nature,
applicability, and scalability of BLESS by developing Breaks Labeling In Situ and Sequencing (BLISS) (Yan
et al., Nat Commun 8:15058, 2017). Here, we first present an overview of existing methods for genome-
wide localization of DSBs, and then focus on the BLESS and BLISS methods, discussing different assay
design options depending on the sample type and application.

Key words DNA double-strand breaks, Genome instability, Next-generation sequencing, Genome
editing, BLESS, BLISS

1 Introduction

The integrity of DNA is constantly challenged by exogenous and

endogenous damaging agents. In multicellular organisms, individ-
ual cells are thought to undergo thousands of lesion events per day,
most of which are promptly repaired by an intricate network of
signaling pathways in a process known as DNA damage response
[1]. Among different lesion types, DNA double-strand
breaks (DSBs) pose a major threat to genomic stability and organ-
ismal homeostasis, since unrepaired or mis-repaired DSBs can result
in a variety of mutations—chromosomal rearrangements as well as
small insertions/deletions (indels) due to the error-prone activity
of the Non-Homologous End Joining (NHEJ) repair process—
that can result in cell death or potentially initiate carcinogenesis.

Marco Muzi-Falconi and Grant W. Brown (eds.), Genome Instability: Methods and Protocols, Methods in Molecular Biology,
vol. 1672, DOI 10.1007/978-1-4939-7306-4_14, © Springer Science+Business Media LLC 2018

167
168 Reza Mirzazadeh et al.

DSBs arise as a consequence of exposure to exogenous damaging

agents, such as ionizing radiation and several chemotherapeutic
drugs, but the vast majority of DSBs is formed during physiological
cellular processes, including DNA recombination, replication, and
transcription [2, 3]. Another source of DSBs that has recently
become under the spotlight are endonucleases used for genome
editing purposes, including zinc-finger nucleases [4], transcription
activator-like effector nucleases (TALENs) [5], and endonucleases
belonging to the clustered, regularly interspaced, short palindromic
repeats (CRISPR) system, such as Cas9 [6] and Cpf1 [7]. While
these enzymatic activities mainly induce DSBs at a specific genomic
location where the desired genome-editing event is planned, the
formation of unspecific, off-target DSBs is not a completely
unlikely event and represents an important concern particularly in
the context of medical applications [8]. Thus, systematic genome-
wide characterization of the off-target activity of designer endonu-
cleases is necessary before they can fulfill their therapeutic potential.
In addition to helping characterize the safety profile of
genome-editing endonucleases, the ability to accurately map the
location of DSBs along the genome is essential to understand the
role of DNA sequence and chromatin structure in the formation,
processing, and repair of these lesions. In recent years, several
methods have been developed to localize DSBs genome-wide by
harnessing the high-throughput capacity of next-generation
sequencing (NGS). A summary of these methods and their main
characteristics is presented in Table 1. Chromatin immunoprecipi-
tation sequencing (ChIP-seq) targeting the phosphorylated his-
tone variant H2A.X (γH2A.X) around DSBs was initially used to
identify fragile sites in yeast [9]. Subsequently, ChIP-seq against
γH2A.X or other repair factors recruited at DSB sites, such as
53BP1, was applied to investigate the process of repair of DSBs
induced by the AsiSI endonuclease in mammalian cells [10].
Recently, ChIP-seq against Topoisomerase IIβ in neuronal stem/
progenitor cells showed that neuronal activity stimulation results in
the formation of DSBs in the promoters of early-response genes,
which is essential to their transcription [11], highlighting the
importance of mapping DSBs at genomic scale in order to gain
new insights into fundamental biological processes.
A major limitation of ChIP-seq, however, is that γH2A.X and
other DSB markers extend around DSBs for thousands or even
millions bases, thus preventing the identification of fragile hotspots
at nucleotide resolution. In order to circumvent this problem, we
pioneered BLESS, which was the first method enabling direct label-
ing of DSBs and identification of their genomic location by NGS
[12]. Using BLESS, we identified numerous aphidicolin-sensitive
regions that recurrently break upon replication stress, and found
that certain repeat element classes, such as alpha satellites, are partic-
ularly prone to break under replication stress. Subsequently, BLESS
 Genome-Wide Profiling of DNA Double-Strand Breaks by the BLESS and BLISS Methods 169

Table 1
Comparison between different methods for DSBs sequencing

Method Detection Main features Sample (input) Reported applications

BLESS Direct In situ DSB blunting and Fixed cells Replication stress-induced
ligation of biotinylated (at least 106 DSBs in mammalian cells
adapters. DSB capture on cells) [12], Cas9 specificity
streptavidin [13, 14]
DSBCapture Direct In situ DSB blunting and Fixed cells DSBs at G-quadruplex-rich
A-tailing. Modified BLESS (at least 106 sites, active genes and
adapters containing cells) transcription start sites
Illumina adapter sequences [15]
END-seq Direct In vivo DSB blunting and Live cells AsiSI-induced DSBs
A-tailing in agarose plugs. (at least 107 resection mapping, RAG
Modified BLESS adapters cells) endonuclease specificity
containing Illumina adapter [16]
sequences
BLISS Direct In situ DSB blunting, Fixed cells, Transcription-associated
A-tailing, and ligation of tissue DSBs in cells and tissue
adapters containing the T7 sections and Cas9 and Cpf1
promoter sequence and (at least 103 specificity [17]
UMIs. Selective DSB cells)
amplification by in vitro
transcription
Digenome- Direct In vitro nuclease digestion of Purified DNA Cas9 and Cpf1 specificity
seq purified genomic DNA and [18, 19]
detection of DSBs by
whole-genome sequencing
ChIP-seq Indirect Capture of chromatin Fixed cells Replication stress-induced
containing DSBs markers (at least 107 DSBs in yeast [9], AsiSI-
such as γH2A.X cells) induced DSBs processing
in mammalian cells [10],
transcription-associated
DSBs in neuronal cells
[14]
GUIDE-seq Indirect In vivo DSB labeling by Transfected live Cas9 and Cpf1 specificity
incorporation of dsDNA cells [20, 21]
oligos through NHEJ-
mediated repair
IDLV Indirect In vivo DSB labeling by Transduced live Cas9 and TALENs
capture random incorporation of cells specificity [22]
integration defective
lentiviral vectors through
NHEJ-mediated repair
LAM- Indirect In vivo induction of DSBs and Live cells Cas9 specificity [23],
HTGTS sequencing of translocation treated to transcription-associated
products originated from induce DSBs in neuronal cells
NHEJ-mediated repair translocations [24]
170 Reza Mirzazadeh et al.

was also applied to profile the off-target activity of CRISPR/Cas9 on

a genome-wide scale [13, 14]. Recently, two new methods,
DSBCapture [15] and END-seq [16], extended the BLESS design
by introducing an A-tailing step, improving the adapter design, and
implementing DSB labeling in vivo based on the use of agarose plugs
(in END-seq). An important limitation of BLESS and its recent
modifications, however, is that a substantial amount of starting mate-
rial is required (typically, more than 107 cells), thus challenging their
application to low-input samples and tissue specimens. Moreover,
these methods are labor-intensive and thus difficult to scale up to
large numbers of samples. To overcome these limitations, we recently
improved the BLESS concept by developing DNA Double-Strand
Breaks Labeling In Situ and Sequencing (BLISS). In BLISS, DSBs are
in situ ligated to a dsDNA oligonucleotide adapter that contains the
T7 promoter sequence enabling selective linear amplification of the
labeled DSBs by in vitro transcription (IVT). In addition, unique
molecular identifiers (UMIs) [25] and sample-specific barcodes are
incorporated at the site of in situ DSB ligation, allowing for multi-
plexing and quantitative mapping of DSBs even in low-input samples.
A side-by-side comparison and detailed description of the BLESS and
BLISS methods is presented in Fig. 1.
In parallel to the BLESS and its derivative methods, other
genome-wide DSB localization approaches have recently emerged,
particularly with the purpose of characterizing the specificity profile of
genome-editing endonucleases. These methods include in vitro
digested whole-genome sequencing (Digenome-seq) [18], GUIDE-
seq [20], integration-deficient lentiviral vector (IDLV) capture [22],
and high-throughput, genome-wide translocation sequencing
(HTGTS) [23]. An overview of these methods and their comparison
to BLESS and BLISS is presented in Table 1. The main distinctive
feature of these methods in comparison to BLESS and its derivatives is
that DSBs are not directly labeled (except for Digenome-seq, in
which purified DNA is digested in vitro). GUIDEseq, IDLV incor-
poration, and HTGTS methods require the activity of NHEJ to label
DSBs, thus potentially missing DSBs that are not repaired through
this pathway. Furthermore, exogenous DNA introduction in the case
of GUIDEseq and IDLV integration can be challenging in primary
cells and in tissues, where transfection efficiency and toxicity in the
former, and delivery in the latter may be limiting factors. Here, we
focus on the BLESS and BLISS methods, and provide a flexible step-
by-step protocol that can be adapted to various applications.

2 Materials

2.1 Common 1. Methanol-free paraformaldehyde (PFA) 16%.

Reagents for BLESS 2. Lysis buffer 1 (LB1): Tris–HCl 10 mM, NaCl 10 mM, EDTA
and BLISS 1 mM, Triton X-100 0.2%, pH 8 at 4
C.
 Genome-Wide Profiling of DNA Double-Strand Breaks by the BLESS and BLISS Methods 171

BLESS BLISS
7
At least 10 cells At least 103 cells

Trypsinization, fixation Attachment, fixation and

and permeabilization permeabilization

Nuclei purification In situ DSB blunting

5'

3'

In situ DSBs blunting

5'
In situ DSB ligation
3'

In situ DSBs ligation

DNA purification
and sonication

DNA purification
and sonication
In vitro transcription

Streptavidin capture

Library preparation
and sequencing

On-beads ligation

Adapter removal
Genomic DNA Sample barcode

Sample barcode (proximal) UMI

Sample barcode (distal) RA5

PCR
XhoI T7 promoter

I-SceI

Poly(T) loop

Biotin
Library preparation
and sequencing Streptavidin

Fig. 1 Side-by-side comparison of BLESS and BLISS workflows. Both the procedures start by fixing cells in
order to stabilize chromatin and prevent the formation of artificial breaks during subsequent steps. Next, cells
172 Reza Mirzazadeh et al.

3. Lysis buffer 2 (LB2): Tris–HCl 10 mM, NaCl 150 mM, EDTA

1 mM, SDS 0.3%, pH 8 at 25
C.
4. Nuclease-free water.
5. Nuclease-free Phosphate-Buffered Saline (10), pH 7.4.
6. CutSmart® buffer (NEB, cat. no. B7204S).
7. Quick Blunting™ Kit (NEB, cat. no. E1201L).
8. NEBNext® dA-Tailing Module (NEB, cat. no E6053L),
OPTIONAL.
9. BSA (50 mg/ml) (e.g., Thermo, cat. no. AM2616).
10. ATP Solution (100 mM).
11. T4 DNA Ligase (e.g., NEB, cat. no. M0202M).
12. High-salt wash buffer (HSW): Tris–HCl 10 mM, NaCl 2 M,
EDTA 2 mM, Triton X-100 0.5%, pH 8 at 25
C.
13. Proteinase K, Molecular Biology Grade (NEB, cat. no.
P8107S).
14. Nuclease-free TE buffer.
15. NEBNext® High-Fidelity 2 PCR Master Mix (NEB, cat. no.
M0541L).

Fig. 1 (continued) are permeabilized in order for in situ DSB labeling reactions to take place in the crowded
environment of cross-linked chromatin. In BLESS, but not in BLISS, a short incubation in the presence of
Proteinase K is used to purify nuclei and render them more accessible for in situ reactions (see Note 1). After
permeabilization, in situ blunting is performed in order to convert the DSB ends that contain an overhang (such
as the intermediates of end-resection formed during homologous recombination repair) to a ligatable
configuration. Blunt DSBs are then ligated in situ using oligo adapters with a different design in BLESS and
BLISS (see Note 2 and Table 2). A key difference between BLESS and BLISS is that, while in BLESS
permeabilization and in situ reactions are performed in suspension, in BLISS all these steps are done on a
solid surface (a microscope slide or coverslip). This avoids the requirement for multiple centrifugations (which
cause progressive sample loss and increased turnaround time) and enables safe processing even of very low-
input samples consisting of few thousand cells, such as rare cell populations and precious clinical specimens.
In addition, BLISS may be performed directly on fixed tissue sections mounted on a microscope coverslip or
slide, opening the possibility to study DSBs in a wide range of clinically relevant samples. After in situ labeling
of DSBs, DNA can be safely extracted and purified (in BLISS, cells or tissue sections are either scraped off the
slide/coverslip or they are captured using Laser Capture Microdissection or less resolved methods such as the
Pinpoint Slide DNA Isolation System™ [Zymo Research, cat. no. D3001]). In BLESS, the genomic sequence
surrounding the DSB ends is selectively captured using affinity purification on streptavidin beads. In BLISS,
DSBs are selectively amplified using IVT driven by the T7 RNA polymerase bound to the T7 promoter sequence
ligated to the DSB ends. Finally, preparation of sequencing libraries using the captured/amplified material is
achieved using standard Illumina technology (see Notes 18 and 20). Major advantages of BLISS over BLESS
include: (1) lower input requirement (few million cells in BLESS, as little as 103 cells in BLISS); (2) faster
turnaround (4 days from fixation to library for BLISS, at least 10 days for BLESS); (3) simpler workflow
(especially due to the avoidance of multiple centrifugations and Proteinase K incubation); (4) improved adapter
design, enabling more quantitative data analysis thanks to UMIs; (5) easy scalability by performing all in situ
reactions in multi-well plates
 Genome-Wide Profiling of DNA Double-Strand Breaks by the BLESS and BLISS Methods 173

2.2 Reagents 1. NEBuffer 2 (NEB, cat. no. B7002S).

for BLESS 2. Dynabeads® MyOne™ Streptavidin C1 (Thermo, cat. no.
65001).
3. I-SceI (NEB, cat. no. R0694L).
4. XhoI (NEB, cat. no. R0146L).
5. NEBNext® DNA Library Prep Master Mix Set for Illumina
(NEB, cat. no. E6040L).

2.3 Reagents 1. Poly-L-lysine (PLL) solution (Sigma, cat. no. P8920-100ML).

for BLISS 2. Nucleus Suspension Buffer (NSB): NaCl 146 mM, Tris–HCl
10 mM, CaCl2 1 mM, MgCl2 21 mM, BSA 0.05%, Nonidet P-
40 0.2%, pH 7.8. OPTIONAL (for tissue-BLISS).
3. T4 Polynucleotide Kinase (e.g., NEB, cat. no. M0201L).
4. MEGAscript® T7 Transcription Kit (Thermo, cat. no.
AM1334).
5. DNase I, RNase-free.
6. RiboSafe RNase Inhibitor (Bioline, cat. no. 65027).
7. AMPure XP (Beckman Coulter, cat. no. A63881).
8. Agencourt RNAClean XP (Beckman Coulter, cat. no.
A63987).
9. T4 RNA Ligase 2, truncated (NEB, cat. no. M0242L).
10. RNaseOUT™ Recombinant Ribonuclease Inhibitor (Thermo,
cat. no. 10777019).
11. SuperScript® IV First-Strand Synthesis System (Thermo, cat.
no. 18091050).
12. TruSeq Small RNA Sample Preparation Kit (Illumina, cat. no.
RS-200-0012, RS-200-0024).

2.4 Consumables 1. Coverslips (e.g., VWR, cat. no. 631-0148).

2. Secure-Seal™ Hybridization Chambers (EMS, cat. no
70333-10).
3. Microcentrifuge tubes 0.5 ml (e.g., Eppendorf RNA/DNA
LoBind).
4. Microcentrifuge tubes 1.5 ml (e.g., Eppendorf RNA/DNA
LoBind).
5. Sapphire Filter tips, low retention (Greiner Bio-One, cat. no.
771265, 773265, 738265, 750265).
6. Cell scrapers.
7. Qubit® dsDNA HS Assay Kit (Thermo, cat. no. Q32851).
8. High Sensitivity DNA Kit (Agilent, cat. no. 5067-4626).
9. RNA 6000 Pico Kit (Agilent, cat. no. 5067-1513).
174 Reza Mirzazadeh et al.

2.5 Equipment 1. Cell counter (e.g., Countess II FL Automated Cell Counter,

Thermo).
2. Cooling incubator (e.g., Binder incubator, Model KB 53).
3. Tabletop centrifuge.
4. Thermoshaker (e.g., Eppendorf® Thermomixer Compact).
5. PCR cycler.
6. Sonication device (e.g., Bioruptor® Plus, Diagenode, cat. no.
B01020001).
7. DynaMag™-2 Magnet (Thermo, cat. no. 12321D).
8. Qubit® 2.0 Fluorometer (Thermo, cat. no. Q32866).
9. Bioanalyzer 2100 (Agilent, cat. no. G2943CA).

3 Methods

In both BLESS and BLISS, suspension or adherent cells are

grown inside cell culture dishes or flasks and, depending on the
experiment, they are treated with the desired DSB-inducing pro-
tocol (for example, a DSB-inducing drug or gene manipulation
using siRNA, zinc-finger nucleases, TALENs, or CRISPR). In
BLISS, adherent cells are typically grown and treated on coverslips
placed in a cell culture dish, while suspension cells are grown and
treated in cell culture flasks, and then spotted onto PLL-coated
coverslips before fixation. For BLISS in large numbers of samples
(for example, to assess the specificity of multiple CRISPR guide
RNAs) cells are grown, treated, and processed directly inside
multi-well plates.

3.1 BLISS Adapters BLISS adapters are generated by annealing a forward oligo with a
reverse complementary oligo, each purified by standard desalting
(see Table 2 and Note 2).
1. Prepare the phosphorylation mix (volumes for 100 μl, to be
adjusted proportionally depending on the sample volume):

(a) Nuclease-free water 58 μl

(b) Upper oligo 100 μM 10 μl
(c) 10 T4 PNK buffer 10 μl
(d) ATP 10 mM 10 μl
(e) T4 polynucleotide kinase 10 U/μl 2 μl

2. Incubate the sample for 1 h at 37
C.

3. Add 10 μl of bottom oligo at 100 μM and mix well.
 Genome-Wide Profiling of DNA Double-Strand Breaks by the BLESS and BLISS Methods 175

Table 2
Sequence of adapters and primers used in BLESS and BLISS

Oligo name Sequence (50 –30 ) Structure

BLESS P TACTAC_CTCGAG_AGTTACGCTAGGGATAACAG Barcode_XhoI_ISce_
adaptera GGTAATATAG_TTT/Bio/TTTT_CTATATTACCCTG ISceI_XhoI_Barcode
TTATCCCTAGCGTAACT_CTCGAG_GTAGTA
BLESS D CGTCGT_CTCGAG_AGTTACGCTAGGGATAACAGG Barcode_XhoI_ISce_
adapterb GTAATATAG_TTTTTTT_CTATATTACCCTGTTAT ISceI_XhoI_Barcode
CCCTAGCGTAACT_CTCGAG_ACGACG
BLESS PCR CCCTAGCGTAACT_CTCGAG_GTAGTA ISceI_XhoI_Barcode
primer P1c
BLESS PCR CTAGCGTAACT_CTCGAG_ACGACG ISceI_XhoI_Barcode
primer P2d
BLISS adaptere Upper GCGTGATG_NNNNNNNN_GATCGTCGG Barcode_UMI_RA5_
oligo ACTGTAGAACTCTGAAC_CCCTATAGTGA T7_Overhang
GTCGTATTACCGGCCTCAATCG_AA
Bottom CGATTGAGGCCGGTAATACGACTCA T7_RA5_UMI_
oligo CTATAGGG_GTTCAGAGTTCTACAG Barcode
TCCGACGATC_NNNNNNNN_CATCACGC
BLISS RA3 TGGAATTCTCGGGTGCCAAGG
adapterf
BLISS RTP GCCTTGGCACCCGAGAATTCCA
primerg
BLISS AATGATACGGCGACCACCGAGATCTACACG
universal PCR TTCAGAGTTCTACAGTCCGA
primerh
BLISS indexing CAAGCAGAAGACGGCATACGAGAT Bold:Index sequence
PCR primeri CGTGATGTGACTGGAGTTCCT
TGGCACCCGAGAATTCCA
a
Same as linker L1 linker in Suppl. Table 5 in [12]. As discussed in Note 2, this adapter may be modified by substituting
the I-SceI sequence with the Illumina P5 adapter sequence, removing the XhoI site, and replacing the barcode with a
longer or different sequence, if multiplexing is needed. A UMI sequence may be added before the barcode to avoid
problems with cluster calling during sequencing, and enable removal of PCR duplicates as in BLISS. In addition, instead
of synthesizing one long self-annealing oligo, one upper and one bottom reverse complementary oligo may be annealed
instead, as during the preparation of BLISS adapters.
b
Same as linker L3 in Suppl. Table 5 in [12]. The same considerations exposed above for the proximal adapter also apply
for the distal adapter. However, in this case, the I-SceI sequence should be replaced with the Illumina P3 adapter
sequence. Moreover, adding a UMI in this adapter is not required.
c
Same as primer P1 in Suppl. Table 5 in [12].
d
Same as primer P3 in Suppl. Table 5 in [12].
e
As explained in Note 2, BLISS adapters are prepared by annealing two complementary oligos containing the T7
promoter sequence, the sequence of the 50 adapter (RA5) contained in Illumina’s TruSeq Small RNA Sample Preparation
Kit, a UMI and a sample barcode. The upper oligos terminates with an A dinucleotide overhang to prevent the formation
of head-to-tail concatemers during in situ ligation. We have successfully used 8 nt and 12 nt UMIs, as well as multiple
barcode sequences [17]. One barcode sequence is shown as an example.
f
Same as RA5 adapter in Illumina’s TruSeq Small RNA Sample Preparation Kit.
g
Same as RTP primer in Illumina’s TruSeq Small RNA Sample Preparation Kit.
h
Same as RP1 primer in Illumina’s TruSeq Small RNA Sample Preparation Kit.
i
Same as RPI1 primer in Illumina’s TruSeq Small RNA Sample Preparation Kit. Different RPI primers are available for
multiplexing when multiple libraries need to be sequenced in the same run.
176 Reza Mirzazadeh et al.

4. In a PCR thermo-cycler, perform denaturation for 5 min at

95
C, then ramp down to 25
C over a period of 45 min
(approx. –1.55
C/min).
5. Store the adapter at 20
C.

3.2 PLL Coating 1. Place the coverslips in a 10 cm dish, cover them with 5 ml PLL
of Coverslips for BLISS solution, and then gently shake the dishes for 15 min (make
sure that coverslips stay covered by liquid).
2. Aspirate the solution and transfer it into a 15 ml tube (it can be
then used up to three times).
3. Wash the coverslips three times with 1 PBS at room tempera-
ture (rt) and once with ethanol (EtOH) 70%.
4. Air-dry the coverslips and proceed with cell spotting and fixa-
tion (if not used immediately, store the coverslips in EtOH 70%
at 4
C and air-dry before use).

3.3 Attachment and 1. Place the desired number of PLL-coated coverslips into a 24-
Fixation of Suspension well cell culture plate.
Cells for BLISS 2. In each well, dispense a cell suspension freshly prepared in 1
(See Note 3) PBS (we usually spot up to 3  105 cells onto a 13 mm coverslip).
3. Let the cells sediment onto the coverslip for 10 min at rt.
4. Slowly add one volume of PFA 8% in 1 PBS equal to the
volume of cell suspension added before onto the coverslip.
5. Incubate for 10 min at rt.
6. Rinse the coverslips twice with 1 PBS at rt.
7. Store the samples in 1 PBS at 4
C or proceed to permeabi-
lization (see Note 4).

3.4 Cell 1. Incubate the samples in LB1 for 1 h at 4
C.

Permeabilization 2. Equilibrate the samples at rt. and then incubate in LB2 for 1 h
(See Note 5) at 37
C.
3. Wash the samples (see Note 6) and proceed to in situ blunting.

3.5 In Situ DSB 1. Equilibrate the samples once with 1 blunting buffer.
Blunting (See Note 7) 2. Prepare the following blunting mix using the Quick Blunt-
ing™ Kit (volumes for 100 μl, to be adjusted proportionally
depending on the sample volume):
(a) Nuclease-free water 75 μl
(b) 10 blunting buffer 10 μl
(c) dNTPs 1 mM 10 μl
(d) BSA 10 mg/ml 1 μl
(e) Blunting enzyme mix 4 μl

3. Incubate the samples for 1 h at rt.

 Genome-Wide Profiling of DNA Double-Strand Breaks by the BLESS and BLISS Methods 177

3.6 OPTIONAL: 1. Wash the samples twice, 5 min each (see Note 6).
In Situ A-Tailing 2. Prepare the following A-tailing mix using the NEBNext dA-
(See Note 8) Tailing Module (volumes for 100 μl, to be adjusted propor-
tionally depending on the sample volume):
(a) Nuclease-free water 84 μl
(b) NEBNext dA-tailing reaction buffer 10 μl
(10)
(c) Klenow fragment (30 ! 50 exo) 6 μl

3. Incubate the samples for 30 min at 37
C.

3.7 In Situ DSB 1. Wash the samples twice, 5 min each (see Note 6).
Ligation (See Note 9) 2. Equilibrate the samples once with 1 T4 ligase buffer.
3. Prepare the following ligation mix (volumes for 100 μl, to be
adjusted proportionally depending on the sample volume):
(a) Nuclease-free water 75 μl
(b) T4 ligase buffer 10 μl
(c) ATP 10 mM 8 μl
(d) BSA 50 mg/ml 2 μl
(e) Proximal adapter 10 μM (see Table 2 and Note 2) 4 μl
(f) T4 ligase 2000 U/μl 1 μl

4. Incubate the samples for 16–18 h at 16
C.

3.8 Removal 1. Wash the samples with HSW for three times, 1 h each, at 37
C.
of Unligated Adapters 2. Quickly rinse the samples with 1 PBS at rt.
(See Note 10)
3.9 Extraction 1. Extract and purify genomic DNA either using silica-based col-
and Purification umns or standard methods based on Proteinase K and alcohol
of Genomic DNA extraction.
(See Note 11) 2. Dissolve the purified DNA in nuclease-free TE buffer.
3. Measure DNA concentration by Nanodrop™ or Qubit®, and
dilute DNA in TE buffer to the concentration recommended
for the sonication instrument used.

3.10 DNA 1. Prepare samples for sonication according to the manufacturer’s

Fragmentation instructions for the system used.
by Sonication 2. Sonicate DNA aiming to achieve a mean fragment size of
(See Note 12) 300–500 bp. An example of settings that we routinely use is
shown in Table 3.
178 Reza Mirzazadeh et al.

Table 3
Sonication options for DNA shearing in BLESS and BLISS

Target size Volume,

Instrument Sample (bp) Settings buffer
Bioruptor® Cells 350 30 s. on/90 s. off, high-mode, 20 cycles 100 μl, TE
Plus Tissue 350 30 s. on/90 s. off, high-mode, 40 cycles 100 μl, TE
Covaris S- Cells 350 Duty 10%, intensity 4, time 30 s., 4 cycle/ 50 μl, TE
series burst 200

3. Depending on the application, concentrate the sonicated DNA

using AMPure XP beads or silica-based columns.
l Checkpoint: check DNA fragment size by agarose gel elec-
trophoresis or Bioanalyzer.

3.11 Affinity Capture 1. Aliquot 5 μl of Dynabeads® MyOne™ Streptavidin C1 suspen-

of DSBs in BLESS sion into a 1.5 ml tube, and wash the beads two times with
(See Note 13) 500 μl of HSW.
2. Dispense a volume of sonicated genomic DNA corresponding
to 20 μg, and then bring the volume up to 600 μl with HSW
buffer freshly supplemented with Triton X-100 0.1%.
3. Incubate the samples for 30 min at 4
C, rotating.
4. Wash the beads three times, each with 600 μl of HSW buffer
freshly supplemented with Triton X-100 0.1%.
5. Resuspend the samples in nuclease-free water.
6. Store the samples on ice before proceeding to the ligation of
the distal adapter.

3.12 Ligation of 1. Prepare the following blunting mix using the Quick Blunt-
Distal Adapters in ing™ Kit (volumes for 100 μl, to be adjusted proportionally
BLESS (See Note 14) depending on the sample volume):
(a) Beads suspension 75 μl
(b) 10 blunting buffer 5 μl
(c) dNTPs 1 mM 5 μl
(d) BSA 10 mg/ml 1 μl
(e) Blunting enzyme mix 4 μl

2. Incubate for 1 h at rt.

3. Wash the samples twice in 600 μl of HSW buffer freshly sup-
plemented with Triton X-100 0.1%.
 Genome-Wide Profiling of DNA Double-Strand Breaks by the BLESS and BLISS Methods 179

4. Resuspend the beads in the following ligation mix (volumes for

100 μl, to be adjusted proportionally depending on the sample
volume):

(a) Nuclease-free water 75 μl

(b) T4 ligase buffer 10 μl
(c) ATP 10 mM 8 μl
(d) BSA 50 mg/ml 2 μl
(e) D adapter 10 μM (Table 2) 4 μl
(f) T4 ligase 2000 U/μl 1 μl

5. Incubate the samples for 16–18 h at 16
C.

3.13 Removal 1. Wash the beads three times, each with 600 μl of HSW buffer
of Adapter Loops in freshly supplemented with Triton X-100 0.1%.
BLESS (See Note 15) 2. Resuspend the beads in the following mix (volumes for 100 μl,
to be adjusted proportionally depending on the sample
volume):
(a) Nuclease-free water 85 μl
(b) 10 I-SceI buffer 10 μl
(c) BSA 10 mg/ml 1 μl
(d) I-SceI 4 μl

3. Incubate for 4 h at 37
C.
4. Spin the beads for 5 min at 15–20,000  g.
5. Transfer as much supernatant as possible to a new 1.5 ml tube.
6. Store the samples at 20
C if not immediately used for PCR.
3.14 PCR 1. For each sample, prepare the following PCR mix (volumes for
Amplification of 50 μl, to be adjusted proportionally, depending on the sample
Captured DSBs in volume):
BLESS (See Note 16)
(a) I-SceI digested DNA 23 μl
(b) Primer P1 and P2 10 μM (Table 2) 1 μl
each
(c) NEBNext® high-Fidelity 2 PCR 25 μl
master mix

2. In a fast-cycling PCR thermo-cycler perform the following

steps:
(a) 98
C, 30 s.
(b) 98
C, 10 s.
(c) 60
C, 30 s.
180 Reza Mirzazadeh et al.

(d) 72
C, 30 s.
(e) 4
C, pause.
3. Repeat steps (b)–(d) for 18–20 times depending on the initial
sample’s size.
4. Purify PCR products using silica-based columns or alcohol
precipitation.
l Checkpoint: check PCR product size by agarose gel electro-
phoresis or Bioanalyzer.

3.15 Adapter 1. Digest the purified PCR products in the following mix
Cleavage and Library (volumes for 100 μl, to be adjusted proportionally depending
Preparation in BLESS on the sample volume):
(See Note 17)
(a) Purified PCR 85 μl
®
(b) 10 CutSmart buffer 10 μl
(c) XhoI 10 U/μl 5 μl

2. Incubate the samples 16–18 h at 37
C.

3. Purify the PCR products using silica-based columns or alcohol
precipitation.
4. Measure DNA concentration and prepare a sequencing library
using the NEBNext DNA Library Prep Master Mix Set for
Illumina, according to the manufacturer’s instructions.

3.16 Selective Linear 1. After sonication, concentrate the samples using AMPure XP
Amplification of DSBs beads and elute DNA in 10 μl of nuclease-free water.
in BLISS (See Note 18) l Checkpoint: run 1 μl of sample on Bioanalyzer using a High
Sensitivity DNA chip.
2. Prepare the following IVT mix using the MEGAscript® T7
Transcription Kit:
(a) Purified DNA 7.5 μl
(b) rNTPs (premixed in equal volumes) 8 μl
(c) 10 T7 polymerase buffer 2 μl
(d) T7 RNA polymerase 2 μl
(e) RiboSafe RNase inhibitor 40 U/μl 0.5 μl

3. Incubate for 14–16 h at 37
C.

3.17 Removal of 1. To each sample add 1 μl of DNAseI.

Genomic DNA and 2. Incubate for 15 min at 37
C.
Amplified RNA (aRNA)
3. Purify the aRNA using Agencourt RNAClean XP beads or
Purification in BLISS
silica-based columns.
(See Note 19)
 Genome-Wide Profiling of DNA Double-Strand Breaks by the BLESS and BLISS Methods 181

4. Elute the aRNA in 6 μl nuclease-free water.

l Checkpoint: run 1 μl of sample on Bioanalyzer using a RNA
6000 Pico chip.

3.18 Library 1. 30 Illumina adapter ligation.

Preparation in BLISS (a) On ice, dilute the RA3 adapter 1:5 in nuclease-free water.
(See Note 20)
(b) Add 1 μl of diluted RA3 to 5 μl of purified aRNA.
(c) In a PCR thermo-cycler, incubate the sample for 2 min at
70
C, and then immediately place it on ice.
(d) Add 4 μl of the following mix:

10 RNA ligase buffer 2 μl

RNaseOUT™ 1 μl
T4 RNA ligase truncated 1 μl

(e) In a PCR thermo-cycler with the lid open, incubate the

sample for 1 h at 28
C.
(f) Place the sample on ice, and add 3.5 μl of nuclease-free water.
2. Reverse transcription (1st strand synthesis).
(a) On ice, prepare a dilution of dNTPs at 12.5 mM in
nuclease-free water.
(b) Add 1 μl of RTP primer to the RNA sample.
(c) In a PCR thermo-cycler, incubate the sample for 2 min at
70
C and then immediately bring back the sample on ice.
(d) Add 5.5 μl of the following mix prepared using the Super-
Script IV First-Strand Synthesis System:

1st strand buffer 5 2 μl

dNTPs 12.5 mM 0.5 μl
DTT 100 mM 1 μl
RNaseOUT™ 1 μl
SuperScript® IV 1 μl

(e) In a PCR thermo-cycler with the lid at 50
C, incubate the

sample for 60 min at 50
C.
3. Library indexing and amplification.
(a) Transfer 20 μl of cDNA sample into a PCR tube on ice.
(b) Add 2 μl of the desired RPI indexing primer (Table 2), and
then 28 μl of the following mix:

Nuclease-free water 1 μl
RP1 primer 2 μl
NEBNext 2 PCR master mix 25 μl
182 Reza Mirzazadeh et al.

(c) In a fast-cycling PCR thermo-cycler perform the following

steps:
98
C, 30 s.
98
C, 10 s.
60
C, 30 s.
72
C, 30 s.
4
C, pause.
(d) Repeat steps (b)–(d) for 14–16 times depending on the
initial sample’s size.
4. Library purification.
(a) Transfer the PCR sample into a 0.5 ml LoBind tube.
(b) Add 40 μl of AMPure XP bead suspension (pre-warmed
at rt. for 30 min), and then mix thoroughly by pipetting
up-down five to six times.
(c) Incubate the sample for 7 min at rt.
(d) Place the sample onto a magnetic stand and wait 2–3 min
until all the beads have attached to the magnet.
(e) With a 200 μl pipette, aspirate the supernatant.
(f) Wash the beads once with 200 μl of ice-cold EtOH 80%,
with the sample still on the magnetic stand.
(g) Aspirate the supernatant.
(h) Repeat the wash once.
(i) Air-dry the beads for 7 min at rt.
(j) Remove the sample from the magnetic stand, and resus-
pend the beads in 20 μl of nuclease-free water.
(k) Incubate for 5 min at rt.
(l) Place the sample on the magnetic stand, and wait 2–3 min
until all the beads have attached to the magnet.
(m) Transfer 18 μl of the cleared solution into a 1.5 ml LoBind
tube.
(n) Store the library at 20
C.
l Checkpoint: run 1 μl of sample on Bioanalyzer using a
High Sensitivity DNA chip.

4 Notes

1. Using Proteinase K to purify nuclei. This step requires opti-

mization for every cell type in order to avoid over-digestion.
For example, in our original paper we found that incubation of
HeLa cells for 7–8 min at 37
C in the presence of Proteinase K
yields purified nuclei that retain their morphology and form a
visible pellet upon centrifugation. Other cell types (such as
Genome-Wide Profiling of DNA Double-Strand Breaks by the BLESS and BLISS Methods 183

U2OS and mouse lymphocytes) required shorter incubation

times (< 5 min). However, since this renders downstream
sample manipulation challenging (the nuclei pellet becomes
almost transparent after Proteinase K digestion), and since we
have found that it does not add a significant benefit (unpub-
lished observations), this step can be safely omitted in BLESS.
We also note that in BLISS, since permeabilization and in situ
reactions are performed in cells attached onto a microscope
slide or coverslip, incubation with Proteinase K is skipped in
order to avoid the detachment of cells from the glass surface.
2. Design and preparation of BLESS and BLISS adapters. A
scheme of BLESS and BLISS adapters is shown in Table 2.
BLESS adapters consist of single, self-annealing oligos that
form a hairpin-like structure to prevent the formation of
head-to-tail concatemers. The proximal (P) oligo (i.e., the
one that is in situ ligated to DSBs) is synthesized to have a
single biotin group attached to the loop and is purified using
HPLC, while the distal (D) oligo is purified by standard desalt-
ing. Recently, a modified adapter design consisting of a hairpin-
forming oligo conjugated with two biotin moieties was used in
END-seq [16]. Self-annealing is carried out by incubating the
oligo at a concentration of 10 μM in 1 T4 ligase buffer for
5 min at 95
C (typically, in a PCR thermo-cycler), followed by
immediate cooling down by placing the sample on ice. In the
original design, the proximal BLESS adapter contained a bar-
code sequence tagging the site of ligation, an XhoI cutting site,
and an I-SceI recognition site [12]. The latter was used to
release the captured biotinylated DSB fragments from strepta-
vidin beads, while XhoI was used to remove the I-SceI
sequence before the final library preparation strep. A disadvan-
tage of this design is that, after XhoI cleavage, the remaining
portion of the proximal adapter (i.e., where the DSB was
0 0
ligated) can be ligated to either the 5 or the 3 Illumina adapter.
Paired-end sequencing is then preferably used in order to
maximize the number of DSB ends detected given a certain
sequencing depth. This limitation was recently mitigated in
DSBCapture [15] using an improved BLESS adapter design.
In DSBCapture Y-shaped adapters similar to the ones routinely
used DNA library preparation for whole-genome sequencing is
used instead, enabling sequencing of only the DSB ends by
single-end sequencing. In contrast to BLESS, where the proxi-
mal adapter is used to selectively capture the genomic DNA
sequence surrounding the DSBs using affinity purification on
streptavidin beads, BLISS adapters have been designed to
enable selective DSB amplification using in vitro transcription
by the T7 RNA polymerase/promoter system. In addition to
the T7 promoter sequence, another special feature of BLISS
184 Reza Mirzazadeh et al.

over BLESS adapters is that they contain a sequence of random

nucleotides (so far we have successfully used adaptors contain-
ing either 8 or 12 random bases) that is used as a UMI during
processing of sequencing reads. In addition, the adapter con-
tains the same RA5 adapter sequence included in Illumina
TruSeq Small RNA library preparation kit, therefore making
single-end sequencing sufficient to identify the original in situ
ligation location. BLISS adapters are obtained by annealing a
forward oligo with a reverse complementary oligo, each pur-
ified by standard desalting (Table 2).
3. Attachment of cells and fixation. In BLESS, adherent cell
cultures are first trypsinized and washed in PBS, and then
fixation is performed in suspension inside 15 ml tubes (for up
to 107 cells) rotating at room temperature. In the original
BLESS protocol, cells are fixed in formaldehyde 2% in culture
medium for 30 min at rt., followed by a quenching step of
5 min in glycine 125 mM. However, we have now shortened
the fixation procedure and only use methanol-free PFA instead
of formaldehyde. Fixation in PFA 4% for 10 min at rt. is
sufficient for both BLESS and BLISS to stabilize chromatin
while enabling DSB accessibility for in situ reactions. For
BLISS in tissues, we have developed two approaches: (1) Tissue
cryopreservation and sectioning. Freshly extracted tissue biop-
sies are first fixed in PFA 4% for 1 h at rt., followed by three
washes in 1 PBS, immersion first in sucrose 15% and then in
sucrose 30% until the tissue sinks. Biopsies are finally embedded
in OCT medium and quickly frozen at 20
C before section-
ing. OCT blocks are cryosectioned into 30 μm-thick slices,
which are then mounted onto microscope slides, dried for
60 min at rt., and then stored at 4
C before further processing.
To reduce reagent use and facilitate sample processing, all steps
from lysis until in situ ligation are performed inside Secure-
Seal™ Hybridization Chambers mounted onto the slide. (2)
Preparation of nuclei suspensions (adapted from [26]). In the
second approach, fresh tissue biopsies are cut into small pieces
and transferred into a volume of NSB inside a 1.5–2 ml tube
(note that the volume of NSB buffer must be adjusted depend-
ing on the biopsy size in order to cover all the tissue frag-
ments). The sample is incubated for 15–40 min until the
tissue fragments become transparent. At this point, examining
a small volume of solution under a bright-field microscope shall
reveal the presence of nuclei in suspension (DAPI staining and
examination under a fluorescence microscope can also be done
to further assess the quality of the nuclei suspension). Nuclei
are then pelleted by centrifugation for 5 min at 500  g and
resuspended in 200–500 μl of 1 PBS. 100 μl of nuclei sus-
pension are dispensed onto a 13 mm PLL-coated coverslip and
Genome-Wide Profiling of DNA Double-Strand Breaks by the BLESS and BLISS Methods 185

incubated for 10 min at rt. 100 μl of PFA 8% 1 PBS are then

gently added and incubated for 10 min at rt., followed by
2 washes in 1 PBS at rt. The samples can be stored in 1
PBS at 4
C up to 1 month before performing BLISS.
4. Storage of fixed samples. Fixed cells can be stored (in tubes in
BLESS or on coverslips in BLISS) in 1 PBS at 4
C for several
days up to a month. To avoid bacterial contamination, NaN3
can be added to the solution to 0.05% final concentration.
5. Cell permeabilization. In BLESS, all the permeabilization
steps are typically done in rotating tubes (max 106 cells per
15 ml tube), and then the cell pellet is transferred into a
1.5–2 ml tube for washings before in situ blunting. Equilibra-
tion at rt. before shifting from buffer LB1 to LB2 is needed to
avoid precipitation of SDS contained in LB2. In the original
BLESS protocol cells were incubated in LB1 for 1 h and in LB2
for 45 min. However, we now routinely perform 1 h incuba-
tions in LB1 and LB2 buffers in both BLESS and BLISS. In
BLISS, cells fixed on coverslips are typically permeabilized and
washed inside 6- or 12-well plates, depending on their size (for
6-well plates, use at least 1.5 ml of buffer per well).
6. Washes. Following permeabilization and in between subsequent
in situ reactions, both the BLESS and BLISS samples should be
washed with a buffer that will not interfere with the next reac-
tion, if not completely removed. In the original BLESS protocol
NEBuffer 2 is used after permeabilization, digestion with Pro-
teinase K, and in situ blunting steps. In addition, in situ blunting
and ligation were each preceded by one equilibration step in 1
blunting and 1 T4 ligase buffer, respectively. In BLISS, NEB-
uffer 2 has been substituted with CutSmart® buffer, which was
meanwhile introduced by New England Biolabs as a universal
buffer for many of their enzymes. Moreover, the equilibration
step in blunting buffer has been removed in BLISS. Washing
volumes depend on the assay type and number of cells. In
BLESS, a pellet containing up to 5  106 cells is washed in a
1.5–2 ml tube with 200 μl of buffer per 106 cells. In BLISS,
approximately 3  105 cells on a 13 mm circular coverslip can be
washed inside a 6-well plate containing 2 ml of 1 CutSmart® of
buffer per well, and shaking to avoid cell drying. For BLISS in
tissue sections, the washes are done by pipetting in and out a
volume of buffer equal to the volume of the Secure-Seal™
chamber covering each section.
7. In situ DSB blunting. In both BLESS and BLISS, the blunting
step is needed to convert the DSBs that have been resected
during homologous recombination repair as well as the breaks
that lack a 50 phosphate group, into a configuration that can be
ligated by a universal oligo adapter. In BLESS, a pellet containing
186 Reza Mirzazadeh et al.

up to 5  106 cells is incubated with 100 μl of blunting mix. In

contrast, for BLISS on coverslips, the in situ blunting procedure
is performed by dispensing 50 μl of blunting mix per sample (for
a 13 mm circular coverslip) on a piece of Parafilm placed inside a
cell culture dish containing a piece of wipe tissue presoaked in
distilled water and wrapped all around the dish edge to prevent
drying. Coverslips are then placed upside-down onto the blunt-
ing mix droplet, and the dish is sealed with Parafilm during the
time of incubation. Alternatively, in situ blunting can also be
performed directly in a multi-well plate (with cells either on
coverslips or directly attached to the plastic) by adjusting the
volumes depending on the well size. Although more costly, this
approach is particularly suited when a large number of samples
need to be processed in parallel (for example, when screening for
off-target DSBs of multiple CRISPR guide RNAs).
8. In situ A-tailing. Addition of a 30 A-tail to blunted DSBs is
expected to increase the efficiency of ligation by providing short
sticky ends. We have performed comparison experiments using
BLISS to detect Cas9-induced DSBs, and demonstrated that A-
tailing increases the number of off-targets that are detected at the
same sequencing depth. A-tailing step is also included in END-
seq and DSBCapture [17, 18]. In BLESS, an optional A-tailing
step may also be included after in situ blunting by incubating a
pellet of up to 5  106 nuclei with 100 μl of A-tailing mix.
9. In situ DSB ligation. Similar to blunting, in situ ligation may
be performed in tubes, on Parafilm or directly in multi-well
plates depending on the assay type. In BLESS, a pellet contain-
ing up to 106 nuclei is incubated with 50 μl of ligation mix. For
BLISS on coverslips, the in situ ligation procedure is performed
by dispensing 50 μl of ligation mix (for a 13 mm circular
coverslip) on a piece of Parafilm placed inside a cell culture
dish containing a piece of wipe tissue presoaked in distilled
water and wrapped all around the dish edge to prevent drying.
Compared to the original BLESS protocol, in both BLESS and
BLISS we now complement the ligation mix with extra BSA
and ATP to increase the efficiency of in situ, similar to the
ligation step in HiC chromosome conformation capture [27].
In all cases, the reaction is carried out at 16
C overnight, with
the goal of ligating as many DSBs as possible. However, we
have not determined the effect of shorter incubation times and
higher temperature on the efficiency of in situ ligation.
10. Removal of unligated adapters. In both BLESS and BLISS,
removal of adapters that have not been ligated to DSBs and of
head-to-head adapter dimers is critical to avoid contamination
of the final library by short adapter sequences, which results in
poor mappability of sequenced reads. In the original BLESS
protocol, the nucleus pellet remaining after in situ ligation is
Genome-Wide Profiling of DNA Double-Strand Breaks by the BLESS and BLISS Methods 187

washed three times with HSW buffer containing NaCl 1 M, by

pipetting the nucleus suspension (200 μl of HSW buffer per
106 nuclei) thoroughly up and down for five to six times at
room temperature. In BLISS, washing cells on coverslips or in
culture plates with the same NaCl concentration and duration
is insufficient, likely because many adapters stick to the posi-
tively charged surface of glass and plastic (especially when PLL
coating is used). Instead, cells are thoroughly washed over a
period of 3 h in HSW buffer containing NaCl 2 M at 37
C,
while shaking (at least three washes, 1 h each). This should
remove most of the unligated adapters and adapter dimers.
However, the washing efficiency needs to be determined
empirically for each surface and coating type. For BLISS in
tissue sections, after the in situ ligation step, the hybridization
chamber covering the tissue is removed, and the slide is washed
with HSW in a washing jar.
11. Extraction and purification of genomic DNA. Once DSBs
have been ligated in situ, genomic DNA can be extracted even
though the extraction procedure causes artificial breaks. In the
original BLESS protocol, DNA is extracted by digesting
washed nuclei in the presence of Proteinase K followed by the
precipitation of genomic DNA with isopropanol. Alternatively,
commercial kits for genomic DNA extraction can be used. In
BLISS, since cells are adherent onto a solid surface, they have
to be manually scraped off and transferred into tubes for con-
venient downstream processing. Typically, coverslips are
quickly rinsed with nuclease-free water, air-dried, and trans-
ferred onto a piece of Parafilm with cells facing upward. For a
coverslip of 13 mm, a digestion mix containing 5 μl of Protein-
ase K in 100 μl of digestion buffer (SDS 1%, NaCl 100 mM,
EDTA 50 mM, Tris–HCl 10 mM, pH 8) is deposited onto it,
and the cells are manually scraped off using a sterile cell scraper.
The solution is then transferred into a 1.5–2 ml tube and
incubated for 16–18 h at 55
C in a thermo-shaker. For sam-
ples consisting of a few thousand cells, it is advisable to use Lo-
bind tubes to reduce DNA adsorption to the tube’s walls. DNA
can then be extracted with phenol-chloroform and precipitated
with isopropanol or ethanol in the presence of glycogen
(0.05–1 μg/μl final conc.), or column-based purification kits.
When cells are grown directly in multi-well plates, the digestion
mix is dispensed and the cells are scraped directly inside each
well. In both BLESS and BLISS, purified genomic DNA
should be dissolved in nuclease-free TE buffer.
12. DNA sonication. In both BLESS and BLISS, genomic DNA
must be fragmented to prevent biases due to different fragment
size during enrichment of the sequences surrounding DSBs by
affinity purification (in BLESS) or IVT (in BLISS). In the
188 Reza Mirzazadeh et al.

original BLESS protocol, we had to apply both enzymatic

digestion with HaeIII and sonication, since with the available
sonication device we could not generate reproducible fragmen-
tation patterns. However, this approach is suboptimal, not only
because more steps and reagents are needed, but also because
the genomic distribution of HaeIII cut sites can interfere with
the detection of DSBs at locations with high density of HaeIII
cut sites (for example, inside gene promoter regions) or where
HaeIII sites and DSBs would be too close (for instance, in
CRISPR applications HaeIII cannot be used since the recogni-
tion site CCGG also identifies the PAM sequence to which the
endonuclease binds). With the widespread diffusion of pro-
grammable sonication devices such as Bioruptor® Plus and
Covaris, we have now abandoned the use of HaeIII both in
BLESS and BLISS, and only apply sonication. Table 3 shows
examples of sonication settings that we routinely use. Follow-
ing sonication, the sample is concentrated to allow the highest
amount possible of DNA to be used as input for streptavidin
capture in BLESS or IVT in BLISS (see Notes 14 and 19).
13. Affinity capture of DSBs on streptavidin beads. In BLESS,
streptavidin beads are used to enrich for the genomic sequences
surrounding the DSBs that were previously in situ ligated with
biotinylated adapters. During the development of the method,
we tested different types of commercially available streptavidin
beads (Dynabeads® M-270, M-280, MyOne™ C1), and found
that C1 beads were very specific and had low unspecific binding
to non-biotinylated DNA. Per 5 μl of beads, we typically use
20 μg of genomic DNA as input (corresponding to approx. 3.3
million diploid genome equivalents), but this amount can be
lowered depending on initial sample size.
14. Ligation of distal adapters in BLESS. In the original BLESS
protocol, a distal adapter (D, see Table 2) is ligated to the free
extremity of captured DSB fragments attached onto streptavidin
beads. However, as already discussed in Note 2, this procedure
requires downstream paired-end sequencing to maximize the
number of ligated DSB ends that can be retrieved. We now
recommend the Illumina 30 adapter sequence to be incorporated
in the D adapter by replacing the I-SceI sequence with it.
15. Removal of adapter loops in BLESS. In the original BLESS
protocol, after distal adapter ligation DSB fragments are
released from streptavidin beads using the I-SceI homing
endonuclease before PCR is performed (see Note 17). Alterna-
tively, PCR can be performed directly on the beads, even
though based on our experience this procedure is less efficient.
16. PCR amplification of captured DSBs in BLESS. In BLESS,
DSB fragments captured by streptavidin must be amplified to
Genome-Wide Profiling of DNA Double-Strand Breaks by the BLESS and BLISS Methods 189

yield a quantity of DNA sufficient to prepare a sequencing

library. In the original protocol, we subjected samples derived
from 5–10 million cells to 18–20 cycles of amplification in order
for the PCR product to be detectable by agarose gel electropho-
resis and ethidium bromide staining. The number of cycles can
be reduced to minimize the amount of PCR duplicates, by using
more sensitive assays such as Agilent’s Bioanalyzer. However,
the optimal number of cycles will still vary depending on the
sample type and amount, and needs to be empirically adjusted.
17. Adapter cleavage and library preparation in BLESS. In the
original BLESS protocol PCR-amplified DSB fragments are
digested with XhoI to cleave off both the proximal and distal
adapter sequences (with the exception of barcodes marking the
proximal and distal ligation position), and the resulting DNA is
finally used to prepare a sequencing library. However, this poses
two major limitations. First, the enzyme may cut inside certain
fragments, producing DNA fragments that only contain either
the proximal or distal barcode sequence (this will reduce the
number of DSB ends effectively identified at a given sequencing
depth). Second, as already discussed in Note 2, since sequencing
adapters are introduced after PCR, the original directionality of
ligation is not preserved, and paired-end sequencing must be
performed to maximize the number of proximal adapter ligation
sites that are sequenced. This increases costs and renders the
procedure very time-consuming. Thus, as recommended above,
the original BLESS adapter design should be modified to
include the P5 and P3 Illumina adapter sequences.
18. Selective linear amplification of DSBs in BLISS. In BLISS,
the DNA sequence surrounding the labeled DSBs is selectively
amplified using IVT driven by the T7 RNA polymerase. Inde-
pendently of the amount of input, we usually perform IVT for
14–16 h, although the incubation time may be shortened if the
input is large enough (> 105 genome equivalents). For multi-
plexing, DNA purified from multiple samples that have been in
situ labeled with different barcodes can be premixed and used
as input for a single IVT reaction. Depending on the concen-
tration of input DNA and the level of multiplexing, the final
IVT reaction volume may be adjusted by proportionally scaling
all the volumes.
19. Removal of genomic DNA and purification of amplified RNA
(aRNA) in BLISS. After IVT, genomic DNA is digested with
DNAseI to avoid carry-over of DNA into the final library. How-
ever, even without this step we have managed to obtain high-
quality libraries, and thus we recommend it only in the case of high
DNA inputs to the IVT reaction (> 105 genome equivalents).
Purification of aRNA generated by IVT can be done using
190 Reza Mirzazadeh et al.

Agencourt RNAClean XP beads or silica-based columns (for

example, RNeasy® MinElute® Cleanup kit, Qiagen, cat. no.
74204).
20. Library preparation in BLISS. A custom-modified protocol
of Illumina TruSeq Small RNA Sample Preparation Kit is used
to prepare BLISS libraries. The protocol has been adapted to
account for low input as well as for the fact that the RA5
adapter sequence is already present in the aRNA. Purchase of
the kit is not necessary, since all the needed components (RA3
adapter and primers, see Table 2) can be purchased separately
(for example, we used Integrated DNA Technologies Inc., and
purify the RA3 adapter by HPLC, and RTP, RP1, and RPI
primers by standard desalting). T4 RNA ligase, RNaseOUT™,
SuperScript® IV are not part of the TruSeq Small RNA Sample
Preparation Kit and are purchased separately (see Reagents). As
in BLESS, the optimal number of PCR cycles depends on the
input size (initial amount of cells or number of genome equiva-
lents used as input in the IVT reaction) and expected number
of DSBs per cell. Typically, for samples of 103–104 cells we
perform 12–16 amplification cycles, but this number needs to
be empirically adjusted to limit the number of PCR duplicates.

5 Additional Considerations

5.1 Choice of Both the BLESS and BLISS protocols described here terminate
Sequencing Platform with the preparation of sequencing libraries that can be sequenced
and Sequencing Depth on several Illumina platforms, including MiSeq, NextSeq, and
HiSeq (Fig. 1). In the original BLESS paper [12], we also used
the Roche 454 system, but since this platform is no longer com-
mercially available and because of the widespread availability of
Illumina platforms, we now only use the latter. The platform choice
depends on the desired sequencing throughput, which ultimately
will be dictated by the sample size (number of cells or genome
equivalents) and by the number of DSBs per cell. Based on our
experience, for a sample of approx. 105 human or mouse cells
processed by BLISS, a sequencing depth of 40–50 million reads
per sample is typically sufficient to identify genomic regions span-
ning 50–100 kb with significant enrichment in DSBs probability, by
comparing a sample treated with a DSBs-inducing condition with a
control. Certain applications, such as identifying rare CRISPR off-
target events, might require a higher depth, and in general the
optimal sequencing depth must be empirically determined based
on the resolution needed.

5.2 Data Analysis There is currently no ready-to-use software package for the analysis of
BLESS and BLISS data, and pipelines similar to the one which we
have developed for BLISS (Fig. 2) must be developed in house.
 Genome-Wide Profiling of DNA Double-Strand Breaks by the BLESS and BLISS Methods 191

BCL2FASTQ SOFTWARE

GENERATE
FASTQ

FASTQC SOFTWARE

QUALITY
CONTROL

CUSTOM SCRIPTS

- 1 mismatch allowed in UMI and barcode FILTER R1

CUSTOM SCRIPTS
- Extract genomic sequence from R1 reads
with UMI and barcode TRIM
FILTERED R1

BWA SOFTWARE
- Align extracted genomic sequence to
reference genome ALIGN
TRIMMED R1

CUSTOM SCRIPTS
- Filter for desired mapping quality
- Append UMIs to mapped reads FILTER
MAPPED R1

CUSTOM SCRIPTS

FILTER UMIs

CUSTOM SCRIPTS

GENERATE
BED

Fig. 2 Sequence reads processing workflow in BLISS. The BCL file obtained after sequencing on an Illumina
platform (MiSeq, NextSeq or HiSeq) is first converted into a FASTQ file and checked for quality using available
192 Reza Mirzazadeh et al.

Following read filtering and alignment to the reference genome,

genomic regions with higher breaking probability (compared to the
rest of the genome or to the same region in a different sample) can be
identified using different approaches. In the original BLESS paper
[12], we applied the hypergeometric probability distribution to
compute the statistical significance of differences in DSBs counts
between corresponding sliding genomic windows in a control sample
versus a sample treated with aphidicolin to induce replication stress.
However, depending on the size of the sliding window, this approach
is computationally intensive. Furthermore, since additional methods
such as the Bonferroni method [28] must be implemented to correct
for multiple hypothesis testing, this approach is rather time-
consuming. In alternative, nearest neighbor clustering algorithms
can be used to identify DSBs hotspots. However, while this approach
can robustly identify DSBs clusters of a given expected size and shape,
such as those induced by CRISPR/Cas9 [13, 14], tailoring it to
identify clusters of mixed DSBs forming through different processes
(e.g., replication stress, transcription, etc.) and repaired by different
mechanisms (NHEJ vs. homologous recombination) is not trivial.
Peak calling algorithms can also be applied, although they were
originally developed for ChIP-seq analysis and are not tailored for
the type of data obtained with BLESS or BLISS. For example,
MACS2 [29] was recently used to analyze DSBCapture data, while
HOMER [30] and SICER [31] were used in END-seq. Finally, a key
advantage in BLISS data analysis is that, thanks to the use of UMIs,
single nucleotide locations that break recurrently in multiple cells can
be robustly identified without the need for clustering algorithms or
peak calling tools. Indeed, simply by filtering genomic locations
based on the number of UMIs associated with each location, we
have been able to identify etoposide-induced DSBs hotspots as well
as recurrent breaking sites in untreated cells [17].

Fig. 2 (continued) software. Afterward, a series of custom scripts written in Unix are used to select the R1
reads that start with the correct pattern of 8 nt UMI followed by the proper barcode. At this point, the genomic
sequence downstream of the barcode is aligned to a reference genome using BWA. Afterward, again using
custom scripts in Unix, the original UMI is associated with the corresponding aligned read (after removing
reads mapping to repetitive regions), and a filter is applied to remove reads mapped to the same or very near
genomic location and having the same UMI (IVT and PCR duplicates). Finally, a BED file is constructed,
containing information about the genomic coordinate of each DSB event labeled by a unique UMI tag
 Genome-Wide Profiling of DNA Double-Strand Breaks by the BLESS and BLISS Methods 193

References
1. Jackson SP, Bartek J (2009) The DNA-dam- 16. Canela A, Sridharan S, Sciascia N et al (2016)
age response in human biology and disease. DNA breaks and end resection
Nature 461:1071–1078 measured genome-wide by end sequencing.
2. van Gent DC, Hoeijmakers JH, Kanaar R Mol Cell 63:898–911
(2001) Chromosomal stability and the 17. Yan WX, Mirzazadeh R, Garnerone S et al
DNA double-strandedbreak connection. Nat (2017) BLISS is a versatile and quantitative
Rev Genet 2:196–206 method for genome-wide profiling of DNA
3. McKinnon PJ, Caldecott KW (2007) DNA double-strand breaks. Nat Commun 8:15058
strand break repair and human genetic disease. 18. Kim D, Bae S, Park J et al (2015) Digenome-
Annu Rev Genomics Hum Genet 8:37–55 seq: genome-wide profiling of CRISPR-
4. Porteus MH, Carroll D (2005) Gene targeting Cas9 off-targeteffects in human cells. Nat
using zinc finger nucleases. Nat Biotechnol Methods 12:237–243. 1 p following 243
23:967–973 19. Kim D, Kim J, Hur JK et al (2016) Genome-
5. Joung JK, Sander JD (2013) TALENs: a widely wide analysis reveals specificities of Cpf1 endo-
applicable technology for targeted genome nucleases in human cells. Nat Biotechnol
editing. Nat Rev Mol Cell Biol 14:49–55 34:863–868
6. Hsu PD, Lander ES, Zhang F (2014) Devel- 20. Tsai SQ, Zheng Z, Nguyen NT et al
opment and applications of CRISPR-Cas9 for (2015) GUIDE-seq enables genome-
genome engineering. Cell 157:1262–1278 wide profiling of off-targetcleavage
7. Zetsche B, Gootenberg JS, Abudayyeh OO by CRISPR-Cas nucleases. Nat Biotechnol
et al (2015) Cpf1 is a single RNA-guided endo- 33:187–197
nuclease of a class 2 CRISPR-Cas system. Cell 21. Kleinstiver BP, Tsai SQ, Prew MS et al
163:759–771 (2016) Genome-wide specificities of CRISPR-
8. Tsai SQ, Keith Joung J (2016) Defining and Cas Cpf1 nucleases in human cells. Nat Bio-
improving the genome-wide specificities technol 34:869–874
of CRISPR-Cas9nucleases. Nat Rev Genet 22. Wang X, Wang Y, Wu X et al (2015) Unbiased
17:300–312 detection of off-target cleavage by CRISPR-
9. Szilard RK, Jacques P-E, Laramée L et al Cas9 and TALENs using integrase-defec-
(2010) Systematic identification of fragile sites tive lentiviral vectors. Nat Biotechnol
via genome-wide location analysis of gamma- 33:175–178
H2AX. Nat Struct Mol Biol 17:299–305 23. Frock RL, Hu J, Meyers RM et al
10. Iacovoni JS, Caron P, Lassadi I et al (2015) Genome-wide detection of DNA dou-
(2010) High-resolution profiling of gamma- ble-stranded breaks induced by engineered
H2AX around DNA double strand breaks in nucleases. Nat Biotechnol 33:179–186
the mammalian genome. EMBO J 24. Wei P-C, Chang AN, Kao J et al (2016) Long
29:1446–1457 neural genes harbor recurrent DNA break clus-
11. Madabhushi R, Gao F, Pfenning AR et al ters in neural stem/progenitor cells. Cell
(2015) Activity-induced DNA breaks govern 164:644–655
the expression of neuronal early- 25. Kivioja T, V€ah€arautio A, Karlsson K et al
response genes. Cell 161:1592–1605 (2012) Counting absolute numbers of mole-
12. Crosetto N, Mitra A, Silva MJ et al cules using unique molecular identifiers. Nat
(2013) Nucleotide-resolution DNA double- Methods 9:72–74
strand break mapping by next-genera- 26. Baslan T, Kendall J, Rodgers L et al
tion sequencing. Nat Methods 10:361–365 (2012) Genome-wide copy number analysis of
13. Ran FA, Cong L, Yan WX et al (2015) In vivo single cells. Nat Protoc 7:1024–1041
genome editing using Staphylococcus aureus 27. Naumova N, Imakaev M, Fudenberg G et al
Cas9. Nature 520:186–191 (2013) Organization of the mitotic chromo-
14. Slaymaker IM, Gao L, Zetsche B et al (2016) some. Science 342:948–953
Rationally engineered Cas9 nucleases with 28. Dunnett CW (1955) A multiple comparison
improved specificity. Science 351:84–88 procedure for comparing several treatments
15. Lensing SV, Marsico G, H€ansel-Hertsch R et al with a control. J Am Stat Assoc 50:1096–1121
(2016) DSBCapture: in situ capture and 29. Zhang Y, Liu T, Meyer CA et al (2008) Model-
sequencing of DNA breaks. Nat Methods based analysis of ChIP-Seq (MACS). Genome
13:855–857 Biol 9:R137
194 Reza Mirzazadeh et al.

30. Heinz S, Benner C, Spann N et al (2010) Sim- 31. Zang C, Schones DE, Zeng C et al (2009) A
ple combinations of lineage-determining tran- clustering approach for identification of
scription factors prime cis-regulatory elements enriched domains from histone modifica-
required for macrophage and B cell identities. tion ChIP-Seq data. Bioinformatics
Mol Cell 38:576–589 25:1952–1958
 Chapter 15

DNA Replication Profiling Using Deep Sequencing

Xanita Saayman, Cristina Ramos-Pérez, and Grant W. Brown

Abstract
Profiling of DNA replication during progression through S phase allows a quantitative snap-shot of
replication origin usage and DNA replication fork progression. We present a method for using deep
sequencing data to profile DNA replication in S. cerevisiae.

Key words DNA replication, Replication forks, Deep sequencing, Replication fork rate, Replication
origins

1 Introduction

DNA replication in eukaryotes is initiated at multiple origins of

replications in a highly defined two-step mechanism [1–4]. Briefly,
origins are “licensed” in G1 by the formation of pre-replicative
(pre-RC) protein complexes. DNA synthesis is then initiated as a
subset of these licensed origins are “fired” in a defined temporal
order throughout S phase [5]. The resulting replication forks travel
bidirectionally from each origin until encountering newly repli-
cated regions emanating from adjacent replication origins, or
until the replisome dissociates from the DNA strands. The latter
case is typically a result of aberrant DNA structures or limiting
replication factors under replication stress conditions.
Due to the inherently mutagenic nature of DNA replication,
the rate and timing at which replication is completed can influence
somatic mutation rates in both yeast and human cells [6, 7]. This
implies that temporal regulation of DNA replication is important in
preserving genomic integrity. Global replication rates are a function
of both origin firing patterns and fork progression rate. While it is
evident that eukaryotes show remarkable plasticity in both para-
meters in response to conditions such as DNA damage [8],

Xanita Saayman and Cristina Ramos-Pérez contributed equally to this work.

Marco Muzi-Falconi and Grant W. Brown (eds.), Genome Instability: Methods and Protocols, Methods in Molecular Biology,
vol. 1672, DOI 10.1007/978-1-4939-7306-4_15, © Springer Science+Business Media LLC 2018

195
196 Xanita Saayman et al.

relatively little is known about how origin activation and fork

progression are temporally regulated [4, 9].
Previous techniques for assessing replication dynamics include
density transfer methods, flow cytometry, chromatin-
immunoprecipitation (ChIP) of replication proteins or nucleotide
analogs, single-stranded DNA accumulation, and single molecule
DNA fiber analysis [10–14]. Additionally, high-density oligonucle-
otide microarrays used in DNA “copy number variation” (CNV)
procedures have been adapted to map replication origins and mon-
itor fork progression [5, 15]. However, microarray-based CNV
introduces intrinsic resolution constraints due to probe size, and
signal disruption due to nonlinear hybridization signal artifacts
[16]. CNV by deep sequencing offers an increasingly accessible
alternative for investigating genome-wide replication patterns on
a population level, and can generate replication profiles with
remarkable spatial and temporal resolution [17].
Here, we describe a method for generating DNA replication
profiles with CNV analysis from deep sequencing of synchronous S.
cerevisiae cultures to investigate replication features (e.g., origin
firing, replication fork progression). To this end, logarithmically
growing cultures are arrested in G1 with mating pheromone and
synchronously released into S phase in the presence of DNA dam-
aging agents. A variety of alternative cell synchronization techni-
ques can be applied including centrifugal elutriation, conditional
mutants, or drug treatment [18, 19]. Flow cytometric analysis is
used to assess synchronization quality and to determine median
sample DNA content for downstream normalization. Genomic
DNA is extracted from samples taken at different times after release
from G1, and sequenced with Illumina technology.
We also present a computational method adapted from Muller
et al. [17] for data extraction from sequencing reads, data proces-
sing and normalization, and replication profile visualization.
Unique reads are first aligned to a reference S. cerevisiae genome
and scaled according to the sample DNA content as determined by
peak-aligned flow cytometry profiles. This normalization technique
is applied to establish a baseline of 1.0 (as described in [17]) for an
intra-sample comparison. The number of uniquely mapped
sequences at each replicating time point is compared to that of
the non-replicating sample, and the resulting enrichment ratio is
smoothed by a least-squares polynomial digital filter to preserve
higher moments and prevent signal distortion [20]. Overlapping
replication profiles are generated to visualize origin firing and repli-
some progression.
 DNA Replication Profiling Using Deep Sequencing 197

2 Materials

2.1 Specialized 1. Qubit Fluorometer (Invitrogen).

Equipment 2. Magnetic Particle Concentrator (e.g., Invitrogen Dynal
MPCTM-S).
3. Thermal Cycler (e.g., BioRad C1000).
4. Real-Time PCR Detection System (e.g., BioRad CFX96).
5. NextSeq 500 High-Throughput Sequencing System
(Illumina).
6. 2100 Bioanalyzer (Agilent).

2.2 Kits 1. Yeast DNA Purification Kit (Epicentre MasterPure).

2. Qubit dsDNA BR Assay Kit (Thermo Fisher Scientific).
3. Nextera Amplicon DNA Sample Preparation Kit (Illumina).
4. Nextera XT Index Kit (24 indices—96 samples) (Illumina).
5. AMPure XP beads (Beckman Coulter).
6. KAPA Library Quantification Kit Illumina platforms (Com-
plete Kit - KK4824, KAPA Biosystems).
7. High Sensitivity DNA Kit (Agilent).

2.3 Plasticware 1. 1.5 mL microfuge tubes (e.g., Eppendorf DNA LoBind).

and Reagents 2. Multiplate Low-Profile 96-Well, Unskirted PCR Plates.
3. Adhesive Seals for 96-well PCR plates.
4. Alpha factor: 5 mg/mL in 95% ethanol, store at
20  C.
5. Pronase: 10 mg/mL in double-distilled H2O, always prepare
fresh.
6. 10% (w/v) Sodium Azide.
7. 1 TE buffer: 10 mM Tris pH 8.0, 1 mM EDTA.
8. 80% (v/v) Ethanol.
9. Buffer P1 Resuspension Buffer: 50 mM Tris–HCl, 10 mM
EDTA, adjust pH to 8.0, 100 μg/mL RNase A final, store at
2–8  C.
10. 1 Library Dilution Buffer: 10 mM Tris–HCl pH 8.0, 0.05%
Tween 20.
11. FACS Buffer: 200 mM Tris–Cl pH 7.5, 200 mM NaCl, 78
MgCl2.
12. 5 mL Polystyrene Round-Bottom tube (e.g., Falcon
352052).13. 2X SYBR solution: 1/5000 SYBR green in 50
mM Tris-Cl pH 7.5.

2.4 Data Analysis 1. FlowJo (http://www.flowjo.com).

Software 2. DNA sequencing files (FASTQ-formatted files).
198 Xanita Saayman et al.

3. FastQC (http://www.bioinformatics.babraham.ac.uk/pro
jects/fastqc/).
4. Bowtie (http://bowtie-bio.sourceforge.net/index.shtml).
5. matplotlib version 1.5.1 (http://matplotlib.org).
6. NumPy (http://www.numpy.org).
7. Python (http://www.python.org).
8. SciPy (http://www.scipy.org).
9. SAMtools version 1.3 (https://github.com/samtools/
samtools).
10. deepTools version 2.3.4 (https://github.com/fidelram/
deepTools).
11. Plotly (https://plot.ly).
12. bigWigToWig (https://genome.ucsc.edu/goldenpath/help/
bigWig.html).
13. wig2bed from BEDOPS (http://bedops.readthedocs.io/en/
latest/index.html).

3 Methods

3.1 Cell The experimental and analytical procedures of this protocol are
Synchronization summarized in Fig. 1.
and DNA Preparation
1. Grow cells overnight in 50 mL of liquid medium to early log
phase (OD600 of 0.2–0.4) in a water bath shaker under pre-
ferred conditions (e.g., 30  C, 200 rpm) (see Note 1).
2. Arrest the cells in G1 by adding alpha factor to 2 μg/mL final
concentration, and incubate for 1 h. Add an additional 1 μg/
mL alpha-factor and incubate for 1 h or until >90% of the cells
have the “shmoo” morphology (see Note 2). Before releasing
the cells from G1 take a sample of at least 2 OD, add sodium
azide to 0.1% (w/v) final concentration, and leave on ice for at
least 15 min (see Note 3).
3. Release the cells from G1 by adding pronase to 100 μg/mL
final concentration (see Note 4).
4. For flow cytometry, harvest at least 0.5 OD600 at the desired
time points, and immediately resuspend in 1 mL 70% EtOH.
Incubate at room temperature for 15 min or 4  C until further
processing (see Subheading 3.2).
5. For library preparations, harvest at least 2 OD600 of the cells at
the desired time points, immediately mix with sodium azide to
0.1% (w/v) final concentration, and leave on ice for at least
15 min.
 DNA Replication Profiling Using Deep Sequencing 199

EXPERIMENTAL ANALYTICAL
PROCEDURE PROCEDURE

G1 ARREST AND RELEASE SEQUENCE AND ALIGN

2X SAMPLE COLLECTION TO REFERENCE GENOME

FLOW DNA NORMALIZE TO READ DEPTH

CYTOMETRY EXTRACTION (RPKM)
RELEASE INTO S PHASE

LOG ENRICHMENT
1
GENOMIC DNA

0
1C 2C
DNA CONTENT -1

CHROMOSOME VIII COORDINATES

DNA FRAGMENTATION
AND TAGMENTATION

NORMALIZE TO DNA CONTENT

(flow cytometry)
TRANSPOSOMES™
1.5
LOG ENRICHMENT

LIBRARY AMPLIFICATION
1.0
0.5

P5
RD1 SP
INDEX 2
0

INDEX 1
P7 CHROMOSOME VIII COORDINATES
RD2 SP

DATA SMOOTHING
polynomial least-squares fitting

LIBRARY CLEAN-UP,
1.0

QUANTIFICATION AND POOLING

LOG ENRICHMENT
0.5
0

CHROMOSOME VIII COORDINATES

30 MIN 60 MIN 90 MIN

HIGH-THROUGHPUT SEQUENCING EARLY FIRING ORIGINS

Fig. 1 Workflow of the experimental and analytical procedures. Asynchronous yeast cultures are arrested in
G1 and released into S phase. DNA is extracted from G1-arrested samples as well as at the desired times
200 Xanita Saayman et al.

6. Extract genomic DNA with MasterPure Yeast DNA Purifica-

tion Kit from Epicentre or similar, according to the manufac-
turer’s instructions. The protocol can be downloaded from
http://www.epibio.com/docs/default-source/protocols/
masterpure-yeast-dna-purification-kit.pdf. Perform the RNase
A treatment included in the kit (see Note 5).
7. Resuspend the samples in 35 μL TE and determine DNA
concentration using a Qubit Fluorometer and the Qubit
dsDNA BR Assay Kit, or similar. Proceed to DNA tagmenta-
tion or store the DNA at
20  C (see Note 6).

3.2 Flow Cytometric Flow cytometric analysis measures bulk cellular DNA content at
Analysis each time sampled for copy number analysis, to verify synchroniza-
tion quality and for downstream data normalization.
1. Starting with 1 mL of ethanol-fixed cells from Subheading
3,1.4, centrifuge at 1200  g at room temperature for 2 min.
2. Remove the supernatant and wash cell pellet with 1 mL dH2O.
Centrifuge again at 1200  g at room temperature for 1 min.
3. Remove the supernatant and resuspend cell pellet in 0.5 mL
50 mM Tris-Cl pH 7.5 containing 10 mg/mL freshly prepared
proteinase K. Incubate for 40 min at 50  C.
4. Centrifuge at 1200  g at room temperature for 1 min.
Remove the supernatant and resuspend cell pellet in 0.5 mL
FACS buffer.
5. Samples can now be stored at 4  C for up to a week.
6. Transfer 100 μL samples to a 5 mL round-bottom tube.
7. Add 0.5 mL 2 SYBR solution (diluted in 50 mM Tris–Cl
pH 7.5).
8. Sonicate each sample at low intensity for 1–3 s.
9. Analyze on flow cytometer, measuring SYBR green fluores-
cence for at least 10,000 cells.

3.3 DNA This step uses Illumina’s Nextera Amplicon DNA Sample Prepara-
Tagmentation tion Kit to fragment the DNA and add adapter sequences to the
ends.
ä

Fig. 1 (continued) following release into S phase. Libraries for each sample are prepared by the addition of the
Transposome complex, which fragments the DNA and attaches transposons to the DNA 50 ends. The libraries
are amplified and barcoded with a unique combination of indexed adapters. The samples are then purified,
quantified, and normalized. Finally, the libraries are pooled together and submitted for high-throughput
sequencing. For data analysis, reads are aligned to a reference S. cerevisiae genome and normalized to
the reads per kilobase per million (RPKM). Each sample is scaled according to the median DNA content, as
determined by flow cytometry. Read counts are then compared to that of the G1-arrested (non-replicating)
sample. Resultant replication profiles are smoothed by the Savitzky-Golay filter and visualized using graphing
software
 DNA Replication Profiling Using Deep Sequencing 201

1. To prevent contamination of the samples, clean the bench

thoroughly with bleach or similar and always wear gloves.
2. In a PCR tube for each sample, add 5 μL of the extracted DNA
at 0.2 ng/μL (1 ng in total), 10 μL Tagment DNA buffer (TD
buffer), and 5 μL Amplicon Tagmentation Mix (ATM). Mix
gently.
3. Centrifuge at 1000  g at room temperature for 1 min.
4. Incubate the reaction in a thermal cycler at 55  C for 5 min and
then hold at 10  C.
5. Add 5 μL Neutralize Tagment buffer (NT buffer) to each
sample and mix gently.
6. Centrifuge at 1000  g at room temperature for 1 min.
7. Incubate the reaction at room temperature for 5 min.

3.4 Library This step uses Illumina’s Nextera XT Index Kit to amplify the
Amplification tagmented DNA by PCR, by adding unique combinations of the
adapters Index 1 (i7) and Index 2 (i5) to each library.
1. In a PCR tube, add to each sample 15 μL of Nextera PCR
Master Mix and 5 μL of each appropriate index primer (i5 and
i7 respectively) (see Note 7). If multiple libraries will be pooled
for sequencing, then each library must have a unique i7 primer
(for single-end reads) or a unique index primer combination
(for paired-end reads).
2. Centrifuge at 280  g for 1 min.
3. Place samples in a thermal cycler and heat at 72  C for 3 min,
followed by 30 s at 95  C. Next, run 12 cycles of 95  C
denaturation for 10 s, 55  C annealing for 30 s and 72  C
extension for 30 s. Allow a final extension of 72  C for 5 min
and hold at 10  C (see Note 8).
4. At this point the library can be left at 2–8  C for up to 2 days, or
proceed to library cleanup.

3.5 Library Cleanup 1. Centrifuge PCR products at 280  g for 1 min at room
temperature. Transfer volume (50 μL) to a LoBind eppendorf
tube, or similar.
2. Bring the AMPure XP beads to room temperature and vortex
for 30 s to ensure that they are evenly dispersed. Add 30 μL of
beads to each PCR product sample (0.6 volume). Mix well.
3. Incubate at room temperature without shaking for 5 min. Place
on a magnetic particle concentrator (e.g., Invitrogen Dynal
MPC™-S Magnetic Particle Concentrator) for 2 min or until
the supernatant has cleared. Discard the supernatant.
202 Xanita Saayman et al.

4. Wash beads with 200 μL fresh 80% ethanol without resuspend-

ing the beads. Incubate on a stand for 30 s or until the super-
natant has cleared. Discard the supernatant.
5. Repeat washing step with 200 μL 80% ethanol. Remove excess
ethanol with a P10 pipette and allow beads to air-dry for
15 min.
6. Remove samples from the magnetic stand and add 52.5 μL of
P1 Resuspension buffer. Mix well and incubate at room tem-
perature for 2 min.
7. Place on a magnetic stand for 2 min or until the supernatant has
cleared.
8. Transfer 48 μL of the supernatant to a fresh LoBind eppendorf
tube.
9. The samples can be stored at
20  C for up to 7 days before
sequencing. Library quantification is recommended to reduce
sequencing cost (see Subheading 3.6).

3.6 Library Accurate library quantification ensures equal representation of each

Quantification library when they are pooled prior to sequencing. In this protocol,
and Pooling DNA quantification is performed by qPCR analysis using the KAPA
Library Quantification Kit for Illumina platforms, which includes
2 KAPA SYBR FAST qPCR Master Mix, 10 Illumina Primer
Premix, and Illumina DNA Standards 1–6.
1. Make three separate 1:1000 dilutions of each library in Library
Dilution Buffer.
2. Prepare the PCR reaction mix taking into account that both the
samples and the DNA standards must have three replicates
each. Aliquot 16 μL of the PCR Reaction Mix into each well.
Add 4 μL of the corresponding DNA Standard or 4 μL of the
1:1000 dilution of each library into the appropriate well of the
96-well plate.
3. Seal the 96-well plate with an adhesive seal and centrifuge at
280  g for 1 min at room temperature.
4. Incubate in the real-time thermocycler for 5 min at 95  C and
then apply 35 cycles of denaturation at 95  C for 30 s and
extension at 60  C for 60 s.
5. Generate the standard curve by plotting the quantification
cycle values (Cq) to the DNA concentrations of the standards.
It should have an efficiency of amplification close to 100% and
R2 > 0.99. Calculate the concentration of each library using
the standard curve as a reference.
6. Analyze the samples on an Agilent 2100 Bioanalyzer using the
High Sensitivity DNA Kit to determine the size of each library,
which will be used to adjust the real concentration of each
 DNA Replication Profiling Using Deep Sequencing 203

sample. The protocol can be downloaded from http://www.

agilent.com/cs/library/usermanuals/Public/G2938-90321_
SensitivityDNA_KG_EN.pdf.
7. With the concentration corrected, normalize the libraries to
30 nM and pool them, taking into account that the lowest
volume added of the most concentrated library should be at
least 2 μL. We pool between 24 and 32 libraries, depending on
the number of reads we are anticipating from the sequencer.
8. As a quality control, the pool must be analyzed on Bioanalyzer.
Finally, perform a qPCR to determine the concentration of the
final pool, using the KAPA Library Quantification Kit.
9. Submit the pooled library for high-throughput sequencing on
the Illumina NextSeq 500 System. At least 107 reads per library
for yeast are necessary for robust analysis. Samples can be
multiplexed in a Single-Read Illumina flow cell as long as
each carries a unique Index 1 (i7) primer. Unique Index
2 (i5) primers can also be used to increase the number of
samples per flow cell, using Paired-End flow cell sequencing.
Single-end reads of 50 bp are sufficient for this application.

3.7 Measuring As described in Muller et al. [17], normalization by flow cytometry

Median DNA Content is applied to establish a baseline of log2 ¼ 1.0 for samples taken at
from Flow Cytometry different time points.
Data 1. Data processing. Using any available package for flow cytome-
try analysis (i.e., FlowJo), upload flow-cytometry standard for-
mat (FCS) files and gate appropriately to exclude debris,
aggregates, and cell doublets. FCS files can then be exported
for downstream analysis using any statistical software package
(i.e., SciPy, R).
2. Peak normalization of flow cytometry data. To control for
fluorescence readout shifts from experimental variation, align
G1 and G2 asynchronous peaks of each time sample strain to
the wild-type control. Apply the same scaling factor to all time
points for each strain.
3. Scale factor determination. Determine the median DNA con-
tent of each time sample relative to the median DNA content of
the G1-arrested cells.

3.8 Quantifying Data processing steps are conducted in a Unix shell environment.
and Visualizing Downstream data analysis is conducted in Python, but similar
Genomic Replication packages are available to run analyses through alternative software
(i.e., MATLAB, R). Input files are expected to be compressed
FASTQ files, containing all raw sequence reads.
204 Xanita Saayman et al.

1. Quality assessment. Retrieve demultiplexed FASTQ-formatted

sequencing files from the Illumina sequencing run. Perform
quality assessment using FastQC or similar software (see Note 9).
2. Read alignment. Index the reference genome in FASTA format
with the bowtie “bowtie-build” indexer function. Align FASTQ
format cells to the reference genome with the bowtie “aligner”
function. Specify output in SAM file format (see Note 10).
3. Read mapping and processing.
(a) Convert SAM files into binary compressed BAM files
using SAMtools “view” function.
(b) Sort and index BAM files using SAMtools “sort” and
“index” functions. This should be done on a per-
chromosome basis to generate individual chromosome
maps if desired. See SAMtools documentation for specify-
ing genomic regions.
(c) Determine number of uniquely mapped reads per speci-
fied bin size using the “bamCoverage” function of deep-
Tools. Normalize to Reads Per Kilobase per Million
mapped reads (RPKM) to normalize for read depth.
Extend reads to read fragment size, as specified by Illu-
mina sequencing conditions (see Note 11).
(d) Compare output bigWig file of each time sample to out-
put bigWig file of G1-arrested sample using “bigwigCom-
pare” function of deepTools. Specify “scaleFactors”
according to factors determined by flow cytometry in
Subheading 3.7 (see Note 12).
(e) For easier data manipulation, convert to BED files using
bigWigToWig executable followed by wig2bed. BED files
can still be uploaded to most genome browsers for manual
inspection or be read by statistical analysis software (i.e.,
Python, R).

3.9 Replication 1. For more readable replication profile visualization, import data
Profile Visualization from BED files and apply a polynomial data smoothing filter
(i.e., Savitsky-Golay digital filter) to reduce noise. This can be
done using the “savgol_filter” function of SciPy in Python or
similar software (see Note 13).
2. Generate overlapping chromosomal maps of smoothed data
from BED files by chromosomal coordinates using Plotly, mat-
plotlib, or similar software.
 DNA Replication Profiling Using Deep Sequencing 205

4 Notes

1. If the culture is too dense (i.e., OD600 > 0.7), dilute to an

OD600 of 0.1–0.2 and allow at least two cell doublings before
adding alpha factor.
2. It is recommended that the cell cycle arrest and release be
confirmed by flow cytometry, taking samples at steps 1, 2,
and 4.
3. Multiple samples can be left on ice for up to 2 h. The amount of
cells needed will depend on the yield of the DNA extraction
and purification, if different methods are chosen. Minimum
DNA final concentration required for the library preparation
is 0.2 ng/μL.
4. At this point, conditions can be customized, for example by
adding a drug treatment or changing the temperature.
5. Some kits specify RNase treatment during the DNA extraction.
If RNase treatment is done after the DNA extraction, an addi-
tional extraction of the DNA with phenol:chloroform followed
by ethanol precipitation is required to remove the RNase A.
6. The DNA concentration should be determined not longer
than 1 day before the library preparation, and is typically
3–4 μg. It is important to use a fluorescence-based method
rather than UV absorbance.
7. Use the recommended combination of the i5 and i7 adapters as
suggested by the Nextera XT DNA Library Prep Reference
Guide.
8. Use the recommended number of PCR cycles to ensure high-
quality results.
9. FastQC will generate an HTML file (readable with a web
browser) containing information on quality including potential
contaminations or technical biases. Full descriptions of analysis
modules as well as common reasons for errors are available at
http://www.bioinformatics.babraham.ac.uk/projects/fastqc/
Help/.
10. Reference genomes in FASTA format are available for down-
load at http://hgdownload.cse.ucsc.edu/downloads.html. If
paired-end sequencing was used, this should be specified at
this point. An alternative to this is uploading sample files to
BWA Aligner, a BaseSpace Lab App developed by Illumina.
This software will index and align FASTQ files against any
given reference file and fix read pairing information and flags.
SAM files are converted to BAM files which are then sorted,
merged, and indexed. If selected, the BWA Aligner app will
calculate alignment statistics.
206 Xanita Saayman et al.

11. Bin size can be modified according to desired resolution. The

output file is in bigWig Track format (compressed binary ver-
sions of WIG files) and can be visualized independently in most
genome browsers (i.e., UCSC Genome Browser, IGV). Visu-
alization at this stage may identify any experimental or analyti-
cal issues up to this point.
12. Chromosome maps can be visualized at this point by uploading
output bigWig files to UCSC Genome Browser, IGV or similar
software to compare replication profiles of various time point
samples.
13. Filter width can be specified but should be maintained below
width of smooth line profiles. Non-polynomial least-square
filters can be applied with loss of peak resolution.

Acknowledgments

This work was supported by the Canadian Institutes of Health

Research and the Canadian Cancer Society Research Institute. We
thank the Donnelly Sequencing Centre for their deep sequencing
expertise.

References
1. Huberman JA, Riggs AD (1968) On the mech- 8. Seiler JA, Conti C, Syed A, Aladjem MI, Pom-
anism of DNA replication in mammalian chro- mier Y (2007) The intra-S-phase checkpoint
mosomes. J Mol Biol 32(2):327–341 affects both DNA replication initiation and
2. Kelly TJ, Brown GW (2000) Regulation of elongation: single-cell and -DNA fiber ana-
chromosome replication. Annu Rev Biochem lyses. Mol Cell Biol 27(16):5806–5818.
69:829–880 doi:10.1128/MCB.02278-06
3. Bell SP, Dutta A (2002) DNA replication in 9. Yamazaki S, Hayano M, Masai H (2013) Rep-
eukaryotic cells. Annu Rev Biochem 71:333–374 lication timing regulation of eukaryotic repli-
4. Masai H, Matsumoto S, You Z, Yoshizawa- cons: Rif1 as a global regulator of replication
Sugata N, Oda M (2010) Eukaryotic chromo- timing. Trends Genet: TIG 29(8):449–460.
some DNA replication: where, when, and how? doi:10.1016/j.tig.2013.05.001
Annu Rev Biochem 79:89–130. doi:10.1146/ 10. van Brabant AJ, Raghuraman MK (2002)
annurev.biochem.052308.103205 Assaying replication fork direction and migra-
5. Raghuraman MK, Winzeler EA, Collingwood tion rates. Methods Enzymol 351:539–568
D, Hunt S, Wodicka L, Conway A, Lockhart 11. Tercero JA, Diffley JF (2001) Regulation of
DJ, Davis RW, Brewer BJ, Fangman WL DNA replication fork progression through
(2001) Replication dynamics of the yeast damaged DNA by the Mec1/Rad53 check-
genome. Science 294(5540):115–121 point. Nature 412(6846):553–557
6. Agier N, Fischer G (2012) The mutational 12. Katou Y, Kanoh Y, Bando M, Noguchi H,
profile of the yeast genome is shaped by repli- Tanaka H, Ashikari T, Sugimoto K, Shirahige
cation. Mol Biol Evol 29(3):905–913. doi:10. K (2003) S-phase checkpoint proteins Tof1
1093/molbev/msr280 and Mrc1 form a stable replication-pausing
7. Stamatoyannopoulos JA, Adzhubei I, Thur- complex. Nature 424(6952):1078–1083
man RE, Kryukov GV, Mirkin SM, Sunyaev 13. Feng W, Collingwood D, Boeck ME, Fox LA,
SR (2009) Human mutation rate associated Alvino GM, Fangman WL, Raghuraman MK,
with DNA replication timing. Nat Genet 41 Brewer BJ (2006) Genomic mapping of single-
(4):393–395. doi:10.1038/ng.363 stranded DNA in hydroxyurea-challenged
 DNA Replication Profiling Using Deep Sequencing 207

yeasts identifies origins of replication. Nat Cell 17. Muller CA, Hawkins M, Retkute R, Malla S,
Biol 8(2):148–155 Wilson R, Blythe MJ, Nakato R, Komata M,
14. Lengronne A, Pasero P, Bensimon A, Schwob Shirahige K, de Moura AP, Nieduszynski CA
E (2001) Monitoring S phase progression (2014) The dynamics of genome replication
globally and locally using BrdU incorporation using deep sequencing. Nucleic Acids Res 42
in TK(þ) yeast strains. Nucleic Acids Res 29 (1):e3. doi:10.1093/nar/gkt878
(7):1433–1442 18. Futcher B (1999) Cell cycle synchronization.
15. Yabuki N, Terashima H, Kitada K (2002) Methods Cell Sci 21(2–3):79–86
Mapping of early firing origins on a replication 19. Smith J, Manukyan A, Hua H, Dungrawala H,
profile of budding yeast. Genes Cells 7 Schneider BL (2017) Synchronization of yeast.
(8):781–789. doi:559 [pii] Methods Mol Biol 1524:215–242. doi:10.
16. Gilbert DM (2010) Evaluating genome-scale 1007/978-1-4939-6603-5_14
approaches to eukaryotic DNA replication. Nat 20. Savitzky A, Golay MJE (1964) Smoothing and
Rev Genet 11(10):673–684. doi:10.1038/ differentiation of data by simplified least square
nrg2830 procedures. Analytical Chem 36(8):1627–1639
 Chapter 16

Quantitative Bromodeoxyuridine Immunoprecipitation

Analyzed by High-Throughput Sequencing (qBrdU-Seq
or QBU)
Joanna E. Haye-Bertolozzi and Oscar M. Aparicio

Abstract
Incorporation into DNA of nucleoside analogs like 5-bromo-20 -deoxyuridine (BrdU) is a powerful tool for
in vivo studies of DNA synthesis during replication and repair. Immunoprecipitation of BrdU-labeled DNA
analyzed by DNA sequencing (BrdU-IP-seq) allows for genome-wide, sequence-specific tracking of repli-
cation origin and replication fork dynamics under different conditions, such as DNA damage and replica-
tion stress, and in mutant strains. We have recently developed a quantitative method for BrdU-IP-seq
(qBrdU-seq) involving DNA barcoding to enable quantitative analysis of multiple experimental samples
subjected to BrdU-IP-seq. After initial barcoding of multiple, individually BrdU-labeled genomic DNA
samples, a pooling strategy is used for all subsequent steps including immunoprecipitation, amplification,
and sequencing, which eliminates sample-to-sample variability in these steps. Parallel processing of an
aliquot of the pooled input sample provides a direct control for the normalization of the data and yields
results that allow quantitative comparisons of the experimental samples. Though developed for the analysis
of S. cerevisiae, this method should be directly adaptable to other model systems.

Key words 5-Bromo-20 -deoxyuridine, Nucleoside analog, DNA replication, DNA barcoding, Data
normalization between experiments

1 Introduction

Incorporation of nucleoside analogs such as 5-Bromo-20 -deoxyur-

idine (BrdU) has been used for decades to study DNA replication
in vivo [1]. Such studies in yeasts have been facilitated by the
development of vectors expressing thymidine kinase and a nucleo-
side transporter to reconstruct the thymidine salvage pathway,
which is lacking in budding and fission yeasts [2–7]. BrdU incor-
poration is now routinely used in studies of replication origin and
replication fork dynamics under normal and stressed conditions,
which are known to affect both origin and fork functions. BrdU
incorporation into DNA may be analyzed by several means, such as
immunofluorescence, which allows for cytogenetic analysis [8],

Marco Muzi-Falconi and Grant W. Brown (eds.), Genome Instability: Methods and Protocols, Methods in Molecular Biology,
vol. 1672, DOI 10.1007/978-1-4939-7306-4_16, © Springer Science+Business Media LLC 2018

209
210 Joanna E. Haye-Bertolozzi and Oscar M. Aparicio

Anti-BrdU Barcoded Barcoded

BrdU Antibody Adapter 1 Adapter 2

Sample 1

Shear BrdU labeled Repair ends, ligate Illumina-compatible

genomic DNA adapters with unique barcodes, QC
Sample 2

Pool samples
(1:1)
IP anti-BrdU

Amplify DNA with Illumina-compatible

Input (2%)
primers, QC, sequence on Illumina

Fig. 1 Scheme of quantitative BrdU-IP-seq analysis. BrdU-labeled genomic DNA from each sample is barcoded
by end-ligation of Illumina-compatible linkers. Samples are pooled, a small fraction of this pool is set aside as
“Input” and the remainder is subjected to immunoprecipitation (IP) with anti-BrdU antibody. The IP and Input
samples are PCR-amplified with indexed primers and sequenced. IP sample reads are normalized against
Input sample reads. Adapted from [14] under Creative Commons Attribution-NonCommercial 4.0 International
License (CC-BY-NC)

DNA fiber labeling, which allows for the analysis of the replication
of individual molecules, though typically without DNA sequence
information [9], and immunoprecipitation (IP), which allows
determination of specific DNA sequences incorporating BrdU
(BrdU-IP-chip or BrdU-IP-seq), but is normally carried out as a
population measure [10–12]. Nevertheless, we note that there is no
inherent reason BrdU-IP-seq cannot be performed on individual
cells; indeed, BrdU-IP is more sensitive than ChIP, for example,
which is amenable to single cell analysis [13].
We have recently developed a quantitative BrdU-IP-seq
(qBrdU-seq or QBU) that allows for direct quantitative compar-
isons of BrdU-incorporation amongst independently generated
experimental samples (Fig. 1) [14]. The key features that distin-
guish the QBU procedure from BrdU-IP-seq are: initial barcoding
of individual BrdU-labeled samples; pooling of multiple experi-
mental samples, including replicates, for IP, amplification, and
sequencing; and parallel analysis of an aliquot of the pooled input
for quantitative normalization of the data. Not only does this
approach eliminate technical variability due to individual sample
processing at each of the above steps and the purification steps in
 Quantitative BrdU IP-seq 211

between, but it reduces labor and expense of multiple, individual

sample preparations. QBU is frequently performed on up to 12
samples concurrently, but should be scalable.
Because BrdU-IP experiments are performed on populations of
cells and events related to DNA replication and repair occur in a
spatio-temporal manner (e.g., differential origin initiation timing
or replication fork stalling/progression), meaningful analysis typi-
cally requires culture synchronization. In the protocol below, we
describe synchronization of S. cerevisiae cells in late G1 phase with
α-factor followed by release into S-phase in the presence of BrdU
with or without hydroxyurea, which arrests replication in early S
phase. Hydroxyurea is often used to provide an early replication
profile from a single time-point, which can reveal deregulation of
normal replication timing patterns or intra-S checkpoint control of
(late and dormant) origin firing. Cells are harvested at the appro-
priate time(s) depending on experimental design and genomic
DNA is isolated. The DNA from each sample is uniquely identified
by barcoded linker addition to the DNA ends (Fig. 1). The samples
are pooled, a small aliquot of this “Input” is set aside, and the pool
is denatured and subjected to IP with antibody against BrdU. The
IP DNA and Input DNA are amplified in parallel with distinctly
indexed primers and sequenced. The resulting reads are de-
multiplexed and mapped to the genome and IP read counts are
divided by the Input read counts for the corresponding sample.
This normalization step corrects for any differences in the initial
amounts of the individual samples pooled together, as well as
possible differences in efficiency of the adapter ligation/barcoding
step. Thus, the normalized data are suitable for quantitative com-
parisons among the pooled samples. The protocol includes quality
control steps to confirm efficient end-linker addition and library
amplification, and outlines a computational pipeline for initial anal-
ysis of the sequence data.

2 Materials

All solutions should be prepared with deionized, nuclease-free

H2O, except growth medium, which should be prepared with
pure, deionized, or distilled H2O. All tips and tubes should be
sterile and nuclease-free. Reagents, kits, and solutions should be
stored according to the manufacturer’s guidelines, or as indicated
below.

2.1 Growth, 1. YEPD broth: 1%(w/v) Yeast Extract, 2%(w/v) Bacto-Peptone,

Synchronization, and 2%(w/v) Dextrose.
Harvesting of Cells 2. 1 mg/mL α-factor (e.g., Sigma T6901) (in 1% DMSO), store
at 20  C.
212 Joanna E. Haye-Bertolozzi and Oscar M. Aparicio

3. Hydroxyurea (e.g., Sigma H8627).

4. 40 mg/mL 5-Bromo-20 -deoxyuridine, (in 100 mM Tris–HCl
pH 7.6), store at 20  C, protected from light.
5. 20 mg/mL Pronase (e.g., Sigma P5147), store at 20  C.
6. Probe sonicator, for disaggregation of cells.
7. 10%(w/v) Sodium azide, store at 4  C.
8. 50 mL sterile polypropylene tubes.
9. 2 mL flat- or round-bottom tubes with gasket-sealed screw-
caps.
10. Tris-buffered saline (TBS): 100 mM Tris–HCL pH 7.6,
150 mM NaCl, store at 4  C.

2.2 Cell Lysis, 1. Glass beads, 425–600 μm (sterile).

Genomic DNA 2. Genomic lysis buffer: 100 mM Tris–HCl pH 8.0, 50 mM
Isolation, and EDTA, 1% SDS, store at room temperature.
Chromatin
3. 5 M NaCl, store at room temperature.
Fragmentation
4. FastPrep 24 Instrument (MP Biomedicals). Vortexers or other
vigorous agitators may be used instead.
5. 26G hypodermic needles.
6. 5 mL polypropylene snap-cap tubes; caps not required.
7. 1.7 mL microcentrifuge tubes.
8. 2 mL Phase-lock gel tubes (Quanta Biosciences 2302830).
Optional: The use of Phase-lock gel is recommended for safety,
reproducibility, and effective separation of contaminants.
9. PCI (Phenol/Chloroform/Isoamyl alcohol, 25:24:1, buffered
for DNA), store at 4  C.
10. TE: 10 mM Tris–HCl pH 7.6, 1 mM EDTA, store at room
temperature.
11. 20 mg/mL RNase A (DNase-free), store at 20  C.
12. 10 SDS-PK: 1 mg/mL Proteinase K (in 200 mM Tris–HCl
pH 7.6, 200 mM EDTA pH 8, 2% SDS, 1 mM CaCl2), store at
20  C.
13. Qiagen QIAquick PCR purification kit (includes buffers PB,
PE, and EB) (see Note 1).
14. Nanodrop spectrophotometer (Cole-Palmer) or alternative
instrument for measurement of DNA concentration in a low
volume.
15. Covaris S2 Sonicator (see Note 2).
16. Covaris microTUBE AFA Fiber Pre-slit, Snap-Cap (Covaris
520045).
 Quantitative BrdU IP-seq 213

2.3 DNA Barcoding/ 1. KAPABiosystems Hyper prep kit, KR0961-v3.15 (see Note 3).
Adapter Ligation 2. Qiagen QIAquick PCR purification kit (includes buffers PB,
PE, and EB) (see Note 1).
3. Barcoded adapters (see Table 1 for sequences, and Note 4 for
annealing protocol).
4. Nanodrop spectrophotometer (Cole-Palmer) or alternative
instrument for measurement of DNA concentration in a low
volume.

Table 1
Oligonucleotides sequences for barcoded adapters. All sequences are given 50 to 30 ; P indicates
50 -phosphate. Sequences are taken from [23].

Oligonucleotide Sequence
Barcode 1a P-CAGTAGATCGGAAGAGCACACGTCT
Barcode 1b ACACTCTTTCCCTACACGACGCTCTTCCGATCTACTGT
Barcode 2a P-GCATAGATCGGAAGAGCACACGTCT
Barcode 2b ACACTCTTTCCCTACACGACGCTCTTCCGATCTATGCT
Barcode 3a P-AGCTAGATCGGAAGAGCACACGTCT
Barcode 3b ACACTCTTTCCCTACACGACGCTCTTCCGATCTAGCTT
Barcode 4a P-CGTAAGATCGGAAGAGCACACGTCT
Barcode 4b ACACTCTTTCCCTACACGACGCTCTTCCGATCTTACGT
Barcode 5a P-TCGAAGATCGGAAGAGCACACGTCT
Barcode 5b ACACTCTTTCCCTACACGACGCTCTTCCGATCTTCGAT
Barcode 6a P-GTCAAGATCGGAAGAGCACACGTCT
Barcode 6b ACACTCTTTCCCTACACGACGCTCTTCCGATCTTGACT
Barcode 7a P-ACTGAGATCGGAAGAGCACACGTCT
Barcode 7b ACACTCTTTCCCTACACGACGCTCTTCCGATCTCAGTT
Barcode 8a P-CTAGAGATCGGAAGAGCACACGTCT
Barcode 8b ACACTCTTTCCCTACACGACGCTCTTCCGATCTCTAGT
Barcode 9a P-TACGAGATCGGAAGAGCACACGTCT
Barcode 9b ACACTCTTTCCCTACACGACGCTCTTCCGATCTCGTAT
Barcode 10a P-GATCAGATCGGAAGAGCACACGTCT
Barcode 10b ACACTCTTTCCCTACACGACGCTCTTCCGATCTGATCT
Barcode 11a P-ATGCAGATCGGAAGAGCACACGTCT
Barcode 11b ACACTCTTTCCCTACACGACGCTCTTCCGATCTGCATT
Barcode 12a P-TGACAGATCGGAAGAGCACACGTCT
Barcode 12b ACACTCTTTCCCTACACGACGCTCTTCCGATCTGTCAT
214 Joanna E. Haye-Bertolozzi and Oscar M. Aparicio

Table 2
Oligonucleotide sequences for library amplification primers. All sequences are given 50 to 30 ; *
indicates phosphorothioate bond, which protects from nucleolytic degradation. Sequences are taken
from [23]

Illumina primers Sequence

Multi/Std primer 1.0 AATGATACGGCGACCACCGAGATCTACACTCTTTCCCTACACG
ACGCTCTTCCGATC*T
Multiplexing primer 2.0 GTGACTGGAGTTCAGACGTGTGCTCTTCCGATC*T
Index primer 1 CAAGCAGAAGACGGCATACGAGATCGTGATGTGACTGGAGTT*C
Index primer 2 CAAGCAGAAGACGGCATACGAGATACATCGGTGACTGGAGTT*C
Index primer 3 CAAGCAGAAGACGGCATACGAGATGCCTAAGTGACTGGAGTT*C
Index primer 4 CAAGCAGAAGACGGCATACGAGATTGGTCAGTGACTGGAGTT*C
Index primer 5 CAAGCAGAAGACGGCATACGAGATCACTGTGTGACTGGAGTT*C
Index primer 6 CAAGCAGAAGACGGCATACGAGATATTGGCGTGACTGGAGTT*C
Index primer 7 CAAGCAGAAGACGGCATACGAGATGATCTGGTGACTGGAGTT*C
Index primer 8 CAAGCAGAAGACGGCATACGAGATTCAAGTGTGACTGGAGTT*C
Index primer 9 CAAGCAGAAGACGGCATACGAGATCTGATCGTGACTGGAGTT*C
Index primer 10 CAAGCAGAAGACGGCATACGAGATAAGCTAGTGACTGGAGTT*C
Index primer 11 CAAGCAGAAGACGGCATACGAGATGTAGCCGTGACTGGAGTT*C
Index primer 12 CAAGCAGAAGACGGCATACGAGATTACAAGGTGACTGGAGTT*C

2.4 Validation of 1. Phusion DNA Polymerase (New England Biolabs M0530L).

Adapter Ligation 2. Multi/Std primer 1.0 (25 μM) (see Table 2 for sequence).
3. Multiplexing primer 2.0 (0.5 μM) (see Table 2).
4. Index primers (25 μM) (see Table 2).

2.5 Pooling and BrdU 1. 10 PBS: 1.37 M NaCl, 27 mM KCl, 100 mM Na2HPO4,
Immunoprecipitation 18 mM KH2PO4; store at room temperature.
2. 2IP buffer: 2PBS þ0.10%(v/v) Triton-X100; prepare fresh.
3. Wash buffer: PBS þ0.05%(v/v) Triton-X100; prepare fresh.
4. TE: 10 mM Tris–HCl pH 7.6, 1 mM EDTA, store at room
temperature.
5. Elution buffer: TE þ0.5%(w/v) SDS; store at room
temperature.
6. Anti-BrdU antibody (Invitrogen 033900).
7. DynaBeads Protein G (Invitrogen 10004D) (see Note 5).
8. Deionized, nuclease-free H2O.
9. Microcentrifuge tube rotator.
 Quantitative BrdU IP-seq 215

10. Magnetic rack for standard microcentrifuge tubes.

11. Qiagen QIAquick PCR purification kit (includes buffers PB,
PE, and EB) (see Note 1).

2.6 Library 1. Phusion DNA Polymerase (New England Biolabs M0530L).

Amplification 2. Multi/Std primer 1.0 (25 μM) (see Table 2 for sequence).
3. Multiplexing primer 2.0 (0.5 μM) (see Table 2).
4. Index primers (25 μM) (see Table 2).
5. AMPure beads (Beckman Coulter A63880).
6. Magnetic rack for standard microcentrifuge tubes.
7. Nanodrop spectrophotometer (Cole-Palmer) or alternative
instrument for measurement of DNA concentration in a low
volume.

2.7 Quality Control 1. BioAnalyzer (Agilent Technologies) (see Note 6).

and Quantification

2.8 DNA Sequencing 1. Illumina sequencing instrument (see Note 7).

2.9 Data Analysis 1. Software for analysis of sequencing data (ea-utils, Bowtie2,
SAMtools, BEDTools2, MACS, R-studio, Diffbind; alterna-
tives are available).
2. High-performance computing platform.

3 Methods

3.1 Growth, 1. Grow 10–20 mL cells per sample to OD ~ 0.5 in YEPD at

Synchronization, and 23  C (see Note 8).
Harvesting of Cells 2. Transfer culture to a 15 or 50 mL sterile polypropylene tube(s)
and pellet cells by centrifugation in a swinging-bucket rotor
at ~1500  g, for 3 min; discard the supernatant.
3. Resuspend cells in equal volume of fresh YEPD containing
α-factor (~5 nM for bar1Δ strains and ~5 μM for BAR1
strains), and transfer culture to a conical flask and incubate
for 4 h at 23  C (or as otherwise appropriate) on a platform
shaker or shaking water bath (~180 rpm) (see Note 9).
4. For synchronous release, pellet cells by centrifugation in 15 or
50 mL sterile polypropylene tubes in a swinging-bucket rotor
at ~1500  g, for 3 min. Discard the supernatant.
5. Resuspend cells in an equal volume of fresh YEPD containing
400 μg/mL BrdU þ200 μg/mL Pronase (for bar1Δ strains, to
degrade α-factor) þ200 mM hydroxyurea (optional). Sonicate
the resuspended cells while still in the centrifuge tube with a
216 Joanna E. Haye-Bertolozzi and Oscar M. Aparicio

probe sonicator to disaggregate cells, which aggregate in

response to α-factor (see Note 10). Transfer to a conical flask
for incubation at 23  C with shaking.
6. Harvest cells at the appropriate time(s) after release (see Note
11). Transfer cells to a 15 or 50 mL screw-cap tube containing
an appropriate volume of 10% sodium azide to achieve 0.1%
final after addition of culture, and mix by inversion. Pellet cells
by centrifugation as above and discard the supernatant.
7. Resuspend cells in 10 mL ice-cold TBS. Pellet cells by centrifu-
gation as above and discard the supernatant.
8. Resuspend cells in 1 mL ice-cold TBS and transfer to 2 mL
screw-cap (FastPrep) tube. Pellet cells in microcentrifuge by
spinning at full speed for ~3 s. Discard the supernatant. Pro-
ceed to DNA isolation or freeze samples at 20  C.

3.2 Cell Lysis, 1. Cell lysis: If cells were frozen, thaw on ice.
Genomic DNA 2. Resuspend cells in 500 μL genomic lysis buffer. To each tube,
Isolation, and add an equal volume of glass beads (a 0.5 mL microcentrifuge
Chromatin tube may be used to measure and dispense the beads).
Fragmentation 3. Disrupt cells by bead-beating with FastPrep (45 s, speed 5.5).
Other instruments may be used for the mechanical disruption
of the yeast cell wall. Efficiency of cell breakage may be deter-
mined by light microscopy.
4. Pellet cells for 1 min at full speed in a microcentrifuge to
collapse foam; add 25 μL 5 M NaCl and close the tube tightly.
Repeat step 3 with FastPrep. Place the tube on ice for ~1 min
to chill and reduce any pressure build-up in the tube for the
next step.
5. Make sure the tube cap is tight, wipe off any ice with a kimwipe,
invert the tube and flick to cause beads and liquid to drop into
cap. Poke two holes near the bottom of the tube using a red-
hot 26G hypodermic needle and insert the tube into a 5 mL
polypropylene tube (snap-cap tube without cap). Spin 2 min at
2000 rpm (~1000  g) in table-top or clinical, swinging-
bucket centrifuge.
6. Transfer sample including pellet into a 2 mL phase-lock tube.
7. Genomic DNA isolation: Add equal volume (~500 μL) of PCI
and extract by shaking vigorously by hand for ~15 s, spin 5 min
at full speed in microcentrifuge
8. Repeat step 7 in the same phase-lock tube and decant the
aqueous phase (after centrifugation) into a microcentrifuge
tube. Add 1 mL 100%EtOH, mix and spin 10 min at full
speed in microcentrifuge.
 Quantitative BrdU IP-seq 217

9. Discard the supernatant, rinse pellet with 1 mL 70%EtOH, and

discard (do not loosen pellet; if it comes loose, centrifuge
again), and dry pellet briefly.
10. Dissolve pellet in 100 μL TE. Heat to 45–50  C to facilitate
DNA solubilization, and once pellet is dissolved, add RNAseA
to 0.2 mg/mL, and incubate for 30 min at 37  C.
11. Add 10 μL 10 SDS-PK and incubate for 30 min at 50  C.
12. Purify over QIAquick PCR purification column according to
the Qiagen protocol, except elute in 120 μL pre-warmed
(55  C) elution buffer, EB.
13. Measure DNA concentration with Nanodrop or other instru-
ment requiring minimal sample volume. The yield from 20 mL
culture (OD600 ~ 1) should be at least 2 μg.
14. DNA fragmentation: Transfer sample to Covaris microTUBE
and fragment DNA using the Covaris S2 sonicator (duty
cycle ¼ 10%, intensity ¼ 4, 200 cycles per burst for 120 s).
These settings yield DNA sheared to an average of 300–400 bp
(see Note 2).
15. Transfer the sheared DNA to a new 1.7 mL microcentrifuge
tube.

3.3 DNA Barcoding/ Samples are now end-labeled with unique barcoded adapters that
Adapter Ligation are compatible with the subsequent library-amplification and
sequencing primers for the sequencing platform to be used (i.e.,
Illumina). These barcodes will uniquely identify the source sample
for every DNA fragment, allowing the samples to be pooled for all
subsequent operations.
1. End repair and A-tailing: We recommend using KAPA Hyper
prep kit (Kapa Biosystems, KR0961-v3.15) according to the
Kapa protocol, except that we carry out half reactions. In a
0.2 mL PCR tube, on ice, combine the following:

0.5 μg fragmented, genomic DNA 25 μL

(use H2O to adjust volume)
End-repair and A-tailing buffer 3.5 μL
End-repair and A-tailing enzyme mix 1.5 μL
Total 30 μL

2. Mix by pipetting and immediately place in thermocycler and

incubate for 30 min at 20  C, followed by 30 min at 65  C, and
hold at 4  C.
3. Adaptor Ligation: To the End-repair and A-tailing reaction (on
ice or in 4  C thermocycler block) add each of the following
reagents from the KAPA kit:
218 Joanna E. Haye-Bertolozzi and Oscar M. Aparicio

Water 2.5 μL
Ligation buffer 15 μL
Adapter (15 μM stock) 2.5 μL
DNA ligase 5 μL


4. Mix by pipetting and incubate for 15 min at 20 C in
thermocycler.
5. Purify each sample using the QIAquick PCR purification kit
according to the Qiagen protocol, eluting in 50 μL pre-
warmed (55  C) EB.
6. Measure DNA concentration with Nanodrop.

3.4 Validation of Before proceeding to pooling and IP, the critical step of adapter
Adapter Ligation ligation is confirmed for individual samples by PCR with library
amplification primers. Whereas it is sufficient to ensure that indi-
vidual samples amplify with qualitatively similar efficiencies as
described here, quantitative PCR may be used to achieve more
exact measurements and potentially to adjust amounts for pooling
in the next section.
1. Pilot amplification with adapter-compatible primers.
In separate 0.2 mL PCR tubes on ice, place 1 μL of each
barcoded sample; prepare an additional tube containing 1 μL of
H2O for “no DNA” control.
2. In a microcentrifuge tube on ice, prepare the following reaction
mix multiplied by the number of samples to be examined plus a
few extra for the “no DNA” control and to account for pipet-
ting losses.

H2O 16.75 μL
5 Phusion HF buffer 5 μL
Multi/Std primer 1.0 (25 μM) 0.5 μL
Multiplexing primer 2.0 (0.5 μM) 0.5 μL
Index primer (25 μM) 0.5 μL
10 mM dNTP mix 0.5 μL
Phusion DNA polymerase 0.25 μL
Total 24 μL

PCR mix (per sample):

3. Dispense 24 μL of the PCR mix to each of the PCR sample
tubes.
4. Place samples in thermocycler and carry out the following PCR
protocol:
 Quantitative BrdU IP-seq 219

100bp no 100bp
ladder BC1 BC2 BC3 BC4 BC5 BC6 DNA ladder

1000bp

500bp

100bp

Fig. 2 Validation of adapter ligation. PCR analysis using Illumina amplification

primers as described in Subheading 3.4 was carried out on six different adapter
ligation reactions, each using a different barcoded adapter (BC1-BC6); an
additional PCR reaction was performed without added template (no DNA). After
addition of gel loading dye, half of each PCR reaction was subjected to electro-
phoresis on a 2% agarose gel.

Step 1: 98  C 1 min.
Step 2: 98  C 10s.
Step 3: 65  C 30s.
Step 4: 72  C 45 s.
Step 5: Repeat steps 2–4, 14–17 more times (use the minimum
number of cycles necessary to amplify enough DNA for
quality control and sequencing below).
Step 6: 72  C 5 min.
Step 7: 4  C hold.
5. Analyze PCR products on agarose gel (see Fig. 2).

3.5 Pooling and BrdU Samples are pooled for IP at a DNA concentration of 1 μg/mL.
Immunoprecipitation 100 ng of each barcoded sample is recommended, but should be
scalable. IP volume should be scaled according to number of sam-
ples pooled, and amount used per sample. It is not necessary to
pool equal amounts of each sample, as parallel analysis of an aliquot
of the pooled Input will allow for correction of DNA amounts
through normalization of the data.
1. In a 1.7 mL microcentrifuge tube (or larger tube for greater
than 15 samples),

Combine (per sample): 100 ng DNA 45 μL

(adjust volume with H2O)
10 PBS 5 μL
Total (per sample) 50 μL

Mix thoroughly.
220 Joanna E. Haye-Bertolozzi and Oscar M. Aparicio

2. Transfer 2% of the total volume of the pooled sample to a new

microcentrifuge tube and set-aside until Subheading 3.6; this is
the “Input” DNA.
3. Heat the pooled sample from step 1 for 5 min at 95  C and
immediately place in an ice water bath for 2 min.
4. Add an equal volume of ice-cold 2IP buffer.
5. Add an appropriate amount of anti-BrdU antibody (e.g.,
1:200) and incubate on rotating wheel for 45 min at 4  C.
6. Add 30 μL of Protein G-Dynabeads (resuspend the slurry
immediately before use) and incubate on rotating wheel for
30 min at 4  C. Place the tube on a magnetic rack to collect
beads, and discard buffer by decanting and/or aspirating,
being careful not to aspirate beads.
7. Remove the tube from magnet and wash beads by adding 1 mL
cold wash buffer by inverting tubes until beads are fully sus-
pended. Place the tube on a magnetic rack to collect beads and
remove wash buffer as above. Repeat this step two more times
with wash buffer. (This and the following steps may be per-
formed at room temperature with cold buffer).
8. Wash beads once as in step 7 with 1 mL ice-cold TE.
9. Elute sample by resuspending beads in 100 μL elution buffer
and incubating 10 min at 65  C in heat block or water bath.
10. Place the tube on a magnetic rack to collect beads and transfer
the eluate to a new microcentrifuge tube.
11. Purify with QIAquick PCR purification kit according to Qia-
gen protocol, except elute from column with 30 μL pre-
warmed H2O (65  C).

3.6 Library High-fidelity Phusion polymerase is used for amplification; Kapa

Amplification polymerase is also recommended. The IP and Input DNA samples
are amplified in separate reactions with a different index primer to
allow combined sequencing on Illumina.
1. Library amplification with Illumina Indexed Primers.
In a 0.2 mL PCR tube on ice, combine the following and
mix:

DNA (volume adjusted with H2O) 34.5 μL

5 Phusion HF buffer 10 μL
Multi/Std primer 1.0 (25 μM) 1 μL
Multiplexing primer 2.0 (0.5 μM) 1 μL
Index primer* (25 μM) 1 μL
*a different Index primer is used for each reaction
 Quantitative BrdU IP-seq 221

10 mM dNTP mix 1.5 μL

Phusion DNA polymerase 1 μL
Total 50 μL

2. Place samples in thermocycler and carry out the following PCR

protocol:
Step 1: 98  C 1 min.
Step 2: 98  C 10 s.
Step 3: 65  C 30 s.
Step 4: 72  C 45 s.
Step 5: Repeat steps 2–4, 14–17 more times (use the minimum
number of cycles necessary to amplify enough DNA for
quality control and sequencing below).
Step 6: 72  C 5 min.
Step 7: 4  C hold.
3. DNA Purification with AMPure beads.
AMPure beads are effective at removal of primers and
larger primer dimers, which are common by-products of PCR
that should be removed prior to DNA sequencing.
Transfer the library amplification reaction to a 1.7 mL
microcentrifuge tube and purify DNA according to the
AMPure protocol using an equal volume of AMPure bead
slurry, and final elution with 20 μL of pre-warmed TE
(65  C). Use the magnetic rack as described for the Dynabeads
in Subheading 3.5.
4. Measure DNA concentration with Nanodrop.

3.7 Quality Control The Agilent Technologies Bioanalyzer permits quantification and
and Quantification determination of DNA fragment size distribution of the amplified
samples. This analysis confirms the library quality based on the size
distribution, expected to show an average fragment size of
~400 bp, as the linker and primer additions add ~130 bp (Fig. 3).
Common artifacts such as primer dimers that will consume
sequence reads will be revealed, allowing their possible elimination
through further purification using AMPure beads. Quantification
at this step allows adjustment of sample concentrations for the
desired sequencing platform.
1. Analyze up to 10 ng DNA on Bioanalyzer (see Note 6).

3.8 DNA Sequencing 1. Sequence samples using an Illumina instrument (see Note 7).
10 ng (~2 ng/μL) of DNA is sufficient.
222 Joanna E. Haye-Bertolozzi and Oscar M. Aparicio

Fig. 3 Quality control and quantification. 1–10 ng of the amplified and purified library from the completion of
Subheading 3.6 was run on a Bioanalyzer chip (High Sensitivity DNA Assay). The tracing shows Intensity in
fluorescence units (FU) on the y-axis and DNA length in base-pairs (bp) on the x-axis of the gel image to the
right. The lower (35 bp) and upper molecular weight markers (10,380 bp) are indicated in green and purple
respectively. Results are summarized below the graph indicating an average fragment size (400 bp) and
concentration (~4 ng/μL).

3.9 Data Analysis Sequenced reads are processed using a series of software programs
that are freely available online. The following analysis pipeline
allows for the extraction of the data, mapping the reads to the
reference genome, quantitative normalization, and data display.
1. ea-utils: Sequenced reads are first processed using the fastq-
multx tool [15, 16]. This de-multiplexes the reads based on the
specified barcode sequences. The first five bases are trimmed
from the 50 end of each read to remove the barcodes for
genome alignment.
2. bowtie2: The S. cerevisiae genome sequence is accessed from
Saccharomyces Genome Database (www.yeastgenome.org)
and indexed using bowtie2, bowtie2-build [17]. The raw reads
are then aligned to the indexed genome using bowtie2.
3. Samtools: Filter reads that map to more than one region in the
genome using samtools view. Sort by coordinates of the refer-
ence genome using samtools sort. Remove PCR duplicates
using samtools rmdup.
4. BEDtools: Change file format by converting files from .bam to .
bed files. Bin the aligned reads into 50 bp nonoverlapping bins
using BEDtools [18, 19].
5. QBU normalization: Binned read counts for IP are divided by
the corresponding binned read counts for Input, for the
 Quantitative BrdU IP-seq 223

corresponding samples (i.e., IP and Input with the same bar-

codes). The resulting data is median-smoothed over a 1000 bp
window for both IP and Input. Replicate data sets may be
averaged for plotting and further analysis. The values from
the IP/Input normalization are unitless but are directly com-
parable between samples in the same pool. Optional: We find it
useful to scale the data from each pool to a maximum value of
“1” for one of the samples (e.g., the wild-type sample). This is
achieved as follows: Any replicates are first averaged. The maxi-
mum bin value in the wild-type data set is determined, and all
bins for all samples in that pool are divided by that maximum
bin value.
Additional analyses: MACS may be used for peak-calling, using
filtered, aligned reads [20]; Diffbind (using DESeq) may be used
for analysis for differential data signal (e.g., IP signal) [21, 22].

4 Notes

1. Alternative kits or methods for DNA purification may be

substituted.
2. Alternative sonicators including probe sonicators may be used,
but will require prior determination of conditions for optimal
DNA shearing. This protocol targets ~300 bp average length.
DNA may also be fragmented by enzymatic digestion (e.g.,
KAPA Hyper Plus kit).
3. Alternative library preparation kits or individual reagents may
be used for Illumina sequencing; alternative sequencing plat-
forms will require alternative library preparation reagents or
kits.
4. Oligonucleotide pairs must be annealed for ligation. Oligonu-
cleotide pairs are diluted (from 100 μM stocks in TE) to 15 μM
using annealing buffer (10 mM Tris–HCl pH 7.6, 50 mM
NaCl, 1 mM EDTA). Place in heat block 5 min at 95  C.
Remove the heat block (with samples) from the heating unit
and allow to cool to room temperature. A thermocycler may be
used with a cooling rate of ~2  C per minute; store at 20  C.
5. Protein A- and anti-IgG-coupled beads may also be used as
appropriate for the anti-BrdU antibody used; sepharose beads
may be used instead of magnetic beads.
6. The Bioanalyzer allows electrophoretic analysis of nanogram
quantities of DNA. If a Bioanalyzer is not available, larger
amounts on the order of 100 ng may be analyzed by agarose
gel electrophoresis using a sensitive dye such as SYBR Green
(Molecular Probes) and an appropriate scanner. Real-time PCR
(using the library amplification primers) may also be used for
224 Joanna E. Haye-Bertolozzi and Oscar M. Aparicio

quantification, and PCR products may be analyzed by gel

electrophoresis.
7. Other sequencing platforms may be used but will require mod-
ification(s) of the method, particularly the library construction
steps.
8. Culturing conditions, including media, temperature, synchro-
nization method, etc., may be modified according to experi-
mental or strain-specific requirements. DNA content analysis
by flow cytometry and/or morphological analysis of budding
is/are recommended to confirm expected cell cycle synchroni-
zation and progression.
9. Monitor cells microscopically during this incubation to con-
firm synchronization and determine the appropriate time for
release into S phase (remove ~0.5 mL of culture and sonicate
gently to disaggregate cells before placing ~10 μL on slide with
cover slip). For α-factor-arrested cells, unbudded cells will
accumulate (>95% for bar1Δ strains after 3–4 h) and change
morphologically from roughly spherical- or oval-shaped cells to
elongated, pear-shaped cells called “schmoos.” Cells are held
arrested ~4 h to allow daughter cells to grow sufficiently to
release synchronously with the larger mothers.
10. Sonicator settings will depend on the instrument and the vol-
ume of the culture and the centrifuge tube. For example, for a
15 mL culture volume in a 50 mL tube, the following settings
are used for a Branson 250 Sonicator: power: 2.5, duty cycle:
30%, six pulses.
11. For hydroxyurea-treated cells, 60–90 min is most commonly
used. For time-course experiments, a large culture may be
continuously labeled with BrdU and sampled at regular inter-
vals, such as 10 min; alternatively, cells from a large unlabeled
culture (i.e., lacking BrdU) may be pulse-labeled for defined
intervals (e.g., 10–15 min) by incubating aliquots removed
from the culture with BrdU.

Acknowledgments

We thank current and previous members of the Aparicio lab for

comments on the protocol. Research in the Aparicio lab has been
funded by NIH R01 GMS05494.

References

1. Dolbeare F (1996) Bromodeoxyuridine: a DNA replication sites and in situ hybridization.

diagnostic tool in biology and medicine, Part Histochem J 28(8):531–575
III. Proliferation in normal, injured and dis- 2. Lengronne A, Pasero P, Bensimon A, Schwob
eased tissue, growth factors, differentiation, E (2001) Monitoring S phase progression
 Quantitative BrdU IP-seq 225

globally and locally using BrdU incorporation prot5385. doi:2010/2/pdb.prot5385 [pii]

in TK(þ) yeast strains. Nucleic Acids Res 29 1101/pdb.prot5385
(7):1433–1442 13. Rotem A, Ram O, Shoresh N, Sperling RA,
3. Vernis L, Piskur J, Diffley JF (2003) Reconsti- Goren A, Weitz DA, Bernstein BE (2015)
tution of an efficient thymidine salvage path- Single-cell ChIP-seq reveals cell subpopulations
way in Saccharomyces cerevisiae. Nucleic Acids defined by chromatin state. Nat Biotechnol 33
Res 31(19):e120 (11):1165–1172. doi:10.1038/nbt.3383
4. Viggiani CJ, Aparicio OM (2006) New vectors 14. Peace JM, Villwock SK, Zeytounian JL, Gan Y,
for simplified construction of BrdU- Aparicio OM (2016) Quantitative BrdU
incorporating strains of Saccharomyces cerevi- immunoprecipitation method demonstrates
siae. Yeast 23(14–15):1045–1051 that Fkh1 and Fkh2 are rate-limiting activators
5. Vickers MF, Mani RS, Sundaram M, Hogue of replication origins that reprogram replica-
DL, Young JD, Baldwin SA, Cass CE (1999) tion timing in G1 phase. Genome Res 26
Functional production and reconstitution of (3):365–375. doi:10.1101/gr.196857.115
the human equilibrative nucleoside transporter 15. Aronesty E (2011) Command-line tools for
(hENT1) in Saccharomyces cerevisiae. Interac- processing biological sequencing data. https://
tion of inhibitors of nucleoside transport with github.com/ExpressionAnalysis/ea-utils.
recombinant hENT1 and a glycosylation- 16. Aronesty E (2013) Comparison of sequence
defective derivative (hENT1/N48Q). Bio- utility programs. Open Bioinform J 7:1–8.
chem J 339(Pt 1):21–32 doi:10.2174/1875036201307010001
6. Hodson JA, Bailis JM, Forsburg SL (2003) 17. Langmead B, Salzberg SL (2012) Fast gapped-
Efficient labeling of fission yeast Schizosacchar- read alignment with bowtie 2. Nat Methods 9
omyces pombe with thymidine and BUdR. (4):357–359. doi:10.1038/nmeth.1923
Nucleic Acids Res 31(21):e134 18. Li H, Handsaker B, Wysoker A, Fennell T,
7. Sivakumar S, Porter-Goff M, Patel PK, Benoit Ruan J, Homer N, Marth G, Abecasis G, Dur-
K, Rhind N (2004) In vivo labeling of fission bin R, Genome Project Data Processing S
yeast DNA with thymidine and thymidine ana- (2009) The sequence alignment/map format
logs. Methods 33(3):213–219. doi:10.1016/j. and SAMtools. Bioinformatics 25
ymeth.2003.11.016 (16):2078–2079. doi:10.1093/bioinformat
8. Green MD, Sabatinos SA, Forsburg SL (2015) ics/btp352
Microscopy techniques to examine DNA replica- 19. Quinlan AR, Hall IM (2010) BEDTools: a
tion in fission yeast. Methods Mol Biol 1300: flexible suite of utilities for comparing genomic
13–41. doi:10.1007/978-1-4939-2596-4_2 features. Bioinformatics 26(6):841–842.
9. Bianco JN, Poli J, Saksouk J, Bacal J, Silva MJ, doi:10.1093/bioinformatics/btq033
Yoshida K, Lin YL, Tourriere H, Lengronne A, 20. Zhang Y, Liu T, Meyer CA, Eeckhoute J, John-
Pasero P (2012) Analysis of DNA replication son DS, Bernstein BE, Nusbaum C, Myers RM,
profiles in budding yeast and mammalian cells Brown M, Li W, Liu XS (2008) Model-based
using DNA combing. Methods 57(2):149–157. analysis of ChIP-Seq (MACS). Genome Biol 9
doi:10.1016/j.ymeth.2012.04.007 (9):R137. doi:10.1186/gb-2008-9-9-r137
10. Peace JM, Ter-Zakarian A, Aparicio OM 21. Anders S, Huber W (2010) Differential expres-
(2014) Rif1 regulates initiation timing of late sion analysis for sequence count data. Genome
replication origins throughout the S. cerevisiae Biol 11(10):R106. Doi:gb-2010-11-10-r106
genome. PLoS One 9(5):e98501. doi:10. [pii] 1186/gb-2010-11-10-r106
1371/journal.pone.0098501 22. Stark R, Brown R (2013) DiffBind: differential
11. Ryba T, Battaglia D, Pope BD, Hiratani I, binding analysis of ChIP-Seq peak data. bio
Gilbert DM (2011) Genome-scale analysis of conductor.Org. http://bioconductor.org/
replication timing: from bench to bioinformat- packages/release/bioc/vignettes/DiffBind/
ics. Nat Protoc 6(6):870–895. doi: inst/doc/DiffBind.pdf
nprot.2011.328 [pii] 1038/nprot.2011.328 23. Dunham JP, Friesen ML (2013) A cost-
12. Viggiani CJ, Knott SR, Aparicio OM (2010) effective method for high-throughput con-
Genome-wide analysis of DNA synthesis by struction of illumina sequencing libraries.
BrdU immunoprecipitation on tiling microar- Cold Spring Harb Protoc 2013(9):820–834.
rays (BrdU-IP-chip) in Saccharomyces cerevi- doi:10.1101/pdb.prot074187
siae. Cold Spring Harb Protoc 2010(2.) pdb
 Chapter 17

Strand-Specific Analysis of DNA Synthesis and Proteins

Association with DNA Replication Forks in Budding Yeast
Chuanhe Yu, Haiyun Gan, and Zhiguo Zhang

Abstract
DNA replication initiates at DNA replication origins after unwinding of double-strand DNA(dsDNA) by
replicative helicase to generate single-stranded DNA (ssDNA) templates for the continuous synthesis of
leading-strand and the discontinuous synthesis of lagging-strand. Therefore, methods capable of detecting
strand-specific information will likely yield insight into the association of proteins at leading and lagging
strand of DNA replication forks and the regulation of leading and lagging strand synthesis during DNA
replication. The enrichment and Sequencing of Protein-Associated Nascent DNA (eSPAN), which measure
the relative amounts of proteins at nascent leading and lagging strands of DNA replication forks, is a step-
wise procedure involving the chromatin immunoprecipitation (ChIP) of a protein of interest followed by
the enrichment of protein-associated nascent DNA through BrdU immunoprecipitation. The isolated
ssDNA is then subjected to strand-specific sequencing. This method can detect whether a protein is
enriched at leading or lagging strand of DNA replication forks. In addition to eSPAN, two other strand-
specific methods, (ChIP-ssSeq), which detects potential protein-ssDNA binding and BrdU-IP-ssSeq,
which can measure synthesis of both leading and lagging strand, were developed along the way. These
methods can provide strand-specific and complementary information about the association of the target
protein with DNA replication forks as well as synthesis of leading and lagging strands genome wide. Below,
we describe the detailed eSPAN, ChIP-ssSeq, and BrdU-IP-ssSeq protocols.

Key words Chromatin immunoprecipitation (ChIP), BrdU immunoprecipitation, Single-strand

DNA library preparation, Next-generation sequencing, eSPAN, ChIP-ssSeq and BrdU-IP-ssSeq

1 Introduction

Eukaryotic DNA replication initiates at multiple DNA replication

origins in a tempo-spatial order [1]. Once started, DNA synthesis
proceeds bi-directionally, with continuous synthesis of leading
strand DNA and discontinuous synthesis of lagging strand or Oka-
zaki fragment [2]. Multiple proteins, including DNA polymerases,
replicative DNA helicase MCM complexes, DNA ligase, proliferat-
ing cell nuclear antigen (PCNA), and single-strand DNA binding
protein (SSB), are involved in DNA synthesis [3–9]. Chromatin
immunoprecipitation (ChIP) coupled with PCR analysis and/or

Marco Muzi-Falconi and Grant W. Brown (eds.), Genome Instability: Methods and Protocols, Methods in Molecular Biology,
vol. 1672, DOI 10.1007/978-1-4939-7306-4_17, © Springer Science+Business Media LLC 2018

227
228 Chuanhe Yu et al.

microarray has been used to map DNA replication origins and to

monitor the replication proteins traveling along with the replica-
tion fork [10, 11]. All these methods cannot differentiate leading
strand from lagging strand. We reason that monitoring of DNA
synthesis and the association of a protein of interest with DNA
replication forks in a strand-specific manner will yield novel insight
into the DNA replication process. Therefore, we developed the
eSPAN method (enrichment and Sequencing of Protein-Associated
Nascent DNA) [12] to measure the relative amount of protein at
leading and lagging strands DNA replication forks (Fig. 1). Using
this method, we show that leading strand DNA Polymerase ε and
lagging strand polymerase δ [13] associate preferentially with lead-
ing and lagging strands of DNA replication forks, respectively
(Fig. 2) [14]. We further showed that PCNA is unloaded from
lagging strands when replication forks stall. These results demon-
strate the efficiency and power of eSPAN. During the development
of the eSPAN method, we also performed nascent DNA strand-
specific sequencing (BrdU-IP-ssSeq) and targeted protein ChIP
strand-specific sequencing (ChIP-ssSeq) (Fig. 1). BrdU-IP-ssSeq
can monitor the leading and lagging strands synthesis indepen-
dently; whereas ChIP-ssSeq can monitor replication protein travel
on the template strand and detect potential protein-ssDNA bind-
ing. These combined methods provide unprecedented information
on the synthesis of leading and lagging strands of DNA replication
forks and the association of proteins with leading and lagging
strands of DNA replication forks. Here, we describe the protocols
for ChIP-ssSeq, BrdU-IP-ssSeq, and eSPAN in budding yeast.
Targeted protein ChIP, labeling DNA with BrdU, and BrdU
immunoprecipitation protocols were adapted from [15, 16] and
the ssDNA library preparation protocol was adapted from [17, 18].

2 Materials

2.1 Yeast Cell 1. YPD medium: Bacto peptone 20 g, yeast extract 10 g, glucose
Culture and Sample 20 g and transfer to a 1 l cylinder. Add 950 ml water. Stir until
Collection fully suspended. Make up to 1 l with water. Autoclave.
2. Yeast strain (W303A strain).
3. Alpha factor (WHWLQLKPGQPMY) stock: 5 mg/ml. Dis-
solve 50 mg in 10 ml 100% ethanol. Store at
20  C.
4. Tris-buffered saline (TBS; 10):1.5 M NaCl, 0.1 M Tris–HCl,
pH 7.6. 24.2 g Tris base and 80 g NaCl. Transfer to a 1 l
cylinder. Add 900 ml and adjust pH to 7.6 with HCl. Stir until
fully suspended. Make up to 1 l with water. Autoclave.
5. PMSF (100): Dissolve 17.4 mg PMSF in 1 ml isopropanol.
 Strand-Specific Analysis of DNA Synthesis and Proteins Association with. . . 229

A BrdU-IP Strand-specific Sequencing (ssSeq) (BrdU-IP-ssSeq)

Release into S phase
Cells at G1 in the presence of HU and
ssSeq (protein ChIP-ssSeq)
BrdU for 45 min protein ChIP
BrdU-IP ssSeq (eSPAN)

B origin

Strand-specific
BrdU IP Sequencing BrdU-IP-ssSeq
Polε
5’ 3’
BrdU
3’ 5’
Pole
Pole ChIP 5’ * Pole 3’
3’ Pole * 5’ +

Reverse-crosslinking
Denaturation
ssSeq

protein ChIP-ssSeq BrdU IP

ssSeq

eSPAN

Fig. 1 An outline of experimental procedures of BrdU IP-ssSeq, ChIP-ssSeq, and eSPAN. (a) A flowchart and (b)
a schematic of the experimental procedures. Yeast cells arrested in G1 were released into fresh medium
containing HU and BrdU at 30  C for 45 minutes. Early S phase cells were collected for BrdU-IP-ssSeq, ChIP-
ssSeq, and eSPAN. After cells were physically broken and chromatin was sheared by sonication, a fraction of
the chromatin (2–5%) was saved as input as well as for BrdU-immunoprecipitation (BrdU-IP). The rest of the
chromatin (95–98%) was used for ChIP of the protein of interest. After the purification of ChIP DNA, ChIP DNA
was separated into two parts, one (13%) was used directly for ssDNA library preparation (ChIP-ssSeq), the rest
(87%) was used for the enrichment and Sequencing of Protein-Associated Nascent DNA (eSPAN) procedure.
For each experiment, four DNA samples including input DNA, BrdU IP DNA, ChIP-ssSeq DNA, and eSPAN DNA
were used to prepare libraries following the single-stranded DNA library preparation protocol. The sequencing
reads of Watson strand (red) and Crick strand (green) were independently mapped to the yeast reference
genome

6. TBS (1) þ PMSF buffer (see Note 1): dilute 10 TBS by
adding nine times water in volume, then add 1/100 volume
PMSF(100) solution.
7. BrdU (e.g., Sigma B5002).

2.2 Chromatin 1. PBS (10): 1.37 M NaCl, 27 mM KCl, 80 mM Na2HPO4,

Immunoprecipitation and 20 mM KH2PO4 pH 7.4. 25.6 g of Na2HPO4·7H2O,
80 g of NaCl, 2 g of KCl and 2 g ofKH2PO4. Make up to 1 l
with water. Autoclave.
2. NaOH (10 N): Add 40 g NaOH to a suitable container. Add
distilled water to make solution up to 100 ml.
3. 37% PFA (freshly made): Weigh 4.7 g of paraformaldehyde and
transfer to a 50 ml falcon tube. Add 25 μl of NaOH (10 N) and
7.5 ml of 1 PBS buffer. Put the tube into a cup containing 2/
230 Chuanhe Yu et al.

B Pol ε ChIP C Pol ε eSPAN

.07 0.0007

.06 0.0006

.05 0.0005

Input %

Input %
.04 0.0004
BrdU IP
12.00 .03 0.0003

.02 0.0002
10.00
.01 0.0001
8.00
.00 0
Input %

ARS607 ARS607+8kb ARS607 ARS607+8kb

6.00

D Pol δ ChIP E Pol δ eSPAN

4.00
0.016 0.0004
2.00 0.014 0.00035
0.012 0.0003
.00
ARS607 ARS607+8kb 0.01 0.00025
Input %

Input %
0.008 0.0002
0.006 0.00015
0.004 0.0001
0.002 0.00005
0 0
ARS607 ARS607+8kb ARS607 ARS607+8kb

Fig. 2 Newly synthesized DNA and DNA Polymerases ε and δ are enriched at replicating DNA using real-time
PCR analysis. (a) BrdU Immunoprecipitation (IP) assay showed that BrdU was incorporated at the early
replication origin ARS607, but not 8-kb away from ARS607 (ARS607 þ 8 k). (b) and (c) DNA Pol ε and δ
associate with replicating DNA. (d) and (e) QPCR analysis of Pol ε and Pol δ eSPAN DNA. ChIP assays were
performed using antibodies against FLAG-tagged Pol2 or Pol32

3 of boiling water and agitate the tube once every 1 min until
PFA is completely dissolved. Bring to 12.5 ml with 1 PBS and
filter through a 0.45 μM filter. Use on the same day.
4. 2.5 M Glycine: dissolve 93.8 g glycine in 500 ml water. Store at
room temperature after autoclaving.
5. Glass beads, 425–600 μm in diameter (e.g., Sigma G8772),
acid washed and heat-sterilized.
6. Hypodermic needles (16 gauge  1 in).
7. 1 M HEPES/KOH, pH 7.5: Add 700 ml water to 238.3 g
HEPES in a 1 l cylinder, stir and adjust pH to 7.5 with KOH.
Make up to 1 l with H2O, filter and store at
20  C.
8. ChIP lysis buffer: 0.1 M HEPES/KOH, pH 7.5; 0.1 M NaCl;
0.01 M EDTA;1% TX-100; 1% Na-deoxycholate (deoxycholic
acid, sodium salt). Add 50 ml 1 M HEPES/KOH, pH 7.5;
10 ml 5 M NaCl; 1 ml 0.5 M EDTA; 5 ml TX-100 and 0.5 g
 Strand-Specific Analysis of DNA Synthesis and Proteins Association with. . . 231

Na-deoxycholate to a 1 l cylinder. Stir and make to 500 ml with

water. Store at 4  C.
9. Protein G-Sepharose beads (GE).
10. Protease inhibitor and antibiotic mix: PMSF (100): add
17.4 mg to 1 ml isopropanol and mix until fully dissolved;
AEBSF (4-(2-Aminoethyl) benzenesulfonyl fluoride hydro-
chloride; 200): add 50 mg into 1 ml water and mix until
fully dissolved; Benzamidine (1000): add 157 mg benzami-
dine into 1 ml water and mix until fully dissolved; Bacitracin
(100): 100 mg bacitracin in 1 ml water and mix until fully
dissolved. Store all at
20  C.
11. 1 M Tris, pH 8.0: Add 121.1 g Tris base to 0.8 l water, adjust
pH to 8.0 with HCl. Bring up to 1 l.
12. Tris/LiCl buffer: 0.5 M LiCl, 0.5% NP-40, 0.1% sodium deox-
ycholate, 0.1 M Tris–HCl (pH 8.0). For 500 ml, add 50 ml of
1 M Tris–HCl pH 8.0, 50 ml of 5 M LiCl (for 5 M LiCl,
dissolve 21.2 g LiCl in 90 ml distilled water. Adjust the volume
to 100 ml. Store at 4  C, 2.5 ml of 100% Nonidet P-40, and
0.5 g of sodium deoxycholate. Adjust the volume to 500 ml
with distilled water.
13. ChIP lysis buffer þ 0.5 M NaCl: add 72 μl 5 M NaCl to 1 ml
ChIP lysis buffer.
14. Tris/EDTA washing buffer: 50 mM Tris, pH 8.0;10 mM
EDTA. Mix 25 ml 1 M Tris, pH 8.0 and 10 ml 0.5 M
EDTA. Bring to 500 ml with H2O.
15. 2 TE buffer: 20 mM Tris, pH 8.0, 2 mM EDTA. 10 ml of
1 M Tris (pH 8.0) and 2 ml EDTA (0.5 M) are made to 100 ml
with distilled water.
16. Anti-flag antibody or specific antibody for your protein of
interest.
17. 10% Chelex-100 (BioRad): Add 1 g Chelex-100 to 10 ml
water. It can be stored at room temperature for several months.
18. Covaris E210 sonicator, or equivalent.

2.3 BrdU 1. BrdU IP buffer: dilute 10 PBS by adding nine volumes of
Immunoprecipitation water, and then add Triton X-100 to 0.0625% (v/v).
2. 1 TE buffer: dilute 2 TE buffer by adding an equal volume
of distilled water.
3. TE þ 1% SDS buffer: Mix 5 ml 10% SDS, 25 ml 2 TE buffer,
and 20 ml distilled water.
4. BrdU antibodies (BD bioscience Cat. No. 555627).
5. E.Coli tRNA (e.g., Sigma Cat. No. R1753).
6. Minielute PCR Purification Kit (QIAgene Cat. No. 28006).
7. SYBR Green PCR Master MIX (ThermoScientific 4,309,155).
232 Chuanhe Yu et al.

3 Methods

3.1 Yeast Cell 1. Grow 10 ml of yeast cells on a shaking platform at 30  C

Synchronization and overnight.
Sample Collection (See 2. Dilute into 100 ml fresh YPD medium to OD600 ~ 0.2 in the
Note 2) morning. When it reaches OD600 ~ 0.4–0.5, add 100 μl α-
factor (5 mg/ml) and continue to culture at 25  C (yeast cells
are released more synchronously at 25  C than at 30  C). After
1.5 h, add another 100 μl α-factor (5 mg/ml) and culture yeast
cells for another 1.5 h. During that time, prepare the 37.5%
paraformaldehyde solution and warm up 100 ml YPD þ 1.52 g
hydroxyurea þ40 mg BrdU medium to 30  C.
3. 15 min before the 3 h arrest, check the cells with a light
microscope to make sure that at least 90% yeast cells are in
G1 phase (pear shaped in morphology).
4. Pellet the cells by centrifugation at 1600  g for 10 min at 4  C.
Wash twice with 20 ml cold H2O. Resuspend cells in pre-
warmed 100 ml YPD containing 200 mM HU and 400 mg/l
BrdU.
5. Continue culturing cells on a shaker for 45 min at 30  C, add
3 ml fresh 37% paraformaldehyde, and incubate with shaking
for another 20 min at 25  C.
6. Quench formaldehyde with 6 ml 2.5 M glycine for 5 min at
25  C on a shaker.
7. Pellet the cells at 1600  g for 10 min. Then wash the cells
twice with 20 ml cold TBS buffer containing 1 mM PMSF.
Resuspend the cell pellet in 1 ml TBS þ PMSF buffer and
transfer into two 2 ml tubes (50 ml cell/tube) with screw
cap. Pellet the cells, remove the supernatant, and store at

80  C.

3.2 ChIP and DNA 1. Dissolve cell pellets in 0.1 ml ChIP lysis buffer with protease
Extraction inhibitors and antibiotic mix and add ~100 μl glass beads. Lyse
by bead beating for 30 s on, 1 min off, in a 4  C cold room.
Repeat bead beating for a total of four times. Punch a hole in
the bottom of the tube with a 16-gauge hot needle. Nest the
tube into a 1.5 ml empty Eppendorf tube and centrifuge at
1600  g for 1 min. All of the liquid and cell debris—without
beads—should now be at the bottom of the new tube.
2. Aspirate the supernatant (chromatin is in the cell debris frac-
tion). Resuspend the cell pellets in 0.25 ml ChIP lysis buffer
with proteinase inhibitor by pipetting up and down several
times.
3. To shear the chromatin, sonicate the resuspended cell lysate
with a Covaris E210 (10 s on, 30 s off) for 25 times in the tubes
 Strand-Specific Analysis of DNA Synthesis and Proteins Association with. . . 233

provided by Covaris (see Note 3). Transfer the lysate to a new

Eppendorf tube and spin at 10,000  g for 5 min at 4  C.
4. Collect the supernatant and transfer to a new Eppendorf tube.
Centrifuge at 10,000  g for another 10 min at 4  C. Save 50 μl
of the clarified lysate as whole cell DNA input and for the BrdU
IP (Fig. 1).
5. For the rest of clarified lysate (about 200 μl), add a suitable
amount of antibody (see Note 4) and incubate on a rotating
platform (20–30 rotations per min) at 4  C overnight.
6. The next day, clarify by centrifuging at 10,000  g for 10 min
at 4  C. Wash protein G beads twice with ChIP lysis buffer.
Transfer the supernatant into an Eppendorf tube containing
25 μl washed protein G beads.
7. Incubate at 4  C for 2 h on a rotating platform (20–30 rota-
tions per min).
8. Spin at 1000  g for 1 min and remove the supernatant. Wash
beads with the following buffers in order: (a) 1 ml cold lysis
buffer; (b) 1 ml cold lysis buffer with 0.5 M NaCl; (c) 1 ml cold
Tris/LiCl buffer; (d) 1 ml room temperature 1 Tris/EDTA
washing buffer. For each wash, rotate 5 min on a rotating
platform, spin the beads at 1000  g for 1 min, and remove
all liquid using a 16G needle.
9. Add 100 μl 10% Chelex to the washed beads. Vortex to mix
well.
For the 50 μl clarified lysate input, also add 0.1 ml 10% Chelex.
10. Boil for 10 min in a 100  C heat block, and then cool at room
temperature.
11. Add 5 μl 20 mg/ml proteinase K to the reaction tube, mix well,
and incubate at 55  C for 30 min while shaking at 600 rpm.
12. Boil beads for another 10 min in a 100  C heat block.
13. Spin at a top speed, save the supernatant.
14. Add 100 μl 2 TE to the pellet, vortex, spin down, save the
supernatant.
15. Mix the supernatants taken above (about 180 μl). For the ChIP
sample, save 30 μl for library preparation (ChIP-ssSeq, Fig. 1),
the rest (about 150 μl) will be subjected to BrdU IP (eSPAN,
Fig. 1).
16. For the input samples, also save 30 μl for input library prepara-
tion (input), the rest 150 μl will be used for BrdU IP (BrdU-
IP-ssSeq) described below

3.3 BrdU IP 1. Samples (including the input and ChIP) (about 150 μl) are
boiled 3 min in a 100  C heat block, and immediately cooled
on ice for 3 min. Add BrdU IP mix (15 μl 10 PBS, 1.35 ml
234 Chuanhe Yu et al.

BrdU IP buffer, 0.5 μl anti-BrdU Ab and 1 μl E.coli tRNA

(10 mg/ml)). Rotate for 2 h at 4  C.
2. Add 20 μl protein G sepharose bead slurry (washed twice with
cold BrdU IP buffer). Incubate at 4  C for 1 h on a rotor.
3. Spin at 800  g for 1 min and remove the supernatant. Wash
the beads three times using 1 ml cold BrdU IP buffer, 3 min
incubation each time, then wash beads once with 1 ml TE
buffer.
4. Add 100 μl elution buffer (1 TE þ 1%SDS), vortex to mix
well. Incubate at 65  C for 15 min. After a short spin, the
supernatant is collected for DNA purification.

3.4 ChIP DNA 1. Purify the recovered DNA (four DNA samples including DNA
Purification from the input sample, protein ChIP, BrdU-IP, and eSPAN for
each experiment) using the Mini-elute PCR purification kit
according to the manufacturer’s protocol.
2. After purification, DNA is eluted with 17 μl EB buffer. At this
point, samples can be stored at 4  C for several days or at

20  C for months.

3.5 Quantitative PCR 1. Dilute 2 μl purified DNA eightfold by adding 14 μl of nuclease-

Analysis of ChIP, BrdU free water. Use the diluted DNA for the analysis using real-time
IP, and eSPAN at DNA PCR. The remaining 15 μl of DNA from each sample is used
Replication Origins for ssDNA library preparation.
2. Make a 12.5 μl reaction with 2.5 μl of diluted DNA, 6.25 μl
SYBR Green PCR Master Mix (ThermoScientific 4,309,155),
0.5 μl primer mix (forward and reverse primer at 1.5 μM,
ARS607 and ARS607 þ 8 kb (see Note 5)) and 3.25 μl water.
3. Perform QPCR: 95  C for 3 min; 40 cycles of 95  C for 10 s,
61  C for 30 s, 72  C for 35 s; and a final melt curve stage,
consisting of 95  C for 15 s, cooling to 55  C and 80 repeats of
heating for 5 s, starting at 55  C and with 0.5  C increments
(Fig. 2).

3.6 ssDNA Library 1. Single-stranded DNA libraries are prepared according to [17].
Preparation and Next- Omit the endonuclease VIII and UDG treatments.
Generation 2. Perform parallel paired-end sequencing with the Illumina Hi-
Sequencing seq 2000 or 2500 platform (see Note 6).

3.7 Data Analysis 1. Align the sequencing reads to the yeast genome (sacCer3)
using Bowtie2 software (example shown in Fig. 3) [19].
2. Separate the Watson and Crick strand reads by Perl codes, then
calculate the genome-wide read coverage of Watson and Crick
strands using BEDTools [20].
 Strand-Specific Analysis of DNA Synthesis and Proteins Association with. . . 235

A B
BrdU IP-ssSeq

Protein ChIP-ssSeq
Pol e

Pole
5’ 3’
Pol δ
3’ 5’
Pole

Pol e

eSPAN
Origin
Pol δ

ARS507

C D
Lagging bias
2.5 1
Pol e Pol δ Leading bias
2
Log2 ratio of Watson/Crick

1.5
Indeterminable
1.5
0.5
1
Lagging/Leading

0.5
0 0 1.0
-0.5
-1
-0.5
-1.5
-2 0.5
-2.5 -1
-20k Origin +20k

Pol e Pol δ

Fig. 3 DNA polymerase Pol ε (Pol ε) and DNA polymerase Pol δ (Pol δ) showing leading and lagging strands
bias at HU-stalled DNA replication forks, respectively. (a) A cartoon showing Pol ε and Pol δ bind to the leading
and lagging strands, respectively. (b) A snapshot of BrdU IP-ssSeq, protein ChIP-ssSeq, and eSPAN peaks at
ARS507 for Pol ε and Pol δ. Red and green colors represent the Watson and Crick strands, respectively. (c) The
average bias pattern of Pol ε and Pol δ eSPAN peaks at HU-stalled DNA replication forks. A 200-bp sliding
window was used to scan from 20Kb upstream to 20 Kb downstream 20 kb of fired replication origins. (d) Dot-
and-box plot showing the bias pattern of Pol ε and Pol δ eSPAN peaks at the 134 individual early origins

3. Use both Watson and Crick strand reads from BrdU-IP-ssSeq

to call replication peaks using MACS software [21].
There are two methods to determine whether a protein of
interest is enriched at the leading and lagging strands of DNA
replication forks.
4. Analysis of the average bias pattern of eSPAN peaks at all fired
origins (Fig. 3c)
(a) Extract the regions with 20 kb surrounding the fired
replication origins using Perl codes.
236 Chuanhe Yu et al.

(b) Use a 200 bp sliding window to scan each replication

origin and calculate the log2 ratio of Watson/Crick strand
reads using Perl codes.
(c) Calculate the average log2 ratio of Watson/Crick strand
reads of all fired replication origins using Perl codes.
(d) email us for the detailed analysis protocol
5. Analysis of the bias pattern of eSPAN peaks at individual origins
(Fig. 3d)
(a) Based on the location of the replication origin and Watson
and Crick strands, split each BrdU peak region into the
following four quadrants: Watson strand on the left (WL),
Watson strand on the right (WR), Crick strand on the left
(CL), and Crick strand on the right (CR).
(b) Count the number of sequence reads in each four quad-
rants using Perl codes.
(c) Calculate the p-value to determine whether sequence
reads of leading-strand (WL þ CR) were significantly
different from lagging-strand reads (WR þ CL) at each
replication fork using the binomial distribution with 10
5
as the cut-off p-value.
(d) Calculate the log2 ratio at each replication origin using the
following formula: log2 (WL þ CR)/(WR þ CL), which
represents the log2 ratio of lagging/leading, to determine
whether the fork exhibited a lagging or leading bias
pattern.
(e) Separate the origins into one of three categories based on
p-values and the log2 ratio of leading/lagging sequence
reads.

Leading strand bias: p < 10
5 and log2 (WL þ CR)/

(WR þ CL) &lt; 0.
Lagging strand bias: p < 10
5 and log2 (WL þ CR)/
(WR þ CL) &gt; 0.
Indeterminable bias: p > 10
5.

4 Notes

1. PMSF is poisonous and not stable in water. TBSþ PMSF buffer

should be made just prior to use.
2. Although unsynchronized cells also work for eSPAN experi-
ments, synchronized early S phase cells work much better. In
 Strand-Specific Analysis of DNA Synthesis and Proteins Association with. . . 237

the presence of hydroxyurea, 2–3 kb regions at early replication

origins replicate, which help generate clean eSPAN results.
3. The sonication condition should be tested experimentally
before ChIP experiments. The average sheared DNA fragments
should be 300 bp–500 bp in length.
4. The amount of each antibody used depends on the target
protein and antibody specificity. It is better to determine the
amount of each antibody before going through the eSPAN
protocol.
5. In synchronized early S phase cells in the presence of hydroxy-
urea, early replication origin ARS607 is replicated, whereas the
distal site (ARS607 þ 8 k), which is 8 kb distal from the
ARS607, is not. Therefore, BrdU and targeted protein ChIP
and BrdU IP are expected to enrich at ARS607 compared to
the ARS607 þ 8 k (Fig. 2).
6. Follow the instructions for multiplex sequencing provided by
Illumina. Replace the first read primer by the custom primer
CL72 (ACACTCTTTCCCTACACGACGCTCTTCC)
because one ligation adaptor is deleted of 4 nt in the library
preparation step [17, 18]. We sequence a maximum of 32
samples per lane.

Acknowledgment

We thank Dr. Albert Serra Cardona for editing this protocol. This
work was supported by NIH grants GM118015 to Z.Z and C.Y. is
supported by the Edward C. Kendall Fellowship.

References
1. Stillman B (2005) Origin recognition and the biochem.71.110601.135425110601.135425
chromosome cycle. FEBS Lett 579 [pii]
(4):877–884. Doi:S0014-5793(04)01538- 5. Waga S, Stillman B (1998) The DNA replica-
8 [pii]10.1016/j.Febslet.2004.12.011 tion fork in eukaryotic cells. Annu Rev Bio-
2. Okazaki R, Okazaki T, Sakabe K, Sugimoto K, chem 67:721–751. doi:10.1146/annurev.
Sugino A (1968) Mechanism of DNA chain biochem.67.1.721
growth. I Possible discontinuity and unusual 6. Bell SD, Botchan MR (2013) The Minichro-
secondary structure of newly synthesized mosome maintenance replicative helicase. Cold
chains. Proc Natl Acad Sci U S A 59 Spring Harb Perspect Biol 5(11):a012807.
(2):598–605 doi:cshperspect.a012807v1 [pii]10.1101/
3. O’Donnell M, Langston L, Stillman B (2013) cshperspect.a012807cshperspect.a012807
Principles and concepts of DNA replication in [pii]
bacteria, archaea, and eukarya. Cold Spring 7. Bell SP, Kaguni JM (2013) Helicase loading at
Harb Perspect Biol 5(7.) doi:10.1101/cshper- chromosomal origins of replication. Cold
spect.a010108a010108 Spring Harb Perspect Biol 5(6.) doi:cshper-
4. Bell SP, Dutta A (2002) DNA replication in spect.a010124 [pii]10.1101/cshperspect.
eukaryotic cells. Annu Rev Biochem a010124
71:333–374. doi:10.1146/annurev.
238 Chuanhe Yu et al.

8. MacAlpine DM, Almouzni G (2013) Chroma- 16. Viggiani CJ, Knott SR, Aparicio OM (2010)
tin and DNA replication. Cold Spring Harb Genome-wide analysis of DNA synthesis by
Perspect Biol 5(8):a010207. doi:cshperspect. BrdU immunoprecipitation on tiling microar-
a010207 [pii]10.1101/cshperspect.a010207 rays (BrdU-IP-chip) in Saccharomyces cerevi-
9. Arias EE, Walter JC (2007) Strength in num- siae. Cold Spring Harb Protoc 2010(2.) pdb
bers: preventing rereplication via multiple prot5385. doi:2010/2/pdb.prot5385 [pii]
mechanisms in eukaryotic cells. Genes Dev 21 10.1101/pdb.prot5385
(5):497–518. doi:21/5/497 [pii]10.1101/ 17. Gansauge MT, Meyer M (2013) Single-
gad.1508907 stranded DNA library preparation for the
10. Aparicio OM, Stout AM, Bell SP (1999) Dif- sequencing of ancient or damaged DNA. Nat
ferential assembly of Cdc45p and DNA poly- Protoc 8(4):737–748. doi:nprot.2013.038
merases at early and late origins of DNA [pii]10.1038/nprot.2013.038
replication. Proc Natl Acad Sci U S A 96 18. Meyer M, Kircher M, Gansauge MT, Li H,
(16):9130–9135 Racimo F, Mallick S, Schraiber JG, Jay F, Pru-
11. Aparicio OM, Weinstein DM, Bell SP (1997) fer K, de Filippo C, Sudmant PH, Alkan C, Fu
Components and dynamics of DNA replication Q, Do R, Rohland N, Tandon A, Siebauer M,
complexes in S. cerevisiae: redistribution of Green RE, Bryc K, Briggs AW, Stenzel U,
MCM proteins and Cdc45p during S phase. Dabney J, Shendure J, Kitzman J, Hammer
Cell 91(1):59–69 MF, Shunkov MV, Derevianko AP, Patterson
12. Yu C, Gan H, Han J, Zhou ZX, Jia S, Chabes N, Andres AM, Eichler EE, Slatkin M, Reich
A, Farrugia G, Ordog T, Zhang Z (2014) D, Kelso J, Paabo S (2012) A high-coverage
Strand-specific analysis shows protein binding genome sequence from an archaic Denisovan
at replication forks and PCNA unloading from individual. Science 338(6104):222–226.
lagging strands when forks stall. Mol Cell 56 doi:10.1126/scien-
(4):551–563. doi:S1097-2765(14)00749-7 ce.1224344science.1224344 [pii]
[pii]10.1016/j.molcel.2014.09.017 19. Langmead B, Salzberg SL (2012) Fast gapped-
13. Pursell ZF, Isoz I, Lundstrom EB, Johansson read alignment with bowtie 2. Nat Methods 9
E, Kunkel TA (2007) Yeast DNA polymerase (4):357–359. doi:nmeth.1923 [pii]10.1038/
epsilon participates in leading-strand DNA rep- nmeth.1923
lication. Science 317(5834):127–130. doi: 20. Quinlan AR, Hall IM (2010) BEDTools: a
DOI 10.1126/science.1144067 flexible suite of utilities for comparing genomic
14. Nick McElhinny SA, Gordenin DA, Stith CM, features. Bioinformatics 26(6):841–842.
Burgers PM, Kunkel TA (2008) Division of doi:10.1093/bioinformatics/btq033btq033
labor at the eukaryotic replication fork. Mol [pii]
Cell 30(2):137–144. doi:10.1016/j.mol- 21. Zhang Y, Liu T, Meyer CA, Eeckhoute J, John-
cel.2008.02.022S1097-2765(08)00168-8 [pii] son DS, Bernstein BE, Nusbaum C, Myers
15. Viggiani CJ, Aparicio OM (2006) New vectors RM, Brown M, Li W, Liu XS (2008) Model-
for simplified construction of BrdU- based analysis of ChIP-Seq (MACS). Genome
incorporating strains of Saccharomyces cerevi- Biol 9(9):R137. doi:10.1186/gb-2008-9-9-
siae. Yeast 23(14–15):1045–1051. doi:10. r137gb-2008-9-9-r137 [pii]
1002/yea.1406
 Chapter 18

Analysis of Replicative Polymerase Usage by

Ribonucleotide Incorporation
Andrea Keszthelyi, Izumi Miyabe, Katie Ptasińska, Yasukazu Daigaku,
Karel Naiman, and Antony M. Carr

Abstract
Mapping the usage of replicative DNA polymerases has previously proved to be technically challenging. By
exploiting mutant polymerases that incorporate ribonucleotides into the DNA with a significantly higher
proficiency than their wild-type counterparts, we and others have developed methods that can identify what
proportion of each DNA strand (i.e., the Watson and Crick strands) is replicated by a specific DNA
polymerase. The incorporation of excess ribonucleotides by a mutated polymerase effectively marks, in
each individual cells, the DNA strand that is replicated by that specific mutated polymerase. Changes to
DNA polymerase usage can be examined at specific loci by Southern blot analysis while a global analysis of
polymerase usage can be achieved by applying next-generation sequencing. This genome-wide data also
provides a direct measure of replication origin efficiency and can be used to indirectly calculate replication
timing.

Key words Pu-seq, Fork restart, rNMP incorporation, Origins, HTP sequencing

1 Introduction

The division of labor for eukaryotic replicative polymerases was

initially mapped in S. cerevisiae using mutation spectra analysis
[1]. In this work, the mutational bias of a specific error-prone
mutant polymerase allows the assignment of specific base changes
to the mis-incorporation of dNMPs on either the Watson or the
Crick strand. An alternative approach was subsequently developed
that used Southern blot analysis to map ribonucleotide incorpora-
tion at specific loci in strains that contained a specific polymerase
mutation that increases the mis-incorporation of rNMPs [2]. Col-
lectively, these approaches demonstrated that Polε synthesized the
leading strand, Polδ synthesized the lagging strand and that this
division of labor between Polε and Polδ was conserved between
S. cerevisiae and S. pombe. The analysis of rNMP incorporation by
Southern blot analysis has also been used to demonstrate that Polδ

Marco Muzi-Falconi and Grant W. Brown (eds.), Genome Instability: Methods and Protocols, Methods in Molecular Biology,
vol. 1672, DOI 10.1007/978-1-4939-7306-4_18, © Springer Science+Business Media LLC 2018

239
240 Andrea Keszthelyi et al.

synthesizes both DNA strands after replication is restarted by

homologous recombination [3]. A limitation of such studies is
that they are focused on a defined locus.
More recently, the basic idea of exploiting polymerase mutants
with a higher promiscuity toward ribonucleotide incorporation has
provided a global picture of replication polymerase usage in both S.
cerevisiae and S. pombe by coupling the approach with next-
generation sequencing [4–8]. The subsequent genome-wide poly-
merase usage data also allowed the identification of origins of
replication throughout the genome and provided a direct measure
of their efficiency (the percentage of the population in which a
given origin is fired) [5]. This is of particular interest because
previous methods that measured replication dynamics genome-
wide were restricted to directly measuring replication timing and
only indirectly provide a measure of replication origin efficiency [9].
In this chapter, we describe two approaches to studying poly-
merase usage. Both are based on increased ribonucleotide incor-
poration by mutant DNA polymerases. The first provides
information on a specific locus and exploits Southern blotting.
The second—which we have termed Polymerase usage sequencing
(Pu-seq)—involves next-generation sequencing and provides a map
of polymerase usage throughout the whole genome. The first
method is cost efficient, but is restricted to a few defined loci. The
second method requires additional financial input but extends the
analysis to the whole genome and provides appreciably better
quantitative information.
Both the methods require strains with specific backgrounds. In
order to ensure specificity of ribonucleotide incorporation to an
individual polymerase, the steric gate of the given replicative
enzyme is mutated [10]. This results in promiscuity toward the
incorporation of ribonucleotides. However, ribonucleotides in the
genome are normally rapidly removed by ribonucleotide excision
repair (RER) [11]. In order to maintain the mis-incorporated
rNMP residues in the DNA, it is thus necessary to disable RER.
We have achieved this by deleting the catalytic subunit of RNase H2
(rnh201) (see Subheading 3.1). Once the ribonucleotides are stably
incorporated into the DNA by the mutant polymerase, DNA can be
prepared and the site of incorporation turned into a strand break in
the DNA backbone. We achieve this by using alkali treatment
(rNMPs are alkali labile), which also makes the DNA single
stranded (see Subheading 3.2). Fragmented ssDNA can subse-
quently be subjected to either Southern blotting with strand-
specific probes to visualize the extent of the fragmentation (see
Subheading 3.3) or it can be the starting point to create a library
for next-generation sequencing (see Subheading 3.4). During the
latter process, the 50 end of each fragment generated by alkali
treatment allows for the precise identification of the site of ribonu-
cleotide incorporation by the mutant polymerase.
 Analysis of Replicative Polymerase Usage by Ribonucleotide Incorporation 241

While this protocol describes the analysis of polymerase usage

by ribonucleotide mapping, ribonucleotides incorporated at physi-
ological levels into the DNA can also be mapped, i.e., without the
presence of specific polymerase mutations. In such experiments the
generation of a library for next-generation sequencing should be
carried out as described below.

1.1 Alkaline The specific fragmentation of the DNA at the site of the ribonu-
Cleavage of DNA cleotides can be achieved by alkaline treatment or by the use of
Containing specific enzymes such as RNase H2. During alkaline treatment the
Ribonucleotides phosphate backbone is cleaved as a result of a hydroxyl radical
attack on the 20 -OH group. The products of this reaction are a
cyclic 20 -30 -phosphate and a 50 -OH end. In contrast, cleavage by
RNase H2 results in 50 -phosphate and 30 -OH ends that are non-
ligatable and which thus require further processing for library
production. Alkaline treatment also denatures the DNA and thus
the analysis of alkali-treated DNA by either Southern blot or by
library production for next-generation sequencing must account
for this.

1.2 Required Genetic Because ribonucleotides are efficiently removed by ribonucleotide

Background excision repair (RER), downregulation of this system is required so
that a sufficient signal-to-noise ratio can be achieved either for
Southern blot analysis or for next-generation sequencing. In our
methodology we delete the rnh201 gene that encodes the catalytic
subunit of RNase H2. Theoretically, other means of ablating RER
would be adequate. The next-generation sequencing method, but
not the Southern blotting method, is sufficiently sensitive that it
can be used to analyze DNA purified from otherwise unaltered
strains to study physiological ribonucleotide incorporation and/
or repair. To follow polymerase usage specifically between the
Watson and Crick strands of the DNA, two specific polymerase
mutations, in Polε and Polδ respectively, are required that incorpo-
rate ribonucleotide with higher proficiency. In this protocol, we
describe a data analysis pipeline for analyzing sequence data derived
from paired strains containing a mutation in each of these two
replicative polymerases (see Subheading 3.4.6). However, the
method can be extended to the third major replicative polymerase
(Polα) and potentially other polymerases involved in replication or
repair.

1.3 Mapping Alkali-treated ssDNA from each of the two polymerase mutants can
Polymerase Usage be run on a denaturing gel and be analyzed using strand-specific
Locally or Genome probes by Southern blot (see Subheading 3.3. and Fig. 1). This
Wide method has been used to establish the polymerase usage during
canonical replication [2] and examine replication polymerase usage
during site-specific replication fork arrest and restart events [3]. In
each case, pairs of strand-specific probes were designed to separately
242 Andrea Keszthelyi et al.

Fig. 1 Detection of alkali-sensitive sites in genomic DNA by Southern blot. (a)

Schematic representation of strand-specific probe labeling. Blue dotted lines
show labeled DNA molecules. (b) Typical result. Left panel: Probed with the
lagging strand. Right panel: Probed with the leading strand. Genomic DNA from
S. pombe wild-type (lanes 1 and 5), rnh201Δ (lanes 2 and 6), rnh201Δcdc6
(polδ)-L591G (lanes 3 and 7), rnh201Δcdc20(polε)-M630F (lanes 4 and 8) are
subjected to alkaline gel electrophoresis without enzyme digestion (lanes 1–4)
or after digestion with XmaI (lanes 5–8)
Analysis of Replicative Polymerase Usage by Ribonucleotide Incorporation 243

reveal either the Watson or the Crick strand of a specific locus

following restriction enzyme digestion.
A far more comprehensive analysis has been achieved by
exploiting next-generation sequencing (see Subheading 3.4). In
this protocol, the position of the nucleotide on the 50 end of each
ssDNA fragment generated by the alkali treatment is mapped. This
nucleotide was originally adjacent to the ribonucleotide before the
alkali cleavage. In order to attach specific adapters (see Subheading
3.4.4, step 2) the alkali-treated ssDNA is first converted to dsDNA
using random primers (see Subheading 3.4.3). To ensure we can
assign strand specificity dUTP is used instead of dTTP in the dNTP
mix. This enables the selective digestion by the USER enzyme mix
of the newly synthesized complementary strand after adaptor liga-
tion (see Subheading 3.4.4, step 4). Index primers are next used to
amplify and create the library. If multiple samples are processed
sample-specific index primers allow multiplex runs. For a schematic
view on the complementary strand synthesis and library prepara-
tion, please refer to Fig. 2.
The library is sequenced on an Illumina sequencing platform to
result in at least ten million reads per sample (see Subheading 3.4.5).
Reads are then aligned to the reference genome and the count
numbers of each position (i.e., the relative amount of incorporated
ribonucleotides) are calculated (see Subheading 3.4.6, steps 1–6).

Fig. 2 Schematic representation of complementary strand synthesis and library

preparation. See details in the text (Subheadings 1 and 3.4)
244 Andrea Keszthelyi et al.

Fig. 3 Typical representation of the Pu-seq specific data analysis. Polε (red) and
Polδ (blue) usage on the Watson (top panel) and on the Crick strand (middle
panel) and origin efficiencies (lower panel)

These calculations can be carried out on data from any genetic

background and bespoke scripts can be then used for detailed
analysis according to the specific aim of the study. Here, we present
a data analysis tool (see Subheading 3.4.6, step 7) that is specific for
Pu-seq and generates datasets on the relative contribution of the
two main replicative polymerases Polε and Polδ. Origins are repre-
sented in this dataset as sharp reciprocal changes in the usage of
Polε and Polδ and the efficiencies of the origins are calculated by
determining the height of these changes. The pipeline presented
here identifies these positions as origins and calculates their effi-
ciency. A typical result from this analysis can be seen in Fig. 3.

1.4 Data Analysis— Next-generation sequencing should be carried out as paired-end

Generating Count sequencing where library fragments are sequenced from both direc-
Numbers tions to allow high accuracy of alignment. Paired-end sequencing
generates two FASTQ files—each containing the 30 or the 50 end
sequence of the fragments. These reads are aligned to the reference
genome using the bowtie2 program that is freely available (see
Subheading 3.4.6, steps 1–4). This process generates a single
SAM file from the two FASTQ files for each strain. The SAM file
contains information on the position where each fragment was
aligned, in which direction they were orientated and statistics on
the quality of each alignment (see https://samtools.github.io/hts-
specs/SAMv1.pdf). This information is subsequently converted
into count numbers by a bespoke perl program (pe-sam-to-bin-
count.pl) which we have deposited on the github website (https://
github.com/yasukasu/sam-to-bincount). Briefly, the program
finds the fragments where both ends were aligned concordantly
and defines the position of the base next to the 50 end of the original
fragment (corresponding to the ribonucleotide). These positions
are then divided into two groups, according to whether they were
 Analysis of Replicative Polymerase Usage by Ribonucleotide Incorporation 245

aligned to the Watson or to the Crick strand. Last, the reads are
sorted by chromosome and position, the genome is divided into
bins with a specified size (300 bp by default) and the numbers of
reads are counted for each bin. In this way two CSV files are
generated—one for the Watson and one for the Crick strand—
where each row corresponds to a bin and each column contains
the name of the chromosome, the middle position of the bin and
the counts of reads within that bin. These CSV files can be gener-
ated from any paired-end sequencing data and can be a starting
point for various bespoke analyses. In the next section, we describe
a Pu-seq specific pipeline that starts from the CSV files generated
from sequences from Polε and Polδ mutant strains that outputs
polymerase usage data for each strand in addition to a list of the
positions and the efficiency of origins.

1.5 Data Analysis— We provide a program written in the R language (https://github.

Pu-Seq-Specific com/yasukasu/Pu-seq.R) where the CSV files described above
Pipeline (that contain the counts of sequenced fragments from Polε and
Polδ mutants—both in an rnh201Δ background) are used as input.
The program generates WIG and BEDGRAPH files containing
data on polymerase usage related to chromosome position and
the efficiency of origins throughout the genome. These files can
be used to visualize the data in a genome browser (e.g., IGV).
To calculate the relative contribution of each polymerase on
each strand at each bin the program inputs four datasets; the counts
(C) on the forward (þ) and reverse () strands from both the Polδ
(δ) and from the Polε (ε) strains (C+δ, Cδ, C+ε, and Cε). These
counts are then normalized to enable a direct comparison between
them: at each position (1) the count is divided by the sum of all the
counts from that dataset (e.g., the normalized counts (N) at posi-
tion i on the forward strand (þ) from Polδ (N+δi) equals the counts
at this position divided by the sum of all the counts from the Polδ
dataset for the forward strand: N+δi ¼ C+δi/Σ C+δ). The ratio (R) of
the usage of each polymerase is calculated using the assumption
that each strand is synthesized by either Polδ or Polε (i.e., ignoring
the relatively small contribution of Polα). The ratio of Polδ on the
forward strand at each position is then calculated as R+δi ¼ N+δi/
(N+δi þ N+εi). The relative contribution of Polδ on the reverse
strand is similarly calculated (Rδi ¼ Nδi/(Nδi þ Nεi). The
corresponding equations are used to generate the datasets for
Polε usage on the forward and reverse strands: (R+εi ¼ N+εi/
(N+εi þ N+δi) and Rεi ¼ Nεi/(Nεi þ Nδi). These ratios can be
smoothed by a moving average specified by the user (see Subhead-
ing 3.4.6, step 7, Option --ma1).
Origins of DNA replication are represented by sharp transitions
between the usages of the two polymerases. These changes appear
as sudden increases in the Polδ usage on the reverse strand (Rδ)
and Polε usage on the forward strand (R+ε). To identify these steep
246 Andrea Keszthelyi et al.

positive slopes the differentials of these two datasets (i.e., the

differences of each neighboring data point) are calculated and
smoothed by a moving average (see Subheading 3.4.6, step 7,
Option --ma2). Positive peaks in a differential dataset represent
sudden change of polymerase usage in relation to chromosome
position. The positions of these peaks are then identified. To locate
origins above the noise, only peaks with a maximum above the 30th
percentile of all peaks are considered and peaks found within four
bins are merged. The positions of these peaks are subsequently
identified in the corresponding original polymerase usage dataset
and the differences between the local maxima and the minima
surrounding each peak (the heights of the positive changes) are
calculated. This difference, multiplied by 100, represents the effi-
ciency, expressed as a percentage, of the origin (i.e., in the popula-
tion what percentage of the cells fired that particular origin). Lastly,
only origins identified in both independent datasets (Polδ usage on
the reverse strand (Rδ) and Polε usage on the forward strand (R+ε)
calculated from each of the matched mutant strains) within four
bins are chosen and the average of their efficiencies from both
datasets is calculated.

2 Materials

2.1 Cell Growth, DNA 1. YE medium: 0.5% Difco Yeast Extract 3% Glucose.
Extraction, and Alkali 2. Liquid nitrogen.
Treatment
3. NIB buffer: 50 mM MOPS (pH 7.2), 17% Glycerol, 150 mM
potassium acetate, 2 mM MgCl2 (see Note 1).
4. Lyticase (Sigma; L4025-1MU).
5. 1% (wt/vol) SDS.
6. Glass slide.
7. Qiagen buffer G2.
8. RNase A.
9. 30% (wt/vol) N-lauroyl sarcosine.
10. 20 mg/ml proteinase K.
11. Qiagen 100/G Genomic-tip.
12. TE.
13. 1 M NaOH.

2.2 Detection of 1. Ethanol.

Alkali-Sensitive Sites 2. 1 N KOH.
in Genomic DNA by
3. Agarose.
Southern Blot
 Analysis of Replicative Polymerase Usage by Ribonucleotide Incorporation 247

4. 10 Alkaline electrophoresis buffer: 500 mM NaOH, 10 mM

EDTA. Prepare just before use. EDTA is necessary to prevent
the precipitation of Mg(OH)2 which may entrap DNA.
5. 1 Alkaline electrophoresis buffer: 50 mM NaOH, 1 mM
EDTA. Prepare just before use.
6. 6 Alkaline loading buffer: 300 mM NaOH, 6 mM EDTA,
15% (w/v) Ficoll400, 0.25% (w/v) Xylene Cyanol FF, 0.15%
(w/v) Bromocresol Green.
7. Neutralization buffer: 1 M Tris–HCl (pH 7.2), 1.5 M NaCl.
8. 0.5 μg/ml EtBr.
9. Uncharged nylon membrane.
10. 10 Standard Taq Reaction Buffer containing 15 mM MgCl2.
11. dNTP (2 mM each) without dCTP.
12. α-32P-dCTP.
13. Taq polymerase.
14. G-50 gel filtration spin column.

2.3 Genome-Wide 1. 2% (wt/vol) TBE gel in 0.5 TBE buffer.

Identification 2. 0.5 μg/ml acridine orange solution (2000 dilution in water
of Incorporated from 10 mg/ml stock).
Ribonucleotides 3. Macherey-Nagel gel extraction kit.
4. Qubit ssDNA assay kit.
5. 3 mg/ml of 8 N random primer.
6. 10 NEB 2.1 buffer.
7. 2 mM each dNTP with dUTP instead of dTTP.
8. T4 polymerase.
9. 0.5 M EDTA.
10. AMPure XP beads.
11. 80% (vol/vol) ethanol.
12. Magnetic rack.
13. High-sensitivity DNA chip.
14. NEBNext Ultra DNA library prep kit for Illumina.
15. NEBNext multiplex oligos for Illumina.
16. NEBNext USER enzyme.

3 Methods

3.1 Strain Generation 1. To follow ribonucleotide incorporation by mutant polymerases

deletion of rnh201, or another means of RNase H2 down-
regulation, is required.
248 Andrea Keszthelyi et al.

2. In our hands the following S. pombe DNA polymerase mutants

showed the optimal balance between a sufficient level of ribo-
nucleotide incorporation and retaining viability:
– Polε: cdc20-M630F.
– Polδ: cdc6-L591G.
– Polα: Pol1-L850 M.
For S. cerevisiae others have used the following equivalent
mutations [6, 7]:
– Polε: pol2-M644G.
– Polδ: pol3-L612G, pol3-L612 M.
– Polα: pol1-L868 M, pol1-Y869A.

3.2 Cell Growth, DNA 1. Primary culture:

Extraction, and Alkali Inoculate one colony of each strain into 10 ml of YE medium in
Treatment a test tube and incubate it overnight at 30  C with constant
rotation.
2. Inoculate 8  107 cells from the primary culture into 1 L of YE
at 30  C with 180 rpm shaking. Allow the culture to grow to
3–5  106 cells/ml. Collect the cells by centrifugation
(6000  g at 4  C, 5 min) (see Note 2).
3. Remove the supernatant and resuspend the cell pellet in 40 ml
H2O. Transfer the cells to Falcon tube and centrifuge again
(4000  g at 4  C, 5 min). After discarding the supernatant
immediately freeze the cells in liquid nitrogen. Cell pellets can
be kept at 80  C for extended periods of time.
4. Resuspend the cell pellet in 2 ml of NIB buffer containing
5 mg/ml lyticase and incubate for 20 min at 37  C. To check
cell lysis, mix 1 μl of cells with 9 μl of 1% (wt/vol) SDS on a
glass slide and observe the cells under the microscope. Lysed
cells should appear hollow. If insufficient cells have lysed con-
tinue incubation for a further 10 min and check again.
5. When lysis is judged sufficient (>70% of cells lyse), add 20 ml of
ice-cold water and mix. After centrifugation (at 4000  g for
10 min) remove the supernatant and gently resuspend the
pellet in 2 ml Qiagen buffer G2.
6. RNase treat the mix by adding 100 μl of 10 mg/ml RNase A,
followed by incubation for 30 min at 37  C.
7. Add 100 μl of 30% (wt/vol) N-lauroyl sarcosine and 100 μl of
freshly prepared 20 mg/ml proteinase K. After 30 min incuba-
tion at 55  C centrifuge the mixture at 4000  g for 15 min at
4  C and transfer the supernatant into an empty falcon tube. To
increase DNA yield, repeat the proteinase treatment on the
remaining pellet; resuspend it in 1 ml Qiagen buffer G2,
50 μl of 30% (wt/vol) N-lauroyl sarcosine, and 50 μl 20 mg/
 Analysis of Replicative Polymerase Usage by Ribonucleotide Incorporation 249

ml proteinase K. Incubate for 30 min at 55  C. After centrifu-

gation, transfer the supernatant to the same tube as the super-
natant from the first proteinase K treatment.
8. Using the combined supernatant, follow the Genomic-tip pro-
tocol for 100/G Genomic-tip described in the Qiagen Geno-
mic DNA Handbook. After precipitation collect the DNA by
centrifugation and resuspend it in 200 μl of TE.
9. Measure the DNA concentration with Nanodrop. The
expected yield is 250–500 ng/μl. DNA can be stored at
20  C for at least 1 month.

3.3 Detection of 1. Digest 5 μg of genomic DNA with appropriate restriction

Alkali-Sensitive Sites enzyme(s) (see Note 3).
in Genomic DNA by 2. Precipitate DNA with ethanol and resuspend in 28 μl of dis-
Southern Blot tilled water.
3.3.1 Alkaline Agarose 3. Add 12 μl of 1 N KOH and incubate for 2 h at 55  C.
Gel Electrophoresis (See 4. In the meantime, add the appropriate amount of agarose to a
Note 1) measured quantity of distilled water (in most cases, a 1% gel is
sufficient). Heat the slurry in a microwave oven until the aga-
rose dissolves.
5. After cooling the agarose solution to 55  C add 0.1 volume of
10 alkaline electrophoresis buffer and pour immediately into
the gel tray (see Note 4). Leave the gel until it is completely set.
6. Mount the gel in the electrophoresis tank. Add freshly prepared
1 alkaline electrophoresis buffer until the gel is completely
covered.
7. Add 8 μl of 6 alkaline loading buffer to the DNA samples
after step 3 is completed.
8. Load the samples into the wells and electrophorese at <5 V/
cm until the xylene cyanol has migrated approximately half of
the length of the gel (see Note 5).
9. Soak the gel in neutralization buffer and incubate for 30 min at
room temperature with gentle shaking.
10. Soak the gel in 0.5 μg/ml EtBr solution and incubate over-
night. Some other staining dyes, such as SYBR Gold, should
also work. This step can be skipped if it is judged that the
visualization of the total DNA used for the subsequent South-
ern transfer is unnecessary.
11. Transfer DNA to an uncharged nylon membrane according to
any standard Southern blot protocol.

3.3.2 Southern 1. Amplify the DNA fragment (100–500 bp) of interest using a
Hybridization with Strand- standard PCR protocol.
Specific Probes
250 Andrea Keszthelyi et al.

2. Purify the PCR product with Qiagen PCR purification or Gel

extraction kit.
3. Prepare the reaction mix for primer extension (see Note 6).
– 1 μl Template (1 μg/μl purified PCR product).
– 1 μl primer (10 μM, see Note 7).
– 2.5 μl 10 Standard Taq Reaction Buffer containing
15 mM MgCl2.
– 1 μl dNTP (2 mM each) without dCTP.
– 2 μl α-32P-dCTP.
– 0.1 μl Taq polymerase.
– Distilled water up to 25 μl.
4. Perform the reaction with the following conditions:
(a) 94  C for 5 min.
(b) 94  C for 1 min.
(c) 55  C for 30 s.
(d) 72  C for 1 min.
(e) Repeat (c) and (d) 25 times.
(f) 72  C for 5 min.
(g) Hold 4  C.
5. Purify the reaction using a G-50 gel filtration spin column.
6. Hybridization can be performed using any standard protocol.
7. Strip the membrane and re-probe with the other strand-specific
probe (Fig. 1).

3.4 Genome-Wide 1. Dilute 20 μg of the DNA in 70 μl of H2O and add 30 μl of 1 M

Identification of NaOH (final concentration of 0.3 M). Incubate for 2 h at
Incorporated 55  C.
Ribonucleotides
3.4.1 Alkali Treatment
of Genomic DNA

3.4.2 Size Selection The protocol can be executed without this step. However, we
of the Alkali-Treated found that size selection results in more efficient and reproducible
Genomic DNA complementary strand synthesis (see Subheading 3.4.3). If the
method is used for the identification of ribonucleotide incorpora-
tion in DNA from wild-type cells (i.e., no downregulation of RNase
H2, and/or no DNA polymerase mutation), it is unlikely that the
concentration of ribonucleotides in the DNA is sufficient to result
in small fragments after alkali treatment. In this case, it is thus not
possible to size select DNA fragments. Therefore, this step should
not be included in the protocol.
 Analysis of Replicative Polymerase Usage by Ribonucleotide Incorporation 251

1. Run 10 μg of DNA (50 μl) on a 2% (wt/vol) TBE gel in 0.5

TBE buffer for 2 h at 100 V (see Note 8).
2. Stain the gel in a 0.5 μg/ml acridine orange solution for 2 h at
room temperature with gentle shaking followed by destaining
in water overnight at room temperature with gentle shaking.
Changing the water a couple of times helps to decrease the
presence of background staining.
3. Visualize the gel under long-wave UV illumination. As acridine
orange stains ssDNA red and dsDNA green a red smear should
be observed.
4. Excise a slice of gel that contains ssDNA fragments between
300 and 500 bp and isolate the ssDNA from the gel slice with
the Macherey-Nagel gel extraction kit as instructed by the
manufacturer (see Note 9).

3.4.3 Complementary In our hand if the protocol is executed without size selection of the
Strand Synthesis (See ssDNA (see Subheading 3.4.2) klenow fragment from BioPrime
Fig. 2) DNA labeling system works with higher efficiency than T4 poly-
merase (see Note 10). If the size selection step was included follow
the protocol below.
1. Measure the ssDNA concentration with a fluorimeter using the
Qubit ssDNA assay kit and dilute 100 ng into a final volume of
30 μl in H2O.
2. Add 5 μl of 3 mg/ml of 8 N random primer and 5 μl of 10
NEB 2.1 buffer to the ssDNA and boil for 5 min at 95  C.
Immediately place it on ice for 5 min.
3. Add 5 μl of 2 mM each dNTP (use dUTP instead of dTTP),
4 μl of H2O and 1 μl of T4 polymerase.
4. Incubate the mixture at 37  C for 20 min. Immediately add 5 μl
of 0.5 M EDTA (pH 8) to stop the reaction.

3.4.4 Library Preparation 1. Purification of dsDNA.

(See Fig. 2) (a) Resuspend the AMPure XP beads by vortexing and add
99 μl (1.8) of beads to the reaction from the previous
step.
(b) Mix well by pipetting up and down and incubate for 5 min
at room temperature. Briefly spin the tube to collect any
sample from the wall of the tube.
(c) Place the tube on a magnetic rack and wait until the beads
separate from the supernatant (~5 min). Carefully remove
and discard the supernatant.
(d) Add 200 μl of freshly prepared 80% (vol/vol) ethanol to
the beads (leave the tube in the magnetic rack) and
252 Andrea Keszthelyi et al.

incubate at room temperature for 30 s, then carefully

remove and discard the supernatant.
(e) Repeat the wash step with 200 μl of freshly prepared 80%
(vol/vol).
(f) Keep the tubes in the magnetic rack and air-dry the beads
with the lid open for 10 min.
(g) Elute the DNA by adding 60 μl of nuclease-free water to
the beads. Mix well by pipetting up and down. Briefly spin
the tube to collect any sample from the wall of the tube.
Place it in the magnetic rack and wait until the solution
clears.
(h) Remove 58 μl of the supernatant and transfer it to a clean
nuclease-free PCR tube.
(i) Run 1 μl on the Bioanalyzer for a quality check using a
high-sensitivity DNA chip (see Note 11).
2. End repair and adaptor ligation.
(a) Transfer 55.5 μl of the purified dsDNA to a PCR tube and
add 6.5 μl of NEBNext end repair reaction buffer (10)
and 3 μl of NEBNext end prep enzyme mix.
(b) Incubate the sample in a thermal cycler for 30 min at
20  C and then for 30 min at 65  C.
(c) Add the followings to the sample:
– 15 μl Blunt/TA ligase master mix.
– 1 μl 10 diluted NEBNext adapter (dilute the adapter
freshly in nuclease-free water).
– 2.5 μl nuclease-free water.
(d) Incubate the mixture for 15 min at 20  C in a thermal
cycler.
3. Size selection of the adaptor ligated fragment.
To ensure a consistent size of insert in the library and to discard
adaptor dimers and other unwanted residuals, a size selection
step is used at this point. For Illumina sequencing the optimal
fragment size is around 400 bp. However, to ensure a sufficient
yield, we have found it works better if we recover fragments of a
slightly higher size range (250–600 bp).
(a) Add 16.5 μl of nuclease-free water to the reaction to
adjust the final volume to 100 μl.
(b) Resuspend the AMPure XP beads by vortexing and add
35 μl (0.35) to the mixture.
(c) Mix well by pipetting up and down and incubate it for
5 min at room temperature.
Analysis of Replicative Polymerase Usage by Ribonucleotide Incorporation 253

(d) Briefly spin the tube to collect any sample from the wall of
the tube.
(e) Place the tube on a magnetic rack and wait until the beads
separate from the supernatant (~5 min) and then carefully
remove and transfer the supernatant to a new tube (see
Note 12).
(f) Add 35 μl (0.26) of resuspended AMPure XP beads to
the supernatant, mix well, and incubate for 5 min at room
temperature.
(g) Briefly spin the tube to collect any sample from the wall of
the tube.
(h) Place the tube on a magnetic rack and wait until the beads
separate from the supernatant (~5 min) and then carefully
remove and discard the supernatant.
(i) Add 200 μl of freshly prepared 80% (vol/vol) ethanol to
the beads (leave the tube in the magnetic rack) and incu-
bate it at room temperature for 30 s, then carefully remove
and discard the supernatant.
(j) Repeat the washing step twice with 200 μl of freshly
prepared 80% (vol/vol) for a total tree washes.
(k) Keep the tubes in the magnetic rack and air-dry the beads
for 10 min with the lid open.
(l) Elute the DNA by adding 25 μl of 0.1 TE, pH 8.0 to
the beads. Mix well by pipetting up and down.
(m) Briefly spin the tube and then place it in the magnetic rack
and wait until the solution is clear (~5 min), transfer 23 μl
to a new tube. As residual beads may affect the following
polymerase reaction be careful not to transfer any beads.
4. USER excision and PCR library enrichment.
(a) Take the Q5 Hot Start Hifi Master mix out of storage and
let it warm up to room temperature.
(b) Transfer 20 μl size-selected DNA to a PCR tube, and add
the followings:
– 3 μl NEBNext USER enzyme.
– 25 μl Q5 Hot Start Hifi Master mix.
– 1 μl Universal PCR primer (25 μM).
– 1 μl Sample-specific index primer (see Note 13)
(25 μM).
(c) Perform a PCR with the following conditions:
– 37  C for 15 min (USER digestion).
– 98  C for 30 s (Initial denaturation).
– 98  C for 10 s (Denaturation).
254 Andrea Keszthelyi et al.

– 65  C for 75 s (Anealing/extension).
– Repeat the above two steps nine times (Cycle numbers
might need optimalization see Note 14).
– 65  C for 5 min (Final extension).
– Hold 4  C.
5. Purification of the library
To ensure that there is no residual primer or adaptor dimer in
the final library, we routinely perform the AMPure purification
twice.
(a) Resuspend the AMPure XP beads by vortexing and add
50 μl (1) to the library.
(b) Briefly spin the tube to collect any sample from the wall of
the tube.
(c) Place the tube on a magnetic rack and wait until the beads
separate from the supernatant (~5 min) and then carefully
remove and discard the supernatant.
(d) Add 200 μl of freshly prepared 80% (vol/vol) ethanol to
the beads (leave the tube in the magnetic rack) and incu-
bate it at room temperature for 30 s, then carefully remove
and discard the supernatant.
(e) Repeat the washing step once with 200 μl of freshly
prepared 80% (vol/vol) for a total of two washes.
(f) Keep the tubes in the magnetic rack and air-dry the beads
for 10 min with the lid open.
(g) Elute the DNA by adding 51 μl of nuclease-free water to
the beads. Mix well by pipetting up and down.
(h) Briefly spin the tube and then place it in the magnetic rack
and wait until the solution is clear (~5 min), transfer 50 μl
to a new tube.
(i) Repeat the clean-up procedure from steps (a) to (f).
(j) Elute the DNA by adding 23 μl of nuclease-free water to
the beads. Mix well by pipetting up and down.
(k) Briefly spin the tube and then place it in the magnetic rack
and wait until the solution is clear (~5 min), transfer 20 μl
to a new tube. As residual beads may affect the following
sequencing reactions, be careful not to transfer any beads.
(l) Dilute 1 μl of the purified library in 4 μl of nuclease-free
water and run 1 μl from this dilution in an Agilent DNA
high-sensitivity chip to assess library quality (see Note 15).

3.4.5 Library Sequencing For Pu-seq analysis, sequence the library on an Illumina high-
throughput sequencing system. We aim for a minimum of ~10
million reads from each strain. We use the NextSeq system with
150 cycle paired-end sequencing. For details, see documentation on
the Illumina webpage (https://support.illumina.com/).
 Analysis of Replicative Polymerase Usage by Ribonucleotide Incorporation 255

3.4.6 Data Analysis Below are commands in Linux syntax which can be used to output
the Pu-seq information from paired end reads stored in FASTQ
formats.
1. Download and install Bowtie2, follow the instructions on the
Bowtie2 homepage (http://bowtie-bio.sourceforge.net/
bowtie2/).
2. Create an index from a reference genome FASTA file with the
following command (see Note 15):

filepath/bowtie2-build filepath/reference_FASTA_file index_-

name

3. Align the paired end reads (R1 and R2 index files) from the
library created from the polymerase δ mutant background,
generating a single SAM file.

filepath/bowtie2 –x filepath/index_name –trim5 1 –trim3 30 -1

pol-d_R1.fastq -2 pol-d-R2.fastq –S pol-d.sam

This command uses parameters that give good genome-wide

alignment of reads to the S. pombe reference genome. They can be
modified for better coverage of, e.g., repetitive regions. For more
options see Bowtie2 online manual (http://bowtie-bio.
sourceforge.net/bowtie2/manual.shtml).
4. Align the paired end reads from the library created from the
polymerase ε mutant background, generating a single SAM file.

filepath/bowtie2 –x filepath/index_name –trim5 1 –trim3 30 -1

pol-e_R1.fastq -2 pol-e-R2.fastq –S pol-e.sam

5. Download the custom-made Perl program pe-sam-to-bin-

count.pl from https://github.com/yasukasu/sam-to-
bincount and use it to extract concordantly aligned reads
from the SAM file generated from the polδ dataset, sort them
according to forward and reverse strand and calculate the num-
ber of reads in each bin.

perl filepath/pe-sam-to-bincount.pl –i pol-d.sam –trim5 1 –

strand –end 1 -n 1 -w 300 -ref filepath/reference_FASTA_file

This outputs two sets of files to the working directory; one for
the reverse and one for the forward strand. The CSV files contain
the numbers of reads in each bin (size of which is set by --w). The
counts are calculated by the position adjacent to the 50 end of the
R1 mate, set by the --end and -n parameters. These parameters
can be modified to extract a different subset of reads. Details of all
of the options can be found by typing in the help command –
256 Andrea Keszthelyi et al.

perl filepath/pe-sam-to-bincount.pl -h

In addition to the CSV files, for each strand an ALN and a

SORT.ALN file will be created. These files are intermediate files for
the final CSV file and contain a list of the position on each read (in
the SORT.ALN file these positions are sorted by chromosome and
position). The ALN and SORT.ALN files are not necessary for
further Pu-Seq analysis.
6. Extract concordantly aligned reads from the SAM file gener-
ated from the polε dataset and sort them.

perl filepath/pe-sam-to-bincount.pl –i pol-e.sam –trim5 1 –

strand –end 1 -n 1 -w 300 -ref filepath/reference_FASTA_file

7. Download the custom-made R program Pu-seq.R from

https://github.com/yasukasu/Pu-seq.R and use it to trans-
form the raw count data from the four CSV files generated in
steps 4 and 5 into polymerase usage data and origin datasets
(see Note 16).

R < filepath/Pu-seq.R –vanilla –slave –args –prefix strain_-

name –df pol-d.e1.fw300.count.csv –dr pol-d.e1.r-w300.count.
csv –ef pol-e.e1.f-w300.count.csv –er pole.e1.r-w300.count.csv
-ma1 3 ma2 3 –ori

Execute this command in the working directory in which the

CSV files from steps 4 and 5 are located. It is very important that
each of the files is entered into the command at the appropriate
option (for example, the CSV files with counts from the polymerase
δ mutant background on the forward and reverse strands should be
entered as --df and --dr, respectively; similarly, for reads from the
polymerase ε mutant background).
A more detailed description of the options and how they can be
modified can be found on https://github.com/yasukasu/Pu-seq.R.
This command will generate four WIG files, describing the
usage of the two polymerase on the forward and reverse strands.
A BEDGRAPH file, with information about the origins of replica-
tion and their firing efficiencies, will also be created.
8. Visualize the WIG files and BEDGRAPH file using a genome
browser, e.g., IGV (https://www.broadinstitute.org/igv/).
We recommend overlaying the usage of polymerases δ and ε on
each strand, creating two panels, one for the forward strand and
another for the reverse (see Fig. 3).
 Analysis of Replicative Polymerase Usage by Ribonucleotide Incorporation 257

4 Notes

1. We have occasionally observed that prolonged incubation with

lyticase in NIB buffer caused degradation of DNA, which
affected reproducibility of our Southern analysis. This problem
can be resolved by replacing 2 mM MgCl2 with 40 mM EDTA
in NIB buffer and using 1 mg/ml Zymolyase 100T as opposed
to lyticase to remove the cell wall.
2. It is important not to let the cultures overgrow and to grow
different strains to the same cell concentration: we have found
that different growth states apparently result in different ribo-
nucleotide incorporation levels.
3. A restriction digestion is not always necessary. However, it
helps to detect alkali sensitivity at a specific genomic locus.
Using the S. pombe strains described above, the alkaline degra-
dation of DNA is obvious within a restriction fragment as small
as 2 kb.
4. Because hydrolysis of the polysaccharides occurs in alkaline
condition at high temperature, NaOH should be added after
cooling the agarose solution to <55  C.
5. The appropriate voltage depends on the equipment to be used.
The gel and buffer temperature should be kept lower than
40–45  C to prevent alkaline hydrolysis of the gel. Gel tanks
with a large buffer volume, or a buffer cooling system, will
allow the gel to be run at high voltages. Therefore, the voltage
should be adjusted adequately to the equipment.
6. Nonradioactive nucleotide analogues such as fluorescein-12-
dUTP should work as well as radioisotopes. In this case,
unmodified nucleotide should be added at the appropriate
ratio. For example, fluorescein-12-dUTP and unmodified
dTTP are frequently used in the ratio of 0.35:0.65.
7. The individual primers used to amplify the template should
work well for the labeling reaction, assuming that the original
PCR is specific. Designing internal primers may help to avoid
the labeling of nonspecific PCR products.
8. To avoid overloading the gel, divide the sample into two and
run 25–25 μl (5–5 μg) per lane.
9. In our hands freezing the gel slice on 20  C for at least 20 min
before extraction increases the final yield of isolated ssDNA.
10. If the size selection step is not performed during the protocol,
we recommend the use of Klenow fragment obtained from the
BioPrime DNA labeling system instead of T4 polymerase. In
this case, starting from 100 ng ssDNA, follow the instruction
manual provided by the supplier but use 2 mM each dNTP mix
258 Andrea Keszthelyi et al.

with dUTP substituted for dTTP instead of the labeled dNTP

mix provided in the kit and reduce the length of incubation on
37  C to 40 min. In our hands, the Klenow fragments obtained
from other manufacturers did not perform as well as the Kle-
now fragment from the BioPrime DNA labeling system.
11. Our typical yield is at least 300–600 pg/μl dsDNA from size-
selected ssDNA and ~100 pg/μl dsDNA from non-size-
selected ssDNA. The size distribution is usually between 300
and 500 bp from size-selected ssDNA and somewhat smaller
(200–400 bp) from a non-size-selected template.
12. Do not discard the supernatant. At this step due to the PEG
concentration added with the AMPure XP only the large
(>600 bp) unwanted fragments are precipitated and bound
to the beads.
13. Multiplex sequencing is possible by using sample-specific index
primers. In this case several samples can be run on the same
sequencing lane. On a Next-seq platform we typically mix
28–30 samples which results in ~10 million paired-end reads
from each sample.
14. A clear peak of 300–600 pg/μl DNA between 200 and 500 bp
should be visible in the bioanalyzer profile. An additional peak
at higher molecular weight could be the result of over-
amplification and thus the number of PCR cycles should be
reduced. On the other hand, if an insufficient quantity of DNA
is present, additional PCR cycles could be applied. Additional
peaks at 50–150 bp indicate the presence of an adapter dimer.
In this case, it is advised to clean up the library again with 1
AMPure beads.
15. Do not copy paste commands from the Windows to Linux, as
this often creates syntax errors.
16. R software environment can be downloaded from https://
www.r-project.org/.

Acknowledgment

We thank Conrad Nieduszynski, Carolin Muller, and members of

the Carr and Murray labs for discussion.

References

1. Pursell ZF et al (2007) Yeast DNA polymerase evolutionarily conserved. PLoS Genet 7(12):
epsilon participates in leading-strand DNA rep- e1002407
lication. Science 317(5834):127–130 3. Miyabe I et al (2015) Polymerase delta repli-
2. Miyabe I, Kunkel TA, Carr AM (2011) The cates both strands after homologous
major roles of DNA polymerases epsilon and recombination-dependent fork restart. Nat
delta at the eukaryotic replication fork are Struct Mol Biol 22(11):932–938
 Analysis of Replicative Polymerase Usage by Ribonucleotide Incorporation 259

4. Keszthelyi A et al (2015) Mapping ribonucleo- 8. Ding J et al (2015) Genome-wide mapping of

tides in genomic DNA and exploring replica- embedded ribonucleotides and other nonca-
tion dynamics by polymerase usage sequencing nonical nucleotides using emRiboSeq and
(Pu-seq). Nat Protoc 10(11):1786–1801 EndoSeq. Nat Protoc 10(9):1433–1444
5. Daigaku Y et al (2015) A global profile of 9. Retkute R, Nieduszynski CA, de Moura A
replicative polymerase usage. Nat Struct Mol (2012) Mathematical modeling of genome
Biol 22(3):192–198 replication. Phys Rev E Stat Nonlinear Soft
6. Clausen AR et al (2015) Tracking replication Matter Phys 86(3 Pt 1):031916
enzymology in vivo by genome-wide mapping 10. Brown JA, Suo Z (2011) Unlocking the sugar
of ribonucleotide incorporation. Nat Struct “steric gate” of DNA polymerases. Biochemis-
Mol Biol 22(3):185–191 try 50(7):1135–1142
7. Reijns MA et al (2015) Lagging-strand replica- 11. Sparks JL et al (2012) RNase H2-initiated
tion shapes the mutational landscape of the ribonucleotide excision repair. Mol Cell 47
genome. Nature 518(7540):502–506 (6):980–986
 Chapter 19

Dynamic Architecture of Eukaryotic DNA Replication Forks

In Vivo, Visualized by Electron Microscopy
Ralph Zellweger and Massimo Lopes

Abstract
The DNA replication process can be heavily perturbed by several different conditions of genotoxic stress,
particularly relevant for cancer onset and therapy. The combination of psoralen crosslinking and electron
microscopy has proven instrumental to reveal the fine architecture of in vivo DNA replication intermediates
and to uncover their remodeling upon specific conditions of genotoxic stress. The replication structures are
stabilized in vivo (by psoralen crosslinking) prior to extraction and enrichment procedures, allowing their
visualization at the transmission electron microscope. This chapter outlines the procedures required to
visualize and interpret in vivo replication intermediates of eukaryotic genomic DNA, and includes an
improved method for enrichment of replication intermediates, compared to previously used BND-cellulose
columns.

Key words Electron microscopy, DNA replication, Psoralen crosslinking, In vivo replication inter-
mediates, Replication fork reversal, ssDNA, Nucleosome position

1 Introduction

Visualizing DNA replication intermediates (RIs) by structural

approaches has proven invaluable to complement standard cell
and molecular biology studies on the DNA replication process.
Owing to the high magnification that can be achieved, transmission
electron microscopy (EM) has helped uncovering the fine architec-
ture of DNA replication forks. Since its original establishment by
Dr. José Sogo at the ETH Zurich, this EM approach improved our
understanding of the physiological process of genomic or episomal
DNA duplication, in different model systems ranging from bacter-
iophages to mammalian cells [1–4]. The same approach was later
extensively used to investigate specific conditions of replication
stress in yeast cells, revealing profound changes of RI architecture
upon fork stalling and other types of genotoxic stress [5–9]. More
recently, several studies have made use of this technique in higher
eukaryotic systems, uncovering the remodeling of replication forks

Marco Muzi-Falconi and Grant W. Brown (eds.), Genome Instability: Methods and Protocols, Methods in Molecular Biology,
vol. 1672, DOI 10.1007/978-1-4939-7306-4_19, © Springer Science+Business Media LLC 2018

261
262 Ralph Zellweger and Massimo Lopes

as a global response to cancer-relevant replication stress and identi-

fying cellular factors orchestrating such a response [10–19]. Over-
all, this approach has played a pivotal role in providing structural
insight in the DNA replication process and has showed the poten-
tial to confirm, disprove, or refine long-standing models and
dogmas in this field.
This chapter will focus on the experimental procedures
required to extract genomic RIs from yeast, mammalian cells, and
Xenopus egg extracts and analyze their fine architecture in vivo
using the Transmission Electron Microscope (TEM). We refined
several experimental steps of previously published protocols [20,
21]. In particular, we optimized a new protocol for RI enrichment
(see Subheading 3.7), which no longer relies on BND cellulose and
which showed improved enrichment of RIs over linear non-
replicating DNA fragments, increasing the speed of the analysis.
The technique described here is in principle reproducible in any lab,
provided that a TEM and a proper high vacuum evaporator of
carbon and platinum/carbon are available.
A crucial prerequisite for these investigations is in vivo psoralen
crosslinking of DNA, achieved by repetitive exposure of living cells
to tri-methyl-psoralen (TMP), followed by irradiation pulses with
365–366 nm monochromatic light. In vivo psoralen crosslinking is
critical for two reasons (Fig. 1): (a) the formation of inter-strand
crosslinks induced by this treatment prevents branch migration of
cellular DNA and generally stabilizes in vivo RIs during DNA
extraction and de-proteinization; (b) the inaccessibility of nucleo-
somal DNA to the crosslinks provides an opportunity (by proper
modification of the technique, see Subheading 3.9.2) to obtain
important information on nucleosome positioning in vivo and
nucleosome dynamics at DNA replication forks [2, 5, 22–24].
Under denaturing conditions, DNA strands are separated wherever
a nucleosome was positioned in vivo and are kept together by the
interstrand crosslinks induced in the linker DNA (Fig. 1), giving to
double-stranded DNA (dsDNA) the appearance of a string of
single-stranded DNA (ssDNA) bubbles (Figs. 1 and 8). In fact,
this technique has not only been coupled to EM, but also to more
standard molecular biology approaches to study in vivo chromatin
structure [25].
After in vivo-psoralen crosslinking, genomic DNA is extracted
from the cells by standard procedures, minimizing mechanical
forces to avoid the shearing of chromosomal DNA. DNA replica-
tion intermediates are then enriched by binding, washing, and
elution in a QIAGEN-genomic tip 20 column, taking advantage
of the high affinity of this resin to ssDNA, which is invariably
present at DNA replication forks. With optimal stringency, this
enrichment procedure allows an approximately 20-fold enrichment
of RIs, although the majority of the recovered molecules are still
linear duplex DNA (see Subheading 3.11, step 3). The non-
Dynamic Architecture of Eukaryotic DNA Replication Forks In Vivo. . . 263

Fig. 1 Graphic representation of the psoralen crosslinking procedure and

advantages for studies on DNA replication intermediates and nucleosome
dynamics (see Subheading 1 for details). For the sake of clarity, linker- and
nucleosomal-DNA are not drawn to scale in the graphical representation

saturating conditions described here for RI enrichment exclude

that the recovered RI pool is biased toward molecules containing
longer ssDNA stretches.
Once the DNA sample is enriched for RIs, it is usually concen-
trated in size-exclusion columns and used for protein-free DNA
spreadings in the presence of the cationic detergent benzyldimethy-
lalkylammoniumchloride (BAC), using water as hypophase, with a
modified version of the method originally described by Vollenwei-
der et al. [26]. The low molecular weight of this spreading agent
(compared to protein-based methods) generally allows a better
visualization of details (i.e., secondary structures) along the DNA
molecules and an easier identification of ssDNA regions.
The monolayer nucleic acid film is then absorbed to freshly
prepared carbon-coated grids and stained with uranyl acetate to
improve the contrast of DNA molecules (in particular of ssDNA,
[26, 27]). The grids are then subjected to flat angle-rotary shadow-
ing with Platinum, which allows for the visualization of individual
DNA molecules over the background granularity of the carbon
264 Ralph Zellweger and Massimo Lopes

support. The reproducibility of the shadowing angle is a crucial

point for the success of this technique. Once a high quality sample
is obtained (in terms of molecule concentration, unfolding, and
contrast over the background), a high number of RIs are photo-
graphed and carefully analyzed for important structural features
(presence of ssDNA, secondary structures, nucleosome dynamics,
etc.). Contour length measurements on the digital files comple-
ment the visual investigation and lead to accurate statistical analysis
of the RI population analyzed.

2 Materials

2.1 In Vivo Psoralen 1. Ice-cold deionized water (S. cerevisiae).

Crosslinking 2. Standard Petri dish, diameter 8.5 cm (S. cerevisiae).
(S. cerevisiae)
3. Ice-cold 1
PBS (mammalian cells).
4. 96-Well plates (Xenopus egg extracts).
5. Ice-cold EB buffer (Xenopus egg extracts).
6. Tissue culture dishes, 60
15 mm (mammalian cells).
7. 4,50 ,8-Trimethylpsoralen (TMP) stock solution: dissolve
200 μg/ml TMP (Sigma, cat. T6137) in Ethanol 100%. Stir
extensively until the compound is dissolved completely. The
solution can be stored at 4  C for at least 1 year. Stir briefly at
RT before each usage. Due to its DNA-modifying potential,
TMP-containing solutions should be handled with gloves, lab-
coat, and protection glasses.
8. Biolink Crosslinker 365 nm, cat. B89110 (see Note 1).
9. Monochromatic 365 nm lamps for Biolinker (replacement UV
bulbs, 8 W, cat. B89270).
10. UVP UVX Radiometer, with 365 nm sensor.
11. Freezing pack, flat, 1.5–2 cm thick.
12. Flat metal support (1 cm), precooled at 20  C. The surface
should be large enough to accommodate several Petri dishes
(to crosslink more samples at the same time), but small enough
to be easily inserted and removed from the available Cross-
linker. This support is placed on the top of the freezing pack, to
prevent heating of the samples during the incubation/irradia-
tion cycles required for psoralen crosslinking.

2.2 In Vivo Psoralen 1. Ice-cold deionized water (S. cerevisiae).

Crosslinking 2. Standard Petri dish, diameter 8.5 cm (S. cerevisiae).
(Mammalian Cells)
3. Ice-cold 1
PBS (mammalian cells).
4. 96-Well plates (Xenopus egg extracts).
 Dynamic Architecture of Eukaryotic DNA Replication Forks In Vivo. . . 265

5. Ice-cold EB buffer (Xenopus egg extracts).

6. Tissue culture dishes, 60
15 mm (mammalian cells).
7. 4,50 ,8-Trimethylpsoralen (TMP) stock solution: dissolve
200 μg/ml TMP (Sigma, cat. T6137) in Ethanol 100%. Stir
extensively until the compound is dissolved completely. The
solution can be stored at 4  C for at least 1 year. Stir briefly at
RT before each usage. Due to its DNA-modifying potential,
TMP-containing solutions should be handled with gloves, lab-
coat, and protection glasses.
8. Biolink Crosslinker 365 nm, cat. B89110 (see Note 1).
9. Monochromatic 365 nm lamps for Biolinker (replacement UV
bulbs, 8 W, cat. B89270).
10. UVP UVX Radiometer, with 365 nm sensor.
11. Freezing pack, flat, 1.5–2 cm thick.
12. Flat metal support (1 cm), precooled at 20  C. The surface
should be large enough to accommodate several Petri dishes
(to crosslink more samples at the same time), but small enough
to be easily inserted and removed from the available Cross-
linker. This support is placed on the top of the freezing pack, to
prevent heating of the samples during the incubation/irradia-
tion cycles required for psoralen crosslinking.

2.3 Psoralen 1. Ice-cold deionized water (S. cerevisiae).

Crosslinking of Sperm 2. Standard Petri dish, diameter 8.5 cm (S. cerevisiae).
DNA (Xenopus Egg
3. Ice-cold 1
PBS (mammalian cells).
Extracts)
4. 96-Well plates (Xenopus egg extracts).
5. Ice-cold EB buffer (Xenopus egg extracts).
6. Tissue culture dishes, 60
15 mm (mammalian cells).
7. 4,50 ,8-Trimethylpsoralen (TMP) stock solution: dissolve
200 μg/ml TMP (Sigma, cat. T6137) in Ethanol 100%. Stir
extensively until the compound is dissolved completely. The
solution can be stored at 4  C for at least 1 year. Stir briefly at
RT before each usage. Due to its DNA-modifying potential,
TMP-containing solutions should be handled with gloves, lab-
coat, and protection glasses.
8. Biolink Crosslinker 365 nm, cat. B89110 (see Note 1).
9. Monochromatic 365 nm lamps for Biolinker (replacement UV
bulbs, 8 W, cat. B89270).
10. UVP UVX Radiometer, with 365 nm sensor.
11. Freezing pack, flat, 1.5–2 cm thick.
12. Flat metal support (1 cm), precooled at 20  C. The surface
should be large enough to accommodate several Petri dishes
(to crosslink more samples at the same time), but small enough
266 Ralph Zellweger and Massimo Lopes

to be easily inserted and removed from the available Cross-

linker. This support is placed on the top of the freezing pack, to
prevent heating of the samples during the incubation/irradia-
tion cycles required for psoralen crosslinking.

2.4 Genomic DNA 1. Spheroplasting buffer: 1 M sorbitol, 100 mM EDTA pH 8.0,

Extraction 0.1% β-Mercaptoethanol, 100 U/ml Lyticase. This buffer is
(S. cerevisiae) (See freshly prepared from the following stocks: 2 M Sorbitol; 0.5 M
Note 2) EDTA pH 8.0; 14.3 M β-Mercaptoethanol (pure liquid); Lyti-
case (Sigma, cat. L4025) 1000 U/ml in water (store 1 ml
aliquots at 20  C).
2. RNaseA stock solution (10 mg/ml in Tris–HCl pH 7.4)
prepared according to [28], freeze aliquots at 20  C.
3. Proteinase K stock solution (20 mg/ml in water), freeze ali-
quots at 20  C.
4. Chloroform/isoamylalcohol 24:1. This solution can be stored
at RT in a hood for unlimited time.
5. Kimble glass tubes (KIMBLE HS No. 45500-30).
6. Solution I: 2% w/v CTAB (cetyltrimethylammoniumbromide),
1.4 M NaCl, 100 mM Tris–HCl pH 7.5, 25 mM EDTA
pH 8.0 (see Note 3). Filter Solution I and Solution II to
avoid the formation of aggregates during DNA preps. These
solutions can be usually stored for 2–3 months at RT. They
should be refiltered or freshly prepared if a precipitate is
detectable.
7. Solution II: 1% CTAB, 50 mM Tris–HCl pH 7.5, 10 mM
EDTA (see Note 3).
8. Solution III: 1.4 M NaCl, 10 mM Tris–HCl pH 7.5, 1 mM
EDTA.
9. Isopropanol.
10. 70% EtOH, stored at RT.
11. 1
TE buffer.

2.5 Genomic DNA 1. Lysis buffer (QIAGEN buffer C1: 1.28 M Sucrose; 40 mM
Extraction Tris–HCl pH 7.5; 20 mM MgCl2; 4% Triton X-100). Dissolve
(Mammalian Cells) 483.14 g sucrose, 4.06 g MgCl2·6H2O, and 4.84 g Tris base in
680 ml ddH2O. Add 42 g Triton X-100 (100%). Adjust the pH
to 7.5 with HCl. Adjust the volume to 1 L with ddH2O. Keep
at 4  C.
2. Digestion buffer (QIAGEN buffer G2: 800 mM Guanidine-
HCl; 30 mM Tris–HCl pH 8.0; 30 mM EDTA pH 8.0; 5%
Tween-20; 0.5% Triton X-100). Dissolve 76.42 g guanidine
HCl, 11.17 g Na2EDTA·2H2O, and 3.63 g Tris base in 600 ml
of ddH2O. Add 250 ml 20% Tween-20 solution and 50 ml 10%
Triton X-100 solution. Adjust the pH to 8.0 with NaOH.
Adjust the volume to 1 L with ddH2O. Store at RT.
 Dynamic Architecture of Eukaryotic DNA Replication Forks In Vivo. . . 267

3. Ice-cold ddH2O.
4. Ice-cold 1
PBS.
5. Proteinase K stock solution 20 mg/ml in ddH2O (Roche, cat.
03115852001). Store at 20  C.
6. Chlorophorm/isoamylalcohol 24:1.
7. Isopropanol.
8. Ethanol 70%.
9. 30 ml glass centrifugation tubes, KIMBLE HS No. 45500-30.
10. Eppendorf centrifuge 5810R, rotor A-4-81.
11. Sorvall Evolution RC, rotor HB-6 swinging
(8000 rpm ¼ 10,459
g).

2.6 Sperm DNA 1. EB EDTA buffer: 50 mM Hepes-KOH pH 7.5, 100 mM

Extraction (Xenopus KCl, 2.5 mM MgCl2,1 mM EDTA.
Egg Extracts) 2. Cy3™-dCTP (GE HealthCare cat. PA35021). Dilute 1:50 in
EB-EDTA buffer and store in aliquots at 20  C.
3. EB EDTA sucrose buffer: EB EDTA +30% sucrose.
4. Proteinase K stock solution 20 mg/ml in ddH2O (Roche cat.
03115852001). Store at 20  C.
5. Chlorophorm/isoamylalcohol 24:1.
6. Ethanol 70%.
7. 1
TE buffer.

2.7 DNA Digestion 1. RNase A (Ribonuclease A Type I-AS, Sigma-Aldrich, R5503)

and Enrichment 10 mg/ml. Store in aliquots at 20  C.
of Replication 2. RNase III (ShortCut RNase III, NEB M0245) 10 mg/ml.
Intermediates Store aliquots at 20  C.
3. QBT equilibration buffer. Dissolve 43.83 g NaCl, 10.46 g
MOPS (free acid) in 800 ml distilled water. Adjust the pH to
7.0 with NaOH. Add 150 ml pure isopropanol and 15 ml 10%
Triton X-100 solution (v/v). Adjust the volume to 1 L with
distilled water.
4. Stock solution 10 mM Tris–HCl pH 8.0, 300 mM NaCl. The
solution can be prepared in large volume (50 ml), stored at RT
in plastic tubes, and reused for different experiments. 5 M NaCl
and 1 M Tris–HCl pH 8.0 stocks are required to prepare this
and the following solutions.
5. Stock solution 10 mM Tris–HCl pH 8.0, 900 mM NaCl can be
prepared in large volume (50 ml), stored at RT in plastic tubes,
and reused for different experiments.
6. Stock solution 10 mM Tris–HCl pH 8.0, 1 M NaCl can be
prepared in large volume (50 ml), stored at RT in plastic tubes,
and reused for different experiments.
268 Ralph Zellweger and Massimo Lopes

7. Stock solution 10 mM Tris–HCl pH 8.0, 1 M NaCl, 1.8%

caffeine (w/v) can be prepared in large volume (50 ml), but
requires extensive incubation at 50  C and stirring or vortexing
to help caffeine dissolution. The stock needs to be filtered
(syringe filter, 0.2 μM), can be stored at RT in plastic tubes,
and can be reused for different experiments during 4–6 months.
A new stock should be prepared when a precipitate is
detectable.
8. QIAGEN Genomic-tip 20/G (QIAGEN, Cat. No. 10223).
9. Millipore size-exclusion columns: Amicon ultra 100K mem-
brane (Millipore, cat. UFC510096).
10. 1
TE buffer.

2.8 Preparation 1. High Vacuum Evaporator MED 020 (BalTec), with two EK
of Carbon-Coated 030 electron guns (C and Pt/C), quartz crystal, thin film
Grids monitor QSG 100, and control unit EVM 030 (see Note 4).
2. Mica Sheets (Plano, mica high grade quality V1, 25
76 mm,
cat. 56).
3. Wolfram cathodes for electron guns (BALTIC, BP2317).
4. Carbon rods (BALTIC, 3
50 mm, cat. BP2217 P ¼ 6).
5. Scotch solution: 20–30 cm of Scotch tape in 100 ml of chloro-
form in a glass bottle (Fig. 2a). After quick stirring of the
bottle, the chloroform turns yellowish, dissolving the tape
adhesive (the cellophane support does not dissolve). This solu-
tion can be stored at RT for at least 1 year.
6. Supporting Teflon-wire mesh stand (Fig. 2b; see Note 5).
7. Filter paper circles (Macherey-Nagel, 4310045, diameter:
45 mm).
8. Filter paper circles (Macherey-Nagel, 431009, diameter:
90 mm).
9. EM-grade water (see Note 6).
10. Copper 3.05 mm grids, 400 mesh (Plano, cat. G2400C;
see Note 7).

2.9 DNA Spreading 1. P2 pipette.

by the “BAC Method” 2. EM-grade water (see Note 6).
3. EtBr stock (10 mg/ml, Sigma, cat. E1510). It can be stored at
4  C for at least 1 year.
4. EtBr working solution (33.3 μg/ml). It is freshly prepared
before each set of spreadings, adding 1 μl of EtBr stock
(10 mg/ml) to 300 μl of EM-grade water.
5. Formamide (pure, Sigma-Aldrich, cat. 47680).
6. Glyoxal: 40% solution in water (Merck, cat. 4910).
 Dynamic Architecture of Eukaryotic DNA Replication Forks In Vivo. . . 269

Fig. 2 Series of photographs showing crucial steps in the preparation of carbon-coated grids (see Subheading
3.8 for details)

7. BAC stock: BenzyldimethylalkylAmmoniumChloride (Bayer-

Leverkusen, n-alkyl mixture: C12H25, 60%; C14H29, 40%;
see Note 8) 0.2% w/v in Formamide. This solution can be
stored at RT for at least 1 year.
8. BAC working solution (BAC 1:20): just before each set of
spreadings, the BAC stock is diluted 1:20 in 1
TE (20 μl of
the diluted stock are prepared in a microfuge tube and are
normally sufficient for all spreadings performed during the
day).
9. Ethanol 100%, molecular biology grade.
10. Uranyl acetate stock: uranyl acetate (UrAc, Fluka, cat. 73943)
5 mM in HCl 5 mM. This stock solution can be stored at 4  C
for at least 1 year. Uranyl acetate has a typical radioactivity of
0.37–0.51 μCi/g. This mild radioactivity level is not sufficient
to be harmful while the material remains external to the body.
The use of standard protective clothing (gloves, glasses, lab
coat) is therefore sufficient to work safely.
270 Ralph Zellweger and Massimo Lopes

11. Uranyl Acetate working solution (0.5 mM UrAc, 0.5 mM

HCl, 90% Ethanol): just before each set of spreadings, the
UrAc stock is diluted 1:10 in 100% Ethanol.
12. Round filter paper (Macherey-Nagel, 431009, diameter:
90 mm).
13. Tissue culture dishes, 60
15 mm (see Note 9).
14. Mica sheets (Plano, mica high grade quality V1, 25
76 mm,
cat. 56).
15. Graphite powder (BalTec, cat. LZ 02096 VN).
16. Fine tweezers with bent points (see Note 10) or Dumont
pinzette (Plano T539).

2.10 Platinum- 1. High Vacuum Evaporator MED 020 (BalTec; see Note 4), with
Carbon Rotary two EK 030 electron guns (C and Pt/C), quartz crystal, thin
Shadowing film monitor QSG 100, control unit EVM 030, rocking rotary
stage, and specimen table (54 mm) for 20 grids (3.05 mm,
Fig. 4a; see Note 11).
2. Carbon/Platinum rods: Carbon rods 2
20 mm (BALTIC
BP2260); platinum insets (BALTIC BP2261).
3. Micrometric control of rocking rotary stage angle: the standard
knob of the MED020 is substituted by a Precision Rotation
Platform PR01 with Adapter Plate PR01A, Thorlabs, Newton,
NJ, USA (see Note 12; Fig. 4c and d, black arrow).

2.11 Visualization 1. Transmission Electron Microscope, connected to a computer-

at the Transmission driven CCD-camera. Minimal resolution required is 1k
1k if
Electron Microscope, the camera is mounted on the bottom port, while it should
Contour Length preferably reach 2k
3.5k if mounted on the side 35 mm port.
Measures, and 2. Software for Camera Control and storage of acquired images in
Statistics original and *.tiff format (i.e., Digital Micrograph).
3. Software package to produce a composite image (montage)
from adjacent images.
4. ImageJ or other software allowing contour length measure-
ments on *.tiff files.
5. Standard statistic applications (i.e., Microsoft Office Excel).
6. Large hard disk capacity or external hard disk for storage of a
high number of large image files (see Note 13).

3 Methods

3.1 In Vivo Psoralen 1. Samples are typically collected at different time points during
Crosslinking synchronization experiments (see Note 14). Every sample cor-
(S. cerevisiae) responds to 4
109–1
1010 cells (400 ml of culture
 Dynamic Architecture of Eukaryotic DNA Replication Forks In Vivo. . . 271

1–2.5
107 cells/ml). Spin cells down at 3200
g for 10 min.

Resuspend the cells in ice-cold water, transfer the suspension to
50 ml tubes. Spin cells down at 3200
g for 5 min at 4  C (see
Note 15).
2. Resuspend the pellet in 20 ml of ice-cold water. Transfer cells
into an 8.5 cm diameter Petri dish. Several samples can be
simultaneously crosslinked with the following procedure.
3. Install the five monochromatic 365 nm lamps in the cross-
linker. Make sure that they are all properly inserted and that
they all light up when starting the crosslinker.
4. Add 1 ml TMP stock solution (10 μg/ml final concentration)
to the cell suspension in the Petri dish. Mix well with a pipette
and incubate for 5 min in the dark on the precooled metal
support (see Note 16). Place the precooled metal support with
the Petri dishes on the freezing pack. Irradiate the sample for
3 min.
5. Repeat step 4 once (see Note 17).
6. Transfer the cell suspension to a 50 ml tube and keep it on ice.
Wash the dish twice with 1 ml water to remove all the cells and
pool the washes in the same 50 ml tube. Spin down the cells at
3200
g for 10 min. Use the cell pellet for DNA extraction (see
Subheading 3.4).

3.2 In Vivo Psoralen 1. Samples can be collected from asynchronously growing cells or
Crosslinking at different time points in synchronization experiments (see
(Mammalian Cells) Note 14). Every sample corresponds to 2.5–5.0
106 cells
(150
20 mm tissue culture dish, 50–80% confluency for
U2OS cells). After standard trypsinization (or collection, for
cells in suspension), transfer the cells to 15 ml Falcon tubes and
spin them down at 600
g for 5 min. Wash the cell pellet once
(by resuspension/centrifugation) with 5 ml ice-cold 1
PBS
(see Note 15).
2. Resuspend the pellet in 10 ml ice-cold 1
PBS. Transfer cells
into a tissue culture dish 60
15 mm. Several samples can be
simultaneously crosslinked with the following procedure.
3. Insert the five monochromatic 365 nm lamps in the crosslinker.
Make sure that they are all properly inserted and that they all
light up when starting the crosslinker.
4. Add 0.5 ml of TMP stock solution (10 μg/ml final concentra-
tion) to the cell suspension in the Petri dish. Mix well with a
pipette and incubate for 5 min in the dark on the precooled
metal surface (see Note 16). Place the metal support with the
Petri dishes on the top of the freezing pack. Irradiate the
sample for 3 min.
5. Repeat step 4 once (see Note 17).
272 Ralph Zellweger and Massimo Lopes

6. Transfer the suspension back to a 50 ml tube and keep it on ice.

Wash the dish twice with 1 ml 1
PBS to remove all cells. Spin
down the cells at 600
g for 5 min. Resuspend the cell pellet in
2 ml 1
PBS and proceed with DNA extraction (see Subhead-
ing 3.5).

3.3 Psoralen 1. Samples are typically collected 30–60 min after sperm nuclei
Crosslinking of addition in the Xenopus extracts. Replicating interphase
Replicating Sperm extracts are prepared as described in [29]. Every sample corre-
DNA (Xenopus Egg sponds to 200–300 μl of extract to which 4000 nuclei/μl are
Extracts) added.
2. Split the sample into 100 μl aliquots in microfuge tubes and
incubate at 23  C.
3. (Optional) To control the timing of replication in the extract,
add 5 μl Cy-3 dCTP to 100 μl of the sample (extract + sperm)
and check visible incorporation by standard immunofluores-
cence microscopy (see Note 18).
4. When 80–90% of the nuclei start incorporating Cy-3 dCTP
(ideally 45–50 min after sperm addition), arrest the DNA
replication by diluting each 100 μl aliquot with 200 μl of cold
EB-EDTA buffer and incubate on ice.
5. Pool the aliquots corresponding to each sample.
6. Underlay the samples with 2 volumes of cold EB-EDTA
Sucrose Buffer.
7. Spin at 8600
g at 4  C for 5 min.
8. Remove the supernatant carefully so as not to disturb the
pellet.
9. Resuspend the pellet in 100 μl of ice-cold EB-EDTA using a
P200 pipette with cut tips.
10. Transfer the samples to a precooled round-bottom microtiter
plate on a precooled metal support.
11. Insert the five monochromatic 365 nm lamp in the crosslinker.
Make sure that they are all properly inserted and that they all
light up when starting the crosslinker.
12. Add 5 μl of TMP stock solution (10 μg/ml final concentration)
to each 100 μl nuclei suspension and mix by pipetting with cut
tips. Incubate for 5 min in the dark on a precooled metal
support. Place the metal support with the microtiter plate on
the top of the freezing pack and irradiate the sample for 3 min.
13. Repeat step 12 two more times.
14. Recover the nuclei suspension from the wells into microfuge
tubes. Centrifuge at 5000
g for 5 min and resuspend nuclei
in 300 μl of EB-EDTA buffer using a precut 1 ml tip. Proceed
with DNA extraction (see Subheading 3.6).
 Dynamic Architecture of Eukaryotic DNA Replication Forks In Vivo. . . 273

3.4 Genomic DNA 1. Resuspend the cells in 5 ml of Spheroplasting buffer. Incubate

Extraction by CTAB at 30  C for 45 min. Invert the tube several times during
Method (S. cerevisiae) incubation. Spin down the spheroplasts at 6000
g for 10 min.
(See Note 2) 2. Resuspend the spheroplasts in 2 ml of distilled water and
rapidly add 2.5 ml of Solution I, 200 μl of RNaseA (10 mg/
ml). Incubate for 30 min at 50  C, then add 200 μl of Protein-
ase K (20 mg/ml). Gently mix the sample.
3. Incubate for 1.5–2 h at 50  C. Invert the tube several times
during incubation. Note that the solution has to become clear
(no clumps!); if necessary, use precut 1 ml tips to break clumps.
If clumps are still present after 2 h, add 100 μl of Proteinase K
(20 mg/ml) and incubate O.N. at 30  C.
4. Centrifuge for 10 min at 3200
g: keep the pellet (cellular
debris) for further extraction (see Subheading 3.4, steps 9–11).
Transfer the supernatant carefully to a 15 ml tube containing
2.5 ml of chloroform/isoamylalcohol 24:1 at RT.
5. Mix by inverting the tubes several times and spin at RT at
3200
g for 10 min. Note that a white protein layer is formed
between the two phases. Carefully remove the clear upper
phase and transfer it to a 30 ml Kimble glass tube.
6. Gently add 10 ml (2 volumes) of Solution II, cover with
parafilm and invert several times. The solution should slowly
become turbid. If necessary, incubate for 10–15 min at RT
until some turbidity is detectable.
7. Centrifuge for 10 min at 12,000
g in a proper swinging
bucket rotor and discard the supernatant.
8. Add 2 ml of Solution III and incubate the solution briefly at
37  C until the pellet is completely dissolved.
9. Resuspend the pellet from step 4 in 2 ml of Solution III, mix
vigorously, and incubate at 50  C for 1 h (use precut 200 μl tips
to help complete resuspension). Check that the solution is
finally homogeneous.
10. Transfer the solution carefully to a 15 ml tube containing 1 ml
of chloroform/isoamylalcohol 24:1 at RT.
11. Mix by inverting the tubes several times and spin at RT at
3200
g for 10 min. Note that a white protein layer is formed
between the two phases. Carefully remove the clear upper
phase (ca. 2 ml) and pool it in the Kimble glass tube from
step 8 (4 ml final volume).
12. Precipitate the DNA by adding 4 ml of isopropanol (RT).
Gently mix the sample to ensure a proper DNA precipitation.
Spin down the DNA at 12,000
g for 10 min in a swinging
bucket rotor.
13. Discard the supernatant and wash the pellet with 1 ml of
Ethanol 70% (RT).
274 Ralph Zellweger and Massimo Lopes

14. Remove as much ethanol as possible and briefly spin the pellet
to accumulate residual ethanol at the bottom of the tube.
Remove it using a 200 μl pipette. Further dry the pellet by
briefly incubating the open Kimble glass tube in a 37  C water
bath.
15. Add 200 μl of 1
TE to the dried pellet. Cover the Kimble
glass tube with parafilm and incubate for 30 min at 37  C (or
O/N at RT) to ensure proper resuspension of the DNA pellet.
Spin briefly and transfer the DNA solution to a microfuge tube
using a precut 200 μl tip, to avoid shearing of genomic DNA.
16. Check the quality and the concentration of DNA preps, as well
as possible RNA contamination (see Note 21) by agarose-gel
electrophoresis. The standard yield, starting with
4
109–1
1010 cells, is therefore 10–60 μg of genomic
DNA from each sample.

3.5 Genomic DNA 1. Add to the cell suspension in 1
PBS (Subheading 3.2, step 6)
Extraction 2 ml of ice-cold lysis buffer and 6 ml of ice-cold ddH2O.
(Mammalian Cells) 2. Mix by inverting the tube several times and incubate on ice for
10 min.
3. Spin the lysed cells at 4  C for 15 min at 1300
g. Discard the
supernatant.
4. Add 1 ml of ice-cold lysis buffer and 3 ml of ice-cold ddH2O.
Resuspend the pelleted nuclei completely by vortexing.
5. Spin the lysed cells at 4  C for 15 min at 1300
g. Discard the
supernatant completely.
6. Add 100 μl of ice-cold PBS and resuspend the nuclei with a cut
200 μl tip. Make sure that no clumps of cells are left.
7. Add 5 ml of digestion buffer. It is crucial NOT to VORTEX or
rotate the tube to avoid clumping of the cells in the center of
the tube. Add 200 μl of proteinase K stock and incubate at
50  C until the solution is clear (1–2 h).
8. Let the sample cool down to RT, then transfer it into a 50 ml
tube containing 5 ml of chlorophorm/isoamylalcohol. Close
the tube properly and invert it vigorously but carefully 30 times
(solution turns milky). Immediately pour sample into a glass
centrifugation tube.
9. Centrifuge at 4  C for 20 min at 10,500
g (phase separation).
10. Carefully transfer upper phase into a new glass tube using a cut
1 ml tip. Add the same volume of isopropanol and mix well to
precipitate the DNA.
11. Centrifuge at 4  C for 10 min at 10500
g.
12. Wash DNA with 70% ethanol. Spin at 4  C for 5 min at
10,500
g.
 Dynamic Architecture of Eukaryotic DNA Replication Forks In Vivo. . . 275

13. Air dry the pellet and resuspend in 1
TE buffer (200–400 μl).
Pay attention not to overdry DNA pellet to prevent irreversible
“agglutination” of DNA molecules. The standard yield is
10–50 μg of genomic DNA for 2.5–5.0
106 cells.

3.6 DNA Extraction 1. Add to the crosslinked nuclei (Subheading 3.3, step 14)
from Sperm Nuclei 1.5 μg/ml final concentration of proteinase K and incubate
(Xenopus Egg for 2 h at 50  C.
Extracts) 2. Add equal volume of phenol:chloroform:isoamylalcohol—
25:24:1 to the above solution and shake vigorously. Spin at
18,500
g for 10 min.
3. Transfer the supernatant to a new microfuge with a cut 1 ml
tip.
4. Add equal volume of 100% isopropanol to the supernatant to
precipitate the DNA. Incubate at 4  C for 10 min. Spin the
sample at 18,500
g for 10 min at 4  C. Discard the
supernatant.
5. Wash pellet with 500 μl of 70% ethanol and spin down the
sample for 5 min.
6. Discard the supernatant and briefly centrifuge at 1700
g.
Remove the residual ethanol carefully with a 20 μl pipette.
7. Incubate the tube at 37  C to evaporate the excess ethanol.
8. Resuspend the pellet in 100 μl TE.
9. Assess quality and quantity of the DNA by UV spectrophotom-
etry and by gel electrophoresis.

3.7 DNA Digestion 1. 10–15 μg (yeast, mammalian cells) or 5 μg of DNA (Xenopus)

and Enrichment is digested with 50–100 U of restriction enzyme in the proper
of Replication buffer for 3–5 h at 37  C. Standard restriction enzymes used
Intermediates are PvuI for S. cerevisiae genomic DNA, PvuII HF for mam-
(See Note 19) malian genomic DNA, and NdeI for Xenopus sperm DNA (see
Note 20). The volume of the restriction reaction is normally
set to 250 μl, but can be increased to account for more diluted
DNA preps (the reaction mix is anyway diluted in step 5 in
Subheading 3.7). If necessary, add small amounts of the proper
RNAse enzymes (see Note 21).
2. Equilibrate QIAGEN-tip 20 by applying 1 ml Buffer QBT.
Allow the column to empty by gravity flow. Flow of buffer
will begin automatically by reduction in surface tension due
to the presence of the detergent in the equilibration buffer
(Triton X-100).
3. Wash the QIAGEN-tip 20 column three times with 1 ml of
10 mM Tris–HCl pH 8, 1 M NaCl.
4. Equilibrate the column six times with 1 ml of 10 mM Tris–HCl
pH 8, 300 mM NaCl.
276 Ralph Zellweger and Massimo Lopes

5. After the required incubation time (3–5 h, see step 1), adjust
the digestion mix to 300 mM NaCl final concentration, by
adding 5 M NaCl stock (check the restriction buffer composi-
tion). Adjust the final volume to 600 μl with 10 mM Tris–HCl
pH 8, 300 mM NaCl.
6. Apply the pre-equilibrated digestion mix and allow it to enter
the resin by gravity-flow.
7. Wash the QIAGEN-tip 20 with two times 1 ml 10 mM
Tris–HCl pH 8, 850 mM NaCl. Collect the flow-through in
2 ml tube and save for an analytical gel.
8. Add to the column 600 μl of 10 mM Tris–HCl pH 8, 1 M
NaCl, 1.8% caffeine at 50  C. Collect the flow-through,
enriched in RIs in 1.5 ml tube.
9. Purify and concentrate the DNA (removing residual RNA,
small linear fragments, and microscopic dirty particles) using
an Amicon size-exclusion column.
Load the 600 μl of elution from step 8 into Amicon column.
Spin the column for 8 min at 9000 rcf.
10. Wash the membrane with 200 μl of 1
TE and spin the column
for 5 min at 9000 rcf. Wash again with 200 μl of 1
TE and
spin the column for 4 min at 8000 rcf until 15–30 μl remains.
11. Invert the Amicon filter and short spin into a fresh Amicon
tube.
Note: Transfer the sample to a fresh 1.5 ml eppendorf tube to
prevent evaporation of the sample solution.
12. Load a 1 μl aliquot on an agarose gel to check DNA quality and
concentration. If necessary, adjust the final volume (by adding
1
TE or concentrating the sample in a standard vacuum
evaporator) to reach the optimal DNA concentration of
10–50 ng/μl (see Note 22).

3.8 Preparation of 1. Cleave a 2 cm
2 cm sheet of mica and place it (with the freshly
Carbon-Coated Grids cleaved surface facing up) on the support plate of the MED
020, at a distance of about 12 cm to the carbon evaporator gun.
Place the quartz sensor as close as possible to the mica. Cover
the mica with the tilting shutter, properly position the glass
vacuum chamber wall, and start the turbomolecular pump of
the MED020.
2. At a vacuum of about 3
105 mbar, preheat the filament of
the carbon electron gun. Adjust voltage and current applied to
the electron gun to reach a constant evaporation rate of
0.03–0.05 nm/s (detected on the QSG 100 thin film moni-
tor). Open the tilting shutter and start measuring the carbon
film thickness on the QSG 100 (see Note 23). When the
thickness readout is 3.5–4.5 nm (50–70 Hz), close the shutter
 Dynamic Architecture of Eukaryotic DNA Replication Forks In Vivo. . . 277

and shut off the electron gun. Generally, 2–3 carbon films can
be produced in series (see Note 24).
3. Remove the carbon-coated mica sheet from the MED 020.
Carbon-coated mica sheets can be stored at this stage up to
4–5 weeks.
4. When ready to transfer the carbon film on the grids, place the
carbon-coated mica sheet (carbon side up) in a Petri dish on
wet filter paper and incubate it at 37–42  C for 30 min–1 h.
5. Spread copper grids on filter paper in a glass Petri dish and
make sure that the glossy side of all grids is facing up. 30–35
grids should be used for each 2 cm
2 cm carbon-coated mica
sheet. Place the Petri dish in a hood and, using a Pasteur
pipette, rinse each grid with one drop of Scotch solution
(Fig. 2a). Air-dry the grids. Repeat the procedure two to
three times to make sure that all grids have been extensively
rinsed in Scotch solution. The tape adhesive will keep the
carbon film attached.
6. Fill the supporting wire mesh stand with EM-grade water
(Fig. 2b). A round filter paper (diameter: 45 mm) is submerged
in water and 30–35 grids (glossy side up) are placed on its
surface in a close and ordered distribution (Fig. 2c).
7. The carbon-coated mica sheet is then removed from the wet
filter paper, briefly dried on its lower side (no carbon) by a
napkin, and slowly lowered into the water (carbon side up) at
an angle of approximately 45 (Fig. 2d), until the carbon film
is completely released and floating on the water surface (see
Note 25). Discard the mica support.
8. The carbon film is finally placed on the grids by carefully
lowering the water level in the supporting Teflon-wire mesh
stand, using an aspirator connected to a vacuum pump. Use
tweezers to correctly position the carbon film on the grids,
while lowering the water level (Fig. 2e).
9. Once the water has been removed, take the filter paper with the
carbon-coated grids (Fig. 2f), cut off the excess of wet filter
paper around the grids, and let the carbon-coated grids dry for
at least one night before using them for DNA spreading experi-
ments. Although some variability has been observed, carbon-
coated grids are usually reliable for DNA absorption for about
4–5 weeks after carbon-film production.

3.9 DNA Spreading 1. For each DNA spreading, distribute up to eight droplets
for EM Visualization (10–15 μl) of EtBr working solution on a piece of parafilm.
Carefully place a carbon-coated grid on the top of each drop,
3.9.1 “Native” DNA
with the carbon-side facing the liquid (Fig. 3a; see Note 26).
Spreading by the “BAC
Incubate grids for 20–45 min. Prevent evaporation by covering
Method”
the parafilm with the lid of the 15 cm Petri dish. Just before
278 Ralph Zellweger and Massimo Lopes

spreading the sample on the hypophase (Subheading 3.9.1,

step 6), take each grid, remove the excess of EtBr solution
contacting the filter paper, and place the dried grids (carbon
side down) on the top of filter paper (Fig. 3b).
2. Using a P2 pipette, mix at the very bottom of a 1.5 ml-micro-
fuge tube: 1 μl of Formamide, 0.4 μl of BAC 1:20 (see
Note 27).

Fig. 3 Series of photographs showing crucial steps in the BAC-DNA spreading method (see Subheading 3.9 for
details)
Dynamic Architecture of Eukaryotic DNA Replication Forks In Vivo. . . 279

3. In a separate drop on the side of the same tube (see Note 28),
mix: 1.5–3 ng of plasmid DNA of known size (3–10 kb; inter-
nal size maker) and 10–50 ng of sample DNA (genomic DNA
enriched for Replication Intermediates, see Subheading 3.7,
step 12 and Note 29). The total volume of the DNA drop
should be 1.25 μl; usually it consists of 0.25 μl plasmid DNA
(5 ng/μl) and 1.0 μl sample DNA. Smaller sample volumes can
be filled up with 1
TE buffer (see Note 29). In case of very
low DNA concentration in the sample, up to 4 μl of the DNA
sample can be added to the spreading. In this case an equal
volume of Formamide should be added to the mix, while
volumes of plasmid DNA and BAC remain unchanged
4. Pour approximately 20 ml (the minimum volume to cover the
surface completely) of EM-grade water in the 15 cm Petri dish:
this is called the hypophase. Cleave a mica sheet (about
1 cm
2 cm) and place it in the water as a ramp, with the
freshly cleaved surface facing up (Fig. 3c).
5. Spin the tube containing the sample for a few seconds in a
microfuge and aspirate it completely in a 10 μl tip. With a
cotton swab sprinkle a few grains of graphite powder (no
graphite flakes!) onto the water surface, in close proximity to
the mica ramp (Fig. 3c).
6. Pipette the entire sample volume out of the tip and let the
droplet touch the ramp few millimeters above the water sur-
face. The BAC-containing drop will immediately slide down
the ramp and spread over the water hypophase. The graphite
powder will mark the border of the monomolecular detergent
film containing the DNA molecules (Fig. 3d; see Note 30).
7. Using fine tweezers take one carbon-coated, EtBr-treated grid
from the filter paper (carbon-side down, Fig. 3e) and pick up
part of the DNA film touching the spreading surface in prox-
imity to the graphite powder (see Note 31). Hold the tweezers
to ensure full parallel contact of grid and surface (see Note 10).
Enough pressure should be applied to ensure full contact
between the carbon and the DNA-containing film (Fig. 3f).
The grid is then removed from the surface and incubated for
15 s for staining in 1 ml of uranyl acetate working solution (in a
flat-bottom 20 ml tube; Fig. 3g). After a brief wash (1–2 s) in
100% Ethanol, air-dry the grid (carbon-side up) on filter paper
(Fig. 3h) and carefully wipe the tweezers with a napkin (see
Note 32).
8. Step 7 is repeated for the other grids, ensuring to collect DNA
from different regions of the BAC film containing the DNA
molecules (see Note 33).
280 Ralph Zellweger and Massimo Lopes

3.9.2 Denaturing DNA To obtain information about in vivo-nucleosome positioning on

Spreading by the “BAC the RIs (see Subheading 1), DNA samples can be denatured just
Method” before the BAC spreading. In this case, the presence of the dena-
turing agents (Formamide and Glyoxal) in the spreading mix is
coupled with a short incubation at 42  C, leading to DNA strand
separation at each of the not-crosslinked regions (nucleosomal
DNA; Figs. 1 and 8; see Note 34).
The spreading procedure is identical to the one described above
(Subheading 3.9.1), with the exception of steps 2 and 3, per-
formed as follows:
2. Using a P2 pipette, mix in a 1.5 ml-microfuge tube: 1.0 μl of
Formamide, 0.2 μl of Glyoxal, and 1.0 μl of DNA sample
(10–50 ng) from Subheading 3.7, step 12. Incubate for
10 min at 42  C in a water bath and chill immediately after in
ice-water.
3. Spin briefly the sample and add: 0.25 μl of plasmid DNA 5 ng/
μl (internal size marker, see Note 35) and 0.4 μl of BAC 1:10.
Immediately proceed with steps 4–8 of the BAC spreading.

3.10 Platinum- 1. Place the grids on the specimen table, taking care that they are
Carbon Rotary properly fixed (by clips or magnetic stripes) and flat (Fig. 4a).
Shadowing Position the specimen table on the rocking rotary stage and the
quartz sensor as close as possible to the specimen table (Fig. 4b).
2. Tilt the rocking rotary stage using the Precision Rotation
Platform PR01, so that the angle between the specimen table
and the Pt/C gun is exactly and reproducibly 3 (Fig. 4c). The
fine micrometric scale on the Rotation Platform enables the
reproducibility of the angle (Fig. 4c and d, black arrow). Cover
the rotary stage with the tilting shutter, properly position the
glass vacuum chamber wall, and start the turbomolecular pump
of the MED 020.
3. At a vacuum of 3
105 mbar (or higher vacuum), preheat the
filament of the Pt/C electron gun. Adjust voltage and current
applied to the electron gun to reach a constant platinum/
carbon evaporation rate of 0.03–0.05 nm/s (see Note 36),
detected on the QSG 100 thin film monitor (see Note 23).
Open the tilting shutter and start the measure of platinum film
thickness on the QSG 100. When the detected thickness is
0.4 nm, start the rotation of the specimen table at the mini-
mum speed (about 20 rpm). During the evaporation time
(4–5 min) keep adjusting voltage and current applied to the
electron gun to maintain the platinum/carbon evaporation
rate constant. When the thickness readout at the QSG 100 is
8–10 nm (1800–2000 Hz), close the shutter, stop rotation,
and shut off the electron gun. After the gun is cooled down,
start the machine again and evaporate another 4–6 nm plati-
num on your grids while rotating (see Note 37).
 Dynamic Architecture of Eukaryotic DNA Replication Forks In Vivo. . . 281

Fig. 4 Series of photographs showing crucial steps in the platinum/carbon

rotatory shadowing procedure (see Subheading 3.10 for details)

4. Remove the grids from the specimen table and store them
properly for transportation. The grids can immediately be ana-
lyzed at the Transmission Electron Microscope or can be (re-)
analyzed after unlimited storage periods.

3.11 Visualization at 1. The grids can be analyzed at any Transmission Electron Micro-
the Transmission scope. Duplex DNA in this technique is expected to appear as a
Electron Microscope, 10 nm thick fiber, while ssDNA thickness should be 5–7 nm
Contour Length [20]. Both molecules should be clearly detectable on the
Measures, and homogenously granular background given by the platinum
Statistics grains deposited on the carbon film (see Note 12).
282 Ralph Zellweger and Massimo Lopes

2. The plasmid DNA molecules added to the spreading mix can

be used as an internal control of DNA absorption by the carbon
film and as an internal size marker. Moreover, these circular
molecules are often partially denatured (even in “native spread-
ings”) by the short incubation with Formamide and Glyoxal:
their partial denaturation offers an easy opportunity to verify
whether ssDNA and dsDNA in the same molecule are easily
distinguishable by their thickness. At this step, if the concen-
tration of the DNA, the quality of the shadowing, and/or the
generally clean appearance of the grid background are not fully
satisfactory, the experiment should be repeated, starting from
the DNA spreading (or from earlier steps in the sample prepa-
ration, in case of recurrent problems, probably resulting from
the DNA sample itself).
3. The extensive analysis of a satisfactory sample is usually per-
formed at 5000–20,000
. Despite the enrichment procedure
described in Subheading 3.7, linear duplex DNA represents at
least 90% of the genomic DNA visualized on the grids (it is
retained on the QIAGEN-tip 20 probably because of local
“breathing” of the DNA duplex and exposure of ssDNA).
The identification of a replication fork requires the recognition
of a “3 leg-junction,” i.e., a contact point from which
three different DNA fibers depart (Fig. 5). These junctions
need to be carefully analyzed at higher magnification
(50,000–250,000
) to distinguish them from occasional over-
laps of two linear molecules in proximity to one of the ends
(“4-leg junctions”). At higher magnifications, some ssDNA is
usually detectable at one or two sides of the junction, reinfor-
cing the interpretation of the junction as a bona fide replication
fork and supporting the identification of the daughter strands
(Fig. 5). Furthermore, at least two of the three legs should be
equal in length (as the genomic DNA has been digested, the
elongation point should be equally distant from the two newly
replicated restriction sites). The estimation of DNA molecule
length is often complicated by the convoluted distribution of
the fibers; this often requires the detailed analysis to be per-
formed only later, once a digital file has been generated. Tech-
nical problems, such as partial restriction digestion or breakage
of DNA molecules during sample preparation, account for the
fraction of replication forks with all three legs different in
length, i.e., asymmetric replicated duplexes. In standard con-
ditions these should not represent more than 25–30% of the
total replicating molecules, but the proportion of asymmetric
forks can increase in conditions of DNA damage, associated
with ssDNA accumulation.
4. Although photo-documentation can be obtained on any TEM
generating photographic negatives (followed by developing
 Dynamic Architecture of Eukaryotic DNA Replication Forks In Vivo. . . 283

Fig. 5 Normal replication fork visualized by in vivo psoralen crosslinking and

TEM. Replicated duplexes (b, c) are of equal length. A ssDNA stretch is clearly
visible in proximity to the fork on one of the two replicated duplexes (presumably
the lagging strand)

and printing of the corresponding micrographs), the use of a

microscope connected to a CCD camera is highly recom-
mended: besides the reduction of the time required for
photo-documentation, this procedure allows us to obtain digi-
tal files that can be promptly used for contour length measures
through Imaging applications (see Subheading 3.11, step 6).
5. As a general rule, 70–100 replicating molecules should be
collected in each experiment to obtain reliable data from the
following statistical analysis. Observed trends need to be repro-
duced in two-three biological replicates. Should the observed
differences in the frequency of specific intermediates be small
and/or not fully reproducible, statistical assessment on data
from three independent experiments is essential.
6. Once the digital files are converted in *.tiff files, contour
length measurements can be performed by standard image
analysis applications. ImageJ has been successfully used, but
other, more specific applications are being tested and could be
more appropriate (and time-saving) for filament recognition
and analysis. Once the measurements are performed, data can
be analyzed by using standard statistical/graphical applications
such as Microsoft Excel or Graphpad Prism.

3.12 Interpretation of Our most recent studies have identified reversed replication forks
Four-Way Junctions (Figs. 6a, b and 7) as frequent replication intermediates under
certain conditions, i.e., upon genetic or pharmacological replica-
tion interference [11–17, 30]. Importantly, the regressed arm can
either be connected to both or only one of the daughter duplexes,
284 Ralph Zellweger and Massimo Lopes

Fig. 6 Compilation of three molecules to assist the identification of reversed forks. (a) and (b) are representa-
tive examples of reversed forks, while (c) shows the accidental crossing of two independent DNA molecules
(see Subheading 3.12 for details)

Fig. 7 Drawing on reversed fork vs Holliday junction and their expected features
in terms of contour length measurements
Dynamic Architecture of Eukaryotic DNA Replication Forks In Vivo. . . 285

and it can be entirely or partially single-stranded. The unambiguous

identification of these replication intermediates is based on contour
length measurements in combination with several additional cri-
teria. As reversed replication forks are four-way structures, acciden-
tal crossings of linear DNA molecules can easily be mistaken for
reversed forks. The incidence of random crossings increases with
the DNA concentration on the grids; therefore, it is necessary to
work with samples with low DNA density to minimize the risk of
misinterpretations.
The most important criterion for the assignment of a four-way
structure as a reversed fork is the junction itself. If the junction is
opened up and allows the clear identification of a romboid struc-
ture, the assignment as a reversed replication fork is usually straight-
forward (Fig. 6a). However, most reversed replication forks display
a collapsed junction with little or no opening (Fig. 6b) and there-
fore require additional parameters for their interpretation. These
parameters are the appearance of the junction, and the orientation
and the length of the arms.
For a careful interpretation of a four-way junction, taking a
high magnification image (135,000–250,000
) of the junction is
recommended. All the arms of a reversed fork are connected to each
other; therefore, they give the junction a “flat” appearance in one
focal plane (Fig. 6b, inset). In contrast, in accidental crossings of
linear DNA, one molecule is on the top of the other, resulting in a
difference in focal plane that can help to discriminate DNA cross-
ings from true DNA junctions already while taking pictures
(Fig. 6c). Moreover, crossings are usually associated with a contin-
uous shadowing of the molecule on the top and a discontinuous
platinum deposition on the bottom molecule (Fig. 6c).
As in an unperturbed replication fork, the length of the two
new replicated duplexes of a reversed fork should be identical
(b ¼ c; Fig. 7). In contrast, as typical for all replication intermedi-
ates, the length of the parental arm (a) is not defined, as is the
length of the regressed (d) arm (a 6¼ b ¼ c 6¼ d; Fig. 7). Conversely,
Holliday junctions connect symmetrically two homologous frag-
ments and will have by definition arms of equal length two by two
(a ¼ b, c ¼ d). Only in those rare cases in which a restriction site on
the parental DNA is located in close proximity to the elongation
point (late replication intermediate), the same restriction site will
also be present on the regressed arm, giving rise to (a ¼ d; b ¼ c)
molecules. The appearance of such molecules would be identical to
a late termination intermediate or to a Holliday Junction; there-
fore, they cannot be unambiguously assigned and should not be
considered in the analysis.
The final criterion for the identification of reversed forks is the
orientation of the arms of the molecule. Reversed forks and other
DNA junctions usually display different orientations of at least
three of the four arms of the molecule, whereas the “arms” in
286 Ralph Zellweger and Massimo Lopes

accidental crossings are often aligned along only two axes, some-
times even in a perpendicular fashion (compare Fig. 6b and c).
As for all other replication intermediates (see Subheading 3.11,
step 3), partial digestion and DNA breakage can lead to reversed
forks displaying no obvious symmetry in the length of the four
arms. Particular attention should be paid to the analysis of the
junction at high magnification (see above) to assign these molecules
as reversed forks. As a general rule, the fraction of asymmetric
reversed forks should never exceed the frequency of asymmetry
observed in the same samples for the population of “normal”
(three-way) replication forks.

3.13 Analysis Denaturing spreadings (Subheading 3.9.2) allow the analysis of

of Replicating nucleosome density on DNA. For this kind of analysis, a sufficient
Chromatin Under degree of psoralen crosslinking is required. 80–90% of the linkers
Denaturing Conditions should contain at least one crosslink, so that 80–90% of detected
ssDNA bubbles (Figs. 1 and 8) represent mononucleosomes of
150–200 bp [2, 22–24]. Nucleosome density is usually expressed
by the so-called r-value, which is calculated as the combined con-
tour length of all nucleosome bubbles in a given stretch of DNA
(Fig. 8a, white), divided by the overall contour length of the same
DNA stretch (Fig. 8b, black). A reduced r-value indicates a reduc-
tion in nucleosome density.

Fig. 8 Example of denaturing molecule (b) and system to measure it (a). The portion of the molecules
organized in nucleosomes is calculated by adding all DNA stretches in the visible ssDNA bubbles (white tracts),
thus excluding gaps of crosslinked DNA (white arrows). This combined length is then divided by the total
length of the DNA tract analyzed (black line), giving rise to the “r-value” (see Subheading 3.13 for details)
 Dynamic Architecture of Eukaryotic DNA Replication Forks In Vivo. . . 287

4 Notes

1. A high efficiency of psoralen crosslinking can be obtained with

shorter irradiation times, using monochromatic 365 nm Hg
lamps [25]. However, with proper tuning of the irradiation
time, this protocol has the advantage of using standard equip-
ment (Biolinker) frequently available in molecular biology
laboratories.
2. High quality genomic DNA for EM analysis has also been
obtained reproducibly using QIAGEN Genomic-tip 20/G, fol-
lowing the manufacturer’s instruction (vortexing steps are sub-
stituted by more gentle inversion of the tubes, to avoid shearing
of genomic DNA). Nevertheless, the procedure described in
detail is the one used for most of the EM data recently published
on yeast replicating genomic DNA [5, 8, 31].
3. Help CTAB dissolution by heating the solutions to about
50  C (do not overheat or boil!).
4. A number of different evaporators are available, with relevant
differences in terms of vacuum chamber design and orientation
of the electron guns in respect to the specimen. All evaporators
can be used to produce carbon films and most of them can be
adapted to perform extremely flat angle Pt/C rotary shadow-
ing. The procedure described here allows us to perform Car-
bon evaporation (Subheading 3.8) and Pt/C rotary shadowing
(Subheading 3.10) with the same machine, currently available
on the market, with minor technical modifications to the appa-
ratus (see Note 12). In our experience, the correct maintenance
of the apparatus and the constant use/order of identical con-
sumable parts (Wolfram cathodes, carbon rods, platinum
insets) is of crucial importance to avoid unexpected problems
during evaporation and to obtain reproducibly good carbon
layers and rotary shadowing.
5. Supporting wire mesh stands, similar to the one shown in
Fig. 2b, can be produced in any workshop. It is composed of
two Teflon rings, the lower of which has holes that allow
controlled water flow out of the central cavity. A wire mesh
stand is placed in-between the two Teflon rings and can accom-
modate the round filter paper where the grids are placed.
Teflon can be substituted by different materials (Plexiglas,
plastic, etc.). However, Teflon offers the best resistance to
acid washes of the apparatus, occasionally required to remove
traces of the carbon films.
6. Excellent results have been obtained with re-distilled water.
More recently, standard MilliQ water (produced by a conven-
tional Millipore apparatus), with resistivity of about 18 MΩ cm
and total organic content of less than eight parts per billion, has
provided satisfactory results. This water is stored in a proper
288 Ralph Zellweger and Massimo Lopes

flask solely dedicated to the EM work, which is directly filled at

the MilliQ water apparatus.
7. Nickel grids can be used as an alternative to copper grids to
make use of magnetic holders for the evaporation machine.
This enables the use of for example the Leica BAF060 which
speeds up the shadowing procedure since the machine has a
permanent vacuum. The handling and procedures are identical
but it is recommended to demagnetize the tweezers to prevent
the nickel grids from sticking to them.
8. The original powder stock of BAC was kindly provided by
Bayer, Leverkusen, Germany. The chemical is now available
also from different suppliers (i.e., Sigma, cat. B6295), but has
never been tested for this specific application.
9. Tissue culture dishes are preferred over standard Petri dishes, as
their surface treatment allows covering the surface with smaller
volumes. A reduced hypophase volume improves the stability of
the detergent-DNA film during the spreading (see also Note 32).
10. Bent tweezers (Fig. 3e and f): to hold the grids as flat as
possible while touching the spreading surface, it is of great
help to slightly bend the tweezers tips; this can be done
mechanically applying pressure on the tips with a proper tool.
Particular care is necessary to avoid the breakage of the tips.
Recently, we also obtained satisfactory results using Dumont
pinzettes (see Subheading 2.8).
11. Better results have been obtained with grid tables that fix
individual grids by small pins (Fig. 4a), compared with those
currently available on the market, where multiple grids are
fixed on the table by metal stripes. Besides the ease in fixing/
removing individual grids, the former tables show reduced
interference with low angel rotary shadowing, by maximizing
the area of each grid effectively exposed to the correct amount
of evaporated Pt. Leica is currently considering to restart pro-
duction of these grid tables. Magnetic grid holders can also be
efficiently used, if Nichel grids are used (see Note 7).
12. The accurate determination of the very flat angle between the
electron gun and specimen table is crucial to observe an opti-
mal difference in thickness between ssDNA and dsDNA. To
facilitate the determination and the reproducibility of the opti-
mal angle in independent experiments, the knob available on
the MED 020 to tilt the rocking rotary stage can be substituted
with a precision rotation platform, providing micrometric con-
trol of this angle (Fig. 4a and d, black arrow).
13. A standard EM experiment, consisting of 100 replicating mole-
cules (by a standard 1k
1k resolution CCD camera), corre-
sponds to several Gigabytes of digital files!
Dynamic Architecture of Eukaryotic DNA Replication Forks In Vivo. . . 289

14. For experiments in S. cerevisiae, it is recommended to extract

genomic DNA from cells synchronized in S-phase, i.e.,
30–180 min after release from an α-factor arrest, depending
on the treatment [5, 9]. While synchronization experiments
have been performed also for EM analysis in mammalian
cells—3-6 h after release from 16 h HU-block [32]—this
does not seem to be strictly required. Although certainly
more time consuming, it is possible to obtain 70–100 RIs
from asynchronous, untreated U2OS cells or MEFs [15].
15. It is important to remove traces of growth media, which would
otherwise absorb part of the monochromatic light in Subhead-
ings 3.1, step 4 and 3.2, step 4.
16. Due to the increasing Ethanol concentration in the suspension,
cells tend to aggregate and deposit to the bottom of the dish.
This does not interfere with the following extraction of geno-
mic DNA, but requires extensive resuspension of the sample
prior to each irradiation.
17. Timing of irradiation and number of crosslinking cycles strictly
depend on the irradiance of the 365 nm bulbs. The observed
irradiance within the crosslinker, measured by UV meter, is
6.2 mW/cm2. This needs to be monitored, as it changes
considerably with the type of apparatus, the age of the bulbs,
and the distance of the samples from the bulbs. The total
irradiation time should be adjusted to apply the same total
irradiation (J/cm2) to the samples as indicated here. Only
doing so, it is possible to obtain optimal frequency of crosslinks
on genomic DNA (1 crosslink in 80–90% of DNA linkers),
particularly important for chromatin studies (see Subheading
3.13).
18. Incorporation of Cy3-dCTP can be monitored by visualization
of nuclei using a standard fluorescence microscope. We con-
sider an appropriate time for nuclear replication when 80–90%
of the nuclei appear positive for Cy3 incorporation.
19. We have tested the efficiency of this RI enrichment protocol
and compared it to standard BND cellulose-mediated enrich-
ment [21], by performing a new round of digestion, RI enrich-
ment, and EM analysis of several genomic DNA samples that
had already been analyzed by EM [17], and for which sufficient
amount of residual genomic DNA was still available in our
stocks. For all the tested samples, we obtained practically iden-
tical results—in terms of ssDNA accumulation at forks and
reversed fork frequency—to those previously obtained proces-
sing the same samples with the standard BND-cellulose enrich-
ment (Fig. 6a in [17]). However, we reproducibly noticed
that—despite similar density of total DNA on our EM
grids—the identification of 70–100 RIs was significantly
290 Ralph Zellweger and Massimo Lopes

(two- to threefold) faster than what typically observed with

BND-enriched DNA. These observations strongly suggest that
this optimized enrichment protocol allows “discarding” a
higher fraction of linear, non-replicating DNA during the
first elution step, yielding a higher enrichment of RI in the
final elution, without introducing any bias on the type of
replication structures and on ssDNA accumulation finally
detected in the sample.
20. While choosing the restriction enzyme to digest genomic
DNA, it is important to consider that: (a) frequent cutters
will lead to smaller RIs (possibly complicating the analysis of
phenomena happening at longer distance from the replication
forks); (b) due to frequent methylation of CpG sequences,
mammalian genomic DNA may be resistant to the action of a
number of common restriction enzymes. It is also recom-
mended to use enzymes releasing blunt DNA ends, to prevent
ssDNA overhangs competition for binding to the BND
cellulose.
21. It is important that residual RNA is removed prior to loading
the sample on the QIAGEN-tip 20, to avoid interference with
the enrichment procedure. This is particularly important for
the extraction procedure described in Subheading 3.4, which
normally leads to the extraction of large amounts of dsRNA,
besides the expected genomic DNA. When residual RNA is
detected on the gel, add to the restriction mix a few units of
RNaseA (for ssRNA; 5 μl of 10 mg/ml in 250 μl reaction) and
RNaseIII (for dsRNA; 1 μl of 1:100 dilution in TE in a 250 μl
reaction; do not digest longer than 3 h to prevent DNA
degradation). They are both usually active in the common
restriction buffers.
22. The DNA recovered in Subheading 3.7, step 12 is normally
1–15% of the total DNA loaded on the QIAGEN-tip 20 col-
umn, depending on the fraction of cells actively replicating
their DNA. If the amount recovered at this step is far from
this proportion, proper aliquots of the flow-through from
Subheadings 3.7, step 6 and 3.7, step 7 can be loaded on a
gel to figure out where the DNA got lost, recover it and repeat
the procedure with appropriate modifications. The accurate
quantification of the DNA recovered in Subheading 3.7, step
12 requires an agarose gel, as the caffeine “contamination” in
the RIs-enriched fraction interferes with standard spectropho-
tometric or fluorimetric measurements of DNA concentration.
23. The readout is calculated correcting for the different position
of the quartz sensor in respect to the specimen (tooling factor).
With the quartz crystal placed next to the sample (see Fig. 4b)
the tooling factor in the QSG 100 should be set to 1. Different
Dynamic Architecture of Eukaryotic DNA Replication Forks In Vivo. . . 291

settings will require the empirical determination of the correct

tooling factor.
24. After two-to-three evaporation sessions, the electron gun over-
heats. Because of the high temperature, carbon films tend to
strongly adhere to the mica sheet and are resistant to the
floating procedure in Subheading 3.8, step 7.
25. Do not proceed lowering the mica in the water unless the
carbon film is visibly detaching from the mica sheet. The
proper coordination of mica sheet submersion and carbon
film detachment is crucial for carbon film integrity and requires
some manual skills and experience.
26. Generally, for each DNA spreading, grids covered with two-to-
three different carbon films are used. As each carbon film has
different absorption properties, this increases the chances of
obtaining at least some grids with an optimal concentration of
absorbed DNA molecules (see Note 31).
27. Formamide acts as a “partially-denaturing” agent, helping to
disentangle and unfold DNA molecules during the spreading
procedures. These conditions are optimal for psoralen-
crosslinked DNA, which is inherently resistant to denaturation;
if uncrosslinked DNA needs to be used, formamide concentra-
tion can be reduced to prevent DNA denaturation. If the DNA
is only briefly exposed to formamide during the procedure (see
Note 30), the vast majority of the psoralen-crosslinked DNA
molecules is still visible as dsDNA, although some “breathing”
is occasionally detected along the duplex and especially at DNA
ends. Full denaturation of the sample requires longer exposure
to these agents at higher temperatures (see Subheading 3.9.2).
28. The mix of the two droplets (by spinning briefly, see Subhead-
ing 3.9.1, step 5) should only be performed immediately
before the spreading and having all necessary material ready,
to minimize DNA exposure to formamide (see Note 29).
29. The final concentration of the DNA on the grid is crucial for a
proper analysis of RIs. This depends on the DNA concentra-
tion in the sample, but also on the carbon absorption proper-
ties and the size of the spreading surface. Therefore, the
spreading procedure often needs to be repeated to obtain
grids with an optimal distribution of DNA molecules. For
DNA samples that are particularly diluted, the DNA volume
can be increased up to 5 μl, provided that formamide/glyoxal
volumes are changed accordingly. BAC/plasmid volumes are
unchanged, so that the size of the spreading surface and the
density of marker molecules are not changed, leading to a
higher frequency of DNA molecules of interest on the EM
grids.
292 Ralph Zellweger and Massimo Lopes

30. The stability of the monolayer can be heavily affected by any

local perturbation of the conditions (hitting the bench, collea-
gues walking by, etc.). It is therefore highly recommended to
perform the BAC spreading procedure in a quiet laboratory, on
an (at least transiently) isolated bench. It is helpful to protect
the working area with a Plexiglas box, open only to the side
facing the operator.
31. The DNA molecules accumulate in close proximity to the edge
of the spreading surface (close to the graphite powder).
32. Ethanol traces on the tweezers heavily disturb the spreading
surface!
33. It is unavoidable that the monolayer surface will shrink and be
perturbed while repeatedly touching it with the grids. The
concentration of the DNA can also vary accordingly. Nonethe-
less, if the spreading surface is clearly unstable from the begin-
ning, it is recommended to repeat the spreading procedure (see
also Notes 9 and 32, on how to improve the spreading surface
stability).
34. Denatured DNA is more difficult to absorb to the carbon films.
It may happen that carbon-coated grids that proved proficient
in absorbing dsDNA do not perform equally well absorbing
denatured (mostly single-stranded) DNA (see Note 37). Due
to the lower thickness, denatured DNA is also more difficult to
visualize at the EM and usually requires optimal contrast both
from uranyl acetate staining and from Pt/C rotary shadowing.
35. It is important that the plasmid marker is added only after
denaturation. Preserving its duplex status, the marker will
serve as an absorption control for the carbon films and will
help assessing the quality of the contrast obtained on the grids
(see Subheading 3.11, step 2).
36. Standard settings for platinum shadowing of a MED020 evap-
oration machine are 1.70 kV, 0.40 mA and tooling factor 1.0
and for carbon evaporation 1.70 kv, 0.70 mA and tooling
factor 0.5. These values can differ from machine to machine
but can be used as a starting point to determine your machine’s
settings.
37. A second round of platinum shadowing provides increase con-
trast to DNA molecules on carbon-coated grids. It is essential
to wait that the Pt/C gun cools down, before the second
round of platinum coating is performed.

Acknowledgments

We are grateful to José M. Sogo for his crucial support while learning
all technicalities of this approach. We wish to thank the whole team at
 Dynamic Architecture of Eukaryotic DNA Replication Forks In Vivo. . . 293

the ZMB (Center for Microscopy and Image Analysis of the Univer-
sity Zurich) for consistently excellent technical assistance, while run-
ning our EM experiments. We are grateful to Arnab Ray Chaudhuri,
Yoshitami Hashimoto, Fabio Puddu, and Vincenzo Costanzo for
their assistance in optimizing this EM approach on Xenopus egg
extracts. We also wish to thank Petr Cejka for suggesting the use of
QIAGEN columns as possible valuable alternative to BND cellulose
for the enrichment of a subpopulation of DNA molecules based on
differential ssDNA content. We are also grateful to Sebastian Ursich
for his recent efforts optimizing this approach and for careful reading
of the manuscript. Work in the Lopes lab is currently financed by the
SNF grant 31003A_169959, the ERC Consolidator grant 617102
(ReStreCa) and the Swiss Cancer League grant KFS-3967-08-2016.

References
1. Inciarte MR, Salas M, Sogo JM (1980) Struc- 9. Sogo J, Lopes M, Foiani M (2002) Fork rever-
ture of replicating DNA molecules of Bacillus sal and ssDNA accumulation at stalled replica-
subtilis bacteriophage phi 29. J Virol tion forks owing to checkpoint defects. Science
34:187–199 (New York, NY) 297:599–602
2. Lucchini R, Sogo JM (1995) Replication of 10. Berti M, Chaudhuri AR, Thangavel S et al
transcriptionally active chromatin. Nature (2013) Human RECQ1 promotes restart of
374:276–280 replication forks reversed by DNA topoisomer-
3. Sogo JM, Stahl H, Koller T, Knippers R (1986) ase I inhibition. Nat Struct Mol Biol
Structure of replicating simian virus 40 mini- 20:347–354. doi:10.1038/nsmb.2501
chromosomes. The replication fork, core his- 11. Follonier C, Oehler J, Herrador R, Lopes M
tone segregation and terminal structures. J Mol (2013) Friedreich’s ataxia-associated GAA
Biol 189:189–204 repeats induce replication-fork reversal and
4. Avemann K, Knippers R, Koller T, Sogo JM unusual molecular junctions. Nat Struct Mol
(1988) Camptothecin, a specific inhibitor of Biol. doi:10.1038/nsmb.2520
type I DNA topoisomerase, induces DNA 12. Neelsen KJ, Zanini IMY, Mijic S et al (2013)
breakage at replication forks. Mol Cell Biol Deregulated origin licensing leads to chromo-
8:3026–3034 somal breaks by rereplication of a gapped DNA
5. Lopes M, Foiani M, Sogo JM (2006) Multiple template. Genes Dev 27:2537–2542. doi:10.
mechanisms control chromosome integrity 1101/gad.226373.113
after replication fork uncoupling and restart at 13. Neelsen KJ, Zanini IMY, Herrador R, Lopes M
irreparable UV lesions. Mol Cell 21:15–27 (2013) Oncogenes induce genotoxic stress by
6. Engels K, Giannattasio M, Muzi-Falconi M mitotic processing of unusual replication inter-
et al (2011) 14-3-3 proteins regulate exonucle- mediates. J Cell Biol 200:699–708. doi:10.
ase 1-dependent processing of stalled replica- 1128/MCB.24.16.7140-7150.2004
tion forks. PLoS Genet 7:e1001367. doi:10. 14. Neelsen KJ, Lopes M (2015) Replication fork
1371/journal.pgen.1001367 reversal in eukaryotes: from dead end to
7. Giannattasio M, Follonier C, Tourriere H et al dynamic response. Nat Rev Mol Cell Biol
(2010) Exo1 competes with repair synthesis, 16:207–220. doi:10.1038/nrm3935
converts NER intermediates to long ssDNA 15. Ray Chaudhuri A, Hashimoto Y, Herrador R
gaps, and promotes checkpoint activation. et al (2012) Topoisomerase I poisoning results
Mol Cell 40:50–62. doi:10.1016/j.molcel. in PARP-mediated replication fork reversal.
2010.09.004 Nat Struct Mol Biol 19:417–423. doi:10.
8. Hashimoto Y, Chaudhuri AR, Lopes M, Cost- 1038/nsmb.2258
anzo V (2010) Rad51 protects nascent DNA 16. Ray Chaudhuri A, Ahuja AK, Herrador R,
from Mre11-dependent degradation and pro- Lopes M (2015) Poly(ADP-ribosyl) glycohy-
motes continuous DNA synthesis. Nat Struct drolase prevents the accumulation of unusual
Mol Biol 17:1305–1311. doi:10.1038/nsmb. replication structures during unperturbed S
1927
294 Ralph Zellweger and Massimo Lopes

phase. Mol Cell Biol 35:856–865. doi:10. 25. Wellinger RE, Lucchini R, Dammann R, Sogo
1128/MCB.01077-14 JM (1999) In vivo mapping of nucleosomes
17. Zellweger R, Dalcher D, Mutreja K et al using psoralen-DNA crosslinking and primer
(2015) Rad51-mediated replication fork rever- extension. Methods Mol Biol 119:161–173
sal is a global response to genotoxic treatments 26. Vollenweider HJ, Sogo JM, Koller T (1975) A
in human cells. J Cell Biol 208:563–579. routine method for protein-free spreading of
doi:10.1083/jcb.201406099 double- and single-stranded nucleic acid mole-
18. Thangavel S, Berti M, Levikova M et al (2015) cules. Proc Natl Acad Sci U S A 72:83–87
DNA2 drives processing and restart of reversed 27. Sogo JM, Thoma F (1989) Electron micros-
replication forks in human cells. J Cell Biol copy of chromatin. Methods Enzymol
208:545–562. doi:10.1083/jcb.201406100 170:142–165
19. Fugger K, Mistrik M, Neelsen KJ et al (2015) 28. Sambrook J, Fritsch EF, Maniatis T (1989)
FBH1 catalyzes regression of stalled replication Molecular cloning, a laboratory manual, 2nd
forks. Cell Rep. doi:10.1016/j.celrep.2015. edn. Cold Spring Harbor Laboratory Press.,
02.028 Cold Spring Harbor, NY
20. Lopes M (2009) Electron microscopy methods 29. Trenz K, Errico A, Costanzo V (2008) Plx1 is
for studying in vivo DNA replication inter- required for chromosomal DNA replication
mediates. Methods Mol Biol 521:605–631 under stressful conditions. EMBO J
21. Neelsen KJ, Chaudhuri AR, Follonier C et al 27:876–885. doi:10.1038/emboj.2008.29
(2014) Visualization and interpretation of 30. Ahuja AK, Jodkowska K, Teloni F et al (2016)
eukaryotic DNA replication intermediates A short G1 phase imposes constitutive replica-
in vivo by electron microscopy. Methods Mol tion stress and fork remodelling in mouse
Biol 1094:177–208. doi:10.1007/978-1- embryonic stem cells. Nat Commun 7:10660.
62703-706-8_15 doi:10.1038/ncomms10660
22. Gasser R, Koller T, Sogo JM (1996) The sta- 31. Sogo JM (2002) Fork reversal and ssDNA
bility of nucleosomes at the replication fork. J accumulation at stalled replication forks owing
Mol Biol 258:224–239 to checkpoint defects. Science 297:599–602.
23. Gruss C, Wu J, Koller T, Sogo JM (1993) doi:10.1126/science.1074023
Disruption of the nucleosomes at the replica- 32. Mojas N, Lopes M, Jiricny J (2007) Mismatch
tion fork. EMBO J 12:4533–4545 repair-dependent processing of methylation
24. Mejlvang J, Feng Y, Alabert C et al (2014) New damage gives rise to persistent single-stranded
histone supply regulates replication fork speed gaps in newly replicated DNA. Genes Dev
and PCNA unloading. J Cell Biol 204:29–43. 21:3342–3355
doi:10.1083/jcb.201305017
 Chapter 20

A Molecular Toolbox to Engineer Site-Specific DNA

Replication Perturbation
Nicolai B. Larsen, Ian D. Hickson, and Hocine W. Mankouri

Abstract
Site-specific arrest of DNA replication is a useful tool for analyzing cellular responses to DNA replication
perturbation. The E. coli Tus-Ter replication barrier can be reconstituted in eukaryotic cells as a system to
engineer an unscheduled collision between a replication fork and an “alien” impediment to DNA replica-
tion. To further develop this system as a versatile tool, we describe a set of reagents and a detailed protocol
that can be used to engineer Tus-Ter barriers into any locus in the budding yeast genome. Because the Tus-
Ter complex is a bipartite system with intrinsic DNA replication-blocking activity, the reagents and
protocols developed and validated in yeast could also be optimized to engineer site-specific replication
fork barriers into other eukaryotic cell types.

Key words Replication fork barrier, DNA replication stress, Tus-Ter

1 Introduction

The E. coli Tus-Ter barrier consists of the high-affinity binding of

the DNA replication-blocking protein, Tus, to a specific 23-bp
“Ter” sequence. The Tus-Ter barrier is polar in E coli, and DNA
replication forks are arrested only when encountering the “restric-
tive” face of the Tus-Ter complex [1]. We demonstrated previously
that short arrays of either 3
Tus-Ter or 7
Tus-Ter barriers can
function as a polar replication pausing system when engineered into
the budding yeast genome [2, 3]. Our recent data indicate that the
strength of replication fork pausing within a defined restriction
fragment is saturated at 14
Tus-Ter barriers. However, because
yeast can apparently overcome replication fork pausing at Tus-Ter
barriers [3], further increasing the number of Ter sites, or engineer-
ing additional tandem >14
Tus-Ter barriers in cis, may be used to
additively delay DNA replication at specific chromosomal loci.
Here, we describe a detailed method for how to create and analyze
budding yeast strains with inducible Tus-Ter barriers. Although
these protocols were developed for engineering Tus-Ter barriers

Marco Muzi-Falconi and Grant W. Brown (eds.), Genome Instability: Methods and Protocols, Methods in Molecular Biology,
vol. 1672, DOI 10.1007/978-1-4939-7306-4_20, © Springer Science+Business Media LLC 2018

295
296 Nicolai B. Larsen et al.

into the yeast genome, the core reagents could easily be adapted for
use in other cell types by utilization of other host-cell specific
selectable markers. Indeed, the Tus-Ter system has also been estab-
lished and independently validated in both mouse and human cell
lines by other groups [2, 4, 5], suggesting that the Tus-Ter complex
can be used as a versatile tool to engineer site-specific DNA replica-
tion perturbation in any cell type. Combined with the recent
advances in genome engineering using CRISPR-Cas9 and short-
homology targeting sequences [6], this could permit detailed ana-
lyses of DNA replication perturbation at specific regions of the
genome in eukaryotic cells in a manner analogous to those devel-
oped in budding yeast and fission yeast [7–9].
The plasmids for engineering Tus-Ter barriers into the yeast
genome contain 14
Ter or 21
Ter sites, in either Leftwards
(L ! R) or Rightwards (L R) fork-blocking orientation, and
are flanked by either the LEU2 or URA3 selectable markers on
either side. Because all 16 of these Ter vectors harbor common
primer sequences flanking the Ter modules, a number of config-
urations of Ter sites and selectable marker(s) can simultaneously be
amplified using the same set of locus-specific primers. These Ter
modules can then be integrated into the yeast genome by homolo-
gous recombination, and the validated strains transformed with a
plasmid that allows the controlled expression of HA-tagged Tus (or
a non-blocking allele of Tus as a control [2]). The choice of two
different selectable markers (LEU2 or URA3) in the Ter vectors
allows two discrete barriers to be engineered/maintained in a given
yeast strain. These Tus-Ter barriers can either be constructed in
trans to allow in vivo comparisons between different locations/
chromosomes, or else constructed in cis to additively disrupt DNA
replication on the same chromosome. Additional combinations/
configurations of Tus-Ter barriers are also theoretically possible
through marker switching or recycling, if required. Furthermore,
the URA3 marker can also be used as a counter-selectable reporter
to assay for replication-fork-arrest-induced mutagenesis at Tus-Ter
barriers (by positive selection for ura3 mutations on plates contain-
ing 5-FOA [10]). The positioning of the URA3 reporter on either
side of the Tus-Ter barrier permits the two regions either side of a
Tus-Ter barrier to be probed for mutagenic outcomes. For exam-
ple, an origin proximal URA3 configuration permits positive selec-
tion for mutagenic outcomes either directly at, or behind, stalled
replication forks. Conversely, an origin-distal URA3 configuration
permits positive selection for mutagenic events associated with fork
resumption, or fork merging within this reporter gene.
Combined with the well-established methodologies to study
DNA replication in yeast [11], this system serves as a “molecular
toolbox” that can be used to create, and then monitor the pheno-
typic consequences of various challenges to the DNA replication
machinery. This will permit the detailed analysis of localized
 A Molecular Toolbox to Engineer Site-Specific DNA Replication Perturbation 297

responses to DNA replication perturbation at discrete regions of

the genome (e.g., early/late replicating regions, fragile sites, sub-
telomeric loci, etc.), as well as analyzing the fate of unreplicated
regions of the genome that persist beyond S-phase [12].

2 Materials

2.1 DNA 1. High fidelity PCR polymerase (e.g., PrimeSTAR HS DNA

Amplification for polymerase from Takara-Clontech).
Construction of Yeast 2. Locus-specific targeting primers (see Note 1 and
Strains with Inducible Subheading 3.1).
Tus-Ter Barriers 3. Ter plasmid to use as a PCR template (see Note 2 and Fig. 1).

2.2 Yeast 1. YEP Broth þ 2% Glucose.

Transformation 2. LiAc-PEG3350 mix (Combine 1.5 ml of 1 M LiAc with 10 ml
50% PEG).
3. Carrier DNA (2 mg/ml herring sperm DNA).
4. CSM-Ura plates.

2.3 Validation of 1. 200 mM LiAc, 1% SDS solution.

Successful Ter Module 2. 96–99% Ethanol.
Integrants
3. Frozen-EZ Yeast Transformation II kit from Zymo Research.

Origin-proximal URA3 reporter Origin-distal URA3 reporter

ARS305
ARS305

14xTer 14xTer

URA3 URA3
ARS305
ARS305

14xTer 14xTer

URA3 URA3

(no pausing) (no pausing)

Fig. 1 Diagram of the Tus-Ter barrier constructs engineered adjacent to the early-firing origin, ARS305, on
ChrIII. The four configurations are shown: permissive, or restrictive, Ter sites, with an origin-proximal, or
origin-distal, URA3 marker. A number of other configurations are possible if using different Ter vectors as PCR
templates. A 21
Ter module can be amplified instead of 14
Ter, or a LEU2 marker can be amplified instead
of URA3. In total, 16
[Ter þ marker] configurations are possible
298 Nicolai B. Larsen et al.

2.4 Synchronization 1. YEP Broth þ 3% Lactate (autoclaved).

of Cells in the G1 Phase 2. Alpha-factor mating pheromone (Synthetic peptide of the fol-
and Induction of Tus lowing sequence: WHWLQLKPGQPMY, >90% purity).
3. Galactose.

2.5 Flow Cytometric 1. RNase A (10 mg/ml).

Analysis 2. Proteinase K (20 mg/ml).
3. Propidium Iodide solution (1 mg/ml).

2.6 DNA Extraction 1. 10% Sodium azide solution (keep at 4  C).

for Two-Dimensional 2. Spheroplast buffer (1 M sorbitol, 100 mM EDTA pH 8, 0.1%
Gel Electrophoresis β-mercaptoethanol, 100 U/ml zymolyase (AMSBIO, UK)).
3. Solution I (2% w/v CTAB, 1.4 M NaCl, 100 mM Tris–HCl
pH 7.6, 25 mM EDTA pH 8.0).
4. Solution II (1% CTAB, 50 mM Tris–HCl pH 7.6, 10 mM
EDTA).
5. Solution III (1.4 M NaCl, 10 mM Tris–HCl pH 7.6, 1 mM
EDTA).
6. Chloroform/isoamylalcohol 24:1.
7. KIMBLE HS Borosilicate glass centrifuge tubes.
8. Isopropanol.
9. 99% ethanol.
10. 10 mM Tris–HCl pH 8.0.

2.7 Two- 1. 2.5 M potassium acetate, pH 6.

Dimensional gel 2. TE buffer.
Electrophoresis (2DGE)
3. Low EEO agarose (e.g., USBiological).
of DNA Extracts
4. Owl A2 Large Gel System (ThermoFisher Scientific).
5. 2DGE Loading Dye (0.83% Bromophenol blue, 50% glycerol).
6. 10 mg/ml Ethidium bromide.
7. TBE buffer.
8. Depurination solution (8 ml of 37% HCL per 1 L water).
9. Denaturing solution (0.5 M NaOH, 1.5 M NaCl).
10. Neutralization solution (1 M Ammonium acetate, 0.02 M
NaOH).
11. 10
SSC.
12. Genescreen Hybridization Transfer Membranes (Perkin Elmer,
USA).
13. UVC crosslinker.
 A Molecular Toolbox to Engineer Site-Specific DNA Replication Perturbation 299

2.8 Hybridization 1. Hybridization solution (6
SSC, 1
Denhardt’s, 1% sarcosyl;

of Membranes For Pre-Hybridization solution, add BSA to 0.1% final
concentration).
2. DNA probe.
3. Rediprime II DNA labeling system (GE Healthcare
Amersham).
32
4. P dCTP (6000 Ci/mmole).
5. Quick Spin Column (Roche).
6. Wash Solution I (2
SSC, 1% SDS; incubate at 65 ˚C).
7. Wash Solution II (0.1
SSC, 0.1% SDS; incubate at 37 ˚C).

3 Methods

The protocol detailed below describes how to engineer site-specific

replication fork stalling at any desired location in the yeast genome.
To date, we have used this protocol to integrate Ter modules into
five discrete locations, and subsequently observed DNA replication
fork pausing when Tus is induced. For the purposes of brevity, we
describe how to integrate and validate DNA replication fork paus-
ing at a number of different modules inserted adjacent (to the
right) of ARS305 on ChrIII: 14
Ter-URA3 and URA3-14
Ter
(Fig. 1). In this scenario, the L ! R barrier will be the “restrictive”
(blocking) orientation, whereas the R L barrier will serve as the
“permissive” (non-blocking) control, relative to the ARS305-
borne forks. Note that the restrictive/permissive configurations of
Tus-Ter barriers will always depend on the activity of local DNA
replication origins, which determine the timing and directionality
of DNA replication in the region of interest. Relative origin usage
and efficiency have been extensively characterized in budding yeast
[13, 14], and these studies can therefore be used as the basis for
strategic engineering of Tus-Ter barriers.

3.1 DNA 1. All 16 Ter plasmids (see Note 2) have the same common flank-
Amplification for ing sequences that can be used as “TMA” (Ter module ampli-
Construction of Yeast fication) PCR primers:
Strains with Inducible TMA-1 Fwd primer: cgactcactatagggcgaattgg.
Tus-Ter Barriers TMA-2 Rev. primer: gccgctctagaactagtggatc.
To target the region to the right of ARS305, 65-bp homology
arms were included (indicated in upper case; see Note 1) to give
the final primer sequences as:
305-TMA-1 Fwd:
GAGCAAGACAAACAGGGCCAGCTGAT
GCATATGTTTTGTGTTGCTTTCCTACGATCAGCTAA
TGCcgactcactatagggcgaattgg.
300 Nicolai B. Larsen et al.

305-TMA-2 Rev.:
AAATGAGTTTTGTCCCACCTTCCCTTTGG
GAAAAGGCAATGTAAATCTTAGAGGCAAGAACCAC
Agccgctctagaactagtggatc.
2. The [Ter þ marker] modules should be amplified using a high-
fidelity PCR enzyme (using ~10–20 pg of Ter vector as a PCR
template), at a high annealing temperature. Typically, we use
PrimeSTAR HS DNA polymerase from Takara-Clontech, with
the following PCR parameters:
(a) 98 ˚C for 90 s.
(b) [98 ˚C for 10 s, 60 ˚C for 15 s, 72 ˚C for 1 min/
kb]
35 cycles.
(c) 72 ˚C for 10 min.
3. In the example shown above, the entire Ter cassette, selectable
marker, and homology arms are amplified as a single PCR
product. However, it is also possible to use “split-URA3
PCR” to generate two 428-bp overlapping PCR products
that can only create a functional URA3 gene by recombining
into the desired locus. If Ter sequences flank the split-URA3
PCR products on either side, then co-transformation of these
(see Subheading 3.2 below) permits up to 42
Tus-Ter barriers
(i.e., 21
Ter-URA3(URA3 promoter þ 1-533) + (105-804 þ URA3
terminator)
URA3-21
Ter ¼ 21
Ter-URA3-21
Ter) to be
engineered within a ~3 kb fragment in a single round of yeast
transformation.
The sequences of the “split-URA3” primers are as follows:
Split-URA3-Fwd: (to be paired with the locus-specific TMA-
2 Rev. primer): TTGGATGTTCGTACCACCAAGGAAT.
Split-URA3-Rev.: (to be paired with the locus-specific TMA-1
Fwd primer): GAGCAATAAAGCCGATAACAAAATCTT
TGTCG.

3.2 Yeast The transformation procedure is adapted from Gietz and Schiestl
Transformation [15] and has been optimized for use in BY4741 strains. This
protocol can also be used for making gene disruptants (see Note 3).
1. Grow cells overnight in YPD.
2. The next day, dilute cells ~1:40 in YPD and grow for 5–6 h at
30 ˚C/200 rpm. Each transformation reaction requires 15 ml
of exponentially growing cells (OD600 reading of 1.0–1.2).
3. Boil 2 mg/ml carrier DNA aliquots for 5 min, and then keep
them on ice.
4. Prepare labeled tubes with 900 μl of LiAc-PEG mix.
5. Harvest cells from 15 ml aliquots of cultures by centrifuging at
900
g for 5 min. Resuspend cells in 1 ml water, transfer the
 A Molecular Toolbox to Engineer Site-Specific DNA Replication Perturbation 301

mixture to 1.5 ml microfuge tubes, and briefly centrifuge.

Remove the water to leave a cell pellet.
6. To each cell pellet, add 150 μl of carrier DNA and (up to) 15 μl
of deletion cassette (or approximately 1 μg of PCR products).
Mix by pipetting and transfer the entire mixture to the Falcon
tubes containing 900 μl LiAc-PEG mix.
7. Incubate the transformation reactions at 30 ˚C/200 rpm for
10 min.
8. Transfer to 42 ˚C/180 rpm for 10–12 min in a water bath
shaker.
9. Transfer the tubes to an ice-water bath to instantly chill them.
10. Add 3 ml of appropriate prechilled selection medium to the
tubes, gently mix, and then centrifuge at 1000 rpm for 1 min.
11. After pouring off the supernatant, gently resuspend the cells in
the ~300 μl of residual media. Plate cells onto CSM–Ura plates
(see Note 3).
12. After 3 days, Ura+ colonies should be detectable. Restreak
single colonies onto fresh CSM–Ura plates.

3.3 Validation of This DNA extraction procedure [16] can also be used to verify
Successful Ter Module successful integration of Ter modules (or gene disruptions) by
Integrants PCR using primers that flank the targeting regions. For those
strains that give correct band sizes indicative of successful integra-
tion, it is recommended to confirm the DNA sequence is correct. In
our experience, the success rate (i.e., no sequence alterations) is
usually >66%.
1. Harvest cells from 200 to 300 μl of stationary phase cultures,
and resuspend in 300 μl of 200 mM LiAc, 1% SDS solution.
2. Incubate for 10 min at 70  C/600 rpm.
3. Add 1 ml of 99% ethanol, and then vortex the samples.
4. Centrifuge the DNA and cell debris at full speed for 5 min.
5. Wash the pellet with 70% ethanol.
6. Dissolve the pellet in 200 μl of water, and incubate samples for
a few hours at 37 ˚C to facilitate resuspension of DNA.
7. Clarify the samples by centrifugation and use 1–2 μl of the
supernatant as a PCR template.
Successfully transformed (and sequence validated) strains can
then be transformed with a plasmid that allows the controlled
expression of plasmid-borne HA-tagged Tus (or a non-blocking
control) from the GAL1-promoter [2]. To transform yeast
with plasmids, we routinely use the Frozen-EZ Yeast Transforma-
tion II kit from Zymo Research, and follow the manufacturer’s
instructions.
302 Nicolai B. Larsen et al.

3.4 Synchronization This protocol has been optimized for BY4741 strains. If using
of Cells in G1 and another strain background, or if a particular deletion mutant is
Induction of Tus slow growing, some optimization of this protocol may be required.
1. Grow cultures in 4 ml of selective medium overnight.
2. The next day, add 1 ml yeast per 25 ml medium. OD600 read-
ings should be ~0.35. Incubate cultures at 200 rpm, 30 ˚C for
~7 h until OD600 reading reaches ~1.30.
3. Set up overnight cultures in baffled flasks. The culture should
comprise no more than 1/4th of the total volume. Add cells
to 0.04 OD units/ml. Grow cultures overnight at 200 rpm,
30 ˚C.
4. After ~17–18 h, the OD600 value should be around ~1.5–1.6.
Add 5 μg/ml alpha factor (see Note 4), and galactose to 2%
final concentration.
5. Incubate yeast cultures for 2.5 h at 200 rpm, 30 ˚C.
6. After arrest/induction, harvest cells by centrifugation
(2000
g for 3 min).
7. Wash cells twice in 100 ml YEP þ 3% lactate þ 2% Gal
(2000
g for 3 min), and resuspend the cells in double the
starting volume (in step 4) of YEP þ 3% lactate þ 2% Gal.
8. Incubate cultures at 30 ˚C, 200 rpm. Harvest cells at fixed time
points for flow cytometry and 2DGE analysis.

3.5 Flow Cytometric Bulk DNA replication by flow cytometry is detectable at approxi-
Analysis mately 20 min after release from alpha-factor arrest, and is usually
completed around 60 min later (i.e., at 80 min).
1. Harvest 1 ml of cells by centrifugation and fix in 70% ethanol
overnight.
2. Wash cells, and resuspend in 1 ml of 50 mM sodium citrate
(pH 7.0).
3. Briefly sonicate cells (see Note 5), and then treat with 0.25 mg/
ml RNase A for 1 h at 50  C.
4. Add Proteinase K to a final concentration of 1 mg/ml, and
incubate for a further 1 h at 50  C.
5. Dilute samples in 50 mM sodium citrate containing 16 μg/ml
of propidium iodide, and incubate at room temperature for a
minimum of 30 min. Remaining samples can be stored at 4  C.
6. Analyze samples using a flow cytometer. We use a Becton
Dickinson FACSCalibur machine, and CellQuest software for
analysis.

3.6 DNA Extraction This protocol is based on the hexadecyltrimethylammonium bro-

for 2DGE mide (“CTAB”) method of DNA extraction [11]. For the analysis
of ARS305-borne replication forks, the optimal time point for
A Molecular Toolbox to Engineer Site-Specific DNA Replication Perturbation 303

2DGE analysis is approx. 30–35 min following release from G1

arrest.
1. Add 0.1% (final w/v) sodium azide to 200 ml aliquots of cells at
predefined time points, and harvest the cells by centrifugation
(900
g, 5 min).
2. Prepare spheroplasts by incubating cells in 5 ml of spheroplast
buffer (1 M sorbitol, 100 mM EDTA pH 8, 0.1% β-mercap-
toethanol, 100 U/ml zymolyase (AMSBIO, UK)) for 1 h at
30  C, 200 rpm.
3. Harvest spheroplasts by centrifugation (3500
g, 10 min),
briefly wash in 5 ml water, and then lyse the spheroplasts in
2 ml water (see Note 6).
4. Add 2.5 ml of Solution I (2% w/v CTAB, 1.4 M NaCl,
100 mM Tris–HCl pH 7.6, 25 mM EDTA pH 8.0) and
200 μl of 10 mg/ml RNase A, and incubate samples at 50  C
for 30 min.
5. Add 200 μl of 20 mg/ml Proteinase K, and incubate samples
for a further 90 min at 50  C.
6. Add a further 100 μl of 20 mg/ml Proteinase K, and incubate
samples overnight at 30  C.
7. Separate the supernatant (S; see steps 8–10 below) and pellet
(P; see steps 11–13 below) by centrifugation (3500
g for
10 min) and treat as follows.
8. (S) Combine the supernatant with 2.5 ml of 24:1 chloroform:
isoamylalcohol in 15 ml Falcon tubes, mix (see Note 7), and
clarify by centrifugation (3500
g for 5 min).
9. (S) Transfer the upper phase to borosilicate glass centrifuge
tubes containing 10 ml of Solution II (1% CTAB, 50 mM
Tris–HCl pH 7.6, 10 mM EDTA), and incubate for 1–2 h,
with occasional gentle mixing.
10. (S) Precipitate the DNA by centrifugation at 8500
g for
10 min. Resuspend the DNA in 2.5 ml of Solution III
(1.4 M NaCl, 10 mM Tris–HCl pH 7.6, 1 mM EDTA).
11. (P) Resuspend the pellet from step 7 in 2 ml of Solution III
and incubate at 50  C for 1 h. Gently mix samples at regular
intervals to facilitate resuspension.
12. (P) Combine with 2.5 ml of 24:1 chloroform:isoamylalcohol in
15 ml Falcon tubes, mix (see Note 7), and clarify by centrifu-
gation (3500
g for 5 min).
13. (P) Combine the upper phase with the resuspended DNA from
step 10.
14. Add an equivalent volume (5 ml) of isopropanol, mix gently,
and then centrifuge at 8500
g for 10 min.
304 Nicolai B. Larsen et al.

15. Briefly wash the pellet in 75% ethanol (8500
g for 1 min).
16. Resuspend the DNA in 10 mM Tris–HCl pH 8.0 (see Note 8).

3.7 Two- For optimal 2DGE images, ~15–20 μg DNA should be cut with
Dimensional gel DNA restriction enzymes that liberate a ~5 kb restriction fragment
Electrophoresis (2DGE) for Southern blot analysis. Restriction fragments of <4 kb or >6 kb
of DNA Extracts generally give poorer-quality images in our experience. For other
considerations regarding restriction enzymes, see Note 9. The basic
principles of this technique are that the first dimension gel separates
DNA restriction fragments on the basis of their size (low voltage;
no ethidium bromide), whereas the second dimension gel separates
DNA restriction fragments on the basis of their shape (high volt-
age; þ ethidium bromide). The types of DNA structures that
can be identified using this technique have been described exten-
sively [11].
1. For each sample, digest 15–20 μg of DNA with appropriate
restriction enzymes for a minimum of 5 h. Typical restriction
digests contain 100 U of enzymes, and a total volume of
650 μl.
2. Add 83 μl of 2.5 M Potassium Acetate solution (pH 6), and
730 μl of isopropanol. Gently mix by inversion, and then
centrifuge at 16,000
g for 10 min.
3. Wash pellets in 75% ethanol, and centrifuge at 16,000
g for
5 min.
4. Air-dry the pellets to remove all residual ethanol, and then
resuspend the DNA in 20 μl of TE buffer. Allow a minimum
of 1 h for resuspension of DNA. In the meantime, prepare the
first dimension gel.
5. Prepare a 0.35% low EEO agarose (US Biological, USA) gel.
The agarose should be 55  C when pouring, and the gels
should be poured in a cold room (see Note 10).
6. Add 5 μl of 2DGE loading dye to each of the DNA samples,
gently mix, and then carefully load onto the first dimension gel.
Leave a spare lane between each sample to allow cutting of gel
slices the next day. Load DNA markers in the final lane.
7. Run first dimension gels overnight at 50 V for 18–21 h in TBE
buffer. The length of time depends on the gel apparatus, and
the size of restriction fragment being analyzed. In our standard
conditions, the linear form of the restriction fragment (~5 kb)
has usually migrated ~10 cm.
8. Once the first dimension is complete, stain the gel with
TBE þ 0.3 μg/ml ethidium bromide for 30 min.
9. Cut 6.5 cm gel strips from first dimension gels that contain the
1N ! 2N restriction fragments (i.e., cut from ~4 ! 12 kb).
Rotate gel slices 90 and arrange (six gel slices per second
 A Molecular Toolbox to Engineer Site-Specific DNA Replication Perturbation 305

dimension gel) in two horizontal lines of three, separated ver-

tically by a minimum of 10 cm. Use a large knife and flexible
plastic ruler to help maneuver the fragile gel slices. Pour a
second dimension 0.90% agarose gel (þ0.3 μg/ml ethidium
bromide) around these.
10. Run second dimension gels at 180 V (in TBE buffer containing
0.3 μg/ml ethidium bromide) for 8 h at 4  C. The DNA from
each sample will be visible as an arc on the second dimension
gels.
11. Cut second dimension gels into 20 cm
10 cm slices, and
transfer the DNA samples to membranes by Southern blotting.
The following wash procedures should be performed prior to
assembling the Southern blot apparatus:

Depurination solution 15 min

Denaturing solution 20 min
Neutralization solution 20 min

12. The Southern blot apparatus (from bottom to top) should be

arranged as follows: Whatman paper wick soaked in 10
SSC,
second dimension Gel, Pre-soaked membrane, 2
soaked
Whattman paper, 2
dry Whattman paper, paper towels
(~10–15 cm), plastic tray, weight. Take care to remove any air
bubbles between the gel and membrane, and use parafilm to
cover any exposed parts of the paper wick.
13. After 16–24 h, the DNA is (faintly) detectable on membranes
using a UV transilluminator. After drying the membranes for
30 min on filter paper, expose both the sides of the membranes
to 2
cycles of 120,000 μJ UVC. Membranes can then be
enclosed in filter paper until they are ready to hybridize.

3.8 Hybridization DNA replication intermediates can be detected using unique

of Membranes locus-specific 32P dCTP (6000 Ci/mmole) radiolabeled probes
(see Note 11).
1. Wash membranes three times in water and add (up to 4) to
hybridization tubes.
2. Add 80 ml of pre-hybridization buffer to each tube, and incu-
bate at 65  C with rotation for >1 h.
3. To make the radiolabeled probe, denature 100 ng of the DNA
probe (in 47.5 μl TE buffer) at 100  C for 10 min, and then
transfer the denatured DNA to an ice:water bath for 5 min.
4. Combine the DNA from step 3 with 2.5 μl of 32P dCTP
(6000 Ci/mmole) in the Rediprime II DNA labeling system.
Incubate at 37  C for >15 min.
306 Nicolai B. Larsen et al.

5. Remove the unincorporated nucleotides using a Quick Spin

column.
6. Boil the flow-through at 100  C for 10 min. Also boil the
carrier DNA aliquots (500 μl of 4 mg/ml herring sperm
DNA).
7. For each hybridization tube, prepare 20 ml of 65  C pre-
warmed hybridization buffer, and combine the following:
(a) >5 μl of radiolabelled probe (split the flow-through
evenly between the hybridization tubes).
(b) Denatured carrier DNA.
8. Pour off the pre-hybridization buffer, and replace with the
hybridization solution from step 7.
9. Incubate overnight at 65  C in a hybridization oven (with
rotation).
10. The next day, pour off the hybridization buffer, and add 50 ml
Wash Solution I. Incubate at 65  C (with rotation) for 10 min.
11. Transfer all of the membranes to a plastic box, and wash them
as follows (at room temperature, with gentle agitation):
500 ml Wash Solution I for 15 min.
500 ml Wash Solution II for 10 min (
3 washes).
12. Dry the membranes on filter paper for 30 min, and then wrap
in a cling film. Transfer the membranes to phosphorscreen
cassettes, and store overnight.
13. Scan the phosphorscreens using a phosphorimaging system
(e.g., Typhoon, Storm, or Phosphorimager). Replication inter-
mediates are usually detectable after 1–2 days.
14. If required, stripping of membranes for subsequent reprobing
can be achieved by incubating the membranes in a boiling
solution of 0.1% SDS for >1 h at 65  C (with rotation).
Membranes can then be probed again using the above
procedure.

4 Notes

1. Ter targeting primers should be PAGE-purified following syn-

thesis to ensure they are the correct sequence. If particular
combinations of 65-bp targeting sequences are inefficient,
then the length of the homology arms can be increased
through an additional round of PCR.
2. Ter vectors should be propagated using Tus-negative bacteria.
Small aliquots of the desired vectors, as well as their complete
sequences, are available on request. The HA-Tus expression
plasmids (with either LEU2, MET15, or HIS3 markers) are also
available on request.
A Molecular Toolbox to Engineer Site-Specific DNA Replication Perturbation 307

3. When plating the transformed cells, the cell density is impor-

tant. We recommend plating 1/3rd and 2/3rd of the cells,
respectively, onto two large selection plates to ensure differing
cell densities. If using this protocol for making gene disruptants
requiring drug selection, resuspend the cells in YPD for 3 h at
30 ˚C, 200 rpm prior to plating. After this time, harvest the
cells by centrifugation at 1000 rpm for 1 min, and then plate
the cells onto large YPD plates containing the desired drug. We
typically use the following markers/drug doses for gene dis-
ruptants: kanMX (400 mg/L G418), hphMX (300 mg/L
Hygromycin B), or natMX (100 mg/L CloNAT).
4. When starting the alpha-factor arrest, the cell cycle distribution
should be approximately 50:50 1C:2C, as judged by flow
cytometry. If the majority of cells in the culture have a 1C
content, modifying the growth conditions to improve aeration
is recommended. Alpha-factor should also be batch-tested to
find the optimal concentration for efficient G1-arrest, and
subsequent release.
5. Over-sonication of cells can cause DNA fragmentation (detect-
able by flow cytometry as a sub-1C peak). We have observed
that different strain backgrounds require different amounts of
sonication to disrupt cell clumps. For BY4741, we typically use
~13 W output for 10 s. However, we recommend that the
sonication conditions are optimized for each strain/sonicating
apparatus using trial runs of logarithmic, G1-arrested, and G2/
M-arrested cells, for validation.
6. During the 5 ml wash step, carefully invert the tubes to wash
the sides of the tubes. This step helps remove any residual
sorbitol that can interfere with later steps. Following addition
of 2 ml water, incubate the samples at 200–300 rpm until a
homogenous mixture is obtained. This can usually take up to
an hour.
7. Shake the tubes three times, and release any pressure due to gas
build-up. Do not shake the samples too vigorously or violently,
but just enough to get a (transiently) even suspension. After
the separation of the phases by centrifugation, take care not to
accidentally take up any of the lower phase while pipetting.
8. Resuspend DNA in a minimum of 500 μl of 10 mM Tris pH 8.
Incubate samples at room temperature for 1–2 h to facilitate
resuspension of DNA, and then store the samples at 4  C. Do
not freeze samples at 20  C, as this can destroy fragile repli-
cation intermediates. Always use cut-tips when transferring the
DNA and avoid vortexing or any mechanical stress.
9. The same membranes can be stripped and reprobed multiple
times to analyze different loci. Therefore, if strains contain two
Tus-Ter barriers, one should choose compatible enzymes that
308 Nicolai B. Larsen et al.

give ~5 kb fragments in each case. Other genomic loci may also

be considered controls for early/late-replicating regions. We
generally find that addition of multiple restriction enzymes
often improves the quality of the 2DGE images, as this helps
remove background signals.
10. First dimension gels are very fragile and should be handled
carefully in all subsequent steps. Take care in particular to
remove the gel combs carefully while the gel is submerged in
TBE buffer. It is recommended that the wells are stained with a
small amount of loading dye prior to loading the DNA samples
to (1) ensure that the wells are intact, and (2) facilitate loading
of samples.
11. This should be a unique and pure PCR product for the region
of the genome to be analyzed. This can be made by doing a
two-step PCR, using “nested” primers for the second step and
low amounts (10–20 pg) of the primary PCR product as
template.

Acknowledgments

Work in the authors’ laboratory is funded by the Danish National

Research Foundation (DNRF115), The European Research Coun-
cil, The Novo Nordisk Foundation, and The Nordea Foundation.

References
1. Hill TM, Marians KJ (1990) Escherichia coli 6. Natsume T, Kiyomitsu T, Saga Y et al (2016)
Tus protein acts to arrest the progression of Rapid protein depletion in human cells by
DNA replication forks in vitro. Proc Natl auxin-inducible degron tagging with short
Acad Sci U S A 87(7):2481–2485 homology donors. Cell Rep 15(1):210–218.
2. Larsen NB, Hickson ID, Mankouri HW doi:10.1016/j.celrep.2016.03.001
(2014) Tus-Ter as a tool to study site-specific 7. Calzada A, Hodgson B, Kanemaki M et al
DNA replication perturbation in eukaryotes. (2005) Molecular anatomy and regulation of
Cell Cycle 13(19):2994–2998. doi:10.4161/ a stable replisome at a paused eukaryotic DNA
15384101.2014.958912 replication fork. Genes Dev 19
3. Larsen NB, Sass E, Suski C et al (2014) The (16):1905–1919. doi:10.1101/gad.337205
Escherichia coli Tus-Ter replication fork barrier 8. Ahn JS, Osman F, Whitby MC (2005) Replica-
causes site-specific DNA replication perturba- tion fork blockage by RTS1 at an ectopic site
tion in yeast. Nat Commun 5:3574. doi:10. promotes recombination in fission yeast.
1038/ncomms4574 EMBO J 24(11):2011–2023. doi:10.1038/sj.
4. Willis NA, Chandramouly G, Huang B et al emboj.7600670
(2014) BRCA1 controls homologous recom- 9. Lambert S, Watson A, Sheedy DM et al (2005)
bination at Tus/Ter-stalled mammalian repli- Gross chromosomal rearrangements and ele-
cation forks. Nature 510(7506):556–559. vated recombination at an inducible site-
doi:10.1038/nature13295 specific replication fork barrier. Cell 121
5. Willis NA, Scully R (2016) Spatial separation of (5):689–702. doi:10.1016/j.cell.2005.03.022
replisome arrest sites influences homologous 10. Iraqui I, Chekkal Y, Jmari N et al (2012)
recombination quality at a Tus/Ter-mediated Recovery of arrested replication forks by
replication fork barrier. Cell Cycle:1–9. doi:10. homologous recombination is error-prone.
1080/15384101.2016.1172149 PLoS Genet 8(10):e1002976. doi:10.1371/
journal.pgen.1002976
 A Molecular Toolbox to Engineer Site-Specific DNA Replication Perturbation 309

11. Liberi G, Cotta-Ramusino C, Lopes M et al 14. Muller CA, Hawkins M, Retkute R et al (2014)
(2006) Methods to study replication fork col- The dynamics of genome replication using
lapse in budding yeast. Methods Enzymol deep sequencing. Nucleic Acids Res 42(1):e3.
409:442–462 doi:10.1093/nar/gkt878
12. Mankouri HW, Huttner D, Hickson ID (2013) 15. Gietz RD, Schiestl RH (2007) High-efficiency
How unfinished business from S-phase affects yeast transformation using the LiAc/SS carrier
mitosis and beyond. EMBO J 32(20): DNA/PEG method. Nat Protoc 2(1):31–34.
2661–2671. doi:10.1038/emboj.2013.211 doi:10.1038/nprot.2007.13
13. Raghuraman MK, Winzeler EA, Collingwood 16. Looke M, Kristjuhan K, Kristjuhan A (2011)
D et al (2001) Replication dynamics of the Extraction of genomic DNA from yeasts for
yeast genome. Science 294(5540):115–121. PCR-based applications. Biotechniques 50
doi:10.1126/science.294.5540.115 (5):325–328. doi:10.2144/000113672
 Chapter 21

Single Cell Gel Electrophoresis for the Detection of Genomic

Ribonucleotides
Barbara Kind, Christine Wolf, Kerstin Engel, Alexander Rapp,
M. Cristina Cardoso, and Min Ae Lee-Kirsch

Abstract
Single cell gel electrophoresis or comet assay enables the quantification of DNA damage such as single-
strand or double-strand breaks on a single cell level. Here, we describe a variant of this method for the
detection of ribonucleotides embedded in genomic DNA. Briefly, cells are embedded in agarose on a
microscopic slide, lysed under high salt and alkaline conditions and then subjected to in situ treatment with
E. coli RNase HII which nicks 50 to a ribonucleotide within the context of a DNA duplex thereby
converting genomic ribonucleotides into strand breaks. After unwinding of genomic DNA using a highly
alkaline buffer, electrophoresis under mild alkaline conditions is performed resulting in formation of comets
due to migration of fragmented DNA toward the anode. Following SYBR Gold staining comets can be
visualized by fluorescence microscopy. In this setting, the length and the intensity of comets formed reflect
the level of genomic ribonucleotides present in a given cell.

Key words Comet assay, Single cell electrophoresis, Alkaline lysis, RNase H2, Ribonucleotides,
Genomic DNA

1 Introduction

Ribonucleotides misincorporated into genomic DNA by replicative

polymerases represent the most frequent DNA base lesion in repli-
cating mammalian cells [1]. In vivo, genomic ribonucleotides are
removed by ribonucleotide excision repair, which is initiated by
ribonuclease H2 (RNase H2) through cleavage 50 of a ribonucleo-
tide followed by strand displacement synthesis by Pol δ, flap
removal by FEN1, and ligation by DNA ligase 1 [2]. If left unre-
paired, ribonucleotides render the DNA backbone susceptible to
DNA strand breaks leading to genome instability [1, 3–5]. Thus,
RNase H2 plays an essential role in the maintenance of genome
integrity. In humans, mutations in the genes encoding the three

Barbara Kind and Christine Wolf are co-first authors.

Marco Muzi-Falconi and Grant W. Brown (eds.), Genome Instability: Methods and Protocols, Methods in Molecular Biology,
vol. 1672, DOI 10.1007/978-1-4939-7306-4_21, © Springer Science+Business Media LLC 2018

311
312 Barbara Kind et al.

RNase H2 subunits (RNASEH2A, RNASEH2B, RNASEH2C)

cause the neuroinflammatory disorder Aicardi-Goutières syndrome
[6]. While complete RNase H2 deficiency in mice is embryonic
lethal due to massive DNA damage [1, 5], RNase H2 mutations
found in patients with Aicardi-Goutières syndrome were shown to
be hypomorphic [7, 8]. In patient cells, increased levels of genomic
ribonucleotides cause chronic low level DNA damage leading to
the activation of the antiviral type I interferon axis [8–10].
Here, we describe a method for the detection and quantifica-
tion of ribonucleotides embedded in genomic DNA of human
fibroblasts based on single cell gel electrophoresis or comet assay
[8]. The basic principle of the comet assay rests on the migration of
negatively charged DNA fragments out of agarose-embedded
nuclei toward the anode during electrophoresis resulting in the
formation of comet tails [11–13]. While electrophoresis under
neutral pH conditions predominantly exposes DNA double-strand
breaks, electrophoresis under highly alkaline denaturing conditions
(pH >13) further exposes single-strand breaks in DNA [12]. Fur-
thermore, the use of specific endonucleases that convert base
lesions into single-strand breaks enables detection of lesion-specific
DNA damage. For example, treatment of cells with 8-oxoguanine
DNA glycosylase prior to electrophoresis allows detection of oxi-
dized bases [14, 15]. In our protocol E. coli RNase HII is used to
nick genomic DNA at sites of embedded ribonucleotides to create
single-strand breaks. The steps involved in the comet assay for the
detection of genomic ribonucleotides are depicted in Fig. 1. Cells
are first embedded in agarose on a microscopic slide and lysed to
remove membranes, cytoplasm, and nucleoplasm including nucleo-
somes. The remaining nucleoids are subjected to RNase HII treat-
ment followed by electrophoresis using a mildly alkaline buffer.
Migration of fragmented DNA leads to the formation of comets,
the sizes of which reflect the extent of DNA fragmentation gener-
ated by RNase HII treatment. The representative images of comets
obtained with this method in human fibroblasts deficient in RNase
H2 are shown in Fig. 2.

2 Materials

Prepare all solutions using ultrapure water (prepared by purifying

deionized water, to attain a sensitivity of 18 MΩ  cm at 25  C) and
analytical grade reagents. Prepare and store all reagents at room
temperature unless indicated otherwise.
1. Low melting Agarose (1%): Dissolve 1 g agarose (type VII, low
gelling temperature; Sigma Aldrich) in 100 ml 1 PBS. Make
aliquots and store at 20  C.
Single Cell Gel Electrophoresis for the Detection of Genomic Ribonucleotides 313

coating of slides

embedding of cells

cell lysis

RNase HII digestion

alkaline unwinding

- +
electrophoresis

staining

microscopy

analysis

Fig. 1 Flow chart depicting steps involved in the modified comet assay for the
detection of genomic ribonucleotides.

2. Normal melting Agarose (0.5%): Dissolve 0.5 g agarose (type

II; Sigma Aldrich) in 100 ml 1 PBS. Make aliquots and store
at 20  C.
3. Normal melting Agarose (1%): Dissolve 1 g agarose (type II;
Sigma Aldrich) in 100 ml 1 PBS. Make aliquots and store at
20  C.
4. Antifade solution with SYBR Gold: Dissolve 2.33 g DABCO
(1,4-diazabicyclo[2.2.2]octane; 20 mM) in 2 ml 1 M Tris–HCl
(pH 8.0) and 80 ml glycerol by heating to 70  C. Add distilled
water to a volume of 100 ml. Add 1 μl of SYBR Gold stock
solution (Thermo Fisher Scientific) to 10 ml Antifade solution
and store in a dark vial at 20  C.
5. Slides (Dakin Fully Frosted Slides, 71876-01; Electron Micros-
copy Sciences).
6. Phosphate-buffered saline (1 PBS): 137 mM NaCl, 2.7 mM
KCl, 10 mM Na2HPO4, 2 mM KH2PO4): Dissolve 8 g NaCl,
0.2 g KCl, 1.44 g Na2HPO4 and 0.24 g KH2PO4 in 800 ml
distilled H2O. Adjust pH to 7.4 with HCl. Add distilled water
to a volume of 1 L.
314 Barbara Kind et al.

a - RNase HII + RNase HII

wild type

mutant

- cathode anode +

c
60
- RNase HII
50 + RNase HII
Olive tail moment

40

30

20

10

0
wild type mutant

Fig. 2 Modified comet assay of RNase H2-deficient human fibroblasts. (a)

Images of comets obtained from primary human fibroblasts of a healthy control
(wild type) and a patient with RNase H2 deficiency (mutant) with and without
RNase HII (/þ RNase HII) treatment. Scale bar 20 μm. (b) Schematic of slide
after electrophoresis and SYBR Gold staining. The red box indicates the area
from which comets should be sampled for analysis. Superimposed comets
should be excluded from analysis. (c) Olive tail moments as determined with
CASP software. At least 50 comets were analyzed per slide. Shown are the
means and SD.

7. Tris-borate-EDTA (10 TBE): 0.9 M Tris base, 0.9 M boric

acid, 20 mM EDTA): Dissolve 108 g Tris, 55 g boric acid and
9.3 g Na2H2EDTA in 800 ml distilled water. Add distilled
water to a volume of 1 L.
 Single Cell Gel Electrophoresis for the Detection of Genomic Ribonucleotides 315

8. Alkaline lysis buffer: 2.5 M NaCl, 0.1 M EDTA, 10 mM Tris.

Dissolve 73.05 g NaCl, 18.61 g Na2H2EDTA and 0.606 g Tris
in 400 ml distilled water. Adjust to pH 10 with 1 N NaOH.
Adjust volume to 500 ml with distilled water. Add 1 ml Triton
X-100 per 100 ml buffer (1%) prior to lysis.
9. RNase H2 buffer: 20 mM Tris–HCl, 10 mM (NH4)2SO4,
10 mM KCl, 2 mM MgSO4. Mix 20 ml 1 M Tris–HCl
(pH 8.0), 10 ml 1 M (NH4)2SO4, 10 ml 1 M KCl and 2 ml
1 M MgSO4. Adjust to pH 8.8 with 1 N NaOH. Add distilled
water to a volume to 1 L. Add 1 ml Triton X-100 per 1 L buffer
(0.1%) prior to use.
10. E. coli RNase HII (M0288; NEB).
11. 0.1% SDS: Dissolve 1 g SDS in 1 L distilled water.
12. Alkaline unwinding buffer: 0.3 M NaOH, 1 mM
Na2H2EDTA. Mix 60 ml 5 N NaOH and 2 ml 0.5 M
Na2H2EDTA with 800 ml distilled water. Adjust pH to 13.1
with 10 N NaOH and add distilled water to a volume of 1 L.
13. Neutral electrophoresis buffer (0.5 TBE).

3 Methods

Carry out all the procedures at room temperature unless otherwise

specified.

3.1 Coating of Slides Pre-warm slides on a hot plate (IKAMAG Rec-G) at 50  C. First
layer: Disperse 50 μl of warm (50  C) 0.5% normal melting agarose
per slide using a second slide to spread the agarose. Let agarose
solidify. Second layer: Disperse 400 μl of warm (50  C) 1% normal
melting agarose per slide. Quickly cover the agarose with a
24  60 mm coverslip to smoothen the surface (see Note 1).
Agarose-coated slides can be stored in a moist chamber at 4  C
for up to 1 week.

3.2 Embedding Melt aliquots of 1% low melting agarose in a water bath at 42  C.

of Cells For cooling of slides, prepare ice box and place metal plate on ice
(see Note 2). Prepare alkaline lysis buffer with 1% Triton X and cool
to 4  C. Trypsinize fibroblasts and centrifuge at 200  g for 5 min
and 4  C. Discard the supernatant, resuspend cells in 1 ml 1 PBS.
Dilute cell suspension to 1  106 cells per ml in 1 PBS. Pre-warm
agarose-coated slides to 42  C on a hot plate and remove coverslip
shortly prior to adding the agarose-cell suspension. Add 150 μl of
cell suspension to 600 μl of 1% low melting agarose and mix
carefully using a pipette with a cut tip to prevent shearing of cells.
Pipette 150 μl of agarose-cell suspension per precoated slide and
quickly cover with a 24  60 mm coverslip. Transfer the slide
immediately onto a cold plate on ice.
316 Barbara Kind et al.

3.3 Cell Lysis Allow the agarose-cell suspension to solidify for 5 min, carefully
remove the coverslip, and let slides air-dry for a few minutes.
Incubate slides in a glass staining jar filled with precooled alkaline
lysis solution at 4  C overnight (see Note 3). Wash slides 10 for
10 min each with 1 PBS.

3.4 RNase HII Equilibrate slides 3 for 20 min each in RNase H2 buffer, drain
Digestion excess buffer by placing slides upright on a paper towel. Dilute E. coli
RNase HII in RNase H2 buffer at 100 U per 50 μl. Add 50 μl of
RNase HII solution per slide. Add 50 μl RNase H2 buffer without
enzyme per no enzyme control slide. Cover slides with a 24  60 mm
coverslip and incubate them in a wet chamber at 37  C overnight.
Stop RNase HII digestion by washing the slides in 0.1% SDS for
10 min. Wash slides 10 for 10 min each with 1 PBS.

3.5 Alkaline Wash slides 2 using alkaline unwinding buffer for a few seconds.
Unwinding Equilibrate cells for 20 min in alkaline unwinding buffer. Wash
slides 10 for 10 min each in 0.5 TBE.

3.6 Electrophoresis Precool neutral electrophoresis buffer to 4  C and pour into an

Under Neutral electrophoresis chamber (Owl Scientific, model 5) just to cover the
Conditions tray holder. Place slides next to each other onto try holder with the
short edges facing the electrodes. Fill up electrophoresis chamber
with buffer. Conduct electrophoresis at 1 V/cm and 15 mA for
25 min (see Note 4).

3.7 Staining of Place slides upright on paper towel for 10 min to drain buffer. Place
Comets slides horizontally on a paper towel, add 200 μl Antifade solution
with SYBR Gold per slide and cover with a coverslip. Store slides in
a dark wet chamber at 4  C for at least 30 min before microscopic
analysis.

3.8 Microscopic Use a fluorescence microscope with an excitation filter at 495 nm

Analysis and an emission filter at 537 nm and a 10 objective. Image
50–100 randomly selected comets per slide using identical imaging
settings. The fluorescence intensity of the comet tail relative to the
head reflects the number of DNA breaks. Quantify the degree of
DNA migration by assessing the Olive tail moment using comet
assay analysis software such as CASP (CaspLab) [16] (see Note 5).
The Olive tail moment is defined as the product of the tail length
and the fraction of total DNA in the tail [11].

4 Notes

1. Avoid formation of bubbles in agarose, which could interfere

with electrophoresis. Per cell line to be analyzed prepare three
slides with RNase HII digestion and three slides without
RNase HII as no enzyme control.
 Single Cell Gel Electrophoresis for the Detection of Genomic Ribonucleotides 317

2. We use the stainless-steel cover of an instrument tray, which is

large enough to fit the number of slides needed for the experi-
ment. One may also use a glass plate.
3. For a Coplin staining jar (Thermo Fisher Scientific) holding
eight slides, 70 ml lysis buffer are needed. Remove and add
washing buffers very slowly to avoid agarose from detaching off
the slides. We use a glass funnel pointed toward the wall of the
staining jar when adding buffer.
4. To eliminate effects of electric field inhomogeneity during
electrophoresis, place replicate sample slides and controls in
alternating positions into the electrophoresis chamber avoiding
the outer rims of the tray holder. Make sure to run enzyme
treated and no enzyme controls in the same electrophoresis
run. Two to three rows of slides can be placed into the electro-
phoresis chamber we use. Run electrophoresis in a 4  C cold
room.
5. Quantification of comets can be done either by visual scoring
or by using image analysis software tools available as commer-
cial packages or freeware (for an overview see www.cometassay.
com).

Acknowledgment

This work was supported by grants by the Deutsche Forschungsge-

meinschaft (KFO 249; LE 1074/4-1 and LE 1074/4-2 to M.L.-K.;
KI 1956/2-1 to B.K.) and the Bundesministerium f€ ur Forschung
und Bildung (grants 02S8355 and 02NUK017D to M.C.C.; grant
02NUK036D to A.R.). B.K. is a recipient of a Maria Reiche fellow-
ship of TU Dresden.

References
1. Reijns MA, Rabe B, Rigby RE, Mill P, Astell 4. Kim N, Huang SN, Williams JS, Li YC, Clark
KR, Lettice LA, Boyle S, Leitch A, Keighren AB, Cho JE, Kunkel TA, Pommier Y, Jinks-
M, Kilanowski F et al (2012) Enzymatic Robertson S (2011) Mutagenic processing of
removal of ribonucleotides from DNA is essen- ribonucleotides in DNA by yeast topoisomer-
tial for mammalian genome integrity and ase I. Science 332:1561–1564
development. Cell 149:1008–1022 5. Hiller B, Achleitner M, Glage S, Naumann R,
2. Sparks JL, Chon H, Cerritelli SM, Kunkel TA, Behrendt R, Roers A (2012) Mammalian
Johansson E, Crouch RJ, Burgers PM (2012) RNase H2 removes ribonucleotides from
RNase H2-initiated ribonucleotide excision DNA to maintain genome integrity. J Exp
repair. Mol Cell 47:980–986 Med 209:1419–1426
3. Nick McElhinny SA, Kumar D, Clark AB, Watt 6. Crow YJ, Leitch A, Hayward BE, Garner A,
DL, Watts BE, Lundstrom EB, Johansson E, Parmar R, Griffith E, Ali M, Semple C, Aicardi
Chabes A, Kunkel TA (2010) Genome instabil- J, Babul-Hirji R et al (2006) Mutations in
ity due to ribonucleotide incorporation into genes encoding ribonuclease H2 subunits
DNA. Nat Chem Biol 6:774–781 cause Aicardi-Goutieres syndrome and mimic
318 Barbara Kind et al.

congenital viral brain infection. Nat Genet 11. Ostling O, Johanson KJ (1984) Microelectro-
38:910–916 phoretic study of radiation-induced DNA
7. Kind B, Muster B, Staroske W, Herce HD, damages in individual mammalian cells. Bio-
Sachse R, Rapp A, Schmidt F, Koss S, Cardoso chem Biophys Res Commun 123:291–298
MC, Lee-Kirsch MA (2014) Altered spatio- 12. Singh NP, McCoy MT, Tice RR, Schneider EL
temporal dynamics of RNase H2 complex (1988) A simple technique for quantitation of
assembly at replication and repair sites in low levels of DNA damage in individual cells.
Aicardi-Goutieres syndrome. Hum Mol Genet Exp Cell Res 175:184–191
23:5950–5960 13. Olive PL, Banath JP (2006) The comet assay: a
8. Gunther C, Kind B, Reijns MA, Berndt N, method to measure DNA damage in individual
Martinez-Bueno M, Wolf C, Tungler V, cells. Nat Protoc 1:23–29
Chara O, Lee YA, Hubner N et al (2015) 14. Collins AR, Duthie SJ, Dobson VL (1993)
Defective removal of ribonucleotides from Direct enzymic detection of endogenous oxi-
DNA promotes systemic autoimmunity. J Clin dative base damage in human lymphocyte
Invest 125:413–424 DNA. Carcinogenesis 14:1733–1735
9. Mackenzie KJ, Carroll P, Lettice L, Tarnauskaitė 15. Kushwaha S, Vikram A, Trivedi PP, Jena GB
Ž, Reddy K, Dix F, Revuelta A, Abbondati E, (2011) Alkaline, Endo III and FPG modified
Rigby RE, Rabe B et al (2016) Ribonuclease H2 comet assay as biomarkers for the detection of
mutations induce a cGAS/STING-dependent oxidative DNA damage in rats with experimen-
innate immune response. EMBO J tally induced diabetes. Mutat Res
35:831–844. doi:10.15252/embj.201593339 726:242–250. doi:10.1016/j.mrgentox.
10. Pokatayev V, Hasin N, Chon H, Cerritelli SM, 2011.10.004
Sakhuja K, Ward JM, Morris HD, Yan N, 16. Końca K, Lankoff A, Banasik A, Lisowska H,
Crouch RJ (2016) RNase H2 catalytic core Kuszewski T, Góźdź S, Koza Z, Wojcik A
Aicardi-Goutières syndrome-related mutant (2003) A cross-platform public domain PC
invokes cGAS-STING innate immune-sensing image-analysis program for the comet assay.
pathway in mice. J Exp Med 213:329–336. Mutat Res 534:15–20
doi:10.1084/jem.20151464
 Chapter 22

Measuring the Levels of Ribonucleotides Embedded

in Genomic DNA
Alice Meroni, Giulia M. Nava, Sarah Sertic, Paolo Plevani,
Marco Muzi-Falconi, and Federico Lazzaro

Abstract
Ribonucleotides (rNTPs) are incorporated into genomic DNA at a relatively high frequency during
replication. They have beneficial effects but, if not removed from the chromosomes, increase genomic
instability. Here, we describe a fast method to easily estimate the amounts of embedded ribonucleotides
into the genome. The protocol described is performed in Saccharomyces cerevisiae and allows us to quantify
altered levels of rNMPs due to different mutations in the replicative polymerase ε. However, this protocol
can be easily applied to cells derived from any organism.

Key words DNA replication, DNA repair, DNA polymerase, Ribonucleotides incorporation, RNase
H, Genome stability, Genomic rNMPs

1 Introduction

During evolution, DNA was selected as the principal molecule to

preserve genetic information likely due to its greater stability com-
pared to RNA, whose 30 hydroxyl group increases its susceptibility
to hydrolysis.
At every cell cycle, genomic DNA is duplicated by DNA poly-
merases, enzymes that are specialized to copy a single-stranded
DNA template and polymerize deoxyribonucleotides (dNTPs)
accordingly, forming a complementary DNA strand. Given the
much greater abundance of rNTPs compared to dNTPs in the
nucleus, DNA polymerases evolved a steric gate to help preventing
rNTPs from entering the active site [1].
However, recent data revealed that large amounts of rNTPs are
incorporated in genomic DNA during replication [2]. The presence
of rNMPs in the chromosomes has physiological roles [3–5] and is

M. Muzi-Falconi and F. Lazzaro are co-last authors.

Marco Muzi-Falconi and Grant W. Brown (eds.), Genome Instability: Methods and Protocols, Methods in Molecular Biology,
vol. 1672, DOI 10.1007/978-1-4939-7306-4_22, © Springer Science+Business Media LLC 2018

319
320 Alice Meroni et al.

normally transient: specific RNase H-based pathways excise them

before mitosis [6]. Failure to remove genomic rNMPs causes repli-
cation stress and genome instability in yeast and human cells
[7–12]. Mutations in the genes coding for RNase H2 in humans
are responsible for the rare Aicardi-Goutieres Syndrome (AGS)
[13]. Intriguingly, cells derived from AGS patients accumulate
rNMPs in their chromosomes and exhibit constitutively activated
DNA damage response and post-replication repair mechanisms [8,
10].
To investigate the mechanisms underlying incorporation and
removal of ribonucleotides in chromosomes and to determine their
effect on genome integrity, it is important to determine rapidly and
semiquantitatively the amounts of ribonucleotides present genomic
DNA. Here, we describe an experimental strategy based on the
approach originally described by Hiller and colleagues [12] and
then in [8]. Briefly, genomic DNA is extracted and treated in vitro
with bacterial RNase HII, which introduces nicks at every ribonu-
cleotides site. These nicks are radioactively labeled taking advantage
of the DNA Polymerase I nick translation capability (Fig. 1). In this
chapter, we describe the procedure starting from the preparation of
genomic DNA from yeast cells and compare the effect of two
mutations affecting the steric gate of pol ε, M644G and M644L,
that respectively increase and decrease rNTPs incorporation [9].

2 Materials

1. Eppendorf tubes 1.5 and 2 mL.

2. Pipettes and tips.
3. Glass Pasteur pipette.
4. MilliQ water.
5. 250 mL glass flasks.
6. Stirrer.
7. Gel electrophoresis apparatus.
8. Power supply.
9. UV transilluminator and digital camera.
10. Plastic wrap.
11. Tape.
12. Scalpel.
13. Thermomixer.
14. Geiger counter.
15. Gel dryer.
 Ribonucleotides Incorporation Assay 321

rNMP 5’
3’
rNMP
rNMP

5’ 3’

RNHII

3’ 5’

5’ 3’

dCTP
DNA-PolI

5’
3’

3’
5’
Fig. 1 Representative scheme for ribonucleotides incorporation assay. RNHII recognizes and cleaves ribonu-
cleotides embedded into genomic DNA (red dot) leaving 50 P-ribonucleotide ends. The DNA-PolI enzyme,
through nick translation, marks RNHII-induced nicks with radiolabeled dCTP

16. 3 MM Whatman blotting papers.

17. Towel papers.
18. Weight (400 g).
19. Phosphorimager and screen.
20. YDER Yeast DNA Extraction kit, materials and reagents listed
in the kit instructions (Thermo Scientific).
21. RNase A 10 mg/mL.
22. Phenol:Chloroform:Isoamyl Alcohol 25:24:1 v/v/v
(Saturated with 10 mM Tris–HCl, pH 8.0, 1 mM EDTA).
Stored at 4  C.
23. 100% ethanol. Stored at 20  C.
24. 3 M sodium acetate, pH 7.0. Stored at 4  C.
25. 70% ethanol. Stored at 20  C.
322 Alice Meroni et al.

26. Agarose powder.

27. 10 mg/mL ethidium bromide (EtBr).
28. TAE 1: 40 mM Tris–HCl, pH 8.0, 20 mM acetic acid, and
1 mM EDTA. Stored at room temperature.
29. Lambda DNA marker.
30. RNase HII 5000 U/mL.
31. 10 ThermoPol® Reaction Buffer: 200 mM Tris–HCl,
100 mM (NH4)2SO4, 100 mM KCl, 20 mM MgSO4, 1%
Triton® X-100 (New England Biolabs).
32. 10 dNTPs mix (without dCTP); 200 μM dATP, 200 μM
dGTP and 200 μM dTTP.
33. NEBuffer 2; 500 mM NaCl,100 mM Tris–HCl, 100 mM
MgCl2, 10 mM DTT (New England Biolabs).
34. DNA Polymerase I 10,000 U/mL.
35. α32P-dCTP, 3000 Ci/mmol.
36. STOP solution: 30% glycerol, 200 mM EDTA, bromophenol
blue, in MilliQ water.
37. TCA 30%.
38. Software for quantification and analysis: ImageQuant and
Microsoft Excel.

3 Methods

3.1 Genomic DNA 1. Isolate yeast genomic DNA using the Y-DER extraction Kit
Preparation according to the manufacturer’s instructions. All the steps are
performed as described in the kit’s instructions, with the fol-
lowing modifications:
(a) Use 50 mL cultures with an OD600 between 0.3 and 0.8.
(b) RNase A 10 mg/mL is diluted 1:1000 in the Y-PER
reagent.
2. Resuspend DNA in 200 μL of MilliQ water and add an equal
volume of Phenol:Chloroform:Isoamyl Alcohol 25:24:1 v/v/v
saturated with 10 mM Tris–HCl, pH 8.0, 1 mM EDTA
(see Note 1).
3. Vortex vigorously for 15 s.
4. Centrifuge at maximum speed for 10 min at RT.
5. Carefully transfer only the aqueous phase (upper phase) to a
new 1.5 mL eppendorf tube. That phase contains DNA. Do
not transfer material from the interface or the lower phase. If
so, repeat the procedure from step 2, adding an equal volume
of Phenol:Chloroform:Isoamyl Alcohol 25:24:1 v/v/v
saturated with 10 mM Tris–HCl, pH 8.0, 1 mM EDTA.
 Ribonucleotides Incorporation Assay 323

6. Precipitate DNA by adding three volumes of ice-cold 100%

ethanol and 1/10 of the volume of sodium acetate 3 M,
pH 7.0. Mix well and keep overnight at 20  C.
7. Spin down the precipitate at maximum speed for 450 at 4  C.
Draw off the supernatant.
8. Wash adding 1 mL of ice-cold 70% ethanol. Spin at maximum
speed for 30 min at 4  C. Draw off the supernatant and dry the
pellet at 42–44  C (see Note 2).
9. Resuspend gently the pellet with 50 μL of MilliQ water.

3.2 Nicking DNA at 1. Quantify genomic DNA by loading 2 μL on a 1% agarose gel in

Ribonucleotide Sites TAE (with 0.67 μg/mL EtBr) next to 1 μL of Lambda DNA
marker. Run for 10 min at 8–10 V/cm. The genomic DNA
band should be compact and easily quantifiable (see Note 3).
2. Normalize DNA in each sample to 25 ng/μL by adding MilliQ
water.
3. Prepare two new 1.5 mL eppendorf tubes per sample. Label the
tubes with the sample name plus “"” and “þ” (e.g., Sample
1 and Sample 1þ) (see Note 4), and transfer 20 μL (500 ng)
of normalized genomic DNA to each tube.
4. Dilute 1:10 the RNase HII in ThermoPol Buffer 1. Prepare
the two reaction mixtures (mix “” and mix “þ”) in new
1.5 mL tubes in excess with respect to the number of samples.
Keep the mixtures on ice.
1 reaction mix recipe (Note that the final reaction volume is
50 μL):

Mix “” Mix “þ”

ThermoPol Buffer 10 5 μL 5 μL
RNase HII / 1 μL
ThermoPol Buffer 1 1 μL /
H2O MilliQ 24 μL 24 μL
Total volume 30 μL 30 μL

5. Vortex briefly and add 30 μL of each mixture to the appropriate

labeled tube containing genomic DNA.
6. Incubate at 37  C with 550 rpm agitation in a thermomixer for
2.30 h.
7. Add 50 μL of MilliQ water and precipitate DNA following
steps 6–8 of section 3.1, consider 100 μL of total volume.
8. Resuspend gently the pellet in 20 μL of MilliQ water.
324 Alice Meroni et al.

3.3 Radioactive 1. Quantify and normalize DNA by loading 2 μL on a 1% agarose

Labeling of Nicks gel in TAE (with 0.67 μg/mL EtBr) next to 1 μL of Lambda
DNA marker. Run for 10 min at 9–10 V/cm (see Note for gel
preparation).
2. Transfer 300 ng of DNA to a new 1.5 mL tube and add MilliQ
water to 15 μL of total volume.
3. Prepare the common DNA Polymerase I reaction mixture,
keep it on ice.
1 DNA Polymerase I reaction mix recipe:

10 dNTPs mix (without dCTP): 2 μL

DNA Pol I 10 buffer NEBuffer 2 2 μL
DNA Pol I 10,000 U/mL 0.5 μL
α P-dCTP 3000 Ci/mmol
32
0.3 μL
H2O MilliQ 0.2 μL

4. Add 5 μL of the DNA Polymerase I reaction mix to the each

sample and incubate at 16  C for 30 min (see Note 5).
5. Add 4 μL of the STOP solution (see Notes 5 and 6).
6. Load 20 μL on a 1% agarose gel in TAE (with 0.67 μg/mL
EtBr). Here the DNA size marker is dispensable.
7. Run at 7 V/cm for 1.20 h.
8. Cut the gel immediately under the bromophenol blue line (see
Note 7).
9. Examine the gel by UV light and photograph it digitally. This
will allow normalization of the radioactive signal with respect
to the DNA loaded in the gel.
10. Soak the gel in TCA 30% for half an hour to precipitate the
DNA. The bromophenol blue turns yellow.
11. Assemble the sandwich on a glass tray: cover completely the
inner part with plastic wrap. Layer three pieces of 3MM What-
man blotting paper larger than the gel. Take out the gel from
TCA 30% and place it on the top of the blotting paper. Place in
order: three more sheets of blotting paper on the gel, a stack of
paper towels, and a weight (0.3–0.5 kg) (Fig. 2).
12. Let the gel dry overnight at room temperature.
13. Remove the paper towels and the blotting papers above the gel.
Transfer the desiccated gel on a new 3MM blotting paper.
14. Dry the gel on the blotting paper using a gel dryer for 20 min
at 80  C. Use more than one blotting paper and cover the gel
with plastic wrap to avoid radioactive contamination of the
instrument.
 Ribonucleotides Incorporation Assay 325

Weight (300 g)

Towels (10 cm)

3 blotting papers

Gel

3 blotting papers

Plastic wrap

Glass Tray

Fig. 2 Scheme to assemble paper sandwich to dry the agarose gel

RNH201 rnh201Δ rnh201Δ rnh201Δ

POL2 POL2 pol2-M644G pol2-M644L
- + - + - + - + RNHII

32
P-dCTP

EtBr

Fig. 3 Visualization of ribonucleotides incorporation assay results. The strains tested are derivatives of a W303
background (MATa ade2–1 trp1–1 leu2–3112 his3–11,15 ura3–1 can1–100 RAD5) with a deletion of gene
coding for the catalytic subunit of RNase H2 (rnh201Δ) combined with wt or mutated POL2 gene. The RNH201
POL2 wt strain is used as control. The radiolabeled signal represents the nicks labeled by PolI. The signal
dependent upon RNHII treatment is proportional to the genomic ribonucleotides levels. The EtBr panel
represents the loading control, acquired before gel dryng and necessary for radioactive quantification

15. Expose the dried gel on a phosphorimager screen for 5–15 min
(see Note 8).
16. Scan the screen in a phophorimager. To quantify the result see
Note 9. An Example is shown in Fig. 3.
326 Alice Meroni et al.

4 Notes

1. DNA has to be clean and as little nicked as possible to achieve

the best resolution. For these reasons it is strongly recom-
mended to clean DNA through phenol:chloroform extraction
and ethanol precipitation. We also found that the YDER prep-
aration yields a genomic DNA with fewer nicks compared with
other methods.
2. Draw off as much supernatant as possible with a glass Pasteur
pipette, this would ensure removing the majority of the etha-
nol, and then place the eppendorf tubes in a heater at
42–44  C. Drying time depends on how much ethanol is left
in the samples. Check the samples after 20 min, if the ethanol is
still there, leave them in the heater and check later. Note that
excessive drying would damage the sample, for this reason
keep checking samples every 20 min until the pellet is
completely dry.
3. To normalize DNA in the samples take a digital image and use
quantifying tools such as ImageQuant or ImageLab (BioRad).
4. Here, each sample is split into two: one half is digested with
purified bacterial RNase HII, the other half is left untreated.
This allows discrimination between the ribonucleotide-
dependent nicks and the nicks generated during DNA
preparation.
5. Labeling the nicks is the key step of this protocol. For this
reason the procedure needs strict standardization. We suggest
proceeding sample by sample adding the reaction mix to each
tube every 15–30 s. After 30 min of incubation, repeat this
procedure with the STOP solution. In this way, all the samples
would be incubated for exactly 30 min at 16  C.
6. As most DNA polymerases, DNA polymerase I needs magne-
sium ions. In this case the high concentration of EDTA in the
STOP solution will stop the reaction, while the glycerol and the
bromophenol blue make the samples ready to be loaded on a
gel.
7. Labeled genomic DNA is loaded on the agarose gel together
with the DNA Polymerase I reaction mix, including the unin-
corporated radioactive nucleotides. The nucleotides migrate
faster than genomic DNA and the long run ensures complete
separation. Free nucleotides migrate immediately below the
bromophenol blue; therefore, the gel is cut immediately after
the run to avoid their diffusion through the gel. The bromo-
phenol blue is also a pH-indicator, used to monitor the change
in the gel pH while soaking in 30% TCA. When it turns yellow
 Ribonucleotides Incorporation Assay 327

the agarose gel pH has become acid so DNA would be pre-

cipitated into it.
8. The exposure time could vary depending on the α32P-dCTP
activity. With fresh and fully active α32P-dCTP the signal is
saturated in 15 min.
9. Signal quantification. To quantify the resulting signal proceed
with the following step. Quantify the bands corresponding to
the genomic DNA, including the ones from the EtBr capture.
Use the Volume Tool of ImageQuant software drawing a rect-
angle around the right band. Normalize each radioactive value
to the corresponding EtBr value. This would ensure the correct
interpretation of the radioactive signal. Then for each sample
subtract the “” signal from the “þ” signal as a background
normalization.

Acknowledgments

This work was supported by grants from AIRC (n.15631) and

Telethon (GGP15227) to M.M.-F., from AIRC and Fondazione
Cariplo (TRIDEO 2014 Id. 15724) to F.L., and from Fondazione
Cariplo (grant number 2013-0798) to P.P.

References
1. Joyce CM (1997) Choosing the right sugar: 8. Pizzi S, Sertic S, Orcesi S et al (2015) Reduc-
how polymerases select a nucleotide substrate. tion of hRNase H2 activity in Aicardi-Gou-
Proc Natl Acad Sci U S A 94:1619–1622 tières syndrome cells leads to replication stress
2. McElhinny SAN, Watts BE, Kumar D et al and genome instability. Hum Mol Genet
(2010) Abundant ribonucleotide incorpora- 24:649–658
tion into DNA by yeast replicative polymerases. 9. Nick McElhinny SA, Kumar D, Clark AB et al
Proc Natl Acad Sci U S A 107:4949–4954 (2010) Genome instability due to ribonucleo-
3. Ghodgaonkar MM, Lazzaro F, Olivera- tide incorporation into DNA. Nat Chem Biol
Pimentel M et al (2013) Ribonucleotides mis- 6:774–781
incorporated into DNA act as strand- 10. G€ unther C, Kind B, Reijns MAM et al (2014)
discrimination signals in eukaryotic mismatch Defective removal of ribonucleotides from
repair. Mol Cell 50:323–332 DNA promotes systemic autoimmunity. J Clin
4. Lujan SA, Williams JS, Clausen AR et al (2013) Invest 125(1):413–424
Ribonucleotides are signals for mismatch repair 11. Reijns MAM, Rabe B, Rigby RE et al (2012)
of leading-strand replication errors. Mol Cell Enzymatic removal of ribonucleotides from
50:437–443 DNA is essential for mammalian genome
5. Dalgaard JZ (2012) Causes and consequences integrity and development. Cell
of ribonucleotide incorporation into nuclear 149:1008–1022
DNA. Trends Genet 28:592–597 12. Hiller B, Achleitner M, Glage S et al (2012)
6. Sparks JL, Chon H, Cerritelli SM et al (2012) Mammalian RNase H2 removes ribonucleo-
RNase H2-initiated ribonucleotide excision tides from DNA to maintain genome integrity.
repair. Mol Cell 47:980–986 J Exp Med 209:1419–1426
7. Lazzaro F, Novarina D, Amara F et al (2012) 13. Crow YJ, Manel N (2015) Aicardi-Goutières
RNase H and postreplication repair protect syndrome and the type I interferonopathies.
cells from ribonucleotides incorporated in Nat Rev Immunol 15(7):429–440
DNA. Mol Cell 45:99–110
 Chapter 23

Mapping Ribonucleotides Incorporated into DNA

by Hydrolytic End-Sequencing
Clinton D. Orebaugh, Scott A. Lujan, Adam B. Burkholder,
Anders R. Clausen, and Thomas A. Kunkel

Abstract
Ribonucleotides embedded within DNA render the DNA sensitive to the formation of single-stranded
breaks under alkali conditions. Here, we describe a next-generation sequencing method called hydrolytic
end sequencing (HydEn-seq) to map ribonucleotides inserted into the genome of Saccharomyce cerevisiae
strains deficient in ribonucleotide excision repair. We use this method to map several genomic features in
wild-type and replicase variant yeast strains.

Key words HydEn-seq, DNA replication, DNA repair, DNA polymerase, Ribonucleotide excision
repair, RNase H2, Next-generation sequencing, Bioinformatics, Genomics

1 Introduction

Ribonucleotides are abundantly incorporated into DNA during

eukaryotic DNA replication [1–4]. These ribonucleotides are nor-
mally removed by RNase H2-dependent ribonucleotide excision
repair (RER) [5, 6]. The presence of ribonucleotides in DNA can
benefit the cell [7–9], and ribonucleotides left unrepaired can lead
to replication stress [10, 11], DNA strand breaks [5, 12], and
autoimmune disease in humans [13–15]. Deletion of the gene
encoding the catalytic subunit of the RNase H2 enzyme,
RNH201, eliminates RER, resulting in persistent genomic ribonu-
cleotides [2, 5, 7, 16–18]. The presence of a 20 -OH group of the
ribose sugar imbedded within a DNA polynucleotide increases the
chances of spontaneous DNA hydrolysis, especially in the presence
of increased pH [19, 20]. Treatment of DNA containing ribonu-
cleotides with alkali results in a strand break on the 30 side of the
ribonucleotide and produces DNA ends containing a 20 -30 -cyclic
phosphate and 50 -OH ends [5, 19, 20]. This provides an opportu-
nity to monitor ribonucleotide incorporation by wild-type forms of

Marco Muzi-Falconi and Grant W. Brown (eds.), Genome Instability: Methods and Protocols, Methods in Molecular Biology,
vol. 1672, DOI 10.1007/978-1-4939-7306-4_23, © Springer Science+Business Media LLC 2018

329
330 Clinton D. Orebaugh et al.

the three major yeast replicative DNA polymerases, Pol a (Pol1),

Pol d (Pol3), and Pol ε (Pol2), and by variant polymerase derivatives
with altered ribonucleotide incorporation propensity [1, 5, 7, 16,
18, 21–23]. This is evident by the detection of abundant alkali-
sensitive sites in the nuclear genome of yeast strains harboring wild
type and the variant replicases in combination with the deletion of
the RNH201 gene. Here, we describe Hydrolytic End-sequencing
(HydEn-seq), a method that enables detection and mapping of
ribonucleotides across the yeast genome using next-generation
sequencing on an Illumina platform [21].

2 Materials

2.1 Genomic DNA 1. Sterile toothpicks.

Purification and 2. YPDA Agar: 20 g dextrose, 20 g Bacto-peptone, 10 g Bacto-
Alkaline Hydrolysis yeast extract, 20 g Bacto-agar, and 2 mL of a 0.5% adenine
sulfate solution brought up to 1 L with deionized water in a
graduated cylinder and sterilized for 20 min in an autoclave.
Pour 30 mL into 10 cm petri dishes and cool at room temper-
ature to solidify.
3. YPDA Broth: 20 g dextrose, 20 g Bacto-peptone, 10 g Bacto-
yeast extract, and 2 mL of a 0.5% adenine sulfate solution
brought up to 1 L with deionized water in a graduated cylinder
and sterilized for 20 min in an autoclave.
4. Erlenmeyer flask.
5. MasterPure Yeast DNA Purification kit (Epicentre).
6. Isopropanol.
7. 70% Ethanol.
8. Microcentrifuge.
9. 1.5 mL microcentrifuge tubes (sterilized).
10. Laboratory oven.
11. 3 M KOH.
12. 3 M Na Acetate pH 5.2.

2.2 Library 1. PCR thermal cycler.

Preparation 2. HighPrep PCR magnetic beads (MagBio Genomics).
3. Magnetic-ring stand (96-well) (ThermoFisher Scientific).
4. Phosphorylation reagents:
(a) T4 Polynucleotide Kinase (10,000 U/mL) (30 phospha-
tase minus) (NEB).
(b) 10 Polynucleotide Kinase buffer (NEB).
(c) ATP (10 mM).
 HydEn-seq Mapping 331

5. Adapter ligation reagents:

(a) ARC 140 Oligonucleotide with 50 amino terminated C6
spacer (50 -/5AmMC6/ACACTCTTTCCCTACAC-
GACGCTCTTCCGATCT-30 ) (100 μM).
(b) T4-RNA ligase (10,000 U/mL) (NEB).
(c) T4-RNA Ligase 10 buffer (NEB).
(d) ATP (2 mM).
(e) PEG 8000 (50% solution).
(f) CoCl3(NH3)6 (10 mM).
6. Second strand synthesis reagents:
(a) ARC 76/ARC 77 annealed primers (ARC 76: 50 -
GTGACTGGAGTTCAGACGTGTGCTCTTCC-
GATCTNNNN*N*N -30 , ARC 77: 50 -AGATCGGAA-
GAGCACACGTCTGAACTCCAGTC*A*C - 30 , [“*”
indicates a phosphorothioate bond]) (10 μM).
(b) T7 DNA Polymerase (unmodified) (NEB).
(c) T7 DNA Polymerase 10 buffer (NEB).
(d) Deoxynucleotide solution mix (2 mM each of dATP,
dCTP, dGTP, and dTTP).
(e) Bovine serum albumin (1 mg/mL).
7. PCR amplification reagents:
(a) ARC 49 oligonucleotide (50 -AATGATACGGCGAC-
CACCGAGATCTACACTCTTTCCCTACAC-
GACGCTCTTCCGATCT-30 ) (2 μM).
(b) Barcoded oligonucleotide (50 -CAAGCAGAAGACGG-
CATACGAGATX(6–8)GTGACTGGAGTTCA-
GACGTGTGCTCTTCCGATCT-30 (X(6–8) is a unique
barcode sequence for each library to be sequenced in the
same pool)) (2 μM).
(c) KAPA HiFi HotStart DNA polymerase ReadyMix (2)
(KAPA Biosystems).

2.3 Quality Check 1. 2100 bioanalyzer (Agilent).

2. DNA 7500 or 1000 Reagents (Agilent).

2.4 DNA Sequencing 1. HiSeq2500 high-throughput DNA sequencer (Illumina).

3 Methods

3.1 Yeast Growth 1. Using a sterile toothpick, streak yeast strain(s) on YPDA agar
and Genomic DNA plates to produce single colonies, and incubate about 48 h at
Purification 30 ˚C.
332 Clinton D. Orebaugh et al.

2. Resuspend about five colonies in 5 mL of YPDA broth and add

a sufficient volume to 50 mL of YPDA broth (about
50–2000 μL) so that an overnight growth at 30 ˚C with shak-
ing at 160 RPM produces a culture in exponential growth
phase (~OD600 ¼ 0.3–0.8).
3. After overnight growth, pellet 15 OD600 units of each culture
by centrifugation, remove all growth media, and purify geno-
mic DNA according to the Epicentre MasterPure Yeast DNA
Purification kit instructions with the omission of the RNase A
treatment step (see Note 1).

3.2 Alkaline 1. Treat genomic DNA with 300 mM KOH for 2 h at 55 ˚C in a

Hydrolysis of Yeast laboratory oven in a total volume of 50 μL (Fig. 1, Step 1).
Genomic DNA 2. Precipitate genomic DNA by combining with 0.2 volumes of
3 M sodium acetate (pH 5.2) and 2.5 volumes of ice-cold 100%
ethanol. Incubate on ice for 5 min, and spin full speed in a
microcentrifuge for 5 min to pellet precipitated DNA. Wash
with 70% ethanol, spin again at full speed for 5 min, aspirate any
remaining ethanol, and allow pellet to air-dry if necessary.
Resuspend DNA pellet in 20 μL of Qiagen buffer EB.

3.3 Phosphorylation 1. Warm MagBio beads at room temperature during this step.
of DNA 50 Termini 2. Assemble the phosphorylation reaction without DNA:
(a) 1 μL T4 Polynucleotide Kinase (30 phosphatase minus)
(b) 2.5 μL 10 Polynucleotide Kinase buffer, 2.5 μL ATP
(10 mM)
3. Denature 19 μL of DNA resuspended from the previous step at
85 ˚C for 3 min, then cool on ice.
4. Add 6 μL of reaction mix to denatured DNA and mix by
pipetting up and down six to eight times.
5. Incubate the reaction for 30 min at 37 ˚C, then heat inactivate
for 20 min at 65 ˚C.
6. Purify phosphorylated DNA with magnetic PCR beads (see
Note 2), add 36 μL (1.8 reaction volumes) of magnetic beads
to the reaction, and mix by pipetting up and down six to eight
times. Incubate at room temperature for 5 min and then pellet
beads with a magnetic rack for 2–3 min or until the solution has
cleared. Remove the supernatant and wash 2 with 170 μL of
70% ethanol. Dry beads at room temperature for ~5 min (It is
important to remove all traces of ethanol from the beads but
not overdry the beads as both the situations will result in
reduced yield). Add 14 μl Qiagen Buffer EB and incubate for
5 min. Transfer 13 μl of the supernatant to a new tube.
 HydEn-seq Mapping 333

Fig. 1 HydEn-seq library construction and amplification. The individual steps of library preparation, from
alkaline hydrolysis of genomic DNA through final amplification, are presented as a flowchart. Complementary
sequences are indicated by light and dark shades of the same color
334 Clinton D. Orebaugh et al.

3.4 Adapter Ligation 1. Assemble the ligation reaction without DNA or enzyme:
to the 50 End (a) 5 μL T4-RNA ligase 10 buffer
(b) 25 μL 50% PEG 8000
(c) 5 μL CoCl3(NH3)6 (10 mM)
(d) 0.5 μL ATP (2 mM)
(e) 0.5 μL ARC 140 oligonucleotide (100 μM) (Fig. 1,
Step 3) (Table 1).
2. Denature 13 μL of DNA eluate from the previous step at 85 ˚C
for 3 min, and then cool on ice.
3. Add 36 μL of reaction mix to denatured DNA and mix by
pipetting up and down six to eight times.
4. Add 1 μL of T4 RNA ligase to each reaction and mix by
pipetting up and down six to eight times.
5. Allow the reaction to incubate overnight at room temperature.
6. Purify ligation product with two magnetic bead washes. For
the first, add 45 μL of magnetic beads, wash twice with 70%
ethanol, and elute with 20 μL of buffer EB. For the second, add
36 μL of magnetic beads, incubate and wash as described in the
previous step, and elute with 14 μL of buffer EB.

3.5 Second Strand 1. Assemble the second strand synthesis reaction without DNA or
Synthesis enzyme:
(a) 2 μL 10 T7 DNA polymerase buffer
(b) 2 μL dNTP mix (2 mM)
(c) 0.8 μL BSA (1 mg/mL)
(d) 2 μL ARC 76/ARC 77 annealed oligonucleotides
(10 μM) (Fig. 1, Step 4), (Table 1).
2. Denature 13 μL of DNA eluate from the previous step at 85 ˚C
for 3 min, and then cool on ice.
3. Add 6.8 μL of the reaction mix to denatured DNA and mix by
pipetting up and down six to eight times.
4. Incubate DNA and reaction mix at room temperature for 5 min
and then add 0.4 μL of T7 DNA polymerase, mix by pipetting
up and down six to eight times with a 10 μL pipette.
5. Incubate the reaction at room temperature for 5 min and then
stop the reaction with the addition of 16 μL of magnetic beads,
mix by pipetting up and down six to eight times.
6. Incubate and wash beads as described in the previous step, and
elute with 11 μL of buffer EB.

3.6 Library 1. Assemble the library amplification reaction without DNA or

Amplification barcoded oligonucleotide:
Table 1
HydEn-seq oligonucleotides

Oligo name Sequence (50 –30 )

ARC 49 AATGATACGGCGACCACCGAGATCTACACTCTTTCCCTACACGACGCTCTTCCGATCT
ARC 76 GTGACTGGAGTTCAGACGTGTGCTCTTCCGATCTNNNN*N*N
ARC 77 AGATCGGAAGAGCACACGTCTGAACTCCAGTC*A*C
ARC 78 CAAGCAGAAGACGGCATACGAGATCGTGATGTGACTGGAGTTCAGACGTGTGCTCTTCCGATCT
ARC 84 CAAGCAGAAGACGGCATACGAGATACATCGGTGACTGGAGTTCAGACGTGTGCTCTTCCGATCT
ARC 85 CAAGCAGAAGACGGCATACGAGATGCCTAAGTGACTGGAGTTCAGACGTGTGCTCTTCCGATCT
ARC 86 CAAGCAGAAGACGGCATACGAGATTGGTCAGTGACTGGAGTTCAGACGTGTGCTCTTCCGATCT
ARC 87 CAAGCAGAAGACGGCATACGAGATCACTGTGTGACTGGAGTTCAGACGTGTGCTCTTCCGATCT
ARC 88 CAAGCAGAAGACGGCATACGAGATATTGGCGTGACTGGAGTTCAGACGTGTGCTCTTCCGATCT
ARC 89 CAAGCAGAAGACGGCATACGAGATGATCTGGTGACTGGAGTTCAGACGTGTGCTCTTCCGATCT
ARC 90 CAAGCAGAAGACGGCATACGAGATTCAAGTGTGACTGGAGTTCAGACGTGTGCTCTTCCGATCT
ARC 91 CAAGCAGAAGACGGCATACGAGATCTGATCGTGACTGGAGTTCAGACGTGTGCTCTTCCGATCT
ARC 93 CAAGCAGAAGACGGCATACGAGATAAGCTAGTGACTGGAGTTCAGACGTGTGCTCTTCCGATCT
ARC 94 CAAGCAGAAGACGGCATACGAGATGTAGCCGTGACTGGAGTTCAGACGTGTGCTCTTCCGATCT
ARC 95 CAAGCAGAAGACGGCATACGAGATTACAAGGTGACTGGAGTTCAGACGTGTGCTCTTCCGATCT
ARC 96 CAAGCAGAAGACGGCATACGAGATTATTGACTGTGACTGGAGTTCAGACGTGTGCTCTTCCGATCT
ARC 97 CAAGCAGAAGACGGCATACGAGATACGGAACTGTGACTGGAGTTCAGACGTGTGCTCTTCCGATCT
ARC 98 CAAGCAGAAGACGGCATACGAGATTCTGACATGTGACTGGAGTTCAGACGTGTGCTCTTCCGATCT
HydEn-seq Mapping

ARC 99 CAAGCAGAAGACGGCATACGAGATCGGGACGGGTGACTGGAGTTCAGACGTGTGCTCTTCCGATCT
ARC 100 CAAGCAGAAGACGGCATACGAGATGTGCGGACGTGACTGGAGTTCAGACGTGTGCTCTTCCGATCT
335

(continued)
Table 1
336

(continued)

Oligo name Sequence (50 –30 )

ARC 101 CAAGCAGAAGACGGCATACGAGATCGTTTCACGTGACTGGAGTTCAGACGTGTGCTCTTCCGATCT
ARC 102 CAAGCAGAAGACGGCATACGAGATAAGGCCACGTGACTGGAGTTCAGACGTGTGCTCTTCCGATCT
ARC 103 CAAGCAGAAGACGGCATACGAGATTCCGAAACGTGACTGGAGTTCAGACGTGTGCTCTTCCGATCT
ARC 104 CAAGCAGAAGACGGCATACGAGATTACGTACGGTGACTGGAGTTCAGACGTGTGCTCTTCCGATCT
ARC 105 CAAGCAGAAGACGGCATACGAGATATCCACTCGTGACTGGAGTTCAGACGTGTGCTCTTCCGATCT
Clinton D. Orebaugh et al.

ARC 106 CAAGCAGAAGACGGCATACGAGATATATCAGTGTGACTGGAGTTCAGACGTGTGCTCTTCCGATCT

ARC 107 CAAGCAGAAGACGGCATACGAGATAAAGGAATGTGACTGGAGTTCAGACGTGTGCTCTTCCGATCT
ARC 140 /5AmMC6/ACACTCTTTCCCTACACGACGCTCTTCCGATCT
Oligonucleotides used for HydEn-seq library construction and amplification. Index sequences are in bold for oligonucleotides ARC (78, 84–91, 93–107). Phosphorothioate
bonds, which provide resistance to nuclease degradation, are indicated by (asterisk). 50 Amino modifier C6, which prevents formation of oligonucleotide concatamers during
adapter ligation, is indicated by (/5AmMC6/). Oligos ARC 76 and ARC 77 are annealed and gel purified prior to use
 HydEn-seq Mapping 337

(a) 25 μL KAPA HiFi HotStart DNA polymerase ReadyMix

(2)
(b) 7.5 μL ARC 49 oligonucleotide (2 μM) (Fig. 1, Step 5)
(Table 1).
2. Add 32.5 μL of reaction mix to 10 μL of DNA eluate from the
previous step.
3. Add 7.5 μL of barcoded oligonucleotide (ARC 78, 84–91,
93–107), (see Note 3) to the mixture and mix by pipetting up
and down six to eight times.
4. Run the library amplification reaction in a PCR thermal cycler
using the following conditions: 98 ˚C hot start for 45 s, then
repeat the following for a total of 18 cycles (melt DNA at 98 ˚C
for 15 s, anneal primers at 65 ˚C for 30 s, and extend primers at
72 ˚C for 30 s), followed by a final extension at 72 ˚C for 60 s,
cool reaction to 4 ˚C and hold.
5. Purify amplified library by adding 40 μL of magnetic beads,
incubate and wash as described in the previous step, and elute
in 20 μL buffer TE.

3.7 Quality Check 1. Measure the concentration of each library using a fluorometric
dsDNA assay. The typical range of concentrations is 1–100 ng/
μL, although lower or higher concentrations may still work.
2. Run 1 μL of each library on an Agilent Bioanalyzer using the
DNA 7500 or DNA 1000 kit according to the manufacturer’s
instructions to determine average fragment size.
3. Determine the molarity of each library by dividing the concen-
tration of each library (g/L) by the average fragment molecular
weight (327 g/nucleotide  fragment length in nucleotides).
4. Pool libraries in equimolar amounts for analysis on an Illumina
HiSeq2500 high-throughput DNA sequencer.

3.8 Sequencing 1. Run pooled libraries on a single lane on an Illumina

HiSeq2500 high-throughput DNA sequencer using paired-
end mode. Save resultant FASTQ files.

3.9 Alignment 1. Trim appropriate adapter sequence from the 30 ends of End1
and Visualization and End2 reads using cutadapt, excluding both mates when
either is trimmed shorter than 15 nt, and utilizing quality
trimming to enhance ability to recognize adapter when it is
present (m 15 -q 10).
2. Concatenate the sequences of all oligos used during library
preparation in a FASTA format, and use this file to construct
a bowtie1 index (bowtie-build).
3. Align trimmed End1 reads to the oligo index using bowtie1,
allowing two mismatches. Utilize –un option to have
338 Clinton D. Orebaugh et al.

unmapped reads written to a separate file. Single- rather than

paired-end alignment is utilized here to achieve a higher align-
ment rate, resulting in a more stringent filter.
4. Extract the mates of all End1 reads passed to the “unhit” file
from the trimmed End2 FASTQ by matching read ID.
5. Align paired “unhit” FASTQs to an appropriate reference
genome using bowtie1, allowing two mismatches and a frag-
ment size of up to 10 kb, retaining unique alignments only,
passing all unmappable and multi-mapped reads to separate
files, and trimming one additional nucleotide from the 30 end
(m1 -v2 –un –max 3 1 –best -X10000). Utilizing the –best
option ensures multi-mapped reads are not erroneously cate-
gorized as unmapped, and 30 trimming allows the successful
alignment of 100% complementary pairs.
6. Perform single-end alignment of remaining unmapped End1
reads to reference genome, allowing two mismatches, retaining
only unique alignments, and passing unmappable and multi-
mapped reads to separate files, as before (m1 -v2 –un –max).
This is included to recover the 50 ends of fragments rendered
unmappable by errors in the End2 read alone.
7. Extract the 50 mapping position of all End 1 reads for align-
ments performed in steps 5 and 6. Shift positions 1 nt
upstream (subtract 1 from coordinate for plus strand, add 1
for minus strand) and determine total read count per position,
per strand, genome-wide. From this information, generate
bedGraph or bigWig files for visualization on the UCSC or
other genome browsers. These counts will be utilized for all
downstream analyses other than normalization.
8. Perform paired-end alignment of multi-mapped reads from
step 5, and single-end alignment of multi-mapped reads from
step 6, reporting the single best alignment, but otherwise
utilizing identical parameters.
9. Extract the 50 mapping positions, shift upstream, and deter-
mine counts genome-wide as in step 7. Combine with counts
derived from unique alignments, and use resulting values when
determining telomere-based normalization factors.

3.10 Generating 1. For a list of strand-specific genomic features of interest, select

Heatmaps/Metagene an appropriate single-nucleotide “anchor” (e.g., Transcription
Plots start sites (TSS), start of Autonomously Replicating Sequence
Consensus Sequences (ACS)).
2. Determine start and end coordinates of desired number of
equally sized bins over a fixed size region relative to each
anchor point, for example, a 1000 to þ999 window, sepa-
rated into 40 bins of 50 nt each. Bins should be arranged in a
strand-specific manner, such that bin 1 of a plus strand feature
 HydEn-seq Mapping 339

should be positioned at anchor 1000 to anchor 951, and

bin 1 of a minus strand feature should be positioned at
anchor þ951 to anchor þ1000.
3. For each bin, determine the total counts of intersecting
HydEn-seq 50 ends, and arrange these values in a matrix, with
each row representing a feature, and each column consisting of
all bins with the same position relative to the anchor. Separate
matrices may be generated for same- and opposite-strand reads,
and this may be performed using raw or normalized counts.
4. Heatmaps may be generated from these matrices using Partek
Genomics Suite, or other appropriate visualization tools
(Fig. 2a).
5. Perform Meta-analyses by summing or averaging the counts in
each column, and plotting these values, utilizing the bin center
as the x-coordinate (Fig. 2b).
6. An additional context may be provided to meta-analyses with a
paired plot of nucleotide frequencies. Extract the strand-
specific genomic sequence, and identify the total count of A,
T, G, and C for each bin. Arrange these counts in a matrix
similar to those described above, determine the aggregate fre-
quency of each nucleotide per column, and plot as in step 5.

3.11 Assessing 1. Identify start and end coordinates of genomic regions of inter-
Ribonucleotide Base est, for example, those determined to have the highest strand
Identity bias for a given variant replicase.
2. Extract the strand-specific genomic sequence for the selected
regions. Determine the total counts of A, T, G, and C to
provide genomic background frequencies.
3. For each position examined, determine the number of HydEn-
seq 50 ends mapped to the same location and strand. Add each
value to the total for the corresponding ribonucleotide.
4. Ribonucleotide frequencies may be visualized and compared to
genomic background and across HydEn-seq data sets by plot-
ting as stacked column/bar charts.

3.12 Normalization 1. Define telomeric regions as sections of the reference genome

Versus Chromosomal that consist of only telomeric repeats and at least one read
50 Ends length from no non-telomeric sequence. The total length of
these regions is
PC
2N  
L telomere ¼ max 0; L i, telomere  L read ,
i¼1

where NC is the chromosome count (16 in S. cerevisiae).

2. Count single- and multi-mapped reads (from Subheading 3.9,
step 8) whose 50 -ends (shifted as per Subheading 3.9, step 9)
map within telomeric regions (Ntelomere) and single-mapped
reads outside of telomeric regions (Nbulk).
340 Clinton D. Orebaugh et al.

Fig. 2 Representative stacked ACS Intensity Plot and ACS Average Meta-Analysis. (a) Intensity plots show
mapped ribonucleotides/million reads 2 kilobases upstream and downstream of 200 vertically stacked
 HydEn-seq Mapping 341

3. The HydEn-seq end density, excluding telomeric regions is

Rbulk ¼ 2L genome
N bulk
L telomere ,

where Lgenome is the genome length (~12 Mbp in S. cerevisiae).

4. The mean number of 50 chromosome ends per telomere is
N telomere ¼ N telomere L telomere R bulk
2N C :
5. Division by N telomere converts end counts per bin into end
counts per position per genome. The inverse of this yields the
estimate of the mean fragment size within the bin. For exam-
ple, the overall mean fragment size (L fragment ), where the bin is
the whole genome, is
N telomere
L fragment ¼ Rbulk ,
and the mean number of ends per genome (Nends) is
2L genome
N ends ¼ L fragment :

3.13 Predicting 1. Calculate the fraction of replication events (F) in which the
Replication Origins bottom strand (b) is replicated as the leading strand in genomic
from HydEn-Seq Maps bin i. This requires two HydEn-seq data sets with opposite
strand biases (lead and lag below). As per [22],

1
F i lead, b ¼ qffiffiffiffiffiffiffiffiffiffiffiffi
O i lag, b
1þ O i lead, b

where
m ij , b
O ij , b ¼
m ij , t
and where mij , k is the normalized (Subheading 3.12) and
background-subtracted (specific to chosen strains) end count
in the k-strand (bottom or top) in the j-biased data set. Esti-
mates improve when using data sets with more extreme strand
biases and when using an average of normalized counts (m ij )
from replicate experiments (Fig. 3).

Fig. 2 (continued) Autonomously Replicating Sequence Consensus Sequences (ACS). Reads mapping to the
right of the ACS on the top strand and to the left of the ACS on the bottom strand are on the leading strand.
Reads mapping to the left of the ACS on the top strand and to the right of the ACS on the bottom strand are on
the lagging strand. The pol2-M644G rnh201Δ strain exhibits a bias toward ribonucleotides mapped to the
leading strand. The pol3-L612G rnh201Δ exhibits a bias toward ribonucleotides mapped to the lagging strand.
Wild-type replicase backgrounds are presented with (RNH201) and without (rnh201Δ) RER. Images were
generated with Partek Genomics Suite. (b) Meta-analysis of mapped ribonucleotides/million reads 2 kilobases
upstream and downstream of 200 ACS. The red line is the top strand and the blue line is the bottom strand
342 Clinton D. Orebaugh et al.

A mi ε-MG,b and mi ε-MG,t mi δ-LM,b and mi δ-LM,t

0.001 0.001

(percent of ChrIII counts)

0.10

HydeEn-seq end counts

0.0008
0.08 0.0008

0.0006
0.06 0.0006

0.0004
0.04 0.0004

0.0002
0.02 0.0002

0
0.00 0
0 60 80 100 120 60 80 100 120
Chromosome III position (kbp)

B Oi j,t = mi j,t /mi j,b

Log10Oi ε-MG,t Log10Oi δ-LM,t
33 3
observed strand bias

00 0

-3
-3 -3
0 60 80 100 120 60 80 100 120
Chromosome III position (kbp)

C Ri lag,b = (Oi lead,t /Oi lag,t)0.5

Log10Ri δ-LM,b
33
replication asymmetry
strand bias due to

00

-3
-3
0 60 80 100 120
Chromosome III position (kbp)

D Fi lag,b = 1/(1+Ri lag,b-1)

Fi δ-LM,b 306URA3 Ty2 307
1.01
nascent lagging strand
strands replicated as
fraction of bottom

0.5
0.5

0.00
0 60 80 100 120
Chromosome III position (kbp)

Fig. 3 Calculating the fraction of replication events in which the bottom strand is replicated as the leading
strand. These panels illustrate the data transformations in Subheading 3.13, step 1. (a) Normalized
 HydEn-seq Mapping 343

2. Predicted origins were defined as regions where Fi lead , b

changed abruptly. Origin calling parameters were set based on
the results from a training set of confirmed Saccharomyces
cerevisiae replication origins and data sets from modified yeast
strains (200 bp bins; smoothed over 9 bins) [21]. As such,
individual parameters may be more or less suitable for other
data sets and especially for other taxa, but general procedures
should be widely applicable. In the original yeast experiments,
origins were defined as positions where either the average slope
(the derivative) of Fi lead , b exceeded 0.00011 fractional units
per bp in an 11-bin window (2.2 kbp) or 0.00016 fractional
units per base pair in at least three of five surrounding bins
(600 bp out of 1 kbp). Based on the training set, these
parameters were selected to define a sufficiently abrupt change
over a region wide enough to exclude random noise.

4 Notes

1. The RNase a treatment step is omitted from genomic DNA

purification to avoid contaminant nuclease activity in the sup-
plied RNase A. The alkali treatment sufficiently degrades any
residual RNA.
2. While the enzymatic steps are performed in PCR tubes or
plates, we recommend transferring the reactions to 96-well
culture plates for incubation and capturing the magnetic
beads with a magnetic ring plate. This will spread the beads
over a larger surface area and they will dry faster. We recom-
mend drying at 30  C, but any temperature between 23  C and
37  C will work. It is important not to over dry the beads, as
this will reduce yield.
3. Use a unique barcode sequence for each library that will be run
on the same lane of the sequencer. We run as many as 24
libraries on each lane, often with six biological replicates of
each of two isolates of the same genotype. In this fashion we
can generate 12 individual measurements of each of two

Fig. 3 (continued) (Subheading 3.12) end counts (mij , k), by strand (bottom ¼ b, orange; top ¼ t, black), for
two data sets with opposite strand biases (here, leading ¼ ε-MG, from pol2-M644G rnh201Δ yeast;
lagging ¼ δ-LM, from pol3-L612 M rnh201Δ yeast). (b) Apparent strand biases (Oij , k) are calculated for
each data set. (c) The apparent biases are combined, removing correlated noise and leaving an estimate of
true bias (Rij , k). (d) This is used to estimate the fraction of replication events in which the bottom strand is
replicated as the leading strand (Fi lead , b). The abrupt transitions used to predict origins (Subheading 3.13,
step 2) are obvious at origins ARS306 and ARS307 (orange). Predictions may be confounded where strains
differ (e.g., URA3 orientation at the AGP1 locus; blue) and at certain other features (transposon Ty2; red). End
counts averaged over multiple data sets (m ij ) may be used
344 Clinton D. Orebaugh et al.

genotypes on a single run. This is especially useful for establish-

ing statistically significant differences between genotypes.

Acknowledgment

We thank Dr. Jessica Williams and Dr. Kin Chan for helpful com-
ments on the manuscript. This work was supported by the Division
of Intramural Research of the US National Institutes of Health
(NIH), National Institute of Environmental Health Sciences (proj-
ect Z01 ES065070 to T.A.K.).

References
1. Nick McElhinny SA, Watts BE, Kumar D, Watt 1016/j.molcel.2013.03.017. S1097-2765
DL, Lundstrom EB, Burgers PM, Johansson (13)00221-9 [pii]
E, Chabes A, Kunkel TA (2010) Abundant 8. Ghodgaonkar MM, Lazzaro F, Olivera-
ribonucleotide incorporation into DNA by Pimentel M, Artola-Boran M, Cejka P, Reijns
yeast replicative polymerases. Proc Natl Acad MA, Jackson AP, Plevani P, Muzi-Falconi M,
Sci U S A 107(11):4949–4954. doi:10.1073/ Jiricny J (2013) Ribonucleotides misincorpo-
pnas.0914857107. 0914857107 [pii] rated into DNA act as strand-discrimination
2. Clausen AR, Zhang S, Burgers PM, Lee MY, signals in eukaryotic mismatch repair. Mol
Kunkel TA (2013) Ribonucleotide incorpora- Cell 50(3):323–332. doi:10.1016/j.molcel.
tion, proofreading and bypass by human DNA 2013.03.019. S1097-2765(13)00223-2 [pii]
polymerase δ. DNA Repair 12(2):121–127. 9. Vengrova S, Dalgaard JZ (2006) The wild-type
doi:10.1016/j.dnarep.2012.11.006. S1568- Schizosaccharomyces pombe mat1 imprint con-
7864(12)00285-6 [pii] sists of two ribonucleotides. EMBO Rep 7
3. Goksenin AY, Zahurancik W, LeCompte KG, (1):59–65. doi:10.1038/sj.embor.7400576.
Taggart DJ, Suo Z, Pursell ZF (2012) Human 7400576 [pii]
DNA polymerase epsilon is able to efficiently 10. Caldecott KW (2014) Molecular biology.
extend from multiple consecutive ribonucleo- Ribose–an internal threat to DNA. Science
tides. J Biol Chem 287(51):42675–42684. 343(6168):260–261. doi:10.1126/science.
doi:10.1074/jbc.M112.422733 1248234
4. Williams JS, Lujan SA, Kunkel TA (2016) Pro- 11. Pizzi S, Sertic S, Orcesi S, Cereda C, Bianchi
cessing ribonucleotides incorporated during M, Jackson AP, Lazzaro F, Plevani P, Muzi-
eukaryotic DNA replication. Nat Rev Mol Falconi M (2015) Reduction of hRNase H2
Cell Biol 17(6):350–363. doi:10.1038/nrm. activity in Aicardi-Goutieres syndrome cells
2016.37 leads to replication stress and genome instabil-
5. Nick McElhinny SA, Kumar D, Clark AB, Watt ity. Hum Mol Genet 24(3):649–658. doi:10.
DL, Watts BE, Lundstrom EB, Johansson E, 1093/hmg/ddu485
Chabes A, Kunkel TA (2010) Genome instabil- 12. Reijns MA, Rabe B, Rigby RE, Mill P, Astell
ity due to ribonucleotide incorporation into KR, Lettice LA, Boyle S, Leitch A, Keighren M,
DNA. Nat Chem Biol 6(10):774–781. Kilanowski F, Devenney PS, Sexton D, Grimes
doi:10.1038/nchembio.424. nchembio.424 G, Holt IJ, Hill RE, Taylor MS, Lawson KA,
[pii] Dorin JR, Jackson AP (2012) Enzymatic
6. Sparks JL, Chon H, Cerritelli SM, Kunkel TA, removal of ribonucleotides from DNA is essen-
Johansson E, Crouch RJ, Burgers PM (2012) tial for mammalian genome integrity and devel-
RNase H2-initiated ribonucleotide excision opment. Cell 149(5):1008–1022. doi:10.
repair. Mol Cell 47(6):980–986. doi:10. 1016/j.cell.2012.04.011. S0092-8674(12)
1016/j.molcel.2012.06.035. S1097-2765 00508-9 [pii]
(12)00599-0 [pii] 13. Coffin SR, Hollis T, Perrino FW (2011) Func-
7. Lujan SA, Williams JS, Clausen AR, Clark AB, tional consequences of the RNase H2A subunit
Kunkel TA (2013) Ribonucleotides are signals mutations that cause Aicardi-Goutieres syn-
for mismatch repair of leading-strand replica- drome. J Biol Chem 286(19):16984–16991.
tion errors. Mol Cell 50(3):437–443. doi:10. doi:10.1074/jbc.M111.228833
 HydEn-seq Mapping 345

14. Shaban NM, Harvey S, Perrino FW, Hollis T 1-mediated removal of ribonucleotides from
(2010) The structure of the mammalian RNase nascent leading-strand DNA. Mol Cell 49
H2 complex provides insight into RNA.NA (5):1010–1015. doi:10.1016/j.molcel.2012.
hybrid processing to prevent immune dysfunc- 12.021. S1097-2765(13)00033-6 [pii]
tion. J Biol Chem 285(6):3617–3624. doi:10. 19. Lipkin D, Talbert PT, Cohn M (1954) The
1074/jbc.M109.059048 mechanism of the alkaline hydrolysis of ribonu-
15. Crow YJ, Leitch A, Hayward BE, Garner A, cleic acids. J Am Chem Soc 76
Parmar R, Griffith E, Ali M, Semple C, Aicardi (11):2871–2872. doi:10.1021/Ja01640a004
J, Babul-Hirji R, Baumann C, Baxter P, Bertini 20. Li YF, Breaker RR (1999) Kinetics of RNA
E, Chandler KE, Chitayat D, Cau D, Dery C, degradation by specific base catalysis of transes-
Fazzi E, Goizet C, King MD, Klepper J, terification involving the 2’-hydroxyl group. J
Lacombe D, Lanzi G, Lyall H, Martinez-Frias Am Chem Soc 121(23):5364–5372. doi:10.
ML, Mathieu M, McKeown C, Monier A, 1021/Ja990592p
Oade Y, Quarrell OW, Rittey CD, Rogers RC, 21. Clausen AR, Lujan SA, Burkholder AB, Ore-
Sanchis A, Stephenson JB, Tacke U, Till M, baugh CD, Williams JS, Clausen MF, Malc EP,
Tolmie JL, Tomlin P, Voit T, Weschke B, Mieczkowski PA, Fargo DC, Smith DJ, Kunkel
Woods CG, Lebon P, Bonthron DT, Ponting TA (2015) Tracking replication enzymology
CP, Jackson AP (2006) Mutations in genes in vivo by genome-wide mapping of ribonucle-
encoding ribonuclease H2 subunits cause otide incorporation. Nat Struct Mol Biol 22
Aicardi-Goutieres syndrome and mimic con- (3):185–191. doi:10.1038/nsmb.2957.
genital viral brain infection. Nat Genet 38 nsmb.2957 [pii]
(8):910–916. doi:10.1038/ng1842
22. Williams JS, Clausen AR, Lujan SA, Marjavaara
16. Lujan SA, Williams JS, Pursell ZF, Abdulovic- L, Clark AB, Burgers PM, Chabes A, Kunkel
Cui AA, Clark AB, Nick McElhinny SA, Kunkel TA (2015) Evidence that processing of ribonu-
TA (2012) Mismatch repair balances leading cleotides in DNA by topoisomerase 1 is
and lagging strand DNA replication fidelity. leading-strand specific. Nat Struct Mol Biol
PLoS Genet 8(10):e1003016. doi:10.1371/ 22(4):291–297. doi:10.1038/nsmb.2989.
journal.pgen.1003016. PGENETICS-D-12- nsmb.2989 [pii]
00710 [pii]
23. Koh KD, Balachander S, Hesselberth JR, Stor-
17. Miyabe I, Kunkel TA, Carr AM (2011) The ici F (2015) Ribose-seq: global mapping of
major roles of DNA polymerases epsilon and ribonucleotides embedded in genomic DNA.
delta at the eukaryotic replication fork are evo- Nat Methods 12(3):251–257. 253 p following
lutionarily conserved. PLoS Genet 7(12): 257. doi: 10.1038/nmeth.3259, nmeth.3259
e1002407. doi:10.1371/journal.pgen. [pii]
1002407. PGENETICS-D-11-01459 [pii]
18. Williams JS, Smith DJ, Marjavaara L, Lujan SA,
Chabes A, Kunkel TA (2013) Topoisomerase
 Chapter 24

Detection of DNA-RNA Hybrids In Vivo

Marı́a Garcı́a-Rubio, Sonia I. Barroso, and Andrés Aguilera

Abstract
DNA-RNA hybrids form naturally during essential cellular functions such as transcription and replication.
However, they may be an important source of genome instability, a hallmark of cancer and genetic diseases.
Detection of DNA-RNA hybrids in cells is becoming crucial to understand an increasing number of
molecular biology processes in genome dynamics and function and to identify new factors and mechanisms
responsible for disease in biomedical research. Here, we describe two different procedures for the reliable
detection of DNA-RNA hybrids in the yeast Saccharomyces cerevisiae and in human cells: DNA-RNA
Immunoprecipitation (DRIP) and Immunofluorescence.

Key words DNA-RNA hybrids, R loop, Genome instability, S9.6 antibody

1 Introduction

DNA-RNA hybrids are highly stable structures formed during

essential cellular functions such as transcription and replication.
Hybrids covering few nucleotides (around 10) are essential for
Okazaki fragment synthesis during DNA replication as well as for
RNA synthesis by RNA polymerases. In a different context, how-
ever, an RNA molecule may invade a double-stranded DNA
(dsDNA) to form a DNA-RNA hybrid and the displaced DNA
strand. These structures are known as R loops.
R loops have been shown to have a physiological role in initia-
tion of replication of mitochondrial DNA or some bacterial plas-
mids, as well as in Immunoglobulin class switch recombination, a
vital process for the generation of the genetic diversity that the
defense systems of vertebrates require, as reviewed in [1]. In these
cases, DNA-RNA hybrids may expand from tens of nucleotides to a
few hundreds, the S regions involved in Ig class switching having
been reported to form hybrids that can exceed even 1 kb in size [2].
However, since a first report showing that yeast cells lacking the

Marı́a Garcı́a-Rubio and Sonia I. Barroso contributed equally to this work.

Marco Muzi-Falconi and Grant W. Brown (eds.), Genome Instability: Methods and Protocols, Methods in Molecular Biology,
vol. 1672, DOI 10.1007/978-1-4939-7306-4_24, © Springer Science+Business Media LLC 2018

347
348 Marı́a Garcı́a-Rubio et al.

RNA biogenesis and export factor THO accumulate R loops

responsible for high levels of transcription-dependent genome
instability [3], different contributions have provided evidence that
DNA-RNA hybrids or R loops can form at higher levels than
previously foreseen, becoming an important threat to genome
integrity [1]. The increasing abundance of factors identified to be
involved in prevention or resolution of DNA-RNA hybrids have
been shown to map all over the genome with particular hotspot
regions, and different reports have demonstrated that they may play
positive and negative roles in transcription. Importantly, however,
it is becoming evident that cells have a large number of proteins
involved in the prevention or resolution of R loops.
In most physiological cases, so far, it has not been either
reported or explored whether DNA-RNA hybrids require in addi-
tion the participation of specific proteins for their formation or
stability. In this sense, we know that the action of the CRISPR-
Cas9 system of genome editing is based on the formation of a
DNA-RNA hybrid indeed [4]. Thus, the possibility that R loops
may be facilitated or stabilized by specific proteins [5] needs to be
further explore in the future. However, any in vivo study on R
loops, whether naturally formed or artificially induced, requires to
be completed with the in vivo detection of DNA-RNA hybrids in
cells. This is so far a bottle neck step in all studies. Even though
there have been several methods used in different studies, the most
extended ones are based on the use of the monoclonal anti-DNA-
RNA hybrid antibody of the S9.6 hybridoma [2, 3, 6–13]. Here,
we provide the step-by-step protocol of the two more-extended
techniques used with this antibody, DNA-RNA immunoprecipita-
tion (DRIP) [14] and immunofluoresecence (IF) analyses to detect
hybrids both in yeast and in human cells [15, 16]. As can be seen, it
is essential in any DNA-RNA hybrid detection techniques to vali-
date that the signals detected correspond unequivocally to DNA-
RNA hybrids by eliminating them by treatment with RNase H
in vitro or by overexpressing RNase H in vivo.

2 Materials

2.1 DNA-RNA 1. Cell cultures.

Inmunoprecipitation 2. 10 cm plates.
(DRIP) in Human Cells
3. PBS Tablets (GIBCO).
4. 1 TE buffer: 10 mM Tris–HCl pH 7.6, EDTA 1 mM pH 8.0.
5. 20% SDS.
6. Proteinase K, recombinant (20 mg/ml, Roche).
7. Phenol:Chloroform:Isoamyl Alcohol 25:24:1 Saturated with
10 mM Tris-HCl, pH 8.0, 0.1 mM EDTA.
 R Loop Detection 349

8. Phase Lock Gel Heavy 2 ml tubes (% Prime GmbH).

9. 3 M Sodium Acetate.
10. Isopropanol.
11. Glass rod.
12. 70% Ethanol.
13. Restriction enzymes (HindIII 2.5 μl, EcoRI 2.5 μl, XbaI 2.5 μl,
SspI 10 μl, BsrGI 5 μl), BSA 2.5 μl and NEB Buffer 2 (New
England BioLabs).
14. StrataClean Resin (Stratagene).
15. 50T5E buffer: 50 mM Tris–HCl pH 7.5, 5 mM EDTA pH 8.0.
16. 50% v/v slurry of Sephadex G-50 (Sigma) in 50T5E buffer
(kept at 4  C).
17. Empty micro Bio-spin chromatography columns (Biorad).
18. RNase H and appropriate 10 buffer (New England BioLabs).
19. Fluorometer apparatus and appropriate cuvettes.
20. Appropriate DNA fragments.
21. anti-DNA-RNA hybrid antibody (S9.6).
22. 1 M Na-Phosphate pH 7.0: 39 ml 2 M monobasic sodium
phosphate NaH2PO4, 61 ml 2 M dibasic sodium phosphate
Na2HPO4, 100 ml H2O.
23. 10 Binding Buffer: 100 mM NaPO4 pH 7.0, 1.4 M NaCl,
0.5% Triton X-100, stored at room temperature.
24. Elution Buffer: 50 mM Tris pH 8.0, 10 mM EDTA, 0.5% SDS.
25. Protein A-coated magnetic beads (Dynabeads, Novex by Life
Technologies).
26. Magnetic rack (Life Technologies).
27. Proteinase K, recombinant (20 mg/ml, Roche).
28. Rotator mixer.
29. Thermomixer.
30. PCR Purification Kit (Quiagen).
31. 1 TE buffer: 10 mM Tris–HCl pH 7.6, EDTA 1 mM pH 8.0.
32. Power SYBR green PCR master mix (Applied Biosystems).
33. Appropriate oligonucleotides (see Note 10).
34. Fast Real-Time PCR System.

2.2 DNA-RNA 1. Cell cultures.

Immunoprecipitation 2. Zymolyase 20 T (USB).
in Yeast Cultures
3. Spheroplasting Buffer: 1 M sorbitol, 2 mM Tris–HCl pH 8.0,
100 mM EDTA pH 8.0, 0.1% v/v beta-mercapto-ethanol.
350 Marı́a Garcı́a-Rubio et al.

4. Solution I: 0.8 mM GuHCl, 30 mM Tris–HCl pH 8.0, 30 mM

EDTA pH 8.0, 5% Tween 20, 0.5% Triton X-100.
5. RNase A stock solution: 10 mg/ml (Roche; see Note 11).
6. Proteinase K, recombinant (20 mg/ml, Roche).
7. Chloroform-Isoamyl alcohol (24:1).
8. Isopropanol (at room temperature).
9. 70% Ethanol (at room temperature).
10. Restriction enzymes (HindIII 2.5 μl, EcoRI 2.5 μl, XbaI 2.5 μl,
SspI 10 μl, BsrGI 5 μl), BSA 2.5 μl and NEB Buffer 2 (New
England BioLabs).
11. StrataClean Resin (Stratagene).
12. 50T5E buffer: 50 mM Tris–HCl pH 7.5, 5 mM EDTA pH 8.0.
13. 50% v/v slurry of Sephadex G-50 (Sigma) in 50T5E buffer
(kept at 4  C).
14. Empty micro Bio-spin chromatography columns (Biorad).
15. RNase H and appropriate 10 buffer (New England BioLabs).
16. Fluorometer apparatus and appropriate cuvettes.
17. Appropriate DNA fragments.
18. anti-DNA-RNA hybrid antibody (S9.6).
19. 1 M Na-Phosphate pH 7.0 (39 ml 2 M monobasic sodium
phosphate NaH2PO4, 61 ml 2 M dibasic sodium phosphate
Na2HPO4, 100 ml H2O).
20. 10 Binding Buffer (100 mM NaPO4 pH 7.0, 1.4 M NaCl,
0.5% Triton X-100, stored at room temperature).
21. Elution Buffer (50 mM Tris pH 8.0, 10 mM EDTA, 0.5%
SDS).
22. Protein A-coated magnetic beads (Dynabeads, Novex by Life
Technologies).
23. Magnetic rack (Life Technologies).
24. Proteinase K, recombinant (20 mg/ml, Roche).
25. Rotator mixer.
26. Thermomixer.
27. PCR Purification Kit (Quiagen).
28. 1 TE buffer: 10 mM Tris–HCl pH 7.6, EDTA 1 mM pH 8.0.
29. Power SYBR green PCR master mix (Applied Biosystems).
30. Appropriate oligonucleotides (see Note 10).
31. Fast Real-Time PCR System.
 R Loop Detection 351

2.3 S9.6 1. DMEM-Dulbecco’s Modified Eagle Medium (GIBCO).

Immunofluorescence 2. 24-well plates.
in Mammalian Cells
3. Tweezers.
4. Round Coverslips.
5. Bovine serum albumin (BSA).
6. PBS Tablets (GIBCO).
7. Blocking Solution: 2% BSA in PBS (Phosphate-Buffered
Saline).
8. anti-DNA-RNA hybrid antibody (S9.6).
9. Nucleolin antibody (Abcam).
10. Ice-cold methanol: 20  C 100% methanol (see Note 16).
11. Alexa Fluor 488 goat anti-rabbit (Invitrogen).
12. Alexa Fluor 594 goat anti-mouse (Invitrogen).
13. Vacuum line.
14. Immu-Mount Mounting Medium (Thermo Scientific).
15. Microscope Slides.
16. Confocal microscope.
17. DAPI solution (1 μg/ml in PBS).

2.4 S9.6 1. Yeast cultures.

Immunofluorescence 2. Formaldehyde 37%.
in Yeast
3. Phosphate buffer 0.1 M pH 6.4.
4. Sorbitol-citrate 1.2 M: 218.6 g/l sorbitol, 7 g/l citric acid,
17.418 g/l K2HPO4.
5. Poly L Lysine Treated Slides.
6. Glusulase (Perkin Elmer).
7. Zymoliase 20T (USBiological): 50 mg/ml in sorbitol 1 M.
8. SDS.
9. PBS-BSA Solution: 1% BSA, 0.04 M K2HPO4, 0.01 M
KH2PO4, 0.15 M NaCl, 0.1% NaN3. Filter sterilized and
stored at 4  C.
10. Digestion Mix: 200 μl of 1.2 M sorbitol-citrate, 20 μl of
glusulase, 2 μl of zymolyase 20T.
11. Alexa Fluor 594 goat anti-mouse (Invitrogen).
12. anti-DNA-RNA hybrid antibody (S9.6).
13. PBS Tablets (GIBCO).
14. DAPI.
15. Fluorescence microscope.
352 Marı́a Garcı́a-Rubio et al.

3 Methods

3.1 DNA-RNA 1. Transfer the content of a 10 cm plate of 90% confluent cells

Immunoprecipitation (3  106 cells) to a 15 ml falcon tube and pellet them at
in Human Cells 400  g, 4  C in a swing rotor for 5 min. Remove the
supernatant.
2. Wash and resuspend cells with 10 ml 1 cold PBS and pellet
them again. Remove the supernatant.
3. Resuspend cells in 1.6 ml of TE 1 and split into two tubes
with a cut-off pipette tip (see Note 1). Add 21 μl of 20% SDS,
and 2.5 μl of Proteinase K to each sample (0.8 ml) and invert
the tube a few times. Incubate overnight at 37  C (see Note 2).
4. Pour the DNA from the previous step into a phase lock 2 ml
tube (see Note 3). Add 1 volume (0.8 ml) of Phenol:Chloro-
form:Isoamyl Alcohol, invert gently a few times, spin down
5 min at 13,000  g (see Note 4).
5. Pour the DNA (top aqueous phase) in a new phase lock tube
and repeat step 4.
6. Add 1/10 volume 3 M NaOAc (160 μl) and 1 volume (1.6 ml)
isopropanol to a 15 ml tube and pour in the DNA (top aqueous
phase, rejoin the samples in one tube) from two phase lock
tubes.
7. Invert gently until the DNA begins to precipitate. Spool DNA
on a glass rod (see Note 5).
8. Wash DNA with 70% Ethanol by allowing the EtOH run down
glass rod.
9. Allow to air dry. Break off the tip with DNA and put it in a
1.5 ml tube with 150 μl TE 1. Do not resuspend the DNA by
over-pipeting/vortexing. Mix several times by gentle shaking
(see Note 6).
10. Digest the DNA overnight using cocktail restriction enzymes
according to supplier’s instructions (add BSA 1, final volume
is 250 μl) (see Note 7).
11. Add 3 μl StrataClean Resin, pipet with cut tips, and spin down
at full speed in a microcentrifuge for 1 min (see Note 8).
12. Purify over Sephadex g-50 column (in 50T5E buffer) (see
Note 9).
13. Split the flow-through containing DNA into two new tubes
(125 μl).
14. Treat the half of the DNA (one tube) with 5 μl RNase H (NEB)
overnight at 37  C. In the meantime, keep the other half at
4  C.
15. Repeat step 12.
 R Loop Detection 353

16. Check DNA concentration (samples treated and untreated

with RNase H).
17. Take 1 μg of digested DNA (treated and untreated with RNase
H) (INPUT) and process immediately as described in steps 25
and 26.
18. Dilute 5 μg of digested DNA (treated and untreated with
RNase H) in 450 μl TE1.
19. Add 51 μl of 10 Binding Buffer.
20. Add 10 μl of anti-DNA-RNA hybrid antibody (Stock
1 mg/ml).
21. Incubate overnight at 4  C on a rotator mixer.
22. Wash 40 μl of Protein A magnetic beads per DRIP sample,
twice with 1 ml of 1 Binding Buffer (10 diluted in TE1).
Resuspend in the original volume (40 μl).
23. Add the prepared magnetic beads to the extracts (40 μl per
reaction) and incubate for 2 h at 4  C on a rotator mixer.
24. Place the precooled magnetic rack on ice, insert the tubes,
aspirate all liquid off, and resuspend the beads in 1 ml of 1
Binding Buffer. Repeat the same procedure to achieve three
washes with 1 Binding Buffer.
25. Add 120 μl Elution Buffer and 7 μl Proteinase K and incubate
at 55  C for 45 min in a shaker. Dilute 1 μg of INPUT samples
(step 17) in 47 μl of TE1 and add 3 μl of Proteinase K,
incubate at 55  C for 45 min.
26. Purify (INPUT and PRECIPITATE) over PCR purification
columns (e.g., from Quiagen) according to manufacturer’s
protocol and elute in 150 μl TE1. Store samples at 20  C.
27. Quantify the enriched DNA fragments in the PRECIPITATE
by real-time quantitative PCR (qPCR). We perform qPCR
using Fast SYBR Green dye and the absolute quantification
protocol in the 7500 Fast Real-Time PCR System (Applied
Biosystems). Standard curves for all pairs of primers are per-
formed for each analysis. All PCR reactions are performed in
triplicate. The enrichment for each PCR of interest is normal-
ized with respect to the corresponding ratios of the INPUT (see
Fig. 1).

3.2 DNA-RNA 1. Grow 100 ml yeast cultures in the appropriate medium and
Immunoprecipitation temperature to an absorbance at 600 nm of about 0.8.
in Yeast Cultures 2. Harvest cells by centrifugation at 4000  g, 4  C for 5 min.
3. Wash cells twice with 20 ml cold H2O.
4. Harvest cells by centrifugation at 4000  g, 4  C for 5 min.
354 Marı́a Garcı́a-Rubio et al.

Fig. 1 Relative amount of R loops in the patient FANCD2 / human PD20 cell line and the corrected PD20
FANCD2+/+ control at four different genes. Results shown correspond to DRIP-qPCR assays using the S9.6
monoclonal antibody with and without RNase H (RNH) treatment. Signal values of DNA-RNA hybrids
immunoprecipitated in each region are normalized to input values. A.U. arbitrary units. (Reproduced from
[16] with modifications)

5. Resuspend the cell pellet in 2.4 ml Spheroplasting Buffer

(freshly made) (add 2 mg Zymolyase to 2.4 Spheroplasting
Buffer). Split the sample into two tubes (see Note 1).
6. Incubate at 30  C for 30 min with rotation.
7. Spin down for 5 min at 4500  g in a microcentrifuge and
discard the supernatant. Remove all the supernatant.
8. Carefully wash the spheroplast pellet by adding 500 μl of cold
H2O without resuspending.
9. Spin briefly and discard the water.
10. Break the spheroplasts by resuspending in 1.125 ml Solution I
(565 μl per tube, rejoin the samples) and transferring in 2 ml
eppendorf tubes (see Note 12).
11. Add 20 μl RNase A.
12. Incubate for 30 min at 37  C.
13. Add 75 μl Proteinase K (20 mg/ml) and incubate at 50  C for
60 min (see Note 13).
14. Spin down for 5 min at 4500  g and transfer the supernatant
to a new tube.
15. Add 800 μl Chloroform-Isoamyl alcohol.
16. Spin down for 5 min at 4500  g.
17. Carefully collect the upper phase with 1 ml cut tips and transfer
into a 2 ml tube.
 R Loop Detection 355

18. Add 800 μl isopropanol to the aqueous phase and invert several
times until genomic DNA precipitates. Spool DNA on a glass
rod (see Note 5).
19. Wash DNA with 70% Ethanol by allowing the EtOH run down
glass rod.
20. Allow to air dry, break off the tip with DNA, and put it in a
1.5 ml tube with 150 μl TE 1. Do not resuspend the DNA by
over-pipetting/vortexing. Mix several times by gentle shaking
(see Note 6).
21. Digest the DNA overnight using cocktail restriction enzymes
according to supplier’s instructions (add 1 BSA, final volume
is 250 μl) (see Note 7).
22. Add 3 μl StrataClean Resin, pipet with cut tips and spin down
at full speed in a microcentrifuge for 1 min (see Note 8).
23. Purify over Sephadex g-50 column (in 50T5E buffer) (see
Note 9).
24. Split the flow-through containing DNA into two new tubes
(125 μl).
25. Treat half of the DNA (one tube) with 8 μl RNase H (NEB)
overnight at 37  C. In the meantime, keep the other half
at 4  C.
26. Repeat step 12.
27. Take 20 μl of digested DNA (treated and untreated with RNase
H) (INPUT) and process immediately as described in step 35.
28. Dilute digested DNA (treated and untreated with RNase H) in
450 μl TE1.
29. Add 51 μl of 10 Binding Buffer.
30. Add 10 μl of anti-DNA-RNA hybrid antibody (Stock
1 mg/ml).
31. Incubate overnight at 4  C on a rotator mixer.
32. Wash 40 μl of Protein A magnetic beads per DRIP sample,
twice with 1 ml of 1 Binding Buffer (10 diluted in TE1).
Resuspend in the original volume (40 μl).
33. Add the prepared magnetic beads to the extracts (40 μl per
reaction) and incubate for 2 h at 4  C on a rotator mixer.
34. Place the precooled magnetic rack on ice, insert the tubes,
aspirate all liquid off, and resuspend the beads in 1 ml of 1
Binding Buffer. Repeat the same procedure to complete three
washes with 1 Binding Buffer.
35. Add 120 μl Elution Buffer and 7 μl Proteinase K and incubate
at 55  C for 45 min in a shaker. Add 27 μl TE1 to INPUT
356 Marı́a Garcı́a-Rubio et al.

samples (step 27), 3 μl Proteinase K and incubate at 55  C for

45 min.
36. Purify INPUT and PRECIPITATE over PCR purification col-
umns according to manufacturer’s protocol and elute in 150 μl
TE1. Store samples at 20  C.
37. Quantify the enriched DNA fragments in the PRECIPITATE
by real-time quantitative PCR (qPCR). We perform qPCR
using Fast SYBR Green dye and the absolute quantification
protocol in the 7500 Fast Real-Time PCR system (Applied
Biosystems). Standard curves for all pairs of primers are per-
formed for each analysis. All the PCR reactions are performed
in triplicated. The enrichment for each PCR of interest is
normalized with respect to the corresponding ratios of the
INPUT.

3.3 S9.6 1. HeLa cells are cultured on coverslips (see Note 14) at a con-
Immunofluorescence centration of 2  105 cells/well in a 6-well plate.
in Mammalian Cells 2. After 24 h, coverslips are transferred to a 24-well plate well with
cold PBS (see Note 15).
3. Aspirate the PBS with the vacuum system and add 1 ml of 100%
ice-cold methanol (see Note 16). Incubate for 7 min at 20  C.
4. Remove methanol (see Note 16) and wash twice with 1 ml/
well of PBS.
5. Incubate the cells with Blocking Solution (1 ml/well) and
incubate overnight at 4  C.
6. Remove Blocking Solution and add anti-DNA-RNA hybrid
(1:500 in Blocking Solution) and Nucleolin (1:1000 in Block-
ing Solution) antibodies in a total volume of 250 μl/well.
Incubate overnight at 4  C.
7. Wash three times in PBS (5 min each) and incubate with Alexa
Fluor 488 goat anti-rabbit and Alexa Fluor 594 goat anti-
mouse secondary antibodies (1:1000 in Blocking Solution)
1 h at room temperature.
8. Wash twice for 5 min each in PBS. Incubate with DAPI solu-
tion for 5 min at room temperature and wash twice more for
5 min each in PBS. Wash once in distilled H2O.
9. Coverslips are removed using tweezers and placed on a
microscope slide with Immumount mounting medium (see
Note 17).
10. The slides are placed flat at room temperature for 24 h and then
stored at 4  C (see Fig. 2).
 R Loop Detection 357

Fig. 2 Representative image of HeLa cells stained with DAPI (DNA), S9.6 (DNA-RNA hybrids), and anti-
nucleolin (nucleoli) antibodies. DNA-RNA hybrids can be observed at mitochondria, nucleoli, and nuclear DNA

3.4 S9.6 1. Collect at least 1 ml of mitotic culture.

Immunofluorescence 2. Add 100 μl of 37% formaldehyde to the media and vortex.
in Yeast
3. Spin the culture for 1 min at 900  g and resuspend in 3.7%
formaldehyde (see Note 18). Fix 10 min at room temperature.
4. Wash twice in 1 ml of 0.1 M Phosphate Buffer pH 6.4.
5. Resuspend in 1 ml of 1.2 M sorbitol-citrate (see Note 19).
6. Spin down and resuspend in digestion mix. Rotate on rack at
30  C for 5 min (see Note 20).
7. Spin down at 900  g for 2 min. Remove the supernatant and
resuspend in 1 ml of sorbitol by inversion (see Note 21).
8. Spin down at 900  g for 3 min. Remove the supernatant and
resuspend in sorbitol by inversion (see Note 21).
9. Clean the slide with a scrubber under distilled water (see Note
22). Let it dry completely.
10. Place 5 μl of 0.1% Poly-L-lysine on each well and let it sit for
5 min (see Note 23).
11. Wash the slide under distilled water and let it dry completely.
12. Prepare a humid chamber by wrapping a petri dish in alumi-
num foil and put a wet paper towel inside.
13. In order to make the cells burst, drop 5 μl of the cells onto the
slide from a height of 30–40 cm (see Note 24). Let cells sit for
10 min.
14. Remove the supernatant by placing the vacuum tip to the side
of the well (see Note 23).
15. Put slide in 4  C methanol (see Note 16) for 3 min followed by
10 s in 4  C acetone. Let it dry completely.
16. Add 5 μl of primary antibody (anti-DNA-RNA hybrid 1:200 in
PBS/BSA). Incubate in a wet chamber for 2 h at room
temperature.
358 Marı́a Garcı́a-Rubio et al.

17. Remove the supernatant and wash five times in PBS/BSA (see
Note 25).
18. Add 5 μl of secondary antibody (anti-mouse Alexa Fluor 546).
Incubate for 2 h at room temperature.
19. Remove the supernatant and wash five times in PBS/BSA.
20. Add 3 μl of mounting medium with DAPI to each sample. Put
on a coverslip. Paint with nail polish around the edges of the
coverslip and let it dry for 10 min.

4 Notes

1. Always pipet DNA with cut tips to preserve DNA-RNA

hybrids. Do not attempt to resuspend DNA by over-pipet-
ting/vortexing.
2. The solution becomes viscous.
3. This step is extremely important. Phase lock gel heavy 2 ml
tubes (5 PRIME GmbH, Hamburg, Deutschland) can be used
to facilitate sample phenol treatment.
4. Wear gloves and a lab coat while working with phenol, because
it is a very dangerous compound. Do it in the hood. Normally,
the aqueous phase forms the upper phase. You should always
dispose of phenol waste in a specially sealed container and
ensure that it is eliminated according to your institution’s
policies for dangerous waste.
5. Use a glass Pasteur pipette to spool the DNA. Do not spin to
precipitate the DNA. Centrifugation results in high loads of
RNA contamination.
6. Do not attempt to resuspend the DNA by vortexing, eventually
the DNA could be resuspended at 30  C (1–2 h). Invert the
tube occasionally.
7. Use Buffer 2.1 (NEB) for the restriction. You do not need to
remove the cut glass rod. We often run 4 μl on gel the next day
to ensure the digestion is complete and the samples are free of
RNA (see Fig. 3)
8. Better processing groups of two or three samples.
9. To prepare Sephadex G-50 columns, place empty micro Bios-
pin chromatography columns in 2 ml eppendorf tubes, add
800 μl of Sephadex G-50 slurry and centrifuge at 2000  g
for 2 min in a microcentrifuge. The Sephadex G-50 columns
are then placed in new 1.5 ml tubes, which serve to collect the
purified DNA after centrifugation at 2000  g for 2 min in a
microcentrifuge.
 R Loop Detection 359

Fig. 3 Genomic DNA digested with a mix of restriction enzymes for DRIP analysis.
4 μl of each digestion reaction were run on agarose gel to confirm that they are
free of RNA

10. We use the Primer Express application software (Applied Bio-

systems) to design the oligonucleotides.
11. The RNase A should be carefully boiled to eliminate residual
DNase activity.
12. Spheroplasts are very difficult to resuspend at this step. Use
1 ml cut tips.
13. At this point if there are not many clamps in the samples you
can proceed further with the extraction. If the preps are “dirty”
you can incubate overnight at 30  C.
14. Use tweezers to place two to three round coverslips on a well
before seeding the cells.
15. Fill the 24-well plate with 1 ml of PBS per well and place one
coverslip per well with the cells on the upper side.
16. Methanol is a hazardous chemical with significant toxic, flam-
mable, and reactive properties that can produce deleterious
impacts on human health and the environment when not
properly handled. Wear lab coat and gloves when working
with methanol and manipulate it always in a chemical fume
hood.
17. Place a drop of about 30 μl of mounting medium on a slide and
place the coverslip with the cells facing the mounting medium.
18. Formaldehyde is a sensitizing agent and a cancer hazard. Wear
gloves and lab coat and work always in a chemical fume hood.
360 Marı́a Garcı́a-Rubio et al.

19. At this step, samples can be stored at 20  C overnight.

20. Do not vortex.
21. Check digestion by mixing equal amounts of cells and 1% SDS.
You should see crystals at the microscope.
22. It is very important for cells to burst properly that the slides are
really clean.
23. Do not touch the well with the tip.
24. You can stand on a chair to get a good height to drop the cells.
25. Lay down a drop and remove it with a vacuum line.

Acknowledgments

A.A.’s lab is funded by Spanish Ministry of Economy and Compet-

itiveness, Junta de Andalucı́a, European Research Council, the
Worldwide Cancer Research and the European Union (FEDER).

References
1. Aguilera A, Garcia-Muse T (2012) R loops: ribosomal RNA synthesis. Genes Dev 24
from transcription byproducts to threats to (14):1546–1558
genome stability. Mol Cell 46(2):115–124 8. Pohjoism€aki JLO, Holmes JB, Wood SR, Yang
2. Yu K, Chedin F, Hsieh CL, Wilson TE, Lieber M-Y, Yasukawa T, Reyes A, Bailey LJ, Cluett
MR (2003) R-loops at immunoglobulin class TJ, Goffart S, Willcox S, Rigby RE, Jackson
switch regions in the chromosomes of stimu- AP, Spelbrink JN, Griffith JD, Crouch RJ,
lated B cells. Nat Immunol 4(5):442–451 Jacobs HT, Holt IJ (2010) Mammalian mito-
3. Huertas P, Aguilera A (2003) Cotranscription- chondrial DNA replication intermediates are
ally formed DNA:RNA hybrids mediate tran- essentially duplex but contain extensive tracts
scription elongation impairment and of RNA/DNA hybrid. J Mol Biol 397
transcription-associated recombination. Mol (5):1144–1155
Cell 12(3):711–721 9. Mischo HE, Gómez-González B, Grzechnik P,
4. Jinek M, Chylinski K, Fonfara I, Hauer M, Rondón AG, Wei W, Steinmetz L, Aguilera A,
Doudna JA, Charpentier E (2012) A program- Proudfoot NJ (2011) Yeast Sen1 helicase pro-
mable dual-RNA-guided DNA endonuclease tects the genome from transcription-associated
in adaptive bacterial immunity. Science 337 instability. Mol Cell 41(1):21–32
(6096):816–821 10. Wahba L, Amon JD, Koshland D, Vuica-Ross
5. Gavalda S, Santos-Pereira JM, Garcia-Rubio M (2011) RNase H and multiple RNA biogen-
ML, Luna R, Aguilera A (2016) Excess of esis factors cooperate to prevent RNA:DNA
Yra1 RNA-binding factor causes hybrids from generating genome instability.
transcription-dependent genome instability, Mol Cell 44(6):978–988
replication impairment and telomere 11. Stirling PC, Chan YA, Minaker SW, Aristizabal
shortening. PLoS Genet 12(4):e1005966 MJ, Barrett I, Sipahimalani P, Kobor MS,
6. Boguslawski SJ, Smith DE, Michalak MA, Hieter P (2012) R-loop-mediated genome
Mickelson KE, Yehle CO, Patterson WL, Car- instability in mRNA cleavage and polyadenyla-
rico RJ (1986) Characterization of monoclonal tion mutants. Genes Dev 26(2):163–175
antibody to DNA.RNA and its application to 12. Gómez-González B, Aguilera A (2007)
immunodetection of hybrids. J Immunol Activation-induced cytidine deaminase action
Methods 89(1):123–130 is strongly stimulated by mutations of the
7. El Hage A, French SL, Beyer AL, Tollervey D THO complex. Proc Natl Acad Sci U S A 104
(2010) Loss of Topoisomerase I leads to R- (20):8409–8414
loop-mediated transcriptional blocks during
 R Loop Detection 361

13. Wahba L, Costantino L, Tan FJ, Zimmer A, 15. Sollier J, Stork CT, Garcı́a-Rubio ML, Paulsen
Koshland D (2016) S1-DRIP-seq identifies RD, Aguilera A, Cimprich KA (2014)
high expression and polyA tracts as major con- Transcription-coupled nucleotide excision
tributors to R-loop formation. Genes Dev 30 repair factors promote R-loop-induced
(11):1327–1338 genome instability. Mol Cell 56(6):777–785
14. Ginno PA, Lott PL, Christensen HC, Korf I, 16. Garcia-Rubio ML, Perez-Calero C, Barroso SI,
Chédin F (2012) R-loop formation is a distinc- Tumini E, Herrera-Moyano E, Rosado IV,
tive characteristic of unmethylated human CpG Aguilera A (2015) The Fanconi Anemia Path-
island promoters. Mol Cell 45(6):814–825 way Protects Genome Integrity from R-loops.
PLoS Genet 11(11):e1005674
 Chapter 25

Analysis of De Novo Telomere Addition by Southern Blot

Diego Bonetti and Maria Pia Longhese

Abstract
Telomere length is maintained in most eukaryotes by the action of a specialized enzyme, the telomerase.
However, the complexity of mechanisms regulating telomeric DNA length as well as the heterogeneity in
length of each telomere in a population of cells has made it very difficult to understand how telomerase is
regulated in vivo. Here, we describe a method developed in Saccharomyces cerevisiae to monitor the addition
of telomeric sequences to a single newly generated telomere in vivo. The primary strain consists of a HO
endonuclease cleavage site that is placed directly adjacent to an 81-base-pair stretch of telomeric DNA
inserted into the ADH4 locus of chromosome VII. Upon cleavage by HO, the de novo DNA end is rapidly
healed by the telomerase enzyme and the analysis of this process allows to gain a mechanistic understanding
of how telomerase action is regulated in the cell.

Key words De novo telomere, HO endonuclease, S. cerevisiae, Southern blot, Telomerase

1 Introduction

Telomeres are specialized nucleoprotein complexes that maintain

the integrity of eukaryotic chromosomes. They are believed to
provide a “cap” to the end of the chromosome, thus preventing
fusions between chromosomes and concomitantly inhibiting activ-
ities that could either degrade chromosomes or inappropriately use
them as substrates for recombination events [1, 2].
Telomeric DNA consists in most species of a simple tandemly
repeated sequence that is guanine-rich in the strand running 50 -30
from the centromere towards the chromosome end (e.g., T2AG3 in
humans; T2G4 in Tetrahymena; TG1–3 in Saccharomyces cerevisiae).
In most eukaryotes, a specialized enzyme called telomerase adds
telomeric repeats to the 30 end of a telomeric G-rich strand. This
process represents the main pathway for maintaining telomeric
DNA repeats at the end of the chromosome [1, 2]. Importantly,
work in S. cerevisiae has shown that the telomerase does not act on
every telomere in each cell cycle, but preferentially extends the
shortest telomeres [3–5].

Marco Muzi-Falconi and Grant W. Brown (eds.), Genome Instability: Methods and Protocols, Methods in Molecular Biology,
vol. 1672, DOI 10.1007/978-1-4939-7306-4_25, © Springer Science+Business Media LLC 2018

363
364 Diego Bonetti and Maria Pia Longhese

Telomeric DNA repeat length has generally served as a readout

for telomerase activity in vivo, because this enzyme is required to
maintain a particular steady-state length of telomeric DNA at the
end of the chromosome. For example, in most yeast S. cerevisiae
strains the length of TG1–3 repeats at the end of the chromosomes
varies between 225 and 375 base-pair (bp) [1, 2]. However, the
length heterogeneity at a given telomere in a population of cells and
the complexity of telomeric DNA length regulation have made it
difficult to understand how telomerase is regulated in vivo.
To circumvent this problem, Gottschling and colleagues have
developed a Saccharomyces cerevisiae system that allows to monitor
in vivo the addition of telomeric DNA sequences by telomerase onto
a single de novo created telomere [6, 7]. This system was based on a
previous finding that a double-strand break generated by the HO
endonuclease can be healed by the telomerase enzyme at a low
frequency when stretches of the T2G4 repeat are present near the
break site [8]. Gottschling and colleagues made the healing more
efficient by constructing a haploid strain carrying the HO recogni-
tion sequence immediately adjacent to an 81 bp telomeric “seed”
sequence of C1–3A/TG1–3 repeats. Upon HO cleavage, the new
DNA end is rapidly elongated in a telomerase-dependent manner
to reach a steady-state length and this elongation is dependent upon
most of the telomere regulating factors identified so far [6, 7]. In
more detail, the ADH4 locus on chromosome VII in strain
UCC5913 is replaced with a 6 kb fragment consisting of the S.
cerevisiae ADE2 gene, 81 bp of TG1–3 telomeric sequences and the
recognition site for the HO endonuclease (Fig. 1). Since a recogni-
tion site for this endonuclease is normally present at the mating type
locus (MAT), the UCC5913 strain carries a mutation that makes this
site uncleavable (MAT-inc). Moreover, HO expression is under the
control of the inducible GAL1 promoter, and the GAL1-HO con-
struct is integrated at the LEU2 locus. As transcription from GAL
promoter is repressed by glucose, cells are grown in raffinose-
containing media followed by galactose addition to induce HO
expression. After GAL1-HO induction, telomeric DNA addition to
the newly generated telomere end can be monitored by Southern
blot analysis with an ADE2 specific probe. It is important to mention
that the distal part of chromosome VII containing the native telo-
mere (ca. 20 kb) is lost at the first cell divisions after the HO cut
(Fig. 1), but no essential genes are present in this region. Figure 2 is a
representative Southern blot showing how this system allows to
monitor telomere addition in wild type cells as well as in sae2Δ
mutant cells that are defective in this process.
Here, we describe a protocol to monitor de novo telomeric
DNA addition by telomerase at a single telomere in haploid yeast
cells carrying the HO system described above. We divide it in three
main steps: (i) generation of the de novo telomere; (ii) genomic
DNA extraction; and (iii) DNA electrophoresis and Southern blot
analysis.
 Detection of De Novo Telomeric DNA Addition 365

Fig. 1 System to detect de novo telomere addition. In UCC5913 strain [MATa-inc

ade2-101 lys2-801 his3-Δ200 trp1-Δ63 ura3-52 leu2-Δ1::GAL1-HO-LEU2
VII-L::ADE2-TG(1–3) -HO site-LYS2], the ADH4 locus on chromosome VII is
replaced with a 6 kb fragment consisting of the S. cerevisiae ADE2 gene and
81 bp of TG1–3 telomeric sequences (zigzag lines) flanking the recognition site
for the HO endonuclease. SpeI cuts 700 bp upstream of the ADE2 ORF starting
codon (ATG to TAA is from TG repeats to centromere direction). A wild-type copy
of the LYS2 gene is present on the distal side of the HO cut site. This region of
about 20 kb does not contain essential genes and is lost upon galactose
induction. The probe used for Southern blot analysis is a 560 bp fragment of
the ADE2 gene located upstream of the ATG (720 bp to 1280 bp). Telomere
addition is indicated as grey zigzag lines)

2 Materials

All solutions are prepared using filtered and deionized ultrapure

water (ddH2O; resistivity 18.2 MΩ cm at 25  C) and analytical
grade reagents. Prepare and store all solutions at room temperature
(unless otherwise indicated).

2.1 HO Endonuclease 1. YEP medium: 2% Bacto peptone, 1% yeast extract, 0.0025%

Induction in Yeast adenine hemisulfate salt. Dissolve in ddH2O. Autoclave.
Cells 2. 30% raffinose: Dissolve D-(þ)-raffinose pentahydrate in
ddH2O. Autoclave or sterilize by filtration. Add sterilized raffi-
nose to YEP medium to 2% final concentration.
3. 30% galactose: Dissolve D-(þ)-galactose 99.0% in ddH2O.
Sterilize by filtration only (see Note 1).
366 Diego Bonetti and Maria Pia Longhese

Fig. 2 Southern blot analysis to detect de novo telomere addition. Wild type and
sae2Δ cells, exponentially growing in raffinose, were shifted to galactose
containing medium at time zero to induce HO expression. Genomic DNA
prepared at the indicated times after galactose addition was digested with
SpeI and separated on an 0.8% agarose gel by about 16 h run at 2 V/cm. DNA
fragments were transferred onto a nylon membrane and hybridized with the
ADE2 specific probe described in Fig. 1. A band of about 3 kb (uncut) can be
detected in the absence of galactose. This band is converted by HO cleavage into
a 0.7 kb fragment (HO-cut). A bracket points out new telomere repeats added to
the exposed TG1–3 telomeric sequences. The band of about 1.6 kb (INT)
represents the endogenous ade2-101 gene that can be used as a loading control

4. SD -Lys plates: Dissolve 6.7 g/L of yeast nitrogen base without

amino acids, 20 mg/L uracil, adenine, histidine, tryptophan,
leucine, arginine and methionine, 50 mg/L threonine, 60 mg/
L phenylalanine and isoleucine, 2% D-(þ)-glucose monohy-
drate. Dissolve in ddH2O. Add 2% Bacto agar. Autoclave,
cool down and pour plates.

2.2 Genomic DNA 1. Spheroplasting solution: 0.9 M sorbitol, 0.1 M ethylenediami-

Extraction netetraacetic acid (EDTA), pH 7.5.
2. Zymolyase solution: Dissolve 2 mg/mL Zymolyase 20T® (see
Note 2) from Arthrobacter luteus (Nacalai Tesque) in sphero-
plasting solution þ14 mM β-mercaptoethanol.
3. 1 TE: 10 mM Tris–HCl, pH 7.5, 1 mM EDTA, pH 7.4.
Autoclave.
 Detection of De Novo Telomeric DNA Addition 367

4. Lysis solution: 2.2% sodium dodecyl sulfate (SDS), 278 mM

EDTA, pH 8.5, 445 mM Tris base. Prepare the lysis solution
just before use by mixing the appropriate amounts of 10% SDS,
0.5 M EDTA, pH 8.5, 2 M Tris base stock solutions.
5. 5 M potassium acetate: Dissolve 5 M potassium acetate in
ddH2O. Autoclave.
6. Ice-cold ethanol: 96% and 70%. Prepare a 70% ethanol solution
by diluting 96% ethanol in ddH2O. Store at 20  C aliquots of
both 70% ethanol and 95% ethanol in glass bottles.
7. RNase A solution: Dissolve 10 mg/mL RNase A, DNase-free,
in 10 mL 10 mM Tris–HCl, pH 7.5, 15 mM NaCl. Heat to
100  C for 5 min and cool down at room temperature. Prepare
small aliquots and store them at 20  C.
8. 2propanol anhydrous, 99.5%.
9. 6 DNA loading buffer: 30% glycerol, 0.25% bromophenol
blue.
10. 1 TAE buffer: 40 mM Tris–acetate, 10 mM EDTA. For 50
TAE buffer stock: for 1 L dissolve 242 g of Tris base in
approximately 600 mL ddH2O. Add 57.1 mL glacial acetic
acid and 100 mL 0.5 M EDTA, and bring final volume to 1 L
with ddH2O. Before use, dilute in ddH2O to a final concentra-
tion of 1.
11. Agarose gel: melt 0.8% agarose in 1 TAE buffer. Cool at
approximately 60  C and add 10 mg/mL ethidium bromide
solution to a final concentration of 1 μg/mL. Pour the gel into
a gel tank and insert a comb.
12. Ethidium bromide solution: prepare a stock of 10 mg/mL
ethidium bromide in ddH2O and store in light-protected con-
tainers (see Note 3).
13. UV lamp with camera.

2.3 Native Gel 1. SpeI restriction enzyme (20,000 U/mL; New England Bio-
Electrophoresis and labs) and 10 buffer supplied from the distributor.
Southern Blot
2.3.1 DNA Digestion

2.3.2 Native Gel 1. Horizontal electrophoresis system with a large gel running
Electrophoresis and chamber (gel size 25  20 cm) and standard 32-tooth comb
Transfer (thickness 1.0 mm and width of teeth 4.0 mm).
2. 1 TAE buffer.
3. Denaturing solution: 0.2 N NaOH, 0.6 M NaCl. Dissolve in
ddH2O just before use.
4. Neutralizing solution: 1 M Trizma base, 1.5 M NaCl. Dissolve
in ddH2O just before use. Adjust pH to 7.4 with HCl.
368 Diego Bonetti and Maria Pia Longhese

5. Nylon hybridization transfer membrane (e.g., GeneScreen®

Plus from Perkin Elmer).
6. 20 SSC buffer: 3 M NaCl, 300 mM sodium citrate. Adjust
the pH to 7.0 with HCl. Autoclave.
7. Whatman™ 3 MM paper.
8. Paper towels.
9. UV cross-linker (e.g., Stratalinker® from Stratagene).

2.3.3 Probe Labeling 1. DNA template. dsDNA for random priming labeling is
obtained by PCR using yeast genomic DNA as a template
(e.g., UCC5913 DNA prepared for Southern blot analysis)
and oligos ADE2-30 50 -ATTTACAGTTTTGATATCTTGGC-
30 and ADE2-50 50 -TTCTAATGTAGATTCTTGTTGTTCG-
30 to amplify a 560 bp region at the ADE2 locus.
2. Gel extraction kit (e.g., QIAgen).
3. Random priming labeling kit (e.g., Decaprime™ II kit from
Ambion®)
4. dATP-α32P or dCTP-α32P. Specific activity: 3000 Ci/mmol.

2.3.4 Filter Hybridization 1. Hybridization oven and hybridization tubes.

2. Hybridization solution: 0.5 M NaPO4 pH 7.2, 1 mM EDTA
pH 7.5, 7% SDS, 1% BSA. Prepare fresh in sterile ddH2O.
3. Washing solution: 0.2 M NaPO4 pH 7.2, 1% SDS. Prepare
fresh in sterile ddH2O.
4. Autoradiography cassette with intensifying screens.
5. Autoradiography films or imaging plates.

3 Methods

3.1 De Novo 1. Streak out the strains to be analyzed on SD -Lys þ 2% glucose

Telomere Generation plates and grow them for 2 days at 25–30  C (see Note 4).
2. Inoculate cells in an appropriate volume of YEP þ raffinose
(2%) medium (see Note 5).
3. Grow the cell culture overnight at 25–30  C.
4. Next day, when the cell culture has grown up to 8  106 cells/
mL (OD600nm 0.5) in an appropriate volume (see Note 5), take
first 50 mL of the cell culture for genomic DNA extraction
(uncut control) and then add galactose to 2% final concentra-
tion. If telomere addition has to be analyzed in a specific cell
cycle phase (e.g., G2/M), first of all synchronize the cell culture
(see Note 6).
5. Grow the cell culture and take 50 mL samples at the desired
time points after galactose addition.
 Detection of De Novo Telomeric DNA Addition 369

3.2 Genomic DNA 1. Pellet the cells by spinning for 3 min at RT at 1600  g in
Extraction 50 mL tubes.
2. Wash the cells in 1 mL spheroplasting solution and transfer the
samples to 1.5 mL microcentrifuge tubes.
3. Spin for 3 min at 1600  g and completely remove supernatant
with a tip.
4. Freeze and store the pellets at 20  C.
5. Thaw the samples at room temperature and resuspend the cell
pellets in 400 μL spheroplasting solution, 14 mM β-
mercaptoethanol.
6. Add 100 μL of Zymolyase solution to each sample and invert
the tube 4–6 times. Incubate the samples at 37  C. After
30 min check the formation of spheroplasts under a light
microscope (see Note 7).
7. When >95% cells become spheroplasts, spin for 1 min at
15,000  g and carefully remove supernatant with a tip.
8. Gently resuspend spheroplasts in 400 μL 1 TE (do not
vortex).
9. Add 90 μL lysis solution (prepared just before use). Immedi-
ately mix by inverting the tube several times and incubate the
samples for 30 min at 65  C. This causes spheroplasts lysis.
10. Add 80 μL 5 M potassium acetate and mix by inverting the
tube several times. Place the tubes on ice for at least 1 h.
11. Spin for 15 min at 15,000  g at 4  C. Transfer the supernatant
to new 1.5 tubes. Discard the pellets.
12. Add 1 mL ice-cold 96% ethanol and mix by inverting several
times the tube. A white cloudy precipitate with the nucleic
acids should form.
13. Spin for 5–10 min at 15,000  g at 4  C and remove the
supernatant.
14. Wash the pellet with 1 mL ice-cold 70% ethanol and immedi-
ately discard the ethanol.
15. Air-dry the pellet.
16. Add 500 μL 1 TE. Let tubes sit for 15 min at room tempera-
ture (or overnight at 4  C), then gently dissolve the pellet (do
not vortex).
17. Once pellets are completely dissolved, add 2.5 μL RNase A
solution to each sample and incubate for 1 h at 37  C.
18. Add 500 μL 2-propanol and invert several times the tube. The
solution should become cloudy.
19. Spin for 15–30 min at 15,000  g at 4  C and remove the
supernatant.
370 Diego Bonetti and Maria Pia Longhese

20. Wash the pellet with 1 mL ice-cold 70% ethanol, invert and
immediately discard the ethanol.
21. Air-dry the pellet.
22. Add 50 μL 1 TE. Let tubes sit for 30 min at room tempera-
ture or overnight at 4  C, then gently dissolve the DNA pellet.
Avoid pipetting or vortexing to prevent DNA shearing (see
Note 8).
23. Load 1 μL of each genomic DNA sample (added to 9 μL of 1
DNA loading buffer) on a 0.8% agarose gel with ethidium
bromide and run in 1 TAE buffer. Check under the UV
lamp the quality of the extracted DNA (see Note 9).

3.3 DNA Native Gel 1. Digest at least 5 μg (up to 10 μg) of DNA for 5–6 h at 37  C
Electrophoresis and with 20 U of SpeI restriction enzyme (New England Biolabs)
Transfer with 1 enzyme buffer in a total volume of 25 μL.
2. In the meanwhile melt 0.8% agarose in 450 mL TAE 1 buffer
and ethidium bromide, and pour into a gel tray. Prepare also a
smaller 0.8% agarose gel to check for the digestion reaction
efficiency.
3. Once the gels are solid, put them in a gel running apparatus
and fill in with TAE 1.
4. Before loading the Southern blot gel, test 1 μL of each diges-
tion reaction (added to 9 μL of 1 DNA loading buffer) on a
0.8% agarose gel with ethidium bromide and run in 1 TAE
buffer (see Note 9).
5. Load now on the 450 mL gel the whole digestion reaction
mixed with 5 μL of 6 bromophenol blue loading dye, as well
as a DNA ladder.
6. Run the gel overnight at 2 V/cm (see Note 10).
7. Once the run is complete remove the gel from the tank and
take a picture under UV light.
8. Soak the gel 30–60 min with gentle agitation in Denaturing
solution.
9. Soak the gel 30–60 min with gentle agitation in Neutralizing
solution.
10. Blot overnight the DNA from the gel onto a nylon neutral
membrane by capillary transfer with 10 SSC buffer as for
standard Southern blot procedure (see Note 11).
11. After overnight transfer quickly soak the membrane in SSC 4
or water (optional).
12. Let the filter air-dry on 3MM paper for at least 30 min.
13. Cross-link the DNA on the membrane with an UV cross-linker
by following the instructions of the manufacturer.
 Detection of De Novo Telomeric DNA Addition 371

3.4 Probe Labeling 1. Insert the filter in a hybridization tube and quickly rinse it with
and Filter ddH2O.
Hybridization 2. Block the filter by incubating 5 h in a hybridization oven at
55  C in 25 mL hybridization buffer (prehybridization step).
3. In the meanwhile, proceed with probe labeling according to
the manufacturer protocol (see Note 12). We generally use
25–50 ng of gel-extracted DNA template.
4. As soon as the prehybridization step is complete, discard the
prehybridization solution and add 25 mL of fresh hybridiza-
tion buffer.
5. Denature the labeled probe for 5–10 min at 98  C and immedi-
ately add it to the hybridization tube (eventually keep it on ice).
6. Incubate overnight at 55  C by gently rotating in the hybridi-
zation oven.
7. Next day remove the hybridization solution (see Note 13).
8. Wash the filter 60 min at 55  C with washing solution.
9. Wash the filter 30 min at 55  C with washing solution.
10. Air-dry the filter on Whatman™ 3MM paper.
11. Expose the filter to an imaging plate or to an autoradiography
film in an autoradiography cassette with intensifying screens
(see Note 14).
12. Develop the filter. See Fig. 2 for a representative gel image.

4 Notes

1. As galactose is reported to isomerize at elevated temperatures,

its sterilization should be obtained by filtration. Autoclaving
should be avoided.
2. There are two commercially available preparations of Zymo-
lyase from Nacalai Tesque: Zymolyase® 20T and Zymolyase®
100T (lytic activity of 20 U/g and 100 U/g, respectively).
Zymolyase100T can also be used but the concentration in the
spheroplasting solution and/or the time of incubation should
be carefully adjusted.
3. Ethidium bromide can be substituted by less-toxic dyes such as
GelRed, or SYBR Safe.
4. The UCC5913 strain has a LYS2 wild type gene on the distal
side (compared to the TG sequences) of the HO cut site and it
is worth streaking out the cells on SD -Lys þ 2% glucose plates
(see Subheading 2) before inoculation. This allows to select for
cells with an intact construct at chromosome VII.
5. Be sure to inoculate a proper amount of cells in YEP þ raffinose
medium to reach a density of at least 6  106 cells/mL the next
372 Diego Bonetti and Maria Pia Longhese

day. Cells grow much slower in presence of raffinose compared

to glucose as a carbon source. Usually the volume of the culture
inoculated overnight should be no less than half of the volume
needed the following day. We usually collect 50 mL of culture
for each time point to be analyzed.
6. To monitor de novo telomere addition in G2/M phase, cells
can be arrested in metaphase by treatment with the drug noco-
dazole that interferes with microtubule polymerization. To this
end, prepare a 100 nocodazole stock by dissolving nocoda-
zole at 1.5 mg/mL in 100% dimethyl sulfoxide (DMSO). Add
the nocodazole stock to the exponentially growing cells in
YEP þ raffinose medium to reach a final concentration of
15 μg/mL. After 2.5–3 h at 25–30  C, about 90% cells should
be arrested as large-budded cells (score at the microscope).
Nocodazole-mediated arrest can be maintained for several
hours.
7. In order to monitor the efficiency of zymolyase digestion,
prepare two drops (5 μL) of the same sample on a microscope
slide and add 1-2 μL of 10% SDS to one of them. Spheroplasts
will lyse in SDS solution. We usually obtain >95% spheroplasts
in 40–50 min.
8. At this step the solution does not often look clear. In this case
ensure that the pellet is fully resuspended, then spin again for
10–15 min at 15,000  g at RT and take the clear supernatant.
9. Yeast genomic DNA is mainly visible as a unique discrete band
of high molecular weight on an agarose gel. After digestion,
depending on the enzyme used and the frequency of cutting
sites in the genome, it usually appears as a pattern of several
bands of different size.
10. In an 0.8% agarose gel the bromophenol blue usually migrates
as DNA fragments of about 700 bp in size. This corresponds to
the cut product to be monitored for de novo telomere addi-
tion. Since the latter will only increase of few dozens base pairs
upon telomerase action, a nice separation of fragment sizing
between 1000 and 700 bp is required. We usually stop electro-
phoresis run when the blue front has migrated about 16–18 cm
from wells. This usually requires at least 16 h run at 55 V for a
25 cm long gel.
11. Briefly, fill a tank with 10 SSC buffer (approximately 1 L).
Place a support in the tank, wet it with some SSC 10 and
immediately create a “bridge” with Whatman™ 3MM paper
onto the support. Pour some more 10 SSC over the 3MM
paper so that it is fully wet and carefully remove air bubbles.
Place the gel on top. Wet a nylon membrane in ddH2O (it
should fit with the size of the gel, e.g., 20  20 cm) and lay it
on top of the gel removing air bubbles. Wet 2 Whatman™
 Detection of De Novo Telomeric DNA Addition 373

3MM papers (e.g., 20  20 cm) in 10 SSC and lay them on

the membrane. Place a stack of absorbent paper towels on top,
place a glass plate and put a 0.5 kg weight. Allow upward
capillary transfer at room temperature for at least 18 h.
12. Clean up of the probe with G50 columns is usually not
required.
13. Probes can eventually be labeled in advance and stored at 4  C.
Probes recovered after hybridization (in hybridization buffer)
can be reused multiple times generally within a month (32P
half-life is about 14 days). In all cases, probes need to be boiled
for 5–10 min before each use.
14. In case autoradiography films are used, the cassette can be
placed at 80  C to further intensify the signal. With freshly
labeled dATP-α32P, we usually obtain a good signal after a 4 h
exposure at 80  C.

Acknowledgments

We thank G. Lucchini for critical reading of the manuscript. D.B.

has been supported by a fellowship from CancerTelSys (grant
01ZX1302) in the E:med program of the German Federal Ministry
of Education and Research (BMBF). Research in Longhese’s lab is
supported by Associazione Italiana per la Ricerca sul Cancro
(AIRC) (grant number 15210) and Cofinanziamento 2015
MIUR/Università di Milano-Bicocca.

References
1. Bonetti D, Martina M, Falcettoni M, Longhese 5. Bianchi A, Shore D (2007) Increased association
MP (2014) Telomere-end processing: mechan- of telomerase with short telomeres in yeast.
isms and regulation. Chromosoma 123:57–66 Genes Dev 21:1726–1730
2. Wellinger RJ, Zakian VA (2012) Everything you 6. Diede SJ, Gottschling DE (1999) Telomerase-
ever wanted to know about Saccharomyces cer- mediated telomere addition in vivo requires
evisiae telomeres: beginning to end. Genetics DNA primase and DNA polymerase α and δ.
191:1073–1105 Cell 99:723–733
3. Marcand S, Brevet V, Gilson E (1999) Progres- 7. Diede SJ, Gottschling DE (2001) Exonuclease
sive cis-inhibition of telomerase upon telomere activity is required for sequence addition and
elongation. EMBO J 18:3509–3519 Cdc13p loading at a de novo telomere. Curr
4. Teixeira MT, Arneric M, Sperisen P, Lingner J Biol 11:1336–1340
(2004) Telomere length homeostasis is achieved 8. Kramer KM, Haber JE (1993) New telomeres in
via a switch between telomerase- extendible and yeast are initiated with a highly selected subset of
-nonextendible states. Cell 117:323–335 TG1-3 repeats. Genes Dev 7:2345–2356
 Chapter 26

Assays to Study Repair of Inducible DNA Double-Strand

Breaks at Telomeres
Roxanne Oshidari and Karim Mekhail

Abstract
The ends of linear chromosomes are constituted of repetitive DNA sequences called telomeres. Telomeres,
nearby regions called subtelomeres, and their associated factors prevent chromosome erosion over cycles of
DNA replication and prevent chromosome ends from being recognized as DNA double-strand breaks
(DSBs). This raises the question of how cells repair DSBs that actually occur near chromosome ends. One
approach is to edit the genome and engineer cells harboring inducible DSB sites within the subtelomeric
region of different chromosome ends. This provides a reductionist and tractable genetic model system in
which mechanisms mediating repair can be dissected via genetics, molecular biology, and microscopy tools.

Key words DNA double strand break (DSB), Telomeres, Subtelomeres, Inducible, Survival, Endo-
nuclease, Nuclear envelope, Chromatin silencing, Homologous recombination (HR), Nonhomolo-
gous end joining (NHEJ)

1 Introduction

During S phase of the cell cycle, the ends of linear eukaryotic

chromosomes cannot be replicated by the replication machinery
[1]. This is a long-appreciated limitation known as the end replica-
tion problem [1]. To prevent rapid erosion of chromosome ends
over rounds of replication, the ends are constituted of highly repet-
itive DNA sequences known as telomeres, whose length is main-
tained by the enzyme telomerase [2, 3]. Telomeres also prevent the
deleterious fusion of various chromosome ends to each other by
preventing the classic DNA repair machinery from recognizing
them as the ends of a DNA double-strand break (DSB) [4]. Telo-
meres can also help initiate silent chromatin structures that then
spread to nearby chromosomal regions known as subtelomeres
[5, 6]. In yeast, telomeric TG1–3 repeats help recruit the Silent
Information Regulator (SIR) protein complex composed of Sir3,
Sir4, and the catalytic subunit Sir2 histone deacetylase [7]. Iterative
cycles of Sir2-mediated deacetylation of histone tails at neighboring

Marco Muzi-Falconi and Grant W. Brown (eds.), Genome Instability: Methods and Protocols, Methods in Molecular Biology,
vol. 1672, DOI 10.1007/978-1-4939-7306-4_26, © Springer Science+Business Media LLC 2018

375
376 Roxanne Oshidari and Karim Mekhail

nucleosomes followed by recruitment of additional SIR complexes

allow silent chromatin to spread within subtelomeres [7]. This form
of chromatin silencing compacts chromatin, making it less accessi-
ble to the DNA repair and recombination machineries and ulti-
mately more stable [7, 8].
Importantly, this chromatin silencing is greatly promoted by
the clustering of telomeres in a handful of foci along the nuclear
envelope [6, 8]. Specifically, perinuclear telomere clustering
increases the local concentration of SIR complexes [8, 9]. Proteins
mediating perinuclear telomere tethering include the Cohibin
complex (composed of Lrs4 and Csm1), Enhancer of silent chro-
matin 1 (Esc1), the inner nuclear membrane protein Monopolar
spindle 3 (Mps3), Ku proteins, and even telomerase [8, 10–13].
Given that chromosome ends are confined to perinuclear domains
that are highly repressive to DNA recombination and repair, what
happens to subtelomeric DSBs? Interestingly, DSBs within the
rDNA repeats of yeast transiently escape their usual Rad52/recom-
bination-repressive perinuclear/nucleolar environment to undergo
repair [14, 15]. In Drosophila, heterochromatic DSBs exit their
Rad51/recombination-repressive subnuclear environment to
undergo homologous recombination (HR)-dependent repair at
the nuclear periphery [16]. Similarly, heterochromatic DSBs relo-
cate outside their domain for repair in cultured human cells [17]. It
was recently shown that telomeric DSBs are transiently relocated to
nuclear pore complexes (NPCs) enriched in factors mediating an
error-prone repair mechanism called break-induced replication
(BIR) [18–22]. This mechanism, also implicated at nontelomeric
BIR-repairable DSBs, allows cells to survive DSBs but compromises
overall genome integrity [18]. Importantly, yeast genetic screens
employing inducible subtelomeric DSBs have revealed that BIR-
dependent repair is dependent on perinuclear telomere tethering
and chromatin remodeling that then allow kinesin motor proteins
to transiently relocate subtelomeric DSBs to specialized nuclear
envelope regions such as nuclear pore subcomplexes critical to key
repair steps [18].
To study subtelomeric DSB repair, one can generate a single
DSB by targeting a restriction endonuclease to a desired chromo-
some end locus [18, 21, 23]. This strategy allows for a site-specific
and controlled approach to DSB investigation. Researchers com-
monly use the I-SceI endonuclease to generate single DSBs [18,
21, 23]. Here, we show how to exploit this approach in concert
with classic yeast genetics to determine DSB repair efficiency as well
as repair pathway choice at chromosome ends [24]. Briefly, the
URA3 reporter gene flanked by two inverted I-SceI restriction
sites is inserted 3.5 kb downstream of the telomere on the left
arm of chromosome XI, or another subtelomeric region as needed.
As an internal control, one can alternatively insert the above-
mentioned DSB site outside of the subtelomeric region on the
 Assays to Study Repair of Inducible DNA Double-Strand Breaks at Telomeres 377

same chromosome arm studied. Cells are then transformed with a

plasmid harboring the I-SceI open reading frame (ORF) under a
galactose-inducible promoter allowing for controlled DSB induc-
tion via plating on galactose-containing media. These cells are
subsequently grown and plated on media supplemented with either
glucose (GLU, no induction) or galactose (GAL, induction). Cells
that repair the induced DSB grow and generate colonies on media
supplemented with galactose, while unsuccessful repair is lethal
[18]. The DSB induction and repair efficiency of varying mutants
can subsequently be calculated and the repair pathway employed by
cells can further be determined via PCR. This method can be
adapted to study any telomere, which can also be visualized via
microscopy for movement and dynamic character.

2 Materials

2.1 Media Media is prepared as detailed in [25].

1. SC-LEU(GAL) agar plates and liquid media.
2. SC-LEU(GLU) agar plates and liquid media.
3. SC and SC-URA agar plates.
4. SC and SC-URA agar plates.
5. Autoclaved ddH2O.

3 Methods

Carry out all procedures at room temperature unless otherwise

indicated. All water used should be sterile ddH2O.

3.1 Preparation 1. Generate the URA3 cassette flanked by two inverted I-SceI cut
of Yeast Strains sites for integration into the preferred subtelomeric or internal
locus. Briefly, PCR amplify URA3 using primers harboring I-
SceI cut sites and subtelomeric sequences that flank the inser-
tion site (Fig. 1a). This allows for targeted insertion of the
cassette via homologous recombination (Fig. 1b). Alterna-
tively, insertion of the cassette can be carried out as described
[21, 23]. For integration into any subtelomere, sequences with
homology to all subtelomeres of S. cerevisiae are listed in
Table 1. As an example in this paper, we will also focus on a
DSB cassette inserted into the left subtelomere of chromosome
XI within the ykl222c ORF 3.5 kb away from the telomere [18,
21]. As a control, the cassette can also be inserted ~64 kb away
from the same telomere (Fig. 1b).
2. Transform strain of interest with the integrating PCR product
via the lithium acetate method [18].
378 Roxanne Oshidari and Karim Mekhail

URA3 URA3
S CS CS CS
C +s +i
L+ ub al + nte
TE TEL ern rna
sub int l

B SubTEL DSB Internal DSB

Distance to TEL 3.5 Kb 64 Kb

Left arm of Chr XI TELXI-L TELXI-L
Insertion site
cs URA3 cs cs URA3 cs
I-SceI-induced DSB site

C LEU2, TRP1, Ampr

D

GF
GF
GF
GF
GF
GF
LacI-GFP

P
P
P
P
P
P
LacO
SubTEL-GFP
Nup49−GFP
Pgal I-SceI cs URA3 cs

Fig. 1 Production and integration of the I-SceI URA3 cassette. (a) Amplify URA3
using primers harboring I-SceI cut sites and sequences homologous to the
desired locus of insertion, allowing for targeted insertion via homologous
recombination. (b) Cassette is inserted 3.5 kb away from subTEL XI-L, and as
an internal control is integrated ~64 kb away from the same telomere. (c) Major
features of the I-SceI expression plasmid pKM97. (d) A LacO array can be
inserted downstream of the I-SceI URA3 cassette to allow for binding of ectopi-
cally expressed LacI-GFP. This allows for visualization of the DSB site by
microscopy

3. Once integration of the cassette is confirmed by plating on SC-

URA(GLU) media and PCR over the integration site, trans-
form strains with the GAL-inducible I-SceI expression plasmid
pKM97 (Fig. 1c) [18, 21, 23].
4. If microscopy-based tracking of the DSB site is desired, a LacO
array can be integrated [26] adjacent to the DSB site allowing
for binding of ectopically expressed GFP-tagged LacI
(Fig. 1d).

3.2 Inducible 1. Grow 3 mL of S. cerevisiae culture in SC-LEU(GLU) medium

Subtelomeric DSB overnight at 30
C to saturation.
Repair Assay 2. Inoculate 300 μL of saturated culture into 5 mL of SC-LEU
(GLU) medium and grow at 30
C for 3 h. or until cells reach
log phase (see Note 1).
3. Pellet 5.0  107 cells at 18,000  g.
4. Wash pellets and resuspend in 1 mL of water.
Table 1
List of overhang sequences for insertion of I-SceI URA3 cassette into any of the subtelomeres of S. cerevisiae

TEL Forwarda Reversea

I-L GCTGGACATTATCCAGTAAAGGCTTGGTAGTAACCATAAT ATGAGTCTTTCAAGTTCTTAGCGTTTCGTACCTGGGTAAT
I-R CCGGAGAGATTCCATATTCCGCACGTCACATTGCCAAATT GCCCTGACACAATATTAAAACATTGTACCACCTCATGTAA
II-L GCTCTGTGTCGACTTTCTTTTCGCCAGGTAAACTTGCTTG GTTACTACACGTAGATGAGCTATCGATTTTTTCTGCATAC
I-IR GATGTTGATTTATCAATTTCCGGTATAGTTTCCGGTATGG CTAGTTCTGAAACAGCGGAACTTCCACTGGATCAAGCTGG
III-L GCTATATGTCCCTACGGCCTCGTTTAGCCCCATCTCGCAT ACTACTTCTCTGTGTGTATATACCTATGTTACGGTATTGC
III-R GTCCATGACCTTAGGTCTTCAATATGAGCCTTCGAAGTAC TCCGTCTCTGCCTTATCACTTCTGGATCTTTGAAACGTTT
IV-L ACGCTGATATCTTTATAATTAACCCGGAATGGAAATTGTG TCACAGGTCTTAATACTCGAAACGCCATTGATGTACTGTG
IV-R CATTGTTCATTCTCTAGTACTCGGAATGCGGAAGACGTAG ATTTTTTGATCATGATCTGAACTCGCCGCCTCACCTGCCA
V-L CAGTGTACAGTTGGATGCGTCTTGTTGTATCATCGACGTA GCTTCAGCTACTCGCTTTTGCCGGTCCTCAGCGCAACGTA
V-R GACAACAACGGCGTTGCGATGTGGAAATGGATCAAAATTC GAGTCCTTCAGCGCGGTTTCCTGCGCTACTTCGTGGTACA
b
V-R TCCAGGGCGGACCAATCCCTTCAAACAAACCACGTGATGG CTATTGAATTTCAAAGTATGCACTTGAAACAACGTGTAGA
VI-L TAAGCGACTGTCTGTCCCCTTCAAGCACAACATAGAAAAC TTGCAGGCCGGAAATCAAGCGATGAATGAGACATCCTTCT
VI-R GAGATGCAGCAAAATAACCACTGGTGTTTAAGAAATAATA AGCAGAGTGCGTACGGGCTCTTCGGTTACTGACGTAATCA
VII-L CTTGTACGGTCGCCATATCATGAGATTACTTGAAGCTTCC AAAGAAAAACGTATAAAACTGGTTACGGTATGAGCTTCAA
VII-R GAAGAGGATCGATCCATTACGAATGAAGAACCCATTATTC CGCAACTTACATACTTTCAAGCCATGCTCGTCCACAGAGG
b
VII-R ATATTACAATATATTGAATAGATCATAATGGTATGACGAT CTATACCACCATGACGTCATTAACTTAAATGTTCCTTAAT
VIII-L CGCGTTACTCTTGAAATTTTTTGATCATGATCTGAACTCG AGTACTCGGAATGCGGAAGACGTAGTGGCAGGTGAGGCGG
VIII-R GATGGATTTCTTGTCAAAAAGCATAGCAATCAACATACTA GAAACTTTCATTTTATTTTTGTAAGTTTCGAAATTAACAA
Assays to Study Repair of Inducible DNA Double-Strand Breaks at Telomeres

IX-L GTGCCGCTCAAAAAGATTGCTTTCTCAAAAGCGTCAAAAT GTCTACCGGCAGTCGCAATTTGGGGGCATAACTAACCTTG

IX-R CTGTATTGAAGATCATCAAATACATACTTCATTAACAGAA CCAGAATTACAGAAGACGATTATTGCAACTGAAGCTCAAT
379

(continued)
Table 1
380

(continued)

TEL Forwarda Reversea

X-L GGATTTCTCCTTGATTAGATTAAACATCCGTGTTGGATAG AGCTCAAACGGTCGGAAGATCTCAGCAGAGGTCTATCCAG
X-R CGCTACTTGAGAAAACTATATGGTTGTACGAAAAAGAAAT ATAGTAATGAGATTAATAAGGCCCAAAGAATATTTTGGCC
XI-L GCAAATTAGAATACATAGGATATTATGATGTGCACCTAAA TAACGTTTGATATGTAAGTTCTAAGTATGTTTCTCTATAT
XI-R TGACATAACAGGATAGGATAAACATGCTCGATAGACCCAC TAAAAAGTACGACACAGCTATATGCGTCGCAAGTTCAAAT
XII-L CTGGAAATCTGGTTCGTCTTCAACAAACGTCAGGAACTTT AAAAAATTGTCGTTTCCGTCATGGTCGGGAAAAATGTACA
XII-R GATGCCGTGAAAGCTTTATTGGTGTCGTCTTGTGCTTGTA ACGCCGTTGGTGTCATCAAATATATCTAAATCCCTTGCAG
XII-Rb AGCGCAGATTCTGCTTCTTCTCAATAGAGTAGCTTAATTA GATTGTCCAGTTTCCGTCTTATCATCATCTAAGAATGTAA
Roxanne Oshidari and Karim Mekhail

XIII-L GTCCAGTTTCCGTCTTATCATCATCTAAGAATGTAATAAT GCATAGCGCAGATTCTGCTTCTTCTCAATAGAGTAGCTTA

XIII-R CGCATAATGGAAATAGCTTCCTTAAAGGATGGAGATAAGG GAGAAAGCTCCCAAGTACTTCCTTGTTCTTGTCATTTTTT
XIV-L GTAACGCCATCGCAACCTTCTTCAATAAAGTTTCTTACAG ATGATCAGGTTGAGCCGATGCGGTTGCTTGAATGTGGCCC
XIV-R CGAGAGCTTTGACCAAAAAGTGTCCCGACTTTATAATGAG CCAAAGCATACCATATAACTCCTATGAATTAAGCGGCGGT
XV-L CGTTCCAACATCAATACCTCTAGAATTTCCATTTTGCGTT AAAACTAGGATCAACATCAGGATAGTTACTACCGTTAGCC
XV-R TATTCCTACATGTTTCTGAATACCGCGAAGGGCTGTCTGG GGAAGTTCCCTCATGTACTGCCGAAAAGTTGCGTATTCAA
XVI-L GGCAATATCAACCACTGTAGGCATATAAAAGCAGTCGTTC TTGCATGCAAGGCTCAGTTTGATAACCTTTGGGAAGAATT
XVI-R CATTCTCAGATGATGATAAGACGGAAACTGGACAATCTTT ATTGTTATGCTTTTTGACAAGAAATCCATCAATATAAACA
a 0
These sequences should be added to the 5 end of the following URA3 amplifying primers harboring cut sites (underlined): Forward: TAGGGATAACAGGGTAAT-
CACGCTTTTCAATTCAATTCATC, Reverse: TAGGGATAACAGGGTAATTTAGTTTTGCTGGCCGCATCTTC
Sequences are designed for integration 3–5 kb from the telomere, however in the event that this region is entirely occupied by an ORF, the bindicates an alternative set of sequences
further from the telomere. They still reside within the subtelomeric region.
 Assays to Study Repair of Inducible DNA Double-Strand Breaks at Telomeres 381

5. Prepare 1:10 serial dilution in sterile 1.5 mL microcentrifuge

tubes by mixing 100 μL of the 5.0  107 cells/mL tube with
900 μL of water to achieve an effective concentration of
5.0  106 cells (see Note 2). Repeat this dilution procedure
four more times to generate 5.0  105 to 5.0  102 cells/mL
samples (Fig. 2a-i).
6. Plate 200 μL of the 5.0  102 cells/mL suspension on four SC-
LEU(GLU) plates (~100 cells per plate) (see Note 3) (Fig. 2a-ii).
7. Plate 200 μL of 5.0  103 and/or 5.0  104 cells/mL suspen-
sion on four SC-LEU(GAL) plates (~1000 and/or
10,000 cells/plate).
8. Incubate the plates at 30
C for 7 days.
9. Calculate the ratio of the number of colonies growing over the
number of cells plated and average the ratios for both glucose
and galactose plates. Standard workflow and example calcula-
tions are shown (Fig. 2a).
10. For a final DSB survival rate, divide the average ratio of galac-
tose plated cells over that of glucose (Fig. 2a-i).
11. Perform t-test analyses to evaluate statistical support where
applicable.

3.3 Assessing DSB 1. Grow 3 mL of S. cerevisiae culture in SC-LEU(GLU) medium

Induction Efficiency overnight at 30
C to saturation.
2. Inoculate 300 μL of saturated culture into 5 mL of SC-LEU
(GLU) medium and 5 mL of SC-LEU(GAL) and grow at
30
C for 3 h. or until cells reach log phase.
3. Pellet 5.0  107 cells at 18,000  g.
4. Wash pellets and resuspend in 1 mL of water, then add the
225 μL cell suspension to the first well of a 96-well plate.
5. Prepare 1:10 serial dilutions by mixing 25 μL of the
5.0  107 cells/mL well with 225 μL of water in the next
well to achieve an effective concentration of 5.0  106 cells/
mL. Repeat four times to achieve a final dilution of
5.0  102 cells/mL (Fig. 2b-i).
6. Spot 3 μL of each dilution on SC and SC-URA plates (Fig. 2b-i).
7. Incubate plates at 30
C for 5–7 days.
8. Qualitatively and quantitatively determine DSB induction effi-
ciency (Fig. 2b-ii).

3.4 DNA Repair 1. To determine the repair mechanism employed by cells to sur-
Pathway Choice vive the induced DSB, perform colony PCR on colonies sur-
Determination viving on galactose-containing plates using primer pairs that
bind just outside the DSB site (Fig. 2c-i; Table 2).
382 Roxanne Oshidari and Karim Mekhail

A (i) (ii)

2
10

10

10

5. 0

5. 0
10
SC-LEU(GLU)

1
0x

0x

0x

0x

0x

0x
5.

5.

5.

5.
21 colonies
for 100
cells plated

SC-LEU SC-LEU
(GLU) (GAL)

SC-LEU(GAL)
Colonie Cells Colonie Cells
count plated count plated
9 100 8 1,000
10 100 7 1,000
75 colonies
12 100 63 10,000
for 10,000
9 100 65 10,000 cells plated
Totals 40 400 143 22,000
CFU 40/400 = 0.01 143/22,000 = 0.0065
DSB survival 0.0065/0.01 = 0.065

B (i) (ii) C
From (i)
SC-LEU (GLU or GAL) P1
SC
liquid cultures
from GLU cs URA3 cs

from GAL P2
7

2
10

5. 0

5. 0

5. 0

5. 0
10

(ii)
1

1
0x

0x

0x

0x

0x

0x
5.

5.

SC-URA AC1-3
from GLU
from GAL ?
P2
cs URA3 cs
DSB induction efficiency
~100%
SC SC-URA

Fig. 2 Standard workflow of the Inducible subtelomeric DSB repair and DSB induction efficiency assays. (a) (i)
Begin with performing a 1:10 serial dilution to a final concentration of 5.0  102 cells/mL. Plate on SC-LEU
(GLU) and SC-LEU(GAL) media and count colonies after 7 days of growth. Calculate colony forming units (CFUs)
and a final survival rate. (ii) An example of plated cells on SC-LEU(GLU) and SC-LEU(GAL). (b) (i) After culturing
in liquid SC-LEU(GLU) and SC-LEU(GAL) media, perform serial dilutions as described previous and spot 3 μL of
each dilution on SC(GLU) and SC-URA(GLU) plates. (ii) Example of DSB induction efficiency plating and the
subsequent determination of induction efficiency. (c) PCR primers for DNA repair pathway choice determina-
tion. The sequences of the primers shown here are included in Table 2 and can be used for subTEL XI-L. (i) For
a subtelomeric DSB site, design two primers just outside the I-SceI URA3 cassette. PCR product with the
expected size indicates an incomplete DSB while a smaller amplicon suggests NHEJ and no amplicon
indicates BIR or a similar pathway. (ii) BIR survivors can be further analyzed using a generic AC1–3 primer
and the more internal P2 primer. Reactions would not yield an amplicon if the BIR event engaged a very long
chromosome arm
 Assays to Study Repair of Inducible DNA Double-Strand Breaks at Telomeres 383

Table 2
Primers used to screen survivors for type of DNA repair employed

Primer Sequence
P1 CTGAGTCTGCACTAGACAAT
P2 ATCTTGATCTCAAAAGCACC
AC1-3 ACCACACACCCACCAC

(a) To perform colony PCR, pick up a small portion of the

colony with a pipette tip and smear along the side of a
PCR tube.
(b) Microwave PCR tubes on high for 1 min 30 s, allow
cooling for 1 min, and repeat.
(c) Carry out preferred PCR protocol with the primers
described in step 1.
2. Electrophorese PCR products in a 1% TAE-agarose gel.
3. Calculate the expected amplicon size, taking into account the
inserted cassette. A PCR product of that size would indicate an
incomplete DSB. A visibly smaller amplicon indicates NHEJ.
No amplicon indicates the BIR or similar pathway.
4. For the system focusing on subTELXI-L, perform colony PCR
with the primers shown in Fig. 2c-i and Table 2. No PCR
product indicates the BIR or similar pathway, while an ampli-
con of ~0.7 kb indicates NHEJ. The production of a ~2 kb
product indicates an incomplete DSB.
5. BIR survivors can be further analyzed using a generic AC1–3
forward primer and the more internal reverse primer used in
step 1 (Fig. 2c-ii; Table 2). The majority of reactions should
produce no amplicon. For subTEL-XI-L, perform the PCR
with the generic AC1–3 primer and P2 (Table 2; see Notes 4
and 5).

4 Notes

1. The volume used for inoculation may require adjustment to

compensate for varying rates of growth among strains. Over-
night cultures may be at different levels of saturation. Alterna-
tively, some mutants may reach log phase much slower than
wild-type cells. All strains should be in log phase at the same
time.
2. Between dilutions, mix the cell suspension by pipetting up and
down to ensure that cells have not settled to the bottom of
tubes. This should also be done before plating.
384 Roxanne Oshidari and Karim Mekhail

3. The 5  102 dilution is generally appropriate for wild-type and

mutants with normal growth. However, more concentrated
dilutions may be required for slower growing mutants. Differ-
ent dilutions can be plated during this assay to establish ideal
conditions. This should also be done when plating on
galactose-containing media.
4. When designing primers, ensure that only the AC1–3 primer
falls within repetitive elements/regions.
5. PCR-products can also be sequenced to confirm the DNA
repair pathway used by surviving cells.

Acknowledgments

We thank Mekhail lab members for comments. K.M. is supported

by grants from the Canadian Institutes of Health Research (CIHR),
the Ontario Ministry of Research and Innovation Early Researcher
Award (MRI-ERA), and the Canada Research Chair (CRC) in
Spatial Genome Organization. The authors have no conflict of
interest.

References

1. Pfeiffer V, Lingner J (2013) Replication of tel- 8. Taddei A, Gasser SM (2012) Structure and
omeres and the regulation of telomerase. Cold function in the budding yeast nucleus. Genet-
Spring Harb Perspect Biol 5(5):a010405. doi: ics 192(1):107–129. doi:192/1/107 [pii].
cshperspect.a010405 [pii]. doi:10.1101/ doi:10.1534/genetics.112.140608
cshperspect.a010405 9. Chan JN, Poon BP, Salvi J, Olsen JB, Emili A,
2. Greider CW, Blackburn EH (1985) Identifica- Mekhail K (2011) Perinuclear cohibin
tion of a specific telomere terminal transferase complexes maintain replicative life span via
activity in Tetrahymena extracts. Cell 43(2 Pt roles at distinct silent chromatin domains.
1):405–413. doi:0092-8674(85)90170-9 [pii] Dev Cell 20(6):867–879. doi:S1534-5807
3. Shampay J, Szostak JW, Blackburn EH (1984) (11)00206-1 [pii]. doi:10.1016/j.devcel.
DNA sequences of telomeres maintained in 2011.05.014
yeast. Nature 310(5973):154–157 10. Taddei A, Hediger F, Neumann FR, Bauer C,
4. Chan SW, Blackburn EH (2003) Telomerase Gasser SM (2004) Separation of silencing from
and ATM/Tel1p protect telomeres from non- perinuclear anchoring functions in yeast Ku80,
homologous end joining. Mol Cell 11 Sir4 and Esc1 proteins. EMBO J 23
(5):1379–1387 (6):1301–1312. doi:10.1038/sj.emboj.
5. Gottschling DE, Aparicio OM, Billington BL, 7600144. 7600144 [pii]
Zakian VA (1990) Position effect at S. cerevi- 11. Bupp JM, Martin AE, Stensrud ES, Jaspersen
siae telomeres: reversible repression of Pol II SL (2007) Telomere anchoring at the nuclear
transcription. Cell 63(4):751–762. doi:0092- periphery requires the budding yeast Sad1-
8674(90)90141-Z [pii] UNC-84 domain protein Mps3. J Cell Biol
6. Mekhail K, Moazed D (2010) The nuclear 179(5):845–854. doi:jcb.200706040 [pii].
envelope in genome organization, expression doi:10.1083/jcb.200706040
and stability. Nat Rev Mol Cell Biol 11 12. Schober H, Ferreira H, Kalck V, Gehlen LR,
(5):317–328. doi:nrm2894 [pii]. doi:10. Gasser SM (2009) Yeast telomerase and the
1038/nrm2894 SUN domain protein Mps3 anchor telomeres
7. Moazed D (2001) Common themes in and repress subtelomeric recombination.
mechanisms of gene silencing. Mol Cell 8 Genes Dev 23(8):928–938. doi:23/8/928
(3):489–498 [pii]. doi:10.1101/gad.1787509
 Assays to Study Repair of Inducible DNA Double-Strand Breaks at Telomeres 385

13. Hediger F, Neumann FR, Van Houwe G, 20. Lydeard JR, Jain S, Yamaguchi M, Haber JE
Dubrana K, Gasser SM (2002) Live imaging (2007) Break-induced replication and
of telomeres: yKu and Sir proteins define telomerase-independent telomere maintenance
redundant telomere-anchoring pathways in require Pol32. Nature 448(7155):820–823.
yeast. Curr Biol 12(24):2076–2089. doi: doi:nature06047 [pii]. doi:10.1038/
S0960982202013386 [pii] nature06047
14. Torres-Rosell J, Sunjevaric I, De Piccoli G, 21. Therizols P, Fairhead C, Cabal GG, Genovesio
Sacher M, Eckert-Boulet N, Reid R, Jentsch A, Olivo-Marin JC, Dujon B, Fabre E (2006)
S, Rothstein R, Aragon L, Lisby M (2007) Telomere tethering at the nuclear periphery is
The Smc5-Smc6 complex and SUMO modifi- essential for efficient DNA double strand break
cation of Rad52 regulates recombinational repair in subtelomeric region. J Cell Biol 172
repair at the ribosomal gene locus. Nat Cell (2):189–199. doi:jcb.200505159 [pii].
Biol 9(8):923–931. doi:10.1038/ncb1619 doi:10.1083/jcb.200505159
15. Mekhail K, Seebacher J, Gygi SP, Moazed D 22. Nagai S, Dubrana K, Tsai-Pflugfelder M,
(2008) Role for perinuclear chromosome teth- Davidson MB, Roberts TM, Brown GW, Varela
ering in maintenance of genome stability. E, Hediger F, Gasser SM, Krogan NJ (2008)
Nature 456(7222):667–670. doi:nature07460 Functional targeting of DNA damage to a
[pii]. doi:10.1038/nature07460 nuclear pore-associated SUMO-dependent
16. Chiolo I, Minoda A, Colmenares SU, Polyzos ubiquitin ligase. Science 322(5901):597–602.
A, Costes SV, Karpen GH (2011) Double- doi:322/5901/597 [pii]. doi:10.1126/sci
strand breaks in heterochromatin move outside ence.1162790
of a dynamic HP1a domain to complete recom- 23. Ricchetti M, Dujon B, Fairhead C (2003) Dis-
binational repair. Cell 144(5):732–744. tance from the chromosome end determines
doi:10.1016/j.cell.2011.02.012 the efficiency of double strand break repair in
17. Jakob B, Splinter J, Conrad S, Voss KO, Zink subtelomeres of haploid yeast. J Mol Biol 328
D, Durante M, Lobrich M, Taucher-Scholz G (4):847–862. doi:S0022283603003152 [pii]
(2011) DNA double-strand breaks in hetero- 24. Chapman JR, Taylor MR, Boulton SJ (2012)
chromatin elicit fast repair protein recruitment, Playing the end game: DNA double-strand
histone H2AX phosphorylation and relocation break repair pathway choice. Mol Cell 47
to euchromatin. Nucleic Acids Res 39 (4):497–510. doi:S1097-2765(12)00656-9
(15):6489–6499. doi:10.1093/nar/gkr230 [pii]. doi:10.1016/j.molcel.2012.07.029
18. Chung DK, Chan JN, Strecker J, Zhang W, 25. Dunham MJ, Gartenberg MR, Brown GW
Ebrahimi-Ardebili S, Lu T, Abraham KJ, Dur- (2015) Methods in yeast genetics and geno-
ocher D, Mekhail K (2015) Perinuclear tethers mics: a CSHL course manual. CSHL, New
license telomeric DSBs for a broad kinesin- and York
NPC-dependent DNA repair process. Nat 26. Lettier G, Feng Q, de Mayolo AA, Erdeniz N,
Commun 6:7742. doi:10.1038/ncomms8742 Reid RJ, Lisby M, Mortensen UH, Rothstein R
19. Chung DK, Mekhail K (2015) Repair by a (2006) The role of DNA double-strand breaks
molecular DNA ambulance. Oncotarget 6 in spontaneous homologous recombination in
(23):19358–19359. doi:10.18632/ S. cerevisiae. PLoS Genet 2(11):e194. doi:10.
oncotarget.5140 1371/journal.pgen.0020194
 Chapter 27

Telomerase RNA Imaging in Budding Yeast and Human Cells

by Fluorescent In Situ Hybridization
David Guérit, Maxime Lalonde, and Pascal Chartrand

Abstract
Telomerase, the enzyme that elongates telomeres in most eukaryotes, is a ribonucleoprotein complex
composed of a reverse transcriptase catalytic subunit (TERT in human, Est2 in the budding yeast S.
cerevisiae), regulatory factors and a noncoding RNA called hTERC (in human) or TLC1 (in budding
yeast). Telomerase trafficking is a major process in the biogenesis and regulation of telomerase action at
telomeres. Due to its higher signal-to-noise ratio, imaging of the telomerase RNA moiety is frequently used
to determine telomerase intracellular localization. Here we describe how to image telomerase RNA in
human and yeast cells using fluorescence in situ hybridization.

Key words Telomerase, Fluorescent in situ hybridization, RNA localization, TLC1 RNA, hTERC

1 Introduction

Telomerase biogenesis and action at telomeres is linked to its

cellular trafficking and localization [1]. The link between telome-
rase RNA maturation and trafficking has been well studied in the
budding yeast S. cerevisiae. The budding yeast telomerase RNA
(TLC1) life cycle includes many events that occur at various subcel-
lular localizations. As part of its biogenesis, this RNA goes to the
nucleolus for its 50 end maturation, and is exported to the cyto-
plasm for assembly of the holoenzyme [2–4]. Telomerase is then
reimported in the nucleus, and its nuclear accumulation depends
on the yKu complex, which interacts with the TLC1 RNA [2, 5, 6].
In human cells, telomerase RNA (hTERC or hTR) localization
depends on its maturation status. After transcription, hTERC
3’end processing possibly occurs in the nucleolus, since 30 end
extended hTERC precursors can be detected in this subnuclear
compartment [7]. hTERC is a small Cajal body RNA (scaRNA)
which accumulates in Cajal bodies in human cancer cells [8, 9].
This localization depends on the interaction between a CAB box
element in hTERC with the Cajal body protein TCAB1 [8, 10].

Marco Muzi-Falconi and Grant W. Brown (eds.), Genome Instability: Methods and Protocols, Methods in Molecular Biology,
vol. 1672, DOI 10.1007/978-1-4939-7306-4_27, © Springer Science+Business Media LLC 2018

387
388 David Guérit et al.

Despite the fact that other techniques exists, such as live cell
imaging [11], RNA fluorescence in situ hybridization (RNA-FISH)
remains the technique of reference to study telomerase RNA local-
ization. As a ribonucleoprotein complex, telomerase localization
can be studied using RNA-FISH by imaging its RNA component.
Properly designed FISH probes can push the sensitivity to image
single telomerase RNA molecules, which is not always possible by
immunofluorescence on proteins [12]. It is also important to note
that RNA-FISH is noninvasive and does not require genetic modi-
fication of its target, which allows the study of endogenous telome-
rase RNA. In budding yeast, a mix of oligonucleotide probes
labeled with an organic dye can be used to detect the endogenous
TLC1 RNA [2]. A standard inverted epifluorescence microscope
equipped with a CCD camera is enough to detect single molecules
of TLC1 RNA. Regarding human cells, the small size of hTERC
RNA (450 nt) and its high GC content, limit the number of probes
that can be designed to detect this RNA, which results in low
signal-to-noise ratio and therefore prevents the detection of single
molecules. Therefore, only the accumulation of hTERC RNA in
large foci or in Cajal bodies can be detected [8, 9].
RNA-FISH on telomerase RNA can easily be combined with
immunofluorescence (IF) to investigate its colocalization with dif-
ferent proteins or telomeres. Indeed, several studies successfully
described the colocalization between telomerase RNA and Cajal
bodies in human cancer cells, or between telomerase RNA and
telomeres in both yeast and human cells using simultaneous
RNA-FISH and immunofluorescence [2, 10, 13–16].

2 Materials

2.1 Preparation of 1. Glass coverslips (22  22 mm #1.5).

Poly-L-Lysine-Coated 2. Six 35 mm well plastic plates.
Coverslips
3. Poly-L-lysine, diluted to 0.01% in DEPC water.

2.2 Probe Design and The probes contain amino-allyl deoxythymidine (N-allyl-dT) mod-
Preparation ified nucleotides, allowing covalent dye coupling to the probe.
Oligonucleotides may vary in length and N-allyl-dT modification
is incorporated every ten base pair for >20 bases oligonucleotide,
and only at the 50 or 30 end for oligonucleotide of about 20 bases.
Tables 1 and 2 include the sequences of the probes used for TLC1
and hTERC FISH in yeast and human cells, respectively.
1. G25 Mini quick spin columns (Roche Life Science).
2. Cy3 or Cy5 monoreactive dye pack (GE Healthcare).
3. Sodium bicarbonate buffer 0.1 M pH 8.8: prepare 0.1 M of
sodium carbonate (Na2CO3) and sodium bicarbonate
 Telomerase RNA Imaging in Budding Yeast and Human Cells by Fluorescent. . . 389

Table 1
Sequences of the modified oligonucleotides for TLC1 FISH

TLC1-1 t*gcgcacacacaagcat*ctacactgacaccagcat*actcgaaattctt*tg
TLC1-2 ct*aataaacaatt*agctgtaacatt*tgtgtgtggggt*gtggtgatggt*aggc
TLC1-3 *ccagagttaacgat*aagatagacat*aaagtgacagcgct*tagcaccgt*
TLC1-4 ttacgt*tcttgatctt*gtgtcattgtt*cagttactgat*cgcccgcaaacct*
TLC1-5 tgcat*cgaaggcat*taggagaagt*agctgtgaat*acaacaccaagat*tca
*Amino allyl modified-T

Table 2
Sequences of the modified oligonucleotides for hTERC FISH

hTERC-1 t*gcgcgcggggagcaaaagcacggcgcct*acgcccttctcagtt*agggttagaca
hTERC-2 gct*gacattttt*tgtttgctct*agaatgaacggt*ggaaggcggcaggccgaggct*t
*Amino allyl modified-T

(NaHCO3) solutions (106 mg Na2CO3 and 84 mg NaHCO3

in 10 mL of DEPC treated water). Add Na2CO3 to NaHCO3
until a pH of 8.8 is reached.

2.3 Preparation 1. 4,6-diamidino-2-phenylindole (DAPI).

of Mounting Media 2. p-phenylenediamine.
3. Glycerol.

2.4 Fluorescent In 1. 32% paraformaldehyde, electronic microscopy grade, single-

Situ Hybridization on use sealed ampules (Electron Microscope Sciences) (see
Yeast Cells Note 1).
2. Phenylmethylsulfonyl fluoride (PMSF). Prepare a 17.5 mg/
mL stock solution in ethanol, keep at 20  C.
3. Vanadyl Ribonucleoside Complex (VRC) 200 mM (New Eng-
land Biolabs #S1402S).
4. β-mercaptoethanol.
5. 70% ethanol solution prepared with DEPC water.
6. Lyticase from Arthrobacter luteus (Sigma-Aldrich #L2524).
Reconstituted in 1 PBS at 25 U/μL. Store 30 μL (600 U)
aliquot at 20  C.
7. Protease inhibitor cocktail tablets (Roche #04 693 132 001).
Dissolve 1 tablet in 1 mL DEPC water for a 50 concentrated
solution.
8. 1 Buffer B: 1.2 M Sorbitol, 0.1 M potassium phosphate
pH 7.5 (200 mL needed per sample, can be kept for a few
month at 4  C).
390 David Guérit et al.

9. 28 mM β-mercaptoethanol, 0.06 mg/mL PMSF, protease

inhibitor cocktail 1. Prepare fresh.
10. Formamide.
11. Escherichia coli tRNA (e.g., Roche #10 109 541 001).
12. Single-stranded DNA from salmon testes (ssDNA) 10 mg/mL
(e.g., Sigma #D9156).
13. Solution of E. coli tRNA and salmon sperm single-stranded
DNA in a 1:1 ratio (5 mg/mL each).
14. Sodium phosphate buffer 1 M, pH 7. Dissolve 11.9 g of
sodium phosphate in 50 mL of DEPC water, adjust pH to 7
and make up to 100 mL with DEPC water, can be kept at room
temperature.
15. Buffer F: 80% formamide, 10 mM sodium phosphate pH 7,
freshly made, 15 μL per sample.
16. Labeled probes, see Subheading 3.2.
17. Parafilm.
18. RNAse free BSA 20 mg/mL.
19. Coplin Jars for 22  22 mm coverslips.
20. Glass plate (16  20 cm).
21. Forceps.
22. 20 SSC. Dissolve 175.3 g NaCl and 88.2 g of sodium citrate
in 1 L of water. Adjust pH to 7 with HCl. Treat with 0.1%
DEPC overnight and autoclave. Keep at room temperature.
23. 2 SSC, 40% formamide in DEPC water, freshly made. Can be
stored at 4  C for 2 days.
24. Buffer H: 4 SSC, 20 mM VRC, 4 μg/μL BSA, 20 U RNAse
OUT, freshly made.
25. 2 SSC, from 20 SSC diluted in DEPC water.
26. 1 SSC, from 20 SSC diluted in DEPC water.
27. Triton X-100 (Sigma-Aldrich).
28. Frosted microscope slides (76  26  1 mm).
29. Transparent nail polish.
30. 20 PBS. Dissolve in 1 L water 2.88 g of sodium phosphate
dibasic (Na2HPO4), 160 g sodium chloride (NaCl), 4 g potas-
sium phosphate dibasic (KH2PO4) and 4 g potassium chloride
(KCl), adjust pH to 7.2 with sodium hydroxide (NaOH). Treat
with 0.1% DEPC overnight and autoclave. Store at room
temperature.
31. 2 SSC, 0.1% Triton X-100, freshly made. Can be stored at
4  C for 2 days.
 Telomerase RNA Imaging in Budding Yeast and Human Cells by Fluorescent. . . 391

2.5 Fluorescence In 1. Coplin jars.

Situ Hybridization on 2. 32% Paraformaldehyde, electronic microscopy grade, single-
Human Cells use sealed ampules (Electron Microscope Sciences) (see
Note 1).
3. 1 phosphate buffered saline (PBS), diluted from 20 PBS in
DEPC water, see Subheading 2.4, item 4.
4. Triton X-100 (Sigma-Aldrich).
5. Fixative solution: 3.2% paraformaldehyde in 1 PBS.
6. Permeabilization solution: 0.5% Triton X-100 in 1 PBS.
7. Formamide.
8. Dextran sulfate, make a 50% W/V solution in DEPC water.
9. Escherichia coli tRNA (Roche #10109541001). Dissolved at
10 mg/mL in DEPC water, store at 20  C.
10. Single-stranded DNA from salmon testes (ssDNA) 10 mg/mL
(e.g., Sigma-Aldrich #D9156).
11. Bovine serum albumin RNase free (e.g., Roche
#10711454001).
12. Vanadyl Ribonucleoside Complex (VRC) 200 mM (New Eng-
land Biolabs #S1402S).
13. 20 SSC, see Subheading 2.4, item 3.
14. 2 SSC in 40% formamide, store at 4  C for maximum 2 days.
15. Solution A, for ten coverslips: mix 80 μL of formamide with
20 μL of 20 SSC.
16. Solution B, for 10 coverslips: mix 20 μL of BSA with 40 μL of
50% dextran sulfate and 2 μL 200 mM of VRC, complete with
38 μL of DEPC water.
17. Large glass plate (16  20 cm).
18. Parafilm.
19. Formamide.
20. Probes as described Subheading 3.2.
21. 20 SSC, see Subheading 2.4, item 3.
22. Frosted microscope slides.
23. Mounting media, see Subheading 3.4.
24. Transparent nail polish.
25. Formamide.
26. 1 PBS, diluted from 20 PBS, see Subheading 2.4, item 4.
27. Tween 20.
28. 20 SSC, see Subheading 2.4, item 3.
29. Wash solution 1: 2 SSC, 40% formamide in DEPC water.
30. 1 PBS, 0.1% Tween 20.
392 David Guérit et al.

3 Methods

3.1 Preparation of Yeast cells do not adhere stably on glass or plastic. Coverslips must
Poly-L-Lysine-Coated be pretreated with a coating agent to assure that cells will stay on
Coverslips the coverslips during the hybridization and washing steps. We use
poly-L-lysine, but alternatively concanavalin A can be used.
At the opposite, human cells adhere easily to glass but repeated
washes during the process have the tendency to remove cells from
coverslips and to alter normal cell shape. This can be prevented by
using, as for yeast, poly-L-lysine-coated coverslips.
1. Coverslips are stripped using hydrochloric acid as follow: in a
1 L glass beaker, immerge a large number of coverslips with
about 250 mL of HCl 0.1 N and cover it with aluminum foil.
2. Boil 30 min in a fume hood.
3. Let the beaker cool down and discard the HCl.
4. Wash the coverslips ten times with DEPC water. Leave at least
100 mL of water in the beaker after the last wash.
5. Cover the beaker filled with the coverslips with aluminum foil
and autoclave it. Once autoclaved, the coverslips can be stored
at 4  C for several months. Alternatively, coverslips can be
stored in 50 mL tubes filled with 75% ethanol, this has the
advantage of a faster drying.
6. Place one coverslip in each well of a six-well plastic plate and let
them dry.
7. Dilute poly-L-lysine at 0.01% in water, filtrate through 0.22 μm
filter and apply 200–500 μL of diluted poly-L-lysine on each
coverslip. The solution should cover at least 75% of the cover-
slip surface.
8. Incubate at room temperature for 5 min to let the poly-L-lysine
adhere.
9. Aspirate the excess of poly-L-lysine and let dry completely. This
will take around 3–4 h. Coverslips may stick to the bottom of
the wells. During the washes, they will eventually get free.
10. Once dried, wash each coverslips with 3 mL sterile water three
times. These washes are important to avoid excess adherence of
yeast cells on the coverslips that will lead to aggregates forma-
tion instead of a single layer of individual cells.
11. After the last wash, rest each coverslip on the wall of the well to
be sure they do not get stuck at the bottom while drying.
Aspirate the excess and let dry completely.
12. The coverslips can be kept at room temperature for several
months inside the six well plates.
 Telomerase RNA Imaging in Budding Yeast and Human Cells by Fluorescent. . . 393

3.2 Probe Design and 1. Using a Speed Vac, lyophilize 10 μg of each probe in RNAse-
Preparation free Eppendorf tubes (do not heat).
2. Resuspend the probes in 35 μL of sodium carbonate buffer.
3. Prepare the cyanine dye: dissolve the content of one vial in
30 μL DEPC water. One vial can be used to label 2 probes.
See manufacturer instruction/ protocol for more details.
4. Add 15 μL of the reconstituted cyanine dye to each probe.
5. Incubate at room temperature for 16–24 h in the dark with
occasional vortexing.
6. Purify the probe using a G25 oligo column.
(a) Prepare the column: open the top of the column and
break the bottom seal. Remove the buffer by spinning
the column in a 2 mL tube at 2500  g for 1 min, discard
the flow through.
(b) Place the column in a new 1.5 mL RNAse free Eppendorf.
Apply the 50 μL labeled probe in the center of the column
and centrifuge at 2500  g for 4 min. Discard the column
and determine the labeling efficiency of the probes (see
Note 2).
7. Keep labeled probes at 80  C in opaque RNAse-free tubes.

3.3 Calculation of 1. Measure the optical density at 260 nm to determine DNA

Fluorophore concentration and at 552 nm (Cy3) or 650 nm (Cy5) for
Incorporation cyanine concentration. Usually a 1/50 dilution is needed and
sufficient to avoid signal saturation.
2. Calculate the incorporation efficiency using the molecular
extinction coefficient (MEC) of the oligonucleotide and cya-
nine dye. Given below is the calculation for Cy3 dye.
3. Cy3 labeling of probes by this method gives usually 70–90%
incorporation efficiency, meaning that at least 3 modified
nucleotides out of 5 have been labeled with Cy3. A reaction
having a labeling efficiency below 60% is insufficient and must
be redone.

A 552nm
½Cy3 ¼
Cy3 MEC
ðA260nm  0:08ðA552nmÞÞ
½Oligo ¼
Oligo MEC
½Cy3
Incorporation efficiency ¼
½Oligo

*The MEC of oligonucleotides is the sum of the MEC of each

nucleotide.
394 David Guérit et al.

MEC (M1.cm1)
Cy3 150,000 (552 nm)
Cy5 250,000 (650 nm)
Adenosine 15,200 (260 nm)
Cytidine 7050 (260 nm)
Guanine 12,010 (260 nm)
Thymidine 8400 (260 nm)

3.4 Mounting Media Mounting media can be easily made; it contains p-

Preparation phenylenediamine as antifading agent. It is also convenient to add
a DNA staining dye such as DAPI or Hoechst.
1. Dissolve 10 mg of p-phenylenediamine in 1 mL of 10 SSC.
2. Add DAPI at the final concentration of 1 μg/mL.
3. Once completely dissolved, add 9 mL of glycerol, store at
20  C in opaque tubes. Discard when it becomes dark brown.

3.5 Fluorescent In Fixation of the cells is performed by cross-linking with paraformal-

Situ Hybridization on dehyde (PFA). PFA is a small chemical compound that diffuses
Yeast Cells easily and rapidly inside the cells, and which cross-links protein–-
protein, protein–nucleic acid, and nucleic acid–nucleic acid inter-
3.5.1 Fixation and actions. Once the cells are fixed, keep them on ice to avoid
Spheroplasting of Yeast proteolytic and RNase activities. Yeast cells have a cell wall outside
Cells their cell membrane that needs to be degraded to allow the probes
to diffuse inside the cell. Several enzymes can be used to degrade
this cell wall, such as lyticase, chitinase, zymolase, and gluculase. We
typically use the lyticase from Arthrobacter luteus that hydrolyzes
poly-β-(1 ! 3)-glucose in the cell wall glucan.
1. Grow yeasts in 50 mL of media until mid-log phase growth is
reached (OD600 between 0.2 and 0.4, starting from 0.05 to
0.1) (see Note 3).
2. Add 6.25 mL of 32% paraformaldehyde directly in the culture
flask. Incubate at room temperature for 45 min with gentle
shaking (see Note 4).
3. Transfer the fixed cells in a 50 mL tube and centrifuge at 4  C at
2500  g for 4 min.
4. Wash cell pellet three times with ice-cold buffer B to remove
the culture media and the PFA. Do not resuspend the cell pellet
when washing.
5. Resuspend gently the cell pellet in 1 mL of spheroplasting
buffer by pipetting (no vortex).
6. Add 600 U of lyticase and mix by pipetting.
 Telomerase RNA Imaging in Budding Yeast and Human Cells by Fluorescent. . . 395

7. Incubate for 8 min in a 30  C water bath. Invert the tube each

minute to avoid sedimentation of the cells (see Note 5).
8. Pellet the cells at 4  C at 2500  g for 4 min.
9. Wash the cell pellet with 1 mL of ice-cold buffer B.
10. Resuspend gently the cell pellet in 750 μL of ice-cold buffer B.
11. Gently pipet 100 μL of the spheroplasting reaction on each
prepared coverslips in a six well plate.
12. Let cells adhere for 30 min at 4  C.
13. Very gently, wash the coverslips with 3 mL of ice-cold buffer B
(see Note 6).
14. Dehydrate the spheroplasts by adding 5 mL of ice-cold 70%
ethanol prepared with DEPC water. Incubate for at least
30 min at 20  C before performing the hybridization. Spher-
oplasts can be kept at 20  C for a maximum of 2 months (see
Note 6).

3.5.2 Probe Preparation 1. Prepare two coverslips per experiment in case one of the two
breaks during handling. The following protocol stands for one
coverslip and can be scaled up.
2. In a 1.5 mL tube, add 1 ng of each labeled probe and complete
up to 10 μL with DEPC water.
3. Add 4 μL of a 5 mg/mL of a solution 1:1 of E. coli tRNA and
salmon sperm single-stranded DNA.
4. Lyophilize in a speed vac.
5. Resuspend the lyophilized probes in 12 μL of buffer F.
6. Keep at room temperature in the dark.

3.5.3 Hybridization 1. Cover a glass plate with a layer of Parafilm. Use the back part of
a forceps to make the Parafilm adhere to the glass. At this point,
all manipulation should be done in dim light.
2. Place the coverslips in a Coplin jar. Make sure the yeast-coated
face of each coverslips face the same side (see Note 7).
3. Wash two times with 8 mL of 2 SSC for 5 min at room
temperature with gentle shaking to rehydrate the cells.
4. Incubate in 2 SSC, 40% formamide for 5 min at room tem-
perature (see Note 8).
5. Heat the probes at 95  C for 3 min.
6. Dilute the probes with 12 μL of buffer H.
7. Drop the probes (24 μL) on the layer of Parafilm on the glass
plate.
8. Lay the coverslip on the probes, with the spheroplasts side
facing the probes (see Note 9).
396 David Guérit et al.

9. Cover the coverslips with another layer of Parafilm. Use the

back of the forceps to seal the two layer of Parafilm together,
making a humidified hybridization chamber. Wrap the glass
plate with aluminum foil and incubate overnight at 37  C.

3.5.4 Washes and 1. All manipulation should be done in dim light. Cover the
Coverslips Mounting Coplin jar in aluminum foil during incubations.
2. Gently put back the coverslips from the glass plate in a Coplin
jar with the spheroplasts side facing the same direction.
3. Wash two times with 8 mL of preheated 2 SSC, 40% form-
amide at 37  C for 15 min.
4. Wash with 2 SSC, 0.1% Triton X-100 for 15 min at room
temperature with gentle shaking.
5. Wash two times with 1 SSC 15 min at room temperature with
gentle shaking (see Note 10).
6. Incubate in 8 mL of 1 PBS, 1 ng/mL DAPI at room temper-
ature for 2 min.
7. Drop 7 μL of mounting media on a microscope slide.
8. Gently take the coverslip and remove the excess of liquid at the
bottom of the coverslip with an absorbent tissue.
9. Lay the coverslip on the mounting media, with the spheroplast
side facing the mounting media.
10. Let stand for 2–3 min and then seal the coverslip by applying
nail polish on the sides. Let dry. Microscope slides can be
stored in the dark at 20  C for up to 3 months, but image
acquisition should be performed as soon as possible to assure
good image quality (Fig. 1).

Fig. 1 FISH on budding yeast TLC1 RNA. Fluorescent in situ hybridization on budding yeast cells with TLC1
specific probes. Endogenous TLC1 RNA foci are visible in the nucleus of yeast cells. Acquisition was performed
with a Nikon upright epifluorescence microscope and a CoolSnap Photometrics CCD camera. DAPI: nuclear
staining, DIC: differential interference contrast. Scale bar ¼ 1 μm
 Telomerase RNA Imaging in Budding Yeast and Human Cells by Fluorescent. . . 397

3.6 Fluorescent In To ensure a good cell adherence and a normal shape, cell plating
Situ Hybridization on should be made the day before the FISH. Also it is important to
Human Cells reach around 70–80% of cell confluency the day of the fixation, thus
the number of cells to plate must be determined empirically for
each cell type by plating several cell densities.

3.6.1 Cell Fixation/ All manipulations are done at room temperature and under a fume
Permeabilization hood since paraformaldehyde is a toxic chemical.
1. Remove coverslips from the 6 well tissue culture plates contain-
ing the cells and place them in a Coplin jar (see Note 11).
2. Fill the Coplin jar with 1 PBS for 5 min, gently rock in your
hands few seconds to remove any trace of serum.
3. Remove the PBS and place 7–8 mL of fixative solution, let
stand for 15 min.
4. Pour off the fixative solution and wash two times with 7–8 mL
of 1 PBS to ensure a complete removal of paraformaldehyde.
5. Remove the last wash solution, then permeabilize the cells with
7–8 mL of permeabilization solution for 5 min.
6. Wash one time with 1 PBS.
7. At this point, the cells can be stored in 1 PBS at 4  C for a few
days. For long-term storage, remove the PBS and add 2 mL of
70% ethanol. Keep at 4  C, which allows the cells to remain
usable for a few weeks.

3.6.2 Probe Preparation 1. All the following steps should be performed in dim light in
order to prevent dye photobleaching a thus a loss of signal.
2. For ten coverslips, mix 10 μL of 10 mg/mL ssDNA and 10 μL
of 10 mg/mL E. coli tRNA in a 1.5 mL Eppendorf tube. Add
200 ng of probe per target RNA. If several probes are used, a
total of 200 ng of the pool of probes should be used.
3. Lyophilize the probes using a speed vac. It should take around
20 min.
4. Add 100 μL of solution A, ensure a proper solubilization by
vortexing the solution, then place the tube in a heater block at
95  C for 3 min.
5. Chill the tube on ice for about 5 min.
6. Add 100 μL of solution B, mix well by vortexing.
7. Probes can be stored on ice for few hours.

3.6.3 Hybridization Hybridization is performed in a buffer containing formamide, in a

moist chamber made of Parafilm. If detection of endogenous
hTERC is required, hybridization should be performed overnight.
If hTERC is overexpressed in the cells, incubation time can be
reduced to 2 h.
398 David Guérit et al.

1. Take coverslips prepared Subheading 3.6.1, they are already in

a Coplin jar.
2. Remove the 1 PBS or the 70% ethanol and fill the Coplin jar
with 7–8 mL of wash solution, incubate at room temperature
for 5 min.
3. Meanwhile clean the glass plate and lay on a piece of Parafilm.
Make the Parafilm adhere to the glass plate by scrubbing it with
a forceps or a pencil.
4. Using a marker, criss-cross the Parafilm to define zones for
coverslips and note the conditions.
5. Drop 20 μL of each probe onto the Parafilm
6. One by one, remove coverslips from the Coplin jar, drain excess
of buffer on a paper and carefully deposit on the Parafilm, cells
facing the Parafilm. Avoid contact between the coverslips (see
Note 12).
7. Lay down a new piece of Parafilm onto the first one, seal using
forceps or pencil to form a perfect moist chamber.
8. Wrap the glass plate in aluminum foil to protect from light and
place at 37  C overnight.

3.6.4 Washes and 1. Prewarm Coplin jars and 50 mL of wash solution at 37  C for at
Coverslip Mounting least 30 min.
2. Take the glass plate containing coverslips out of the incubator
and move one by one the coverslips into the prewarmed Coplin
jars. Incubate for 30 min at 37  C.
3. Remove wash solution and add a new volume of prewarmed
wash solution, incubate at 37  C for another 30 min.
4. Remove wash solution and fill the Coplin jar with 1 PBS,
0.1% Tween 20 containing 1 μg/mL of DAPI. Incubate at
room temperature for 5 min.
5. Remove wash solution and fill the Coplin jar with 1 PBS,
0.1% Tween 20. Incubate at room temperature for 5 min.
6. Remove wash solution and fill the Coplin jar with 1 PBS.
Incubate at room temperature for 5 min (see Note 10).
7. Annotate microscope slides and drop 20 μL of mounting media
for each coverslip.
8. Remove coverslips from the Coplin jar, rapidly drain the PBS
on a paper towel and apply onto the microscope slide, cell
facing the mounting media.
9. Seal coverslips with few drops of nail polish, let dry in the dark.
10. To remove any traces of salt, coverslips can be whipped with a
microfiber paper soaked with a drop of lens cleaning solution.
11. Slides are now ready to image. They can be kept up to a week in
the dark at 4  C or a few month at 20  C. Human TERC foci
 Telomerase RNA Imaging in Budding Yeast and Human Cells by Fluorescent. . . 399

Fig. 2 FISH on human telomerase (hTERC) RNA. (a) HeLa cells (hTERC positive cells) and Wi38-VA13 cells
(hTERC negative cells) were plated and processed for hTERC FISH, as described in this protocol. Endogenous
hTERC in HeLa cells appears as foci in the nucleus. (b) hTERC FISH was performed on HeLa cells followed by
immunofluorescence with anti-hCoilin-Alexa 647 coupled antibody (hCoil) to detect Cajal bodies. Acquisition
was performed with a Zeiss upright epifluorescence microscope and Ixon EM-CCD camera. Images are
maximal projections. Scale bar ¼ 5 μm

can be imaged using an epifluorescence or a confocal micro-

scope (Fig. 2a). Colocalization experiments can also be per-
formed using antibodies (to detect Coilin, as in Fig. 2b).

4 Notes

1. Formaldehyde quality is very important to preserve small

details and low contrasts by FISH. We strongly recommend
using EM-grade formaldehyde from Electron Microscope
Science.
400 David Guérit et al.

2. During probe purification, dropping probe solution far from

the center of the G25 Sephadex column will results in the
contamination of the probes with free dye molecules, leading
in a false incorporation calculation. In this case, a second puri-
fication can be done with a new column.
3. When using yeast strains having the ade2 marker, adding
20 μg/mL of adenine in the media decreases autofluorescence
coming from the accumulation of phosphoribosyl-
aminoimidazole, the adenine intermediate giving the pink
color to ade2 yeast colonies. For the same reason, yeast
grown in YEPD medium present more autofluorescence
(YEPD is poor in adenine).
4. To image simultaneously a GFP-tagged protein and the TLC1
RNA by FISH in fixed yeast cells, grow the cells at room
temperature to get a stronger GFP fluorescence. Also, spin
the cells at 2500  g for 4 min and resuspend the cells in
50 mL of RNAse-free PBS before adding the PFA. Keeping
the pH of the yeast culture at 7.0 during fixation help preserves
the GFP fluorescence [17].
5. Spheroplasting is a step that needs to be optimized by each
user. If the cell wall is not enough degraded, the probes will not
diffuse inside the cells. On the other hand, if the cell wall is too
much digested, the cell may lose its integrity and shape. Each
reaction needs to be followed using a differential interference
contrast (DIC) microscopy. Typically each 2–3 min, apply
3.5 μL of the spheroplasting reaction between a coverslip and
a microscopy slide, and observe under the 20 objective. Cells
that still have an intact cell wall appear brighter and have a halo
around them. Cells that have a degraded cell wall loose the halo
and turn darker. When more than 75% of the cells show a
degradation of their cell wall, stop the spheroplasting reaction.
6. Drop the liquid carefully on the side of the well, avoiding
putting it directly on the coverslip to make sure not to disperse
the spheroplasts.
7. Mark the side of the Coplin jar with a marker pen to remember
on which face are the spheroplasted yeasts.
8. The concentration of formamide can be adjusted to have better
signal if needed. If the background is too high, the concentra-
tion of formamide can be adjusted up to 50%.
9. Avoid making bubbles to assure a uniform hybridization and
make sure not to move the coverslips once laid down to avoid
displacing the spheroplasts from the coverslips.
10. After this wash, a standard immunofluorescence protocol can
be performed if a colocalization experiment with a protein is
needed [18].
 Telomerase RNA Imaging in Budding Yeast and Human Cells by Fluorescent. . . 401

11. Mark the side of the Coplin jar with a marker pen to remember
on which face are the cells.
12. Avoid making bubbles to assure a uniform hybridization and
make sure not to move the coverslips once laid down to avoid
displacing the cells from the coverslips.

Acknowledgments

This work was funded by a grant from the Canadian Institutes for
Health Research (CHIR) MOP-89768 to P.C. M.L is supported by
a fellowship from the Fonds de Recherche du Québec-Santé
(FRQS). P.C holds a Research Chair from FRQS.

References

1. Schmidt JC, Cech TR (2015) Human telome- share a common Cajal body-specific localiza-
rase: biogenesis, trafficking, recruitment, and tion signal. J Cell Biol 164(5):647–652.
activation. Genes Dev 29(11):1095–1105. doi:10.1083/jcb.200310138
doi:10.1101/gad.263863.115 9. Zhu Y, Tomlinson RL, Lukowiak AA, Terns
2. Gallardo F, Olivier C, Dandjinou AT, Wellin- RM, Terns MP (2004) Telomerase RNA accu-
ger RJ, Chartrand P (2008) TLC1 RNA mulates in Cajal bodies in human cancer cells.
nucleo-cytoplasmic trafficking links telomerase Mol Biol Cell 15(1):81–90. doi:10.1091/
biogenesis to its recruitment to telomeres. mbc.E03-07-0525
EMBO J 27:748–757 10. Venteicher AS, Abreu EB, Meng Z, McCann
3. Wu H, Becker D, Krebber H (2014) Telome- KE, Terns RM, Veenstra TD, Terns MP,
rase RNA TLC1 shuttling to the cytoplasm Artandi SE (2009) A human telomerase holo-
requires mRNA export factors and is important enzyme protein required for Cajal body locali-
for telomere maintenance. Cell Rep 8 zation and telomere synthesis. Science 323
(6):1630–1638. doi:10.1016/j.celrep.2014. (5914):644–648. doi:10.1126/science.
08.021 1165357
4. Teixeira TM, Förstemann K, Gasser SM, 11. Gallardo F, Laterreur N, Cusanelli E, Ouenzar
Lingner J (2002) Intracellular trafficking of F, Querido E, Wellinger RJ, Chartrand P
yeast telomerase components. EMBO Rep 3 (2011) Live cell imaging of telomerase RNA
(7):652–659 dynamics reveals cell cycle-dependent cluster-
5. Ferrezuelo F, Steiner B, Aldea M, Futcher B ing of telomerase at elongating telomeres. Mol
(2002) Biogenesis of yeast telomerase depends Cell 44(5):819–827
on the Importin Mtr10. Mol Cell Biol 22 12. Bajon E, Laterreur N, Wellinger Raymund J
(17):6046–6055. doi:10.1128/mcb.22.17. (2016) A single Templating RNA in yeast telo-
6046-6055.2002 merase. Cell Rep 12(3):441–448. doi:10.
6. Pfingsten JS, Goodrich KJ, Taabazuing C, 1016/j.celrep.2015.06.045
Ouenzar F, Chartrand P, Cech TR (2012) 13. Jády BE, Richard P, Bertrand E, Kiss T (2006)
Mutually exclusive binding of telomerase Cell cycle-dependent recruitment of telome-
RNA and DNA by Ku alters telomerase recruit- rase RNA and Cajal bodies to human telo-
ment model. Cell 148:922–932 meres. Mol Biol Cell 17(2):944–954. doi:10.
7. Nguyen D, Grenier St-Sauveur V, Bergeron D, 1091/mbc.E05-09-0904
Dupuis-Sandoval F, Scott Michelle S, Bachand 14. Tomlinson RL, Abreu EB, Ziegler T, Ly H,
FA (2015) Polyadenylation-dependent 30 end Counter CM, Terns RM, Terns MP (2008)
maturation pathway is required for the synthe- Telomerase reverse transcriptase is required
sis of the human telomerase RNA. Cell Rep 13 for the localization of telomerase RNA to cajal
(10):2244–2257. doi:10.1016/j.celrep.2015. bodies and telomeres in human cancer cells.
11.003 Mol Biol Cell 19(9):3793–3800. doi:10.
8. Jady BE, Bertrand E, Kiss T (2004) Human 1091/mbc.E08-02-0184
telomerase RNA and box H/ACA scaRNAs
402 David Guérit et al.

15. Abreu E, Aritonovska E, Reichenbach P, Cris- 17. Gallardo F, Chartrand P (2011) Visualizing
tofari G, Culp B, Terns RM, Lingner J, Terns mRNAs in fixed and living yeast cells. In:
MP (2010) TIN2-tethered TPP1 recruits Gerst EJ (ed) RNA detection and visualization:
human telomerase to telomeres in vivo. Mol methods and protocols. Humana Press,
Cell Biol 30(12):2971–2982. doi:10.1128/ Totowa, NJ, pp 203–219. doi:10.1007/978-
mcb.00240-10 1-61779-005-8_13
16. Stern JL, Zyner KG, Pickett HA, Cohen SB, 18. Chartrand P, Bertrand E, Singer RH, Long
Bryan TM (2012) Telomerase recruitment RM (2000) Sensitive and high-resolution
requires both TCAB1 and Cajal bodies inde- detection of RNA in situ. Methods Enzymol
pendently. Mol Cell Biol 32(13):2384–2395. 318:493–506
doi:10.1128/mcb.00379-12
 Chapter 28

Methods to Study Repeat Fragility and Instability

in Saccharomyces cerevisiae
Erica J. Polleys and Catherine H. Freudenreich

Abstract
Trinucleotide repeats are common in the human genome and can undergo changes in repeat length.
Expanded CAG repeats have been linked to over 14 human diseases and are considered hotspots for
breakage and genomic rearrangement. Here, we describe two Saccharomyces cerevisiae based assays that
evaluate the rate of chromosome breakage that occurs within a repeat tract (fragility), and a PCR-based
assay to evaluate tract length changes (instability). The first fragility assay utilizes end-loss and subsequent
telomere addition as the main mode of repair of a yeast artificial chromosome (YAC). The second fragility
assay relies on the fact that a chromosomal break stimulates recombination-mediated repair. In addition to
understanding the role of fragility at repetitive DNA sequences, both assays can be modified to evaluate
instability of a CAG repeat using a PCR-based assay. All three assays have been essential in understanding
the genetic mechanisms that cause chromosome breaks and tract-length changes at unstable repeats.

Key words Chromosome break, Fragility, Stability, Yeast artificial chromosome (YAC), CAG repeat

1 Introduction

Stretches of repetitive DNA sequence are common in eukaryotic

genomes and are capable of forming alternative DNA structures,
such as hairpin loops. These alternative DNA structures effectively
function as a physical barrier, making it difficult for error free DNA
replication and repair to occur, and can result in changes in repeat
length, chromosomal breakage, and genome instability. One type
of repetitive DNA sequence that is common in the human genome
are CNG repeats, which when expanded, have been causally asso-
ciated with over 16 human diseases including Huntington’s disease
(HD), myotonic dystrophy (DM), spinocerebellar ataxias (SCAs),
and fragile X syndrome (FXS) [1, 2].
Long regions of CAG/CTG and CGG/GCC trinucleotide
repeats break and undergo aberrant repair at a higher frequency
than nonrepetitive DNA, which can result in mutation and changes
in repeat copy number. Breakage of the DNA strands (referred to

Marco Muzi-Falconi and Grant W. Brown (eds.), Genome Instability: Methods and Protocols, Methods in Molecular Biology,
vol. 1672, DOI 10.1007/978-1-4939-7306-4_28, © Springer Science+Business Media LLC 2018

403
404 Erica J. Polleys and Catherine H. Freudenreich

herein as “fragility”) and changes in repeat copy number (referred

to herein as “instability”) are two sides of the same coin. For
example, if normal replication or repair is impaired when encoun-
tering a repeat tract, a stalled fork, nick, gap, or double-strand break
(DSB) can occur. If the fork is restarted or if there is healing within
the nick, gap or break, repeat units can be lost or gained. Alterna-
tively, the DNA lesion may be repaired by a less conservative form
of repair that leads to deletion of flanking sequence, loss of a
chromosome arm, or other chromosome rearrangements. This
dramatic loss of genetic material can be easily monitored in S.
cerevisiae by loss of a neighboring genetic marker, and is indicative
that a break in the chromosome occurred. Such a marker loss assay
is an underestimate of the “true” rate of breakage, as it only
measures the incidents that are healed after loss of flanking DNA
sequence, and not those that heal within the repeat, or those that
fail to repair resulting in cell death. Nonetheless, such genetic
fragility assays are very useful for comparative studies and can be
used to determine if the presence of a gene or condition is involved
in replication or repair in the context of repetitive DNA. In con-
trast, a physical assay for trinucleotide repeat instability captures all
the events that result in loss or gain of the number of repeats (at
least those of a size that can be detected by the sensitivity of the
assay), but will usually miss events that result in loss of flanking
DNA sequence that includes one of the PCR primer binding sites.
Because of the clonal nature of yeast growth, repeat instability
can easily be monitored over multiple generations. The PCR-based
instability assay (Subheading 3.1) ideally begins with a parent col-
ony with a known starting repeat tract length, determined by PCR
using a subset of cells in the colony as a template (colony PCR; see
Subheading 3.2). Subsequently, the cells in the parent colony are
grown for a set number of cell divisions in liquid media to allow
repeat tract length changes, plated and allowed to grow into daugh-
ter colonies. Colony PCR is then used to determine the repeat tract
length in multiple individual daughter colonies in order to obtain a
frequency of instability (Fig. 1). The goal is to use the tract length
in the daughter colony as an output of the length present in the cell
at the time of plating (which will contain one or two genomes,
depending on cell cycle phase). This method is set up to ignore any
changes that likely occur during the growth of the daughter colony,
and only count the primary tract length changes that occurred in
the cells during logarithmic growth in culture. It should be noted
that there are other available genetic assay systems that can be used
to select for repeat expansion or contraction events which are
covered in this issue and previous papers [3, 4]. Such selection
systems differ from the PCR-based instability assay as they capture
only the events that yield a phenotype in the selective system.
However they allow for the study of rare repeat length changes,
such as those that occur in short (premutation) length CAG or
 Measuring Repeat Fragility and Instability in Yeast 405

Fig. 1 Schematic of CAG repeat instability assay (a) In this system, CAG repeats have either been integrated
into a (CAG)n-URA3 YAC (YAC CF1) or another chromosomal location. Colony PCR across the repeat is utilized
to monitor the frequency of expansions and contractions in daughter cells. (b) In the (CAG)n-URA3 YAC
instability assay, strains are plated for single colonies on YC -Leu -Ura plates and individual parent colonies
are tested for CAG tract length. Parent colonies of the correct tract length are then grown in YC -Leu -Ura for
6–7 cell divisions before plating each individual culture for single colonies on YC -Leu -Ura plates. Finally PCR
is performed on approximately 25 daughter colonies for each parent colony to assess the frequency of
changes in tract length sizes compared to the starting repeat size. At least four parent colonies should be
tested for a total of at least 100 daughter colonies. P parental, C contraction, E expansion, M marker

CGG repeats (e.g., the Lahue lab system) or for large-scale repeat
expansions (e.g., the Mirkin lab system).
To determine the rate at which breakage occurs within a repeat
tract, two unique yeast systems have been developed. The first is a
yeast artificial chromosome (YAC) system wherein the YAC VS5 is
modified such that the potential fragile sequence is integrated
between a telomere seed sequence (G4T4) and a URA3 marker
gene [5, 6] (Fig. 2). The version of this YAC with an integrated
CAG repeat tract has been termed the (CAG)n-URA3 YAC, also
referred to in the literature as YAC CF1. One feature of this YAC
system is that other modifications are easily made by replacing the
right end of the YAC (after G4T4) with the desired sequence. We
have used this technique to add additional genetic markers, pro-
moters, terminators, etc. in addition to modifying the type of
potential fragile element. For example, other derivatives of this
YAC have been made that contain the CAG repeat in the opposite
orientation (CTG)n [7], AT repeats that are present in the
FRA16D common fragile site [8], the expanded ATTCT repeat
present in SCA10 [9], and inverted repeats [10].
Another important feature of the YAC is that it contains little
homology to any of the natural yeast chromosomes. This feature,
together with the telomere seed sequence placed proximal to the
potential fragile sequence, ensures that the primary mode of
406 Erica J. Polleys and Catherine H. Freudenreich

Fig. 2 Schematic of the YAC breakage assay (a) In this system, a potential fragile sequence (such as an
expanded CAG repeat tract) is integrated onto a yeast artificial chromosome between a telomere seed
sequence (G4T4) and the URA3 gene. Breaks that occur within the fragile sequence are subject to resection
and telomere addition, which results in loss of the URA3 gene and renders cells 5-FOAR. The YAC additionally
contains a LEU2 marker gene, which allows for maintenance of the YAC, a centromere (CEN4) and an origin of
replication (ARS1). The blue line indicates the pYIP5 plasmid backbone, the black line indicates lambda DNA,
and the purple line corresponds to a pUC18 plasmid backbone. (b) In the YAC end loss assay, strains are plated
for single colonies on YC -Leu -Ura plates and 11–13 individual colonies of the correct tract length (if testing
an unstable repeat sequence) are grown in YC -Leu for 6–7 cell divisions. A portion of each individual culture is
plated onto YC -Leu 5-FOA plates, and a portion of the culture is pooled and serially diluted to obtain single
colonies on YC -Leu, which serves as the total viable cell count

healing is by telomere addition rather than recombination. Healing

via telomere addition is ideal as relatively few genes affect the
efficiency of telomere addition. Known examples are the subunits
of the enzyme telomerase, which are needed for telomere addition,
and the Pif1 helicase, which inhibits promiscuous telomere addi-
tion [5]. Thus, in most cases an increase in YAC end loss is caused
by increased breakage rather than decreased healing. The position
of the potential fragile sequence on the YAC is such that on the
proximal side of the sequence there is a 108 bp backup (G4 T4)13
telomeric seed sequence from Oxytrica, which serves as an efficient
substrate for S. cerevisiae telomerase, but does not recombine with
 Measuring Repeat Fragility and Instability in Yeast 407

the natural yeast telomere sequence [11]. Therefore, if breakage

within the fragile sequence occurs, resection and degradation of the
broken end exposes the G4T4 sequence as a template for telome-
rase. The G4T4 seed provides an efficient pathway for healing of
breakage events, which is different from another commonly used
assay for end loss of chromosome V [12], where healing can occur
by a variety of mechanisms including telomere addition onto short
naturally occurring GT sequences or recombination with other
chromosomes.
In the YAC fragility assay, breakage that results in loss of the
right arm of the YAC renders cells ura3- and resistant to 5-
flourorotic acid (5-FOA). The number of cells that are 5-FOA
resistant serves as a measure of the rate of breakage, and the effect
of a potential fragile sequence can be compared to a control
sequence of the same length and base composition, or a YAC that
does not contain the fragile sequence. One additional design con-
sideration of the YAC is the distance between the URA3 gene and
the telomere, such that the placement of the URA3 gene promoter
is far enough away from the telomere not to be subject to telomere
position effect. Elsewhere on the YAC, on the left arm there is a
LEU2 marker gene, which allows for maintenance of the YAC, a
yeast origin of replication (ARS1), and a centromere (CEN4).
Finally, an ade3-2p allele can be used to screen for cells that have
obtained two YACs (which occurs occasionally by nondisjunction)
in strains with an ade3 background, but it is not necessary to use
this feature for most applications.
For the original (CAG)n-URA3 YAC, the CAG repeat tract is
integrated such that the CAG repeat is on the lagging strand
template, which has been shown to be the orientation less prone
to repeat tract contractions [13]. For the CAG repeat, it has been
established that there is a length dependent increase in breakage as
measured by 5-FOA resistance [6]. This YAC system can also be
used to study CAG repeat stability using similar conditions used for
measuring breakage. These instability assays revealed a length
dependent increase in contractions and expansions, though con-
tractions occur with greater frequency than expansions in a repli-
cating yeast population [6]. The CAG instability frequency
measured on this YAC is similar to that measured at the chromo-
some II LYS2 locus for the same orientation with respect to repli-
cation [13, 14], though it is known that factors such as distance
from an origin, transcription level, and chromatin structure can
alter instability frequencies (see ref. 1 for review).
The second system for measuring repeat fragility is a direct
duplication recombination assay (DDRA), which is a chromosomal
assay where a potential fragile sequence has been integrated adja-
cent to a URA3 reporter and has been placed between a full length
LYS2 gene and a duplication of the 30 end of LYS2 on chromosome
II [15, 16] (Fig. 3). If breaks occur within the integrated sequence,
408 Erica J. Polleys and Catherine H. Freudenreich

Fig. 3 Schematic of the direct duplication recombination assay (DDRA). (a) In this chromosomal system, a
potential fragile sequence and the URA3 gene are integrated between a direct duplication of 708 bp of the 30
 Measuring Repeat Fragility and Instability in Yeast 409

resection occurs and repair results from single strand annealing of

the duplicated regions of the LYS2 gene. This results in the loss of
the URA3 gene and resistance to the drug 5-FOA. In this system, it
was noted that there is a length dependent increase in 5-FOA
resistance which was due to breaks within or very close to an
integrated CAG repeat tract [15], and expanded CGG/CCG
repeats also induced length-dependent fragility [16]. This system
is ideal for understanding the role that recombination plays in the
repair of breaks that occur within a repeat tract. Also, as it is located
in the middle of a natural yeast chromosome, the replication pro-
gram is known, and the location is far from elements such as the
centromere and telomere that could influence chromatin structure
or nuclear positioning. A modification of the DDRA utilizes the
same principle, but with a duplication of the ADE2 gene, designed
such that the starting strain is Ade- and recombination products
will become Ade þ 5-FOAR ([17]; Fig. 2b). This modification
allows for selection of recombinants, eliminating other events that
can generate FOAR, and the cassette can be integrated anywhere in
the genome. It is important to note that many genes that are
involved in replication or repair are also required for efficient
recombination, which makes it more difficult to determine whether
a gene is required to prevent fragility or facilitate healing in this
system, as compared to the YAC end loss system.
These assay systems are unique as they can be utilized to
monitor both fragility and instability of long stretches of repetitive
DNAs under the same conditions (genomic location, strain back-
ground, transcriptional status, etc.). Taken together, both assays,
the YAC end loss system and the chromosomal direct duplication
recombination assay, have provided novel insight into the mechan-
isms of fragility and instability of repetitive DNA sequences and can
be utilized to understand the breakage and repair of any sequence
of interest.

Fig. 3 (continued) end of the LYS2 gene, rendering cells LYSþ and FOAS. Breaks that occur within the fragile
sequence are subject to resection and recombination, which renders cells Lys þ and FOAR. (b) In the ADE2
DDRA, a potential fragile sequence and the URA3 gene are integrated between 50 and 30 portions of the ADE2
gene, rendering cells Ade- and FOAS. Breaks that occur within the repeat are subject to resection and repair
via single strand annealing between the 968 bp duplicated region of ADE2, thus rendering cells Ade þ and
FOAR. (c) Schematic of DDRA protocol. Cells are initially plated on YC -Ura media and then plated for single
colonies on YEPD (or other nonselective media) to allow loss of URA3. Ten individual colonies are selected and
each resuspended in water. A portion of each colony suspension is then plated onto ten separate 5-FOA plates
(Lys or -Ade, depending on the assay) and 100 μL of each colony suspension is pooled, diluted and plated
for single colonies onto YC plates for a total cell count
410 Erica J. Polleys and Catherine H. Freudenreich

2 Materials

2.1 Determination 1. Starting colonies that have been checked for tract length by
of Repeat Instability colony PCR (see Subheading 3.2).
Frequency 2. OD spectrophotometer.
3. Cuvettes.
4. 5 mL glass test tubes.
5. P20 or P200 pipette tips.
6. Roller drum at 30  C.
7. Media (standard yeast media, see ref. 18 for recipes):
For YAC repeat instability assays:
(a) YC -Leu -Ura liquid media.
(b) YC -Leu -Ura plates.
For DDRA repeat instability assays:
(c) YC -Ura plates.

2.2 Colony PCR to 1. Taq polymerase.

Determine CAG Repeat 2. Taq polymerase buffer.
Tract Length
3. 20 mM magnesium sulfate (MgSO4) solution.
4. 5 CG buffer (Empire Genomics cat# IDL028).
5. 10 mM dNTPs.
6. Primers that span the CAG repeat (see Note 9).
(a) Forward primer: 50 CCC AGG CCT CCA GTT TGC 30
(starts 74 bp before CAG repeat on (CAG)n-URA3
YAC).
(b) Reverse primer: 50 TAA TAC GAC TCA CTA TAG GG 30
(starts 79 bp after CAG repeat on (CAG)n-URA3 YAC).
7. 2% Metaphor agarose or other high resolution gel system.
8. Refrigerator.
9. Power supply.
10. Electrophoretic gel box.
11. 10 mg/mL ethidium bromide.

2.3 Determination 1. Starting colonies that have been checked for repeat tract length
of Chromosome by colony PCR (see Subheading 3.2 for CAG repeats).
Fragility by a YAC end 2. OD spectrophotometer.
Loss Assay
3. Cuvettes.
4. 5 mL glass test tubes.
5. P20 or P200 pipette tips.
 Measuring Repeat Fragility and Instability in Yeast 411

6. Roller drum at 30  C.
7. Media (standard yeast media, see ref. 18 for recipes):
(a) YC -Leu liquid media.
(b) YC -Leu -Ura plates.
(c) YC -Leu plates.
(d) YC -Leu 5-FOA plates.

2.4 Direct 1. Media (standard yeast media, see ref. 18 for recipes):
Duplication (a) YC -Ura plates.
Recombination Assay
(b) YEPD plates.
(DDRA)
(c) YC -Lys 5-FOA plates (LYS2 assay) or YC -Ade 5-FOA
plates (ADE2 assay).
(d) YC plates.

3 Methods

3.1 Determination 1. Plate or streak for single colonies on YC -Leu -Ura, which will
of Repeat Instability select against breakage events, from a master patch (see Note 1).
Frequency 2. Perform colony PCR (see Subheading 3.2) on 3–5 colonies to
identify two parent colonies with the desired CAG tract length
that have no other visible repeat sizes. The colonies should
ideally be less than 2 weeks old. Usually two separate colonies
from each strain are done in parallel.
3. Using a pipet tip, suspend a small amount of the yeast colony
(from the same region of the colony tested for colony PCR)
into 1 mL of YC -Leu -Ura media. The ideal starting OD600 is
between 0.02 and 0.04 (see Note 2).
4. Grow the culture for 6–7 divisions at 30  C with constant
agitation in a roller drum (see Note 3).
5. Check 100 μL aliquots of your cultures (diluted in 900 μL
diH2O) to determine the final OD600. Multiply this reading
by 10. Grow the culture to whatever OD600 will give 6–7
divisions based on your starting OD600, usually ~OD 1.2–5.0
(1.0).
6. Take a 10 μL aliquot from the culture and add to a microfuge
tube containing 90 μL diH2O. Dilute tenfold three more times
(to 104), and plate 100 μL on a YC -Leu -Ura plate. Do this
for each culture, making sure to keep cultures separate.
7. Incubate for 3 days at 30  C.
8. Do colony PCR on ~25 colonies from each YC -Leu -Ura plate
(see Notes 4 and 5).
412 Erica J. Polleys and Catherine H. Freudenreich

9. Keep track of which PCR reactions came from which plate/

parent colony and which daughter colony (#1–25). Mark the
lanes with expansions (E) and contractions (C) on the picture
(see Note 6).
10. Repeat the assay until ~100 daughter colonies from four parent
colonies have been tested. More assays can be added if greater
statistical significance is desired.
11. Stability data is calculated as frequency of change/total number
of reactions and are either represented as: number of contrac-
tions/total number of reactions ¼ frequency of contractions or
number of expansions/total number of reactions ¼ frequency of
expansions. Statistical significance is determined by a Fisher’s
exact test using the raw numbers from the sum of reactions
performed (for example 2 out of 100 versus 12 out of 100
would be compared) (see Notes 7 and 8).

3.2 Colony PCR to 1. Select individual colonies and clearly mark which PCR reac-
Determine CAG Repeat tions will come from which colony.
Tract Length 2. Combine a bulk PCR reaction mix on ice and aliquot 12.5 μL
per PCR tube. The 1 PCR reaction is:

Sterile diH2O 5.8 μL

10 buffer (w/o MgCl2): 1.25 μL
20 mM MgSO4: 1.25 μL
6.25 pmol primer 1: 0.625 μL (of 10 pmol/μL stock)
6.25 pmol primer 2: 0.625 μL (of 10 pmol/μL stock)
10 mM dNTPs: 0.25 μL
Mix thoroughly, then add,
Taq polymerase: 0.2 μL
5 CG-Rich buffer: 2.5 μL
Total reaction mix is 12.5 μL

Using a pipette tip, add a small amount of yeast cells from the
center of the colony into the PCR reaction mix. When mixing
the cells with reaction mix, either place a gloved finger on
the wider end of pipette tip or use a pipetteman and push the
plunger completely to ensure all liquid entering the tip by
surface tension is expelled (see Notes 9–11).
 Measuring Repeat Fragility and Instability in Yeast 413

3. PCR cycle profile:

95  C 40
95  C 3000 35 times
54  C 10
68  C 30

68  C 70
4 C Hold

4. Once the PCR protocol is complete, add an appropriate

amount of loading dye to the 12.5 μL reaction and run out
the entire PCR reaction on a 2% Metaphor gel or other sizing
gel system. Run Metaphor gels in 1 TAE at 70 V for 90 min
and load markers on both ends of the gel; a third marker in a
middle lane is helpful for gel systems that do not run evenly
across the gel box (see Notes 12 and 13).
5. Stain the gel with ethidium bromide for 30 min. Transfer gel
tray to diH2O and photograph the gel using a UV transillumi-
nator (see Note 14).
6. In Adobe Photoshop or similar program draw a line between
the marker bands on each side, put a curve in if necessary, and
decide if an expansion or contraction occurred based on
whether the band falls above or below the line (see Notes 15
and 16).

3.3 Determination 1. Plate for single colonies on YC -Leu -Ura, which will select
of Chromosomal against breakage events, from a master patch.
Fragility by a YAC End 2. If the potential fragile sequence is an unstable repeat able to be
Loss Assay sized via PCR, do colony PCR on ~20 colonies to check for the
appropriate repeat tract length.
3. Inoculate colonies of correct tract length for the fluctuation
assay. Colonies should be as fresh as possible.

3.3.1 Methods 1. 10 colony protocol (preferred):

of Inoculation For this methodology there will be 11–13 unique colonies that
are inoculated into 11–13 separate cultures.
(a) Pipet 1 mL of YC -Leu into 11–13 5 mL test tubes.
(b) Using a pipet tip, inoculate 1–3 tubes with a small number
of cells and take the OD600 (using the entire 1 mL culture;
these are referred to as the “OD600 cultures”). The start-
ing OD600 should be ~0.02–0.06. Once the OD600 has
been taken, return the 1 mL culture to the initial test tube
(see Note 17).
414 Erica J. Polleys and Catherine H. Freudenreich

(c) Inoculate the remaining tubes with approximately the

same amount of cells by eye, so that they will also have a
starting OD600 of ~0.02–0.06. These are referred to as the
“experimental cultures” (see Note 18).
2. 1 colony protocol:
For this methodology there will be one colony that is divided
into ten experimental cultures and 1–3 OD600 cultures.
(a) Suspend one entire colony with the correct tract length in
1.3 mL of YC -Leu media.
(b) Take 100 μL of the resuspended culture and inoculate into
11–13 5 mL test tubes that contains 0.9 mL of YC -Leu
media.
(c) Take the 1 mL OD600 culture out of the test tube and
measure the OD600. The OD600 should be between ~0.02
and 0.06. Return the 1 mL culture to the initial test tube
(see Notes 19 and 20).
l Allow the cultures to grow with constant agitation in a
spinning roller drum for 6–7 doublings at 30  C (see
Note 21).
l In order to determine the number of doublings, take
100 μL of the OD600 cultures and resuspend into
900 μL of YC -Leu media and measure the final OD.
Multiply this reading by 10. If the ideal starting OD600
was ~0.02, the appropriate final OD600 range is
1.25–2.5 (see Note 3). Once the correct OD600 has
been obtained, discard the 1–3 OD600 tubes.
l For the remaining ten cultures, spread 100 μL from
each of the ten tubes onto ten YC -Leu 5-FOA plates
using a consistent number of sterile glass beads or a
sterile glass spreader. The amount plated can be
adjusted up or down for strains with a higher or lower
rate of FOAR; just remember to adjust the number of
colonies counted to the number of mutants expected
per 100 μL in the rate calculations.
l To determine the viable total cell count, combine
100 μL from each of the ten undiluted cell cultures
and mix together. Serially dilute the mixed cultures in
sterile water (usually to 104 and 105) and plate
100 μL on of each dilution onto YC –Leu plates.
Total cell count plates should be plated in duplicate.
l Incubate the plates at 30  C for 4 days.
l Count colonies on the YC -Leu 5-FOA plates.
l Order the plate counts from low to high to determine
the median number of colonies for each set (i.e., each
 Measuring Repeat Fragility and Instability in Yeast 415

ten plate assay), and average the number of colonies on

the two “total cell count” plates to determine the
number of viable cells in 100 μL. Determine the rate
of 5-FOAR by using a fluctuation rate calculator such as
FALCOR (http://www.keshavsingh.org/protocols/
FALCOR.htmL) [19] or FluCalc (see this issue;
http://flucalc.ase.tufts.edu/) using either the Maxi-
mum Likelihood Method (preferred, if the spread is
not too great) or the Method of the Median. See ref. 20
for an excellent discussion of calculating fluctuation
rates (see Notes 21–24).
l The assay should be repeated at least three times per
strain, and standard error and confidence intervals
calculated.

3.4 Direct 1. Patch strain to test onto a YC -Ura plate and allow to grow for
Duplication 2 days at 30  C.
Recombination Assay: 2. Plate for single colonies on YEPD at 30  C for 3 days such that
A Recombination there are 50  10 colonies per plate (this step allows recombi-
Based Method to nation events to occur).
Assess Chromosome 3. Randomly select ten colonies and resuspend half of each colony
Breakage in 400 μL diH2O (see Note 25).
4. Plate a portion of the suspension on a 5-FOA plate with the
goal being to get ~50–200 colonies as the median after
4–5 days of growth at 30  C. For the LYS2 assay (Fig. 3a),
this should be a YC -Lys 5-FOA plate; For the ADE2 assay,
(Fig. 3b), this should be a YC -Ade 5-FOA plate (see Note 26).
5. To determine the viable cell count combine 100 μL of each of
the ten colony suspensions. Plate a dilution that will yield
~30–300 colonies single colonies per YC plate. Total cell
count plates should be plated in duplicate (see Note 27).
6. Count the number of colonies on the 5-FOA plates. Determine
the number of mutants and the total number of cells per a
certain volume, typically per 10 ul, by multiplying the number
of colonies growing on 5-FOA and the average of the two total
cell count plates by the dilution factors used. To determine the
rate of recombination, use the method of maximum likelihood
or the method of the median and a fluctuation analysis calcula-
tor (e.g., FALCOR [19] or FluCalc; see Subheading 3.3).

4 Notes

4.1 Determination 1. This protocol is written for a CAG tract on the CF1 YAC, but
of Repeat Instability can be adapted to other locations or situations.
Frequency
416 Erica J. Polleys and Catherine H. Freudenreich

2. The low starting OD600 ensures that cells have only been
grown in log phase before being plated. The method is
designed to test for instability occurring during log phase
growth. The OD600 range could be altered according to the
purpose of the experiment.
3. For strains with poor division potential, the total number of
divisions can be altered to fewer divisions, e.g., 3–5 divisions.
Wildtype strains done as a comparison should be grown for the
same number of divisions as mutant strains.
4. Pick colonies in as unbiased approach as possible: either pick a
section of the plate or pick representative colonies that reflect
the natural variability of colony size on the plate.
5. The PCR reactions do not have to be done all at once—choose
the number by how proficient you are at colony PCR, but do all
within 2 weeks. Keep track of which PCR reactions came from
which plate.
6. For lanes where both a starting size band and new size band are
visible, an expansion can be counted if it is clearly visible
(intensity about 30% or more compared to starting size). A
contraction can be counted if the intensity of the band is at least
equal to the starting size band. The logic is that these events are
likely have been present in the plated cell or occurred in the first
cell division (expanded repeats PCR less efficiently than shorter
repeats, thus the difference in cut-off criteria; (see ref. 13 for
examples). The goal is to count events that occurred during the
culture growth, and not those that occurred during the growth
of the plated cell into a colony.
7. Typically, 100–200 PCR reactions in total are used. If a set of
reactions from the same parent show a clear difference from the
others (a jackpot), this assay should be dropped. For example,
an undetected contraction in the starting parent colony could
lead to an unusually high frequency of “contractions” in the
daughters, which are not independent events that occurred
during the culture growth. These often show up as many con-
tractions of the same size.
8. The instability analysis can be combined with a fragility assay by
analyzing the repeat tract length of the daughter colonies
plated for the total cell count. Even though these plates do
not select for the presence of the repeat (e.g., the YC -Leu
plates for the YAC assay), usually the rate of breakage is low
enough that most colonies tested will have retained a repeat
tract. One disadvantage to this method is that the starting
parent colony and the daughter colonies are no longer corre-
lated, reducing the ability to detect jackpot events.
 Measuring Repeat Fragility and Instability in Yeast 417

4.2 Colony PCR to 9. In order to obtain robust tract lengths PCR, primers should be
Determine CAG Repeat designed such that there is a region of nonrepetitive flanking
Tract Length sequence included on both sides of the repeat. Generally, we
recommend between 40 and 80 bp of nonrepetitive sequence
be included on each side of the repeat.
10. Taq polymerase and CG buffer should be added last to the
master mix after everything else. CG rich buffer should not be
vortexed.
11. The amount of cells added to the PCR reaction should be no
larger than the tip of the p200 or p20 pipette tip. Too few cells
will result in a very faint PCR product, while too many will
result in a background smear in the lane.
12. The 2% Metaphor agarose gels should be made according to
the manufacturer’s instructions.
13. Instead of using 2% Metaphor gels for separation of fragments,
PCR reactions can be run on a denaturing PAGE gel or other
high resolution gel system for size analysis, with appropriate
DNA detection.
14. Ethidium bromide can be added directly to Metaphor gel if
preferred. The ideal concentration is 15 μL 10 mg/mL stock in
300 mL 1 TAE.
15. The ideal molecular weight marker used should have bands
that change in size every 100 bp or less. Adding an additional
marker in the middle of the gel helps with sizing the CAG
repeat.
16. If the PCR is of poor quality, then expansions may not have
amplified well and could have been underestimated; therefore,
only use good-quality PCR reactions to determine expansion
frequencies. A 2% Metaphor gel can resolve fragments that
differ by 2%, i.e., 8 bp for a fragment size of 400 bp or about
3 (CAG) repeats added or subtracted for a (CAG)85 repeat
tract.

4.3 Determination of 17. Usually the “OD600 cultures” are only used for taking the
Chromosomal Fragility OD600 throughout the experiment and not plated to be
Using a YAC End Loss counted, due to risk of contamination during the process of
Assay taking the OD600.
18. The method is designed to test for fragility occurring during
log phase growth. The OD600 range could be altered according
to the purpose of the experiment.
19. The 1 colony method is less preferred, but can be used when
finding 10 starting colonies of the same repeat tract length is
technically difficult.
20. For wild type, 6–7 divisions will take approximately ~16–18 h.
418 Erica J. Polleys and Catherine H. Freudenreich

21. Make sure you enter the number of mutants and total cells for
the same volume—usually per 100 μL (per 1 mL can be used for
very low rates).
22. For experiments with zeros or a large spread between low and
high numbers of mutants (happens with lower rate assays), use
the Method of the Median to calculate the rate.
23. Certain mutants could alter URA3 gene function or FOAR by
increasing the mutation rate of URA3, altering URA3 expres-
sion, or affecting the toxicity of 5-FOA, resulting in increased
FOAR that is not due to chromosome breakage. To confirm
that there is loss of the right arm of the YAC, PCR amplify the
URA3 gene from 1 colony off of each YC -Leu 5-FOA plate
(ten colonies total) to confirm there is loss of the right arm of
the YAC. This should be done for three assays (a total of 30
independent colonies checked). The expected end loss fre-
quency for wild type is 90–100%. Make sure to also run your
starting strain as a positive control.
24. Verify the structure of the YAC in 20 independent FOAR
colonies (plated from independent cultures) by Southern
blot, using a lambda probe to the YAC and digestion with
BstEII (see ref. 6).

4.4 Direct 25. Establish a standard size of colony to use to allow equal gen-
Duplication erations of growth between colonies and strains.
Recombination Assay 26. Generally, for rates in the 105 range, approximately 10 μL of
the colony suspension is plated onto 5-FOA. A pilot experi-
ment can determine the best amount to plate. Keep in mind
that the more fragile the sequence, the less should be plated.
27. The dilution plated for the viable cell count depends on the size
of the colonies that were picked for the assay. For initial assays it
is best to plate a series of dilutions, such as 104 and 105.

Acknowledgments

Work in our laboratory is currently supported by grants from the

National Institute of Health (GM105473) and National Science
Foundation (MCB 1330743). We would like to thank past and
present members of the C. H. Freudenreich and V.A. Zakian labs
for their contributions to the development of these assays.

References
1. Usdin K, House NC, Freudenreich CH (2015) 2. McMurray CT (2010) Mechanisms of trinucle-
Repeat instability during DNA repair: Insights otide repeat instability during human develop-
from model systems. Crit Rev Biochem Mol ment. Nat Rev Genet 11(11):786–799.
Biol 50(2):142–167. doi:10.3109/10409238. doi:10.1038/nrg2828. PubMed PMID:
2014.999192. PubMed PMID: 25608779; 20953213; PMCID: 3175376
PMCID: 4454471
 Measuring Repeat Fragility and Instability in Yeast 419

3. Shishkin AA, Voineagu I, Matera R, Cherng N, 12. Chen C, Kolodner RD (1999) Gross chromo-
Chernet BT, Krasilnikova MM, Narayanan V, somal rearrangements in Saccharomyces cerevi-
Lobachev KS, Mirkin SM (2009) Large-scale siae replication and recombination defective
expansions of Friedreich’s ataxia GAA repeats mutants. Nat Genet 23(1):81–85. doi:10.
in yeast. Mol Cell 35(1):82–92. doi:10.1016/ 1038/12687. PubMed
j.molcel.2009.06.017. PubMed PMID: 13. Freudenreich CH, Stavenhagen JB, Zakian VA
19595718; PMCID: 2722067 (1997) Stability of a CTG/CAG trinucleotide
4. Dixon MJ, Bhattacharyya S, Lahue RS (2004) repeat in yeast is dependent on its orientation in
Genetic assays for triplet repeat instability in the genome. Mol Cell Biol 17(4):2090–2098.
yeast. Methods Mol Biol 277:29–45. doi:10. PubMed PMID: 9121457; PMCID: 232056
1385/1-59259-804-8:029. PubMed 14. Su XA, Dion V, Gasser SM, Freudenreich CH
5. Schulz VP, Zakian VA (1994) The saccharomy- (2015) Regulation of recombination at yeast
ces PIF1 DNA helicase inhibits telomere elon- nuclear pores controls repair and triplet repeat
gation and de novo telomere formation. Cell stability. Genes Dev 29(10):1006–1017.
76(1):145–155. PubMed doi:10.1101/gad.256404.114. PubMed
6. Callahan JL, Andrews KJ, Zakian VA, Freuden- PMID: 25940904; PMCID: 4441049
reich CH (2003) Mutations in yeast replication 15. Freudenreich CH, Kantrow SM, Zakian VA
proteins that increase CAG/CTG expansions (1998) Expansion and length-dependent fra-
also increase repeat fragility. Mol Cell Biol 23 gility of CTG repeats in yeast. Science 279
(21):7849–7860. Epub 2003/10/16. (5352):853–856. PubMed
PubMed PMID: 14560028; PMCID: 207578 16. Balakumaran BS, Freudenreich CH, Zakian VA
7. Kerrest A, Anand RP, Sundararajan R, Bermejo (2000) CGG/CCG repeats exhibit
R, Liberi G, Dujon B, Freudenreich CH, orientation-dependent instability and
Richard GF (2009) SRS2 and SGS1 prevent orientation-independent fragility in Saccharo-
chromosomal breaks and stabilize triplet myces cerevisiae. Hum Mol Genet 9
repeats by restraining recombination. Nat (1):93–100. PubMed
Struct Mol Biol 16(2):159–167. doi:10. 17. Paeschke K, Capra JA, Zakian VA (2011) DNA
1038/nsmb.1544. PubMed Epub 2009/01/ replication through G-quadruplex motifs is
13. doi: nsmb.1544 [pii] promoted by the Saccharomyces cerevisiae Pif1
8. Zhang H, Freudenreich CH. An AT-rich DNA helicase. Cell 145(5):678–691. doi:10.
sequence in human common fragile site 1016/j.cell.2011.04.015. PubMed PMID:
FRA16D causes fork stalling and chromosome 21620135; PMCID: 3129610
breakage in S. cerevisiae. Mol Cell. 2007;27 18. Dunham MJ, Dunham MJ, Gartenberg MR,
(3):367–379. doi: 10.1016/j.molcel.2007.06. Brown GW (2015) Methods in yeast genetics
012. PubMed PMID: 17679088; PMCID: and genomics : a Cold Spring Harbor Labora-
2144737 tory course manual/Maitreya J. Dunham, Uni-
9. Cherng N, Shishkin AA, Schlager LI, Tuck versity of Washington, Marc R. Gartenberg,
RH, Sloan L, Matera R, Sarkar PS, Ashizawa Robert Wood Johnson Medical School, Rut-
T, Freudenreich CH, Mirkin SM (2011) gers, The State University of New Jersey, Grant
Expansions, contractions, and fragility of the W. Brown, University of Toronto. 2015 edi-
spinocerebellar ataxia type 10 pentanucleotide tion. Cold Spring Harbor, New York: Cold
repeat in yeast. Proc Natl Acad Sci U S A 108 Spring Harbor Laboratory Press. xvii, 233
(7):2843–2848. doi:10.1073/pnas. pages
1009409108. PubMed PMID: 21282659; 19. Hall BM, Ma CX, Liang P, Singh KK (2009)
PMCID: 3041125 Fluctuation analysis CalculatOR: a web tool for
10. Lu S, Wang G, Bacolla A, Zhao J, Spitser S, the determination of mutation rate using
Vasquez KM (2015) Short inverted repeats are Luria-Delbruck fluctuation analysis. Bioinfor-
hotspots for genetic instability: relevance to matics 25(12):1564–1565. doi:10.1093/bioin
cancer genomes. Cell Rep. doi:10.1016/j.cel formatics/btp253. PubMed PMID:
rep.2015.02.039. PubMed 19369502; PMCID: 2687991
11. Pluta AF, Zakian VA (1989) Recombination 20. Rosche WA, Foster PL (2000) Determining
occurs during telomere formation in yeast. mutation rates in bacterial populations. Methods
Nature 337(6206):429–433. doi:10.1038/ 20(1):4–17. doi:10.1006/meth.1999.0901.
337429a0. PubMed PubMed PMID: 10610800; PMCID: 2932672
 Chapter 29

Quantitative Analysis of the Rates for Repeat-Mediated

Genome Instability in a Yeast Experimental System
Elina A. Radchenko, Ryan J. McGinty, Anna Y. Aksenova,
Alexander J. Neil, and Sergei M. Mirkin

Abstract
Instability of repetitive DNA sequences causes numerous hereditary disorders in humans, the majority of
which are associated with trinucleotide repeat expansions. Here, we describe a unique system to study
instability of triplet repeats in a yeast experimental setting. Using fluctuation assay and the novel program
FluCalc we are able to accurately estimate the rates of large-scale expansions, as well as repeat-mediated
mutagenesis and gross chromosomal rearrangements for different repeat sequences.

Key words Trinucleotide repeats, Repeat expansions, Repeat-induced mutagenesis, Fluctuation assay,
Expansion rate, Mutation rate, MSS-MLE, FluCalc

1 Introduction

The fluctuation test has been a cornerstone of the field of microbial

genetics since the Luria–Delbr€ uck experiment, which determined
that mutations arise spontaneously during growth of a culture prior
to selection [1]. The mathematical methods underlying mutation
rate calculations are complicated by the fact that mutations sponta-
neously arise by chance at different generations in the culture, and
subsequently accumulate with each cell division [1, 2]. This pro-
duces fluctuations in the frequency of mutants in each independent
experiment. The mathematical methods of accounting for this
natural fluctuation have improved over time, with innovations to
account for variables such as the relative fitness of the mutant
population, partial plating of the culture, or variation in the total
population size across the cultures in the experiment [3–5].
Laboratory methods have also evolved since the days of Luria
and Delbr€ uck. An experimental strain typically contains a forward-
selection reporter, which allows the detection of mutant cells on

Marco Muzi-Falconi and Grant W. Brown (eds.), Genome Instability: Methods and Protocols, Methods in Molecular Biology,
vol. 1672, DOI 10.1007/978-1-4939-7306-4_29, © Springer Science+Business Media LLC 2018

421
422 Elina A. Radchenko et al.

particular selective media. A fluctuation test usually begins with

growth of several parallel cultures in nonselective conditions,
where a small number of inoculated cells grow to saturation. In
our experiments, we first spread cells on a Petri dish, where a single
cell grows to form a clonal colony. Individual colonies are then
resuspended in water and various dilutions are plated onto both
selective and nonselective solid media. The number of colonies
grown on each plate is used to determine the number of mutations
per cell per generation, also known as μ or the mutation rate.
Notably, however, different colonies on selective media do not
necessarily arise from the same type of mutation. A reporter gene
may be inactivated by a point mutation, a small insertion or dele-
tion causing a frameshift, a larger deletion removing part or all of
the gene, or by chromosomal events such as translocations or the
loss of a nonessential arm of the chromosome. Furthermore, muta-
tions can occur within the body of the gene, or in an important
regulatory region. Finally, reporter gene inactivation can also occur
indirectly, for instance via mutations or downregulation of genes
responsible for the uptake of a selective drug. These different events
do not necessarily have the same consequences in terms of colony
formation, with a regulatory change producing a weaker selective
pressure than a deletion of the entire marker. Variations on the
fluctuation assay may focus on one or more of these types of
mutations. A classic example is the frameshift reversion assay,
which begins with a selective marker inactivated by a 1 bp insertion,
and selects for those mutations that restore the frame of translation.
One unusual type of mutation is microsatellite instability.
Tracts of short repetitive sequences (repetitions of 2–12 bases)
can mutate in their length, with the number of repeats
either expanding or contracting. Expansions of microsatellite
sequences have been linked to a variety of human disorders [6].
The instability of microsatellite sequences can be attributed to the
formation of various non-B-form DNA secondary structures,
depending on the particular sequence. Repeats can be added a
few at a time, or by a single large-scale event. Different microsatel-
lite sequences may be more or less prone to large or small-scale
events, as well as chromosomal fragility and repeat-induced muta-
tions, all of which differ in their underlying mechanisms and rates
[7]. Importantly, the rates of these various events appear to increase
exponentially with the length of the starting repeat tract [8, 9].
Depending on the particular repeat, the rate of such events can be
higher than the average point mutation rate of the genome by many
orders of magnitude.
In the protocol below, we describe the use of our experimental
system to select for large-scale expansions of H-DNA-forming
(GAA)n repeats in S. cerevisiae [8, 10]. The yeast cells carry a
selectable cassette consisting of a modified URA3 gene, which is
counter-selectable on media containing the drug 5-fluoroorotic
 Quantitative Analysis of the Rates for Repeat-Mediated Genome Instability. . . 423

Fig. 1 Experimental system to study GAA repeats expansions in yeast. The selectable cassette contains
flanking sequences from chromosome III (depicted in black and grey), URA3 flanking and coding sequences
from chromosome V (green), intron sequence inside URA3 gene derived from ACT1 gene from chromosome VI
(blue) and TRP1 flanking and coding sequences from chromosome IV (red)

acid (5-FOA) [see Fig. 1]. This cassette is located ~1 kb downstream

of the early-firing replication origin ARS306. The URA3 gene has
been modified to include 100 GAA repeats within a 974 bp artificial
intron. Due to the limitations of the yeast spliceosome, an intron
exceeding ~1000 bp is no longer spliced, leading to 5-FOA resis-
tance in strains bearing our genetic construct. It is important to
note that this threshold is not absolute, but rather splicing effi-
ciency appears to diminish in an exponential manner as the intron
length grows past 1000 bp. Thus, we designed our system to begin
with the overall intron length close to this threshold, such that even
relatively short expansions should be sufficient to induce a level of
5-FOA resistance distinguishable from the background.
Importantly, however, URA3 can become inactivated not only
due to a repeat expansion, but also by point mutations, deletions,
translocations, downregulation, or even indirectly. In practice, each
of these situations creates varying degrees of 5-FOA resistance,
resulting in a range of colony sizes. The smallest 5-FOA-resistant
colonies are typically formed owing to indirect resistance or slight
URA3 downregulation. The largest colonies typically result from
the strongest inactivating events, such as nonsense mutations and
large deletions. Expansions typically occupy a middle ground, as
they severely but not totally reduce URA3 expression, with the
strength of this effect depending on the size of the expansion.
Very large expansions may be the result of secondary expansions,
i.e., expansions that occurred following a primary expansion.
To distinguish between these different events, we perform a PCR
on various (ideally all) 5-FOA-resistant colonies. Using primers
flanking the repetitive tract, we can measure the size of each expan-
sion, or infer the presence of a deletion/translocation (if the PCR
product is absent) or a point mutation (if there is no change in
repeat length).
A major advantage of our system is that it can be adapted to
favor the recovery of certain types of events, simply by altering the
concentration of 5-FOA in the selective media. The weakest events,
including small-scale expansions, can be observed by growing on a
low concentration of 5-FOA. The concentration of 5-FOA can be
optimized to reduce these background colonies, favoring the recov-
ery of large-scale primary expansions. If the concentration of
424 Elina A. Radchenko et al.

5-FOA is increased further, this can eliminate most expansions,

highlighting only the most damaging types of mutations, including
deletions, translocations and point mutations.
Indeed, there are a number of possible modifications to the
fluctuation test, and to our expansion system, that can be made. For
instance, the starting repeat length can be increased or decreased
[8], or the (GAA)n repeats can be swapped out for other sequences
such as SCA10 pentanucleotide (ATTCT)n repeats, or telomeric
repeats [11, 12]. The URA3 promoter can be swapped for an
inducible GAL1 promoter to study the effects of transcription
[10]. The entire cassette can be moved to different locations,
such as the nonessential arm of chromosome V, allowing for recov-
ery of arm-loss events (double-strand breaks that were not
repaired) [13]. In addition to forward selection for repeat expan-
sions, reverse-selection to a strong Ura+ phenotype can be used to
study contractions of the repetitive tracts [14]. Thus, our system is
highly adaptable to the study of various molecular mechanisms.
As written, the protocol below describes the steps used for
measuring the rate of primary expansions in our wild-type strain
with the (GAA)100 selectable cassette located ~1 kb downstream of
the early-firing origin ARS306. This is perhaps the most challeng-
ing use of the assay, as it requires optimization of 5-FOA concen-
tration and growth times, as well as careful scrutiny of the
distribution of repeat lengths. We have included a number of
notes in the protocol to guide with this optimization, as well as
suggestions for where the assay can be modified. Variations of this
protocol can be found in our previous publications [8–12, 14], but
this document represents our most up-to-date and refined version
of the assay.
Finally, we introduce here a new tool, named FluCalc, for
calculating mutation rates from these fluctuation tests (see Fig. 2).
Although several mathematical tools such as SALVADOR, or web-
tools such as FALCOR and bz-rates were designed for accurate
estimation of mutation rate [15–18], they appeared either unsuit-
able or inconvenient for our experiments. FluCalc uses the Ma-
Sandri–Sarkar maximum likelihood estimation (MSS-MLE) equa-
tions for calculation, regarded as one of the most accurate estima-
tors available [19]. MSS-MLE is also applicable across most
mutational spectra [20]. FluCalc calculates the values for mutant
frequencies, mutations per culture, mutation rates and 95% confi-
dence intervals of mutation rates. FluCalc also automatically incor-
porates plating efficiency calculations to account for dilutions.
FluCalc displays a full description of all equations used in the
calculations, and shows all values of the intermediate variables at
each step. FluCalc is presented in a web format suitable for all
modern browsers, and does not rely on plugins that become out-
dated and broken. FluCalc allows one to download and save the
data as a text file, which can be subsequently reopened by the
 Quantitative Analysis of the Rates for Repeat-Mediated Genome Instability. . . 425

Fig. 2 Screenshot of the FluCalc interface. A user needs to enter the number of colonies on both selective (Csel)
and complete media (Ccom), a dilution factor (Dsel and Dcom), and a volume of cultures (Vsel and Vcom) plated on
both media. A volume of the initial culture (Vtot) can be changed from its default value of 200 μL. Raw values
and corrected values for m—number of mutations; μ—mutation rate; CI—confidence intervals for μ
426 Elina A. Radchenko et al.

program or imported into a spreadsheet program. This feature is

especially handy when processing and comparing large amounts of
data. We hope that FluCalc will be useful for many researchers
wanting an accurate and efficient tool for mutation rate calcula-
tions. The tool is available at http://ase.tufts.edu/biology/labs/
mirkin/resources/.

2 Materials

All experiments are described based on the Saccharomyces cerevisiae

WT strain CH1585 (MATa, leu2Δ1, trp1Δ63, ura3Δ52, and
his3Δ200) with (GAA)100 repeats inserted in the intron-containing
URA3 gene on chromosome III (see Note 1).

2.1 Nonselective 1. 15% glycerol stock solution: 300 μL of autoclaved 50% glycerol
Growth Period solution þ700 μL of YPD.
2. YPD þ Uracil plates (example makes 1 L): 1 L dH2O, 10 g
yeast extract, 20 g peptone, 20 g dextrose, 20 g agar. Auto-
clave, allow to cool slightly, then add 24 mL uracil from a
2 mg/mL stock solution. Pour into Petri dishes, solidify over-
night, seal and store at 4 C.

2.2 Selective Growth 1. 5-FOA media (example makes 1 L): Autoclaved portion:
Period 500 mL dH2O, 20 g agar, 20 g dextrose, 2 g synthetic com-
plete amino acid supplement minus uracil with added adenine,
6.7 g yeast nitrogen base without amino acids or carbohydrates
and with ammonium sulfate. Nonautoclaved portion: 500 mL
dH2O, 1 g 5-FOA, 50 mg uracil—shake at 37 C until 5-FOA is
completely dissolved and use a 0.2 μm bottle-top vacuum filter
to sterilize the solution. Allow autoclaved portion to cool
slightly, then combine both portions, mix on stirrer and imme-
diately pour into Petri dishes. Let the media solidify overnight
and use the following day (see Note 2).
2. YPD media (example makes 1 L): 1 L dH2O, 10 g yeast extract,
20 g peptone, 20 g dextrose, and 20 g agar. Autoclave, allow to
cool slightly, and pour into Petri dishes. Let the media solidify
overnight, seal and store at 4 C.

2.3 Colony PCR 1. Lyticase solution: 2.5 mg/mL lyticase (L2524 Sigma
2000 units/mg protein) in 0.9 M sorbitol, 0.1 M EDTA
(pH 7.4) (see Note 3). Zymolyase (e.g., 100 T Zymo Research)
performs equally well.
2. PCR primers (see Note 4).
3. PCR mastermix (example makes 24 μL):
 Quantitative Analysis of the Rates for Repeat-Mediated Genome Instability. . . 427

H2O 13.5 μL
5 Green GoTaq buffer (Promega) 5 μL
10 mM dNTPs 0.25 μL
10 μM Forward Primer 2.5 μL
10 μM Reverse primer 2.5 μL
Sibenzyme Taq 0.25 μL

2.4 Gel 1. 10TBE stock solution: 0.89 M Tris, 0.89 M borate, 0.02 M
Electrophoresis EDTA. Store at room temperature.
2. Ultrapure high-resolution agarose.
3. NEB Quick-Load 100 bp or 50 bp DNA ladder.
4. 1% ethidium bromide solution.

3 Methods

3.1 Nonselective 1. Streak a small amount of yeast from 15% glycerol frozen stock
Growth Stage onto a YPD plate supplemented with uracil such that distinct
(See Note 5) individual colonies form (see Note 6).
2. Incubate for 40 hours at 30 C (see Note 7).

3.2 Selective Growth 1. Fill a sterile 96-well polypropylene, U-well microtiter plate
and Total Cell Count with 200 μL of sterile deionized water in the 1st row and
(See Fig. 3) 180 μL in rows 2–5.
2. Use a small plastic inoculating loop to scoop up an average-
sized colony and resuspend it in the 1st row of the plate. Repeat
until 8–12 similarly sized colonies have been resuspended, one
per well (see Note 8). Mix thoroughly with a multichannel
pipette.
3. Perform 10 serial dilutions, transferring 20 μL from the 1st to
2nd rows, and then into subsequent rows, mixing thoroughly
at each step (see Note 9).
4. Mark corresponding YPD and 5-FOA plates with the strain
name and colony number.
5. Plate 100 μL from the first row onto a 5-FOA plate and 100 μL
from the 5th row onto a YPD plate for each sample (see Note
10) and spread until dry with a large disposable plastic inocu-
lating loop.
6. Spin down microtiter plate to collect yeast cells for PCR analy-
sis, discard supernatant, add 1.5 μL lyticase solution, mix thor-
oughly and go to Subheading 3.3.3 (see Note 11).
428 Elina A. Radchenko et al.

Fig. 3 Scheme of the fluctuation test

7. Incubate 5-FOA and YPD plates at 30 C for 3 days (see

Note 12).
8. Count individual colonies on all plates (see Note 13).
9. (Optional) Using a flatbed scanner, scan images of all plates.
Load images into OpenCFU colony counting software. Adjust
threshold and radius values until all 5-FOA-resistant colonies
are included and all background growth is excluded from the
count (see Note 14).

3.3 PCR Analysis of 1. If necessary, randomly subdivide 5-FOA plates into 1/2, 1/4
Repeats (See Note 15) or 1/8 size wedges containing no less than 4 and no more than
16 colonies within a wedge from each plate. Count total num-
3.3.1 Option A
ber of colonies within random wedge of each 5-FOA plate.
(See Note 16)
2. Pick all colonies from within the selected 5-FOA plate or
wedge using a pipette tip and resuspend one colony in each
well of a 96-well PCR plate containing 1.5 μL lyticase solution
per well.
 Quantitative Analysis of the Rates for Repeat-Mediated Genome Instability. . . 429

3.3.2 Option B (See Note 1. For each 5-FOA plate, choose four 5-FOA-resistant colonies
17) and Option C (See Note for PCR analysis.
18) 2. Pick colonies using a pipette tip and resuspend one colony in
each well of a 96-well PCR plate containing 1.5 μL lyticase
solution per well.

3.3.3 Continue from 1. Incubate cells in lyticase solution for 8–15 min at 37 C
Option A, B or C 2. Add 50 μL of PCR-grade water to all wells. Resuspend if
necessary.
3. Incubate plate at 100 C for 5 min.
4. Spin plate at 2500–3000  g for 2–5 minutes to pellet cell
debris. The solution containing yeast genomic DNA can be
stored for up to 12 months at 20 C and repeatedly analyzed.
5. Prepare PCR master mix for the appropriate number of
samples.
6. Add 12–24 μL of PCR master mix into each well of a new 96-
well PCR plate.
7. Add 1 μL of genomic DNA supernatant to each PCR reaction.
8. Run PCR program:
(a) 95 C—15 s.
(b) 94 C—15 s.
(c) 72 C—2 min.
(d) Go to Step 2  30.
(e) 72 C—2 min.
(f) 12 C—hold.

3.4 Gel 1. Prepare a 1.5% agarose gel using ultrapure agarose and 0.5
Electrophoresis TBE (see Note 19). Add 2 μL of ethidium bromide per 100 mL
(See Fig. 4) of TBE.
2. Load samples along with 100 bp DNA ladder, run gel, and
image.
3. Calculate expansion sizes by subtracting the starting length of
the unexpanded PCR product from expanded PCR product
sizes and dividing by 3 (for triplet repeats) (see Note 20).
4. For each individual experimental strain, visualize the distribu-
tion of expansion sizes by graphing the percentage of expanded
products that fall into 10-repeat bins (0–10 repeats added,
11–20 repeats added, etc.) (see Note 21).

3.5 Expansion 1. Insert the following information into FluCalc to obtain an

Rate Calculation accurate expansion rate with corresponding 95% confidence
(See Note 22) interval:
430 Elina A. Radchenko et al.

Fig. 4 Representative gel electrophoresis of the PCR analysis of 5-FOA-resistant colonies, originated in a
strain containing our selectable cassette with the (GAA)100 repeat. L—ladder, e—expansion, d—deletion,
m—mutation, c—contraction with a simultaneous mutation in the URA3 body, de—double expansion. The
initial length of PCR product containing (GAA)100 repeat is ~350 bp that is marked by arrow

(a) Vtot—the total volume of the culture following resuspen-

sion of colonies in step 2 of Subheading 3.2 (typically
200 μL).
(b) Csel—the observed number of colonies with expanded
repeats from each plate or sector (adjusted to reflect the
results of the gel electrophoresis) (see Note 23).
(c) Ccom—the observed number of colonies on nonselective
media.
(d) Dsel—the dilution factor used for plating on selective
media (e.g., if liquid from the second row of a 10 serial
dilution was plated on 5-FOA, then Dsel ¼ 10).
(e) Dcom—the dilution factor used for plating on nonselective
media (e.g., if liquid from the fifth row of a 10 serial
dilution was plated on YPD then Dcom ¼ 10,000).
(f) Vsel—the volume of culture plated per plate or sector on
selective media (see Note 24).
(g) Vcom—the volume of culture plated per plate on nonse-
lective media.
2. Download and save the data in FluCalc.txt format so that it can
be subsequently reopened by the program.
3. Click on the “Calculate” button to get results for the expansion
rate, presented as either raw values or corrected values
(see Note 25). Frequencies for each plate will also be reported
(see Note 26). 95% confidence intervals are also reported (see
Note 27).
 Quantitative Analysis of the Rates for Repeat-Mediated Genome Instability. . . 431

4 Notes

1. The strain used for the example in this protocol was described
and thoroughly examined in Shah et al. 2012 [9]. A key feature
of this strain is the “balanced” intron length just under the
splicing threshold length, allowing a single expansion to inacti-
vate the URA3 gene. The rate of expansions for a GAA100
repeat was determined to be ~2  105 expansions per cell
per generation. Altering the starting repeat length or sequence
can dramatically change this rate. Our protocol is ideal for
measuring rates in the range of 106 to 103. Measuring
rates below this range is difficult due to the larger number of
cells required for plating. Measuring rates above this range
results in large natural fluctuations, making this type of assay
inappropriate. We recommend adjusting the starting length of
the repeat until the rate is close to 105. This will allow the
investigation of various mutants or experimental conditions
that raise or lower the rate within the measurable range.
2. 5-FOA will remain effective when stored at 4 C for a period of
weeks. However, as water slowly evaporates from the plates,
this will raise the effective concentration of 5-FOA, which can
shift the selection toward larger expansions, point mutations,
and deletions. Thus it is important, if determining the rate of
expansions in a mutant strain, to concurrently retest the wild-
type strain to help control for subtle differences in the media.
Plastic wrap can be used to reduce evaporation, but it is best to
use the plates within days. The above recipe is for 0.1% 5-FOA
media, which is ideal for the selection of primary repeat expan-
sions. To select for deletions/point mutations/translocations,
a concentration of 0.15% works well. The strength of the 5-
FOA media can also be increased by lowering the pH, though
we find this method of adjustment to be less precise.
3. The pH of the lyticase solution should be kept around 7.4–7.5
and is critical for the activity of the enzyme. This solution
should be stored in small aliquots at 20 C.
4. Primers should be closely flanking the repeats, and should have
Tm above 72 C. The Tm should be high to discourage sec-
ondary structure formation which can lead to PCR failures that
create faint or smeared bands. The PCR product should be as
small as possible to facilitate accurate length determination.
PCR products with longer nonrepetitive flanks can sometimes
help to reduce PCR failures, but these longer products will not
be as accurate for measuring small changes in length.
5. If assessing any experimental variables (such as temperature
sensitivity, chemical perturbance, or gene induction), it is
important to introduce them during this stage of growth.
432 Elina A. Radchenko et al.

6. Extra uracil is added to the media to prevent cells from becom-

ing dependent on the URA3 gene, which will be inactivated by
an expansion. If petites are particularly frequent following
unfreezing, the strains can first be patched onto YPEG plate
for 1 day. The YPEG patches can then be used to streak for
single colonies on YPD þ Uracil. For strains in which the
URA3 promoter is replaced by the galactose-inducible GAL1
promoter, start with a 5 mL overnight culture in YPD þ Raffi-
nose to remove glucose inhibition of the GAL1 promoter.
Dilute the overnight culture in deionized water and plate on
YPD or YPGal, supplemented with uracil.
7. This amount of time can be extended for any slow-growing
strains or alternate growth conditions. However, it is important
not to grow the colonies for too many generations, which can
result in a higher probability of secondary expansions. This is
because a primary expansion will increase the next generation’s
starting repeat length, which will in turn have an exponentially
higher rate of expansion.
8. It is important to pick colonies that are of the same size. Most
methods to calculate mutation rates (including MSS-MLE) rely
on an assumption that culture sizes are the same (within 20% of
the coefficient of variation) across the experiment [5, 20]. To
decrease the inter-culture variation, do not choose the largest
or smallest colonies on the plate—it is better to aim for average
sized colonies. When testing our wild-type GAA100 strain, we
choose single colonies that are about 1 mm in diameter when
uncrowded. The starting colony size can be adjusted in antici-
pation of higher or lower expansion rates. After colonies have
been counted (Step 8 in Subheading 3.2), if there is a YPD
plate that deviates significantly from the median, both the YPD
and corresponding 5-FOA plate should be excluded from fur-
ther analysis.
9. Accurate pipetting is critical throughout the serial dilution, as
small random errors can be magnified after each step. Accuracy
can be improved through the use of low-retention tips, by
discarding tips between each suspension, and by visually check-
ing for air bubbles, droplets stuck in the tips, or other pipetting
errors.
10. This recommendation is applicable to our wild-type GAA100
strain, and should result in uncrowded single colonies. Both
the volume and choice of dilution can be altered for different
mutant backgrounds or growth conditions.
11. PCR analysis of the starting colonies should confirm that
expansions did not take place prior to the nonselective growth
period. All future analysis depends on these cells beginning
from the same initial repeat length. Encountering expanded
Quantitative Analysis of the Rates for Repeat-Mediated Genome Instability. . . 433

or contracted clones at this stage should be rare, but strains

with very high rates of instability may require careful
monitoring.
12. The growth period may be altered if necessary. Alterations to
the temperature or media components at this stage will not
affect the rate at which mutations were generated during the
nonselective growth phase, but may affect the recovery of those
events for counting. Only change these conditions if you have
reason to believe you are not recovering certain events. Also, if
using a galactose-inducible strain, the 5-FOA media must also
contain galactose in order for the URA3 gene to be active and
thus counter-selectable.
13. For YPD plates, all colonies must be counted. For 5-FOA
plates, there will be a certain amount of small background
colonies, which do not have expansions or other mutations of
interest. A rule of thumb is that if the colony is too small to be
easily picked up with a pipette tip, it cannot be analyzed by
PCR and therefore should not be counted. It is possible to
empirically determine for each strain the relationship between
colony size and mutation status. In any case, it is important to
be consistent between those colonies counted and those cho-
sen for PCR analysis.
14. The OpenCFU program is freely available at http://opencfu.
sourceforge.net/. An alternative is the Image J program with a
cell-counter plugin (https://imagej.nih.gov/ij/index.html).
Use a consistent image size for every plate (a 300 dpi scan
will produce 1000  1000 pixels per plate). Image the under-
side of the plate with the top cover removed. This will produce
the most consistent and clear images, allowing proper compar-
ison of colony sizes across different plates.
15. The goal of the PCR analysis is to determine which 5-FOA
resistant colonies contain expansions, as well as the size of
individual expansions. An ideal solution would involve PCR
analysis of every 5-FOA-resistant colony on a plate. To achieve
a manageable number of colonies on each selective plate (e.g.,
~10 per plate) you may need to dilute cells further or even plate
several dilutions on 5-FOA plates in a given experiment. While
performing PCR for each colony on every 5-FOA plate may
become prohibitively costly and laborious, it eliminates sam-
pling error from confidence intervals for the mutation rates.
Because of this consideration, we recommend to PCR all
5-FOA resistant colonies; if it is not practically doable, three
options to make PCR analysis feasible are presented below.
16. Option A is to plate a lower density of cells on each plate, or to
count and PCR every colony only from a sector of each plate.
This approach eliminates any kind of bias (especially by size)
434 Elina A. Radchenko et al.

that might occur while choosing colonies for PCR analysis. Of

course, the subdivision of each plate must occur randomly in
order to avoid unintentionally choosing a sector with larger or
smaller colonies. One solution is to make a “Wheel-of-For-
tune,” marking sectors on the top cover of a Petri plate, then
place the “wheel” over each plate, spin, and mark each colony
that lands within the prechosen sector. If there are a large
number of colonies in the sector, a smaller sector may be
used. Note, that the size of a sector should be the same for
every 5-FOA plate within a batch. If very small sectors are
needed, it may be better to perform the fluctuation test again
with a lower concentration of cells. Option A also has the
advantage of being mathematically simpler when performing
mutation rate calculations. The caveat is that by plating fewer
cells, the fluctuations in rate may be higher due to the greater
number of dilutions and pipetting steps. Counting from sec-
tors of a plate also assumes that the cells were spread evenly.
17. Option B is to PCR a set number of colonies on each plate, and
to apply that percentage to the total number of 5-FOA-resis-
tant colonies. Option B has several advantages. This method
allows a higher plating density, provided that the colonies are
not overcrowded to the point where they affect the growth of
their neighbors. It also has the advantage of producing a pre-
dictable number of PCR reactions per fluctuation test, allowing
more high-throughput processing. A caveat is that the user
must choose a few colonies from each plate to represent the
entire population, potentially introducing bias. Some of this
bias can be eliminated, since we know roughly the relationship
between colony size and mutation type. To best account for
the various events that cause 5-FOA resistance, choose colonies
by size in proportion to their approximate representation in the
population, i.e., if one out of four colonies are larger than
average, then choose three average sized colonies and one
larger colony for analysis. Although this tactic is not completely
void of observer’s bias, it might provide a comprehensive anal-
ysis of the population. It is important not to choose any small
background colonies if they were not included in the total
colony count. Another caveat of Option B is that the additional
step of the percentage calculation can affect the mathematics of
the mutation rate calculation. Rate estimations can be skewed,
particularly if expansions (or other events analyzed) do not
represent the majority of the population. Because FluCalc
applies plating efficiency adjustments, the calculated mutation
rate in Option A is expected to be lower than in Option B. It is
best, therefore, to be consistent in choosing either Option A, B
or C for all experiments.
Quantitative Analysis of the Rates for Repeat-Mediated Genome Instability. . . 435

18. In Option C, as in Option B, a fixed number of colonies per

plate are analyzed. However, this step is treated as the mathe-
matical equivalent of a dilution, with a different dilution factor
for each plate. Like Option A, the actual number of observed
events can then be used for each calculation, though this
requires applying an additional plating efficiency correction to
each plate. The calculations accounting for this “variable dilu-
tion” are not currently integrated into FluCalc. See Note 25
and refs. 12, 14 for more details.
19. In our current system, the repetitive tract starting length PCR
is 350 bp, and expansions will typically be in the range of
375–1000 bp. A 1.5% agarose gel is appropriate for this size
range. Switching to a 2–2.5% agarose gel in 1 TAE buffer can
help provide better resolution when analyzing small-scale
changes [14]. Voltage and running time should be adjusted
for clear separation of bands within this size range, and size
estimation to the nearest 25 bp. Bands can also be analyzed
electronically, for example via the Bio-Rad QuantityOne pro-
gram, for greater accuracy. We also recommend large format
gels with multichannel-compatible wells, due to the high vol-
ume of PCR samples that must be analyzed per assay.
20. In some cases, a gel lane may contain two or more bands,
indicating the presence of a secondary expansion or contrac-
tion that occurred following plating on 5-FOA (as opposed to
secondary expansions that occurred during the nonselective
growth stage, which will only produce a single band). We
typically attribute the preplating event to the darkest band on
the gel (representing the greatest number of cells within the
colony), with the caveat that shorter repeat lengths typically
result in a more successful PCR reaction. On the other hand, an
extremely long expansion may push the limits of the PCR
reaction, creating a smeared band. We typically use the darkest
part of the smear as the length of the preplating expansion. If
one of the two bands has the original size of the repeat, the
expansion most likely occurred postplating and the colony
should be counted as nonexpanded.
21. We have found that primary expansions occur in a normal
distribution with a median of 50–60 repeats added [9]. Only
experiments in which the expansion size histogram follows a
unimodal distribution will produce an accurate estimate of
mutation rate. A bimodal distribution may indicate the pres-
ence of secondary expansion events. Secondary expansions can
be selected for if the 5-FOA concentration is too high. Second-
ary expansions become more and more prevalent as starting
repeat tract length increases, in mutant strains with high expan-
sion rates, or if the nonselective growth period is too long
(because primary expansions will increase the likelihood of
436 Elina A. Radchenko et al.

further expansions in the next generation). It is also not

uncommon for the distribution to reveal that smaller-scale
expansions were not allowed to grow due to a higher 5-FOA
concentration. Expansion rates calculated in nonnormally
distributed assays will thus reflect an uncertain mixture of
primary, secondary, and missed events.
22. This example protocol is used to determine the rate of repeat
expansions. However, FluCalc can be used to determine rates
for any type of observed mutation, depending on which set of
Csel values are entered. For example, the user can instead enter
the number of colonies with an observed deletion, or the total
number of 5-FOA-resistant colonies.
23. For Options A and C, enter the number of colonies with
expansions after running gel electrophoresis. For Option B,
the numbers entered for Csel should be the number of observed
5-FOA-resistant colonies multiplied by the overall percent of
expansions (i.e., total number of colonies with expansions
divided by total number of analyzed colonies for all cultures
in the experiment).
24. If counting sectors (Option A), assume an even distribution of
liquid sample during plating (e.g., if 100 μL of solution was
plated on 5-FOA and all colonies from a 1/8 wedge were
analyzed for expansions then Vsel ¼ 12.5 μL)
25. Corrected values are raw values corrected for plating efficiency,
resulting in a more accurate representation of the mutation rate
per cell per generation. If using Option C for PCR analysis,
plating efficiency calculations must be performed manually.
This is because a fixed amount of PCRs represent a variable
amount of colonies, and thus the plating efficiency correction is
not the same for each plate. We denote the following calcula-
tion method as VZ-MLE, or variable z maximum likelihood
estimation. This method also relies on the assumption that the
cultures were of roughly equal size (see Note 8). We hope to
fully integrate VZ-MLE into FluCalc in the near future. At
present, additional steps must be performed to employ Flu-
Calc: Enter Csel values as the actual number of observed expan-
sions (e.g., 2 out of 4) obtained for each individual plate. Enter
all other values (Vtot, Dsel, Vsel, Ccom, Dcom and Vcom) as
defined in step 1 of Subheading 3.5 and press “Calculate” to
obtain the raw mutation rate and confidence intervals. Next,
calculate the z-factor for each culture by dividing the number
of colonies analyzed by PCR by total number of 5-FOA resis-
tant colonies in the culture, which is CselDselVtot/Vsel. From
there, determine the plating efficiency coefficient (P ¼ 1-z/zln
(z)) for each plate, and then the average value of P (AVP). Then
multiply the rate and confidential limits obtained with FluCalc
by AVP.
 Quantitative Analysis of the Rates for Repeat-Mediated Genome Instability. . . 437

26. Step 5 of Subheading 3.2 describes resuspending and plating

dilutions of each colony on corresponding 5-FOA and YPD
plates. Thus the number of entries for Csel and Ccom should be
the same, and FluCalc will determine individual frequencies for
each corresponding set of plates, making the assumption that
the first 5-FOA plate corresponds to the first YPD plate, and so
forth. If the number of entries for Dcom and Ccom are not the
same, FluCalc assumes that individual 5-FOA and YPD plates
do not correspond, and instead uses the mean of Ccom for each
calculation.
27. The reported confidence intervals are accurate for any given
experiment, but they may not fully reflect the inherent nature
of the fluctuation assay. Due to the relatively high baseline rate
of expansion events, the median expansion rate of our wild-
type GAA100 strain does indeed fluctuate from trial to trial.
This natural fluctuation is usually within 50% of the typical
rate, however we do caution at making conclusions for mutant
strains with any less than a threefold difference from the wild-
type. Despite all of the complications and caveats, we find this
assay to be a reliable and consistent method to determine
expansion rates.

Acknowledgments

We thank Alexander A. Shishkin and Kartik A. Shah for their

invaluable contributions in developing cassettes to study repeat
instability, and for developing experimental protocols for the selec-
tion and PCR procedures, Timofei S. Bondarev for developing
FluCalc program, and Durwood Marshall for statistical consulting.
This study was funded by NIH grants GM105473 and GM60987
to S.M.M and RFBR grant #15-04-08658 and research project in
the Centre for Molecular and Cell Technologies (Research Park,
Saint-Petersburg State University) for A.Y.A.

References

1. Luria SE, Delbruck M (1943) Mutations of 4. Zheng Q (2015) A new practical guide to the
bacteria from virus sensitivity to virus resis- Luria–Delbr€
uck protocol. Mutat Res
tance. Genetics 28(6):491–511 781:7–13. doi:10.1016/j.mrfmmm.2015.08.
2. Drake JW (1991) A constant rate of spontane- 005
ous mutation in DNA-based microbes. Proc 5. Zheng Q (2016) A second look at the final
Natl Acad Sci U S A 88:7160–7164. doi:10. number of cells in a fluctuation experiment. J
1073/pnas.88.16.7160 Theor Biol 401:54–63. doi:10.1016/j.jtbi.
3. Foster PL (2006) Methods for determining 2016.04.027
spontaneous mutation rates. Methods Enzy- 6. Mirkin SM (2007) Expandable DNA repeats
mol 409:195–213. doi:10.1016/S0076-6879 and human disease. Nature 447
(05)09012-9 (7147):932–940. doi:10.1038/nature05977
438 Elina A. Radchenko et al.

7. Kim JC, Mirkin SM (2013) The balancing act 14. Aksenova AY, Han G, Shishkin AA et al (2015)
of DNA repeat expansions. Curr Opin Genet Expansion of interstitial telomeric sequences in
Dev 23(3):280–288. doi:10.1016/j.gde. yeast. Cell Rep 13(8):1545–1551. doi:10.
2013.04.009 1016/j.celrep.2015.10.023
8. Shishkin AA, Voineagu I, Matera R et al (2009) 15. Zheng Q (2002) Statistical and algorithmic
Large-scale expansions of Friedreich’s ataxia methods for fluctuation analysis with SALVA-
GAA repeats in yeast. Mol Cell 35(1):82–92. DOR as an implementation. Math Biosci 176
doi:10.1016/j.molcel.2009.06.017 (2):237–252. doi:10.1016/S0025-5564(02)
9. Shah KA, Shishkin AA, Voineagu I et al (2012) 00087-1
Role of DNA polymerases in repeat-mediated 16. Zheng Q (2016) rSalvador 1.5: an R tool for
genome instability. Cell Rep 2(5):1088–1095. the Luria–Delbr€ uck fluctuation assay. http://
doi:10.1016/j.celrep.2012.10.006 eeeeeric.github.io/rSalvador. Accessed 7 July
10. Shah KA, McGinty RJ, Egorova VI et al (2014) 2016
Coupling transcriptional state to large-scale 17. Hall BM, Ma CX, Liang P et al (2009) Fluctu-
repeat expansions in yeast. Cell Rep 9(5): ation AnaLysis CalculatOR: a web tool for the
1594–1602. doi:10.1016/j.celrep.2014.10.048 determination of mutation rate using Luria-
11. Cherng N, Shishkin AA, Schlager LI et al Delbruck fluctuation analysis. Bioinformatics
(2011) Expansions, contractions, and fragility 25(12):1564–1565. doi:10.1093/bioinformat
of the spinocerebellar ataxia type 10 pentanu- ics/btp253
cleotide repeat in yeast. Proc Natl Acad Sci U S 18. Gillet-Markowska A, Louvel G, Fischer G
A 108(7):2843–2848. doi:10.1073/pnas. (2015) bz-rates: a web tool to estimate muta-
1009409108 tion rates from fluctuation analysis. G3
12. Aksenova AY, Greenwell PW, Dominska M Bethesda 5(11):2323–2327. doi:10.1534/g3.
et al (2013) Genome rearrangements caused 115.019836
by interstitial telomeric sequences in yeast. 19. Sarkar S, Ma WT, Sandri GH (1992) On fluc-
Proc Natl Acad Sci U S A 110 tuation analysis: a new, simple and efficient
(49):19866–19871. doi:10.1073/pnas. method for computing the expected number
1319313110 of mutants. Genetica 85(2):173–179. doi:10.
13. Schmidt KH, Pennaneach V, Putnam CD et al 1007/BF00120324
(2006) Analysis of gross-chromosomal rearran- 20. Rosche WA, Foster PL (2000) Determining
gements in Saccharomyces cerevisiae. Methods mutation rates in bacterial populations. Meth-
Enzymol 409:462–476. doi:10.1016/S0076- ods 20(1):4–17. doi:10.1006/meth.1999.
6879(05)09027-0 0901
 Chapter 30

Measuring Dynamic Behavior of Trinucleotide Repeat Tracts

In Vivo in Saccharomyces cerevisiae
Gregory M. Williams and Jennifer A. Surtees

Abstract
Trinucleotide repeat (TNR) tracts are inherently unstable during DNA replication, leading to repeat
expansions and/or contractions. Expanded tracts are the cause of over 40 neurodegenerative and neuro-
muscular diseases. In this chapter, we focus on the (CNG)n repeat sequences that, when expanded, lead to
Huntington’s disease (HD), myotonic dystrophy type 1 (DM1), and a number of other neurodegenerative
diseases. We describe a series of in vivo assays, using the model system Saccharomyces cerevisiae, to determine
and characterize the dynamic behavior of TNR tracts that are in the early stages of expansion, i.e., the so-
called threshold range. Through a series of time courses and PCR-based assays, dynamic changes in tract
length can be observed as a function of time. These assays can ultimately be used to determine how genetic
factors influence the process of tract expansion in these early stages.

Key words Trinucleotide repeat, Expansion, Contraction, Polymerase chain reaction, Saccharomyces
cerevisiae, DNA replication, Repeat tract dynamics, Microsatellite instability

1 Introduction

Trinucleotide repeat (TNR) sequences, like all repetitive sequences,

pose a problem for DNA metabolism. Stretches of DNA containing
repetitive nucleotides (mono-, di-, tri-, tetra-, etc.), known as
regions of microsatellites are inherently difficult to replicate accu-
rately. These repetitive sequences are hotspots for genetic mutation
due to the increased frequency of replicative polymerase slippage
events that can occur during replication through these regions
[1–4]. This leads to an elevated rate of change in the number of
repeats and is known as microsatellite instability. If left unrepaired,
these slippage events are mutagenic and can be detrimental to the
cell. (CNG)n TNR sequences present an additional hurdle to accu-
rate DNA replication because of their propensity to form secondary
structure (due to C–G base pairing) when the DNA is single-
stranded, as can happen during replication, particularly during
lagging strand synthesis [5, 6]. While the mismatch repair

Marco Muzi-Falconi and Grant W. Brown (eds.), Genome Instability: Methods and Protocols, Methods in Molecular Biology,
vol. 1672, DOI 10.1007/978-1-4939-7306-4_30, © Springer Science+Business Media LLC 2018

439
440 Gregory M. Williams and Jennifer A. Surtees

(MMR) system, particularly Msh2-Msh3, typically directs repair of

these slippage events [7–9], these structures are refractory to MMR
[10]. In fact, Msh2 and Msh3 have been shown to promote TNR
expansions in a variety of systems [6, 11–18].
The addition of TNRs leads to expansion of the TNR tract and
can be within a coding region or a noncoding region of DNA.
Expansions of TNR sequences are associated with several hereditary
neurological diseases, including Huntington’s disease (HD), myo-
tonic dystrophy (DM1), and Friedrich’s ataxia (FRDA) [18–21].
As the tract size increases, the expression and/or function of a
protein is increasingly compromised, eventually leading to disease
(Fig. 1). The transition from a stable repeat tract to an expanded,
pathogenic tract is a critical step in the development of TNR-
mediated disease, but is poorly understood. However, it is known
that a small change in tract size, with the addition of only a few
repeats, can mean the difference between stability and instability.
Therefore, at this critical stage, a small reduction in tract length, or
in the probability of expansion, has significant implications for the
prevention and treatment of disease. Tract lengths substantially
affect age of onset, penetrance of the disease, and severity of disease
symptoms [22–24].
Initial expansions bring the tract into a “threshold-range” in
which the tracts are not pathogenic but are increasingly unstable
and are particularly prone to expansions [15, 17, 22, 25] (Fig. 1).
Such intermediate alleles for HD (~27-35 CAG repeats) have
recently been reported in ~6% of the general population [23], and
these individuals are carriers for disease. Small increases in tract

# of Repeats in Tract
Expanded past threshold into
Pathogenic affected range, individuals
symptomatic for disease

Threshold-Length Increased susceptibility to

further TNR expansion

Normal Unaffected range, TNR stably

maintained

TNR Tract

Fig. 1 TNR tracts within the normal range (which is tract-dependent) are stably maintained within that range.
However, through mechanisms that remain unclear, a TNR tract can expand, increasing the number of repeats
within the tract. Initially this brings the tract into a “threshold-length” range, in which these somewhat longer
tracts are not pathogenic, but are increasingly susceptible to expansion; these individuals are carriers for
disease. Once a tract has expanded sufficiently, it crosses a threshold; tracts above this threshold (which is
disease-specific) are pathogenic and cause disease. As the size of the tract increases, it becomes increasingly
unstable and prone to changes in length, particularly expansions
 Measuring Dynamic Behavior of Trinucleotide Repeat Tracts. . . 441

length make these individuals, and their offspring, at risk for dis-
ease. As a result, it is of interest to understand the dynamic behavior
of TNR tracts within the threshold or intermediate range. Histori-
cally, the field has relied on end-point experiments, including single
sperm typing [26–30] to assess changes in tract length. These
studies led to modeling that predicted that threshold-length TNR
tracts will preferentially increase in increments smaller than the
repeats themselves [31–33]. Using the dynamic system described
in this chapter, we recently demonstrated that threshold length
tracts are highly dynamic, they increase (or decrease) in increments
as small as a single repeat unit and the changes in tract length are
biased toward expansion [17], consistent with our in vitro observa-
tions [6].
Assessing dynamic length changes in TNR tracts is challenging
in higher eukaryotic systems [34]. Our laboratory has developed a
set of genetic assays, in Saccharomyces cerevisiae, designed to assess
repeat tract length with one repeat resolution under several differ-
ent growth conditions, ranging from large heterogeneous cell
populations in liquid culture down to repeat dynamics originating
from a single cell. In this protocol, based on a system developed by
the Lahue lab [37], we focus specifically on (CAG)n and (CTG)n
repeat tracts within the threshold range, i.e., those poised to
expand—or contract. We describe a series of related approaches
that enable monitoring of the tract length as a function of time,
using the model organism Saccharomyces cerevisiae. Using these
approaches, we can detect dynamic tract length changes and start
to understand the underlying mechanism(s) (including MMR) that
lead to tract instability. These assays can also be used to determine
the effect of a therapeutic drug on repeat tract dynamics as a
function of time.

2 Materials

Make up all buffers with deionized water (dH2O). Store at room

temperature unless otherwise noted.

2.1 General 1. 10
TBE: For 1 L: 108 g Tris base, 55.0 g boric acid, 7.44 g
Reagents ethylenediaminetetraacetic acid (EDTA). The 0.5
TBE solu-
tion is 45 mM Tris–borate, 1 mM EDTA.
(a) Add each of the measured components into a beaker
containing a stir bar and 500 mL of deionized water and
place beaker onto a stir plate.
(b) Allow components to dissolve in solution (~1 h). Heat can
be used to assist components into solution (not to exceed
60  C).
442 Gregory M. Williams and Jennifer A. Surtees

(c) Bring dissolved solution up to a final volume of 1 L with

dH2O. Store at room temperature.
(d) To make polyacrylamide gel running buffer, dilute the
10
TBE buffer down to 0.5
TBE (950 mL
dH2O þ 50 mL 10
TBE).
2. Solutions for lithium acetate transformations:
(a) 1 M lithium acetate (LiAc), pH 7.5 (adjusted with a few
drops of diluted acetic acid). Autoclave.
(b) 10
TE (100 mM Tris–HCl pH 7.5, 10 mM EDTA,
pH 7.5). Autoclave.
(c) 50% PEG (50% w/v). Filter-sterilize.
(d) LiAC–TE Mix: 1 mL of 1 M LiAc, 1 mL of 10
TE, and
8 mL of sterile dH2O. Make fresh.
(e) 40% PEG Mix: 1 mL of LiAc, 1 mL 1
TE, and 8 mL of
50% PEG. Make fresh.
3. Chromosome Prep Buffer:
(a) 8.68 mL of H2O–Triton X-100 solution (8.48 mL of
H2O and 200 μL Triton X-100).
(b) 1 mL of 10% SDS.
(c) 200 μL of 5 M NaCl.
(d) 100 μL of 1 M Tris pH 8.0.
(e) 20 μL of 0.5 M EDTA.

2.2 Polymerase 1. PCR Reaction Mixture:

Chain Reaction (PCR) (a) 25 mM MgCl2 (to 2 mM final concentration in reaction).
(b) 10
Taq Buffer (200 mM Tris pH 8.8, 100 mM
(NH4)2SO4, 100 mM KCl, and 1% Triton X-100).
(c) dNTP mix (2.5 mM each).
(d) Taq Polymerase (see Note 1).
(e) Primer SO295: 50 AAACTCGGTTTGACGCCT
CCCATG.
(f) Primer SO296: 50 AGCAACAGGACTAGGATGA
GTAGC.

2.3 Expansion 1. Synthetic Complete (SC) Plates (1 L): 7.0 g Yeast Nitrogen
Assays and Time Base without amino acids (þAmmonium Sulfate) (Difco),
Courses 0.87 g Amino Acid drop-out mix (histidine), 20.0 g agar
(US Biological), 950 mL sterile dH2O.
(a) Add stir bar and all measured contents into a 2 L flask,
capping the top with aluminum foil.
(b) Autoclave at 121  C for 40 min.
Measuring Dynamic Behavior of Trinucleotide Repeat Tracts. . . 443

(c) Remove from autoclave and add 50.0 mL of sterile 40%

glucose to flask.
(d) Allow flask to cool before pouring into plates. Store at
room temperature or at 4  C.
2. Liquid Synthetic Complete Media (1 L): 7.0 g Yeast Nitrogen
Base without amino acids (þAmmonium Sulfate) (Difco),
0.87 g of Histidine Amino Acid drop-out mix, 950 mL sterile
dH2O.
(a) Add all contents into a 1 L screwcap bottle.
(b) Autoclave at 121  C for 40 min.
(c) Remove from autoclave, add 50.0 mL of sterile 40% glu-
cose to bottle and allow media to cool to room tempera-
ture before use.
(d) Liquid media can be stored for long periods of time
without addition of 40% glucose.
3. Synthetic Complete (SC) Plates containing 5-FOA (1 L):
(a) In a 2 L flask, add a stir bar, 700 mL of deionized water,
20.0 g of Agar (US Biological), and 20.0 g of Glucose (US
Biological). Autoclave at 121  C for 40 min.
(b) In a separate beaker containing a stir bar and 200 mL of
deionized water, add 7.0 g Yeast Nitrogen Base without
amino acids (þammonium sulfate) (Difco), 0.87 g of
Histidine Uracil Amino Acid drop-out mix, 50 mg of
uracil, and 1.0 g 5-FOA (0.1%).
(c) Stir contents of beaker on a stir plate until dissolved
(~1–2 h). Bring the total volume up to 300 mL using
dH2O.
(d) Filter-sterilize the 5-FOA solution using a 0.2 μm filter.
(e) Once the flask of agar and glucose cools to approximately
80  C, add the 5-FOA solution to the flask, allowing it to
completely mix on the stir plate. Immediately pour into
plates.
(f) Allow the plates to cool, then store at 4  C. Note that 5-
FOA is light sensitive.
4. 12% Polyacrylamide, 0.5
TBE Gels
(a) Acrylamide–bis solution (29:1).
(b) 10
TBE.
(c) 10% ammonium persulfate (APS).
(d) TEMED.
(e) Horizontal gel apparatus, 16 cm
14 cm glass plates, thin
spacers, and a 20-well comb.
(f) 10
polyacrylamide loading dye (0.25% bromophenol
blue, 0.25% xylene cyanol, and 30% glycerol).
444 Gregory M. Williams and Jennifer A. Surtees

(g) Low molecular weight DNA Marker (e.g., NEBiolabs).

(h) Plastic staining tray.
(i) Ethidium bromide.

2.4 Southern Blotting 1. Denaturation buffer: 0.5 N NaOH and 1.5 M NaCl.
2. Neutralization buffer: 0.5 M Tris–HCl pH 7.0, 3.0 M NaCl.
Start with ~700 mL dH2O, bring Tris and NaCl into solution.
Adjust pH to 7.0 with concentrated HCl and then bring final
volume (1 L) with dH2O.
3. 10
SSC: dissolve 87.7 g of NaCl and 44.1 g of sodium citrate
in ~800 mL dH2O. Adjust pH to 7.0 with a few drops of 10 N
NaOH and adjust to 1 L dH2O. This concentrated stock of
NaOH is required to quickly adjust the pH.
4. 1.85 MBq (50 μCi) [α-32P] dCTP (~111 TBq/mmol,
370 MQq/mL).
5. Random Priming DNA labeling kit (e.g., TaKaRa).
6. Stratalinker or equivalent UV crosslinker.
7. HYBAID hybridization oven, or similar.
8. Hybridization bottles.
9. STORM PhosphorImager, or equivalent.
10. Storage phosphor screen.

3 Methods

3.1 Construction of 1. A URA3 reporter system construct is used to select for expan-
Strains Containing TNR sion events in vivo [35–37]. TNR substrates are integrated into
Tracts each strain of interest, using plasmids that contain the reporter
construct, tract sequences and sequences to integrate the sys-
tem at the LYS2 locus [6, 17, 37]. Plasmids encoding (CAG)25,
(CTG)25, and scrambled (C,A,G)25 or (C,T,G)25 are used
[36–38]. Each repeat tract is cloned into the regulatory region
controlling expression of the URA3 reporter gene. When the
distance between the TATA box and the initiator ATG for the
URA3 gene is increased beyond 29 repeats, URA3 is no longer
expressed, making the cells resistant to 5-FOA (see Note 2)
(Fig. 2).
2. Each plasmid is digested with Bsu36I to linearize and is then
transformed, by the lithium acetate method [39], to allow
integration at the LYS2 locus by homologous recombination.
3. For each transformation, set up a small culture in liquid media
from a single colony. Incubate overnight in a 30  C shaker.
4. From the saturated culture, dilute 1:30 to 1:50 to start a new
20–30 mL liquid culture in a 125 mL sterile flask. Incubate in a
 Measuring Dynamic Behavior of Trinucleotide Repeat Tracts. . . 445

Fig. 2 TNR reporter system used to select expansions. The URA3 gene is used as
a reporter for expansions on the lagging strand [36]. The Schizosaccharomyces
pombe adh1 promoter is fused upstream of the URA3 reporter gene. This
promoter has a stringent requirement for the distance between TATA and ATG
(55–125 bp) for function. Insertion of a trinucleotide repeat (TNR) makes URA3
expression dependent on repeat tract length. Cells with an unexpanded TNR tract
will express URA3 and exhibit sensitivity to the drug 5-FOA, which is toxic to the
cell in the presence of the Ura3 gene product. Expansions within the TNR tract of 4
or more repeats alters the transcriptional start site, effectively inactivating URA3
expression and resulting in 5-FOA resistance. The TNR construct is integrated into
the chromosomal LYS2 locus by homologous recombination with HIS3 as a
selectable marker. Single integration is confirmed by Southern blotting

30  C shaker until the culture reaches mid-logarithmic growth

(OD600 ~ 0.5–0.8).
5. Collect the cells by centrifugation (3000
g). Wash the pellet
in 1 mL sterile dH2O. Transfer to a microcentrifuge tube.
6. Collect the cells by centrifugation (16,000
g) and discard the
supernatant. Wash the cells with 1 mL LiAc–TE mix. Collect
the by centrifugation again and resuspend the pellet in 250 μL
LiAc–TE mix.
7. For each transformation, mix 50 μL of the yeast suspension,
~150 ng digested plasmid and 10 μL sonicated salmon sperm
DNA (10 mg/mL). Always do a negative control with no
plasmid DNA added.
8. After mixing, add 600 μL 40% PEG mix.
9. Incubate reactions in a 30  C shaker for at least 30 min. Longer
incubation times are fine.
10. Incubate the reactions in a 42  C water bath for 15 min to heat
shock.
11. Briefly centrifuge the reaction (30 s), discarding the superna-
tant. Resuspend the cells in 200 mL 1
TE, pH 7.5. Select for
transformants by plating on SC histidine. Incubate in 30  C
cabinet for ~3 days.
12. Individual transformants were struck out to single colonies on
SC histidine plates. A single colony from each candidate was
446 Gregory M. Williams and Jennifer A. Surtees

frozen down for storage in 20% glycerol until ready to test for
single integration of the reporter construct.
13. Perform Southern blotting (see below) to ensure that a single
copy of the reporter construct has been integrated [37]. This
will also confirm that the integration has occurred at the cor-
rect chromosomal locus.

3.2 TNR Expansion 1. Using a sterile wooden applicator, take a small amount of cells
Assay from a frozen glycerol (20%) stock of the strain that is to be
tested. Make a small patch at the top of a SC histidine plate.
Using a new wooden applicator, start toward the bottom of the
patch and make a small vertical line toward the bottom of the
plate, streaking a small amount of cells away from the patch.
With a new wooden applicator, cross this vertical line once and
zig-zag the applicator back and forth across the plate, separat-
ing your patch of cells out to single cells on the plate. Place
plate into the 30  C incubator for 3–4 days allowing the single
cells to grow into ~2 mm colonies.
2. Select three individual colonies and mark them on the back of
the plate, labeling them 1, 2, 3.
3. Perform colony PCR on these colonies to confirm that the
repeat tract has not expanded. In this system, the starting
repeat tract size is 25 repeats (75 bp). Perform PCR as follows:
(a) Combine, per reaction: 12.3 μL sterile dH2O, 1.6 μL
25 mM MgCl2, 2 μL 10
Taq Buffer, 2 μL dNTP mix
(2.5 mM each), 0.8 μL of 5 pmol/μL SO295, 0.8 μL of
5 pmol/μL SO296, and 0.5 μL Taq Polymerase (see Note 1).
(b) Using a sterile 200 μL pipette tip, take a very small sample
of the colony by barely touching the tip to the colony to
obtain a small number of cells on the tip. Resuspend the
cells in 20 μL of your PCR mix, making sure to put the tip
directly into the PCR mix and twirl. Make sure that all
liquid that makes its way into the pipette tip through
capillary action is cleared back into the PCR tube, by
pushing down on the top of the tip with your finger.
(c) 5 min incubation at 95  C; 35 cycles of 2 min at 95  C;
1 min at 53  C, 1 min at 72  C. End with a 10-min
incubation at 72  C.
4. Digest the PCR reactions using SphI and AflII. Make a master
digestion mix to be added directly to the 20 μL PCR reaction,
as follows. The 10 μL mix for 1 reaction, i.e., the amount to be
added to 1 PCR reaction is: 0.2 μL SphI (NEB), 0.2 μL AflII
(NEB), 3.0 μL NEB Buffer 2.1, and 6.4 μL sterile dH2O.
(a) Add 10 μL of digestion mix to each 20 μL PCR reaction.
(b) Incubate digestion reaction for 1 h and 30 min in a 37  C
water bath.
 Measuring Dynamic Behavior of Trinucleotide Repeat Tracts. . . 447

A.

SphI SphI
SO295 AflII SO296

(CNG)25
B.

150bp →

100bp →
75bp →

50bp →

Fig. 3 Digestion of TNR tract PCR. (a) The TNR tract length can be determined by
PCR, using primers upstream and downstream of the repeat sequence (SO295
and SO296), resulting in a PCR fragment of 188 bp. The amplified regions of DNA
immediately upstream and downstream of the repeat tract can be digested with
SphI, freeing the TNR tract. (b) Lane 2 shows that SphI digestion results in three
distinct bands, with a doublet at 75 bp (the 25 repeat TNR tract) and 73 bp and a
41 bp fragment. To eliminate the doublet for easier assessment of the TNR tract
size, the 73 bp fragment is further digested with AflII, resulting in a 59 bp
fragment, as seen in Lane 3. A 14 bp fragment is not visible on this gel. Lane 1
shows the low molecular weight marker (LMWM, NEBiolabs)

(c) Store digested reactions at 20  C, if necessary.

5. Run the digested reaction on a 12% polyacrylamide, 0.5
TBE
gel to confirm the tract size (Fig. 3):
(a) Make and pour gels using a full piece of glass
(16 cm
14 cm), notched piece of glass, thin spacers,
and a 20-well comb.
(b) Clean the inside of the glass plates with ethanol and a
Kimwipes to ensure that they are free of any dust or debris.
Insert spacers between the two cleaned pieces of glass and
clamp the outside using six binder clips; two clips per side.
Insert the 20-well comb. The outside edges of the glass
can be sealed with 0.5% agarose in dH2O, making sure to
add extra agarose to the corners.
(c) Make the gel solution (30 mL, 2 gels): 16.16 mL of sterile
dH2O, 12 mL of acrylamide–bis solution (29:1), 1.5 mL
of 10
TBE, 0.3 mL of 10% ammonium persulfate (APS),
and 32 μL of TEMED. Make sure to add the components
in the order that they are listed, with the APS and the
TEMED added last. The solution will begin to solidify
448 Gregory M. Williams and Jennifer A. Surtees

within 7–10 min. The gel solution can be added into the
sealed glass plates using a 5 mL serological pipette, placing
the tip of the pipette in the corner of the clamped plates.
(d) After gels solidify, remove the bottom spacer and sub-
merge into 0.5
TBE in the gel apparatus, making sure
that there are no bubbles between the glass plates below
the gel. Assemble the gel apparatus and fill with 0.5

TBE.
(e) Using a gel loading tip, rinse each well to make sure that
there is nothing in it.
(f) Add 3 μL of 10
polyacrylamide loading dye (0.25%
bromophenol blue, 0.25% xylene cyanol, and 30% glyc-
erol) to each sample. Mix well.
(g) Load 10 μL of your digested sample into each well. You
should have ~20 μL remaining, which can be stored at
20  C.
(h) Make sure to load a plasmid control (last lane) as well as
the Low Molecular Weight Marker (10 μL of loading dye,
1–2 μL of marker).
(i) Run gels for 2 h at 250 V.
(j) Stain the gel for 10 min in 250 mL of deionized water
containing 0.5 μg/ml ethidium bromide.
6. After identifying an unexpanded tract from a colony, remove
this colony from the plate using a sterile flat end toothpick, and
resuspend it in 100 μL of sterile dH2O in a microcentrifuge
tube. Vortex the tube to resuspend the colony, and serial dilute
the resuspension to 104 (10 μL of undiluted resuspension into
990 μL of sterile deionized water, vortex, 10 μL of 102
dilution into 990 μL of sterile deionized water to get 1 mL of
104 dilution). Make sure to vortex in between dilutions.
7. Plate 20 μL of the 104 dilution onto a SC histidine plates
and incubate at 30  C for 3–4 days, allowing the colonies to
grow to ~2 mm in diameter.
8. Select ten colonies to test in the TNR expansion assay. Using a
ruler, measure the diameter of the colonies on the plate, select-
ing colonies that are ~2 mm in size. It is important to make sure
that all of your selected colonies are approximately the same
size; colonies of similar size will have undergone roughly the
same number of cell divisions.
9. Remove 2 mm colonies from the plate using a sterile flat ended
toothpick and resuspend in 100 μL of sterile deionized water
and resuspend by twisting the toothpick in the microcentrifuge
tube. After the colony is resuspended in water, discard the
toothpick, close the lid on the microcentrifuge tube, label,
 Measuring Dynamic Behavior of Trinucleotide Repeat Tracts. . . 449

and vortex to further resuspend the colony. Dilute the resus-

pension to 104 (see step 3), making sure to save each of your
serial dilutions. Each colony that has been resuspended will be
plated on both selective and permissive plates to determine an
accurate expansion frequency.
10. Plate 20 μL of the 104 dilution onto SC histidine and spread
using sterile glass beads, making sure that the resuspension is
evenly spread on each plate (see Note 3). These are the permis-
sive plates; all cells will grow. These plates are important to
determine the total number of cells tested.
11. Plate 75 μL of the undiluted resuspension onto SC histidine
plates containing 5-FOA. These are the selective plates, only
allowing growth of cells that lack URA3 expression. In this
system, this typically means that the strain contains an
expanded TNR tract greater than or equal to 29 repeats (see
Note 4). Spread the resuspension on the plate using sterile
glass beads.
12. Place both the selective and permissive plates into a 30  C
incubator and allow them to grow for 3–4 days until the
majority of colonies on the plates reach a diameter of
~1–2 mm so that they are easy to count. It may be necessary
to allow the selective plates to grow a day longer to allow for
sufficient growth in the presence of 5-FOA.
13. Remove plates from the incubator and count the number of
colonies on all plates. After counts have been obtained for both
selective and permissive plates, the numbers for each colony
will be subjected to analysis in order to determine the expan-
sion frequency of the strain.

3.3 Mutation Rate 1. To calculate apparent expansion rates in the TNR expansion
Calculations assay, colonies on selective and permissive plates are counted.
Expansion rates are calculated by iterative formulas, as described
by Drake [40]. The 95% confidence intervals are determined
from tables of confidence intervals for the median [41, 42].
2. To calculate the mutation rates for the time course experiments
(see Subheadings 3.4–3.6), the probability of a change in tract
size was treated as a binomial distribution with p ¼ the propor-
tion of tracts with a change in length and q ¼ the proportion of
tracts with no length change.
3. The rate of tract change was defined as the number of changes
observed/the number of cells examined per generation. It is
therefore important to determine the doubling time (or genera-
tion time) of each strain prior to calculating mutation rates.
Growth curve experiments in both liquid media and on plates
are used to determine the number of generations for these calcula-
tions (see Subheading 3.9). To calculate 95% confidence intervals
(95% C.I.) on the mutation rates, the F statistic are used [43, 44].
450 Gregory M. Williams and Jennifer A. Surtees

3.4 Liquid Time 1. Perform a liquid time course to assess tract length and tract
Course Protocol length changes in a large, heterogeneous cell population. By
performing assays with replicative and stationary phase cul-
tures, it is possible to determine the contribution of DNA
replication to tract instability in this system (Fig. 4).
2. Perform TNR expansion assay (see Subheading 3.2) to select
for colonies with expanded repeat tracts.

A. B.
5-FOA

Stationary Phase Logarithmic Phase

SC-His

C1. C2.

D.
SC-His
dIH2O

24 HRS 24 HRS
gDNA
E1. E2.

Continue SC-His Continue

through dIH2O F. through
duration of duration of
time course time course

Fig. 4 Schematic of the liquid time course experiment protocols. (a) Individual colonies with TNR tract
expansions are selected on plates containing 5-FOA; the tract length increase was confirmed by colony PCR.
(b) A single colony with an expanded tract is used to inoculate a 106 mL large culture that was grown to
saturation over a 72-h period (18 generations) at 30  C. Five microliter of this starting culture is removed and
genomic DNA (gDNA) was isolated to obtain time point zero. Then, parallel stationary (c1) and logarithmic (c2)
phase cultures are established and propagated for 14 days by taking 200 μL of the 24-h saturated culture and
adding it to a new flask containing 10 mL of SC His media (e1, e2). (d) gDNA was prepared from each culture
every 24 h (6 generations) and subjected to PCR to evaluate tract length. (f) A sample from each log- and
stationary-phase culture is diluted 104 and plated to isolate individual colonies that are subjected to colony
PCR to assess tract length. This is an end point experiment that allows for the observation of additional
expansions as a function of time as well as general trends in the cell population. Colony PCR on colonies
plated from specific time points permits the analysis of individual tract lengths from single colonies. Fitness
effects of tract lengths are mitigated by limiting the number of cell divisions (6 per time point), although we
previously determined that the tract length does not affect growth rate with our strains [17]. This experiment is
similar to that of a mutation-accumulation experiment and indicates that additional expansions are observable
in a 2-week time span
 Measuring Dynamic Behavior of Trinucleotide Repeat Tracts. . . 451

3. To confirm the presence of an expansion, perform colony PCR

on colonies from selective plates (see Subheading 3.2, step 3).
4. Digest the DNA with SphI and AflII (see Subheading 3.2, step
4) and run samples on a 12% polyacrylamide, 0.5
TBE gel (see
Subheading 3.2, step 5). This will also allow you to determine
the initial size of the expanded tract (see Note 5).
5. Resuspend confirmed expanded colonies in 106 mL of liquid
SC histidine (see Note 6) and grow the culture to saturation
in a 30  C shaker for 3 days (72 h) (Fig. 4).
6. From this saturated culture, remove 5 mL to isolate genomic
DNA (see Subheading 3.5) that will be used as a template for
PCR to determine the average tract size in the culture. This is
the Day 0 time point.
7. Then take 200 μL and add it to a fresh flask containing 10 mL of
liquid SC histidine (1:50 dilution) to propagate the next time
point. Be sure to use sterile technique when sampling the culture
or transferring culture to new flasks. Place this flask in the 30  C
shaker for 24 h. This is your logarithmic phase, or replicative
culture. The cells will grow and divide in this culture (Fig. 5a).

Fig. 5 Liquid time course experiments of wild-type (CTG)25 logarithmic-phase (a) and stationary-phase (b)
cultures to determine population tract dynamics, as described in Fig. 4. The initial expansion size is indicated
in yellow at time point zero. The numbers across the top of the gel indicate the day of the time course. The
brackets at the top of the gel (a, left panel) indicate the progressive accumulation of a larger TNR tract with the
concomitant loss of the initial expansion size toward the end of the time course. Stationary-phase repeat
tracts (b, left panel) remain stable. A sample from Day 14 log-phase (b, right panel) and stationary-phase (b,
right panel) culture were diluted and PCR was performed on individual colonies to assess individual tract
lengths within the population. The length of the tracts was determined using a standard curve measured from
the known bands of the low molecular weight marker (LMWM) run with each gel (see Note 5). The number
above or below each tract indicates the number of repeats within each tract. C, PCR on a TNR plasmid control
is performed alongside each time course as a marker for the unexpanded 75 bp tract (25 repeats). Expansions
are observed in the actively replicating log-phase cultures (a), while stationary-phase tracts remained stable
through the duration of the time course (b)
452 Gregory M. Williams and Jennifer A. Surtees

8. Using sterile technique, pour the remainder of the saturated

culture into a 50 mL conical tube. Collect cells by centrifuga-
tion at 3000
g for 5 min. Remove the supernatant, making
sure to keep the pellet formed at the bottom of the tube. Add
10 mL of sterile dH2O to the tube, pipetting the pellet back
into solution. Collect the resuspended cells by centrifugation
(3000
g, 5 min.) and remove the supernatant. Repeat this
wash step with 10 mL of sterile dH2O, again dumping out the
supernatant. Resuspend the pellet in 10 mL of sterile dH2O
and add to 90 mL of sterile dH2O in a 500 mL sterile flask. This
will serve as your stationary phase culture. The cells will survive
but will exhibit very little growth or division without nutrients.
Incubate this stationary phase culture at 30  C for 24 h
(Fig. 5b).
9. Every 24 h, remove both logarithmic phase and stationary
phase cultures from the 30  C shaker. Remove 5 mL from
each culture to extract genomic DNA, which will be used as
template for PCR to determine TNR tract size. Remove 10 μL
from each culture and add it to a microcentrifuge tube contain-
ing 990 μL of sterile deionized water, resulting in a 102
dilution of your culture. Vortex the dilution and serial dilute
it down to 104 by removing 10 μL of the 102 dilution and
adding it to another microcentrifuge tube containing 990 μL
of sterile deionized water. Vortex, and plate 20 μL of the 104
dilution directly on synthetic media plates lacking histidine and
spread using the glass bead method. Incubate plates at 30  C
for 3 days. To determine the range of expansions within the
population, as opposed to the average, colony PCR can be
performed on colonies to determine tract sizes from individual
colonies from the liquid culture. This can be done at each time
point throughout the time course. We typically plate cells on
Days 1, 7, and 14 to monitor the health of the culture as well as
tract size (Fig. 5).
10. Take 200 μL of your newly saturated logarithmic phase cul-
tures and add them to a flask containing 10 mL of fresh SC
histidine and incubate in a 30  C shaker for 24 h.
11. The cells in the stationary phase cultures should be washed and
resuspended in fresh sterile dH2O every 48 h. Using sterile
technique, collect the cells by centrifugation at 3000
g for
5 min. Resuspend each pellet in fresh sterile dH2O, making
sure to keep the total volume accurate: (100 mL) 
(days
5 mL) ¼ current volume. Changing the water in
these cultures helps prevent cell death.
12. Repeat steps 7–9 throughout the entirety of the time course.
13. Throughout the time course, use the isolated genomic DNA
(gDNA) as a template for PCR to determine the average tract
size in the culture (Fig. 5).
 Measuring Dynamic Behavior of Trinucleotide Repeat Tracts. . . 453

3.5 Genomic DNA 1. Collect cells by centrifugation a 3000
g and discard superna-
Preparation tant. You should have a pellet at the bottom of the tube.
2. Add 500 μL of sterile dH2O to each tube and vortex to
resuspend the pellet. Transfer resuspension to a microcentri-
fuge tube.
3. Centrifuge for 30 s at 16,000
g.
4. Remove tubes from microfuge and pour out the supernatant.
Vortex the pellet in the residual liquid to resuspend.
5. Add 200 μL Chromosome Prep Buffer.
6. Add 200 μL of well-mixed phenol–chloroform (1 part phenol
and 1 part chloroform, vortex to mix, and spin down at
1200
g for 5 min. Store at 4  C, cover tube with aluminum
foil, it is light sensitive).
7. Add 0.3 g of acid-washed glass beads (see Note 7).
8. Vortex tubes at 400
g in the Mixmate (Eppendorf) in cold
room for 3 min.
9. Add 200 μL of 1
TE, pH 8.0 to each vortexed tube.
10. Centrifuge for 5 min at 16,000
g.
11. While waiting for your samples to spin down, set up clean
microcentrifuge tubes and add 10 μL of 4 M ammonium
acetate to each tube.
12. When your samples having finished the centrifugation, transfer
the supernatant into the ammonium acetate-containing tubes.
Add 1 mL of 95% ethanol (EtOH) to each tube and invert two
times. Your reactions are stable at this point.
13. Spin down your samples for 2 min in the microfuge at
16,000
g.
14. Pour out EtOH, making sure not to pour out the pellet at the
bottom of the tube. Add 1 mL of 70% EtOH to wash each
pellet. Gently invert your tubes a few times.
15. Briefly centrifuge your samples for 30 s at 16,000
g. Remove
residual EtOH with your pipetteman.
16. Leave the tops on your samples open and let stand for
15–30 min at room temperature for the EtOH to evaporate
and your samples to dry. Alternatively, you can use a vacuum
apparatus to dry your samples; allow 10 min to dry under
vacuum. Before moving on to the next step, make sure that
all EtOH is evaporated from your samples, as the presence of
EtOH can affect downstream applications.
17. Resuspend your dry pellet in 50 μL of 1
TE (pH 8.0) with
50 μg/mL RNase. NEVER vortex your gDNA. This gDNA
can be used as a template for PCR, using 1 μL of genomic DNA
sample, instead of a colony, in each 20 μL PCR reaction (see
Subheading 3.2, step 3).
454 Gregory M. Williams and Jennifer A. Surtees

A.

5-FOA
Transfer
B.

SC-His

24 HRS

C.
PCR

SC-His

24 HRS

D.
PCR

SC-His

Continue through duration of time course

Fig. 6 Schematic of the colony time course experiment protocol. (a) Individual
colonies with TNR tract expansions are selected on plates containing 5-FOA; the
tract increase is confirmed by colony PCR in each case. (b) A single colony with
an expanded tract is transferred to nonselective medium (SC His) and allowed
to continue to grow. Plates are then places back into the 30  C incubator. (c)
Each colony is subjected to colony PCR every 24 h (~10 generations), sampling
from around the perimeter of the colony, where the cells continued to grow.
Plates are returned to the 30  C incubator, and colony PCR performed again after
24 h (d). This approach allows examination of the dynamics of a single TNR tract
as a function of time. Because the PCR is derived from a sub-sample of an
actively dividing colony, this is necessarily examining a mixed population, as
evidenced by the multiple bands observed in some lanes, although the colony is
derived from the same original cell

3.6 Colony Time 1. Perform a colony time course experiment to assess tract
Course Protocol dynamics in a single colony, derived from a single cell, over
time (Fig. 6).
2. Perform TNR expansion assay to select for colonies with
expanded repeat tracts.
3. Perform colony PCR on colonies from selective plates (see
Subheading 3.2, step 3), digesting with SphI and AflII (see
Subheading 3.2, step 4), and run samples on a 12% polyacryl-
amide gel to determine the initial size of the expanded tract (see
Subheading 3.2, step 5).
 Measuring Dynamic Behavior of Trinucleotide Repeat Tracts. . . 455

4. After confirming the size of the initial TNR tract expansion,

select colonies for the time course experiment. Different sized
expansions can be selected to test, i.e., threshold length tracts
or affected range tracts. Mark these colonies on the back of the
plate.
5. Using a sterile flat end toothpick, swipe the entire colony off
from selective plate containing 5-FOA and deposit it onto a
new SC histidine plate and label the colony appropriately. It is
important to remove confirmed colonies from the selective
plate to allow for dynamic expansion and contraction events.
Maintaining the colonies on plates containing 5-FOA will pre-
vent growth of cells that sustain contractions that result in
recovery of URA3 expression, thereby potentially creating
bias toward expansions.
6. After transferring colonies to new plates, perform colony PCR
on each of the colonies to obtain the “time point zero” for the
time course experiment (see Subheading 3.2, steps 3 and 4).
Place the plates back into the 30  C incubator for 24 h.
7. After 24 h, remove the plates and perform colony PCR on each
individual colony (see Subheading 3.2, steps 3 and 4). When
taking a small amount of the colony with a sterile tip for colony
PCR, you may sample from either the same portion of the
colony (i.e. 12 o’clock), or different portions of the colony
(i.e., 12, 3, 6, and 9 o’clock) at different time points. Because
the cells are always sampled from the perimeter of the colony,
the cells from the colony continue to grow and fill in the
sampled space. The growth rate on solid media can be deter-
mined (see Subheading 3.12) to determine the generation time
of each strain on these plates (see Note 8).
8. Return plates to the 30  C incubator for 24 h.
9. Repeat steps 6 and 7 for 14 days. If necessary, transfer colonies
to new plates by swiping the entire colony off of the old plate
and depositing it onto a fresh synthetic medium plate lacking
histidine.
10. Colony PCR reactions for each time point should be digested
with SphI and AflII to assess tract length over time (Fig. 7).

3.7 Microcolony 1. Perform a microcolony time course to examine tract dynamics

Time Course Protocol starting from a single cell, i.e., a completely homogeneous
starting point (Fig. 8).
2. Perform TNR expansion assay to select for colonies with
expanded repeat tracts.
3. Perform colony PCR on colonies from selective plates, digest-
ing with SphI and AflII (see Subheading 3.2, steps 3 and 4),
456 Gregory M. Williams and Jennifer A. Surtees

Fig. 7 Progressive expansion events wild-type colony time course. Expansion events were selected in the
wild-type (CTG)25 background and confirmed by PCR. Individual colonies were followed over a 14-day time
period; colony PCR was performed on each colony every 24 h to amplify the TNR tract (see Fig. 6). At the end
of the time course, each PCR reaction was digested and resolved on a 12% polyacrylamide gel and stained
with EtBr. The numbers across the top of the gel indicate the day of the time course. The lanes marked C in
each panel indicate the 75-bp tract amplified from the TNR plasmid control. An uncropped gel is shown in this
figure to highlight the different non-tract digestion products that appear below the TNR tract. The yellow arrow
indicates the initial expansion product

and run samples on a 12% polyacrylamide gel to determine the

initial size of the expanded tract (see Subheading 3.2, steps 5).
4. After confirming the size of the initial TNR tract expansion,
select a 2 mm colony from the selective plate and resuspend it
in 200 μL of sterile dH2O (Fig. 8).
5. Plate 50 μL of this resuspension onto the top left side of a fresh
SC histidine. Tilt the plate so that the resuspension drips
down toward the bottom of the plate in a straight line. Tilt
the plate back and forth so that the cells are evenly distributed
along the streak. Allow this streak to dry for approximately
5–7 min (Fig. 8b, 9A).
6. After streak has nearly dried, draw three small boxes vertically
to the right of the streak of cells and label them 1, 2, and 3. This
is where you will place your single cells for the time course
(Fig. 8b, 9C).
7. Using a tetrad dissection microscope, i.e., a microscope with a
micromanipulator (see Note 9), focus on individual cells under
30
magnification (Fig. 9B). Switch to 150
magnification
and identify single cells within the streak of cells. Choose
unbudded cells so that you are starting with a homogeneous/
single tract.
 Measuring Dynamic Behavior of Trinucleotide Repeat Tracts. . . 457

A.

5-FOA

B. C. E.
PCR
~8-10 generations

D.

SC-His SC-His

SC-His
-
F.

SC-His
Continue through
duration of time
course

Fig. 8 Schematic of the microcolony time course experiment. (a) Individual colonies with TNR tract expansions
were selected on plates containing 5-FOA; the tract increase was confirmed by colony PCR in each case. (b)
Each confirmed expanded colony is then resuspended in 200 μL of sterile deionized water and serial diluted
down to 102. Fifty microliters of this dilution is spotted and smeared onto a nonselective media plate (SC
His). Single cells from the plated smear are then isolated using a micromanipulator and placed into boxes
drawn onto the back of the plate. (c) Plates are then placed into a 30  C incubator and allowed to grow for
approximately 15 h, resulting in a microcolony that has undergone ~8–10 generations (~250–1000 cells/
microcolony). (d) Using the micromanipulator, a single cell is removed from the microcolony to propagate the
next time point, while the remainder of the microcolony is swiped off the plated and used in colony PCR to
determine tract length (e). Plates are then returned to the 30  C incubator for ~15 h to allow the new
microcolony to grow (f), continuing the aforementioned steps through the duration of the time course

8. Carefully swipe the single cell off the plate with the microma-
nipulator and place it into one of the boxes you drew vertically
next to the streak of cells. Place a single cell in each of the three
boxes you have drawn. Make sure not to puncture the plate
with the cell and glass rod, this will deposit the cell into the agar
where it will not grow well (Fig. 9B).
458 Gregory M. Williams and Jennifer A. Surtees

Fig. 9 (A) After confirming an expanded TNR tract through colony PCR, the colony is swiped off the plate,
resuspended in sterile deionized water and serial dilute down to 102. Fifty microliter of the 102 dilution is
then plated onto SC His and the plate tilted back and forth to smear the cells to singles and the plate is
brought to the micromanipulator. (Ba) The plate is placed onto the micromanipulator microscope and the
smear of cells is observed at 30
magnification. Once a patch of single cells is identified, the magnification of
the microscope is increased to 150
(Bb), and single, nonbudding cells are removed from the plate using the
glass rod of the micromanipulator (Bc) and placed into the corners of a box drawn on the back of each plate
(Bd). Four individual cells can be placed in each box (Be). (Ca) Plates are then placed into a 30  C incubator for
~15 h, resulting in a microcolony with 250–1000 cells in each corner of the box. The micromanipulator is used
to smear the microcolony under 150
magnification (Cb), allowing for an accurate count of cells within each
microcolony (Cc). A single cell from the microcolony is selected and placed in a newly drawn box to propagate
the next time point (Cd). The perimeter of the remaining microcolony is marked by puncturing the media with
the micromanipulator glass rod, allowing for easy identification away from the microscope (Ce). (D) The
remainder of each microcolony is swiped off the plate using a 200 μL pipette tip and resuspended in PCR
master mix to assess tract length changes

9. After filling all three boxes with single cells, incubate each plate
at 30  C, allowing cells to undergo 8–10 rounds of replication,
resulting in a microcolony of approximately 250–1000 cells.
Determination of the generation time for each strain (see
below) will reveal how long the cells should be incubated. It
 Measuring Dynamic Behavior of Trinucleotide Repeat Tracts. . . 459

is a good idea to check the plates periodically throughout the

day to count cells in the microcolony and confirm the growth
rate. Typically, 15–20 h are required for 8–10 generations in
our strain background on synthetic media [17].
10. After 8–10 generations, draw three new boxes in a different
color marker adjacent to your initial boxes. This new box is
where you will transfer four individual cells to propagate your
next microcolony.
11. Using the micromanipulator, carefully smear the microcolony
with the needle to obtain a better estimate of the cell count.
Then select four individual colonies for your next time point
and place one individual cell in each of the four corners of the
next box. Four cells are necessary to assure you have at least one
microcolony that grows to test for your next time point
(Fig. 9C).
12. After four cells have been placed in the new box, return to the
original microcolony and puncture the agar around the
smeared microcolony. These holes surrounding the microcol-
ony will help identify the location of the microcolony on the
plate and should be visible without use of the microscope. This
will be useful removing the microcolony from the plate for
PCR (Fig. 9C, D).
13. Repeat steps 9–11 for the rest of your time courses, making
sure to mark each microcolony that will be swiped off the plate
and used for template in PCR.
14. After single cells have been transferred to generate the next
microcolony, use a sterile 200 μL pipette tip to harvest each
microcolony from the plate and resuspend it in 20 μL of the
colony PCR mix (see Subheading 3.2, step 3). Make sure that
all of the PCR mix in the tip is pushed back out into the PCR
mix. You can do this by pushing down on the top of the tip
with your finger.
15. Make sure that your tubes are well labeled.
16. Using the microscope, check that each microcolony was har-
vested. There will be a small number of cells that did not make
it onto the tip; this is ok. If there appears to be a substantial
number of cells still present on the plate, mark the area sur-
rounding the microcolony again and reswipe the plate using a
sterile 200 μL pipette tip (see Note 10).
17. Place plates back into the 30  C incubator for another 8–10
generations (Fig. 8f).
18. After incubation, once again draw boxes horizontally on the
plate, making sure to carefully label the time point. Look at
each of the propagated microcolonies and observe their size.
Out of the four microcolonies that you have grown, chose the
460 Gregory M. Williams and Jennifer A. Surtees

Fig. 10 Dynamic changes in TNR tracts starting from a single cell. Expansion events are selected in the wild-
type (CTG)25 background and confirmed by PCR. Individual cells from these colonies are isolated and allowed
to undergo 8–10 rounds of replication, resulting in a microcolony approximately 250–1000 cells in size (see
Fig. 8). A single cell from the microcolony is then taken to propagate another microcolony for the time course.
The remainder was swiped off the plate and used in a PCR reaction to determine tract length. PCR reactions
are digested and resolved on a 12% polyacrylamide gel and stained with EtBr; the images have been inverted
for ease of viewing. The lanes marked C in each panel indicate the 75-bp tract amplified from the TNR plasmid
control. The numbers across the top of the gels indicate the time point. In this microcolony time course, the
initial expansion is shown at time point zero, and dynamic changes in the length of the TNR tract are observed
as a function of time, with the final tract length larger than the starting tract length. Multiple bands are
observed at some time points, likely due to cellular heterogeneity

single microcolony that fits the criteria the best (250–1000

cells in size).
19. Repeat steps 10–15 for each microcolony for the duration of
the time course (Fig. 8b–f).
20. Microcolony PCR reactions for each time point should be
digested with SphI and AflII (see Subheading 3.2, steps 3
and 4) to assess dynamic tract behavior (Figs. 8e, 10, 11).

3.8 Southern Southern blotting is used to confirm single integration of the TNR
Blotting—Making the reporter construct at the correct chromosomal location. It is also
Probes used to confirm that the PCR products obtained following PCR in
each of the different time course experiments actually contain a
TNR tract.
Two different probes are used.
1. Southern blotting is used to confirm single integration of the
tract (a) and to determine tract size following an assay (b).
(a) For a probe to determine proper integration of the
reporter construct, a plasmid preparation of pBL94 [6,
37] is used. The plasmid is digested with BamHI to
release a 1.2 kb fragment that contains lys2 [37].
(b) For a tract specific probe to confirm the presence of the
TNR tract following PCR of time course samples, a plas-
mid preparation of pBL169 [17, 36] is used. PCR amplify
the TNR tract in 2 or more 50 μL PCR reactions using
primers SO295 and SO296 (see Subheading 2). Digest
 Measuring Dynamic Behavior of Trinucleotide Repeat Tracts. . . 461

Fig. 11 Southern blotting using a tract specific probe. Wild-type (CTG)25 (a) and msh3Δ (CTG)25 (b) colony PCR
time courses were performed as described. Digested PCR samples were resolved on a 12% polyacrylamide
gel, stained with EtBr, and imaged (a, b: top panels, as in Fig. 7). DNA from the gel was then transferred to a
neutral Nytran membrane, crosslinked, and subjected to Southern blot using a tract-specific TNR probe (a, b:
bottom panels).Wild-type Southern blot (a, bottom panel) shows progressive expansion of the TNR tract over
time. In contrast, the Southern blot of time course performed in the absence of the MMR factor MSH3 (msh3Δ)
(b, bottom panel) confirms that the TNR tract is stable in this background. Numbers across the top indicate
days of the time course. The lanes marked C in each panel indicate the 75-bp tract amplified from the TNR
plasmid control

each of the PCR reactions with the restriction enzyme

SphI to isolate the repeat tract.
2. Both probes are electrophoresed and purified prior to use.
(a) For the integration probe, gel purify the 1.2 kb fragment
after electrophoresis through a 1% agarose gel.
(b) For the TNR-specific probe, gel purify the 75 bp PCR
product after electrophoresis through a 12% polyacryl-
amide, 0.5
TBE gel, and staining with ethidium bro-
mide (see Subheading 3.2, step 5).
3. For both probes, place the excised bands into a clean micro-
centrifuge tube, making sure to include as little agarose or
polyacrylamide as possible. Weigh each of the excised bands
on a scale, using an empty microcentrifuge tube as your zero.
Use a gel extraction kit (e.g., Omega Bio-Tek E.Z.N.A. Gel
Extraction Kit) to remove your fragment from the gel. For the
small, tract-specific probe, pool the multiple bands into a single
column to increase the concentration of your final extracted
DNA fragment. In the final elution step of the gel extraction
protocol, use a small volume of sterile dH2O (30 μL) to
increase the overall DNA content of the extraction. Sterile
462 Gregory M. Williams and Jennifer A. Surtees

dH2O is used instead of the provided elution buffer to assure

proper labeling using a random primer DNA labeling kit.
4. Radiolabel the purified probes with the Random Primer DNA
Labeling kit. For this specific labeling kit, 150 ng of template
DNA is required, and it is designed for the use of 1.85 MBq
(50 μCi) [α-32P] dCTP (~111 TBq/mmol, 370 MQq/mL).
5. Follow the protocol provided in the Random Primer DNA
Labeling Kit. Once the tract specific probe is properly labeled,
it can be stored in a designated radioactivity approved freezer at
20  C until it is ready to be used.

3.9 Southern 1. To assess both tract integration and tract size, DNA is first
Blotting: Gel and electrophesed through a gel.
Transfer (a) For confirmation of integration, isolate gDNA and digest
with BamHI. Electrophorese through a 0.8%, 0.5
TBE
agarose gel.
(b) For time course tract confirmation, digest PCR reactions
with SphI and AflII and electrophorese the products
through a 12% polyacrylamide, 0.5
TBE gel and stain
the gel with ethidium bromide (see Subheading 3.2,
step 5). Take a picture of the gel so that you can orient
yourself later.
2. In a clean Pyrex dish, denature the gel at room temperature for
60 min in 500 mL of denaturation buffer. Make sure that the
gel is completely immersed in the buffer. The entire gel needs
to be at the high pH for the DNA to be denatured uniformly.
3. Neutralize the gel in 500 mL of neutralization buffer at room
temperature for 30 min. Remove the buffer after 30 min and
repeat the neutralization step. At the end of the two washes, the
pH of the gel should be 8 or less. This can be tested by
touching a pH strip to the side of the gel.
4. Place two weighted tip racks in the middle of a baking pan
(reservoir). Fill reservoir with 1
SSC to ½ in. deep, making
sure that the buffer does not reach the top of the tip boxes.
5. Cut two sheets of Whatman 3MM paper the width of the gel;
these will be used as the wick to bring the 1
SSC buffer
through your gel, effectively transferring the DNA to a neutral
Nytran membrane. Drape these pieces of Whatman paper over
the top of the tip boxes, hanging the ends into the 1
SSC-
filled reservoir. Wet the entire wicks with 1
SSC.
6. Upon flooding the surface of the wick with 1
SSC, remove
any trapped air bubbles by rolling a Pasteur pipette across
surface. Bubbles in the transfer setup could affect the efficient
transfer of DNA to the membrane. Before putting your gel
onto the transfer apparatus, make sure to notch one corner of
 Measuring Dynamic Behavior of Trinucleotide Repeat Tracts. . . 463

the gel to identify the orientation of the gel when transferred. A

corner should also be notched on the neutral Nytran mem-
brane. Carefully invert and lay gel on top of the wick. Remove
any air bubbles between gel and the wick. It is helpful to
continuously wet the gel with 1
SSC. Invert the gel because
the DNA is often more concentrated near the bottom.
7. Place strips of cut plastic (sheet protectors) or rolled plastic
wrap on all four sides of the gel to prevent lateral diffusion of
your DNA.
8. Cut a piece of neutral Nytran membrane to fit the size of the gel
and label the back of the membrane that is not in contact with
the gel. Wet the membrane briefly in deionized water and then
allow it to soak in 1
SSC for 5 min.
9. Flood the gel surface once again with 1
SSC and carefully lay
the neutral Nytran membrane on top, making sure to match up
the notch of the gel with the notch of the membrane, again
being careful to remove any air bubbles that may form.
10. Cut four pieces of Whatman paper to the exact size of the gel.
Soak these pieces for 5 min in 1
SSC, before laying them on
top of the gel. Once again, make sure that all bubbles are
removed by rolling a Pasteur pipette over top of the transfer
set up.
11. Place a stack of paper towels (400 tall) cut slightly larger to cover
the entirety of the gel transfer area. Place a weight on top of the
paper towels. In our transfer system, we use two large alumi-
num blocks to evenly distribute weight across the gel. Wrap the
entire reservoir and transfer set-up with plastic wrap to help
prevent evaporation of the 1
SSC buffer. Allow the transfer to
proceed overnight (~15–20 h) [37].
12. After the overnight transfer, unwrap the reservoir and remove
the gel. Image the gel to make sure that all of the DNA has
transferred from the gel to the Nytran membrane.
13. Dry the membrane between to pieces of filter paper under the
fumehood.
14. Once the membrane is sufficiently dry, wrap the membrane in
plastic wrap.
15. Crosslink the wrapped blot at 1600 μJ, using a Stratalinker or
equivalent UV crosslinker. The DNA should be stably main-
tained on the neutral Nytran membrane and can be stored at
4  C until ready to hybridize with a tract specific probe.

3.10 Southern 1. Following the transfer, remove the radiolabeled probe from the
Blotting: Hybridization radioactive materials freezer and let it thaw on ice behind a
Protocol Using α32P- Plexiglass shield.
dCTP Labeled Tract- 2. Heat the thawed probe to 95  C for 5 min and then place on ice
Specific Probe for 5 min.
464 Gregory M. Williams and Jennifer A. Surtees

3. Add 5 μL of radiolabeled probe [0.008 mCi] to 10 mL of

Hybridization Buffer in a 250 mL screwcap bottle that contains
the neutral Nytran membrane blot that is to be probed.
4. Screw the bottle cap on tight and wrap the top several times
with Parafilm to prevent any leaking.
5. Add the sealed Southern Blotting tube into a rotating
HYBAID hybridization oven at 65  C and allow it to incubate
overnight (~15–20 h).
6. The following day, remove the screwcap bottle form the
HYBAID oven. Open up the screwcap bottle and dump the
Hybridization Buffer and unbound radiolabeled probe into an
appropriate radioactive liquid waste container, making sure the
blot itself stays within the screwcap bottle.
7. Add 250 mL of Wash Buffer to the bottle to wash blot. Screw
the cap on tight and once again wrap the top of the bottle
several times with Parafilm to prevent leaking. Put the sealed
tube back into the HYBAID oven at room temperature for
20 min.
8. After 20 min, remove the bottle from the HYBAID oven and
empty the wash buffer into an approved radioactive liquid
waste container.
9. Repeat steps 6–8 for a second wash. Depending on how fresh
the probe is, it may be necessary to allow for longer washes to
improve the quality of the final Southern Bolt image. Upon
completion of the second wash, survey the Hybaid oven with
the Geiger counter to assure there is no radioactivity present in
or around the oven.
10. Remove the washed blot from the screwcap bottle and allow it
to air-dry on a piece of Whatman paper on the radioactivity
bench behind plexiglass.
11. After the probed blot has dried, wrap the blot with plastic wrap
and expose it to a Phosphorimager screen (Molecular Dynam-
ics). Make sure to note the orientation of the blot on the screen
so that it can be correctly interpreted later on. Radioactive label
that has undergone less than one half-life can be exposed for
4–8 h and will show clear resolution when imaged using a
STORM PhosphorImager (Fig. 11). After the blot is imaged
and the results obtained, the blot should be disposed in an
approved dry radioactive waste container.

3.11 Generation To calculate rates of change, and to normalize and compare indi-
Time in Liquid Culture vidual time courses, it is necessary to determine the number of
generations a culture or colony has undergone during a particular
time period. Generation times, or doubling times, can be deter-
mined in liquid media or upon growth on agar plates.
 Measuring Dynamic Behavior of Trinucleotide Repeat Tracts. . . 465

1. Streak out the TNR-containing strain of interest on SC

histidine plates to generate single colonies. SC histidine
ensures that the colonies maintain the TNR reporter system.
Place plates in the 30  C incubator and allow them to grow for
~3 days, until colonies are ~2 mm in diameter.
2. Select ~2 mm colony using a sterile wooden stick and resus-
pend the colony in 10 mL of liquid SC histidine in a 125 mL
sterile flask. Place the flask in a 30  C shaker overnight to allow
saturation of the culture.
3. After ~16 h, remove the flask from the shaker and measure the
density of the culture. Dilute the saturated culture 1:10
(100 μL of culture þ900 μL of media) in SC histidine in a
microcentrifuge tube (see Note 11). Using a spectrophotome-
ter, measure and record the optical density at 600 nm (OD600)
of the dilution. Multiply this number by 10 (to account for the
dilution) to determine the OD600 of the saturated culture.
4. Dilute the saturated culture to an OD600 of 0.1 with liquid SC
histidine in fresh 250 mL flask. The final culture volume
should be 50 mL.
5. From the newly diluted culture, remove 1 mL using a pipetman
and measure the OD600. This will be used as the time point
zero.
6. Place the diluted culture in the 30  C shaker. After 45 min,
remove the flask from the shaker and remove 1 mL to measure
the OD600. Optical density should be measured in this manner
every 45 min for the duration of 10 h.
7. At the end of the 10 h period, allow cultures to saturate
overnight in the 30  C shaker.
8. The next morning, steps 1–10 can be repeated for a second set
of time points. At the end of the 2-day period, optical density
measurements can be plotted on a graph as a function of time
to extrapolate a generation (or doubling) time.

3.12 Generation 1. Streak out the strain of interest to single colonies on SC

Time on Solid Media histidine plates and allow the colonies to grow to ~2 mm in
a 30  C incubator. Select a 2 mm colony from the plate using a
sterile flat end toothpick and resuspend it in 200 μL of sterile
deionized water.
2. Plate 50 μL of this resuspension onto the top left side of a fresh
synthetic medium plate lacking histidine. Tilt the plate so that
the resuspension drips down toward the bottom of the plate in
a straight line. Tilt the plate back and forth so that the cells are
evenly distributed along the streak. Allow this streak to dry for
approximately 5–7 min.
466 Gregory M. Williams and Jennifer A. Surtees

3. After the streak of cells has nearly dried, draw a single box to
the right of the streak of cells. This is where you will be putting
your single cells for the growth curve.
4. Using a microscope equipped with a micromanipulator (e.g.,
tetrad dissection microscope), invert the plate holder so that
the streak is on your right hand side.
5. First, focus on the line of cells under 30
magnification. Switch
to 150
magnification and identify single cells within the
streak of cells. Select cells that are unbudded to avoid skewing
the apparent generation time.
6. Using the micromanipulator, carefully move a single cell from
the plate to one of the boxes you drew vertically next to the
streak of cells. Place four individual cells, one per corner, into
each box. Make sure not to puncture the plate with the cell; this
will affect the growth rate.
7. To ensure representative growth rates, repeat steps 2–6 using
several different colonies from the same strain. This ensures
that the final growth curve results represent cells from several
independent colonies from the same strain background.
8. After filling all of the boxes with single cells, incubate each plate
at 30  C.
9. Regularly check the individual cells to observe budding of
daughter cells. Budding cells should be probed with the micro-
manipulator to determine whether the cells have completely
divided. When division is complete, the daughter cell is easily
removed from the mother cell with the needle.
10. As soon as the daughter cell is completely separated from the
mother, remove the daughter cell and place it just below the
mother cell on the plate. The single cells selected for the
growth curve experiment are unlikely to be at precisely the
same stage of the cell cycle. Therefore the time required to
complete the first cell division should not be included in the
growth curve. Once the first daughter cell is separated from the
mother, the time for the growth curve begins (see Note 12).
11. Every 30 min, for 10 h, remove the plates from the incubator
and observe the cells morphology, looking for daughter cell
budding. Note the amount of time it took for the daughter cell
to completely separate from the mother and restart the timer.
In order to record doubling times accurately, it is best to follow
fewer cells at a time, and repeat the experiment several times
with different starting colonies.
12. At the end of the 10 h period, median and mean doubling
times can be calculated.
 Measuring Dynamic Behavior of Trinucleotide Repeat Tracts. . . 467

4 Notes

1. We use lab-purified Thermus aquaticus (Taq) polymerase.

Briefly, Taq is overexpressed in Escherichia coli (BL21), har-
vested by centrifugation and the pellet is frozen in liquid nitro-
gen. The pellet is thawed and resuspended in Buffer A (50 mM
HEPES pH 8.0, 50 mM glucose, 1 mM EDTA) containing
lysozyme (Worthington) (4 mg/mL) and then incubated at
room temperature for at least 15 min, until the cells are lysed
and appear gooey. Buffer B (10 mM HEPES pH 8.0, 50 mM
KCl, 1 mM EDTA, 0.5% Tween 20, and 0.5% NP-40) is added
and the lysate incubated in a 75  C water bath for 45 min, with
periodic inversion. The lysate was centrifuged at 20,000
g
and the supernatant is transferred to a new tube. The superna-
tant is treated with DNaseI (Worthington), in the presence of
MgCl2 and MnCl2, at 75  C for 45 min, centrifuged and the
supernatant kept. The supernatant is dialyzed again three
changes of dialysis buffer (20 mM Tris pH 7.5, 200 mM KCl,
and 1 mM EDTA), then diluted 2
with storage buffer
(50 mM Tris pH 7.5, 50 mM KCl, 1 mM DTT, 0.5% Tween
20, and 0.5% NP-40) containing 50% glycerol and then
2
again with storage buffer containing 75% glycerol. The
final protein is aliquoted and stored at 80  C.
2. This elegant system was developed, verified, and characterized
by the Lahue lab [35, 36, 45]. It is beautifully and comprehen-
sively described in a previous Methods in Molecular Biology
publication [37].
3. Pour 8–10 spreading glass beads on each plate. Shake/rotate
the plate, keeping it on the benchtop surface to move the beads
around to cover the plate. Sometimes it is helpful to add 50 μL
of sterile dH2O to each plate before adding the glass beads as it
helps with the spreading process. A stack of ~5 plates can be
plated at the same time.
4. Mutations in the URA3 locus will also result in resistance to 5-
FOA [37]. It is therefore necessary to check for false positives
by performing colony PCR, digesting the DNA and electro-
phoresing the DNA to image the gel, as described above.
5. The following approach can be used to determine TNR tract
lengths:
(a) After running digested PCR reaction on a 12% polyacryl-
amide gel, stain and image gel (see Subheading 3.2, step
5). The digested TNR sequences should be clearly visible.
For the quantification of tract length, it is helpful to image
your gel as straight as possible, making sure to have a level
bottom edge.
(b) From a predetermined point on the gel (i.e., the bottom
of each well), measure the distance, in millimeters (mm),
468 Gregory M. Williams and Jennifer A. Surtees

between the bands of known size in the low molecular

weight marker used on the gel. These measurements (y-
axis) are plotted versus the known size of each band (x-
axis) to generate a standard curve from which tract lengths
can be interpolated.
(c) Using the same starting point of reference that was used in
step 2, measure the known TNR tract band in each lane of
the gel. Use the standard curve graph to determine the
tract length in base pairs of each fragment. Divide this
number by three to report the number of repeats within
the tract.
6. Synthetic media lacking histidine is used throughout all the
time courses to ensure that the TNR reporter construct is
maintained [37].
7. For acid-washed beads, we purchase unwashed Sigma
425–600 μm beads (G9268) and acid-wash in the laboratory.
Pour the beads into a 1 L beaker containing 5 M nitric acid in
the fumehood. Stir occasionally for 1–2 days. Wash with ~5
volumes of dH2O and then with 5 volumes of 18 Ω dH2O.
Remove as much liquid as possible and bake at 450  F for ~2 h.
Store the beads in a 20  C freezer.
8. It is worth taking the extra time to do a few colony PCR
reactions per colony per time point in case one or more colony
PCR reactions are not successful due to the use of too many
cells or too little cells in solution.
9. A 25 μm or 50 μm glass needle (Cora Styles Lab Supplies) is
attached to the dissection arm using putty. This is what will be
used to move cells around on the plate.
10. Removing marked microcolonies from plates with pipette tips
is easiest and most efficient when the pipette tip is held and
swiped across the plate at a low angle, i.e., close to parallel to
the plate.
11. A 1:10 dilution is necessary to get an accurate reading of your
saturated culture.
12. Although tedious, it is best to only have 4–5 single cells per
plate. Observing too many cells on a single plate, each with
potentially different starting times, may result in the plate
being out of the incubator for too long. Ideally, the plate
should be incubating at 30  C for 30 min at a time.

Acknowledgements

We are particularly grateful for the generosity of Dr. Robert Lahue

in providing reagents and for many useful discussions and com-
ments in developing this protocol. We are also grateful for discus-
sions with Dr. Catherine Freudenreich, Dr. Robert Bambara, Dr.
 Measuring Dynamic Behavior of Trinucleotide Repeat Tracts. . . 469

Lata Balakrishnan and Dr. Eric Alani. We also thank Alyssa Crick
Williams for her assistance with the artwork. Work in the Surtees lab
is supported by the National Institutes of Health (GM087459 to J.
A.S.) and the American Cancer Society (RSG-14-235-01 to J.A.S.).
J.A.S. is an ACS Research Scholar.

References
1. Arana ME, Kunkel TA (2010) Mutator pheno- 13. van den Broek WJAA, Nelen MR, Wansink
types due to DNA replication infidelity. Semin DG, Coerwinkel MM, te Riele H, Groenen
Cancer Biol 20:304–311 PJTA, Wieringa B (2002) Somatic expansion
2. Ellegren H (2004) Microsatellites: simple behaviour of the (CTG)n repeat in myotonic
sequences with complex evolution. Nat Rev dystrophy knock-in mice is differentially
Genet 5:435–445 affected by Msh3 and Msh6 mismatch–repair
3. Umar A, Kunkel TA (1996) DNA-replication proteins. Hum Mol Genet 11:191–198
fidelity, mismatch repair and genome instability 14. Foiry L, Dong L, Savouet C, Hubert L, Te
in Cancer cells. Eur J Biochem 238:297–307 Riele H, Junien C, Gourdon G (2006) Msh3
4. Strand M, Prolla TA, Liskay RM, Petes TD is a limiting factor in the formation of intergen-
(1993) Destabilization of tracts of simple erational CTG expansions in DM1 transgenic
repetitive DNA in yeast by mutations mice. Hum Genet 119:520–526
affecting DNA mismatch repair. Nature 15. Gannon A-MM, Frizzell A, Healy E, Lahue RS
365:274–276 (2012) MutSβ and histone deacetylase com-
5. Gordenin DA, Kunkel TA, Resnick MA (1997) plexes promote expansions of trinucleotide
Repeat expansion [mdash] all in flap? Nat repeats in human cells. Nucleic Acids Res
Genet 16:116–118 40:10324–10333
6. Kantartzis A, Williams GM, Balakrishnan L, 16. Halabi A, Ditch S, Wang J, Grabczyk E (2012)
Roberts RL, Surtees JA, Bambara RA (2012) DNA mismatch repair complex MutSβ pro-
Msh2-Msh3 interferes with Okazaki fragment motes GAA·TTC repeat expansion in human
processing to promote trinucleotide repeat cells. J Biol Chem 287:29958–29967
expansions. Cell Rep 2:216–222 17. Williams GM, Surtees JA (2015) MSH3 pro-
7. Sia E, Kokoska R, Dominska M, Greenwell P, motes dynamic behavior of Trinucleotide
Petes T (1997) Microsatellite instability in repeat tracts in vivo. Genetics 200:737–754
yeast: dependence on repeat unit size and 18. Schmidt MHM, Pearson CE (2016) Disease-
DNA mismatch repair genes. Mol Cell Biol associated repeat instability and mismatch
17:2851–2858 repair. DNA Repair 38:117–126
8. Romanova NV, Crouse GF (2013) Different 19. Mirkin SM (2007) Expandable DNA repeats
roles of eukaryotic MutS and MutL complexes and human disease. Nature 447:932–940
in repair of small insertion and deletion loops in 20. Kovtun IV, McMurray CT (2008) Features of
yeast. PLoS Genet 9:e1003920 trinucleotide repeat instability in vivo. Cell Res
9. Kunkel TA, Erie DA (2015) Eukaryotic mis- 18:198–213
match repair in relation to DNA replication. 21. Usdin K, House NCM, Freudenreich CH
Annu Rev Genet 49:291–313 (2015) Repeat instability during DNA repair:
10. Crouse GF (2016) Non-canonical actions of insights from model systems. Crit Rev Biochem
mismatch repair. DNA Repair 38:102–109 Mol Biol 50(2):1–26
11. Owen BAL, Yang Z, Lai M, Gajec M, Badger J 22. Paulson HL, Fischbeck KH (1996) Trinucleo-
d, Hayes JJ, Edelmann W, Kucherlapati R, Wil- tide repeats in neurogenetic disorders. Annu
son TM, McMurray CT (2005) (CAG)n-hair- Rev Neurosci 19:79–107
pin DNA binds to Msh2-Msh3 and changes 23. Kay C, Collins JA, Miedzybrodzka Z, Madore
properties of mismatch recognition. Nat Struct SJ, Gordon ES, Gerry N, Davidson M, Slama
Mol Biol 12:663–670 RA, Hayden MR (2016) Huntington disease
12. Manley K, Pugh J, Messer A (1999) Instability reduced penetrance alleles occur at high fre-
of the CAG repeat in immortalized fibroblast quency in the general population. Neurology
cell cultures from Huntington’s disease trans- 87:282–288
genic mice. Brain Res 835:74–79 24. Langbehn DR, Hayden MR, Paulsen JS (2010)
CAG-repeat length and the age of onset in
470 Gregory M. Williams and Jennifer A. Surtees

Huntington disease (HD): a review and valida- dystrophy type 1 and Huntington disease. J R
tion study of statistical approaches. Am J Med Soc Interface 10(88):20130605
Genet B Neuropsychiatr Genet 153B:397–408 34. Du J, Campau E, Soragni E, Jespersen C, Got-
25. Concannon C, Lahue RS (2014) Nucleotide tesfeld JM (2013) Length-dependent
excision repair and the 26S proteasome func- CTG·CAG triplet-repeat expansion in myo-
tion together to promote trinucleotide repeat tonic dystrophy patient-derived induced plu-
expansions. DNA Repair 13:42–49 ripotent stem cells. Hum Mol Genet
26. Zhang L, Leeflang EP, Yu J, Arnheim N (1994) 22:5276–5287
Studying human mutations by sperm typing: 35. Miret JJ, Pessoa-Brandão L, Lahue RS (1997)
instability of CAG trinucleotide repeats in the Instability of CAG and CTG trinucleotide
human androgen receptor gene. Nat Genet repeats in Saccharomyces cerevisiae. Mol Cell
7:531–535 Biol 17:3382–3387
27. Leeflang EP, Zhang L, Tavare S, Hubert R, 36. Miret JJ, Pessoa-Brandão L, Lahue RS (1998)
Srinidhi J, MacDonald ME, Myers RH, de Orientation-dependent and sequence-specific
Young M, Wexler NS, Gusella JF, Arnheim N expansions of CTG/CAG trinucleotide repeats
(1995) Single sperm analysis of the trinucleo- in Saccharomyces cerevisiae. Proc Natl Acad Sci
tide repeats in the Huntington’s disease gene: 95:12438–12443
quantification of the mutation frequency spec- 37. Dixon M, Bhattacharyya S, Lahue R (2004)
trum. Hum Mol Genet 4:1519–1526 Genetic assays for triplet repeat instability in
28. Leeflang EP, Tavaré S, Marjoram P, Neal COS, yeast. In: Kohwi Y (ed) Trinucleotide repeat
Srinidhi J, MacFarlane H, MacDonald ME, protocols, vol 277. Humana Press, New York,
Gusella JF, de Young M, Wexler NS, Arnheim pp 29–45
N (1999) Analysis of Germline mutation spec- 38. Alani E, Sokolsky T, Studamire B, Miret JJ,
tra at the Huntington’s disease locus supports a Lahue RS (1997) Genetic and biochemical
mitotic mutation mechanism. Hum Mol Genet analysis of Msh2p-Msh6p: role of ATP hydro-
8:173–183 lysis and Msh2p-Msh6p subunit interactions in
29. Martorell L, Gámez J, Cayuela ML, Gould FK, mismatch base pair recognition. Mol Cell Biol
McAbney JP, Ashizawa T, Monckton DG, Bai- 17:2436–2447
get M (2004) Germline mutational dynamics 39. Gietz D, Jean AS, Woods RA, Schiestl RH
in myotonic dystrophy type 1 males: allele (1992) Improved method for high efficiency
length and age effects. Neurology 62:269–274 transformation of intact yeast cells. Nucleic
30. Castel AL, Cleary JD, Pearson CE (2010) Acids Res 20:1425
Repeat instability as the basis for human dis- 40. Drake JW (1991) A constant rate of spontane-
eases and as a potential target for therapy. Nat ous mutation in DNA-based microbes. Proc
Rev Mol Cell Biol 11:165–170 Natl Acad Sci U S A 88:7160–7164
31. Higham CF, Morales F, Cobbold CA, Haydon 41. Dixon W, Massey F (1969) Introduction to
DT, Monckton DG (2012) High levels of statistical analysis. McGraw Hill, New York
somatic DNA diversity at the myotonic dystro- 42. Nair KR (1940) Table of confidence interval
phy type 1 locus are driven by ultra-frequent for the median in samples from any continuous
expansion and contraction mutations. Hum population. SankhyÄ: Indian J Stat (1933-
Mol Genet 21:2450–2463 1960) 4:551–558
32. Morales F, Couto JM, Higham CF, Hogg G, 43. Foster PL, Judith LC, Paul M (2006) Methods
Cuenca P, Braida C, Wilson RH, Adam B, del for determining spontaneous mutation rates.
Valle G, Brian R, Sittenfeld M, Ashizawa T, In: Methods in enzymology, vol 409. Aca-
Wilcox A, Wilcox DE, Monckton DG (2012) demic Press, New York, pp 195–213
Somatic instability of the expanded CTG triplet
repeat in myotonic dystrophy type 1 is a herita- 44. Zar JH (1999) Biostatistical analysis, 4th edn.
ble quantitative trait and modifier of disease Prentice Hall, Upper Saddle River, NJ
severity. Hum Mol Genet 21:3558–3567 45. Rolfsmeier ML, Dixon MJ, Pessoa-Brandão L,
33. Higham CF, Monckton DG (2013) Modelling Pelletier R, Miret JJ, Lahue RS (2001) Cis-
and inference reveal nonlinear length- elements governing trinucleotide repeat insta-
dependent suppression of somatic instability bility in Saccharomyces cerevisiae. Genetics
for small disease associated alleles in myotonic 157:1569–1579
 Chapter 31

The Detection and Analysis of Chromosome Fragile Sites

Victoria A. Bjerregaard, Özg€
un Özer, Ian D. Hickson, and Ying Liu

Abstract
A fragile site is a chromosomal locus that is prone to form a gap or constriction visible within a condensed
metaphase chromosome, particularly following exposure of cells to DNA replication stress. Based on their
frequency, fragile sites are classified as either common (CFSs; present in all individuals) or rare (RFSs;
present in only a few individuals). Interest in fragile sites has remained high since their discovery in 1965,
because of their association with human disease. CFSs are recognized as drivers of oncogene activation and
genome instability in cancer cells, while some RFSs are associated with neurodegenerative diseases. This
review summaries our current understanding of the nature and causes of fragile site “expression”, including
the recently characterized phenomenon of telomere fragility. In particular, we focus on a description of the
methodologies and technologies for detection and analysis of chromosome fragile sites.

Key words Common fragile site, Rare fragile site, Telomere fragility, Mitosis, DNA replication stress,
Karyotype, Fluorescence in situ hybridization (FISH), Ultrafine DNA bridge (UFB), Micronuclei,
53BP1 body

1 Introduction

The faithful replication of DNA is fundamental to all organisms as it

ensures that the genetic material is passed on to the next generation
of cells. In human cells, the genetic information is stored in 23 pairs
of chromosomes (22 pairs of autosomes and one pair of sex chro-
mosomes). The replication and segregation of these 46 chromo-
somes is orchestrated by the cell division cycle. This cycle comprises
interphase, during which the chromosomes are replicated, and
mitosis when the chromosomes are segregated. During mitosis,
the duplicated chromosomes are first condensed (compacted) and
then evenly distributed to the two daughter cells. Mitosis is sub-
divided into six subphases: prophase, prometaphase, metaphase,
anaphase, telophase, and cytokinesis. In metaphase, which is when
chromosomes are fully condensed and adopt the iconic X-shaped
structure, it is well documented that some chromosomal loci tend
to form a gap, break or constriction that is not stained by the most
commonly used DNA dye, DAPI. These gaps/breaks are generally

Marco Muzi-Falconi and Grant W. Brown (eds.), Genome Instability: Methods and Protocols, Methods in Molecular Biology,
vol. 1672, DOI 10.1007/978-1-4939-7306-4_31, © Springer Science+Business Media LLC 2018

471
472 Victoria A. Bjerregaard et al.

referred to as chromosomal fragile sites, and are particularly preva-

lent in cells exposed to agents that partially inhibit DNA synthesis
[1]. Based on their frequency in population, these loci are classified
as either common fragile sites (CFSs), which are present in all
individuals, or rare fragile sites (RFSs), which exist in less than 5%
of individuals [2, 3]. To date, more than 200 CFSs [4], and up to
33 RFS [5] have been identified in human cells.
Interest in chromosomal fragile sites has remained high since
their discovery over 50 years ago [6]. This is because CFSs are
recognized as a driver of genome instability in cancer cells [7],
while some RFSs are associated with neurodegenerative diseases,
such as the Fragile X syndrome [8–11]. One model for the appear-
ance (usually termed “expression”) of fragile sites is that they result
from a failure in chromatin condensation during early mitosis due
to the incomplete replication of the locus [5, 12]. This failure to
fully condense the locus, and hence to make it insufficiently com-
pact to be visualized using DAPI staining, helps to explain why the
apparently broken fragile site is never detached fully from the host
chromosome. Instead, the site appears to be loosely connected,
perhaps by undetectable DNA “tethers”; see below. It is also rele-
vant that all of the known cell culture conditions that promote
expression of fragile sites perturb DNA replication.
Great progress has been made in recent years toward under-
standing why and how fragile sites arise. This progress can be
summarized in the following areas: (1) cataloging of the genomic
features common to different fragile sites (i.e., CG/AT content,
repeats features, location on the chromosomes, or evolutionary
conservation locus) [13, 14]; (2) defining the cell culture condi-
tions that induce fragile site expression [15–18]; (3) examining the
replication program within fragile sites [19]; (4) analyzing how
gene transcription can influence the replication of fragile sites [20,
21]; (5) identifying the proteins that are recruited to fragile sites in
either S phase or mitosis [22–25]; and, more recently, (6) investi-
gating how fragile site replication is completed, which is now
known to occur in early mitosis [26]. This review discusses our
current understanding of the nature and causes of chromosome
fragility, with a particular focus on the methodologies used for the
detection and analysis of these unstable chromosomal loci.

2 The Induction and Detection of Fragile Sites

The first chromosomal fragile site to be defined in human cells was

described by Dekaban et al. in 1965 through an analysis of the
karyotypes of blood lymphocytes [6]. In the following years, cyto-
geneticists remained puzzled by the fact that these fragile sites were
observed in some laboratories but not in others. It was not until
1977 that it was shown that the type of medium in which the blood
 The Detection and Analysis of Chromosome Fragile Sites 473

lymphocytes were cultured played a role in the induction of fragile

site expression [27]. By 1984, 17 heritable fragile sites had been
identified, amongst which 14 were known to be induced by cell
culture medium deficient in folic acid [28], or by the so-called
“thymidylate stress” (e.g., by exposure to 5-fluorodeoxyuridine,
FUdR, a potent inhibitor of thymidylate synthetase) [29]. At that
time, it was noticed that thymidylate stress often produces fragile
site “hot spots” that occur nonrandomly in the genome [30]. It was
subsequently discovered that aphidicolin (APH), an inhibitor of
replicative DNA polymerases (but not polymerases β or γ), could
also induce fragility at several nonrandom sites [15]. The term
“common fragile site” was then introduced to describe the sites
that can be induced by both APH and thymidylate stress, while
“rare fragile sites” were defined as those induced by thymidylate
stress, but not APH [15]. However, soon after, it was discovered
that some CFSs could also be induced by FUdR, although at a
much lower frequency [16]. It was therefore suggested that the
fragile site classification should be based on their frequency in the
population, rather than on the form of induction [16]. To date, it is
known that many agents can induce the expression of CFSs and/or
RFSs. These include APH, BrdU, 5-azacytidine, caffeine, and
FUdR for CFSs [1, 16], and FUdR, thymidine, distamycine A,
BrdU, APH, and 5-azacytidine for RFSs [18].
Conventionally, fragile sites are detected in blood lymphocytes
that are cultured for approximately 72 h in medium supplemented
with 2% phytohemoagglutinin [4]. Various fragile-site “inducers”
can then be added to the cell culture medium according to the type
of fragile site of interest. For example, to induce the expression of
CFSs in lymphocytes, the cells are generally incubated with APH
(0.2 μM) for 24 h, and then arrested for 2 h with the mitotic spindle
inhibitor Colcemid (0.04 μg/ml) to trap the cells in metaphase
when the chromosomes are fully condensed [4]. To observe the
chromosomes under a microscope, the cells are swollen with hypo-
tonic buffer, fixed with methanol–acetic acid (3:1), dropped onto
slides, and stained by a DNA dye (usually DAPI). For more detailed
analysis of chromosome structure, such as gaps or constrictions, G-
banding (digestion of chromosomes with trypsin followed by
Giemsa staining that yields a series of lightly and darkly stained
bands) or Q-banding (a fluorescence based technique using quina-
crine staining) are commonly performed [31].
With the increase in interest in fragile site research, a number of
variations have been introduced to the above standard procedure.
First, CFSs are now widely studied in human fibroblasts and various
cancer cells [24–26, 32], as well as in mouse or rat cells [33–36]. As
a result of this change, the conditions and procedures for cell
culture, fragile site induction, and cell harvesting may vary from
study to study. For example, to harvest mitotic cells from adherent
cell lines, a method called “mitotic shake-off” is frequently used to
474 Victoria A. Bjerregaard et al.

enrich for cells in prometaphase [23]. More recently, because of the

need to define which proteins bind to and process CFSs, immuno-
fluorescence (IF) analysis, EdU incorporation, or Fluorescence in
situ hybridization (FISH), or a combination of these, are widely
performed on metaphase chromosomes [23–25]. Such develop-
ments require that the chromosome preparation step be modified
such that the detection of proteins bound to a fragile site is still
feasible. For example, in the conventional cytogenetic protocol, the
mitotic cells are fixed with methanol–acetic acid (3:1). This can
interfere with IF analysis, which requires that protein conformation
be preserved as much as possible for antibody binding. To over-
come this obstacle, a commonly used method is to centrifuge
mitotic cells onto slides (instead of dropping them) using a “cytos-
pin” apparatus, and then to fix the metaphases with paraformalde-
hyde for analysis by IF and then by FISH [37]. It is important to
bear in mind also that the protocol for this type of analysis can vary
according to the antibodies and FISH probes used [22–24]. There-
fore, each step of this protocol needs to be optimized. Figures 1 and
2 show examples of metaphase chromosomes analyzed by IF and an
EdU incorporation assay (to mark sites of active DNA synthesis), or
by IF combined with FISH to define the location of the FRA16D
CFS locus in human cells.
The detection of fragile sites using FISH has enabled research-
ers to define the fragility of specific loci in greater detail, particularly
in cases where fragility might be difficult to detect using simple
DNA staining (Fig. 3). In addition, this method has confirmed the
identity of a new class of fragile sites: telomere fragile sites. Telo-
meres are the sequences located at the ends of eukaryotic chromo-
somes. Telomeres in human cells contain a repeated TTAGGG
sequence that ranges in length from 4 to 14 kb [38]. Interest in
telomeres has surged since the discovery of the telomere-specific
reverse transcriptase, telomerase [39], and, the fact that the main-
tenance of the telomeric repeat sequence, either by telomerase [40]
or by an alternative mechanism (ALT) [41], is essential for the
long-term proliferation of cells in culture and in vivo. In the past
decade, emerging evidence has shown that telomeric repeats pose a
challenge to the DNA replication process, which can generate the
so-called “telomere fragile site” (TFS) instability that is similar to
what is seen at APH-induced CFSs. Indeed, pronounced TFS
expression is seen in cancer cells or immortalized cells, especially
when the cells are treated with APH [42] (Fig. 4). The widespread
usage of a peptide nucleic acid (PNA) FISH probe that binds
specifically to the telomere repeat sequence has been instrumental
in the analysis of this type of fragile site, as conventional chromo-
some staining methods cannot detect minor changes at the tip of
chromosomes [43]. Equally importantly, the usage of telomeric
PNA probes is compatible with IF analysis [42, 44–47].
 The Detection and Analysis of Chromosome Fragile Sites 475

Fig. 1 A representative example of the colocalization of FANCD2 and EdU with a

common fragile site. The U2OS cells were treated with APH (0.4 μM),
synchronized using RO3306 and then pulse-labeled with EdU as described
previously [26]. Metaphase cells were collected by mitotic shake-off, and then
spun onto coverslips for metaphase chromosome preparation. The metaphase
chromosomes were stained with FANCD2 antibody (green), EdU (click IT
reaction) (red), and observed using DAPI (blue). In the upper panel, the blue
arrow indicates a chromosome fragile site. In the lower panel, the yellow arrow
indicates the colocalization of FANCD2 and EdU at the fragile site, as revealed by
the yellow signal in this merged image

3 The Characterization of Fragile Sites

CFSs cover megabase-pair-long regions and tend to contain or

overlap with large genes. In addition, they are located in evolution-
ally conserved chromosomal regions, such as in syntenic regions of
mouse and human chromosomes [1]. Several hypotheses have been
put forward to explain how completion of CFS replication is
delayed compared to that in other regions of the genome. For
example, evidence has been provided that CFS expression can be
caused by either a lack of replication origins within the locus [19],
or by collisions between replication forks and the long gene tran-
scripts located there [48]. Furthermore, the molecular events lead-
ing to the fragile site expression are beginning to be revealed,
following the findings that key components of DNA repair and
476 Victoria A. Bjerregaard et al.

Fig. 2 A representative example of the analysis of colocalization of FANCD2 with

a specific common fragile site: FRA16D. GM06865 lymphocytes were cultured
and treated with APH as described previously [4]. Metaphase spreads were
prepared by the “Cytospin” method. The metaphase chromosomes were
subjected to IF analysis with FANCD2 antibody (red) followed by FISH analysis
using a BAC probe that binds specifically to the FRA16D locus (green). DNA was
observed with DAPI (blue). The yellow arrow indicates FANCD2 “twin foci”
colocalized with FRA16D FISH probes where the merged signal is yellow

Fig. 3 A representative example of the analysis of a rare fragile site, FRAXA,

using FISH. GM09237 lymphocytes, which contain over 900 CGG repeats in the
promoter region of FMR1 gene, were cultured and treated with FUdR as
described previously [58]. Metaphase spreads were prepared by the
conventional “dropping” method. The metaphase chromosomes were
subjected to FISH analysis using a BAC probe that binds specifically to FRAXA
(green). DNA was stained with DAPI (blue). The yellow arrow indicates the FRAXA
FISH probe signal that is apparently detached from the X chromosome
 The Detection and Analysis of Chromosome Fragile Sites 477

Fig. 4 A representative example of the analysis of telomere fragility. U2OS cells were treated with APH
(0.4 μM) and RO3306 as described previously [26]. Metaphase cells were collected by mitotic shake-off and
dropped onto slides for metaphase FISH analysis with a PNA telomere probe following the manufacturer’s
instructions (Panagene, Korea). The chromosomes were stained with DAPI (blue) and the telomeres were
detected with FAM labeled PNA telomere probe (green). The blown-up images below show examples of
atypical telomere signals

DNA damage response pathways, including γH2AX [22], phos-

pho-DNA-PKcs [22], and SMC1 [49], are recruited to CFSs.
We showed recently that the Fanconi anemia proteins,
FANCD2 and FANCI, bind specifically to CFS loci in cells treated
with a low dose of APH irrespective of whether the chromosome is
broken or not in metaphase [23]. These proteins form a focus on
each sister chromatid at the location of the CFS. Most interestingly,
when cells harboring these foci are analyzed in anaphase, the foci
are often linked by a DNA bridge called an ultrafine DNA bridge
(UFB). It is possible that at CFS loci where DNA replication is
disrupted or otherwise delayed, the sister chromatids are still inter-
linked by regions of underreplicated DNA when cells enter ana-
phase. This would then lead to the formation of a UFB. Most of the
time, these replication intermediates can be processed or resolved
by proteins including PICH, BLM, and RPA [50]. However, some
of the bridges/UFBs may lead to detrimental consequences, such
478 Victoria A. Bjerregaard et al.

as the formation of micronuclei or 53BP1 bodies in the next G1

phase [51]. We showed recently that underreplicated CFSs could
be detected and processed by structure-specific endonucleases
(e.g., the MUS81-EME1 complex) in early mitosis, and that this
promotes CFS expression [25]. This discovery is consistent with
the notion that the expression of CFSs is not an accidental event,
but instead is a programmed and regulated process. Moreover, we
have proposed that CFS expression is beneficial for the mainte-
nance of genome stability because it prevents much more hazard-
ous events, such as irreversible chromosome missegregation. Most
recently, we have demonstrated that, following treatment of cells
with APH, entry into the mitotic prophase triggers the recruitment
of MUS81-EME1 to CFSs, and the nuclease activity of MUS81
then promotes POLD3-dependent DNA synthesis at CFSs. When
analyzed on metaphase chromosomes, the sites of new DNA syn-
thesis in mitosis correspond to the expressed CFSs, indicating that
the gap/break in the chromosome is the location of a repair event
that delays condensation of the locus. This DNA repair-based
replication “salvage” pathway serves as a mechanism to minimize
chromosome missegregation and nondisjunction, particularly in
cancer cells with elevated levels of chromosome instability
(CIN þve) [26].
In the case of RFSs, most of the studies have been focused on
the FRAXA (fragile X) locus that is associated with Fragile X syn-
drome. This disorder is characterized by a spectrum of intellectual
disabilities, as well as atypical physical characteristics (including an
elongated face, large or protruding ears, and large testicles), and
behavioral characteristics such as stereotypic movements (e.g.,
hand-flapping), and social anxiety. It is well established that Fragile
X syndrome is caused by the expansion of the CGG trinucleotide
repeat that affects the expression of the Fragile X mental retardation
1 gene (FMR1) [52, 53]. In the past three decades, great progress
has been made in deciphering the mechanism underlying the
expansion of the repeats that denote such folate-sensitive RFSs
[54–56]. However, the mechanism underlying the fragility of
these loci and other CGG trinucleotide repeat regions remains
unknown. Undoubtedly, the methodologies that have been devel-
oped to define the fragility of CFSs can be applied to the analysis of
this type of fragile site.

4 Concluding Remarks

In the past five decades, a considerable amount has been learned

about chromosome fragile sites in the human genome. The meth-
ods involved in the detection and analysis of chromosome fragile
sites have progressed from the original cytogenetic-based techni-
ques to fluorescence based ones (i.e., IF and FISH) (summarized in
Fig. 5). This has allowed us to understand why fragile sites are
 The Detection and Analysis of Chromosome Fragile Sites 479

Fig. 5 A flowchart of the methodology involved in the induction, detection, and analysis of human chromosome
fragile sites. References for the methods applied are indicated where appropriate
480 Victoria A. Bjerregaard et al.

unusually “fragile”, and the strategies utilized by cells to cope with

this fragility to ensure the faithful replication and segregation of
DNA during the cell division cycle. In addition to the methods
mentioned above, great strides have been made in imaging tech-
nologies that allow for more sensitive, faster, and reliable detection
of genomic loci. Moreover, high content array CGH technology
has enabled researchers to detect DNA copy number changes in an
interval of 1 marker per 1737 bases cross the whole human genome
with a input of DNA as little as 250 ng (e.g., via the CytoScan® HD
platform developed by Affymetrix).
It seems likely that chromosome fragile sites will remain an
intriguing feature of human cells and a fascinating topic for research
for the following reasons: (1) CFSs exist in every individual and
many are in conserved chromosome regions in mammals. This begs
the question as to why that is the case, and whether there is any
advantage for the cells to maintain those apparently fragile loci
during evolution? (2) Several different types of DNA synthesis
inhibitors can induce the expression of either CFS or RFS, and
hence it would be interesting to find out why that is the case. Is
this because of the DNA sequence differences at each fragile site, or
because different inhibitors initiate a different DNA damage
response? (3) Errors arising during the repair or replication at
fragile sites following replication stress (that can be either endoge-
nous, such as oncogene associated, or exogenous, such as folate
deficiency) lead to chromosome translocations or an abnormal
number of chromosomes. This can, in turn, drive the development
of several debilitating disorders (e.g. cancer, cognitive dysfunction,
and infertility). It will be a high priority to identify the causes of
those errors. (4) The recently identified fragility at telomeric sites
raises questions regarding which factors increase or decrease the
expression of TFSs in cancer cells, considering that there is evidence
to show that APH can increase the expression of TFSs in ALT cells
[57]. The answer to this question would potentially allow us to
develop novel anticancer drug treatments given that cancer cells
have shown unusual telomere length maintenance either via an
active telomerase or the ALT mechanism.

Acknowledgment

We thank Theresa Wass, Malgorzata Clausen, and Wei Wu for

experimental help. Work in the authors’ laboratory is supported
by a Faculty Ph.D. fellowship from the University of Copenhagen
(V.A.B.), the European Union (O.O., I.D.H., and Y.L.), The
Danish National Research Foundation (DNRF115; I.D.H. and Y.
L.), and The Nordea Foundation (I.D.H.).
 The Detection and Analysis of Chromosome Fragile Sites 481

References
1. Durkin SG, Glover TW (2007) Chromosome sites in human-chromosomes. Hum Genet 67
fragile sites. Annu Rev Genet 41:169–192 (2):136–142
2. Kremer EJ et al (1991) Mapping of DNA insta- 16. Rao PN, Heerema NA, Palmer CG (1988)
bility at the Fragile-X to a trinucleotide repeat Fragile sites induced by Fudr, caffeine, and
sequence P(Ccg)N. Science 252 aphidicolin - their frequency, distribution, and
(5013):1711–1714 analysis. Hum Genet 78(1):21–26
3. Sutherland GR, Baker E, Richards RI (1998) 17. Kuwano A, Murano I, Kajii T (1990) Cell type-
Fragile sites still breaking. Trends Genet 14 dependent difference in the distribution and
(12):501–506 frequency of excess thymidine-induced com-
4. Mrasek K et al (2010) Global screening and mon fragile sites - lymphocytes-T and skin
extended nomenclature for 230 aphidicolin- fibroblasts. Hum Genet 84(6):527–531
inducible fragile sites, including 61 yet unre- 18. Sutherland GR (1991) Chromosomal fragile
ported ones. Int J Oncol 36(4):929–940 sites. Genet Analysis Techn Appl 8(6):161–166
5. Sutherland GR (2003) Rare fragile sites. Cyto- 19. Letessier A et al (2011) Cell-type-specific repli-
genet Genome Res 100(1–4):77–84 cation initiation programs set fragility of the
6. Dekaban A (1965) Persisting clone of cells with FRA3B fragile site. Nature 470(7332):120–123
an abnormal chromosome in a woman previ- 20. Le Tallec B et al (2013) Common fragile site
ously irradiated. J Nucl Med 6(10):740–746 profiling in epithelial and erythroid cells reveals
7. Richards RI (2001) Fragile and unstable chro- that most recurrent cancer deletions lie in frag-
mosomes in cancer: causes and consequences. ile sites hosting large genes. Cell Rep 4
Trends Genet 17(6):339–345 (3):420–428
8. Bell MV et al (1991) Physical mapping across 21. Wilson TE et al (2015) Large transcription
the fragile X: hypermethylation and clinical units unify copy number variants and common
expression of the fragile X syndrome. Cell 64 fragile sites arising under replication stress.
(4):861–866 Genome Res 25(2):189–200
9. Fu YH et al (1991) Variation of the CGG 22. Schwartz M et al (2005) Homologous recom-
repeat at the fragile X site results in genetic bination and nonhomologous end-joining
instability: resolution of the Sherman paradox. repair pathways regulate fragile site stability.
Cell 67(6):1047–1058 Genes Dev 19(22):2715–2726
10. Pieretti M et al (1991) Absence of expression 23. Chan KL, Palmai-Pallag T, Ying S, Hickson ID
of the FMR-1 gene in fragile X syndrome. Cell (2009) Replication stress induces sister-
66(4):817–822 chromatid bridging at fragile site loci in mito-
11. Verkerk AJ et al (1991) Identification of a gene sis. Nat Cell Biol 11(6):753–760
(FMR-1) containing a CGG repeat coincident 24. Naim V, Wilhelm T, Debatisse M, Rosselli F
with a breakpoint cluster region exhibiting (2013) ERCC1 and MUS81-EME1 promote
length variation in fragile X syndrome. Cell 65 sister chromatid separation by processing late
(5):905–914 replication intermediates at common fragile
12. El Achkar E, Gerbault-Seureau M, Muleris M, sites during mitosis. Nat Cell Biol 15
Dutrillaux B, Debatisse M (2005) Premature (8):1008–U1270
condensation induces breaks at the interface of 25. Ying S et al (2013) MUS81 promotes common
early and late replicating chromosome bands fragile site expression. Nat Cell Biol 15
bearing common fragile sites. Proc Natl Acad (8):1001–1007
Sci U S A 102(50):18069–18074 26. Minocherhomji S et al (2015) Replication
13. Handt O, Sutherland GR, Richards RI (2000) stress activates DNA repair synthesis in mitosis.
Fragile sites and minisatellite repeat instability. Nature 528(7581):286–290
Mol Genet Metab 70(2):99–105 27. Sutherland GR (1977) Fragile sites on human
14. Fungtammasan A, Walsh E, Chiaromonte F, chromosomes: demonstration of their depen-
Eckert KA, Makova KD (2012) A genome- dence on the type of tissue culture medium.
wide analysis of common fragile sites: what Science 197(4300):265–266
features determine chromosomal instability in 28. Sutherland GR (1979) Heritable fragile sites
the human genome? Genome Res 22 on human chromosomes I. Factors affecting
(6):993–1005 expression in lymphocyte culture. Am J Hum
15. Glover TW, Berger C, Coyle J, Echo B (1984) Genet 31(2):125–135
DNA polymerase-alpha inhibition by aphidico- 29. Glover TW (1981) FUdR induction of the X
lin induces gaps and breaks at common fragile chromosome fragile site: evidence for the
482 Victoria A. Bjerregaard et al.

mechanism of folic acid and thymidine inhibi- 44. Vannier JB, Pavicic-Kaltenbrunner V, Petal-
tion. Am J Hum Genet 33(2):234–242 corin MIR, Ding H, Boulton SJ (2012)
30. Sutherland GR (1983) The fragile X chromo- RTEL1 dismantles T loops and counteracts
some. Int Rev Cytol 81:107–143 telomeric G4-DNA to maintain telomere
31. Keagle MB, Gersen SL (2013) Basic cytogenet- integrity. Cell 149(4):795–806
ics laboratory procedures. In: Keagle MB, Ger- 45. Vannier JB et al (2013) RTEL1 is a replisome-
sen SL (eds) The principles of clinical associated helicase that promotes telomere and
cytogenetics, 3rd edn. Springer ScienceþBusi- genome-wide replication. Science 342
ness Media, New York, pp 53–65 (6155):239–242
32. Le Tallec B et al (2011) Molecular profiling of 46. Wan BB et al (2013) SLX4 assembles a telo-
common fragile sites in human fibroblasts. Nat mere maintenance toolkit by bridging multiple
Struct Mol Biol 18(12):1421–1423 endonucleases with telomeres. Cell Rep 4
33. Elder FFB, Robinson TJ (1989) Common (5):861–869
fragile sites do not correlate with cancer chro- 47. Sarek G, Vannier JB, Panier S, Petrini JHJ,
mosome breakpoints in the laboratory mouse Boulton SJ (2015) TRF2 recruits RTEL1 to
and rat. Cancer Genet Cytogen 41 telomeres in S phase to promote T-loop
(2):246–246 unwinding. Mol Cell 57(4):622–635
34. Elder FFB, Robinson TJ (1989) Rodent com- 48. Helmrich A, Ballarino M, Tora L (2011) Colli-
mon fragile sites - are they conserved - evidence sions between replication and transcription
from mouse and rat. Chromosoma 97 complexes cause common fragile site instability
(6):459–464 at the longest human genes. Mol Cell 44
35. Helmrich A, Stout-Weider K, Hermann K, (6):966–977
Schrock E, Heiden T (2006) Common fragile 49. Musio A et al (2005) SMC1 involvement in
sites are conserved features of human and fragile site expression. Hum Mol Genet 14
mouse chromosomes and relate to large active (4):525–533
genes. Genome Res 16(10):1222–1230 50. Liu Y, Nielsen CF, Yao Q, Hickson ID (2014)
36. Helmrich A et al (2007) Identification of the The origins and processing of ultra fine ana-
human/mouse syntenic common fragile site phase DNA bridges. Curr Opin Genet Dev
FRA7K/Fra12C1–relation of FRA7K and 26:1–5
other human common fragile sites on chromo- 51. Lukas C et al (2011) 53BP1 nuclear bodies
some 7 to evolutionary breakpoints. Int J Can- form around DNA lesions generated by mitotic
cer 120(1):48–54 transmission of chromosomes under replica-
37. Sullivan BA, Warburton PE (1999) Studying tion stress. Nat Cell Biol 13(3):243–253
the progression of vertebrate chromosomes 52. Colak D et al (2014) Promoter-bound trinu-
through mitosis by immunofluorescence and cleotide repeat mRNA drives epigenetic silenc-
FISH. In: Warburton PE (ed) Chromosomes ing in fragile X syndrome. Science 343
strutural analysis: a practical approach. Oxford (6174):1002–1005
University Press, Oxford, pp 81–101 53. Gerhardt J et al (2014) The DNA replication
38. de Lange T et al (1990) Structure and variabil- program is altered at the FMR1 locus in fragile
ity of human chromosome ends. Mol Cell Biol X embryonic stem cells. Mol Cell 53(1):19–31
10(2):518–527 54. Lee DY, McMurray CT (2014) Trinucleotide
39. Greider CW, Blackburn EH (1985) Identifica- expansion in disease: why is there a length
tion of a specific telomere terminal transferase threshold? Curr Opin Genet Dev 26:131–140
activity in Tetrahymena extracts. Cell 43(2 Pt 55. Su XA, Dion V, Gasser SM, Freudenreich CH
1):405–413 (2015) Regulation of recombination at yeast
40. Bodnar AG et al (1998) Extension of life-span nuclear pores controls repair and triplet repeat
by introduction of telomerase into normal stability. Genes Dev 29(10):1006–1017
human cells. Science 279(5349):349–352 56. Usdin K, House NC, Freudenreich CH (2015)
41. Bryan TM, Englezou A, Dalla-Pozza L, Dun- Repeat instability during DNA repair: insights
ham MA, Reddel RR (1997) Evidence for an from model systems. Crit Rev Biochem Mol
alternative mechanism for maintaining telo- Biol 50(2):142–167
mere length in human tumors and tumor- 57. Root H et al (2016) FANCD2 limits BLM-
derived cell lines. Nat Med 3(11):1271–1274 dependent telomere instability in the alterna-
42. Sfeir A et al (2009) Mammalian telomeres tive lengthening of telomeres pathway. Hum
resemble fragile sites and require TRF1 for Mol Genet 25(15):3255–3268
efficient replication. Cell 138(1):90–103 58. Kumari D et al (2009) The role of DNA dam-
43. Lansdorp PM et al (1996) Heterogeneity in age response pathways in chromosome fragility
telomere length of human chromosomes. in Fragile X syndrome. Nucleic Acids Res 37
Hum Mol Genet 5(5):685–691 (13):4385–4392
 Chapter 32

Imaging of DNA Ultrafine Bridges in Budding Yeast

Oliver Quevedo and Michael Lisby

Abstract
DNA ultrafine bridges (UFBs) are a type of chromatin-free DNA bridges that connect sister chromatids in
anaphase and pose a threat to genome stability. However, little is known about the origin of these
structures, and how they are sensed and resolved by the cell. In this chapter, we review tools and methods
for studying UFBs by fluorescence microscopy including chemical and genetic approaches to induce UFBs
in the model organism Saccharomyces cerevisiae.

Key words Anaphase bridges, UFBs, Fluorescence microscopy, Yeast, Ultrafine bridges

1 Introduction

The failure to fully separate sister chromatids during mitosis can

lead to the appearance of DNA anaphase bridges, which is a poten-
tial source of genomic instability, a hallmark of cancer cells [1–4].
Thus, removal of sister chromatid linkages is critical in order to
avoid the formation of anaphase bridges, and specialized mechan-
isms exist in the cell to ensure that sister chromatids are no longer
held together upon anaphase entry.
So far, two main types of anaphase bridges have been described
[5, 6]: chromatin bridges and ultrafine bridges (UFBs). Contrary
to the former, UFBs do not contain histones and other chromatin
components and cannot be visualized by staining with conventional
DNA dyes such as DAPI or Hoechst. Nevertheless, they can be
visualized by monitoring a subset of proteins known to bind UFBs,
such as the BLM helicase (Sgs1 in budding yeast), PICH, or
FANCM (Mph1 in budding yeast), among others [7–9]. Strikingly,
UFBs are observed in up to 30% of dividing cells in unperturbed
conditions [8].
Four main types of linkages between sister chromatids are
thought to lead to the appearance of anaphase bridges [10]
(Fig. 1): (A) linkages mediated by the cohesin complex, (B) catena-
tions and other topological entanglements, (C) the presence of

Marco Muzi-Falconi and Grant W. Brown (eds.), Genome Instability: Methods and Protocols, Methods in Molecular Biology,
vol. 1672, DOI 10.1007/978-1-4939-7306-4_32, © Springer Science+Business Media LLC 2018

483
484 Oliver Quevedo and Michael Lisby

A B

C D
branch
migration

Fig. 1 Schematic overview of the main types of linkages that lead to the appearance of anaphase bridges
during mitosis. (a) Sister chromatids held together by the cohesin complex. (b) Catenations between sister
chromatids. (c) Underreplicated regions. (d) Holliday junctions that can branch migrate to hemicatenanes

underreplicated regions at mitotic entry, and (D) the presence of

recombination intermediates that have not been removed before
anaphase/telophase. Hence, anaphase bridges may contain single-
stranded gaps, stalled replication forks, or replication/recombina-
tion intermediates. As mentioned above, the linkages established
between sister chromatids are removed by a number of mechanisms
involving separase (removal of the cohesin complex), condensin
and type II topoisomerases (removal of catenations and other
topological constraints), DNA helicases such as BLM/Sgs1 and
the type I topoisomerase Top3 (branch migration and removal of
hemicatenanes), structure-selective endonucleases (removal of
recombination intermediates), and unscheduled DNA synthesis
during mitosis promoted by proteins such as TopBP1/Dpb11
(removal of underreplicated regions). These mechanisms work in
a coordinated and regulated manner according to cell cycle pro-
gression, and their misregulation leads to the persistence of ana-
phase bridges [11]. Notably, failures in the removal of cohesion and
the presence of recombination intermediates during mitosis lead
mainly to the accumulation of chromatin bridges [10, 12–17],
whereas catenanes and topological entanglements can potentially
lead to both chromatin bridges and UFBs [6, 10, 12–19], and the
presence of underreplicated regions during mitosis leads mainly to
UFBs [6, 15, 20].
In humans, two types of UFBs have been described [6]: (1)
Fanconi anemia (FA)-associated UFBs, originating from regions of
the genome in which the completion of replication can be prob-
lematic, such as the chromosomal common fragile sites (CFSs), and
induced by inhibition of replication [5, 21, 22]; and (2) non-FA-
associated UFBs, originated from DNA catenations and induced by
inhibition of Topoisomerase 2 [6]. Whether a similar distinction
can be made in S. cerevisiae using the FANCM homologue, Mph1,
remains to be tested.
 Imaging of UFBs in Yeast 485

1.1 Budding Yeast as Saccharomyces cerevisiae has a long track-record as a powerful

a Model to Study genetic and cell biological model system [23]. However, when it
Anaphase Bridges comes to imaging of anaphase bridges, the closed mitosis of bud-
ding yeast poses a technical challenge. In contrast to higher eukar-
yotes, there is no disassembly of the yeast nuclear envelope during
mitosis. Thus, a nuclear protein displaying a bridge-like structure
connecting the separating nuclei in mitosis could reflect the nucle-
oplasm contained in the so-called anaphase tube rather than a
protein-bound DNA bridge. To overcome the challenge of distin-
guishing DNA anaphase bridges from nucleoplasm, it is necessary
to use additional markers that provide an indication of the segrega-
tion status of the chromosomes. For instance, since relaxation of
the spindle tension only occurs after the sister chromatids have
been fully resolved during anaphase/telophase, monitoring the
lengthening of the mitotic spindle by time-lapse microscopy pro-
vides a good indication of sister chromatid separation. A spindle
pole body (SPB) marker, such as Spc110 fused to a fluorescent
protein, provides an accurate tool for monitoring the spindle
length. Using this marker, together with another fluorescent pro-
tein fused to a nuclear localization signal (NLS), as a marker for the
nucleoplasm, it was shown that the resolution of the Dpb11-bound
DNA anaphase bridges takes place concomitant with relaxation of
the mitotic spindle, whereas the anaphase tube remains until abscis-
sion [15]. Another indicator of the presence of DNA bridges in the
division plane during mitotic exit is the activation of the NoCut
checkpoint [24]. This checkpoint senses the presence of both chro-
matin bridges and UFBs [15, 24]. Activation of the NoCut check-
point causes a delay in cytokinesis [24]. This delay can be
monitored by the use of a plasma membrane marker fused to a
fluorescent protein as an indicator of abscission [25] (Fig. 2). Both

A B
Ultrafine bridge Chromatin bridge

Dpb11-YFP NLS-RFP YFPCFP Dpb11-YFP NLS-RFP YFPCFP

Ina1-CFP Ina1-CFP
Hoechst Spc110-CFP DIC Hoechst Spc110-CFP DIC

Fig. 2 Examples of anaphase bridges. Cells (ML734-9B: MATa tTA(tetR-VP16)-tetO2-DPB11-4ala-YFP::KanMX

NLS-yEmRFPrv::URA3 SPC110-CFP::KAN INA1-4ala-CFP) were grown in SC þ Ade to mid-log phase and
stained with Hoechst before imaging (see Subheading 3.1 and Notes 1 and 2). (a) Ultrafine bridge. (b)
Chromatin bridge. Yellow arrowheads indicate anaphase bridges. Purple arrowhead indicates Hoechst-
strained chromatin bridge. Scale bar, 3 μm
486 Oliver Quevedo and Michael Lisby

Table 1
Protein markers for analysis of anaphase bridges. After tagging with a fluorescent protein (see Note
2), these proteins constitute fluorescent reporters to visualize UFBs, chromatin bridges, nucleoplasm,
nucleolus, and the NoCut checkpoint delay

UFB markers Dpb11, Rfa1, Sgs1, Top3, Top2, Ddc2, Nop1 (nucleolus), Net1
(nucleolus)
Chromatin markers Hta1, Hta2, Htz1, Nhp10, Rsc1
Anaphase tube/ Nup49 (nuclear pore complex), NLS-XFPa (nucleoplasm)
nucleoplasm marker
NoCut checkpoint Ina1, 2xPH-XFPa (plasma membrane)
a
XFP indicates any fluorescent protein

the PH domain and the Ina1 protein, fused to CFP, have been used
to monitor abscission [15, 25]. Other markers that can be used to
determine the structure of the anaphase tube are proteins that
localize to the nuclear envelope, such as the nuclear pore complex
subunit Nup49. This protein does not mark the nucleoplasm, but
allows for visualization of the anaphase tube [15]. Thus, combining
a spindle marker with a nucleoplasm/nuclear envelope marker is a
good approach to distinguish proteins bound to DNA bridges in
anaphase from those bridge-like structures that reflect the nucleo-
plasm contained in the anaphase tube. As a marker for anaphase
DNA bridges, there is a set of proteins that are known to bind
UFBs: Dpb11, the Sgs1-Top3-Rmi1 complex, RPA, and Top2
(Table 1). Chromatin bridges can be distinguished from UFBs by
including a chromatin marker such as histone H2A or by staining
DNA with DAPI or Hoechst, which do not detect UFBs. Finally,
the presence of DNA in UFBs can be demonstrated by the incor-
poration of and staining for the nucleoside analogue 5-ethynyl-20 -
deoxyuridine (EdU) [15].

1.2 Induction of Even if UFBs can be detected in unperturbed cells, it can be useful
Anaphase Bridges to apply conditions that increase their frequency, either by the use
of mutants or by the use of drugs that are known to interfere with
DNA metabolic processes. For instance, the use of the thermosen-
sitive allele top2-1 at the permissive temperature was shown to
increase the number of UFBs by ~4-fold in asynchronous cultures
[15]. As for the use of drugs to induce an increase in the frequency
of UFBs, both methyl methanesulfonate (MMS) and hydroxyurea
(HU) can be used. Transient exposure (70 min) to 0.03% MMS
leads to a ~5-fold increase of UFBs [15]. Similarly, exposure to
5.6 mM HU for 2 h leads to a ~4-fold increase in the frequency of
yeast cells with UFBs in asynchronous culture (unpublished data).
 Imaging of UFBs in Yeast 487

2 Materials

1. SC þ Ade medium: synthetic complete medium supplemented

with 100 μg/ml adenine [26].
2. SC þ Ade patches (SC þ Ade/LMPA): SC þ Ade with 1% low
melting-point agarose (V2111; Promega).
3. DAPI solution: 10 mg/ml stock solution in deionized water
(1000).
4. Hoechst 33258 solution: 5 mg/ml stock solution in deionized
water (1000) (B2883; Sigma-Aldrich).
5. α-factor mating pheromone.
6. Hydroxyurea.
7. Wax: equal volumes of molten bees’ wax, petroleum jelly (VWR
Scientific), and lanolin (Sigma).
8. Paraformaldehyde (PFA): For a 3.7% PFA solution, add 3.7 g
of PFA to approximately 70 ml of deionized water. Heat to
60  C in a water bath. Add a few drops of 1 M NaOH until the
PFA dissolves. Cool to room temperature and add 10 ml of
10 PBS. Adjust the pH to 7.4 and bring to a final volume of
100 ml with deionized water.

3 Methods

3.1 Live Cell Staining 1. Grow cells in SC þ Ade medium.

of Chromatin Bridges 2. Transfer 1 ml of culture to a separate tube. Add DAPI (to a final
Using Conventional concentration of 10 μg/ml) or Hoechst 33258 (to a final
DNA Dyes concentration of 5 μg/ml) and incubate cells with shaking for
10 min at the same temperature used in step 1.
3. Spin down cells at 1000 rpm, wash once with SC þ Ade, and
resuspend in 50–100 μl SC þ Ade.
4. Transfer cells to a microscope slide for imaging under the
microscope (see Note 3).

3.2 Visualization of Since their discovery in the 1970s by Hartwell and colleagues in
Anaphase Bridges in their study of the genetic control of the cell division cycle [27], the
Metaphase/Anaphase/ thermosensitive alleles of genes encoding proteins controlling dif-
Telophase Using ferent aspects of the cell cycle progression have been extensively
cdc20, cdc14 or cdc15 used to synchronize yeast cultures. Examples of those alleles are the
Thermosensitive cdc20-1 allele that has been used to block the cell cycle in metaphase
Mutants [27]; the cdc14-1 allele used to synchronize cells at the anaphase-to-
telophase transition [27]; and the cdc15-2 allele used to block cells
in telophase [27] (Fig. 3).
488 Oliver Quevedo and Michael Lisby

cdc20

cdc14

cdc15

Fig. 3 Schematic representation of Dpb11-YFP localization in a metaphase,

anaphase, or telophase cell cycle arrest. (a) Metaphase arrest by disruption of the
Cdc20 function. (b) Anaphase arrest by disruption of the Cdc14 function. In this case it
is possible to observe bulges in the bridge that have been shown to colocalize with
unresolved regions of the ribosomal DNA [37]. (c) Telophase arrest by disruption of
the Cdc15 function. Yellow, Dpb11-YFP signal. Orange, cytoplasm

1. Grow cells in SC þ Ade until early log phase

(OD600 ¼ 0.2–0.5) at 25  C (permissive temperature) with
moderate shaking.
2. Incubate cells at 37  C (restrictive temperature) with moderate
shaking for 3 h. Then, collect a sample and proceed as in steps 3
and 4 above (see Subheading 3.1) in order to image directly under
the microscope. If a DNA dye is used for chromatin staining,
proceed as stated above (see Subheading 3.1). It is crucial that the
incubation temperature is 37  C during the staining.
3. To release cells from the cell cycle arrest, transfer the culture
back to 25  C with moderate shaking and collect samples at
appropriate times for imaging under the microscope, proceeding
as stated before. If a DNA dye is used for chromatin staining,
proceed as stated above (see Subheading 3.1). In this case, the
incubation temperature during the staining should be 25  C.

3.3 Visualization of As mentioned above, haploid cells carrying the thermosensitive

Anaphase Bridges in top2-1 allele have a higher frequency of UFBs [15], making this
Metaphase/Anaphase/ mutant suitable for studying this type of bridges. However, it is not
Telophase Using the possible to use a combination of top2-1 and the thermosensitive
Top2-1 cdc20/cdc14/cdc15 mutants to block the cell cycle in metaphase/
Thermosensitive anaphase/telophase, since the incubation at the restrictive temper-
Mutant ature required for these cell cycle arrests would inactivate the
 Imaging of UFBs in Yeast 489

Topoisomerase 2 [12], leading to an increase of chromatin bridges

instead of UFBs [15]. Thus, an alternative strategy to inactivate
CDC20, CDC14, and CDC15 must be applied. For instance, the
methionine repressible MET3 promoter has been used to switch off
the expression of CDC20 and thus synchronize cells in metaphase
followed by release into anaphase upon removal of methionine
[28]. For the disruption of Cdc14 and Cdc15 function, different
conditional non-thermosensitive alleles have been successfully used
in yeast, such as the mouse ornithine decarboxylase for the deple-
tion of Cdc14 [29] and the Cdc15-as1 mutant protein that can be
inhibited by the ATP analog “PP1 analog 8” [30, 31]. In these
cases, the protocol to visualize UFBs is similar to the one described
(see Subheadings 3.1 and 3.2), but keeping the temperature at
25  C and inducing the disruption of any of the conditional alleles
immediately after the release from G1 (see Subheading 3.4).

3.4 Time-Course to 1. Grow MATa haploid cells in SC þ Ade until early log phase
Monitor the Formation (OD600 ¼ 0.2–0.5) at 25  C/30  C with moderate shaking.
and Resolution of 2. Add the α-factor pheromone to the culture to synchronize cells
Anaphase Bridges in G1. If the strain that is being used is bar1Δ, then use a final
Induced by Mild concentration of 50 ng/ml and an incubation time of 3 h to
Replication Stress allow all cells to reach the G1 arrest. If the strain used is BAR1,
then use a final concentration of 5 μg/ml. Incubate cells for 1 h
and then add fresh α-factor (half of the amount used before).
Incubate cells for an additional hour and verify the arrest under
the microscope.
3. To release cells from the G1 arrest, wash cells twice with 1
volume of fresh medium. Then, resuspend pellet in 1 volume
fresh medium containing 5.6 mM HU and incubate cells at
25  C/30  C with moderate shaking. Collect samples at appro-
priate time points for imaging. If a DNA dye is used for
chromatin staining, proceed as stated above (see Subheading
3.1). Spin cells at 1000 rpm, wash once with SC þ Ade, and
resuspend cells in 0.1 volume of SC þ Ade.
4. Transfer cells to a microscope slide for imaging under the
microscope (see Note 3).

3.5 Preparation of 1. Prepare a microscope slide with a patch of SC þ Ade solid

Samples for Time- medium. For preparation of the patch, add 200 μl of melted
Lapse Experiments SC þ Ade containing 1% low melting point agarose to the
center of a microscope slide. Then cover with another slide
and wait 10 min for the agarose to solidify. Finally, remove the
top slide very carefully in order to avoid breakage of the
SC þ Ade patch (Fig. 4).
490 Oliver Quevedo and Michael Lisby

Fig. 4 Scheme for the preparation of SC þ Ade patches on slides for time-lapse microscopy. (a) Add 200 μl
melted SC þ Ade/LMPA to the slide; (b, c) cover with another slide, spreading the SC þ Ade as much as
possible without breaking the patch. Allow to solidify for 10 min; (d) carefully remove the top slide, trying to
avoid breaking the patch, and add 2.8 μl of cell suspension; (e, f) cover the cells with a coverslip; (g) remove
the excess of SC þ Ade/LMPA; (h, i) use wax to carefully seal the edges of the coverslip and proceed to
imaging

2. If a synchronized culture is to be used, proceed first with the

synchronization of the culture in the desired cell cycle stage as
stated above (see Subheadings 3.2–3.4).
3. Immediately after release, take 500 μl of cell culture and spin
down cells at 1000 rpm.
4. Discard supernatant leaving 50–100 μl of medium in the tube.
Then, resuspend the pellet in the remaining medium.
5. Take 2.8 μl of cell suspension and carefully pipette onto the
SC þ Ade patch on the microscope slide.
6. Cover with a coverslip. Finally, seal the edges using melted wax
and proceed to imaging.
 Imaging of UFBs in Yeast 491

3.6 EdU Staining of 1. Grow cells in SC þ Ade until early log phase
Anaphase Bridges (OD600 ¼ 0.2–0.5) at 25  C/30  C with moderate shaking.
2. Add 20 μM EdU to the medium and incubate cells for 3 h with
moderate shaking (see Note 4).
3. Collect cells by centrifugation at 1000 rpm.
4. Fix cells by resuspension in 3.7% PFA for 15 min at room
temperature with shaking.
5. Quench PFA with 0.3 M glycine for 15 min at room
temperature.
6. Wash fixed cells in PBS.
7. Stain with 10 μg/ml DAPI or 5 μg/ml Hoechst 33258 for
10 min.
8. Proceed to detection as described in the Click-iT EdU Alexa
Fluor 594 imaging kit manual (Invitrogen).

4 Notes

1. Microscopy was performed on a wide-field AxioImager Z1

(Carl Zeiss MicroImaging, Inc) equipped with a 100 objec-
tive lens (Zeiss PLAN-APO, NA 1.4), a cooled Orca-ER CCD
camera (Hamamatsu, Japan), differential interference contrast
(DIC), and a Zeiss HXP120C illumination source. Images
were acquired and processed using Volocity (PerkinElmer)
software. Images were pseudocolored according to the approx-
imate emission wavelength of the fluorophores.
2. Proteins were tagged with the cyan fluorescent protein (CFP,
clone W7) [32], yellow fluorescent protein (YFP, clone 10C)
[33], and red fluorescent proteins (RFP, clone yEmRFP) [34]
as described [35].
3. The thickness of the coverslip should match the objective lens
and the volume of the sample should match the area of the
coverslip to restrict all cells to the same focal plane. In this
study, a cell suspension of 2.8 μl was mounted with a no. 1.5
coverslip (24 mm  24 mm).
4. For incorporation of EdU, cells should be engineered to
express thymidine kinase and a nucleoside transporter [36].

Acknowledgments

This work was supported by The Danish Agency for Science, Tech-
nology and Innovation, the Villum Foundation, and the Danish
National Research Foundation (DNRF115). We thank Camilla
Colding and Giedrė Bačinskaja for comments on the manuscript.
492 Oliver Quevedo and Michael Lisby

References
1. Draviam VM, Xie S, Sorger PK (2004) Chro- bridges to prevent genome instability. J Cell
mosome segregation and genomic stability. Biol 204:45–59
Curr Opin Genet Dev 14:120–125 16. Spence JM, Phua HH, Mills W et al (2007)
2. Gollin SM (2005) Mechanisms leading to Depletion of topoisomerase IIalpha leads to
chromosomal instability. Semin Cancer Biol shortening of the metaphase interkinetochore
15:33–42 distance and abnormal persistence of PICH-
3. Bannon JH, Mc Gee MM (2009) Understand- coated anaphase threads. J Cell Sci
ing the role of aneuploidy in tumorigenesis. 120:3952–3964
Biochem Soc Trans 37:910–913 17. Gallina I, Christiansen SK, Pedersen RT et al
4. Hanahan D, Weinberg RA (2011) Hallmarks (2016) TopBP1-mediated DNA processing
of cancer: the next generation. Cell during mitosis. Cell Cycle 15:176–183
144:646–674 18. Bhalla N, Biggins S, Murray AW (2002) Muta-
5. Chan KL, Hickson ID (2009) On the origins tion of YCS4, a budding yeast condensin sub-
of ultra-fine anaphase bridges. Cell Cycle unit, affects mitotic and nonmitotic
8:3065–3066 chromosome behavior. Mol Biol Cell
6. Chan KL, Hickson ID (2011) New insights 13:632–645
into the formation and resolution of ultra-fine 19. Broderick R, Nieminuszczy J, Blackford AN
anaphase bridges. Semin Cell Dev Biol et al (2015) TOPBP1 recruits TOP2A to
22:906–912 ultra-fine anaphase bridges to aid in their reso-
7. Baumann C, Körner R, Hofmann K et al lution. Nat Commun 6:6572
(2007) PICH, a centromere-associated SNF2 20. Pedersen RT, Kruse T, Nilsson J et al (2015)
family ATPase, is regulated by Plk1 and TopBP1 is required at mitosis to reduce trans-
required for the spindle checkpoint. Cell mission of DNA damage to G1 daughter cells.
128:101–114 J Cell Biol 210:565–582
8. Chan KL, North PS, Hickson ID (2007) BLM 21. Naim V, Rosselli F (2009) The FANC pathway
is required for faithful chromosome segrega- and BLM collaborate during mitosis to prevent
tion and its localization defines a class of ultra- micro-nucleation and chromosome abnormal-
fine anaphase bridges. EMBO J 26:3397–3409 ities. Nat Cell Biol 11:761–768
9. Kaulich M, Cubizolles F, Nigg EA (2012) On 22. Vinciguerra P, Godinho SA, Parmar K et al
the regulation, function, and localization of the (2010) Cytokinesis failure occurs in Fanconi
DNA-dependent ATPase PICH. Chromosoma anemia pathway-deficient murine and human
121:395–408 bone marrow hematopoietic cells. J Clin Invest
10. Dı́az-Martı́nez LA, Giménez-Abián JF, Clarke 120:3834–3842
DJ (2008) Chromosome cohesion - rings, 23. Botstein D, Fink GR (2011) Yeast: an experi-
knots, orcs and fellowship. J Cell Sci mental organism for 21st Century biology.
121:2107–2114 Genetics 189:695–704
11. Holm C, Goto T, Wang JC et al (1985) DNA 24. Norden C, Mendoza M, Dobbelaere J et al
topoisomerase II is required at the time of (2006) The NoCut pathway links completion
mitosis in yeast. Cell 41:553–563 of cytokinesis to spindle midzone function to
12. Holm C, Stearns T, Botstein D (1989) DNA prevent chromosome breakage. Cell
topoisomerase II must act at mitosis to prevent 125:85–98
nondisjunction and chromosome breakage. 25. Mendoza M, Norden C, Durrer K et al (2009)
Mol Cell Biol 9:159–168 A mechanism for chromosome segregation
13. Strunnikov AV, Hogan E, Koshland D (1995) sensing by the NoCut checkpoint. Nat Cell
SMC2, a Saccharomyces cerevisiae gene essen- Biol 11:477–483
tial for chromosome segregation and conden- 26. Sherman F (2002) Getting started with yeast.
sation, defines a subgroup within the SMC Methods Enzymol 350:3–41
family. Genes Dev 9:587–599 27. Hartwell LH, Mortimer RK, Culotti J et al
14. Uhlmann F, Lottspeich F, Nasmyth K (1999) (1973) Genetic control of the cell division
Sister-chromatid separation at anaphase onset cycle in yeast: V. Genetic analysis of cdc
is promoted by cleavage of the cohesin subunit mutants. Genetics 74:267–286
Scc1. Nature 400:37–42 28. Roccuzzo M, Visintin C, Tili F et al (2015)
15. Germann SM, Schramke V, Pedersen RT et al FEAR-mediated activation of Cdc14 is the lim-
(2014) TopBP1/Dpb11 binds DNA anaphase iting step for spindle elongation and anaphase
progression. Nat Cell Biol 17(3):251–261
 Imaging of UFBs in Yeast 493

29. Jungbluth M, Renicke C, Taxis C (2010) Tar- green fluorescent protein. Science
geted protein depletion in Saccharomyces cer- 273:1392–1395
evisiae by activation of a bidirectional degron. 34. Keppler-Ross S, Noffz C, Dean N (2008) A
BMC Syst Biol 4:176 new purple fluorescent color marker for genetic
30. Bishop AC, Ubersax JA, Petsch DT et al studies in Saccharomyces cerevisiae and Can-
(2000) A chemical switch for inhibitor- dida albicans. Genetics 179:705–710
sensitive alleles of any protein kinase. Nature 35. Silva S, Gallina I, Eckert-Boulet N et al (2012)
407:395–401 Live cell microscopy of DNA damage response
31. D’Aquino KE, Monje-Casas F, Paulson J et al in Saccharomyces cerevisiae. Methods Mol Biol
(2005) The protein kinase Kin4 inhibits exit 920:433–443
from mitosis in response to spindle position 36. Viggiani CJ, Aparicio OM (2006) New vectors
defects. Mol Cell 19:223–234 for simplified construction of BrdU-
32. Heim R, Tsien RY (1996) Engineering green Incorporating strains of Saccharomyces cerevi-
fluorescent protein for improved brightness, siae. Yeast 23:1045–1051
longer wavelengths and fluorescence resonance 37. Quevedo O, Garcı́a-Luis J, Matos-Perdomo E
energy transfer. Curr Biol 6:178–182 et al (2012) Nondisjunction of a single chro-
33. Ormö M, Cubitt AB, Kallio K et al (1996) mosome leads to breakage and activation of
Crystal structure of the Aequorea victoria DNA damage checkpoint in G2. PLoS Genet
8:e1002509
 Chapter 33

Detection of Ultrafine Anaphase Bridges

Anna H. Bizard, Christian F. Nielsen, and Ian D. Hickson

Abstract
Ultrafine anaphase bridges (UFBs) are thin DNA threads linking the separating sister chromatids in the
anaphase of mitosis. UFBs are thought to form when topological DNA entanglements between two
chromatids are not resolved prior to anaphase onset. In contrast to other markers of defective chromosome
segregation, UFBs cannot be detected by direct staining of the DNA, but instead can be detected using
immunofluorescence-based approaches. Due to the fact that they are short-lived and fragile in nature, UFBs
can be challenging to detect. In this chapter, we describe methods that have been optimized for successful
detection of UFBs. We also provide guidelines for the optimization of UFBs detection depending on the
antibody and the cell line to be used.

Key words Ultrafine anaphase bridges, Common fragile sites, Centromeres, Telomeres, Genome
stability, Chromosome segregation

1 Introduction

The removal of all DNA entanglement between the sister chroma-

tids is a prerequisite for the accurate segregation of the daughter
genomes in mitosis. Nevertheless, certain genomic loci can remain
entangled at the onset of anaphase. In anaphase, such persistent
DNA entanglements are converted into ultra fine DNA threads
bridging the two segregating genomes. These structures are the
so-called “ultrafine anaphase bridges” (UFBs). In contrast to other
segregation defects, such as lagging chromatin or chromatin
bridges, UFBs cannot be stained by conventional DNA dyes.
Instead, they can be detected by immunofluorescence, using anti-
bodies raised against UFB-associated factors such as the SNF2-
family DNA translocase, PICH, and the RecQ DNA helicase
BLM [1, 2].
Several subclasses of UFBs have been identified. These are
defined by the genomic locus from which they originate. The
most common are the centromeric-UFBs (CEN-UFBs) arising
due to persistent catenation of the centromeric chromatin.

Marco Muzi-Falconi and Grant W. Brown (eds.), Genome Instability: Methods and Protocols, Methods in Molecular Biology,
vol. 1672, DOI 10.1007/978-1-4939-7306-4_33, © Springer Science+Business Media LLC 2018

495
496 Anna H. Bizard et al.

CEN-UFBs are abundant at anaphase onset but are rapidly resolved

by topoisomerase IIα during anaphase. Hence, their number rap-
idly drops as the cells progress through anaphase, unless cells are
treated with Topoisomerase II inhibitor (such as ICRF-193) [2–5].
A second source of UFBs is from common fragile sites (CFS). In
contrast to CEN-UFBs, CFS-UFBs are quite rare in an unper-
turbed cell cycle, but can be specifically induced by exposing the
cells to DNA replication stress [6]. Besides centromeres and CFSs,
other critical regions of the genomes that have been shown to be
prone to the formation of UFBs are telomeres and the ribosomal-
DNA loci [7, 8].
There is some evidence to suggest that UFBs that are not
resolved during mitosis can cause the cell to abort cytokinesis,
potentially generating aneuploidy [5]. Hence, UFBs might play a
role in human diseases associated with genome instability [9, 10].
Despite this, the analysis of UFBs has been hampered by a lack of
standardized methods for their detection. It is noteworthy that
UFBs can be challenging to visualize when routine immunofluores-
cence protocols are used. This is due largely to three main issues.
First, UFBs are short-lived and can only be observed during the
anaphase and telophase of mitosis, which each represent only a very
short stage of the cell division cycle. Second, UFBs are relatively
fragile, and can be broken if the cells are not handled carefully prior
to, during and even after being fixed. Finally, UFBs are very thin,
and hence the intensity of immunostaining can easily drop below
the level of detection for a conventional fluorescence microscope.
Accordingly, successful detection of UFBs relies on classical
immunofluorescence protocols, in which some critical steps are
executed in a way that preserves the full range of structures present
in an anaphase cell. Depending on the cell line and the antibodies
used to detect UFBs, we routinely use three different permeabiliza-
tion procedures that differ from one another by the timing of the
application of Triton to the cells (i.e., before, during, or after the
fixation step for pre-permeabilization, co-permeabilization, and
post-permeabilization procedures, respectively). The importance
of the permeabilization procedure choice is exemplified in Fig. 1,
which shows representative images of UFBs obtained with antibo-
dies directed against PICH and BLM. While the anti-PICH anti-
body allows good detection of UFBs independently of the fixation
procedure used, detection using the anti-BLM antibody requires
either co-permeabilized or pre-permeabilized fixed cells. It is note-
worthy that the influence of the permeabilization protocol on the
quality of the detected UFBs also greatly depends on the cell line
studied. In this chapter, we describe in greatest detail the pre-
permeabilized protocol, which is the most widely used and sensi-
tive. The co-permeabilization and post-permeabilization protocols
are also described as alternative protocols (see Note 1).
 Anaphase Bridges 497

PC DAPI PICH BLM Merged

POST
Extraction

CO
PRE

Fig. 1 Representative images of UFBs detected using the anti-PICH (green) or anti-BLM (red) antibody, in
HT1080 cells fixed with the post-permeabilization (top), co-permeabilization (middle), or pre-permeabilization
(bottom) procedures. DNA was stained with DAPI (blue). PC shows the phase contrast images of each
anaphase B cell. Images were taken with a conventional wide field fluorescence microscope. Scale bar;
10 μm

Finally, while UFBs can be detected in virtually all cell lines, it is

noteworthy that some cell lines are more suitable for UFB analysis
than others. For routine UFB analysis, it is, therefore, advisable to
perform the analysis with adherent cell lines that can be grown
directly on glass coverslips, such as HT1080, U2OS, or fibroblasts
such as GMO00637. Nevertheless, it is possible to successfully
detect UFBs when using suspension cell lines. In this case, cells
should be seeded onto a poly-lysine-coated surface immediately
before the fixation (see Note 3).

2 Materials

2.1 Slides 1. Sterile coverslips 22  22 mm (Thickness No. 1.5).

2. Microscope slides 75  25  1 mm.
3. For humidified chambers: Plastic petri dishes; Whatman 3MM,
and Parafilm discs at the dimensions of the petri dishes.

2.2 Buffers 1. 20 Phosphate Buffer Saline (20 PBS): 2.8 M NaCl, 54 mM
and Solutions KCl, 130 mM Na2HPO4, and 30 mM KH2PO4, pH adjusted
to 7.4 with 1 M HCl. Filter and store and room temperature.
2.2.1 Stock Solutions
2. 37 wt. % paraformaldehyde (PFA) solution (stabilized with
10–15% methanol) (Sigma, 25,2549). Store at room tempera-
ture protected from light.
498 Anna H. Bizard et al.

3. 10 PEM Buffer: 200 mM PIPES pH 6.8, 10 mM MgCl2, and

100 mM EGTA. Filter and store at room temperature.
4. 10% Triton X-100, diluted in ddH2O. Store at room
temperature.
5. 30% (w.v) bovine serum albumin (BSA), in ddH2O. Filter and
store and 4  C.
6. Vectashield mounting medium for fluorescence, (without
DAPI) (Vectors Lab, H-1000).
7. DAPI (1 mg/ml in ddH2O). Make small aliquots and store at
20  C protected from light.
8. Transparent nail polish.

2.2.2 Working Solutions 9. Pre-permeabilization Buffer A: 0.2% Triton in 1 PEM Buffer.

(to Be Prepared for Prepare 1 ml per coverslip.
Immediate Use and 10. Pre-permeabilization Buffer B: 0.1% Triton, 8% PFA in 1
Depending on the Chosen PEM Buffer). Prepare 2 ml per coverslip.
Permeabilization
11. Co-permeabilization Buffer: 0.4% Triton, 8% PFA, in 1
Procedure)
PEM). Prepare 1 ml per coverslip.
12. Post-permeabilization Buffer: 8% PFA, in 1 PBS). Prepare
1 ml per coverslip.
13. Post-staining Fixation Buffer: 4% PFA in 1 PBS). Prepare
200 μl per coverslip.

2.2.3 Other Solutions 14. 1 PBS (1/20 dilution of the stock; in ddH2O). Prepare 20 ml
(Can Be Prepared in per coverslip. Store at room temperature.
Advance) 15. PBSAT Buffer: 3% BSA, 0.5% Triton X-100 in 1 PBS). Filter,
aliquot in 50 ml Falcon tubes and store at 20  C.
16. 1 DAPI (1/10.000 dilution of the stock; in 1 PBS). Store
at 4  C protected from light.

2.3 Antibodies 1. Mouse monoclonal anti PICH antibody (clone 14226-3; Milli-
pore, 0.41540). Store a 4  C. Recommended working dilution:
1/50 in PBSAT.
2. Rabbit polyclonal anti BLM antibody (Abcam, ab2175). Ali-
quot and store at 20  C. Recommended working dilution: 1/
200 in PBSAT.
3. Alexa Fluor® 488 goat anti-mouse IgG (Life Technologies,
A11001). Store at 4  C protected from light. Recommended
working dilution: 1/500 in PBSAT.
4. Alexa Fluor® 568 goat anti-rabbit IgG (Life Technologies,
A11001). Store at 4  C protected from light. Recommended
working dilution: 1/500 in PBSAT.
 Anaphase Bridges 499

3 Methods

3.1 Cell Culture for The protocol described below has been optimized for an asynchro-
Adherent Cell Lines nously growing population of HT1080 cells (see Note 2). Some
adjustments might be required if other adherent cell lines were to
be used. If suspension cells are being used, see Note 3.
The following protocol describes the procedure for preparing
six coverslips to be fixed simultaneously.
DAY 1
1. Place a sterile coverslip in each well of a 6-well plate (see
Note 4). Add 1 ml of fresh medium to the wells. Ensure that
the medium covers the coverslips, and that the coverslips are
well seated on the bottom of each well by applying a little
pressure on each coverslip using a pipette tip or sterile tweezers.
2. From an approximately 80% confluent, T75 flask of a healthy
growing population of cells, aspirate the medium, wash the
cells once with PBS, add 2.5–3 ml of trypsin and incubate for
5 min at 37  C. Once the cells have detached, neutralize the
trypsin by adding 5 ml of complete medium to the flask.
Transfer the cells to a 15 ml Falcon tube and harvest the cells
by centrifugation at 300  g for 5 min at room temperature.
3. Aspirate the supernatant and resuspend the cell pellet in 1 ml of
fresh medium using a Gilson P1000 pipette (or similar).
Pipette the suspension up and down several times, with the
tip touching the bottom of the tube in order to dislodge cell
aggregates (see Note 5).
4. Add culture medium in order to reach a cell density of
0.3  106 cells/ml (see Note 6).
5. Distribute 1 ml of cells to each well of the 6-well plate.
6. Homogenize the cell suspension by pipetting up and down
once, and immediately place the plate in the incubator.
7. Let the cells adhere to the surface and proliferate for 40–48 h
until they reach approximately 80–90% confluence (see Note 7).

3.2 Cell For the co-permeabilization and post-permeabilization protocols,

Permeabilization see Notes 8 and 9 respectively.
and Fixation DAY 3
(Pre-permeabilization Unless otherwise stated, each subsequent step is performed at room
Protocol) temperature (see Note 10).
1. Using a conventional inverted light microscope, verify that the
culture has reached the expected confluence (see Note 11) and
that the mitotic cells can be observed at a frequency compatible
with the subsequent analysis (see Note 12).
500 Anna H. Bizard et al.

2. Prepare sufficient quantity of pre-permeabilization Buffer A

and pre-permeabilization Buffer B as described in Subheading
2.2, items 9 and 10.
3. Set up all reagents required for the permeabilization and the
fixation of cells in a fume hood (Buffers prepared in Subhead-
ing 3.2, step 2; PBS 1; and appropriate liquid waste disposals)
(see Note 13).
4. Take the 6-well plate out of the incubator and place in a fume
hood. Avoid agitation of the cells as much as possible (see Note
14). Once the plate has been placed in the hood, it should not
be moved until the end of the fixation procedure. Proceed
immediately to fixation.
5. Using a P1000, remove 1 ml of medium from the side of each
well. The coverslips are now covered by only 1 ml of medium
(see Note 15).
6. Add 1 ml of PBS down one side of each well. Immediately take
out 1 ml of medium/PBS mixture from the opposite side of
each well (see Note 16).
7. Repeat step 6 one more time. The cells are now immersed in
1 ml of PBS (see Note 17).
8. Proceed to pre-permeabilization by adding 1 ml of pre-per-
meabilization Buffer A down the side of each well.
9. Carefully mix the buffer in each well by transferring 1 ml from
one side to the opposite side of each well. At this stage, the
pipetting needs to be quick, but gentle, smooth and regular.
Expect this procedure to require approximately 30 s for the
entire plate.
10. Incubate for 60 s at room temperature (see Note 18).
11. Add 2 ml of pre-permeabilization Buffer B to each well. Main-
tain the same frequency and order as in step 8 to ensure that
each well is being pre-permeabilized for approximately the
same amount of time.
12. Incubate for 15 min at room temperature. Do not move the
plate until fixation is over.
13. Once the cells are fixed, discard the 4 ml from each well into a
dedicated chemical waste container. Rinse the coverslips once
with 4 ml PBS.
14. Wash the coverslips further with 2 ml PBS for 5 min. Repeat
three times.
15. Further permeabilize and saturate the coverslips by incubating
them overnight at 4  C in 1 ml sterile PBSAT (see Note 19).
The coverslips can be kept in PBSAT at 4  C for up to 1 week
before being stained (see Note 20).
 Anaphase Bridges 501

Fig. 2 Phase contrast images of asynchronously growing HT1080 cells fixed with
the post-permeabilization (top left), co-permeabilization (top right), or pre-per-
meabilization (bottom left) procedures. An example of cells fixed following
excessive pre-permeabilization is also shown (bottom right). In each panel,
arrowheads indicate cells in anaphase B. Images were taken with a conventional
wide field microscope. Scale bar; 50 μm

16. Using a conventional inverted light microscope, verify that the

cells have been fixed properly (see Fig. 2 and Note 21). Be
particularly rigorous with this step when the pre-permeabilized
procedure has been used (see Notes 22 and 23).

3.3 Staining DAY 4

1. Prepare a humidity chamber by placing a wet disc of Whatman
3MM paper, covered by a disc of Parafilm into a petri dish (see
Note 24). Ensure that the surface of the Parafilm is flat.
2. Place the coverslip onto the Parafilm with the cells facing up (see
Note 25). Immediately cover the cells with 200 μl PBSAT (see
Note 26).
502 Anna H. Bizard et al.

3. Working on ice, dilute the primary antibody in ice cold PBSAT

(100 μl per coverslip). Centrifuge the diluted antibody at
5000  g for 2 min at 4  C.
4. Remove as much as possible of the PBSAT from the coverslip
and then immediately cover it with 75–100 μl of the diluted
antibody. Ensure that the antibody solution spreads evenly over
the entire surface of the coverslip (see Note 27).
5. Close and seal the petri dish with Parafilm. Incubate overnight
at 4  C.

DAY 5
6. After the primary antibody incubation is completed, wash the
coverslip with 200 μl PBSAT for 10 min. Repeat the washing
step twice more.
7. While waiting for the completion of the last wash step, dilute
the secondary antibody in ice cold PBSAT (100 μl per cover-
slip). This procedure should be performed on ice and the
solution protected from light wherever possible. Centrifuge
the diluted antibody at 5000  g for 5 min at 4  C.
8. Once the last wash is completed, remove as much as possible of
the PBSAT from the coverslip and immediately cover it with
75–100 μl of the diluted secondary antibody. Ensure that the
antibody solution spreads over the entire surface of the cover-
slip (see Note 27).
9. Incubate for 2 h at room temperature and protect from light.
10. After the secondary antibody incubation is completed, wash
the coverslip with 200 μl PBSAT for 15 min. Repeat the wash
step three times.
11. Perform one more wash with PBS for 10 min. Eventually
proceed with post-staining fixation (see Note 28).
12. Cover the coverslip with 200 μl of DAPI solution that is
diluted 1/10,000 in PBS.
13. Incubate for 5 min at room temperature and protect from
light.
14. Wash twice with PBS for 5 min.
15. Label the slides and apply a drop (approx. 10 μl) of Vectashield
mounting medium.
16. Once the washes are completed, remove the PBS from the
coverslip and carefully rinse the coverslip with 200 μl of
ddH2O.
17. Drain as much of the liquid as possible from the coverslip and,
using tweezers, carefully place the coverslip onto the drop of
mounting medium with the cells facing down.
 Anaphase Bridges 503

18. Place the slide upside down on the piece of tissue paper, and
allow any excess mounting medium to drain away for
15–30 min (the coverslip is now sandwiched between the tissue
paper and the slide).
19. Seal the coverslip to the slide by spreading nail polish around
the edges of the coverslip.
20. The slides must be kept at 4  C and protected from the light
until analysis (see Note 29).
21. Using a conventional fluorescence microscope, verify the qual-
ity of the staining, particularly of mitotic cells. Besides the
presence of UFBs (see Fig. 1), the staining can be validated by
a rapid analysis of characteristic staining pattern of the used
antibody (see Note 30 for PICH, and Note 31 for BLM).
22. Proceed to the analysis of UFBs, by counting the number of
UFBs per anaphase B cell (see Note 32). Results are typically
expressed as the proportion of anaphase B with at least 1 UFBs
and as the average number of UFBs per anaphase B (see Note 33).

4 Notes

1. In order to maximize the chance of successful UFB detection,

each time a new combination of cell line and antibody is to be
used, it is advisable to test all three of the fixation protocols.
Fixation protocol optimization: When using a new combina-
tion of cells and antibodies, proceed to a test experiment, in
which the cells are fixed using each of the three protocols, and
then stained with the desired antibodies. Assess the quality of
the fixation and of the staining for each fixation procedure. If
two protocols show an equivalent quality of staining, always
favor post-permeabilization and co-permeabilization over the
pre-permeabilization protocol because these protocols are
more reproducible, and preserve better the different structures.
2. The frequency of anaphases present in an asynchronously
growing population of cells is more than enough to allow
quantitative UFBs analysis. Nevertheless, it is sometimes pref-
erable to analyze a population of cells synchronously progres-
sing throughout mitosis after release from a nocodazole block.
The synchronization protocol needs to be optimized according
to the cell line. Briefly, after the cells have been arrested in
prometaphase in the presence of nocodazole, the mitotic cells
are detached from the surface by tapping the side of the flask
with the palm of the hand three or four times (generally termed
mitotic shake-off). The detached cells are then washed twice in
warm PBS. The cells are then resuspended in prewarmed
504 Anna H. Bizard et al.

medium and seeded on poly-lysine-coated coverslips as

described for suspension cells (see Note 3).
3. When using cells that grow in suspension, proceed as fol-
lows: on the day of the fixation place a poly-lysine-coated
coverslip into each well of a 6-well plate; add 1 ml of medium
to each well and ensure that the coverslips are well seated on the
bottom of the wells. Distribute 1 ml of cell suspension into
each well (approx. 0.5  106 per well) and let them settle for
15–30 min. Proceed to “fixation and staining” as described in
Subheadings 3.2 and 3.3.
4. Coverslips are usually supplied pre-sterilized, and therefore do
not require sterilization—as long as the box has not been
opened in a non-sterile environment. If the sterility of the
coverslips has been compromised, transfer them into a glass
petri dish, and microwave them for 15 s at maximum intensity.
5. This step ensures that any cells clumps or aggregates are dis-
persed, and allows the formation of a homogeneous layer of
cells on the surface of the coverslip.
6. Optimize according to the doubling time of the cell line to be
used. We generally aim to allow the cells to go through two
rounds of division between the seeding and the fixation steps
(see Note 5).
7. It is important that the cells are actively proliferating at the time
of fixation. A high cell density is important to maximize the
number of mitotic events that can be scored per sample.
8. Alternative protocol: Co-permeabilization. Use the pre-per-
meabilization protocol through Subheading 3.2, step 7, and
then add 1 ml of co-permeabilization Buffer down the side of
the well. Carefully mix by transferring 1 ml from one side of the
well to the other, and incubate for 15 min. Proceed to the wash
steps described in the main protocol (from Subheading 3.2,
step 13).
9. Alternative protocol: Post-permeabilization. Use the pre-
permeabilization protocol through Subheading 3.2, step 5,
and then fix the cells by adding 1 ml of post-permeabilization
Buffer. Incubate for 15 min at room temperature and then
proceed to the steps described in the main protocol (from
Subheading 3.2, step 13).
10. Be aware that the pre-permeabilization and co-permeabiliza-
tion efficiency can be affected by even small variations of tem-
perature. Keep the samples at 20–25  C wherever possible. It is
possible to conduct the fixation at 4  C, but this can under
some circumstances induce microtubule depolymerization and
lead to the release of the tension holding the UFBs.
 Anaphase Bridges 505

11. The cells should be 80–90% confluent. If the confluence is less

than 50%, incubate the culture further until it reaches a good
confluence (see Fig. 2, for examples of expected confluence). If
the cells are overgrown, it is preferable to start again from
Subheading 3.1, step 1.
12. Depending on the cell type, expect approximately 5% of the
cells to be mitotic (see Fig. 2, for examples of expected mitotic
density). A high mitotic index not only maximizes the number
of events that can be scored per sample, but also is a good
indicator of the “healthiness” of the cells.
13. The success of this procedure essentially relies on accurate
timing and smooth manipulation of the cells. Therefore, it is
important to organize properly the fume hood and to ensure
that all materials and buffers are carefully organized and are
easily accessible.
14. It is essential to keep in mind that mitotic cells are poorly
adhesive to the surface and that UFBs are short-lived and
fragile in nature. Hence, the coverslips must be processed as
gently as possible, before, during, and after the fixation
process.
15. In order to minimize the perturbations applied on the cells
while they are being washed and fixed, always add buffer from
the same side of the well, and always remove buffer from the
opposite side of the well. Always proceed with the different
wells of the plate in the same order. By using a Gilson P1000
for this and the following steps helps to maintain a regular
rhythm during periods when medium is added or removed.
16. During this and the following steps, the fact that the cells
remain constantly immersed in solution is critical in order to
avoid the detachment of mitotic cells from the coverslip.
17. The fraction of remaining media is negligible for the following
steps.
18. The incubation time might need to be adjusted depending on
the cell type used. It usually varies between 30 s and 2 min.
19. As an alternative, the coverslips can be saturated in PBSAT for
2 h at room temperature.
20. For longer storage, keep coverslips in PBS with 0.02% sodium
azide. Seal the plate with Parafilm to prevent evaporation.
21. Regardless of the procedure used, verify that the mitotic cells
have not been detached from the coverslips during the fixation
(see Fig. 2 for examples of cells after each different fixation
procedure). The number of mitotic cells visible at this stage
should be similar to what was observed prior to fixation.
Moreover, the mitotic cells should appear perfectly fixed to
the coverslips. If the mitotic cells are not static upon moving
506 Anna H. Bizard et al.

the plate, it is likely that the procedure has not been performed
gently and/or quickly enough. These mitotic cells will be
particularly difficult to image. It is recommended to repeat
the fixation. Moving mitotic cells can also suggest a too exten-
sive pre-permeabilization (see Note 23).
22. The extent to which the cells have successfully been pre-per-
meabilized is critical for UFBs analysis. Typically, the cytoplasm
of interphase cells should be still visible, while the contours of
mitotic cells should be clearly marked (see Fig. 2 for examples).
If the pre-permeabilization step is too extensive, the staining
of some of the bridges will be too weak to be detected (see
Note 23).
23. If there are indications that the pre-permeabilization is too
extensive, it is preferable to repeat the fixation with a modified
pre-permeabilization step. It is most likely that a reduction in
the incubation time (Subheading 3.2, step 10) will improve
pre-permeabilization. Otherwise, one might consider, either to
decrease the concentration of Triton in the pre-permeabiliza-
tion buffer, or to select one of the other fixation procedures.
24. We usually can place two and four coverslips, respectively, into
each 10 and 15 cm petri dish. Careful labeling of the samples is
essential in this case.
25. During the staining part of the protocol the manipulation of
the coverslip is minimized, in order to avoid mitotic cells being
detached from the coverslips or UFBs from being broken or
distorted.
26. It is crucial to prevent the coverslips from drying out, as this
will irreversibly affect the quality of the staining. Hence, the
time during which the coverslips are not covered by buffer
should be minimized.
27. At this stage, if the antibody does not spread evenly over the
entire surface of the coverslip, it might indicate that the cover-
slip has dried out from the previous step. The quality of the
staining will be irreversibly impaired on any dried region of the
coverslip.
28. Post-staining fixation. At this stage, proceeding to a second
fixation step can help to increase the stability of the slides over
time. For this, remove as much PBS as possible from the
coverslip and cover the cells with 200 μl post-staining fixation
Buffer. Incubate for 5 min at room temperature and wash three
times for 5 min with PBS. Proceed to DAPI staining as indi-
cated in the main protocol (Subheading 3.3, step 12).
29. Normally, the slides can be kept up to 1 month at 4  C. It is
highly recommended, however, to proceed to the analysis as
soon as possible, as the quality of the staining may decline
 Anaphase Bridges 507

overtime. In order to increase the “working-life” of the stored

slides, it is advisable to perform a post-staining fixation (see
Note 26).
30. Besides its localization on UFBs, PICH localizes to the cyto-
plasm of interphases cells, and to the centromeres of mitotic
cells [2, 11]. Typically, staining using the antibody directed
against PICH is relatively clean but is often quite faint (see
Fig. 1).
31. Besides this localization to UFBs, BLM localizes at PML bod-
ies in a subset of interphase cells [12]. The antibody directed
against BLM overall gives more intense staining than does the
antibody directed against PICH. However, the signal from the
soluble fraction of the cells can interfere with the proper visu-
alization of BLM on UFBs. For this reason, it is recommended
to proceed with the co-permeabilized or the pre-permeabilized
procedure when using this antibody (see Fig. 1).
32. It is important to restrict analysis to the anaphase B stage of
mitosis. Otherwise, UFBs will be over-represented from the
very early stage of anaphase due to the presence of centromeric
UFBs. Anaphase B can be distinguished from anaphase A by
the distance between the segregating DNA masses, as well as by
the general form of the cell (ovoid more than spherical). How-
ever, it is also important to make the distinction between
anaphase B and telophase (during which the DNA is partially
decondensed and the daughter cells are clearly visible) (see
Figs. 1 and 2 for examples of anaphase B cells).
33. In an untreated, asynchronously growing population of cells,
expect between 5 and 20% of the anaphase B cells to contain at
least 1 UFB.

Acknowledgments

Work in the authors’ laboratory is supported by The European

Research Council and The Danish National Research Foundation
(grant DNRF115).

References
1. Chan KL, North PS, Hickson ID (2007) BLM required for the spindle checkpoint. Cell 128
is required for faithful chromosome segrega- (1):101–114. doi:10.1016/j.cell.2006.11.041
tion and its localization defines a class of ultra- 3. Wang LH, Schwarzbraun T, Speicher MR,
fine anaphase bridges. EMBO J 26 Nigg EA (2008) Persistence of DNA threads
(14):3397–3409. doi:10.1038/sj.emboj. in human anaphase cells suggests late comple-
7601777 tion of sister chromatid decatenation. Chro-
2. Baumann C, Korner R, Hofmann K, Nigg EA mosoma 117(2):123–135. doi:10.1007/
(2007) PICH, a centromere-associated SNF2 s00412-007-0131-7
family ATPase, is regulated by Plk1 and
508 Anna H. Bizard et al.

4. Hengeveld RC, de Boer HR, Schoonen PM, de through processing of late-replicating interme-
Vries EG, Lens SM, van Vugt MA (2015) Rif1 diate structures. Nucleic Acids Res 40
is required for resolution of ultrafine DNA (15):7358–7367. doi:10.1093/nar/gks407
bridges in anaphase to ensure genomic stability. 9. Vinciguerra P, Godinho SA, Parmar K, Pellman
Dev Cell 34(4):466–474. doi:10.1016/j. D, D’Andrea AD (2010) Cytokinesis failure
devcel.2015.06.014 occurs in Fanconi anemia pathway-deficient
5. Nielsen CF, Huttner D, Bizard AH, Hirano S, murine and human bone marrow hematopoie-
Li TN, Palmai-Pallag T, Bjerregaard VA, Liu Y, tic cells. J Clin Invest 120(11):3834–3842.
Nigg EA, Wang LH, Hickson ID (2015) doi:10.1172/JCI43391
PICH promotes sister chromatid disjunction 10. Nera B, Huang HS, Lai T, Xu L (2015) Ele-
and co-operates with topoisomerase II in mito- vated levels of TRF2 induce telomeric ultrafine
sis. Nat Commun 6:8962. doi:10.1038/ anaphase bridges and rapid telomere deletions.
ncomms9962 Nat Commun 6:10132. doi:10.1038/
6. Chan KL, Palmai-Pallag T, Ying S, Hickson ID ncomms10132
(2009) Replication stress induces sister- 11. Kaulich M, Cubizolles F, Nigg EA (2012) On
chromatid bridging at fragile site loci in mito- the regulation, function, and localization of the
sis. Nat Cell Biol 11(6):753–760. doi:10. DNA-dependent ATPase PICH. Chromosoma
1038/ncb1882 121(4):395–408. doi:10.1007/s00412-012-
7. Nielsen CF, Hickson ID (2016) PICH pro- 0370-0
motes mitotic chromsome segregation: identi- 12. Yankiwski V, Marciniak RA, Guarente L, Neff
fication of a novel role in ribosomal DNA NF (2000) Nuclear structure in normal and
disjunction. Cell Cycle 15(20):2704–2711 Bloom syndrome cells. Proc Natl Acad Sci U
8. Barefield C, Karlseder J (2012) The BLM heli- S A 97(10):5214–5219. doi:10.1073/pnas.
case contributes to telomere maintenance 090525897
 Chapter 34

A Chromatin Fiber Analysis Pipeline to Model DNA Synthesis

and Structures in Fission Yeast
Sarah A. Sabatinos and Marc D. Green

Abstract
Chromatin fibers, first described by Jackson and Pombo (J Cell Biol 140(6):1285–1295, 1998) are
prepared from cells lysed on glass coverslips, and require minimal equipment to produce. Since the DNA
is not previously treated with denaturing agents, proteins are left intact and may be used to model other
DNA-based processes. Such an analysis can be daunting, without a rigorous method for analysis. We
describe a pipeline for chromatin fiber use to model DNA replication complexes. Full protocols for
chromatin fiber preparation and staining are presented. Further, we have developed an analysis algorithm
for One Dimensional Data—Boolean Logic Operations Binning System (ODD-BLOBS). This freely
available software defines replication and protein tracts, measures their lengths, and then correlates repli-
cated areas with protein distributions. Our methods and analysis are tested in Schizosaccharomyces pombe
(fission yeast) but may be applied to model replication structures across multiple organisms.

Key words Chromatin fibers, Replication tracts, Fission yeast, Fiber spreads, Replication tract analysis

1 Introduction

Schizosaccharomyces pombe is an excellent model organism to study

replication dynamics and complexes, but high-resolution imaging is
difficult in yeast nuclei due to their small size. In general, replication
complex studies are challenging in many model organisms due to
compaction of DNA within nuclei. By releasing DNA from nuclear
confines into chromatin spreads or fibers, individual replicated areas
can be assessed with better resolution. DNA fibers, also called
combed, extended or stretched DNA, have been extensively used
to examine replication dynamics such as origin firing and fork
stability (e.g., [1, 2]). Methods to prepare DNA fibers typically
purify and isolate DNA free of proteins, and then extend the fibers
along treated glass surfaces to produce uniformly aligned, straight
and stretched DNA for analysis.
Chromatin fibers, also called spread-fibers or fiber-spreads, are
a related but distinct technology from DNA fibers. As with DNA

Marco Muzi-Falconi and Grant W. Brown (eds.), Genome Instability: Methods and Protocols, Methods in Molecular Biology,
vol. 1672, DOI 10.1007/978-1-4939-7306-4_34, © Springer Science+Business Media LLC 2018

509
510 Sarah A. Sabatinos and Marc D. Green

fibers, chromatin fibers can be used to detect DNA synthesis by

adding nucleoside analogs such as 5-bromo-20 -deoxyuridine
(BrdU). Distinct from DNA fibers, chromatin fibers retain protein
components and epigenetic domains. Chromatin fibers do not
require specialized equipment or extensive preparation. They are
prepared by lysing whole cells on a slide and allowing the DNA to
extend along a glass surface. They retain considerable detail that can
be imaged using high-resolution microscopy. Chromatin fibers
have been used in multiple organisms to examine origin firing and
replication fork progression by examining synthesis alone (e.g., [3,
4]). By adding the dimension of protein detection, chromatin fibers
have been used to characterize chromatin domains (e.g., [5–7]),
histone modifications relative to DNA synthesis (e.g., [8]), and
replication fork components (e.g., [9]).
We describe protocols to prepare chromatin fibers from S.
pombe, and then detect proteins and study the data with a new
analysis algorithm. The basic chromatin fiber method uses cells
that have been blocked in early S-phase, and then released into
media containing a nucleoside analog to complete replication.
Two-color labeling during replication block and release is possible
using nucleoside pairs of either 5-iodo-20 -deoxyuridine (IdU) and
5-chloro-20 -deoxyuridine (CldU) (e.g., [10, 11]), or 5-bromo-
20 deoxyuridine (BrdU) and 5-ethynyl-20 -deoxyuridine (EdU) (e.
g., [4, 12]). Nucleoside analogs can then be detected by immuno-
fluorescence (IdU, CldU, and BrdU) or click chemistry (EdU) to
visualize regions of DNA replication on the spread fibers. Proteins
are detected using immunofluorescence and their spatial arrange-
ment can be compared to nucleoside analog incorporation,
providing a unique opportunity to assess replication factor distri-
butions. Further, fluorescence in situ hybridization (FISH) (e.g.,
[13, 14]) can be used to compare protein localization on chromatin
fibers relative to specific DNA loci. Together, these methods
expand the analysis of DNA replication to consider the size, com-
position and distribution of replication forks and factors at synthe-
sized and/or specific areas of a genome.
We also describe an analysis algorithm to systematically corre-
late DNA synthesis with protein location on DNA fibers. Named
“One Dimensional Data—Boolean Logic Operations Binning Sys-
tem” (ODD-BLOBS), our analysis algorithm models fiber data as
linear distributions, investigating replicated areas and the regions
surrounding the ends of replicated tracts. The ends of replicated
tracts may be adjacent to replication forks, and increased replication
fork proteins might be found at replicated tract tips. ODD-BLOBS
quickly analyzes data from many samples to examine protein distri-
bution patterns and frequencies. User-defined changes allow data
smoothing to close signal gaps, and alter the size of tract ends.
ODD-BLOBS allows users to build a model of protein distribution
around replicated regions and presumed fork zones.
 A Chromatin Fiber Analysis Pipeline to Model DNA Synthesis and Structures. . . 511

2 Materials

2.1 DNA Fibers 1. Analog-incorporating strains must be used to detect DNA

synthesis. The strain must have a thymidine kinase (tk) gene
so that BrdU is converted into a useable nucleoside triphos-
phate. The kinase is supplied from an integrated hsv-tk cassette
(e.g., [15]) or on a plasmid (e.g., [10]). Transporters such as
the human equilibrative nucleoside transporter 1 (hENT1)
gene facilitate uptake of the analog and allow more efficient
incorporation at much lower concentrations of BrdU in cul-
tures [10, 15, 16].
2. 5-bromo-20 deoxyuridine (BrdU) solution at 5 mg/mL in ster-
ile water. If the strain has hENT1, BrdU is added at 50 μg/mL;
if not 200 μg/mL is required. BrdU stock solution is stored at
20  C and must be protected from light.
3. Hydroxyurea (HU, FW 76.5 g/mol) solution is made at 1 M in
water and stored in the dark at 4  C up to 1 week. For a
reversible early-to-mid S-phase block in S. pombe, stock HU is
diluted to 10–15 mM (see Note 1).
4. Phosphate buffered saline (PBS 1) is made with 137 mM
NaCl, 2.7 mM KCl, 10 mM Na2HPO4 and 1.8 mM
KH2PO4 in deionized water. A 10 solution can be made
and autoclaved or filter sterilized and stored at room
temperature.
5. Zymolyase Mix: 1 M sorbitol, 60 mM EDTA, and 100 mM
sodium citrate, pH 6.9–7.0 (see Note 2). Before use add
0.5 mg/mL zymolyase 20T, 1.0 mg/mL lysing enzymes and
100 mM 2-mercaptoethanol. If a large volume is required over
time (e.g., time course), prepared zymolyase mix can be kept
on ice between time points.
6. Zymolyase 20T (Seikagaku Corporation, Japan) is added to
0.5 mg/mL in zymolyase mix. This enzyme is supplied as a
powder, stored at 4  C, and is added to a master-mix volume of
item 5 solution before use.
7. Lysing enzymes (Sigma) are added at 1.0 mg/mL in zymolyase
mix to digest cell walls. This dried powder is kept at 4  C, to be
weighed and added to Zymolyase mix (item 5) when needed,
similarly to Zymolyase 20T.
8. Lysing solution is made of 50 mM Tris–HCl pH 7.4, 25 mM
EDTA, 500 mM sodium chloride, 0.1% Nonidet P-40 (Sigma),
and 0.5% (w/v) sodium dodecyl sulfate (SDS). Before lysis,
3 mM 2-mercaptoethanol is added to a suitable volume of
lysing solution, and the solution is heated to 70  C before use.
9. 4% paraformaldehyde (PFA) fixative is obtained as a 16% stock
solution, from various manufacturers (e.g., Pierce,
512 Sarah A. Sabatinos and Marc D. Green

ThermoFisher). It is supplied in 10 mL aliquots in brown glass

ampoules. To make 40 mL of a 4% solution, mix entire con-
tents of an ampoule with 30 mL of 1 PBS. Excess 4% PFA
solution can be stored at 4  C up to 1 month.
Alternatively, 4% PFA solution can be made by mixing 2 g PFA
powder in 50 mL of phosphate buffered saline and adding 4 M
sodium hydroxide added to pH to 7.0–7.5. Carefully heat and
stir the solution until all PFA powder is in solution and do not
boil. To prevent scorching the PFA, we recommend a double
boiler on a hot plate using an exterior beaker with water. Allow
the solution to cool and store at 4  C up to 1 month, or aliquot
and freeze at 20  C indefinitely (see Note 3).
10. Poly-L-lysine-treated coverslips. Treated coverslips can be pur-
chased premade. Alternatively, they are easily made from #1.5
glass coverslips (see Note 4). Coverslips are first washed in a
dilute solution of residue-free detergent (e.g., Sparkleen,
Fisher) and then rinsed in several changes of water and then
deionized water. The cleaned slides are soaked for 5 min in
100% ethanol and air-dried. Slides are submerged in a 0.1% to
0.01% poly-L-lysine solution (e.g., Sigma P8920) for 5 min at
room temperature. Coated slides are baked at 60  C for
approximately 1 h, cooled, and are stored in an airtight con-
tainer at 4  C indefinitely.

2.2 1. Phosphate buffered saline (1 PBS) as above.

Immunofluorescence 2. 2 N HCl, in water. After dilution from stock HCl, do not store
2 N HCl for longer than 4 months.
3. 0.1 M sodium tetraborate (Na2B4O7) solution (pH 8.5).
Na2B4O7 is dissolved in water and then pH adjusted to 8.5
with 4 N HCl. Filter-sterilize and store at room temperature,
protected from light.
4. Blocking buffer: 10% fetal calf serum, 10% bovine serum albu-
min, 0.05% Tween 20 detergent (Sigma) are diluted in 1 PBS
and filter sterilized. Store at 4  C up to 1 month.
5. Primary antibody for BrdU is rat BU1/75 (ICR1) (Abcam),
diluted to 1:100 in blocking buffer that contains any protein-
detecting antibodies (item 6).
6. Primary antibody concentration for protein detection is depen-
dent on the tag/protein of interest. All primary antibodies are
diluted in blocking buffer, together with the BrdU antibody
(item 5). We recommend a starting concentration of approxi-
mately 10 the usual concentration for western blot detection.
For example, if an antibody is used at 1:1000 for a western blot,
we would test 1:100 concentration of the same antibody for
chromatin fiber immunofluorescence. Alternatively, follow the
manufacturer’s recommendations for immunofluorescence.
 A Chromatin Fiber Analysis Pipeline to Model DNA Synthesis and Structures. . . 513

Validated antibodies and concentrations we have tried include:

mouse PCNA at 1:100 (PC-10 clone, Santa Cruz Biotechnol-
ogy), mouse anti-MYC tag at 1:100 (9E10 clone, Covance),
mouse anti-HA tag at 1:100 (16B12 clone, Roche), mouse
anti-FLAG at 1:100 (M2, Sigma), rabbit anti-Mcm4 at 1:100
[17], and yeast-specific anti-phospho histone H2A at 1:100
(07-745, EMD Millipore).
7. Secondary antibodies against each primary antibody are diluted
in blocking buffer together at 1:500 each (2 μL each antibody
in 1 mL of block). We typically use chicken anti-rat Alexa Fluor
488 (Invitrogen), goat anti-rabbit Alexa Fluor 546 (Invitro-
gen) and donkey anti-mouse Alexa Fluor 647 (Invitrogen). We
recommend changing the detection colors for all epitopes, i.e.,
goat anti-rabbit 647 and donkey anti-mouse 546 in a repeat of
the same experiment and primary antibodies (see Note 5).
Fluorescent-conjugated antibodies should be used in foil
wrapped tubes, incubated in the dark and shielded from light
as much as possible to prevent fluorophore degradation.
8. Antifade Mount is 50% glycerol in water with 0.1% PPD (p-
phenylenediamine dihydrochloride) and DAPI (40 ,6-diami-
dino-2-phenylindole dihydrochloride) at a final concentration
of 1 μg/mL (see Note 6).
9. DAPI stock solution is a 1 mg/mL solution in water, stored at
20  C. DAPI solutions should be kept in foil-wrapped tubes
and shielded from the light as much as possible.
10. VALAP, a 1:1:1 (w/w/w) mixture of petroleum jelly, lanolin
and paraffin wax that is used to seal coverslips onto slides (see
Note 7). VALAP solidifies below 60  C and must be melted
and mixed before use. VALAP can be stored in a beaker, melted
on a hotplate at low temperature, and applied to seal cover glass
onto slides using a thin wood rod.

2.3 One Dimensional ODD-BLOBS software can be downloaded from www.

Data: Boolean Logic SabatinosLab.net/ODD-BLOBS. ODD-BLOBS runs in the
Operations and Libre Office freeware suite (www.LibreOffice.org, or legacy ver-
Binning System sions of MS Office that support VBA). ImageJ (http://imagej.
(ODD-BLOBS) Analysis nih.gov/ij/, [18]) or some other image analysis software is
required to parse chromatin fiber images into intensity profiles
along line segments.

2.4 Specialized 1. We use 22  30 mm, #1.5 cover glasses for fiber preparation,
Equipment which are easier to manipulate than smaller square cover glass.
Cover glasses can be difficult to label and handle, but we find
that the image brightness and quality is considerably improved
on coverslips compared to slides.
514 Sarah A. Sabatinos and Marc D. Green

2. Humid chamber for sample incubations. A plastic container

with lid (e.g., food storage container) can be used, with damp
paper towels on the bottom and a level support on top to hold
samples (e.g., microfuge tube rack). The chamber can be
wrapped in aluminum foil to protect samples from light.
3. Specialty incubation jars and racks are available that fit
22  30 mm cover glasses and are useful in poly-lysine coating,
washing, and incubation steps. One supplier of these is Elec-
tron Microscopy Supplies. Regularly sized Coplin jars and/or
slide-washing baskets may be helpful for washing steps, but are
not essential.

2.5 Equipment Green et al. discuss imaging chromatin fibers and the reader is
Considerations for referred there for additional discussion. Confocal, widefield, and
Imaging super-resolution modalities should all produce good results when
high numerical aperture objectives and appropriate filter sets are
employed [19].

3 Methods

3.1 BrdU 1. Grow cells in an appropriate volume of minimal media to

Incorporation 1–5  106 cells/mL (OD595 ~ 0.3–0.6). Add hydroxyurea
(HU) to 12 mM, and grow for an additional 3 h at 30  C to
block in early/mid-S-phase. If desired, remove samples of the
asynchronous, blocked and experimental samples to confirm
HU effect using flow cytometry DNA staining (see [20]), and
DAPI/aniline blue staining for morphology assessment [19].
2. Remove HU by filtration onto a filter, and washing with
2 volumes of media without drug. Resuspend cells in 1 volume
of fresh, prewarmed (30  C) minimal media.
3. Alternative to step 2: remove HU by centrifugation. Pour
culture into a sterile 50 mL conical tube and centrifuge at
500  g for 5 min. Decant media, and resuspend pelleted
cells in 1 volume of fresh media. Centrifuge as above. Repeat
a second wash. Resuspend cells in 1 volume of fresh, pre-
warmed (30  C) minimal media.
4. Remove an unlabeled sample (10 mL culture), or partition the
culture to include a non-BrdU incorporated control released
for the same length of time. Add BrdU to 50 μg/mL and grow
for 15–45 min at 30  C. We suggest a 30-min release in BrdU
to start. Under these conditions, cells complete replication by
60 min post-release. Note that high BrdU concentrations and
longer incubation times cause cell toxicity (e.g., [16]).
 A Chromatin Fiber Analysis Pipeline to Model DNA Synthesis and Structures. . . 515

3.2 Preparation of Ahead of time, prepare poly-L-lysine-treated coverslips, zymolyase

Chromatin Fibers mix with enzymes and 2-mercaptoethanol, and heat lysis buffer
(containing 3 mM 2-mercaptoethanol) to 70  C. Zymolyase mix
with enzymes and 2-mercaptoethanol can be stored on ice until
required.
1. Grow cells as appropriate to the experiment, e.g., exponential
growth, or during/after temperature or drug treatment. BrdU
is added to label new DNA synthesis. Cultures should be in
mid-exponential phase (OD595 ~ 0.3–0.9, approximately
1–2  107 cells/mL).
2. Remove 10 mL of culture to a round bottom snap-cap tube.
Add sodium azide to 0.2%, mix cells, and incubate on ice for
5 min. Centrifuge at 500  g for 5 min in a swinging bucket
tabletop centrifuge. Decant supernatant.
3. Wash cells once in 1 PBS, centrifuge as above. Decant
supernatant.
4. Wash cells once in 1 mL zymolyase mix without enzymes or
1 mL water, and centrifuge as above.
5. Resuspend the cells in zymolyase mix with 2-mercaptoethanol
and enzymes. Incubate at 37  C for 15 min to spheroplast cells.
Check for digestion by phase contrast microscopy, both with
and without a drop of 10% SDS. Cells that are digested will be
spherical without added SDS. Addition of 10% SDS to the
sample will cause spheroplast lysis, leaving broken cells and
clear capsules (“ghosts”). Continue spheroplasting cells until
>70% of cells in the sample are digested. If buffer pH is correct,
this takes 10–15 min, but could require up to 30 min at 37  C
for hard-to-digest samples.
6. Centrifuge cells as above and decant.
7. Add 0.5 mL of 1 PBS to each tube. Resuspend spheroplasts
by gently shaking.
8. Pipette 25 μL of the cells onto the short (22 mm) end of a poly-
lysine-coated coverslip; a second line (25 μL) can be added,
with approximately 2 cm in between the two lines.
9. Let cells settle approximately 10 min at room temperature
(cover to protect from light). Tip the cover glass to one side
and blot away excess solution. Allow the samples to dry for
approximately 5 min, lying flat at room temperature, until the
edges are just dry but the center of the sample lines are slightly
damp.
10. Pipette 25 μL of hot (70  C) lysing solution (containing 3 mM
2-mercaptoethanol) onto the prepared and adhered cells. Incu-
bate flat, at room temperature for 3–5 min. Tip the coverslips
to a 15 angle and slowly increase to 30 . Use a pipette tip to
516 Sarah A. Sabatinos and Marc D. Green

channel the lysis solution down the center of the coverslip in

the direction of flow. This allows the DNA fibers to stretch out
of the lysed cells. Once all liquid has drained, dry samples
vertically at room temperature, approximately 3 min (see
Note 8).
11. Fix fibers in 4% paraformaldehyde for 10 min at room temper-
ature. Alternatively, pipette fixative onto the fibers, sit 10 min,
and then tip off PFA solution (see Note 9).
12. Rinse slips by dipping into a jar of 1 PBS. Repeat rinse. Blot
the excess PBS onto paper towels, and then allow the coverslips
to air-dry approximately 5 min.
13. Heat fix fibers onto the coverslips by placing onto a low-tem-
perature heat block at approximately 30–40  C for 5–10 min.
14. Store fixed fibers in the dark at 20  C for several weeks.

3.3 This immunofluorescence protocol detects both labeled DNA syn-

Immunofluorescence thesis tracts and proteins of interest on the chromatin fibers. Fiber
DNA is first denatured to facilitate BrdU–antibody interactions.
After washing and blocking, primary antibodies to BrdU and the
desired protein epitopes are added. Standard immunofluorescence
techniques are used to wash and detect antibodies, and slides are
mounted in the presence of DAPI to detect DNA.
1. Wet samples in 1 PBS. Protect BrdU-incorporated samples
from light as much as possible during this and subsequent
steps.
2. To denature DNA, immerse samples in 2 N HCl for 15 min at
room temperature. Neutralize slides in 0.1 M Na2B4O7 (pH 9)
for approximately 5 min, and then wash slides three times in 1
PBS.
3. Block slides in blocking buffer. Place coverslips (fiber-side up)
into a humid chamber and pipette 100 μL of blocking buffer
onto the samples. Distribute and cover with a piece of Parafilm,
cut to fit and cover the sample area. Avoid bubbles between the
sample and the Parafilm cover (see Note 10). Incubate samples
in blocking buffer in the dark for 30 min at 37  C, or for 1 h at
room temperature. During this incubation step, dilute primary
antibodies into blocking buffer (step 4).
4. Remove the Parafilm covers and tip off excess block from each
sample. Do not let the samples dry completely. Pipette 100 μL
of primary antibody solution onto the samples and cover with
Parafilm slips (as in step 3). Keep samples in dark, humid
chamber.
5. Incubate in primary antibody for 1 h at room temperature, or
overnight at 4  C.
 A Chromatin Fiber Analysis Pipeline to Model DNA Synthesis and Structures. . . 517

6. Remove Parafilm covers and blot off excess primary antibody.

Transfer samples into a Coplin jar with 1 PBS and wash for
3–5 min at room temperature. Repeat wash three times total.
Washes can be performed in Coplin jars or racked and washed
in larger containers. Protect samples from light during wash
steps. During wash incubations dilute secondary antibodies in
blocking buffer for use in step 7.
7. After the final wash, wipe excess PBS from the backs of samples,
and place in the humid chamber. Pipette 100 μL of diluted
secondary antibodies onto the samples and cover with a Paraf-
ilm coverslip (see Note 10).
8. Incubate in secondary antibodies for 1 h at room temperature.
9. Wash samples three times in 1 PBS, as in step 6. If desired,
add 0.1% Tween-20 to the first and second washes; this may
help to decrease background. The third wash should be 1PBS
only wash (no Tween 20).
10. Briefly air-dry sample slips in the dark at room temperature
until samples are damp but not wet. Do not overdry.
11. Mount samples, fiber-side down, onto glass slides using Anti-
fade Mount with DAPI at 1 μg/mL. A commercial alternative
is SlowFade gold with DAPI (Invitrogen), or other noncuring
mounting media. Seal edges with VALAP or nail varnish.
12. Using DAPI (DNA) fluorescence, look for whole cells and
then broken cells with chromatin fibers spilling out of lysed
nuclei. Single-stranded DNA antibodies may be used to detect
fibers and might provide a better signal for thin fibers.
13. Store slides protected from light at 4  C, or at 20  C.

3.4 Imaging DNA Parameters related to primary–secondary antibody combinations

Fibers (exposure times, etc.) must be considered for high-resolution
imaging. Negative controls that were not treated with primary
antibody are used to determine the baseline fluorescence from
nonspecific secondary antibody interactions. By testing no-primary
controls, samples exposed to primary antibody will show epitope-
specific patterns using the same settings. We also recommend pri-
mary antibody dropouts to ensure that two or more primary anti-
body patterns are not linked to each other.
1. Well-spread Chromatin Fibers are planar, within the axial reso-
lution of conventional and super-resolution visible-light
microscopy. Thin, likely single chromatin fibers will have a
very weak DAPI signal (see Fig. 1a), so much so it will often
go unnoticed in a field if some unspread nuclei or unlysed cells
are present. It may be necessary to find the fiber-spreads using
other color channels, e.g., tracts of BrdU synthesis (see
Note 11).
518 Sarah A. Sabatinos and Marc D. Green

Fig. 1 Example of fiber imaging and line tracing results. (a) Merged image of fibers prepared using the
described protocol, to detect DNA (DAPI, blue), DNA synthesis (BrdU, green), cohesin (Rad21-HA, orange),
Cdc45 (GFP tag, red). Bundles are brightly staining and composed of multiple DNA fibers. Single fibers have
weaker DAPI signal, and the BrdU signal is frequently punctate in replicated tracts; an arrow indicates a
replicated tip where a putative DNA replication fork site is hypothesized to occur. An unreplicated area is
identified with a line. Shown is a line tracing to acquire raw fluorescent intensity values on all channels with
pixel position information. All work was performed in strain FY 3841 (h+ rad21-3HA-kanMX6 cdc45-YFP-
ura4+ leu1–32::[hENT1+ leu1+(pJAH29)] his7–366::[hsv-tk+ his7+(pJAH31)] ura4- ade6–210). (b) An area of a
sample fiber pseudocolored for BrdU (blue), cohesin (red) and Cdc45 (green), with associated line tracing of
raw intensities. Values are plotted for each pixel along the tracing, with BrdU and cohesin scale on the left axis,
Cdc45 intensity scale on the right axis. (c) Thresholding was performed for all signals, converting the
fluorescent intensities into ON (above threshold value) or OFF (below threshold) for each place along the
line. Pixels from (b) were converted into distance along the line in μm, using the system conversion of
0.1092 μm/px (specific to microscope acquisition parameters)
 A Chromatin Fiber Analysis Pipeline to Model DNA Synthesis and Structures. . . 519

2. Samples are exposed and data acquired using multicolor z-

stacks. While optical sectioning is not obviously required for
such thin objects, it is helpful to capture z-chromatic
aberration.
3. Multicolor z-stacks are deconvolved in softWoRx with the
constrained iterative method and the manufacturer’s supplied
OTF for the Olympus 60 NA 1.4 PlanApo lens.

3.5 Parsing DNA ODD-BLOBS must be supplied with a linear array of intensity data
Fiber Data into One for each color channel in the fiber images. This data can be parsed
Dimension with almost any microscopy image software. In most programs this
analysis is called “Line Profiling” or similar, and involves the user
tracing a line along the fiber. The line may be straight or curved, but
is 1 pixel wide. The output is a trace of intensity values for each
color at each pixel along the line object that can be saved to a
delimited text file with columns for each color. Automatic fiber
detection is nontrivial and falls outside the scope of this article,
but the manual process is simple such that naive operators can
obtain quantitative amounts of One Dimensional Data from
images.
In ImageJ, the Analyze > PlotProfile operation can be per-
formed on any line object drawn on an active window (see Fig.
1a). A helpful macro for automating all colors in an image in one
operation is ‘Multi color line profile plot’ (Kees Straatman, 2014,
http://www2.le.ac.uk/colleges/medbiopsych/facilities-and-
services/cbs/lite/aif/software-1/imagej-macros#Multi%20color
%20profile), which is similarly executed on any line-drawing in an
open file.
1. Open file in ImageJ, using the LOCI Bioformats Importer if
needed so that channels are properly read in. Locate DAPI-
stained fibers. Activate the line tool in the Main Window. Trace
a line along a straight segment of fiber. Open the Ana-
lyze > Tools > ROI manager and “Add” the trace to the
ROI manager. Use the ROI manager tick box “show all” to
identify all traces and avoid duplication.
2. Draw additional line segment traces and “Add” them to the
ROI manager. When all desired traces are made Save the ROI
file under the ROI manager “more” button.
3. On the ROI manager, disable “show all” and select one line
segment.
4. Using Plugins > Macros > Run, activate ‘Multi color line
profile plot’
5. Export or copy/paste the intensity values to the ODD-BLOBS
spreadsheet.
520 Sarah A. Sabatinos and Marc D. Green

6. Repeat parsing and exporting line segments as required, Addi-

tional segments from the same experimental data type may be
appended to one ODD-BLOBS sheet if a one (or more) row
gap is left between data segments so that tracts across fibers are
not artificially concatenated.
7. ODD-BLOBS reserves Column B for pixel number/identity;
Column C should be DNA intensity values (e.g., DAPI); Col-
umn D should be the marker of replication tracts (e.g., BrdU
intensity values); Column E—protein #1 intensity values; Col-
umn F—protein #2 intensity values. Columns G to J are
reserved for pixel coordinates, if available.

3.6 Analyzing DNA To model DNA replication forks from chromatin fibers, we hypoth-
Fiber Data esize that replication forks are most likely to be found near the tips
of replicated tracts. Our analysis examines analog-incorporated
DNA (tracts) and associated protein (“blobs”) (see Fig. 2). A chro-
matin fiber is unreplicated (no analog signal), replicated (within an
analog region), or tip-proximal (putative replication fork zone).
Anywhere in a “replicated” zone might be called the “middle” of
that zone, and we do not consider the middle of un-replicated
tracts. Thus, replicated tracts have “Middles” and “Ends/Tips”.
The simplest DNA fiber analysis is done using analysis software
(e.g., ImageJ) to measure replicated tract lengths. However, asso-
ciating proteins with replication becomes tedious. Further, defining
where a protein is relative to replicated/unreplicated areas and the
tips is imprecise without measurement. If the pixel size is known,
we can measure and model how much space around a replicated
tract tip might be a region wherein a replication fork was active (see
Fig. 2).
We developed ODD-BLOBS to take linear fiber data, calculate
replicated tract lengths, and correlate protein location relative to
tract middles or tips/ends. Deconvolved, 2D-projected images are
analyzed using a line-drawing tool to acquire fluorescence intensi-
ties on all channels at a given point in the image. DNA staining by
DAPI or another DNA dye indicates the line that should be traced.
Data from all fluorescent channels is saved in a delimited file format
(see Note 12). The 1D/linear data is imported to LibreOffice and
an ODD-BLOBS worksheet. A threshold is set for each channel,
converting signal into a binary on or off datum. ODD-BLOBS uses
Boolean logic to calculate how many “On” pixel-signals make up a
tract-event, and where protein is relative to synthesized tracts (see
Fig. 2). Additional parameters allow the user to model smoothing
to remove gaps in signal, and define a larger or smaller window of
“tip” size for the purpose of modeling replication fork structures
(see Fig. 3).
1. Perform basic labeling of channels in cells C24:F24 and docu-
ment the housekeeping information in the comments section
 A Chromatin Fiber Analysis Pipeline to Model DNA Synthesis and Structures. . . 521

Fig. 2 The logic of ODD-BLOBS. (a) Pixels are represented as boxes that are ON (colored, or above threshold
signal) or OFF. A group of 4 green (ON) pixels is followed by four uncolored (OFF) pixels, distinguishing a
replicated tract from an unreplicated area. Between the replicated and unreplicated pixels is the tip of the
replicated tract, which we call the end and putative “Fork zone”. (b) “Tip/End windows” are determined as a
number of pixels around the tip of the replicated tract. A 1  1 pixel (px) window means that 1 pixel on each
side of the tip encompasses the putative “Fork zone”. This 1  1 px window is the minimum for ODD-BLOBS
analysis. A 2  3 px window here means that two replicated pixels and three unreplicated pixels on either side
of the end of the tract define the window of the putative “Fork zone”. (c) In the case where small,
nonresolvable gaps in signal occur (see Note 15) the “SmoothIt” algorithm closes the gaps. SmoothIt is
user-defined (i.e., “n pixels”) to close gaps of less than or equal to n. (d) A 1 pixel protein (“blob”) within the
end window, spans only the end and is called purely End or “Fork-zone” associated. Note that DNA synthesis
tracts are determined on Channel 1 (green), and protein on a second Channel 2. (e) A larger protein blob may
span both the End and Replicated regions. (f) A complex example of 3 tracts (green) and associated protein
blobs (red). The six tract ends are defined as putative “Fork-zones”. There are six identified protein blobs,
under a 2  2 pixel end-window definition. These include End only; End spanning to Replicated (end þ Rep/
mid-tract); End spanning to Unreplicated (end þ Unrep); Replicated only (Rep only); Unreplicated only (Unrep
only)
 522
Sarah A. Sabatinos and Marc D. Green

Fig. 3 Sample ODD-BLOBS input and output. (a) Data from line traces on all channels, with positional information, are input into columns B to J. (b) User defined
analysis parameters include thresholds and “values for SmoothIt” defining the gap size to be closed for each channel. These may be different for each channel.
“SmoothIt” closes a gap equal to or less than the defined number, but does not have to be used. The number, average size, and standard deviation of replicated
tracts are reported in the middle of this area. The “window size” of ends is set at right, and must be 1 or greater. (c) Data output of number of instances of different
kinds of events and number of pixels comprising those events
A Chromatin Fiber Analysis Pipeline to Model DNA Synthesis and Structures. . . 523

C11:C12 such as epitope identity (see Note 13) (example,

Fig. 3a).
2. Select a cutoff threshold for each channel’s fluorescence inten-
sity that determines presence or absence of signal in cells P23:
R23 (example Fig. 3b, right side).
3. In cell W15, tell ODD-BLOBS how many unreplicated pixels
proximal to the end of a replicated tract should be considered
part of the “tip” (see Note 14) (Fig. 3b, right top).
4. In cell W18, tell ODD-BLOBS how many replicated pixels
proximal to the end of a replicated tract should be considered
part of the “tip” (Fig. 3b, right bottom). The number of pixels
on either side of a tract end, steps 3 and 4, can be indepen-
dently set and becomes a modeling parameter.
5. Optionally, use cells P15:R15 to tell ODD-BLOBS for each
channel to close/smooth gaps in signal equal to or less then a
given size. This smoothing parameter is provided to smooth
local artifacts that could be introduced by binary thresholding
(see Notes 14 and 15). The number of pixels to smooth in
“SmoothIt” is designated for each signal (BrdU, proteins)
independently (Fig. 3b), and becomes a modeling parameter.
6. On the ODD-BLOBS spreadsheet call the “SetUpSheet”
macro with the supplied button. This macro performs some
basic layout procedures to enumerate the linear data and adjust
cell reference formulas to accommodate the number of rows of
data the user has provided.
7. Optionally call the “SmoothIt” macro using the button
provided.
8. Call the “FillFiberSheet” macro using the button provided.
9. ODD-BLOBS returns counts of protein tracts for all proteins
individually, relative to tracts (Fig. 3c). The first five columns
(Z to AD) are for above-threshold detections of Protein #1
(here Rad21, cohesin) along DAPI signal. The second set of
five columns (AE to AI) describes Protein #2. The last six
columns (AJ to AO) describe the total instances of Protein 1
and 2 colocalizations (column AJ), and their distributions rela-
tive to replicated tracts (AK to AO).
10. Distributions of proteins can be calculated relative to the total
number of BrdU tracts (found in Columns U and AU). A
parameter “Number of tracts decorated with protein” is
described relative to END/“Fork”, Replicated, or Repli-
cated-spanning END associations.
11. The total number of positive pixels is calculated below the
number of instances of association (Fig. 3c). This may allow
modeling if a protein complex is asymmetrically located or
sized in replicated/unreplicated areas.
524 Sarah A. Sabatinos and Marc D. Green

12. Protein association around tract ends may extend into Repli-
cated or Unreplicated areas. If a protein blob is only found in a
Replicated or Unreplicated area it is described independently
(Fig. 3c, Columns AC/AD; AH/AI). If the protein at a tip
extends past the defined tip window into purely Replicated or
Unreplicated areas (defined in steps 3 and 4) it is indepen-
dently counted (Fig. 3c, Columns AA/AB; AF/AG). These
parameters may be useful in cases where proteins are associated
at the fork and spread outward such as single stranded DNA.
13. If the user wishes to rerun ODD-BLOBS with different set-
tings in steps 2–5 both steps 7 and 8 should be rerun in order.
We recommend having a separate spreadsheet tab for separate
parameters.

4 Notes

1. Hydroxyurea is toxic. Handle with care and dispose of waste

HU under appropriate safety guidelines.
2. Zymolyase activity is pH dependent. An acidic pH is required
for efficient enzyme activity. If neutral or basic buffers are used
zymolyase digestion requires significantly more time.
3. Paraformaldehyde is toxic in powder and liquid forms. Wear
appropriate safety gear, work in a fume hood if possible and
dispose of reagent and waste with appropriate caution and
under regional safety guidelines.
4. #1.5 cover glass has an average thickness of 0.17 μm, which is
the thickness that most microscope objectives assume and cor-
rect for.
5. “Fluor swap” changes the secondary antibodies and colors
used to detect primary antibodies. For example, if chicken
anti-mouse Alexa Fluor 488 and donkey anti-rabbit Alexa
Fluor 546 were used in one experiment, the two colors could
be reversed (anti-mouse Alexa Fluor 546 and anti-rabbit Alexa
Fluor 488). Additionally, the host species of the secondary
antibody can be changed. Altering the batch and color of
secondary antibodies helps to identify and eliminate spurious
interactions and patterns.
6. 50% glycerol is used for the samples described in these proto-
cols, where samples are mostly dry. Alternatively, mount with
90% glycerol can be used.
7. VALAP is an excellent choice for sealing microscope samples,
since it is nonreactive and nontoxic. One preparation, made in a
small glass beaker with 10–20 g of each component, will last a
year or longer with moderate to heavy use. A thin bead of
melted VALAP is sufficient to completely seal around the
coverslip.
A Chromatin Fiber Analysis Pipeline to Model DNA Synthesis and Structures. . . 525

8. An alternative lysis method is to immerse the coverslip into a

Coplin jar or conical tube containing the 70  C lysis solution.
Incubate the coverslip in the lysis solution for 30 s, and then
remove slowly and vertically. To maintain temperature while
lysing multiple samples, the lysing solution can be kept in a
beaker with 70  C water.
9. An alternative fixative for these samples is 30% methanol with
10% formaldehyde, although we prefer 4% paraformaldehyde.
10. Carefully pipette the blocking buffer onto the cover glass
sample side up, and avoid adding bubbles. Place the Parafilm
cover onto the blocking buffer to exclude bubbles, which may
interfere with sample contact with blocking buffer and
antibodies.
11. Finding fibers using antibody-dependent channels carries a risk
of biased data collection toward epitope-positive events (i.e.,
only BrdU-incorporated DNA, as opposed to total DNA).
This is problematic if the percentage of replicated fibers is
desired, and is less of a problem if patterns are being examined.
To detect all DNA in an unbiased manner, DNA-antibody
methods are recommended.
12. Some image analysis programs save data as text and not num-
bers (e.g., *.slk files generated in softWorx Line Tool). To
convert text information into numbers in Libre Office, open
the *.slk file and select the affected columns of data to set the
cell format as “Number”. Choose Edit ! Find & Replace, and
enter ^[0–9] in the “Search for” box. In the Replace area, enter
“&”. Check on “Regular Expressions” and “Current selection
only”, and then click Replace All. These steps convert the data
into numbers in the spreadsheet.
13. Copy and paste data into the ODD-BLOBS spreadsheet, with
the following parameters: Column B—pixel number/identity;
Column C—DNA intensity values; Column D—synthesis/
BrdU intensity values, collected at 528 nm (green) emission;
Column E—protein #1 intensity values, collected at 617 nm
(orange) emission; Column F—protein #2 intensity values,
collected at 675 nm (far red) emission; Columns G to J—
pixel coordinates.
14. According to the Abbe limit of 0.61(λ/NAobj) in visible epi-
fluorescent microscopy, two objects will require approximately
200 nm separation to be resolved as distinct, using 488 nm
excitation and a state-of-the-art objective (NA 1.4).
15. “SmoothIt” closes gaps in signals. Assuming that the system is
sampling at the Nyquist rate, it is not possible to resolve a
single pixel gap. Pixel gaps might be introduced by threshold-
ing in ODD-BLOBS analysis. SmoothIt can eliminate pixel
gaps that are artifacts of binary thresholding. The number of
pixels smoothed can be set and used to model tract parameters.
526 Sarah A. Sabatinos and Marc D. Green

Acknowledgments

We wish to thank Susan Forsburg (University of Southern Califor-

nia, USA) for her support of this project and for providing strain
FY3841. The Sabatinos lab is supported by NSERC DG RGPIN/
04405-2015.

References
1. Patel PK, Arcangioli B, Baker SP, Bensimon A, 10. Sivakumar S, Porter-Goff M, Patel PK, Benoit
Rhind N (2006) DNA replication origins fire K, Rhind N (2004) In vivo labeling of fission
stochastically in fission yeast. Mol Biol Cell 17 yeast DNA with thymidine and thymidine ana-
(1):308–316 logs. Methods 33(3):213–219
2. Scorah J, McGowan CH (2009) Claspin and 11. Luke-Glaser S, Luke B, Grossi S, Constantinou
Chk1 regulate replication fork stability by dif- A (2010) FANCM regulates DNA chain elon-
ferent mechanisms. Cell Cycle 8 gation and is stabilized by S-phase checkpoint
(7):1036–1043 signalling. EMBO J 29(4):795–805
3. Jackson DA, Pombo A (1998) Replicon clus- 12. Bradford JA and Clarke ST (2011) Dual-pulse
ters are stable units of chromosome structure: labeling using 5-ethynyl-20 -deoxyuridine
evidence that nuclear organization contributes (EdU) and 5-bromo-20 -deoxyuridine (BrdU)
to the efficient activation and propagation of S in flow cytometry. Curr Protoc Cytom Chapter
phase in human cells. J Cell Biol 140 7 Unit 7 38
(6):1285–1295 13. Sullivan BA (2010) Optical mapping of pro-
4. Sabatinos SA, Green MD, Forsburg SL (2012) tein-DNA complexes on chromatin fibers.
Continued DNA synthesis in replication check- Methods Mol Biol 659:99–115
point mutants leads to fork collapse. Mol Cell 14. Haaf T, Ward DC (1994) Structural analysis of
Biol 32(24):4986–4997 alpha-satellite DNA and centromere proteins
5. Sullivan BA, Karpen GH (2004) Centromeric using extended chromatin and chromosomes.
chromatin exhibits a histone modification pat- Hum Mol Genet 3(5):697–709
tern that is distinct from both euchromatin and 15. Hodson JA, Bailis JM, Forsburg SL (2003)
heterochromatin. Nat Struct Mol Biol 11 Efficient labeling of fission yeast Schizosacchar-
(11):1076–1083 omyces pombe with thymidine and BUdR.
6. Blower MD, Sullivan BA, Karpen GH (2002) Nucleic Acids Res 31(21):e134
Conserved organization of centromeric chro- 16. Sabatinos SA, Mastro TL, Green MD, Fors-
matin in flies and humans. Dev Cell 2 burg SL (2013) A mammalian-like DNA dam-
(3):319–330 age response of fission yeast to nucleoside
7. Ross JE, Woodlief KS, Sullivan BA (2016) analogs. Genetics 193(1):143–157
Inheritance of the CENP-A chromatin domain 17. Forsburg SL, Sherman DA, Ottilie S, Yasuda
is spatially and temporally constrained at JR, Hodson JA (1997) Mutational analysis of
human centromeres. Epigenetics Chromatin Cdc19p, a Schizosaccharomyces pombe MCM
9:20 protein. Genetics 147(3):1025–1041
8. Bailis JM, Luche DD, Hunter T, Forsburg SL 18. Rasband WS. ImageJ. U.S. National Institutes
(2008) Minichromosome maintenance pro- of Health, Bethesda, Maryland, USA, http://
teins interact with checkpoint and recombina- imagej.nih.gov/ij/. 1997–2011
tion proteins to promote s-phase genome 19. Green MD, Sabatinos SA, Forsburg SL (2015)
stability. Mol Cell Biol 28(5):1724–1738 Microscopy techniques to examine DNA repli-
9. Cohen SM, Chastain PD 2nd, Cordeiro-Stone cation in fission yeast. Methods Mol Biol
M, Kaufman DG (2009) DNA replication and 1300:13–41
the GINS complex: localization on extended 20. Sabatinos SA, Forsburg SL (2015) Measuring
chromatin fibers. Epigenetics Chromatin 2 DNA content by flow cytometry in fission
(1):6 yeast. Methods Mol Biol 1300:79–97
 Chapter 35

Long-Term Imaging of DNA Damage and Cell Cycle

Progression in Budding Yeast Using Spinning Disk Confocal
Microscopy
Riccardo Montecchi and Etienne Schwob

Abstract
Live cell imaging can monitor biological processes in time and space by providing quantitative measure-
ments of cell behavior on a single-cell basis and in live conditions. However the illumination required to
visualize fluorescently tagged endogenous proteins often perturbs cellular physiology, a problem particu-
larly acute for yeast cells that are small, highly photosensitive and with scarce protein content. Analyzing the
activation of the DNA damage response (DDR) in various yeast mutants or growth conditions, as well as its
consequences for cell cycle progression and cell viability over extended periods of time therefore requires a
special microscopy setup that does not by itself create DNA damage or perturb cell growth. Here, we
provide a quick guide, strains and advice for imaging the DDR in S. cerevisiae for extended time (3–12 h)
using spinning-disk confocal microscopy in conditions of limited photobleaching and photodamage. DDR
is a conserved mechanism that allows the cell to respond to various stresses, especially those altering DNA
integrity or topology. Acquiring time-lapse images of the DDR at high temporal and spatial resolution is of
great interest, in particular when studying the effects of mutations or drugs which compromise genomic
stability and cell cycle progression.

Key words S. cerevisiae, Yeast, Spinning disk microscopy, Yokogama CSU-X1, DNA damage
response, Rad52-GFP, mCherry-Tub1, Microscopy, Recombination foci, Phototoxicity

1 Introduction

Live cell imaging is a powerful method that provides four-

dimensional information by correlating spatial three-dimensional
(x, y, z) acquisition to time (t). This contributed to reveal the
dynamics of numerous biological and biochemical events, the
nature of which was previously obscure [1–3]. However time-
lapse microscopy still remains an invasive approach. Although cells
have evolved protection mechanisms that allow them to handle

Electronic Supplementary Material: The online version of this chapter (doi: 10.1007/978-1-4939-7306-4_35)
contains supplementary material, which is available to authorized users.

Marco Muzi-Falconi and Grant W. Brown (eds.), Genome Instability: Methods and Protocols, Methods in Molecular Biology,
vol. 1672, DOI 10.1007/978-1-4939-7306-4_35, © Springer Science+Business Media LLC 2018

527
528 Riccardo Montecchi and Etienne Schwob

stressful situations, transmitted-light microscopy exposes the sam-

ple to overheating, drying, and photodamage. The main cause of
photodamage is the accumulation of light-induced reactive oxygen
species (ROS), such as singlet oxygen that is extremely reactive to
organic molecules, including those forming biological structures.
This cellular stress will either be undetected when cells are imaged
for a short period of time or remain under-appreciated when cells
are seen dividing, but becomes obvious when cells are imaged over
several generations [4, 5]. Many other factors besides illumination,
such as temperature, humidity, cell cycle stage and the presence of
intracellular chromophores or deleterious mutations will cumula-
tively increase the stress that cells experience during time-lapse
imaging.
Researchers therefore constantly try to develop solutions to
reduce the fluorescence excitation light delivered to cells during
time-lapse acquisition, while trying to maintain a sufficient signal-
to-noise ratio for each monitored fluorescent protein. One such
solution is spinning disk microscopy where the incident excitation
light is temporally and spatially segmented at high frequency so to
avoid continuous sample illumination. When combined with a
sensitive EM-CCD camera, this method allows imaging of live
cells over long periods of time under conditions of low photo-
bleaching and photodamage [6, 7]. Yeast cells are particularly
photosensitive during live cell imaging due to their small size; low
protein content; and the presence of intracellular chromophores
that absorb light and generate ROS. For instance, the light irradi-
ance that yeast cells receive during fluorescence microscopy with a
100x objective can be 103 to 106 more intense than direct sunlight
[8], a dose yeast cells are certainly not equipped to deal with.
Indeed Sedat and colleagues have calculated that the irradiance a
yeast can tolerate without adverse long-term effects is ~0.5 μW/μ
m2·s, about 100 less than the dose given during traditional wide-
field fluorescence microscopy [4]. Another problem is that micros-
copy laser output varies with time due to misalignment, acousto-
optical tunable filter (AOTF) overheating, etc., which means that
the same illumination settings will deliver different light intensities
to the sample during different imaging sessions. To deal with this
issue it is recommended to record and reset the actual light intensity
delivered by the system using a slide power meter (e.g., Thorlabs
S170C). However the maximum light irradiance a cell can cope
with also depends on its genetic make up and growth condition,
which calls for using a cellular proxy for photodamage.
In this chapter we provide a pragmatic solution to quickly assess
photodamage during time-lapse acquisition of yeast cells within
each user’s environment, and a guide to perform dual fluorescence
imaging for several cell generations of S. cerevisiae using spinning
disk confocal microscopy (Yokogawa CSU-X1) in conditions that
minimize the stress caused to cells. Cells are maintained immobile
 Visualizing DDR by Long-Term Spinning Disk Microscopy 529

and viable for at least 12 h under a thin layer of agarose in synthetic

complete medium poured in glass-bottom FluoroDishes. Although
cell growth and doubling time are still the most common criteria
used to assess phototoxicity during time-lapse acquisition [4, 9], we
noticed when using a reporter of DNA repair and recombination
(Rad52-GFP) that even fast dividing cells can experience important
stress. Thus we suggest monitoring the DNA damage response
(DDR) as an additional sensor for microscopy-induced photodam-
age. Indeed, it is well known that ROS accumulation leads to a
variety of damage to proteins and DNA [10, 11], which can lead to
a strong activation of the DDR. Here we use a diploid yeast strain
expressing GFP-tagged Rad52 from its endogenous locus that
produces bright nuclear foci upon DNA damage and DSB repair
[12]. The strain also carries an ectopic copy of mCherry-TUB1 to
label microtubules and monitor cell cycle progression [13]. Alter-
native read-outs for DNA damage, such as Rfa1-GFP, Ddc1-GFP
or Mre11-GFP can also be used [14]. Wild-type strains and various
mutants were observed for up to 12 h or eight generations with
little or no photobleaching or photodamage using an Andor CSU-
X1 spinning disk microscope equipped with an EM-CCD camera.
The strain expressing Rad52-GFP is useful for everyone to rapidly
determine the highest illumination condition that does not cause
overt photodamage to cells in their own microscopy setup. Long-
term cell imaging using spinning disk microscopy is valuable for
determining the duration of the DDR and its consequences for cell
cycle progression, cell division, and the fate of progeny. Protocols
for determining the position, colocalization and mobility of DNA
double-strand breaks (DSB) have been described recently else-
where [15].

2 Materials

2.1 Yeast Culture 1. Yeast strains.

and Sample l E3416: MATa/α, RAD52-GFP/RAD52-GFP, URA3::
Preparation mCherry-TUB1/URA3::mCherry-TUB1 (W303)
l E3410: MATa/α, cdc6–1/cdc6–1, RAD52-GFP/RAD52-
GFP, URA3::mChe-TUB1/URA3::mChe-TUB1
l E5357: MATa/α, mec1Δ/mec1Δ, sml1Δ/sml1Δ, RAD52-
GFP/RAD52-GFP, URA3::mCherry-TUB1/URA3::
mCherry-TUB1
2. SC-D medium (synthetic complete + dextrose): 2 g/L Yeast
nitrogen base (w/o amino acids, w/o NH4SO4), 5 g/L
NH4SO4, 20 mL/L 50 AAA mix (10 g/L threonine; 5 g/
L lysine, leucine, tryptophan, and phenylalanine; 3 g/L isoleu-
cine and methionine; 2.5 g/L histidine, adenine, and uracil;
530 Riccardo Montecchi and Etienne Schwob

2 g/L arginine), buffered with10 mM Tris–HCl pH 7.5 and

2% dextrose (see Note 1).
3. 1% agarose (e.g., UltraPure, InVitrogen) in SC-D medium.
4. Tissue culture dish with glass bottom. 35 mm dish diameter,
0.17 mm glass thickness, non-pyrogenic, gamma sterilized
(e.g., FluoroDish, World Precision Instrument).

2.2 Image A number of suitable spinning disk confocal microscopes exist. We

Acquisition describe the equipment used in our group.
1. Andor Nikon Spinning disk microscope CSU-X1, equipped
with 488 nm (60 mW), 561 nm (50 mW) solid state laser
lines and EM-CCD camera (Andor iXon Ultra).
2. Nikon 100/1.45 oil objective with objective heater and con-
troller unit (Bioptech).
3. Environmental chamber and control (Okolab H301-T-UNIT-
BL-PLUS).
4. Acquisition software (Andor IQ3).

2.3 Image 1. Fiji Is Just ImageJ software [16].

Processing 2. Pure Denoise plug-in for Fiji (http://bigwww.epfl.ch/
algorithms/denoise/).

3 Methods

3.1 Yeast Culture 1. Grow cells at 30  C in buffered SC-D to a concentration of 107

and Sample cells/mL.
Preparation 2. Centrifuge 1 mL cells for 4 min at 5000 rpm (2400  g) and
resuspend in 100 μL SC-D.
3. Melt an aliquot of 1% agarose in SC-D by microwaving (see
Note 2).
4. Pour 800 μL of the melted medium and spread it evenly on the
bottom glass of a FluoroDish (see Notes 3 and 4). Let the
agarose layer harden 5 min at 4  C with lid open. Then take it
back to RT (see Note 5).
5. Gently lift up the agarose layer with a 100 μL pipetman tip held
in one hand.
6. With the other hand deposit 8 μL of the sample in 8–10 small
drops on the glass surface of the FluoroDish (Fig. 1a).
7. Gently put the agarose layer back in place, making sure that no
air bubble gets trapped.
 Visualizing DDR by Long-Term Spinning Disk Microscopy 531

Fig. 1 Imaging chamber schematic. (a) The agarose-SCD pad (yellow) is lifted and several droplets of yeast
cells are deposited on the glass bottom of a FluoroDish. The agarose pad is then gently put back in place so
that cells are immobilized and growing in monolayer. (b) The FluoroDish is clamped on the plate holder within
a closed imaging chamber set at the right temperature and containing a water reservoir for humidification.
Covering the FluoroDish with glass improves transmitted light images. Objective heater is required to maintain
the sample at the set temperature

3.2 Image 1. Set the temperature of the chamber and objective heater to an
Acquisition effective temperature of 30  C on the sample, at least 30 min
before acquisition (see Note 6).
2. Place the FluoroDish in a humidified chamber insert in order
to prevent drying (Fig. 1b).
3. Focus on the cells using the transmitted light mode and set the
Perfect Focus System (PFS) to keep cells in focus during the
entire duration of the movie.
4. For each field of view, program the acquisition loop to take one
image with Differential Interference Contrast (DIC) and 11 z-
stacks of 0.67 μm step-size in the GFP and mCherry channels
(see Note 7). It is highly recommended that a DIC-alone field-
of-view is taken as a negative control for laser-induced
photodamage.
5. When using the RAD52-GFP, mCherry-TUB1 tester strain
(E3416), set the green and red laser power at the minimum
intensity (mW) necessary to barely see the nucleoplasmic
Rad52-GFP signal (soluble pool) and the mCherry-Tub1
labeled mitotic spindle (see Note 8).
6. Perform a 4–12 h acquisition by taking one image (11 z-stacks)
every 1 or 2 min.
7. Process the images using SUM slices or MAX projection for
optimal signal intensity.
8. Use the Pure Denoise plug-in for Fiji to denoise images and
increase the signal to noise ratio.
532 Riccardo Montecchi and Etienne Schwob

Fig. 2 Persistent Rad52-GFP foci in large budded cells as a marker of photodamage and/or DDR activation. (a)
Wild-type diploid yeast cells (E3416) grown and imaged at 30  C for 6 h (shown frame is at 96 min) using a
spinning disk confocal microscope under low illumination conditions. Green, Rad52; red, tubulin. Only a
fraction of the cells in S phase (small bud, no spindle) contain Rad52 foci. (b) cdc6–1 temperature-sensitive
cells (E3410) grown and imaged in the same conditions show persistent Rad52 foci in the majority of
metaphase-arrested cells (large bud, short spindle). Shown frame is at 146 min. A similar phenotype is
seen in WT cells exposed to higher laser illumination (see Movie 2)

9. Rad52 forms nuclear foci in a fraction of S-phase cells [12], but

these foci are faint, transient (2–15 min) and appear in small
budded cells (Fig. 2a).
l The presence of persistent Rad52-GFP foci (>30 min) in
large budded cells with short spindle is a sign of important
photodamage (Fig. 2b).
l In unstressed cells the mitotic spindle remains in its meta-
phase state (1–2 μm) only for a short time before under-
going rapid extension during anaphase (see Movie 1). A
stressed cell will show a slightly elongated metaphase spindle
present at the bud neck for considerably longer time before
engaging anaphase (see Movie 2).
l Large, bright, and long-lasting Rad52-GFP foci gradually
accumulate in number over time in cells experiencing
photodamage, because photodamage is cumulative.
l Cells blocked in G2 or metaphase will have a bud that keeps
growing in size, whereas dead cells will stop increasing their
cell mass. After many hours of growth without division, cells
will eventually burst.
l Some damaged cells can undergo mitosis and divide in spite
of the presence of a bright Rad52-GFP focus, indicating that
cell division is not a good indicator for the absence of stress.
 Visualizing DDR by Long-Term Spinning Disk Microscopy 533

Fig. 3 Comparison of DIC alone and laser-illuminated field of views reveals long-term toxicity of high light
irradiance. (a, b) Wild-type yeast cells (E3416) were imaged every 2 min for 4 h at 30  C on a spinning-disk
confocal microscope either without (a) or with 11  300 ms exposure with 488 nm light (GFP) and 561 nm
light (mCherry), both set at 20% laser power (b). The right panel shows that cells receiving laser light contain
Rad52 foci and divide less rapidly than those exposed to transmitted light only. (c, d) Cells were treated
identically, except that laser power was set at 5%. Cells divided equally well regardless of laser light exposure
and contained little or no Rad52 foci, indicating that these conditions are adequate for live cell imaging with
limited photodamage

10. At the end of the movie (or after 6–12 h of growth) compare
the number of cell divisions in the DIC-only field of view to
that in the laser-illuminated regions (Fig. 3). If cell density is
roughly equal then the conditions of imaging are adequate
with little or no induced photodamage.
534 Riccardo Montecchi and Etienne Schwob

11. Using these conditions it is then possible to study the effect of

drugs (e.g., Movie 3, WT + camptothecin) or mutants (Movie 4,
mec1Δ + camptothecin) on the activation of the DDR and cell
cycle progression. However, keep in mind that drugs and muta-
tions can have additive effects with the light-induced photo-
damage, and may show a different response in non-illuminated
conditions. Conducting experiments in parallel with live or fixed
cells that are not permanently exposed to strong light is there-
fore recommended.

4 Notes

1. Use SC-D medium for better signal–noise ratio, avoiding

the high autofluorescence of cells grown in rich medium
(YPD). Keeping the pH near neutrality (pH 7.5) improves
GFP stability [17].
2. Prepare a solution of 1% agarose in SC-D beforehand, and keep
1 mL aliquots at 4  C. Briefly microwave to melt before pour-
ing. If drugs are to be added to the medium, let it first cool
down to ~50  C.
3. Start pouring from the border between glass and plastic; this
helps even distribution of the medium. A small agarose bump
will not affect the acquisition, as light is collected from the
bottom.
4. We found FluoroDish with agarose top layer to be good solu-
tion for long and short acquisition times of budding yeast cell
division. Yeast cells are more immobile than on concanavalin-
treated glass, while growing for a long time in two dimensions,
thus facilitating pedigree analysis. Avoid acquiring fields that
are close to the edge of the agarose layer, however, as it might
shrink and move slightly due to drying. For long-term acquisi-
tion, place wet paper pads or water drops near the FluoroDish
edge.
5. Since cold is known to de-polymerize microtubules, wait for
the dish surface to reach room temperature before loading
the sample. Short-term storage of the FluoroDish-Agarose-
SCD is possible at 4  C, if sealed with Parafilm to prevent
drying.
6. Immersion objectives are heat sinks for the sample in contact
with the lens. Effective temperature under the objective can be
4  C lower than the set temperature of the microscope envi-
ronmental chamber. An objective heater should be placed at
least 30 min before the acquisition in order to reach the
expected temperature. Temperature sensors placed near the
 Visualizing DDR by Long-Term Spinning Disk Microscopy 535

sample (below the agar pad) should be used to set the correct
temperature before starting the experiment.
7. In our setup, we found QUAD dichroic emission filters (440/
40, 521/21, 607/34 and 700/45 nm) produced images with
better signal to noise ratio.
8. By pressing autoscale during acquisition, it should be possible
to see the nucleoplasmic Rad52-GFP and the mitotic spindle in
the mCherry channel. If not, increase the laser intensity.
Performing short acquisitions (2–3 frames), and immediately
checking the final result with ImageJ (SUM slice or MAX
projection) can also help setting the correct laser intensity.
Change field of view after setting the parameters.

Acknowledgments

We thank D. Tamborrini and S. Piatti (CRBM, Montpellier) for the

initial set-up of yeast live cell imaging, V. Georget (BioCampus
MRI Imaging Facility) for endless support, as well as R. Rothstein
(Columbia U., New York) and E. Schiebel (ZMBH, Heidelberg)
for providing yeast strains and constructs. RM was a Master student
from University Milan-Bicocca (Italy) holding an Erasmus+ stipend
from the EU. ES thanks Association pour la Recherche contre le
Cancer (ARC) and Agence Nationale de la Recherche (ANR-14-
CE11-0007-03) for funding.

References
1. di Pietro F, Echard A, Morin X (2016) Regula- technology and future trends. Methods Cell
tion of mitotic spindle orientation: an Biol 123:153–175
integrated view. EMBO Rep 17:1106–1130 8. Ettinger A, Wittmann T (2014) Fluorescence
2. Park S, Greco V, Cockburn K (2016) Live live cell imaging. Methods Cell Biol 123:77–94
imaging of stem cells: answering old questions 9. Frigault MM, Lacoste J, Swift JL, Brown CM
and raising new ones. Curr Opin Cell Biol (2009) Live-cell microscopy – tips and tools. J
43:30–37 Cell Sci 122:753–767
3. Miné-Hattab J, Rothstein R (2013) DNA in 10. Jena NR (2012) DNA damage by reactive spe-
motion during double-strand break repair. cies: mechanisms, mutation and repair. J Biosci
Trends Cell Biol 23:529–536 37:503–517
4. Carlton PM, Boulanger J, Kervrann C et al 11. Davies MJ (2003) Singlet oxygen-mediated
(2010) Fast live simultaneous multiwavelength damage to proteins and its consequences. Bio-
four-dimensional optical microscopy. Proc Natl chem Biophys Res Commun 305:761–770
Acad Sci U S A 107:16016–16022 12. Lisby M, Rothstein R, Mortensen UH (2001)
5. Magidson V, Khodjakov A (2013) Circum- Rad52 forms DNA repair and recombination
venting photodamage in live-cell microscopy. centers during S phase. Proc Natl Acad Sci U A
Methods Cell Biol 114:545–560 98:8276–8282
6. Inoué S, Inoué T (2002) Direct-view high- 13. Khmelinskii A, Lawrence C, Roostalu J, Schie-
speed confocal scanner: the CSU-10. Methods bel E (2007) Cdc14-regulated midzone assem-
Cell Biol 70:87–127 bly controls anaphase B. J Cell Biol
7. Oreopoulos J, Berman R, Browne M (2014) 177:981–993
Spinning-disk confocal microscopy: present
536 Riccardo Montecchi and Etienne Schwob

14. Lisby M, Rothstein R (2004) DNA damage 16. Schindelin J, Arganda-Carreras I, Frise E et al
checkpoint and repair centers. Curr Opin Cell (2012) Fiji: an open-source platform for
Biol 16:328–334 biological-image analysis. Nat Methods
15. Horigome C, Dion V, Seeber A et al (2015) 9:676–682
Visualizing the spatiotemporal dynamics of 17. Alkaabi KM, Yafea A, Ashraf SS (2005) Effect
DNA damage in budding yeast. Methods Mol of pH on thermal-and chemical-induced dena-
Biol Clifton NJ 1292:77–96 turation of GFP. Appl Biochem Biotechnol
126:149–156
 Chapter 36

The CellClamper: A Convenient Microfluidic Device

for Time-Lapse Imaging of Yeast
Gregor W. Schmidt, Olivier Frey, and Fabian Rudolf

Abstract
Time-lapse fluorescence imaging of yeast cells allows the study of multiple fluorescent targets in single cells,
but is often hampered by the tedious cultivation using agar pads or glass bottom wells. Here, we describe
the fabrication and operation of a microfluidic device for long-term imaging of yeast cells under constant or
changing media conditions. The device allows acquisition of high quality images as cells are fixed in a two-
dimensional imaging plane. Four yeast strains can be analyzed simultaneously over several days while up to
four different media can be flushed through the chip. The microfluidic device does not rely on specialized
equipment for its operation. To illustrate the use of the chip in DNA damage research, we show how
common readouts for DNA damage or genomic instability behave upon induction with genotoxic chemi-
cals (MMS, HU) or induction of a single double-strand break using induced CRISPR-Cas9 expression.

Key words Microfluidic, Perfusion, Long-term imaging, Fluorescence microscopy, Single-cell

analysis, S. cerevisiae, DNA damage, CRISPR-Cas9, MMS, Hydroxyurea, Genomic instability

1 Introduction

Fluorescence time-lapse microscopy is instrumental for studying the

spatial and temporal aspects of cell biology in vivo. Cellular processes
are most commonly visualized by the use of genetically encoded
reporter proteins. Fusing a fluorescent protein to a protein of interest
provides a direct readout of its abundance and localization. Seminal
time-lapse imaging studies used protein fusions to elucidate the
spatio-temporal relationships between the budding yeast DNA dam-
age checkpoint and repair proteins [1–3]. The timing and localiza-
tion of these proteins remain useful reporters in live cell studies of the
DNA repair process [4, 5]. Alternatively, fluorescent proteins can be
used as transcriptional reporters by fusing their coding sequence to a
promoter of interest [6, 7]. More elaborate reporter constructs allow
for a monitoring of phosphorylation events upon DNA damage and
therefore studies on the kinetics of the DNA damage response
signaling became possible [8, 9].

Marco Muzi-Falconi and Grant W. Brown (eds.), Genome Instability: Methods and Protocols, Methods in Molecular Biology,
vol. 1672, DOI 10.1007/978-1-4939-7306-4_36, © Springer Science+Business Media LLC 2018

537
538 Gregor W. Schmidt et al.

The quality of time-lapse imaging data is determined largely by

the experimental setup. Classically, time-lapse imaging of yeast is
performed by embedding cells in agar pads. The agar keeps the cells
in place and provides nutrients for the duration of the experiment.
When studying DNA damage responses, induction of DNA dam-
age can be achieved by chemical treatment of cells directly before
embedding in the agar pad [3], or alternatively by exposing cells to
UV-light using the microscope. While the agar pad has served
researchers well, there are some drawbacks. The quality of the
images is compromised by the autofluorescence of the media and
the shape of the agar drop distorts the light path. Also, use of agar
pads limits the duration of an experiment to <12 h and does not
allow to dynamically administer genotoxic agents. One can partially
bypass these nuisances by using glass bottom plates coated with
concanavalin-A or poly-L-lysine for the cultivation of yeast cells
[10]. While allowing for media exchange and better image quality,
the obtained data is limited to the cells present at the beginning of
the experiment as newborn cells usually do not adhere to the
coating.
Microfluidic technology can overcome these limitations [11].
A recent study aimed at deciphering the spatial and temporal
dynamics of the yeast proteome in response to DNA damage can
illustrate the potential as well as the drawbacks of microfluidic
devices [12]. In this study, DNA damage was repeatedly induced
by pulsatile administration of genotoxic chemicals or UV-light
allowing to study the initial relocalization kinetics of the proteins
as well as their behavior when adapting to prolonged DNA damage.
Continuous perfusion of microfluidic devices ensures rapid media
exchange and enables long observation time through a constant
supply of nutrients and removal of excess cells. Additionally, micro-
fluidic devices allow better image quality as background fluores-
cence of the media can be excluded by trapping the cells between a
glass cover slip and the microfluidic structure made of biocompati-
ble and non-fluorescent polydimethylsiloxane (PDMS). In this
case, high image quality can even be obtained when strongly auto-
fluorescent media such as YPD are used.
Although many microfluidic devices for research have been
developed and their advantages are evident, widespread use of
microfluidics is hindered for several reasons. Commercial devices
are either expensive or cannot be easily adapted to different experi-
mental questions. Published devices have to be self-made and
therefore require equipment and knowledge of the microfabrica-
tion process. Recently, custom-made microfabricated wafers or
PDMS chips have become commercially available, allowing the
experimentalist to bypass the fabrication process. However, estab-
lishing a working setup still requires dedicated equipment (e.g.,
plasma oven, microfluidic pump) and a substantial amount of time
and knowledge which are not common in most biological labs.
 The CellClamper: A Convenient Microfluidic Device for Time-Lapse. . . 539

Cross-section of chip
Vacuum 3.8 µm height
20 µm height
200 µm height

Inlets Outlet
Vacuum

Cross-section
PDMS
yeast cell Culture chamber
glass slide

Mixer Cell culture pad

Fig. 36.1 Design of the microfluidic device for long-term culture and high-resolution imaging of yeast cells.
Cells are trapped between a PDMS culture pad and a glass slide. The chip allows simultaneous culturing of
four different yeast strains. A vacuum channel (green) surrounds the fluidic channel network (blue), and
facilitates a tight seal of the PDMS to the glass and continuous removal of air bubbles from the system. Media
can be switched or mixed from four separate inlets. A microfluidic herringbone mixer facilitates mixing,
despite laminar flow, for flow rates up to 100 μL/min. Design files are available upon request

These challenges can be overcome by engineering microfluidic

devices which are simple to operate and do not rely on plasma
bonding or actively driven pumping mechanisms.
Recently, we engineered a microfluidic device that can be oper-
ated with materials present in any biological lab [13] (Fig. 36.1). In
the device, yeast cells are immobilized by clamping them between a
glass slide and a PDMS culturing pad. The individual pads are
surrounded by continuous media flow allowing the exchange of
media and application of genotoxic chemicals. The media compo-
sition can be dynamically controlled and mixed from four different
media reservoirs. In our microfluidic device, the PDMS and glass
slide are kept together using a vacuum channel, obviating the need
for plasma bonding used in conventional PDMS based devices. This
allows for loading of cells by directly dispensing them on the
microfluidic structure before assembling it with the coverglass. In
each device, up to four different strains can be cultured simulta-
neously in four separate cell culture chambers. The vacuum chan-
nel, which is placed around the fluid filled channel system,
continuously removes bubbles from the system ensuring long-
term stability of the media flow. Perfusion of the device can be
established using gravity flow or microfluidic pumps. Here, we
describe the fabrication and operation of our microfluidic device
in a step-by-step manual and show how DNA damage caused by
genotoxic agents or induction of CRISPR-Cas9 can be measured
inside the device. While the microfluidic device described in this
540 Gregor W. Schmidt et al.

protocol was designed for culturing yeast (S. cerevisiae or S. pombe),

it can be adapted to work with bacteria or mammalian cells by
adjusting the height of the culturing pads.

2 Materials

Prepare all solutions using ultrapure water and use analytical grade
reagents. Prepare and store all reagents at room temperature
(unless indicated otherwise). Diligently follow all waste disposal
regulations when disposing waste materials. Wear protective cloth-
ing and glasses when handling solvents. Work with microfluidic
devices requires a dust-free work environment.

2.1 Fabrication of l Particle free and sterile cleanroom gloves (see Note 1).
Microfluidic Chip l Particle free cleanroom wipes.
l Silanized microfabricated wafer (see Notes 2 and 3).
l Scotch tape.
l Air spray gun with luer lock connector.
l 0.22 μm sterile filter with luer lock connector.
l Petri dishes (diameter: 150 mm).
l Polydimethylsiloxane (Sylgard 184).
l Plastic container and spatula.
l Desiccator.
l Level surface (optional: in convection oven at 80 ∘C).
l Razor blade or knife.
l Biopsy punch (outer diameter: 0.63 mm).

2.2 Preparation of l Microfluidic tubing (Tygon LMT-55, inner diameter: 0.51 mm,
Media Reservoirs and wall thickness: 0.85 mm).
Microfluidic Tubings l Blunt tip luer lock needle straight (inner diameter: 0.34 mm,
outer diameter: 0.64 mm).
l Blunt tip luer lock needle 90∘ bent (inner diameter: 0.34 mm,
outer diameter: 0.64 mm).
l Scissors.
l Flat pliers.
l Pincer pliers.
l Sterile disposable syringes with luer lock tip (10–60 mL).
l 70% ethanol in spray bottle.
l 70% ethanol in falcon tube.
l Growth media.
l 50 mL falcon tubes.
 The CellClamper: A Convenient Microfluidic Device for Time-Lapse. . . 541

l Sterile disposable bottle top filters (pore size: 0.22 μm).

l Aluminum foil.
l Optional: Sterile hood.
l Optional: Microfluidic syringe pump and glass syringes with luer
lock tip (10–60 mL, ILS Innovative Labor Systeme GmbH,
Mittelstrasse 37, 98714 Stuetzerbach, Germany).

2.3 Setting Up a l Acetone, isopropanol, and ultrapure water in squeeze bottles.

Time-Lapse Imaging l Tube rack which can accommodate syringe OR microfluidic
Experiment Using the pumps with matching syringe holders.
Microfluidic Device l Motorized inverted fluorescence microscope with high NA oil
immersion objective (40–100 magnification).
l Vacuum pump.
l Self-adhesive paper tape.
l Nail polish (clear).
l Cover slip to accommodate microfluidic chip (thickness:
0.13–0.16 mm). We use 60 mm  24 mm (Thermo-Fischer
Scientific, catalog number size: Q10143263NR1). Differently
sized cover slips can be used but should measure at least 50 mm
 14 mm to fit microfluidic chip.
l Microscope slide holder fitting the cover slip with adjustable
screws for levelling.
l Kimtech wipes.
l Fume hood.
l Yeast culture in exponential growth phase.
l Foldback clips.
l Small volume pipette (able to dispense  0.4 μL).
l 1.5 mL Eppendorf tubes.
l Microcentrifuge.

3 Methods

3.1 Fabrication of Work in a dust-free environment and use cleanroom gloves for all
Microfluidic Chip steps. PDMS molding of one wafer yields seven microfluidic chips
which can be re-used as often as  30 times (see Note 4). Therefore
these steps do not need to be carried out for each experiment.
1. Using Scotch tape, build a barrier around the silicon wafer.
Blow any dust off the wafer using the air spray gun (see Note 5).
Place wafer into polystyrene Petri dish.
542 Gregor W. Schmidt et al.

2. In a plastic beaker vigorously mix PDMS monomer and curing

agent in a ratio of 10:1 (w:w) using a plastic spatula. Prepare
30 g of PDMS mixture per wafer (see Note 6).
3. Degas PDMS mixture for  30 min in a desiccator, until
solution is free of bubbles (see Note 7).
4. Place wafer inside the Petri dish on a scale and pour 25 g of
degassed PDMS onto the wafer.
5. Place Petri dish with wafer in a desiccator for  30 min to
remove trapped air bubbles (see Note 8).
6. Place Petri dish with wafer on a level surface in a convection
oven and incubate for at least 1 h at 80 ∘C to cure PDMS (see
Note 9).
7. Remove Scotch tape from wafer. Carefully peel the PDMS from
the wafer by slowly releasing it from each side.
8. Cut single microfluidic devices from PDMS slab using a razor
blade or knife.
9. Punch inlet and outlet holes using the biopsy punch (see
Note 10). The chip allows attachment of four different media
reservoirs through separate inlets. Punch only inlet holes that
you need, as they cannot be closed afterwards.
10. Store chips in petri dish at room temperature until further use.

3.2 Preparation of Handle all tubings and connectors using cleanroom gloves to pre-
Microfluidic Tubings vent contamination with particles. Clean workspace using wet
and Media Reservoirs paper towels to create a dust-reduced environment or work in a
for Gravity Driven sterile hood. Microfluidic tubing can be re-used. Therefore, steps
Operation of the 1–5 do not need to be carried out for each experiment.
Microfluidic Device 1. Loosen the plastic connector of a 90∘ bent blunt tip luer lock
needle by gently squeezing the connector using flat pliers
(Fig. 36.2b). Pull the metal tubing from the connector
(Fig. 36.2c, d). Remove any residual glue from the metal
tube using pincer pliers (Fig. 36.2e, f, see Note 11).
2. Cut microfluidic tubing for vacuum connection (10 cm), flu-
idic inlets (80 cm), and fluidic outlet (20 cm) using scissors.
3. For each microfluidic tubing insert a bend metal tube (from
step 1) at one end.
4. For vacuum tubing and inlet tubings insert straight blunt tip
luer lock needle into the other end (see Note 12).
5. Sterilize microfluidic tubing by flushing it with 1 mL of 70%
ethanol using a syringe with a luer lock tip.
6. Prepare yeast growth media as usual and filter sterilize using
disposable bottle top filters (see Note 13).
 The CellClamper: A Convenient Microfluidic Device for Time-Lapse. . . 543

Fig. 36.2 Fabrication of microfluidic connectors. (a) A 90∘ bent blunt tip luer lock needle. (b) Gently squeeze
the seam where the metal tubing is glued into the connector using flat pliers. (c) Pull the metal tubing from the
connector. (d) Extracted metal tubing. (e) Remove residual glue from the metal tubing using pincer pliers. (f)
Ready to use metal tube

7. Disassemble the syringe you want to use as a media reservoir,

by removing the piston from the syringe body. For setting up
the experiment using syringe pumps (see Note 14).
8. Attach microfluidic tubing to the syringe using the luer lock
connector.
9. Place syringe in a tube rack.
10. Fill media into syringe and close top of the syringe using
aluminum foil.
11. Let media flow through the tubing until all residual ethanol in
the tubing is removed. Close tubing using a foldback clip.
12. Store end of microfluidic tubing in a paper towel soaked with
70% ethanol.
13. Transfer rack with syringe(s) to microscope (Fig. 36.3).

3.3 Setting Up a Wear cleanroom gloves, protective glasses, and labcoat.

Time-Lapse Imaging
1. In a fume hood, clean the PDMS chip from all sides using
Experiment Using the
acetone from a squeeze bottle. Rub the structured side of the
Microfluidic Device chip using your fingers to remove any PDMS particles from the
fabrication process. Finally, hold the chip at its sides using
thumb and forefinger and give it a final rinse (see Note 15).
2. Rinse the chip with isopropanol and water from squeeze bot-
tles, while holding it between thumb and forefinger (see
Note 15).
3. Dry the PDMS chip by blowing air from an air spray gun onto
the chip surface. To avoid particle contamination from pressur-
ized air supply, mount a sterile filter on the air spray gun. Hold
544 Gregor W. Schmidt et al.

Media reservoir
Environmental chamber

80cm 20cm
Microfluidic chip

Coverslip
∆h
x40
1.3 NA

Camera

Excitation
light source
Waste container

Fig. 36.3 Overview of microfluidic chip setup. Media is delivered by gravity flow
from a media reservoir. The flow rate is determined by the resistance of the
microfluidic system (tubing and microfluidic chip) and the height difference (Δh)
between the media reservoir liquid level and the end of the outlet tubing
(seeNote 27). The media reservoir should be located outside the environmental
chamber to prevent formation of bubbles and limit evaporation

the chip against a light source to check if the cleaning proce-

dure was successful. If the chip shows traces of dust of particles,
repeat previous steps (see Note 16).
4. Store chip with the structured side facing up on a cleanroom
paper in a clean environment (e.g., cell culture hood).
5. Clean cover slip by rinsing with acetone, isopropanol, and
water from spray bottles. Hold the coverglass between thumb
and forefinger. Dry cover slip using air spray gun and check
cleanliness against light source. Store clean coverglass on a
cleanroom paper.
6. Pipette 0.4 μL of cell solution (1  107 cells/mL) into the
center of each culturing chamber of the microfluidic chip
(Fig. 36.4). The PDMS chip is hydrophobic and the cell sus-
pension will form a standing droplet on the chip surface (see
Notes 17 and 18).
7. Position the coverglass  3 mm above the microfluidic chip
surface. Bring the coverglass into contact with the PDMS on
one end of the microfluidic chip. Suspend the coverglass on the
other end using a pipette tip. Slowly lower the coverglass and
let it adhere to the PDMS over the whole chip surface.
Steps 6–7 should be performed in <1 min to prevent signifi-
cant evaporation of cell media (see Notes 19 and 21).
 The CellClamper: A Convenient Microfluidic Device for Time-Lapse. . . 545

Fig. 36.4 Loading of cells into the microfluidic device. (a) 0.4 μL cell solution is pipetted into each culturing
chamber (Step 6). (b) A coverglass is brought into contact with the PDMS chip (Step 7). The cell solution
spreads in the culturing chamber. (c) The vacuum tubing (top), inlet (left) and outlet (right) tubing is connected
to the chip (Step 13) and perfusion is started. The microfluidic device fills with media. Trapped air bubbles will
be removed by diffusion of air through the PDMS to the vacuum channel

8. Pick up the PDMS-coverglass sandwich and place it glass-side

down in a polystyrene Petri dish to transfer it to the microscope
(see Note 20).
9. (Optional) If available preheat environmental chamber of
microscope to desired temperature  2 h before starting the
experiment, to allow thermal equilibration of the equipment.
10. Place vacuum pump in vicinity of the microscope. From the
vacuum pump a tubing with a luer connector at its end should
be laid into the microscope and fixed with self-adhesive paper
tape.
11. The outlet tubing should be fixed on the microscope stage
using self-adhesive paper tape. The end of the tubing should
be placed into a waste container and fixed with self-adhesive
paper tape (see Note 22).
12. Place the microfluidic chip on a flat part of the microscope
stage.
13. Connect the microfluidic tubing by pressing the metal con-
nectors into the access holes of the microfluidic chip. Connect
tubings in the following order (see Note 23):
(a) vacuum tubing
(b) outlet tubing
(c) inlet tubing(s) (see Note 24).
14. Carefully place the microfluidic chip into the microscope slide
holder (see Note 25).
15. Remove the foldback clip from the media reservoir with which
you want to perfuse the cells (see Notes 26–28).
16. Place microfluidic chip in stage holder and fix microfluidic
tubings on the microscope stage using self-adhesive paper
tape. Try to fix tubing without creating tension on the micro-
fluidic device (see Note 29).
546 Gregor W. Schmidt et al.

17. Fix the stage holder on the microscope stage using self-
adhesive paper tape (see Note 30).
18. (Optional) If using an immersion objective, raise the objective
until it is located  1 mm below the coverglass. Move the
microscope stage so that the objective is at the edge of the
coverglass. Add immersion liquid to the gap between the cov-
erglass and the objective (see Note 31).
19. Move the microscope stage so that the objective is located
below a cell culture chamber. Bring sample into focus by slowly
moving the objective upwards and observing the sample
through the eyepiece or the camera live-stream. The cells
should show a strong membrane diffraction pattern when
they are slightly out-of-focus (see Note 32).
20. Fix the microfluidic device inside the stage holder by applying
one drop of nail polish to each corner of the coverglass. Leave
the doors of the environmental chamber open for  30 min to
let nail polish solvent escape (see Note 33).
21. Use the adjustment screws of the stage holder to level the
microfluidic chip with the imaging plane (see Note 34).
22. Leave microfluidic device to thermally equilibrate for at least
30 min (see Note 35).
23. Program multi-position time series measurement using the
microscope software. If available enable hardware or software
autofocussing routine, to keep sample in focus over the whole
course of the experiment (see Note 36).
24. After the experiment, disassemble the microfluidic device by
closing the media lines with foldback clips. Gently remove all
paper tape and pull out all tubings from the microfluidic chip.
Rinse microfluidic tubings with ultrapure water and 70% etha-
nol and store in a Petri dish until further use. Disassemble the
microfluidic device by gently peeling the PDMS chip from the
coverglass (see Note 37). Vigorously rinse the PDMS chip
using ultrapure water to remove cells and blow dry using the
air spray gun. Inspect cleanliness under a microscope. Store the
PDMS chip in a Petri dish until further use.

3.4 Live-Cell To illustrate the possibility of inducing DNA damage in the micro-
Microscopy of DNA fluidic device, we cultured strains containing a fluorescent protein
Damage Induced by fused to either Mre11, Ddc2, or Rad52. The cells were grown until
Chemicals or CRISPR/ they formed a small colony and DNA damage was induced by
Cas9 Induction switching the media. First, we cultivated the cells using low fluores-
cent SD media and induced DNA damage using 0.03% MMS. The
relocation pattern of all three proteins were as reported in the
literature (Fig. 36.5). As the culture pads of the microfluidic chip
reduce the background fluorescence by excluding media from the
fluorescence excitation volume, the chip allows to capture images
 The CellClamper: A Convenient Microfluidic Device for Time-Lapse. . . 547

Fig. 36.5 Detection of DNA damage by imaging of fluorescent fusion proteins in the microfluidic chip. DNA
damage was induced chemically by perfusing media containing either 0.03% methyl methanesulfonate (MMS)
or 0.2 M hydroxyurea (HU). Site specific double-strand breaks were generated at the URA3 locus using an
inducible CRIPRS-Cas9 construct (unpublished). DNA damage was detected by nuclear localization of
fluorescent signals in strains carrying fluorescent protein fusions with Mre11, Ddc2, or Rad52. Images are
overlays of brightfield and fluorescent channels. The dynamic range of the fluorescence channel has been
equally adjusted for each pair of images (induced image and non-induced control image), to make the signal
visible

with high signal-to-noise ratio even in strongly autofluorescent

media. We cultured cells in YPD and induced DNA damage by
either applying 0.3 M hydroxyurea or introducing a single double-
strand break using an inducible CRISPR-Cas construct (unpub-
lished) targeting the URA3 locus. In all cases, relocation of the
fluorescent protein fusions was swift and well detectable
(Fig. 36.5). This demonstrates that using our device allows micros-
copy assays to be performed independent of the autofluorescent
properties of the media.
Next, we wanted to illustrate how dynamic exposure of cells to
genotoxic agents can be monitored using the microfluidic device.
We loaded cells carrying an Mre11 fluorescent protein fusion and
cultured them in YPD for  2 h until they formed small colonies.
We then exposed the cells to a pulse of YPD + 0.2 M hydroxyurea
for 16 h, and finally released genotoxic stress by switching back to
YPD. The formation of nuclear Mre11 foci was scored by calculat-
ing the change of the nuclear fluorescence interquartile range
(Fig. 36.6a). Foci become visible in a small subset of cells within a
short period of time and after 2 h the majority of the cells show
nuclear foci and stop dividing. Within  1 h after hydroxyurea is
removed, nuclear foci disappear and division restarts for a subpop-
ulation of cells. Newborn cells exhibit no nuclear foci. However, for
548 Gregor W. Schmidt et al.

A 0.1

Mre11 foci index [AU]

0.08 YDP + 0.2 M hydroxyurea YPD 0.15

Frequency [%]
0.06
0.1
0.04
0.05
0.02

0 0
0 4 8 12 16 20 24
Time [h]
B 1.4
Ratio GFP/tdTomato [AU]

0.3
YDP + 0.2 M hydroxyurea YPD
0.25

Frequency [%]
1.2
0.2

0.15
1
0.1

0.05
0.8
0 4 8 12 16 20 24
Time [h]

Fig. 36.6 Imaging DNA damage in an arrest-and-release experiment using

hydroxyurea. Cells were loaded into the microfluidic device and cultured in
YPD for 2 h. To induce DNA damage, the cell was exposed to a pulse of
YPD + 0.2 M hydroxyurea for 16 h. After the pulse the media was switched
back to YPD. Images were acquired in 5 min intervals. (a) A strain carrying an
Mre11 fluorescent fusion protein was imaged. The Mre11 foci index was
calculated for each single cell from the interquartile range of the nuclear
fluorescent fusion protein signal ((nucq75 nucq25)/nucq75). At each timepoint
the normalized histogram of the resulting foci indexes is plotted. (b) A strain
carrying an Rnr3-GFP fusion protein and an Rpl39pr-tdTomato expression
construct was imaged. For each cell, the median Rnr3-GFP signal was divided
by the median signal of the constitutively expressed Rpl39pr-tdTomato con-
struct. The normalized histogram of the resulting values is plotted for each
timepoint

many old cells nuclear foci remain present (visible at the upper right
corner of the plot). As the fraction of newborn cells in the field of
view increases, the average Mre11 nuclear foci index returns to the
same level as before exposure to hydroxyurea.
To illustrate the behavior of a protein abundance reporter, we
repeated the hydroxyurea arrest-and-release experiment using the
parent strain from a genome wide screen for genomic instability. In
this strain, expression of an Rnr3-GFP fusion protein is induced in
 The CellClamper: A Convenient Microfluidic Device for Time-Lapse. . . 549

response to DNA damage. To account for cell-to-cell variability, the

strain also contains a constitutively expressed Rpl39pr-tdTomato
reporter construct, which is used to normalize the Rnr3-GFP
signal. The onset of Rnr3-GFP expression starts about 2 h after
induction and reaches a plateau after 5.5 h. Upon removal of
hydroxyurea, the GFP signal decays as a function of the division
time. As restart of division is not instantaneous, there is a slight
delay until the GFP-signal starts to vanish. Taken together, these
experiments illustrate the usability of the device for imaging yeast
cells under changing media conditions.
In all cases, the cells were cultivated in an environmental cham-
ber at 30∘ and images were acquired using a Nikon Ti-E micro-
scope equipped with a 40 objective (oil immersion, NA 1.3). The
microscope was controlled using the software YouScope [14]. For
accurate cell segmentation, brightfield images were taken above
and below the focal plane (5 AU, Nikon Perfect Focus System)
and divided using ImageJ. Division of brightfield images corrects
illumination and background aberrations, and enhances the diffrac-
tion pattern of the cell boundary. The divided images were seg-
mented and quantified using CellX [15]. The extracted intensities
were plotted using Matlab.

4 Notes

1. Always use cleanroom gloves when handling microfluidic com-

ponents. Conventional gloves (even powder free ones) contain
particles which can cause background fluorescence, blocking of
the microfluidic channels and interfere with adhesion of PDMS
to glass.
2. Design files for microfluidic devices are available upon request.
If a cleanroom with SU-8 processing capabilities is available,
wafers can be fabricated and silanized as described [16]. If
microfabrication facilities are not available, silanized custom
wafers can be ordered from a microfabrication company, e.g.,
l MicruX Fluidic (Severo Ochoa Building, Julian Claveria s/
n, 33006 Oviedo (Asturias), SPAIN)
l SIMTech Microfluidics Foundry (Singapore Institute of
Manufacturing Technology, 71 Nanyang Drive, Singapore
638075)
l FlowJEM (http://www.flowjem.com/)
3. Store wafers in a cool, dry, and dark place. A humid atmosphere
will compromise the silanization of the wafer. Exposure to heat
and sunlight will be detrimental for adhesion of the SU-8 on
the wafer.
550 Gregor W. Schmidt et al.

4. The microfluidic device can be re-used as often as  30 times.

However, hydrophobic chemicals can be absorbed by the
PDMS of the microfluidic chip and the tygon tubing. There-
fore, care has to be taken if parts of the fluidic setup are re-used
to exclude cross-contamination from previous experiments.
5. Always handle the wafer by only touching it on its edges. Never
touch or attempt to clean the wafer surface using solvents. Dust
that cannot be blown off using the air spray gun will get
embedded in the PDMS slab and removed when demolding,
leaving a clean wafer surface.
6. Immediately clean spilled PDMS using a wipe soaked with
isopropanol. Cured PDMS is not soluble, and has to be
removed physically.
7. Take care to stop the vacuum pump before PDMS foam spills
over the edge of the plastic beaker. Place a piece of aluminum
foil in the desiccator to contain possible PDMS spills.
8. Take care to stop vacuum pump before PDMS foam spills over
the edge of the Scotch tape.
9. If a convection oven is not available, place a Petri dish onto a
level surface and leave for 24 h at room temperature to cure
PDMS.
10. Take care to punch holes perpendicular to the chip surface.
Otherwise it will be difficult to connect the microfluidic tubing
later.
11. Try not to squeeze or bend the metal tubing itself with the
pliers, as this might collapse the channel inside the tubing and
block it.
12. Store tubing in a Petri dish if they are not to be immediately
used.
13. Filtering is carried out to prevent clogging of the microfluidic
device due to particles which might be present in convention-
ally prepared media.
14. If the microfluidic device is to be perfused using syringe
pumps, follow these instructions. Syringe pumps allow precise
control of the flow rate at the different inlets. Using syringe
pumps media can be mixed on-chip and cells can be exposed to
precisely timed trajectories of different media conditions. Glass
syringes should be used with syringe pumps, as they can be
more rigidly connected to the syringe pump holder than plastic
syringes. Also, glass syringes have faster response times and
show less hysteresis when the flow rates are changed.
(a) In a sterile hood, disassemble the glass syringes by pulling
out the plunger.
The CellClamper: A Convenient Microfluidic Device for Time-Lapse. . . 551

(b) Spray the syringe body and the plunger with generous
amounts of 70% ethanol from a spray bottle, to sterilize
the syringe. Leave the parts to dry in the sterile hood for 
30 min.
(c) When the syringe is completely dry, reassemble it and
push the plunger all the way into the syringe body. Con-
nect a sterile filter (pore size: 0.22 μm) to the syringe.
(d) Through a microfluidic tubing, draw media through the
filter into the syringe by pulling on the plunger until the
syringe is full.
(e) Bubbles will be present at the plunger and wall inside the
syringe. Hold the syringe with its tip is facing upwards
while holding it on the plunger. Using your thumb and
forefinger quickly rotate the syringe along its longitudinal
axis, to release bubbles from the plunger and wall, which
will move upwards towards the tip of the syringe.
(f) Once all bubbles have been collected at the tip of the
syringe, remove the sterile filter and connect a freshly
sterilized microfluidic tubing, while keeping the syringe
tip facing upwards. Push the plunger to eject all bubbles
from the syringe, until the microfluidic tubing is filled
completely filled with growth media.
(g) Store end of microfluidic tubing in a paper towel soaked
with 70% ethanol.
(h) Transfer syringe to microfluidic syringe pump located at
the microscope.
(i) Syringe pumps should be located outside the environ-
mental chamber of the microscope, since heating of the
media inside the syringes will lead to out-gassing of dis-
solved air and formation of bubbles inside the syringes.
Bubbles inside the syringes should be avoided to allow fast
and precise control of the flow rates.
(j) The microfluidic device should be perfused with
5–100 μL/min. Faster flow rates allow quicker media
exchange and switching, and facilitate efficient removal
of excess cells. At slower flow rates the limited syringe
volume can be used for longer experiments. At flow
rates >100 μL/min the cells underneath the culturing
pads get flushed away by the media flow.
15. Avoid letting the solvents run from your gloves onto the chip,
as it will leave traces which interfere with chip adhesion to the
cover slide.
16. Try not to let water dry on the chip, but rather blow it off the
chip, as it might leave traces.
552 Gregor W. Schmidt et al.

17. The cell concentration determines the number of cells per

culturing pad in the microfluidic chip. Higher cell concentra-
tions (up to 10–20  107 cells/mL) can be used if cells are to
be cultured for a short time only (1–4 h) and more cells are
needed for the analysis. For loading of single cells, cells should
be in the exponential phase of growth. Determine cell concen-
tration of yeast culture using a Coulter Counter, counting
chamber, or calibrated photometer. To achieve the desired
cell concentration transfer 1 mL of cell solution into a
1.5 mL Eppendorf tube. Sediment cells by centrifuging the
tubes for 2 min at 1000g. Remove enough supernatant to
reach desired cell concentration after resuspending the cell
solution. Take care not to remove cells when removing super-
natant. Resuspend cells by carefully flicking the tube or by a
short vortexing pulse at a low speed setting. Depending on the
culture conditions and growth phase cells might sediment as a
pellet, or might float at the top of the media solution. If cells
are used at later cultivation stages they might form clumps and
get loaded as cell colonies. Also, cells in later phases of cultur-
ing might stick to PDMS strongly.
18. Use low adhesion pipette tips to allow efficient transfer of the
whole cell culture volume to the microfluidic chip surface.
Depending on the pipette, pipette tip, and media composition
the actual volume that gets transferred to the chip might vary.
Adjust the setting on the pipette accordingly. The cell culture
chambers should not be completely filled with cell suspension,
as this can lead to cross-contamination of strains between
chambers.
19. If culture chambers are filled <50% the pipetted volume was
too small or evaporation during chip loading was too high.
Practice loading of the device.
20. Always handle the assembled chip by holding it between two
fingers on the coverglass. Avoid any bending of the microflui-
dic device, as this can lead to adhesion of the culturing pads to
the coverglass.
21. The chip structure should be clearly visible if the PDMS-glass
sandwich is held against a light source. Particle or dust con-
tamination is directly visible as defects in the glass-PDMS
interface. If contamination is observed disassemble PDMS
and coverglass. Discard the coverglass. Clean the PDMS chip
and a new coverglass as described earlier.
22. Make sure the outlet tubing will not escape the waste container
during stage movement, as this will cause spillage on the optical
table. The outlet tubing should have enough slack, so it can be
comfortably connected to the microfluidic chip without ten-
sion. The end of the outlet tubing should be  5 cm below the
stage level.
The CellClamper: A Convenient Microfluidic Device for Time-Lapse. . . 553

23. Immediately before pressing the tubing into the access holes,
sterilize them using a Kimtech wipe soaked with 70% ethanol.
The ethanol will evaporate immediately and leave a thin film of
water on the metal pin, which facilitates insertion of the tubing
into the PDMS. Position the microfluidic tubing above the
access hole with one hand. With the other hand, gently press
on top of the bent metal tube to insert it into the PDMS. Try
not to touch the PDMS chip itself, as the PDMS might adhere
to the coverglass in the regions of the culturing pads.
24. If several media reservoirs are to be connected to the chip,
connect the media reservoir which is to be used at the begin-
ning of the experiment, first. Then remove the foldback clip
from this media reservoir and let the media flow into the
microfluidic device until it starts to flow out of the other inlet
holes. Then close the media reservoir with the foldback clip
again and connect other media reservoirs. When all reservoirs
are connected proceed with next step.
25. Make sure the cover slide fits into the slide holder without
tension. Tension on the coverglass leads to bending of the
microfluidic device, which impedes alignment of the sample
with the imaging plane of the microscope.
26. Make sure the liquid level of the reservoir is 5 cm above the
microfluidic chip, to allow media flowing into the chip. The
flow rate during filling of the microfluidic device is higher
because an empty system has a lower resistance. If the flow
rate during filling is too high, delamination of the chip and
leakage into the vacuum channel is possible.
27. The sum of the length of the microfluidic inlet and outlet
tubing determines the flow rate through the microfluidic
device. For a tubing length of 100 cm (80 cm inlet tubing +
20 cm outlet tubing) and a height difference of 10 cm
between the reservoir liquid level and the end of the outlet
tubing, the flow rate will be  10 μL/min. The flow rate scales
linearly with tubing length and height difference.
28. If you observe media leaking into the vacuum channel, close
the inlet tubing again. In this case the chip was not cleaned
thoroughly enough and a particle contamination has created a
connection between the media channel and the vacuum chan-
nel. Disassemble the chip and repeat the loading procedure.
29. If there is tension on the tubing, the microfluidic chip might
delaminate from the coverglass which will result in leakage of
media into the vacuum channel. Also, tension on the tubing
may cause bending of the microfluidic device (see Note 25)
30. Usually the slide holder will have some play when it is inserted
in the microscope stage. To prevent moving of the slide holder
during time-lapse imaging, it needs to be fixed with adhesive
tape.
554 Gregor W. Schmidt et al.

31. Be generous with the amount of immersion liquid. When

performing multi-position imaging, the objective will move
over extended areas of the microfluidic device (distances of
20–30 mm), and the immersion liquid might not be able to
follow the objective if the volume is too low.
32. If cells do not show a strong diffraction pattern, the cell culture
pads have collapsed onto the coverglass (probably while the
tubings were inserted). In this case disassemble the microflui-
dic device and repeat loading procedure.
33. Use only clear nail polish, as colored nail polish can contain
fluorescent molecules which can increase background fluores-
cence during the measurement.
34. Move the stage so that one culture pad with a high number of
cells is in the field of view. Move the objective downwards to
slightly de-focus the sample. Then, slowly move the objective
upwards and observe on which part of the field-of-view the
cells come into focus first. The adjustment screws on this side
of the chip have to be screwed inwards. Repeat the procedure
until all cells come into focus simultaneously.
35. In the first 30 min the microfluidic device and the stage holder
might still move 50–200 μm, which is detrimental for time-
lapse imaging.
36. If immersion liquid or hardware autofocus cannot follow the
stage movement during multi-position imaging, reduce the
stage speed. Make sure that all positions can be imaged in the
desired imaging interval.
37. Take care not to break the coverglass, as the shards might
damage the PDMS structure. Discard the coverglass.

Acknowledgements

This work was supported by the FP7 grant no 28995 ITN ISO-
LATE and the Ambizione grant 142440 of the Swiss National
Science Foundation to Olivier Frey. We thank Andreas Hierlemann
and Jörg Stelling for their support. We acknowledge Mjriam
B€achler and Andreas Cuny for testing the microfluidic device and
Jennifer Ewald and Tania Roberts for critical reading of the
manuscript.

References
1. Lisby M, Rothstein R, Mortensen UH (2001) 2. Lisby M, Mortensen UH, Rothstein R (2003)
Rad52 forms DNA repair and recombination Colocalization of multiple DNA double-strand
centers during S phase. Proc Natl Acad Sci breaks at a single Rad52 repair centre. Nat Cell
USA 98(15):8276–8282. doi:10.1073/ Biol 5(6):572–577. doi:10.1038/ncb997
pnas.121006298
 The CellClamper: A Convenient Microfluidic Device for Time-Lapse. . . 555

3. Lisby M, Barlow JH, Burgess RC, Rothstein R replication in mammalian cells. PLoS One 7
(2004) Choreography of the DNA damage (9). doi:10. 1371/journal.pone.0045726
response: spatiotemporal relationships among 10. Bush A, Chernomoretz A, Yu R, Gordon A,
checkpoint and repair proteins. Cell 118 Colman-Lerner A (2012) Using cell-ID 1.4
(6):699–713. doi:10.1016/j.cell. with R for microscope-based cytometry. Curr
2004.08.015 Protoc Mol Biol (Suppl 100):1–27.
4. Silva S, Gallina I, Eckert-Boulet N, Lisby M doi:10.1002/0471142727.mb1418s100
(2012) Live cell microscopy of DNA damage 11. Duncombe TA, Tentori AM, Herr AE (2015)
response in Saccharomyces cerevisiae. Methods Microfluidics: reframing biological enquiry.
Mol Biol 920:433. doi:10.1006/ Nat Rev Mol Cell Biol 16(9):554–567.
cbir.1999.0447. arXiv:1011.1669v3 doi:10.1038/ nrm4041
5. Tkach JM, Yimit A, Lee AY, Riffle M, Costanzo 12. Dénervaud N, Becker J, Delgado-Gonzalo R,
M, Jaschob D, Hendry JA, Ou J, Moffat J, Damay P, Rajkumar AS, Unser M, Shore D,
Boone C, Davis TN, Nislow C, Brown GW Naef F, Maerkl SJ (2013) A chemostat array
(2012) Dissecting {DNA} damage response enables the spatio-temporal analysis of the yeast
pathways by analysing protein localization and proteome. Proc Natl Acad Sci USA 110
abundance changes during {DNA} replication (39):15842–15847. doi:10.1073/
stress. Nat Cell Biol 14(9):966–976. pnas.1308265110
doi:10.1038/ncb2549 13. Frey O, Rudolf F, Schmidt GW, Hierlemann A
6. Gasch AP, Huang M, Metzner S, Botstein D, (2015) Versatile, simple-to-use microfluidic
Elledge SJ, Brown PO (2001) Genomic cell-culturing chip for long-term, high-
expression responses to DNA-damaging agents resolution, time-lapse imaging. Anal Chem 87
and the regulatory role of the yeast ATR homo- (8):4144–4151. doi:10.1021/ac504611t
log Mec1p. Mol Biol Cell 12(10):2987–3003. 14. Ferry MS, Razinkov IA, Hasty J (2011) Micro-
doi:10.1091/mbc.12.10.2987 fluidics for synthetic biology: from design to
7. Birrell GW, Brown JA, Wu HI, Giaever G, Chu execution. In: Methods in enzymology, vol
AM, Davis RW, Brown JM (2002) Transcrip- 497. Elsevier, Amsterdam, pp 295–372
tional response of Saccharomyces cerevisiae to 15. Lang M, Rudolf F, Stelling J (2012) Use of
DNA-damaging agents does not identify the YouScope to implement systematic microscopy
genes that protect against these agents. Proc protocols. In: Current protocols in molecular
Natl Acad Sci USA 99(13):8778–8783. biology, vol 14. Wiley, New York, pp 1–23.
doi:10.1073/ pnas.132275199 doi:10.1002/0471142727. mb1421s98
8. Johnson SA, You Z, Hunter T (2007) Moni- 16. Mayer C, Dimopoulos S, Rudolf F, Stelling J
toring ATM kinase activity in living cells. DNA (2013) Using CellX to quantify intracellular
Repair 6(9):1277–1284. doi:10.1016/j. events. In: Current protocols in molecular biol-
dnarep.2007.02. 025 ogy, vol 14, pp 1–20. doi:10.1002/
9. Burgess A, Lorca T, Castro A (2012) Quanti- 0471142727.mb1422s101
tative live imaging of endogenous DNA
 Chapter 37

Characterization of Structural and Configurational

Properties of DNA by Atomic Force Microscopy
Alice Meroni, Federico Lazzaro, Marco Muzi-Falconi,
and Alessandro Podestà

Abstract
We describe a method to extract quantitative information on DNA structural and configurational properties
from high-resolution topographic maps recorded by atomic force microscopy (AFM). DNA molecules are
deposited on mica surfaces from an aqueous solution, carefully dehydrated, and imaged in air in Tapping
Mode. Upon extraction of the spatial coordinates of the DNA backbones from AFM images, several
parameters characterizing DNA structure and configuration can be calculated. Here, we explain how to
obtain the distribution of contour lengths, end-to-end distances, and gyration radii. This modular protocol
can be also used to characterize other statistical parameters from AFM topographies.

Key words Atomic force microscope/microscopy (AFM), DNA, Mica, DNA conformation

1 Introduction

Across more than three decades, atomic force microscopy (AFM)

has become a technique of choice for the quantitative investigation
of biomolecules such as DNA, proteins, and their complexes (for an
overview, see [1] and references therein). The success of AFM relies
on its ability to provide nanometer spatial resolution in XY, sub-
nanometer resolution in Z, as well as on the capability of imaging
biological samples in their physiological conditions. Since the
advent of AFM, DNA has been the privileged target of innumerable
studies, due to its paramount biological relevance (for a review, see
[2]). Being a semirigid, charged, strong polyelectrolyte, DNA con-
centrates in itself a wealth of interesting physics, and has a great
potential for nanobiotechnological applications [3, 4]; for this
reasons, DNA has been an ideal benchmark for biophysical studies
[5–17]. Such studies mostly rely on the statistical characterization
of structural and configurational properties of a population of
double stranded DNA molecules with fixed length. Several

Marco Muzi-Falconi and Grant W. Brown (eds.), Genome Instability: Methods and Protocols, Methods in Molecular Biology,
vol. 1672, DOI 10.1007/978-1-4939-7306-4_37, © Springer Science+Business Media LLC 2018

557
558 Alice Meroni et al.

parameters describe the equilibrium configuration of DNA mole-

cules: contour length, end-to-end distance, rise per residue, radius
of gyration, bending angle distribution along the DNA backbone,
persistence length, to cite the most important. The statistical
mechanics of semirigid polymers describes the distribution of
these parameters in equilibrium conditions [5]. Structural and
configurational changes of DNA molecule can occur as a conse-
quence of modification of the DNA environment (ionic strength,
pH, surface charge density) [6, 7, 18], as well as a consequence of
internal changes at the base pairs level (such as mispaired and
damaged bases [19, 20]); it follows that high-resolution imaging
of DNA molecules, as provided by AFM, can be a very valuable
complement of biomolecular studies.
Here, we describe a general protocol for acquiring high-
resolution images of DNA molecules, either linear fragments or
plasmids, using an atomic force microscope operated in Tapping
(or intermittent contact ) Mode in air. In Tapping (or intermittent
contact) Mode, a sharp oscillating probe periodically touches the
surface under investigation [21]. During each gentle tap, lateral
forces are minimized, and this provides high spatial resolution and
overall noninvasiveness of the measurement. DNA must be well
attached to a smooth, flat substrate, so to obtain well-contrasted,
well-resolved topographic maps. At the same time, the sample
preparation procedure must preserve as much as possible the native
DNA characteristics, so to avoid trapping the molecules in out-of-
equilibrium configurations. We present all the steps required for
extracting quantitative information on the structural and configu-
rational properties of a population of DNA molecules: the prepara-
tion of DNA samples on mica surface; the imaging in air by AFM in
Tapping Mode; the preparation of images for the analysis and the
digitization of the DNA traces; the calculation of selected statistical
parameters describing the state of the system (contour length,
radius of gyration, and end-to-end distance). The approach here
described is rather general and modular, and can be easily imple-
mented to add the calculation of other statistical descriptors of the
DNA structural properties, such as the persistence length or the
bending angle distribution. The protocol here presented can be
easily adapted to other imaging modes and conditions.

2 Materials

1. Cyanoacrylate glue (such as Loctite 406 or similar).

2. Steel magnetic disks, diameter 11–15 mm, thickness
0.2–0.5 mm.
 Characterization of Structural and Configurational Properties of DNA. . . 559

3. Mica disks of the highest quality grade (V1, ruby muscovite),

with thickness 0.2–0.6 mm and diameter 6–12 mm. (see Notes
1 and 2).
4. Ethanol (laboratory grade,
95% v/v).
5. Eppendorf tubes, 1.5 mL.
6. Plastic tubes, 50 mL.
7. Pipettes with plastic tips (p20, p200, and p1000).
8. Scotch-like adhesive tape (such as the Magic Transparent
Tape).
9. Ultrapure MilliQ (typical resistivity ρ ¼18.2 MΩ cm at 25  C)
to prepare samples and solution for AFM imaging.
10. Nitrogen from a reservoir (purity
99.999%).
11. Blotting paper to remove drops of water after rinsing.
12. Tweezers.
13. AFM tips for dynamic (or Tapping) Mode (see Note 3).
14. Deposition buffer: 2–5 mM MgCl2, 10 mM NaCl, 10 mM
HEPES-Na pH 7.5 in MilliQ H2O (see Note 4).
15. Stock DNA in MilliQ water (at least 0.05 ng/μL).

3 Methods

3.1 Sample A mica disk must be glued to a rigid support, whose dimensions
Preparation and geometry can vary according to the instrument specifications.
In most cases this support can be a metal disk with diameter
3.1.1 Substrate
11–15 mm (at least 2 mm larger than the mica disk) and thickness
Preparation
0.5–1 mm (see Note 5).
1. Clean the mica and the metallic surfaces with ethanol and dry
all surfaces using blotting paper.
2. Glue the mica disk to the steel disk using cyanoacrylate-based
adhesive (see Note 6).
3. Prepare at least 3–4 similar substrates, to be able to prepare
many samples in parallel for quickly testing different deposition
conditions.

3.1.2 DNA Deposition on 1. Freshly cleave the mica surface to be used for DNA deposition
Mica (see Note 7).
2. Dilute DNA to a final concentration of 0.05 ng/μL in the
deposition buffer (so to have typically a 1–2 nM DNA solution)
(see Note 8).
3. Put a 15–20 μL drop of diluted DNA solution on the freshly
cleaved mica (see Note 9). Avoid touching the mica surface
with the pipette tip.
560 Alice Meroni et al.

4. Incubate at room temperature for 2–10 min (longer incubation

times will provide higher surface densities of DNA molecules).
5. Gently rinse the sample dropwise several times, using each time
up to 1 mL of MilliQ water to remove the exceeding and
loosely bound molecules. During the rinsing procedure, keep
the sample tilted to help water slipping away.
6. Blot the remaining water placing the corner of a piece of blot
paper next to the lower mica disk edge. Make sure the paper is
not in contact with the surface.
7. Dry the sample under a gentle stream of clean nitrogen from a
reservoir. Set a mild stream flux and keep the nozzle a few
centimeter away from the surface (see Note 10).
8. The DNA sample is ready to be imaged (see Note 11 for
troubleshooting).

3.2 AFM Imaging in AFM imaging is aimed at collecting a statistically meaningful num-
Air in Tapping Mode ber of high-quality, well-contrasted, high-resolution images of
DNA molecules, which must then be digitized so to produce a
large collection of molecular traces to be further analyzed. The
general requirements are:
1. to collect in a single scan a reasonable number of molecules, so
to obtain a good statistical sample (several hundred molecules)
within 5–10 images;
2. to have a good sampling resolution overall in each image, in
order to calculate accurate values of conformational parameters
from the molecular traces.
Additional information can be found in Note 12.
Typical surface density of molecules on good samples is
15–40 molecules/μm2, depending on the DNA length. Imaging
of DNA samples in air can be performed in dynamic mode (usually
called Tapping Mode, intermittent-contact or oscillating mode).
Recently, new imaging modes based on a vertical tip-sample
approach have been developed, which provide accurate control of
applied force, low-invasiveness in air as well as in liquid and high
spatial resolution (described briefly in Note 13); these modes could
represent an alternative to the Tapping Mode, that we will consid-
ered hereafter.
We suggest the following scanning parameters and conditions
for imaging DNA in air in Tapping Mode with a well-calibrated
instrument (see Note 14 on AFM calibration):
1. Mount a rigid cantilever for dynamic or Tapping Mode.
2. Set a free oscillation amplitude (target amplitude) of 10 nm or
less. Smaller amplitudes provide high-quality and less-invasive
imaging conditions, but require very clean, nonadhesive
surfaces.
 Characterization of Structural and Configurational Properties of DNA. . . 561

3. Initially, after engaging the tip on the sample surface, adjust the
minimal amplitude (force) setpoint to track the surface while
keeping the scan size to <1 nm.
4. Initially, set a small scan size (100–500 nm) and optimize the
gains and the amplitude setpoint to achieve optimal tracking
conditions at the lowest applied force (see Note 17). Good
tracking is witnessed by a good overlap of topographic profiles
in both scan directions. Check the tracking in correspondence
of points where a sudden change in surface slope occurs (typi-
cally, at the mica–DNA border). Poor tracking results in a loss
of contact when the tip is crossing the DNA molecule downhill.
Increase gradually the scan size up to 2 μm  1 μm (aspect-ratio
2:1, see Note 15).
5. Set the sampling resolution to 2048  512 points (number of
points per line  number of lines). This choice provides a
sampling resolution of 1 nm/pixel and 2 nm/pixel in the fast
and in the slow scan directions, respectively.
6. Set the scan rate to 1–4 Hz (see Note 16).
7. We suggest acquiring up to five images in each location,
according to the simple scheme described below, then to with-
draw the tip and engage some 100–500 μm away. Three–four
different macroscopic locations will provide several hundred
molecules for the statistical analysis. The image acquisition
scheme is the following: in each location, acquire the first
image with no offsets in X and Y directions, at (0,0). Acquire
the other images at points (X0,0), (X0,0), (Y0,0), (Y0,0),
set X0 ¼ 3 μm, Y0 ¼ 2 μm, so to avoid overlap among the scan
areas (see Note 18).
8. In order to maintain stable imaging conditions, in particular to
minimize capillary adhesion at the tip–sample interface, the
AFM head and the sample with the scanning stage can be
hosted in a small chamber, inside which a dry N2 atmosphere
is maintained, with relative humidity below 5%.
Figure 1a shows a typical topographic map of DNA molecules
(727 bp) on mica, imaged in air in Tapping Mode, according to the
described methodology.

3.3 Data Analysis Images must be prepared for the analysis, by removing standard
(See Note 19) artifacts related to the image formation process and by removing
high-frequency noise, typically related to the feedback loop opera-
3.3.1 Image
tion. Artifacts typically manifest themselves as baselines superim-
Preprocessing
posed to the true topographic profiles. Identification and
subtraction of the baseline from each topographic profile is essential
for the accurate analysis of the AFM images. Details on the origin of
artifacts and on the baseline subtraction procedures are provided in
Note 20.
562 Alice Meroni et al.

Fig. 1 727 bp DNA molecules on mica and details of the single-molecule analysis of AFM images. (a) Top-view
AFM image showing molecules equilibrated on a mica surface, after gentle dehydration of the sample (the
vertical range is 1 nm; heights increase from dark to bright colors). (b) A representative DNA molecule with
highlighted the relevant parameters used for the characterization of configurational properties, as described in
Subheading 3.3

1. Apply a global plane-fitting of the first order to the image, so to

get rid of the global tilt of the sample. This operation will level
the height values in the image, offering a complete overview of
the molecules and of the overall image quality.
2. Apply a line-by-line flattening to the image by subtracting
higher-order polynomials (usually up to the third order is
enough) through suitable masks excluding features that do
not belong to the reference substrate.
3. Check the quality of the flattening by looking at the height
histogram of the image: a sharp Gaussian peak centered around
the average height z0 of the substrate (typically z0 ¼ 0 nm) with
FWHM of a few Å should be present; DNA molecules will
typically contribute a broader short tail in the height
distribution.
4. Apply a median filter (with 3  3, maximum 5  5 kernel) to
the flattened image, to smooth high-frequency noise.

3.3.2 Tracing DNA Once the images have been flattened and smoothed, the set of
Molecules spatial coordinates {xi,yi}i¼1:N defining the backbone of each mole-
cule for each particular experimental condition (i.e., for each par-
ticular length of the DNA molecules) must be determined. Here N
is the total number of molecules in all the AFM images that can be
used for the analysis. The tracking can be done manually or by
means of (semi)automatic algorithms; some of them are freely
available upon request to the developers (see Note 21). Here we
describe how to manually trace the molecules using ImageJ/Fiji, an
open-source software written in Java and supported by a broad
community of scientists.
 Characterization of Structural and Configurational Properties of DNA. . . 563

1. Export topographical maps in a format compatible with the

ImageJ software (i.e., .tif). Resize the image to include the
scanned area only.
2. Import the AFM image into ImageJ/Fiji and define image size
and sampling resolution.
(a) Define the image size using the command: Image !
Scale, and set width and height (in pixels).
(b) Define the scale using the command: Analyze ! Set Scale,
and set the scale (in nanometer).
3. Use the “Segmented Line” command to draw by hand the
backbone of the molecule. Keep possibly a constant distance
between each point that should not be too far or too close from
each other. Consider to maintain a point-to-point distance
from 1/2 to 2/3 of the molecule width.
4. Save the molecule trace using the command: “Save As XY
Coordinates”. With this operation ImageJ/Fiji saves an
ASCII (.txt) file containing the XY coordinates of the selected
track, columnwise.
An example of the manual tracking of a DNA molecule is
shown in Fig. 1b.

3.3.3 Basic Structural Once the spatial coordinates of each molecule have been obtained,
Analysis several statistical parameters, describing the structural and configu-
rational properties of DNA in the studied conditions, can be deter-
mined (see among the others [5, 6, 8, 9, 11, 15, 22]). First, the
contour length of DNA depends on the conformation adopted.
Generally, DNA assumes the B-form in physiological conditions,
but it can adopt other different conformations, like A- and Z-form.
Moreover, DNA can display more open or compact configurations
depending on its persistence length, which in turn is sensitively
dependent on the ionic strength of the buffer and on the nature
of the surrounding ions. Here we focus our attention on a selection
of parameters that can be readily obtained by the XY coordinates:
the contour length Lc, the end-to-end distance R, the gyration
radius Rg [5, 8, 22, 23]. All these parameters depend on DNA
form, base-pair composition, and persistence length; therefore,
their accurate statistical determination can provide important infor-
mation on the structural properties of a DNA population under
study. Typically, the configurational parameters are evaluated from
digitized traces, and then the average values with standard devia-
tions are calculated from the distributions of these parameters.
1. Calculate Lc, R, and Rg according to the following equations
(refer to Fig. 1b; see also Note 22):
564 Alice Meroni et al.

X
N 1 X
N 1 qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
 2ffi
Lc ¼ li ¼ ðx iþ1  x i Þ2 þ y iþ1  y i ð1Þ
i¼1 i¼1
qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
R¼ ðx N  x 1 Þ2 þ ðy N  y 1 Þ2 ð2Þ
vffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
u
u X N
Rg ¼ t1=N r 2i ð3Þ
i¼1

where:

 2
r 2i ¼ ðx i  x CM Þ2 þ y i  y CM ð4Þ
!
  X
N X
N
x CM ; y CM ¼ 1=N xi; yi ð5Þ
i¼1 i¼1
In Eq. 5 (xCM,yCM) are the coordinates of the center of mass of
the molecule.
2. Calculate the histograms of the values of the above observables,
in order to represent their statistical distributions. Always look
first at distributions, and then calculate suitable estimators of
the true values and their dispersions.
3. Calculate mean (median) values, standard deviations and stan-
dard deviations of the mean, or other suitable statistical esti-
mators, depending on the particular statistical properties of the
given observable [5].
An example of the analysis performed on DNA according to the
presented protocol is shown in Fig. 2.

4 Notes

1. The typical substrate used for DNA immobilization is mica, a

mineral belonging to the sheet silicate groups. There are many
varieties of mica and Muscovite is the most used form. It is
constituted by tetrahedral sheets of (Si,Al)2O5 ionically linked
by a central layer of Al2(OH)2 [24]. The net negative charge of
the basal oxygen between these double layers is balanced by a
layer of hexagonally coordinates cations (Kþ in Muscovite).
This negatively charged layer becomes exposed after the stan-
dard cleavage procedure and the dissociation of K+ ions. The
most prominent characteristic of mica is the nearly perfect
cleavage, due to its intrinsic atomic structure [25]. Due to its
chemical composition, the outmost layer of mica after cleavage
 Characterization of Structural and Configurational Properties of DNA. . . 565

Fig. 2 Distribution of contour length (Lc) values and assessment of the form of DNA molecules upon
dehydration on mica, according to the proposed protocol. Three populations of DNA molecules have been
investigated, with lengths 464 bp, 645 bp, and 727 bp, respectively. (a) The distribution of contour length
values is calculated according to Eq. 1. About 150, 80, and 80 molecules have been traced, respectively, for
the 464 bp, 645 bp, and 727 bp populations. The measured average lengths agree with those expected for the
B-form of DNA within 5–10%. (b) A linear fit of the curve Lc vs. bp provides a value of the rise per residue
parameter of 0.349 nm/bp, confirming unambiguously that DNA molecules, despite the dehydration, are in the
B form. The discrepancy between absolute observed values and the expected ones could be due to the partial
transition towards the A-form that DNA faces when deposited on mica [22], with the A-form domains likely
located at the DNA free ends [8]

is negatively charged in humid air and in particular in water.

Freshly cleaved mica could hence provide negative, ultra-flat
and clean surfaces, functional for high-quality AFM measure-
ments [26]. By using divalent positive ions (Mg2+, Ca2+, Mn2+,
Ni2+, and Zn2+) [7], or molecules carrying a positive charge, as
in the case of natural or artificial polyamines (poly-L-lysine and
poly-L-ornithine [27, 28]) it is possible to bind the negatively
charged DNA backbone to the mica surface, to the purpose of
imaging DNA by AFM.
2. A cheaper alternative to precut mica (and Teflon) disks is
represented by mica (and Teflon) sheets of the same quality
and thickness, from which disks of the desired diameter can be
obtained using a hole punch.
3. Rigid cantilevers for dynamic modes must be used. Typical
parameters characterizing these cantilevers are: resonance fre-
quency f  300 kHz; single-crystal silicon tips with radius of
curvature R < 10 nm; force constant k  40 N/m; optionally, a
gold or aluminium reflective coating on the back of the
cantilever.
4. All the stock saline solutions are prepared starting from the
powder that is dissolved in MilliQ water. Filtering should
566 Alice Meroni et al.

possibly be avoided, as in general the entities that are removed

are very large compared to DNA and would not represent a
serious issue. The sample has to be as clean as possible in order
to achieve the best condition of imaging, and clean buffers are a
fundamental prerequisite. For this reason all dilutions and
samples must be prepared using ultrapure MilliQ water based
buffers. Never use simple (bi)distilled water, and never sterilize
buffers.
5. If imaging in liquid is envisaged, it can be useful to put in
between the mica disk and the metal support a Teflon spacer,
about 2 mm larger than the mica disk and up to 2 mm thick,
aimed at blocking the spread of the liquid used as imaging
buffer, so that a stable droplet is obtained for in-liquid imaging.
In this case, it is better to use different adhesive depending on
the surfaces to be bound together. Mica on Teflon: two-
component epoxy glue; Teflon on metal disk: cyanoacrylate
glue. This metal–Teflon–mica substrate can be used of course
also for imaging DNA in air.
6. A tiny amount of glue must be used, initially placed in the
middle of the disk. Apply a gentle pressure so that the glue is
distributed uniformly in between the mica disk and the sup-
port. Carefully dose the amount of glue so to avoid it spreading
outside the mica disk area; this will likely cement the mica layers
together from the side and will disturb the stripping procedure.
7. Mica disks should be always freshly cleaved immediately before
the deposition of DNA, using soft adhesive tape to peel the
topmost layers away. In this way it is possible to create a flat
atomically smooth clean surface free of contaminants. To this
purpose, firmly attach the scotch tape over the supported mica
surface and remove it so to peel the topmost layers away.
Repeat the operation using a clean portion of tape until a
uniform thin circular layer remains on it (repeat in any case
2–3 times). To achieve a homogeneous stripping no air bubbles
must be present below the adhesive tape. Apply a constant
tension to the adhesive tape while peeling the mica. Remove
the tape in one continuous movement.
8. DNA must be used freshly prepared, resuspended in MilliQ
water. For best results, it can be stored at 4  C for possibly no
more than 1–2 weeks. Repeated thawing and freezing damages
the DNA backbone [29].
9. The layer used for the imaging has to be cleaved just before
sample deposition. It is a good practice to protect the sur-
face from the contact with air, if it is not used immediately.
10. When significant amount of water remains on the surface upon
drying, in the form of water islands, it may be the case that
something went wrong either during the deposition of DNA
Characterization of Structural and Configurational Properties of DNA. . . 567

or the cleavage of the mica disk. Sticky/dirty/overcoated sur-

faces retain water, indeed. Such surfaces may result from the
deposition of contaminated/degraded solutions or from rem-
nants of the scotch tape adhesive.
11. In principle, the incubation time of the DNA solution on mica
determines the number and density of molecules on the mica
surface [5, 10, 12], and typically a few minutes are enough to
obtain optimal imaging and analysis conditions. It may happen
however that poor reproducibility and deviations from the
expected behavior during sample preparation are observed.
This problem typically occurs when the DNA concentration
in solution and the incubation time are changed in the effort of
obtaining the desired density of molecules. Usually, this anom-
alous behavior is also accompanied by the poor quality of the
deposited molecules (condensed in blobs, totally or partially, or
with small blobs at the free ends; aggregated or associated in
complex two-dimensional structures or networks; etc.). All
these can be symptoms that either the molecules in solution,
or the buffer, or both, have some problems. For instance, a bad
PCR reaction can produce weak DNA molecules with open,
and therefore, sticky ends; because of the intrinsic DNA com-
plementarity, several molecule ends will anneal between each
other assembling networks, which will be mostly washed away,
but also remain on the mica surface to some extent. A dirty
buffer containing nanoscopic contaminants (see Notes 4
and 8) can promote DNA denaturation or aggregation, with
similar effects. Large complexes will be typically washed away,
as mentioned; therefore, only the minoritarian fraction of small
objects will remain on the surface, with a density largely inde-
pendent on both initial DNA concentration and incubation
time. If similar issues are faced, it is usually wise to first prepare
fresh clean buffers, then if needed fresh DNA stocks. Although
only rarely observed in our experience, similar problems can
also be due to poor quality of the mica surface (only use mica of
the highest grade) or to issues in mica cleavage (including
those due to the poor quality of the scotch tape, which can
leave residues of glue on the mica surface).
12. AFM measurements can be carried out in air, as described in
the present manuscript, as well as in a suitable saline buffer.
Imaging DNA in liquid does not necessarily provides more
accurate information, as long as one focuses on structural
data, because the latter mostly rely on the accurate characteri-
zation of lengths along the DNA backbone, rather than on the
measurement of heights. Tip convolution affects only at minor
extent such measurements, at least for relatively long molecules
(>100 bp). Moreover, there is evidence that dehydration
required for sample preparation has little impact on the DNA
568 Alice Meroni et al.

properties, probably due to the fact that the truly DNA–mica

interface remains always partially hydrated (by a few mono-
layers of water) [5, 9, 28]. Nevertheless, the fine conforma-
tional changes induced by dehydration, as well as by the
different sample preparation methods and choices of the bridg-
ing cations are still matter of discussion [8, 22].The general
methodology presented here applies irrespective to the imag-
ing method adopted (air, liquid, Tapping Mode, Peak Force
Tapping, etc.). For some specific indications in the case of
imaging in liquid, see Note 5.
13. Imaging methods based on the vertical approach of the tip
towards the sample have recently been introduced [30], and
represent valid alternatives to dynamic (tapping) modes and
contact modes, especially in liquid. The general idea, besides
specificities related to the different implementations, is to
record a set of force curves on a grid spanning the scan area,
with a carefully controlled maximum force setpoint; during
imaging in fluid, the maximum force can be kept on the
10–100 pN level, while lateral forces are minimized thanks to
the vertical approach mode, similarly to Tapping or other
dynamic modes. In Tapping Mode, however, peak forces are
significantly higher.
14. Calibration of the piezo-scanners should be checked periodi-
cally (every 6–12 months) by imaging the surface of a calibra-
tion grating, with repeated morphological features of
appropriate dimensions (in the present case, the XY period
should be 1 μm, the depths 10–200 nm). Unless a certified
grating is used, the XYZ accuracy after the calibration proce-
dure can be reasonably assumed to be 2%.
15. The number of scanned lines impacts on the acquisition time of
a single image. Setting an aspect-ratio 2:1 allows reducing the
number of scanned lines, i.e., the acquisition time, without
affecting dramatically the image resolution in the slow scan
direction. Given a target sampling resolution in the image,
the scanned area can be kept large, for the sake of a better
statistics, by increasing the scan size and the aspect-ratio value,
and setting suitable values of the points per line and number of
lines parameters. The main limit to be considered is that the
scanning speed cannot grow arbitrarily, otherwise the feedback
loop will not be able to track accurately the surface.
16. Scanning speed (which changes when either scan rate or scan
size are changed) should not be too small, otherwise drifts
could produce significant distortions in the image. At the
same time, too high scan speeds will challenge the feedback
loop of the AFM and determine inaccurate tracking.
17. Typically the integral gain has the highest impact on image
quality. Each instrument has its own gain optimal settings.
Characterization of Structural and Configurational Properties of DNA. . . 569

Increase the integral gain until oscillations appear in the height

and error signals. Then decrease the integral gain just below
the critical value. Increase the proportional gain until the qual-
ity of the image starts worsening. Setting high gains gives
advantages in terms of tracking stability and scanning velocity
that overcompensates the introduction of high-frequency noise
in the image; the latter can be effectively removed a posteriori
by applying a median filter with a 3  3 kernel on the image.
18. Drifts must be minimized in order to obtain accurate topo-
graphic maps. Drifts can be due to thermal equilibration of the
system components (sample, laser, cantilever, electronics as
well as the scanner) or to mechanical hysteresis of the piezo
elements. When a different scan area is selected by applying
offsets to the piezo, the latter will typically keep some memory
of the previous static deformation, resulting in a constant drift
across the new image. In order to minimize this effect and
completely refresh the scanner motion, after setting the new
offsets it is effective to reduce significantly the number of lines
and complete a few low-resolution images moving quickly up
and down, until the hysteresis is lost. Also reverting a few times
the slow scan direction (up-bottom/bottom-up) helps remov-
ing the hysteresis of the piezo scanner.
19. The data analysis procedures described in general terms in this
manuscript can be implemented by means of custom routines,
as well as by means of commercial and open-source software.
Basic image processing tools are typically included in the con-
trol software of the AFM. The authors have developed their
own libraries of data analysis routines in the MATLAB environ-
ment (many research groups use their own libraries). Some
open-source software packages are listed below (this list is
neither meant to be complete, nor is it expected to remain up
to date for a long time):
(a) ImageJ/Fiji, http://rsb.info.nih.gov/ij/
(b) Gwyddion, http://gwyddion.net/
(c) WSxM, http://www.wsxmsolutions.com/
(d) FiberApp, http://www.fsm.ethz.ch/publications-list/
software.html
(e) Image SXM, https://www.liverpool.ac.uk/%7Esdb/
ImageSXM/
20. Mica is atomically smooth and overall flat. Deviations from a
flat baseline can therefore be attributed to scanning artifacts,
with the obvious exception of those due to the presence of a
DNA molecule. A global tilt of the sample adds a linear baseline
to each profile, and globally a plane to the topographic map as a
whole. Tubular scanners will add an approximately parabolic
570 Alice Meroni et al.

baseline to each profile (a bow), because the displacement of

the sample placed on top of the tube follows a curved trajectory
rather than a linear one (ideally an arc of circumference).
Generally, each scanner will add its own polynomial distortion
to the AFM topographic map. Drifts of different nature can
add additional shifts between adjacent profiles (approximately
constant along each profile, i.e., along the fast-scan direction,
but appreciable along the slow-scan direction); the presence of
these line-by-line deviations usually requires the application of
line-by-line polynomial subtraction (also known as flattening)
rather than simply the subtraction of a two-dimensional plane,
or paraboloid, or higher-order surface (also known as plane-
fitting). It is essential, in order to accurately determine the
image baseline, to mask (i.e., not to consider in the polynomial
fit) all the surface features that do not belong to the flat
reference substrate (the DNA molecules, surface defects, . . .).
Masks are typically built by thresholding (flooding) algorithms,
determining image segmentation; only the substrate is consid-
ered for the flattening. The absence of masking in the fitting
procedure will introduce artifacts in the topographic maps (the
ubiquitous black stripes), because in the presence of bumps
and/or depressions the fitted polynomial typically deviates
from the baseline. Masking is typically a feature of the analysis
software. The flattening process is described in Fig. 3.
21. A semiautomatic tracing algorithm is in principle preferable to
the manual tracing of DNA molecules, as the latter is more
prone to introduce bias from the operator. An automatic algo-
rithm can of course introduce systematic errors, although these
would be the same for all molecules, irrespective to the operator.
Ivan Usov’s FiberApp is a free comprehensive suite of
(MATLAB) routines with a GUI for “Tracking and Analyzing
Polymers, Filaments, Biomacromolecules, and Fibrous Objects”
(available at http://www.fsm.ethz.ch/publications-list/soft
ware.html [31]).
22. The formulas reported in the text for the calculation of basic
structural parameters provide in general accurate results as long
as the digitization of the molecules in the image is not poor,
which in the case of DNA means that the point-to-point dis-
tance should be of the order of 0.5–1 nm. Separations signifi-
cantly smaller than 0.5–1 nm are unreasonable, considered that
the nominal width of DNA is 2 nm; oversampling can intro-
duce spurious high-frequency components to the trace, that
will affect the angle distribution as well as the calculated con-
tour length. The contour length Lc is the most sensitive param-
eter, indeed. Optimized estimators for Lc have been developed,
and can be used instead of Eq. 1 [32–34].
 Characterization of Structural and Configurational Properties of DNA. . . 571

Fig. 3 Overview of the image pre-processing procedure. The process starts with (a) a raw AFM image, where
the sample tilt and line-by-line distortions hinder the target topographical features (the DNA molecules in this
case); after (b) a global plane-fitting of the first order, and a series of line-by-line flattening of the (c) first, and
(d) third order, the baseline is effectively removed and the molecules emerge, well-contrasted with respect to
the smooth, flat substrate. In (e) the mask built to apply the third-order flattening is shown. This logical mask
assigns a value of 1 to the points that must be considered for the fitting, i.e., those belonging to the substrate,
and 0 elsewhere. In (f) the distributions of surface heights after the first and third order flattening are
compared. A well-shaped, nearly symmetric dominant mode, representing the height values of the substrate,
is typical of a properly flattened image

Acknowledgments

We thank Francesca Borghi for support in AFM analysis. A.P.

thanks the Dept. of Physics of the University of Milano for financial
support under the project “Piano di Sviluppo dell’Ateneo per la
Ricerca 2014. Linea B: Supporto per i giovani ricercatori”. Work in
M.M-F lab is supported by AIRC (n.15631) and Telethon
(GGP15227).
572 Alice Meroni et al.

References
1. Alessandrini A, Facci P (2005) AFM: a versatile 12. Fang Y, Spisz TS, Wiltshire T, D’Costa NP,
tool in biophysics. Meas Sci Technol 16: Bankman IN, Reeves RH, Hoh JH (1998)
R65–R92. doi:10.1088/0957-0233/16/6/ Solid-state DNA sizing by atomic force micros-
R01 copy. Anal Chem 70:2123–2129. doi:10.
2. Kalle W, Strappe P (2012) Atomic force 1021/ac971187o
microscopy on chromosomes, chromatin and 13. Santos S, Stefancich M, Hernandez H, Chiesa
DNA: a review. Micron 43:1224–1231. M, Thomson NH (2012) Hydrophilicity of a
doi:10.1016/j.micron.2012.04.004 single DNA molecule. J Phys Chem C
3. Podgornik Rudolf (2011) Physics of DNA. 116:2807–2818. doi:10.1021/jp211326c
http://www.fmf.uni-lj.si/~podgornik/down 14. Thomson NH, Santos S, Mitchenall LA, Stu-
load/physics-of-DNA-1.1.pdf chinskaya T, Taylor JA, Maxwell A (2014)
4. Pinheiro AV, Han D, Shih WM, Yan H (2011) DNA G-segment bending is not the sole deter-
Challenges and opportunities for structural minant of topology simplification by type II
DNA nanotechnology. Nat Nanotechnol DNA topoisomerases. Sci Rep 4:6158.
6:763–772. doi:10.1038/nnano.2011.187 doi:10.1038/srep06158
5. Rivetti C, Guthold M, Bustamante C (1996) 15. Wiggins PA, van der Heijden T, Moreno-
Scanning force microscopy of DNA deposited Herrero F, Spakowitz A, Phillips R, Widom J,
onto mica: equilibration versus kinetic trapping Dekker C, Nelson PC (2006) High flexibility
studied by statistical polymer chain analysis. J of DNA on short length scales probed by
Mol Biol 264:919–932. doi:10.1006/jmbi. atomic force microscopy. Nat Nanotechnol
1996.0687 1:137–141. doi:10.1038/nnano.2006.63
6. Podestà A, Indrieri M, Brogioli D, Manning 16. Savelyev A, Materese CK, Papoian GA (2011)
GS, Milani P, Guerra R, Finzi L, Dunlap D Is DNA’s rigidity dominated by electrostatic or
(2005) Positively charged surfaces increase nonelectrostatic interactions? J Am Chem Soc
the flexibility of DNA. Biophys J 133:19290–19293. doi:10.1021/ja207984z
89:2558–2563 17. Lia G, Indrieri M, Owen-Hughes T, Finzi L,
7. Pastré D, Piétrement O, Fusil S, Landousy F, Podesta A, Milani P, Dunlap D (2008) ATP-
Jeusset J, David M-O, Hamon L, Le Cam E, dependent looping of DNA by ISWI. J Bio-
Zozime A (2003) Adsorption of DNA to mica photonics 1:280–286
mediated by divalent counterions: a theoretical 18. Fang Y, Hoh JH, Spisz TS (1999) Ethanol-
and experimental study. Biophys J induced structural transitions of DNA on
85:2507–2518. doi:10.1016/S0006-3495( mica. Nucleic Acids Res 27:1943–1949.
03)74673-6 doi:10.1093/nar/27.8.1943
8. Japaridze A, Vobornik D, Lipiec E, Cerreta A, 19. Parker SCJ, Margulies EH, Tullius TD (2008)
Szczerbinski J, Zenobi R, Dietler G (2016) The relationship between fine scale DNA struc-
Toward an effective control of DNA’s submo- ture, GC content, and functional elements in
lecular conformation on a surface. Macromo- 1% of the human genome. Genome Inform
lecules 49:643–652. doi:10.1021/acs. 20:199–211
macromol.5b01827 20. Knips A, Zacharias M (2015) Influence of a cis,
9. Buzio R, Repetto L, Giacopelli F, Ravazzolo R, syn-cyclobutane pyrimidine dimer damage on
Valbusa U (2014) Symmetric curvature DNA conformation studied by molecular
descriptors for label-free analysis of DNA. Sci dynamics simulations. Biopolymers
Rep 4:6459. doi:10.1038/srep06459 103:215–222. doi:10.1002/bip.22586
10. Bustamante C, Rivetti C (1996) Visualizing 21. Garcı́a R, Perez R (2002) Dynamic atomic
protein-nucleic acid interactions on a large force microscopy methods. Surf Sci Rep
scale with the scanning force microscope. 47:197. doi:10.1016/S0167-5729(02)
Annu Rev Biophys Biomol Struct 00077-8
25:395–429. doi:10.1146/annurev.bb.25. 22. Rivetti C, Codeluppi S (2001) Accurate length
060196.002143 determination of DNA molecules visualized by
11. Valle F, Favre M, De Los RP, Rosa A, Dietler G atomic force microscopy: evidence for a partial
(2005) Scaling exponents and probability dis- B- to A-form transition on mica. Ultramicro-
tributions of DNA end-to-end distance. Phys scopy 87:55–66. doi:10.1016/S0304-3991(
Rev Lett 95:158105. doi:10.1103/Phy 00)00064-4
sRevLett.95.158105 23. Shi Y (1996) Statistical mechanics of the exten-
sible and shearable elastic rod and of DNA. J
 Characterization of Structural and Configurational Properties of DNA. . . 573

Chem Phys 105:714–731. doi:10.1063/1. 29. Bustamante C, Vesenka J, Tang CL, Rees W,
471927 Guthold M, Keller R (1992) Circular DNA
24. Kuwahara Y (1999) Muscovite surface struc- molecules imaged in air by scanning force
ture imaged by fluid contact mode AFM. Phys microscopy. Biochemistry 31:22–26
Chem Miner 26:198–205. doi:10.1007/ 30. Adamcik J, Berquand A, Mezzenga R (2011)
s002690050177 Single-step direct measurement of amyloid
25. Ostendorf F, Schmitz C, Hirth S, K€ uhnle A, fibrils stiffness by peak force quantitative nano-
Kolodziej JJ, Reichling M (2008) How flat is mechanical atomic force microscopy. Appl Phys
an air-cleaved mica surface? Nanotechnology Lett 98:193701. doi:10.1063/1.3589369
19:305705. doi:10.1088/0957-4484/19/ 31. Usov I, Mezzenga R (2015) FiberApp: an
30/305705 open-source software for tracking and analyz-
26. Thundat T, Allison DP, Warmack RJ, Brown ing polymers, filaments, biomacromolecules,
GM, Jacobson KB, Schrick JJ, Ferrell TL and fibrous objects. Macromolecules
(1992) Atomic force microscopy of DNA on 48:1269–1280. doi:10.1021/ma502264c
mica and chemically modified mica. Scanning 32. Sanchez-Sevilla A, Thimonier J, Marilley M,
Microsc 6:911–918 Rocca-Serra J, Barbet J (2002) Accuracy of
27. Bussiek M, M€ ucke N, Langowski J (2003) AFM measurements of the contour length of
Polylysine-coated mica can be used to observe DNA fragments adsorbed on mica in air and in
systematic changes in the supercoiled DNA aqueous buffer. Ultramicroscopy 92:151–158.
conformation by scanning force microscopy in doi:10.1016/S0304-3991(02)00128-6
solution. Nucleic Acids Res 31:e137. doi:10. 33. Rivetti C (2009) A simple and optimized
1093/nar/gng137 length estimator for digitized DNA contours.
28. Podesta A, Imperadori L, Colnaghi W, Finzi L, Cytom Part A 75:854–861. doi:10.1002/cyto.
Milani P, Dunlap D (2004) Atomic force a.20781
microscopy study of DNA deposited on poly 34. Gomez AI, Cruz M, Cruz-Orive LM (2016)
L-ornithine-coated mica. J Microsc On the precision of curve length estimation in
215:236–240. doi:10.1111/j.0022-2720. the plane. Image Anal Stereol 35:1. doi:10.
2004.01372.x 5566/ias.1412
 Chapter 38

Genome-Wide Quantitative Fitness Analysis (QFA) of Yeast

Cultures
Eva-Maria Holstein, Conor Lawless, Peter Banks, and David Lydall

Abstract
We provide a detailed protocol for robot-assisted, genome-wide measurement of fitness in the model yeast
Saccharomyces cerevisiae using Quantitative Fitness Analysis (QFA). We first describe how we construct
thousands of double or triple mutant yeast strains in parallel using Synthetic Genetic Array (SGA)
procedures. Strains are inoculated onto solid agar surfaces by liquid spotting followed by repeated photog-
raphy of agar plates. Growth curves are constructed and the fitness of each strain is estimated. Robot-
assisted QFA, can be used to identify genetic interactions and chemical sensitivity/resistance in genome-
wide experiments, but QFA can also be used in smaller scale, manual workflows.

Key words SGA, QFA, High-throughput, Yeast, Genetic interaction, Genome-wide, Fitness,
Robotic, Quantitative, Population model, Image analysis, Growth curve

1 Introduction

Comparing how well specific strains grow in a carefully chosen

environment or genetic background can reveal how different
genes interact in living cells. For instance, deletion of EXO1, affect-
ing a nuclease, or RAD9, affecting a checkpoint protein, each
suppress telomere induced growth defects of budding yeast cdc13-
1 mutants [1]. Quantitative Fitness Analysis (QFA) allows genome-
wide identification of similar suppressing and enhancing genetic
interactions [2, 3]. For example, QFA revealed that nonsense
mediated mRNA decay proteins affect the fitness of cdc13-1
mutants in a way similar to Exo1 and Rad9 [2]. QFA can also be
used for drug interaction screens [4].
QFA evolved from qualitative spot test assays that were manu-
ally performed and scores and were not photographed [5]. The
procedure we describe here is now high-throughput automated,
documented, and quantitative. However, we and others, also apply
some of the procedures on lower scale, manually performed experi-
ments [3].

Marco Muzi-Falconi and Grant W. Brown (eds.), Genome Instability: Methods and Protocols, Methods in Molecular Biology,
vol. 1672, DOI 10.1007/978-1-4939-7306-4_38, © Springer Science+Business Media LLC 2018

575
576 Eva-Maria Holstein et al.

Typically, to perform a genome-wide QFA, we begin with

Synthetic Genetic Array (SGA), crossing a query strain, containing
one or more mutations, to a library of single gene deletions and
DAmP (Decreased Abundance by mRNA Perturbation) alleles [2,
6]. Strains derived from SGA are grown to saturation in liquid
medium, in 96-well plates. Liquid cultures are next spotted onto
solid agar plates in 384-format arrays and repeatedly photographed
during growth. The images are analyzed to generate quantitative
cell density estimates which are used to create growth curves and
ultimately to measure the fitness of each culture. This approach
allows fitness of thousands of strains grown in parallel to be
measured and compared. We describe detailed protocols for SGA
and QFA using query strains containing a single gene deletion
(abcΔ), a point mutation (xyz-1) or both (abcΔ xyz-1), any of
which can grow on the same selection media. Since we often cross
temperature sensitive mutations to libraries, our SGA protocols
have been optimized to work at 20
C.

2 Materials

2.1 Stock Solutions Figure 1 gives an overview of the SGA procedure and in particular
and Media the media used at each stage, as indicated below (MRN ¼ Media
recipe number).
1. Ethanol: 70% in water.
2. Geneticin (G418): dissolve in water at 200 mg/mL, filter-
sterilize and store in 1 mL aliquots at 20
C.
3. Nourseothricin (clonNAT): dissolve in water at 100 mg/mL,
filter-sterilize and store in 1 mL aliquots at 20
C.
4. Hygromycin B: dissolve in water at 300 mg/mL, filter-sterilize
and store in 1 mL aliquots at 20
C.
5. Canavanine: dissolve in water at 100 mg/mL, filter-sterilize
and store in 0.5 mL aliquots at 20
C.
6. Thialysine: dissolve in water at 100 mg/mL, filter-sterilize and
store in 0.5 mL aliquots at 20
C.
7. Amino-acid supplement powder mixture for synthetic media
(CSM DO): 3 g adenine, 2 g uracil, 2 g inositol, 0.2 g para-
aminobenzoic acid, 2 g alanine, 2 g arginine, 2 g asparagine,
2 g aspartic acid, 2 g cysteine, 2 g glutamic acid, 2 g glutamine,
2 g glycine, 2 g histidine, 2 g isoleucine, 2 g leucine, 2 g lysine,
2 g methionine, 2 g phenylalanine, 2 g proline, 2 g serine, 2 g
threonine, 2 g tryptophan, 2 g tyrosine, and 2 g valine.
8. Drop-out (DO) powder mixture: combination of the ingredi-
ents described in Subheading 2.1, item 7 minus the appropri-
ate supplement: rhk DO: Arg/His/Lys. rhlk DO: Arg/
 QFA 577

Query Library

YPD_N lawn 384 format YPD_G

MRN 3 2 days 20˚C 2 days 30˚C MRN 2

YPD_N 1536 format 1536 format YPD_G

MRN 3 2 days 20˚C 2 days 30˚C MRN 2

YPD
12h-24h 23˚C Mating
MRN 1

YEPD_GN
36h 30˚C Diploid selection
MRN 4

ESM _G Sporulation
5 days 23˚C
MRN 5

SDM_rhk_CT
2 days 20˚C HSI
MRN 6

SDM_rhk_CT
2 days 20˚C HSII
MRN 6

SDM_rhk_CTG Meiotic Progeny

2 days 20˚C HSIII Selection
MRN 7

SDM_rhlk_CTGH 2 days 20˚C HSIV

MRN 8

SDM_rhlk_CTGNH 2 days 20˚C HSV

MRN 9

Fig. 1 Flowchart illustrating the steps of the SGA procedure. Query strain and
library strains are pinned in a 1536 format, followed by mating, diploid selection,
and induction of sporulation. The meiotic progeny undergoes several selection
stages to select for double mutants expressing the STE2pr-SP-his5 reporter,
LEU2 and resistance to canavanine and thialysine (can1Δ and lyp1Δ), G418,
cloNAT and Hygromycin B

His/Leu/Lys. For 1 L of medium: 2 g of the DO powder

mixture.
9. Amino-acid supplement for sporulation medium (ESM DO):
2 g histidine, 10 g leucine, 2 g lysine, and 2 g uracil. For 1 L:
0.1 g of amino-acid supplement powder mixture.
578 Eva-Maria Holstein et al.

10. Dextrose: 40% in water, autoclave and store at room

temperature.
11. Adenine: 0.5% in water, autoclave and store at room
temperature.
12. YEPD (YPD, liquid): For 1 L add 10 g yeast extract, 20 g
peptone to 935 mL. Autoclave (121
C for 15 min), cool down
to ~65
C and add 50 mL of 40% dextrose solution and 15 mL
of 0.5% adenine solution, gently invert bottle until mixture is
homogenous and store at room temperature. (MRN 1).
13. YEPD (YPD, solid): For 1 L add 10 g yeast extract, 20 g
peptone, and 20 g agar to 935 mL. Autoclave (121
C for
15 min), cool down to ~65
C, add 50 mL of 40% dextrose
solution and 15 mL of 0.5% adenine solution, gently invert
bottle until mixture is homogenous and pour plates. (MRN 1).
14. YEPD þ G418 (YPD_G) for library strain selection: For 1 L
add 1 mL of 200 mg/mL G418 stock solution (final concen-
tration 200 mg/L) to YEPD medium (Subheading 2.1, item
13), gently invert bottle until the mixture is homogenous and
pour plates. (MRN 2).
15. YEPD þ clonNAT (YPD_N, solid) for query strain selection:
For 1 L add 1 mL of 100 mg/mL clonNAT stock solution
(final concentration 100 mg/L) to YEPD medium (Subhead-
ing 2.1, item 13), gently invert bottle until the mixture is
homogenous and pour plates. (MRN 3).
16. YEPD þ G418/clonNAT (YPD_ GN, solid) for diploid selec-
tion: For 1 L add 1 mL of 200 mg/mL G418 stock solution
(final concentration 200 mg/L) and 1 mL of 100 mg/mL
clonNAT stock solution (final concentration 100 mg/L) to
autoclaved YEPD medium (Subheading 2.1, item 13), gently
invert bottle until the mixture is homogenous and pour plates.
(MRN 4).
17. Enriched sporulation medium (ESM_G, solid) for sporulation:
For 1 L add 10 g potassium acetate, 1 g yeast extract, 0.5 g
dextrose, 0.1 g amino-acid supplement powder mixture for
sporulation to 500 mL of water. In a separate bottle add 20 g
agar to 500 mL water. Autoclave both solutions at 121
C for
15 min. Add the agar to the media, cool down to ~65
C and
add 250 μL of G418 stock solution (final concentration
50 mg/L), mix thoroughly and pour plates. (MRN 5).
18. (SD/MSG) His/Arg/Lys þ canavanine/thialysine
(SDM_rhk_CT, solid) for haploid I and II (HS I and HS II)
selection stages: For 1 L add 1.7 g yeast nitrogen base, 1 g
MSG, 2 g amino-acid supplement powder mixture (rhk DO),
450 mL water and autoclave at 121
C for 15 min. In a separate
bottle add 20 g agar to 500 mL water and autoclave at 121
C
 QFA 579

for 15 min. Combine autoclaved solutions, add 50 mL 40%

glucose, cool medium to ~65
C, add 0.5 mL canavanine
(100 mg/L) and 0.5 mL thialysine (100 mg/L) stock solu-
tions, mix thoroughly and pour plates. (MRN 6) (see Note 1).
19. (SD/MSG) His/Arg/Lys þ canavanine/thialysine/G418
(SDM_rhk_CTG, solid) for HS III selection stage: same as
MRN 6 (Subheading 2.1, item 18) with 1 mL G418
(200 mg/L) stock solution added before mixing and pouring
plates (MRN 7).
20. (SD/MSG) His/Arg/Lys/Leu þ canavanine/thialysine/
G418/Hygromycin B (SDM_rhlk_CTGH, solid) for HS IV
selection stage: For 1 L add 1.7 g yeast nitrogen base, 1 g
MSG, 2 g amino-acid supplement powder mixture (rhlk DO),
450 mL water and autoclave at 121
C for 15 min. In a separate
bottle add 20 g agar to 500 mL water and autoclave at 121
C
for 15 min. Combine autoclaved solutions, add 50 mL 40%
glucose, cool medium to ~65
C, add 0.5 mL canavanine
(100 mg/L), 0.5 mL thialysine (100 mg/L), 1 mL Hygro-
mycin B (300 mg/L) and 1 mL G418 (200 mg/L) stock
solutions, mix thoroughly and pour plates. (MRN 8).
21. (SD/MSG)  His/Arg/Lys/Leu þ canavanine/thialysine/
G418/clonNAT/Hygromycin B (SDM_rhlk_CTGNH,
solid) for HS IV selection stage and QFA spotting: same as
MRN 8 (Subheading 2.1, item 20) with 1 mL clonNAT
(100 mg/L) stock solution added before mixing and pouring
plates (MRN 9).
22. (SD/MSG) His/Arg/Lys/Leu þ canavanine/thialysine/
G418/clonNAT/Hygromycin B (SDM_rhlk_CTGNH, liq-
uid) for QFA: For 1 L add 1.7 g yeast nitrogen base, 1 g
MSG, 2 g amino-acid supplement powder mixture (rhlk
DO), 900 mL water and autoclave at 121
C for 15 min.
Add 50 mL 40% glucose, cool medium to ~65
C, add
0.5 mL canavanine (100 mg/L), 0.5 mL thialysine (100 mg/
L), 1 mL clonNAT (100 mg/L), 1 mL Hygromycin B
(300 mg/L), and 1 mL G418 (200 mg/L) stock solutions,
mix thoroughly and store at room temperature (MRN 9).

2.2 Plates and Tubes 1. Rectangular polycarbonate Greenlab SBS permaplates (https://
sites.google.com/site/greenlabuk/permaplates). These plates
are stackable and autoclavable. After use, the agar and cultures
are discarded, disinfected using Virkon (1%, at least 30 min),
washed in water, and autoclaved at 121
C for 15 min. Alterna-
tively, disposable OmniTrays (Nunc) can be used.
2. 96-well, flat bottom culture plates and low evaporation lids.
3. 15 mL glass test tubes with Bacti-Caps.
4. Single deep well reservoir containing 96 pyramidal bottoms.
580 Eva-Maria Holstein et al.

2.3 Robotic Systems 1. S&P Robotics Inc BM3-SC with 96, 384, and 1536 pin
robotic pin tools.
2. Beckman Biomek FX with 96 pin robotic pin tool (V&P Scien-
tific magnetic mounting pin tool, 2 mm pin diameter).
3. S&P Robotics automated imager.

2.4 Software 1. Colonyzer image analysis software: http://research.ncl.ac.uk/

colonyzer/
2. QFA R package: http://qfa.r-forge.r-project.org/

2.5 Strains for SGA The SGA protocol is based on that of Tong and Boone [6]. We have
adapted it in two major ways. First, because we often use tempera-
ture-sensitive mutants, many of the steps are performed at 20
C.
Secondly, we use extra selectable markers. All starter strains contain
three markers LEU2, natMX, and hphMX as well as can1Δ and
lyp1Δ to allow us to assess query strains with up to three gene
deletions. For example, query strains containing recessive ts muta-
tions are flanked by two selectable markers (LEU2 and hphMX). For
SGAs where we use query strains with a single gene replaced with
the natMX cassette, we still ensure that LEU2 and hphMX are
present in the query strains. By completing all experiments on
media containing the same selection agents, we can more closely
match environments between QFA experiments. Matching envir-
onments is important for detecting genetic or chemical interac-
tions. For example, in order to calculate genetic interaction
strength (GIS) [2], a control CDC13 strain and an experimental
cdc13-1 query strain, containing the same set of markers, are
crossed to the same library, on the same media, before comparing
fitnesses at the same temperatures [2].

2.5.1 Libraries 1. Single gene deletion yeast library with each gene disruption
being replaced with the antibiotic resistance cassette
kanMX [6].
2. Decreased Abundance by mRNA Pertubation (DAmP) yeast
library with the 30 -UTR being disrupted with the antibiotic
resistance cassette kanMX [7].

2.5.2 Universal Control A universal control strain, with LEU2, hphMX and natMX, is often
Strain crossed to the library to measure the effects of library mutations on
fitness under different conditions (e.g., of temperature or chemi-
cals). This allows for direct comparison of fitness between strains
containing a query mutation and control strains lacking the query
mutation on the same media.
DLY9326: MATalpha can1Δ::STE2pr-Sp-his5 lyp1Δ::hphMX::
LEU2::natMX his3Δ leu2Δ ura3Δ met15Δ
 QFA 581

2.5.3 Single Gene Single genes are deleted with natMX.

Deletion Starter Strain Source strain: DLY7325: MATalpha can1Δ::STE2pr-Sp-his5
(e.g., abcΔ::natMX) lyp1Δ::hphMX::LEU2 his3Δ1 leu2Δ ura3Δ met15Δ
Example query strain: DLY10276: MATalpha can1Δ::STE2pr-
Sp_his5 lyp1Δ::hphMX::LEU2 smt3Δ::natMX his3Δ1 leu2Δ ura3Δ
met15Δ

2.5.4 Point Mutation Point mutation, and if necessary control WT allele, is flanked with
Starter Strain (LEU2::xyz- LEU2 and hphMX. Source strain: DLY7329: MATalpha can1Δ::
1::hphMX) STE2pr-Sp-his5 lyp1Δ::natMX his3Δ1 leu2Δ ura3Δ met15Δ
Example query strain: DLY8205: MATalpha can1Δ::STE2pr-
Sp-his5 LEU2::cdc13-1::hphMX lyp1Δ::natMX his3Δ1 leu2Δ ura3Δ
met15Δ

2.5.5 Point Mutation and Point mutation, and if necessary control WT allele, is flanked with
Single Gene Deletion LEU2 and hphMX. Single gene is replaced with natMX.
Starter Strain (LEU2::xyz- Source strain: DLY3798: MATalpha lyp1Δ can1Δ::STE2pr-
1::hphMX abcΔ::natMX) his5Δ his3Δ1 leu2Δ ura3Δ met15Δ
Example query strain: DLY6722: MATalpha LEU2::cdc13-1::
hphMX exo1Δ::natMX
lyp1Δ can1Δ::STE2pr-his5Δ his3Δ1 leu2Δ ura3Δ met15Δ

3 Methods

3.1 Sterilization 1. Rinse the pin tool for 10 s in revolving water reservoir filled
Procedure for Pin with sterile water.
Tools Before Each 2. Move the pin tool to the revolving brush station containing
Pinning or Spotting 350 mL 70% ethanol for 20 s.
Step 3. Sonicate the pin tool for 20 s in the sonicator containing
3.1.1 S&P Robotics Inc 350 mL 70% ethanol.
BM3-SC 4. Allow pin tool to dry over the fan for 40 s.

3.1.2 Beckman Biomek 1. Rinse the pin tool in the wash station filled with sterile water
FX (see Fig. 2).
2. Transfer the pin tool to the brush station, and rotate for two
full circles in 70% ethanol.
3. Let the pin tool sit in the single deep well reservoir with 96-well
pyramidal bottoms containing 70% ethanol.
4. Allow pin tool to dry over the fan on the VP550 docking
station for 45 s.

3.2 SGA Each 384 format library plate contains neutral strains (e.g., his3Δ)
around the edges of the plate because edge cultures have a growth
advantage and because image analysis is more difficult at plate
edges. Therefore there are 308 experimental deletion mutants per
582 Eva-Maria Holstein et al.

Fig. 2 Layout of the FX robot deck prior to spotting

384-format plate. These strains are pinned to a 1536 format,

resulting in four replicates for each genotype. Figure 1 gives an
overview of the various stages of the SGA procedure to generate
double mutants.

3.2.1 Revival of Libraries 1. Let the plates thaw completely for 1 h at room temperature.
from Frozen Stocks 2. Wipe the 384-well library plates dry with disposable paper.
Take the plastic lid off, wipe the inside with 70% EtOH, asepti-
cally remove the covering film from the frozen 384-well plates
and put the plastic lid back onto the plates. (see Note 2).
3. Replicate the glycerol stocks of yeast from the 384-well plates
onto solid YEPD þ G418 (YEPD_G, Media recipe no. 2) agar
plates using the 384 pin robotic pin tool on the BM3.
4. Reseal the 384-well plates with fresh sealing film using a rubber
roller and return to 80
C.
5. Incubate the transferred colonies on the YEPD þ G418 plates
at 30
C for 2 days.

3.2.2 Documenting SGA 1. Each SGA screen is assigned a unique screen number in a
Procedures database. In addition, the database documents starter strain
number and genotype, library used, screen description and
the name of the person conducting the screen are stored. See
Fig. 3 for details.
2. For each step of the SGA procedure the media used, date,
media batch number as well as the incubation temperature is
 QFA 583

Fig. 3 Representation of SGA metadata in a database. A unique screen number is assigned to each SGA screen
(e.g., screen number 71). Starter strain number and genotype, library, screen description and initials of three
people engaged with the screen are recorded. For each SGA stage, media used, date of the experiment
conducted, media batch number, and incubation temperature are recorded. Any comments or problems
arising during the SGA procedure can also be stored in the database

documented. Any problems or comments arising during any of

the steps of the SGA procedure can be stored in the database
(see Note 3). See Fig. 3 for an example screenshot of the
database we currently use.
3. Each SGA stage (mating, diploid selection, sporulation, and
haploid selection I–V) is imaged in an S&P Robotics spImager
before the next pinning step takes place. Images are saved with
systematic, fixed length filenames (e.g.,
K000134_027_011_2012-05-05_12-22-56.JPG) including
incubator ID (K in the example filename), unique batch ID
(000134), incubator temperature (027), plate number (011),
date and time (2012-05-05_12-22-56).

3.2.3 Cultivation of the 1. Pool several colonies of the query strain and inoculate them
SGA Query Strain and into two 15 mL glass tubes, each containing 5 mL of YEPD.
Library (see Note 4).
2. Allow the strains to grow on a wheel overnight at 20
C.
3. Remove 10 μL with a sterile tip and check by microscopy for
contamination. Discard contaminated cultures.
4. Pour the culture of one tube over a rectangular YEPD plate
containing cloNAT (MRN 3) and ensure that the culture is
spread evenly across the plate by gentle agitation.
5. Transfer the liquid to a second plate by pouring and spread the
culture evenly again. Repeat for a third, fourth and fifth plates.
The lawns are allowed to grow for 2 days at 20
C.
584 Eva-Maria Holstein et al.

6. Repeat steps 4 and 5 depending on how many lawns you

require (see Note 4).
7. Pick the three most evenly grown lawns and robotically pin the
lawn containing the query strain in a 1536 format onto 6 fresh
YEPD þ clonNAT plates per lawn (MRN 3) using the 1536
pin robotic pin tool and incubate at 20
C for 2 days.
8. Pin the library from the 384 format to a 1536 format onto
fresh YEPD þ G418 plates using the 384 pin robotic pin tool
(MRN 2) and incubate at 30
C for 2 days.

3.2.4 Mating the Query 1. Pin the query strains (1536 format) onto a YEPD plate using a
Strain with the Library 1536 pin robotic pin tool.
2. Pin the library (1536 format) on top of the pinned query
strains on the YEPD plates from step 1 using the 1536 pin
robotic pin tool.
3. Incubate the YEPD plates containing the query strains and the
DMA array strains at 23
C for 12–24 h.

3.2.5 MATa/alpha Diploid 1. To select for diploids, resulting MATa/alpha zygotes are
Selection and Sporulation pinned onto YEPD þ G418/clonNAT (MRN 4) plates using
the 1536 pin robotic pin tool.
2. Incubate the plates at 30
C for 1–2 days (ideally 36 h) (see
Note 5).
3. To induce sporulation, pin the resulting diploid cells onto
enriched sporulation medium containing G418 (MRN 5)
using the 1536 pin robotic pin tool.
4. Incubate the plates at 23
C for 5 days (see Note 6).
5. Pick cells from the edge of a colony on one of the sporulation
plates and distribute them in 10 μL of water on a cover slide.
Examine for tetrads under a light microscope and determine
the tetrad frequency. A frequency of >1% is sufficient.
6. Repeat step 5 for 2–3 plates to ensure that all cultures
sporulated.

3.2.6 MATa Meiotic 1. For the first round of haploid selection (HSI), pin sporulated
Progeny Selection cells onto (SD/MSG)  His/Arg/Lys þ canavanine/thialy-
(Haploid Selection I þ II) sine plates (MRN 6, HSI) using the 1536 pin robotic pin tool.
(See Note 7) 2. Incubate the pinned plates at 20
C for 2 days.
3. Repeat steps 1 and 2 for a second round of haploid selection
(HSII).

3.2.7 MATa kanMX 1. Pin the resulting cells from Subheading 3.2.6 onto (SD/MSG)
Meiotic Progeny Selection His/Arg/Lys þ canavanine/thialysine/kanMX plates
(Haploid Selection III) (MRN 7, HS III) using the 1536 pin robotic pin tool.
(See Note 8) 2. Incubate the pinned plates at 20
C for 2 days.
 QFA 585

3.2.8 MATa kanMX/ 1. Pin the cells from Subheading 3.2.7 onto (SD/MSG)  His/
hphMX Meiotic Progeny Arg/Lys/Leu þ canavanine/thialysine/kanMX/hphMX
Selection (Haploid plates (MRN 8, HS IV) using the 1536 pin robotic pin tool.
Selection IV) (See Note 9) 2. Incubate the pinned plates at 20
C for 2 days.

3.2.9 MATa kanMX/ 1. Pin the cells from Subheading 3.2.8 onto (SD/MSG)  His/
hphMX/natMX Meiotic Arg/Lys/Leu þ canavanine/thialysine/G418/clonNAT/
Progeny Selection (Haploid Hygromycin B plates (MRN 9, HS V) using the 1536 pin
Selection V) (See Note 10) robotic pin tool.
2. Incubate the pinned plates at 20
C for 2 days.

3.3 QFA SGA plates in 1536 format contain four biological replicates (R1-
R4) of 308 independent gene deletions per plate as well as his3::
3.3.1 Cultivation of Yeast
kanMX on the edges of the plate (Fig. 4). All strains are cultivated
Strains in Liquid Media
in 96-well culture plates in liquid and grown until saturation prior
to spotting. Using a 96-pin tool, 1536 strains from a final SGA
plate are transferred to 16 96-well culture plates using a BM3
robot. Figure 4 gives an overview of robotic inoculation from
1536 format to 96 format and the colony positions in the 96-well
format. The sterilization procedure for the BM3 is described in
Subheading 3.1.2.
1. Fill each well of 96-well culture plates (16 96-well culture
plates per 1536 format SGA plate) with 200 μL of
SDM_rhlk_CTGNH media (MRN 9). (see Notes 11–14).
2. Load the source (solid agar plates in 1536 format) and destina-
tion plates (96-well culture plates) onto the stackers of the
BM3 robot. A schematic overview of the order of plates in
the stackers is shown in Fig. 5.
3. Using a 1 mm 96-pin tool, transfer 384 strains (R1) from each
final 1536 format SGA solid agar plate into four 96-well culture
plates (Fig. 4).
4. After the transfer, visually check each 96-well culture plate to
confirm that each inoculated well contains a white dot (pooled
cells from the pin) at the bottom of the plate. Missing empty
wells in the 96-well culture plate should correspond to missing
colonies on the SGA plate.
5. Place the inoculated 96-well culture plates in zip lock plastic
bags. Seal bags and place in a temperature controlled incubator
at 20
C for 2 days (see Note 15).

3.3.2 Spotting Yeast Cultures are spotted onto solid agar plates in 384 format using a
Cultures 2 mm 96-pin tool on a Beckman Biomek FX with automated
carousel (allows handling of many plates). 96-well culture plates
and solid agar plates are placed in carousel stacker. Saturated cul-
tures in 96-well plates are resuspended by orbital shaking at
 Final SGA plate: 4 biological replicates (R1-R4) of
308 independent gene deletions (A-P) in 1536
format. his3::KANMX edge cultures
R1 R2 R1 R2 R1 R2 R1 R2 are highlighted in grey.
A B C D 1 1536-format
R3 R4 R3 R4 R3 R4 R3 R4 solid agar plate

R1 R2 R1 R2 R1 R2 R1 R2
E F G H
R3 R4 R3 R4 R3 R4 R3 R4
R1 R2 R1 R2 R1 R2 R1 R2
I J K L
R3 R4 R3 R4 R3 R4 R3 R4
R1 R2 R1 R2 R1 R2 R1 R2
M N O P
R3 R4 R3 R4 R3 R4 R3 R4
96-pin tool used to inoculate 1536 yeast
colonies into 16 96-well plates containing
liquid media.

A C B D E G F H
R1 R1 R1 R1 R1 R1 R1 R1
R1 R1 R1 R1 R1 R1 R1 R1
I K J L M O N P

R1
A C
R2 R2
R2 R2 R2
I K 16 96-format
A C liquid culture
R3 R3 plates
R3 R3 R3
I K
A C
R4 R4 R4
R4 R4
I K Q1 Q2 Q3 Q4
96-pin tool used to spot
saturated cultures onto solid agar plates
in 384 format.
4 384-format
solid agar
plates
R1 R2 R3 R4

Fig. 4 Schematic overview of the QFA spotting procedure: inoculation from 1536 colony plate to 96-well liquid
plates to spotting in 384 format. Strains grown in 1536 format on solid agar plates are inoculated into 96-well
culture plates. Individual genotypes are indicated by letters (A–P). At least four independent replicates (R1–R4)
of each genotype are present on each 1536 plate. Using a 96-pin tool, every fourth colony in a row and column
(e.g., A(R1), C(R1), I(R1), K(R1)) is picked and transferred to a 96-well culture plate. Four 96-well culture plates
(Q1–Q4) are required to inoculate one replicate of each genotype (e.g., all R1s). Each color (green, yellow,
blue, and red) represent one of the four quadrants. To transfer all 1536 colonies of one final SGA plate, 16 96-
well culture plates are required. Strains are grown to saturation (e.g., 2 days at 20
C). All four quadrants
(Q1–4) from one repeat (e.g., R1) are spotted onto the same solid agar plate in 384 format in the same pattern
as the final 1536 format SGA plate
 QFA 587

gripper 1 gripper 2
pin tool

conveyor belt

side 1

stacker 1 stacker 2

side 4 side 2

stacker 1 stacker 2
source: destination:
6 24 side 3
1536 plates 96-well plates top view of the carousel

side 1

Fig. 5 Overview of plate loading in the BM3 stacker. For each 1536 format source plate, four 96-well culture
destination plates containing 200 μL of selective media are placed on the stacker, allowing the transfer of one
replicate of each genotype into liquid media. Each stacker permits the transfer from six 1536 format solid agar
plates to 24 96-well culture plates. The 96-well culture plates corresponding to each 1536 source plates are
indicated by the same color

1000 rpm for 20 s on a deck-mounted Variomag Teleshake.

The pin tool is cleaned and sterilized prior to each spotting (see
Subheading 3.1.2). A typical deck layout is shown in Fig. 2 (see
Notes 16–18).
1. Place the 96-well culture plates from Subheading 3.3.1 in the
stacker of the Biomek FX. Cut off corners of the 96-well
culture plates face inward.
2. Place sterile rectangular solid agar plates (MRN9) in the
stacker. Cut off corners of GreenLab plates (on short edge of
one side of the plate) face inward and lids upward. Cut off
corners of Nunc plates (on long edge of one side of the plate)
face right and lids upward.
3. Use a 2 mm 96 pin robotic pin tool to spot four 96-well culture
plates onto rectangular solid agar plates in 384 format (Fig. 4)
(see Note 19).
4. Spotted rectangular plates are transferred to an S&P Robotics
automated imager. (see Note 20).

3.3.3 Imaging High-throughput imaging takes place in an S&P Robotics auto-

mated imager. The robot repeatedly takes the plates from a
temperature-controlled incubator, removes the plate lid and
588 Eva-Maria Holstein et al.

transfers the plate to an evenly illuminated space underneath a

digital single-lens reflex camera (DSLR) with a 60 mm macro lens
mounted on it. The imager computer controls the camera to cap-
ture an image at high resolution. To allow more direct comparison
between images, images are captured in Manual mode (i.e., fixed
aperture, shutter speed, ISO and color temperature) and with
autofocus switched off. The image file is stored with a systematic
filename. After image capture the plate is re-lidded and returned to
the incubator until the next time point. We typically grow cells at
fixed temperatures ranging from 20 to 37
C and take images every
4 h. (see Notes 21 and 22).
1. Set the incubator to the desired temperature.
2. Using the Robosoft software controlling the S&P robotics
automated imaging system, enter the cycle interval of image
capture, temperature, length of the experiment, batch number,
and number of plates. Assign an Image Data Directory for
storing plate images from the experiment.
3. Assign a unique barcode for each experiment. Image filenames
are constructed as during SGA (see Subheading 3.2.2, step 3).
4. Take the first image immediately after placing the solid agar
plates in the incubator to allow for a zero time point growth
measurement.
5. Plates are incubated and imaged for up to 5 days until colony
growth saturation. (see Note 23).

3.3.4 Experimental Metadata about each experiment are recorded in a tab-delimited

Description File text file, with a row for every plate. This can be created program-
matically or built by hand with any text editor or spread sheet
software. Alternatively a preexisting file can be altered. Using
Excel, for example, set up a spreadsheet using the headers detailed
in Fig. 6 and Note 24 (see Note 24). The file can then be saved as a
tab-delimited file (File ! Save As ! Text (Tab-delimited) (*.txt)).

3.3.5 Library Description The position of each genotype in the library is recorded in a tab-
File delimited Library Description file. This metadata file is required for
data analysis in Subheading 3.3.6 and can be reused for multiple
experiments. An example Library Description file is shown in Fig. 7
(see Note 25).

3.3.6 Data Analysis Following repeated imaging of yeast cultures as they grow on solid
agar plates (Fig. 8), Colonyzer is used to estimate culture cell
densities from photographs [8]. Colonyzer locates cultures on
plates and tracks them over time, correcting for lighting effects
on plate images and estimates culture cell densities from the sum
of pixel intensities within each spot. It also estimates culture area as
well as the shape and color of each culture. Colonyzer software is
 QFA 589

Barcode Start.Time Treatment Medium Screen Library Plate RepQuad Client ExptDate User PI Inoc Condion
J000231_030_001 2013-12-11_10-45-01 30 SDM_rhlk_CTGNH cdc13-1 SDLV4_384 1 1 EMH '2013/12/10' EMH DAL CONC SDM_rhlk_CTGNH
J000231_030_002 2013-12-11_10-45-02 30 SDM_rhlk_CTGNH cdc13-1 SDLV4_384 2 1 EMH '2013/12/10' EMH DAL CONC SDM_rhlk_CTGNH
J000231_030_003 2013-12-11_10-45-03 30 SDM_rhlk_CTGNH cdc13-1 SDLV4_384 3 1 EMH '2013/12/10' EMH DAL CONC SDM_rhlk_CTGNH
J000231_030_004 2013-12-11_10-45-04 30 SDM_rhlk_CTGNH cdc13-1 SDLV4_384 4 1 EMH '2013/12/10' EMH DAL CONC SDM_rhlk_CTGNH
J000231_030_005 2013-12-11_10-45-05 30 SDM_rhlk_CTGNH cdc13-1 SDLV4_384 5 1 EMH '2013/12/10' EMH DAL CONC SDM_rhlk_CTGNH
J000231_030_006 2013-12-11_10-45-06 30 SDM_rhlk_CTGNH cdc13-1 SDLV4_384 1 2 EMH '2013/12/10' EMH DAL CONC SDM_rhlk_CTGNH
J000231_030_007 2013-12-11_10-45-07 30 SDM_rhlk_CTGNH cdc13-1 SDLV4_384 2 2 EMH '2013/12/10' EMH DAL CONC SDM_rhlk_CTGNH
J000231_030_008 2013-12-11_10-45-08 30 SDM_rhlk_CTGNH cdc13-1 SDLV4_384 3 2 EMH '2013/12/10' EMH DAL CONC SDM_rhlk_CTGNH
J000231_030_009 2013-12-11_10-45-09 30 SDM_rhlk_CTGNH cdc13-1 SDLV4_384 4 2 EMH '2013/12/10' EMH DAL CONC SDM_rhlk_CTGNH
J000231_030_010 2013-12-11_10-45-10 30 SDM_rhlk_CTGNH cdc13-1 SDLV4_384 5 2 EMH '2013/12/10' EMH DAL CONC SDM_rhlk_CTGNH
J000231_030_011 2013-12-11_10-45-11 30 SDM_rhlk_CTGNH cdc13-1 SDLV4_384 1 3 EMH '2013/12/10' EMH DAL CONC SDM_rhlk_CTGNH
J000231_030_012 2013-12-11_10-45-12 30 SDM_rhlk_CTGNH cdc13-1 SDLV4_384 2 3 EMH '2013/12/10' EMH DAL CONC SDM_rhlk_CTGNH
J000231_030_013 2013-12-11_10-45-13 30 SDM_rhlk_CTGNH cdc13-1 SDLV4_384 3 3 EMH '2013/12/10' EMH DAL CONC SDM_rhlk_CTGNH
J000231_030_014 2013-12-11_10-45-14 30 SDM_rhlk_CTGNH cdc13-1 SDLV4_384 4 3 EMH '2013/12/10' EMH DAL CONC SDM_rhlk_CTGNH
J000231_030_015 2013-12-11_10-45-15 30 SDM_rhlk_CTGNH cdc13-1 SDLV4_384 5 3 EMH '2013/12/10' EMH DAL CONC SDM_rhlk_CTGNH
J000231_030_016 2013-12-11_10-45-16 30 SDM_rhlk_CTGNH cdc13-1 SDLV4_384 1 4 EMH '2013/12/10' EMH DAL CONC SDM_rhlk_CTGNH
J000231_030_017 2013-12-11_10-45-17 30 SDM_rhlk_CTGNH cdc13-1 SDLV4_384 2 4 EMH '2013/12/10' EMH DAL CONC SDM_rhlk_CTGNH
J000231_030_018 2013-12-11_10-45-18 30 SDM_rhlk_CTGNH cdc13-1 SDLV4_384 3 4 EMH '2013/12/10' EMH DAL CONC SDM_rhlk_CTGNH
J000231_030_019 2013-12-11_10-45-19 30 SDM_rhlk_CTGNH cdc13-1 SDLV4_384 4 4 EMH '2013/12/10' EMH DAL CONC SDM_rhlk_CTGNH
J000231_030_020 2013-12-11_10-45-20 30 SDM_rhlk_CTGNH cdc13-1 SDLV4_384 5 4 EMH '2013/12/10' EMH DAL CONC SDM_rhlk_CTGNH

Fig. 6 Representative Experimental Description file. Experimental metadata for each plate in a QFA experiment
are recorded in separate rows in an Experiment Description file. A typical, genome-wide experiment with four
replicates of each spot includes 60 plates. In this example the experiment included 20 plates: all four replicate
cultures from five library plates. See Subheading 3.3.4 for interpreting column labels and details on file
construction

written in the Python programming language and once both are

installed, Colonyzer can also be used as a command-line tool.
Installation instructions and documentation can be found at
http://research.ncl.ac.uk/colonyzer/
We have developed a software package for QFA to read meta-
data from Library Description and Experiment Description files
and to associate metadata with each culture. The software, devel-
oped in the statistical modeling language R, can also be used to
construct growth curves from cell density estimates and to fit
population models (e.g., the logistic model) to curves, giving a
range of fitness estimates, including intrinsic growth rate. The
QFA R package also produces a wide range of visual reports.
These quality control reports: growth curve plots annotated with
estimated culture fitnesses, plots showing correlation between fit-
ness estimates for replicate cultures, plots examining spatial pat-
terns within and between plates and HTML reports containing
annotated previews of raw image data underlying fitness estimates.
More importantly, there are also functions for generating text
reports tabulating fitness estimates for each culture and fitness
summaries for replicates of each genotype examined.
By matching appropriately paired QFA screens: one control
screen of a library of strains crossed with the universal control strain
(see Subheading 2.5.2) and a second query screen containing the
same strains in the control crossed with a query mutation or grown
under a different condition, we can measure genetic interaction
strength (GIS). The QFA R package includes functions for estimat-
ing GIS as well as p-values for the statistical significance of genetic
interaction for each genotype in an interaction screen. The package
590 Eva-Maria Holstein et al.

Library ORF Plate Row Column Notes

SDLV4_1536 YBR184W 1 3 27 NA
SDLV4_1536 YBR184W 1 3 28 NA
SDLV4_1536 YAR029W 1 3 29 NA
SDLV4_1536 YAR029W 1 3 30 NA
SDLV4_1536 YBR217W 1 3 31 NA
SDLV4_1536 YBR217W 1 3 32 NA
SDLV4_1536 YBR174C 1 3 33 NA
SDLV4_1536 YBR174C 1 3 34 NA
SDLV4_1536 YBL043W 1 3 35 NA
SDLV4_1536 YBL043W 1 3 36 NA
SDLV4_1536 YAL059W 1 3 37 NA
SDLV4_1536 YAL059W 1 3 38 NA
SDLV4_1536 YBL011W 1 3 39 NA
SDLV4_1536 YBL011W 1 3 40 NA
SDLV4_1536 YBL001C 1 3 41 NA
SDLV4_1536 YBL001C 1 3 42 NA
SDLV4_1536 YBR036C 1 3 43 NA
SDLV4_1536 YBR036C 1 3 44 NA
SDLV4_1536 YBL102W 1 3 45 NA
SDLV4_1536 YBL102W 1 3 46 NA
SDLV4_1536 YOR202W 1 3 47 NA
SDLV4_1536 YOR202W 1 3 48 NA
SDLV4_1536 YOR202W 1 4 1 NA
SDLV4_1536 YOR202W 1 4 2 NA
SDLV4_1536 YAL018C 1 4 3 NA
SDLV4_1536 YAL018C 1 4 4 NA
SDLV4_1536 YBR210W 1 4 5 NA

Fig. 7 Representative Library Description file. The physical location (Plate, Row and Column) of each strain
genotype (library ORF) in a QFA experiment is recorded in the library description file. This file contains one row
per library spot, typically ~6000 rows per library. Once constructed, this library file can be reused for every
QFA experiment using the specified library. Optionally, it can be used to record the contents of multiple
libraries, by specifying different sets of data annotated with different Library names. See Subheading 3.3.5 for
details on file construction

also includes functions for interactive visualization of evidence for

genetic interaction in pairs of screens. Installation instructions and
documentation can be found at: http://qfa.r-forge.r-project.org/
Steps to measure fitness and genetic interaction strengths and
visualize results using the QFA R package are outlined below.
Following programming convention, the names of functions
(named sections of programs that carry out specific tasks) are
written in fixed-width font (e.g., colonyzer.read) (see Note 26).
QFA 591
592 Eva-Maria Holstein et al.

3.3.7 Measuring Fitness 1. Navigate to the Image Data Directory (see Subheading 3.3.3,
step 2) containing the images captured during the experiment
and run Colonyzer. This will generate two new sub-directories,
one containing all analyzed images (Output_Images) and one
containing all data files (Output_Data).
2. Go to the Output_Images folder and confirm alignment of the
spots within the rectangular Colonyzer segments. If automatic
alignment is poor, consider using a Colonyzer calibration file to
assist with spot location. Colonyzer calibration files can be built
using the Parameteryzer software which is installed alongside
Colonyzer.
3. Open the QFA package in R. Load the Experimental Descrip-
tion file (Subheading 3.3.4), Library Description file (Sub-
heading 3.3.5) and Colonyzer output files (Subheading
3.3.6.1.2) using colonyzer.read. This function associates meta-
data with each spot, grouping cell density data into growth
curves.
4. Export the R data frames as QFA Raw data text files to an
experimental results directory (write.table).
5. Use the QFA package to fit the logistic population model to
observations (qfa.fit) and to generate .pdf report files including
plots of growth curves and model fits together with fitness
estimates for each spot (qfa.plot) (see Note 27).
6. Export the R data frames as QFA Fitness data text files to an
experimental results directory (write.table).

3.3.8 Measuring Genetic 1. Select an appropriately matched pair of query and control
Interaction Strength screens. For example, if using a temperature sensitive query
mutation xyz-1 and an array of strains from the single deletion
library, abcΔ, then compare fitnesses from the double mutant
screen xyz-1 abcΔ with the single mutant control screen XYZ
abcΔ, where strains have been grown in the same environments
(e.g., on the same medium and at the same temperature).
2. Load the QFA fitness data text files for both the query and
control screens from the experimental results directory (read.
delim).
3. Carry out genetic interaction (or epistasis) analysis (qfa.epi), as
described in [2].


Fig. 8 Schematic overview of imaging, growth curve analysis, GIS calculation, and visualization. Spotted
colonies are photographed every 4 h for up to 5 days. Growth curves (cell density estimates over time) are
generated for each spot by image analysis and a generalized logistic model (black or red line) is fitted to the
cell density estimates (circle or triangle). Evidence for genetic interaction can be visualized as points on a 2D
scatter plot comparing fitnesses of two matched QFA screens that deviate from a regression through the origin
(model of genetic independence)
 QFA 593

4. Generate plots for visualizing genetic interaction (qfa.epiplot)

and tab-delimited GIS text files summarizing genetic interac-
tion strengths and significance of interaction for each array
mutation (report.epi). These tabulated data can be sorted by
GIS and filtered by q-value (p-value after correction for multi-
ple testing). For example, users often classify “hits” as array
mutations which have a q-value <0.05. Hits can then be ranked
by GIS for follow-up. Note that both strongly negative and
strongly positive GIS are informative: check both the top and
bottom of ranked lists for interesting hits.
5. Optionally, load GIS text files into iRVis, an interactive visuali-
zation tool included in QFA R package to allow rapid browsing
of screen data. Specific instructions for installing and using the
iRVis tool can be found at http://qfa.r-forge.r-project.org/
visTool/

4 Notes

1. Monosodium glutamic acid (MSG) is used as a nitrogen source

in synthetic media containing G418 or clonNAT as ammonium
sulfate hinders the function of G418 and cloNAT. Ammonium
sulfate is used as a nitrogen source in any synthetic media that
does not contain G418 or clonNAT.
2. Libraries and query strains are stored at 80
C in 15% glyc-
erol. Query strains used for any SGA are always revived from
80
C. A master plate stock of the libraries can be kept in a
384-format on YEPD þ kanMX at 4
C. The master plate stock
is propagated once a month onto a new YEPD þ kanMX plates.
The library is revived from the frozen stock every 3 months.
For an SGA, the library must be freshly re-pinned from the
master plate stock onto a new YEPD þ kanMX plate.
3. Examine the plates after each SGA stage for any of the
following:
l Fungal or bacterial contamination occurring at any stage of
the SGA is dealt with by repeating the pinning from the
previous uncontaminated stage of the SGA or by excluding
the plate from the analysis.
l Gaps in the plate due to strain inviability, or due to being
intentionally left blank, are expected to be consistent in the
subsequent SGA stages. Growth in previously empty spaces
can be indicative of a potential mix up of plates or
contamination.
l Pinning artifacts such as large empty patches on part of the
plate are indicative of alignment issues of the pin tool and
need to be resolved before continuing the procedure.
594 Eva-Maria Holstein et al.

l The single gene deletion strain can1Δ::kanMX in the library

is expected to lose viability during the meiotic progeny
selection stages as the haploid meiotic spore progeny harbor
either can1Δ::kanMX of the library strain or can1Δ::STE2pr-
Sp_his5 of the query strain. Viability is indicative of an
incorrect genotype or of contamination.
l Evidence for genetic linkage at the end of the SGA proce-
dure. Query strains containing a mutated gene residing
close to a gene that is deleted in the library strain should
not produce double mutants when crossed due to genetic
linkage. Therefore such double mutant colonies should not
be visible on the final SGA stage (HS V).
4. One culture is used as a backup in case of contamination.
Typically, one 5 mL overnight culture is used to create five
lawns of which three lawns are taken forward to the next step
(pinning to 1536 format). Each lawn is used to create six plates
in 1536 format. Take into account how many library plates will
be used in the experiment to determine how many cultures are
required for the procedure.
5. Timing of the diploid selection and sporulation is crucial.
Ideally incubate diploid selection plates for 36 h at 30
C
prior sporulation. Diploid selection plates cannot be stored at
4
C prior to sporulation. If necessary (e.g., weekend) store the
plates after mating at 4
C and then re-pin on diploid selection
plate. At all other stages the plates can, if required, be kept at
4
C for a couple of days before the next step.
6. Temperature is crucial for sporulation and should be kept
between 22 and 24
C. Sporulated plates can be kept at 4
C
for up to 4 months without significantly affecting spore viabil-
ity. >1% of spore frequency is required to continue with mei-
otic progeny selection.
7. HSI and HSII allow germination of MATa haploids containing
can1Δ::STE2pr-Sp-his5 and lyp1Δ markers. Diploids containing
a wild-type copy of CAN1 and LYP1 are not viable.
8. HSIII allows growth of MATa haploids containing can1Δ::
STE2pr-Sp-his5, lyp1Δ and kanMX markers.
9. HSIV allows growth of MATa haploids containing can1Δ::
STE2pr-Sp-his5, lyp1Δ, kanMX and hphmx markers.
10. HSV allows growth of MATa haploids containing can1Δ::
STE2pr-Sp-his5, lyp1Δ, kanMX, hphmx and clonNAT markers.
11. If QFA cannot be performed following SGA, the final SGA
plates can be stored up to a month at 4
C. Stored plates need
to be repinned onto a fresh final SGA plate
(SDM_rhlk_CTGNH, media no. 9) and incubated for 2 days
at 20
C prior the QFA procedure.
 QFA 595

12. Using this protocol, 1 L of liquid media can fill around 48 96-
well culture plates. 16 96-well culture plates are required to
culture every colony from a 1536 format SGA plate. For auto-
mated plate filling, a Biomek WellMate Dispenser can be used
to consistently dispense 200 μL into each well.
13. The number of solid agar plates required for the QFA will
depend on the number of technical replicates, conditions (tem-
peratures and drug concentrations) and the number of final
SGA plates in 1536 format. Four solid agar plates are required
for each SGA plate using one condition and no technical
replicates.
14. Before starting the procedure, bear in mind how many plates
can be run at once on the BM3 robot. Each 1536 format plate
requires four 96-well plates to transfer one biological repeat of
each strain on the plate. Each BM3 stacker can hold a maxi-
mum of 24 96-well culture plates. Therefore, up to six final
SGA plates in 1536 format can be used on each of the four sides
of the robot. For details of setting up the plates, see Fig. 5.
When setting up the robot, always inoculate a complete repeat
at once, do not split within repeats.
15. Liquid cultures can be kept at 4
C and reused for 1 month.
16. In addition to concentrated spotting, it is possible to spot
diluted cultures. To do this, saturated cultures in the 96-well
culture plates are diluted in 200 μL sterile water in a second 96-
well plate using a 2 mm pin tool (approximately 1:70 dilution).
Diluted cultures are then spotted onto solid agar plates.
17. A useful workflow is to spot one repeat at a time.
18. Prior to spotting, take into consideration the maximum capac-
ity of the imaging incubators.
19. When spotting overnight, the start can be delayed to ensure
that spotting finishes in the morning to ensure that the solid
agar plates are freshly spotted and can be transferred to the
incubator at time point 0.
20. When using a humidified incubator, wiping the bottom of the
plates with a one in ten dilution of washing up liquid prevents
condensation forming on the bottom of the plates which shows
up during imaging. Plates incubated in a nonhumidified incu-
bator are generally not affected by condensation problems.
21. Before entering the cycle interval time, ensure that the robots
are capable of imaging all the plates in the incubator within the
set time frame. When imaging a huge amount of plates, a
longer cycle interval might be required. It takes on average
2 min for one plate to be imaged and returned to the incubator.
22. If the experiment is split into several batches, ensure that all
repeats are imaged in the same incubator as the intensity of the
596 Eva-Maria Holstein et al.

light bulbs and the color of the diffuser changes over time,
which might affect analysis. The same consideration applies to
control QFAs that will be compared to another QFA.
23. Image plates for up to 5 days. Incubating the plates for too
long increases the risk of fungal or bacterial contamination.
Also, agar will eventually dry and crack, making spot location
and tracking more difficult.
24. Column headers for the experimental description file:
Barcode—A unique identifier for each plate. Barcodes can be
generated automatically by imaging robot software, for
example, (see Subheading 3.3.3, step 3).
Start.Time—The starting time of the experiment, should be an
accurate estimate of the inoculation time for that plate.
The time format should be yyyy-mm-dd_hh-mm-ss.
Treatment—Information about external conditions applied to
plates, such as incubation temperature or radiation.
Medium—Contents of the agar used in the screen (e.g., nutri-
ents or drugs added).
Screen—An easily interpretable screen name.
Library—Library name (should correspond to a Library name
in the Library Description file, see Subheading 3.3.5).
Plate—Libraries can have thousands of yeast strains across
multiple different plates. The Plate Number column is
used to link the barcode of the plate to the correct library
plate; this allows us to map strains to specific positions on
specific plates. The number should correspond to a plate
number in the Library Description file (see Subheading
3.3.5).
RepQuad—Genome-wide QFA experiments in 384 format are
derived from one of four replicates on final SGA plates
(1536 format). This number is a record of which of those
four replicates the QFA plate was derived from.
Client—The individual who requested the screen.
ExptDate—A more legible date for use in reports and plots.
User—The operator who performs the screen.
PI—The group leader of the client.
Inoc—Shorthand label for the inoculation density of the
experiments. This could be CONC (no dilution prior to
spotting) or DIL (dilution prior to spotting), for example.
Condition—The most important environmental attribute.
Depending on the design of the experiment, this could
be the temperature applied to the plate or the composition
of the media.
 QFA 597

25. Column headers for the library description file:

Library name—This must match the library name specified in
the Experimental Description file in Subheading 3.3.4.
ORF—Systematic open reading frame name specifying geno-
type of spot on library plate.
Plate—Library plate number.
Row—Spot row number.
Column—Spot column number.
Notes—Any notes that might be relevant.
26. For more detailed information on how to use QFA package
functions, including what information they require to work
and what their output will be, please refer to the QFA R
package documentation. For instance, once the QFA package
has been loaded, to access information about a specific func-
tion, type “?” into the R terminal followed by the function
name (e.g., ?colonyzer.read) and press Enter. Package docu-
mentation also includes a complete example of fitness and GIS
estimation from metadata files and Colonyzer output.
27. The QFA package allows edge colonies to be discarded as the
decrease in neighboring colonies leads to more available nutri-
ents and image analysis difficulties near plate walls. The QFA
package also allows for stripping of specific genotypes if
required (e.g., failed SGA, linkage).

References

1. Zubko MK, Guillard S, Lydall D (2004) Exo1 5. Downey M et al (2006) A genome-wide screen
and Rad24 differentially regulate generation of identifies the evolutionarily conserved KEOPS
ssDNA at telomeres of Saccharomyces cerevisiae complex as a telomere regulator. Cell 124
cdc13-1 mutants. Genetics 168(1):103–115 (6):1155–1168
2. Addinall SG et al (2011) Quantitative fitness 6. Tong AH, Boone C (2006) Synthetic genetic
analysis shows that NMD proteins and many array analysis in Saccharomyces cerevisiae. Meth-
other protein complexes suppress or enhance ods Mol Biol 313:171–192
distinct telomere cap defects. PLoS Genet 7(4): 7. Breslow DK et al (2008) A comprehensive strat-
e1001362 egy enabling high-resolution functional analysis
3. Banks AP, Lawless C, Lydall DA (2012) A quan- of the yeast genome. Nat Methods 5
titative fitness analysis workflow. J Vis Exp (66): (8):711–718
e4018 8. Lawless C et al (2010) Colonyzer: automated
4. Andrew EJ et al (2013) Pentose phosphate path- quantification of micro-organism growth char-
way function affects tolerance to the G- acteristics on solid agar. BMC Bioinformatics
quadruplex binder TMPyP4. PLoS One 8(6): 11:287
e66242
 Chapter 39

Rewiring the Budding Yeast Proteome using Synthetic

Physical Interactions
Guðjón Ólafsson and Peter H. Thorpe

Abstract
Artificially tethering two proteins or protein fragments together is a powerful method to query molecular
mechanisms. However, this approach typically relies upon a prior understanding of which two proteins,
when fused, are most likely to provide a specific function and is therefore not readily amenable to large-scale
screening. Here, we describe the Synthetic Physical Interaction (SPI) method to create proteome-wide
forced protein associations in the budding yeast Saccharomyces cerevisiae. This method allows thousands of
protein–protein associations to be screened for those that affect either normal growth or sensitivity to drugs
or specific conditions. The method is amenable to proteins, protein domains, or any genetically encoded
peptide sequence.

Key words Green fluorescent protein (GFP), GFP-binding protein (GBP), Nanobody, Chromobody,
Selective ploidy ablation, Protein-protein interactions (PPI), High-throughput screen

1 Introduction

Fusing proteins together by linking their open reading frames is a

powerful method to query the role of specific protein–protein
interactions. For example, fusing Cdc13 with Cdc2 in fission
yeast is sufficient to drive cell cycle progression [1] and fusing
spindle assembly checkpoint components arrests cell cycle progres-
sion in budding yeast [2]. These studies underlie the notion that a
protein’s function often depends upon its interacting partners and
more broadly that many cellular functions are controlled by the
spatial organization of proteins [3], a concept familiar to the cell
signaling field [4]. However, such studies rely upon making a
separate genetic construct for each fusion, which limits the ability
to test large numbers of possible fusions. To address this issue we
have made use of two resources. First, a GFP binding protein (GBP,
also called nanobody, nanotrap, or chromobody—the latter when
linked to a fluorophore) derived from an alpaca antibody heavy
chain [5–11]. Genetically encoded GBPs are typically 11–13 kDa

Marco Muzi-Falconi and Grant W. Brown (eds.), Genome Instability: Methods and Protocols, Methods in Molecular Biology,
vol. 1672, DOI 10.1007/978-1-4939-7306-4_39, © Springer Science+Business Media LLC 2018

599
600 Guðjón Ólafsson and Peter H. Thorpe

Fig. 1 The cartoon illustrates binding between a query protein (blue) tagged with
GFP (green) and a target protein (orange), tagged with GBP (pink) and RFP (red)

and bind to GFP with high-affinity in vitro (Kd < 1 nM). Impor-
tantly for this approach, proteins tagged with GBP bind to GFP-
tagged proteins in vivo (for examples see [12] or [13] (Fig. 1)).
Thus fusing a protein of interest—a target protein—with the GBP
allows it to associate with any GFP-tagged query protein. Second,
in the budding yeast, Saccharomyces cerevisiae, a genome-wide
library of strains was constructed each encoding a separate endoge-
nously tagged GFP protein [14]. Our Synthetic Physical Interaction
(SPI) method introduces a target protein tagged with GBP into the
GFP library, each of which contains a separate GFP-tagged query
protein, to create thousands of protein–protein associations. Each
interaction is then assayed for growth or sensitivity to a particular
condition.

2 Materials

2.1 GFP-Binding There are various peptide sequences derived from camelid antibo-
Protein (GBP) dies that bind to GFP. The first GBP clone was isolated by Uhlrich
Rothbauer, Heinrich Leonhardt and colleagues [5, 6], but more
recently a large repertoire of GBPs have been isolated [9]. This new
array of GBPs theoretically enable the SPI method to employ two
GBPs that bind to different epitopes of GFP and so recruit either
two copies of the same protein or two different proteins to GFP.
The sequence encoding GBP is sufficiently short (<400 bp) that it
can be introduced at either the 50 or 30 end of open reading frames
to create amino- or carboxy-tags of GBP target proteins with ease
and usually does not perturb their function. We use a plasmid to
drive expression of the GBP-tagged target protein.

2.2 Plasmids for In principle any suitable plasmid could be used to introduce the
Synthetic Physical GBP-tagged target protein into the GFP strains. We have used a
Interactions single copy CEN plasmid based upon pWJ1512 [15] (Fig. 2). This
plasmid contains a LEU2 marker gene for selection and a CUP1
promoter to drive expression of the GBP construct. The CUP1
promoter is active without copper, but its expression can be ele-
vated with added copper [16]. The strains of the GFP collection are
auxotrophic for leucine, uracil, and methionine. However, the
selective ploidy ablation method to transfer plasmids uses selection
 Rewiring the Budding Yeast Proteome using Synthetic Physical Interactions 601

Fig. 2 An example of the methodology used to create plasmids for the SPI screen
is shown. (a) GBP–RFP is inserted into a suitable plasmid to create a GBP–RFP
control plasmid. (b) A gene of interest (GOI) is separately inserted into the same
plasmid to create a target protein control. (c) The gene of interest (GOI) is then
inserted into the GBP–RFP plasmid (a) to create a construct encoding a fusion
between a target protein and GBP–RFP

against URA+; consequently plasmids encoding LEU2 or MET15

could be used in addition to any drug selection markers (such as
kanamycin, hygromycin, or nourseothricin). In addition to a fusion
plasmid (Fig. 2c), we typically use a GBP plasmid (Fig. 2a) and
“Gene of Interest” plasmid (Fig. 2b) as controls in the SPI screens.

2.3 Yeast Strains/ The GFP collection of strains is currently distributed by Thermo-
GFP Collection Fisher Scientific (MA, USA) and contains 4159 strains. All strains
are derivatives of BY4741 (MATa his3Δ1 leu2Δ0 met15Δ0 ura3Δ0)
[14, 17]. The library is delivered in frozen glycerol stocks in 96-well
plates and after thawing can be readily transferred onto rectangular
agar plates using pinning tools.

2.4 Universal Donor The MATα Universal Donor Strain (W8164-2B) is available from
Strain for Selective the Rothstein lab [15], MATα CEN1-16GCS can1-100 his3-11,15
Ploidy Ablation leu2-3,112 LYS2 met17 trp1-1 ura3-1 RAD5 (where CEN1-16GCS
indicates that all the centromeres contain a counterselectable Kluy-
veromyces lactis URA3::GAL promoter cassette [18]). This strain
can be grown with glucose as a carbon source and is readily trans-
formed with plasmids for the SPI screen.
602 Guðjón Ólafsson and Peter H. Thorpe

2.5 Media and Stock 1. YPD liquid media: 20 g/L Bacto peptone, 10 g/L Yeast extract
Solutions and 2% (weight by volume) glucose. Dissolve these ingredients
in water and sterilize by autoclaving.
2. YPD solid agar media: 20 g/L Bacto peptone, 10 g/L Yeast
extract, 2% (weight by volume) glucose and 20 g/L Bacto agar.
Dissolve ingredients in water (agar will not dissolve) then
sterilize and melt agar by autoclaving. Allow to cool to
~55  C prior to pouring plates. Each rectangular plate should
take ~50 ml of media (see Note 1).
3. Leucine amino acid mix: Amino acid supplement mixes,
including drop-outs, such as the “minus leucine” media
described here are available commercially. We use a mix con-
taining the following: 20 mg/L adenine sulfate, arginine sul-
fate, histidine–HCl, methionine, tryptophan, and uracil;
30 mg/L isoleucine, lysine–HCl, and tyrosine; 50 mg/L phe-
nylalanine and 150 mg/L valine.
4. GAL Leu solid agar media: Prepare a solution containing
1.7 g/L yeast nitrogen base without amino acids, 5 g/L
ammonium sulfate (see Note 2), 41 mg/L Leucine amino
acid mix (described in item 3), 20 g/L Bacto agar. Dissolve
ingredients in water (agar will not dissolve) then sterilize and
melt agar by autoclaving. After the media has cooled to
55–60  C, add 100 ml of 20% (weight by volume) sterile
galactose solution to each L of media, mix and pour as for
YPD plates (see Note 6 for diploid assay).
5. GAL Leu 5-FOA solid agar media: Mix 20 g Bacto agar with
500 ml of water in a 1-L bottle, melt using an autoclave.
Prepare a 500 ml solution in water containing 1.7 g yeast
nitrogen base without amino acids, 5 g ammonium sulfate
(see Note 2), 41 mg Leucine amino acid mix (from item 3),
30 mg uracil and 750 mg 5-fluoro-orotic acid (5-FOA) and 2%
(weight by volume) galactose. Allow this solution to dissolve
with stirring for approximately 1 h at room temperature, then
filter-sterilize through a 0.22 μm filter and place in a 60  C
incubator or water bath. When the autoclaved agar has cooled
to 55–60  C, add in the warmed solution, mix well by gentle
stirring and pour the plates immediately.

2.6 Plates for SPA All plates described in this method are the Singer Plus plates (Singer
Instruments Ltd, Somerset, UK), which are designed for use with
the ROTOR robotic pinning platform. However, when using other
pinning tools, any 12 cm by 8 cm sterile rectangular plates are
suitable; for example the Nunc™ OmniTray™ (Thermo Scientific,
MA, USA).

2.7 Pinning Robot The ROTOR robotic pinning platform (Singer Instruments Ltd,
Somerset, UK) allows 96, 384, and 1536 arrays of yeast colonies on
 Rewiring the Budding Yeast Proteome using Synthetic Physical Interactions 603

rectangular agar plates to be copied from plate to plate using sterile

pins. It is also simple to convert between 96, 384, and 1536 arrays.
If hand pinning tools are used, it is very important to ensure that an
even pinning pressure is applied in each copying step.

2.8 Plate Imaging In principle, most desktop scanners that function in transmission
mode are suitable for collecting images of plates. Images should be
at least 300 dpi. It is possible to use specialist hardware such as the
Phenobooth™ (Singer Instruments Ltd, Somerset, UK) to capture
images of the plates.

3 Methods

3.1 Selective Ploidy We use the following general method for SPI screens (Fig. 3)
Ablation
1. The arrayed GFP strains should be copied onto rectangular
YPD agar plates. We typically have a copy of the GFP library
stored at 4  C on YPD plates, with 384 strains on each plate.
Prior to the screen we copy this library onto fresh YPD plates at
1536 colonies/plate density, i.e., four copies of each of the 384
strains on each plate.
2. Grow up 5–10 ml cultures of the UDS strain (W8164-2B) each
containing a separate plasmid for the SPI assay (e.g., GBP
alone, kinase alone, kinase-GBP, and kinase mutant-GBP) in
selective media at 30  C, shaking, overnight. Collect the cells
by centrifugation at 5000  g for 5 min. Each lawn provides
sufficient UDS cells to mate with four arrays of GFP strains, if
there are more than four plates of GFP strains then additional
lawns will need to be created (see Note 3).
3. Resuspend the cell pellets in ~400 μl of growth media and plate
all of this onto YPD plates. Using 4 mm sterile glass beads to
spread the cells onto the rectangular agar plates gives a more
even lawn of cells than a glass spreader tool.
4. Grow these lawns overnight in 30  C incubator.
5. Use a pinning tool or robotic pinner (e.g., Singer ROTOR
pinning robot) to copy the GFP strains onto fresh YPD plates
(see Note 4).
6. Overlay these YPD plates with the cells from the lawns by
pinning in the same way as step 5.
7. Grow these “mating” plates for 6–7 h at 30  C (see Note 5).
8. Copy the colonies by pinning onto GAL-Leu rectangular agar
plates (see Note 6).
9. Grow for 24 h in a 30  C incubator.
10. Copy the colonies by pinning onto GAL-Leu 5-FOA rectan-
gular agar plates.
604 Guðjón Ólafsson and Peter H. Thorpe

Fig. 3 The key steps involved in performing a SPI screen are illustrated with a
daily timeline on the left side. Cultures of the UDS strain containing the SPI
plasmids are grown and plated to create lawns. In parallel, the GFP library is
prepared on agar plates. The GFP strains are mated separately with each of the
UDS plasmid strains by copying them together on an YPD plate. These arrays are
then copied sequentially onto galactose leucine (GAL LEU) and then galac-
tose leucine þ 5FOA (GAL LEU þ 5FOA) medium. All incubations are done at
30  C. The resulting arrays of strains are then imaged to compare controls with
experiments
 Rewiring the Budding Yeast Proteome using Synthetic Physical Interactions 605

11. Grow for 24–72 h (depending on colony sizes) in a 30  C

incubator.
12. Capture images of the plates.

3.2 Data Analysis It is possible to assess the strongest interactions by eye (Fig. 4a),
however, we recommend a more quantitative analysis using imag-
ing. The plates from the SPI screen are imaged with a desktop
scanner in transmission mode (colonies should appear dark on a
light background) or plate imager (see Note 7). The resulting
images are analyzed computationally to determine colony sizes:
1. We employ the “CM engine” (Colony Measurement engine), a
freely available ImageJ plugin and part of the ScreenMill suite
of software [19]. CM engine analyses images and reports the
size of each colony on the plate in pixels, colony circularity and
location coordinates. We use the default settings (see Note 8).
The output file from the CM engine is a “Log file,” which lists
the measurements for each colony.
2. The “Log file” is uploaded to the next part of ScreenMill: The
Data Review engine (DR engine), which is available online:
http://www.rothsteinlab.com/tools/screen_mill/dr_engi-
ne_setup. Here the comparisons between the colonies on the
control plates and the experiment plates are performed. For an
easy readout of the data we recommend generating a “Key
file.” The Key file is a list ascribing each position on each
plate with a particular query GFP-tagged gene. We run the
DR engine with the default settings (see Note 9).
3. Plate based screens often suffer from spatial anomalies, such as
colonies growing larger at the top of the plate than the bottom
[20, 21]. We use a simple smoothing algorithm to spatially
adjust the relative growth rates on plates [13] (Fig. 5). The
PERL script is freely available online: http://www.source-
forge.net/projects/zspatialcorrect/files/spatial1_0.plx/
download.
The resulting data output includes the Log Growth Ratio
(LGR) or z-score for each query protein to assess specific interac-
tions that affect growth (Fig. 4b). The LGR is the natural log of the
average colony size on the controls divided by the average colony
size of the experiment. Hence, higher LGRs indicate a stronger
growth defect. To understand how a specific LGR equates with a
growth difference, we calculate the theoretical difference in the
number of yeast cells relative to colony size assuming that colonies
are hemispheres (Fig. 6a). This works well in practice and shows
that a typical LGR of 0.4 equates with ~80% more cells on the
control than the experiment (Fig. 6b). The formula we derive for
converting LGR to the ratio of the number of yeast on control and
606 Guðjón Ólafsson and Peter H. Thorpe

Fig. 4 An example of images of plates resulting from a SPI screen is shown. (a) Each plate has the same array
of 384 GFP strains, each copied in quadruplicate (hence 1536 colonies in total). On each separate plate the
GFP strains contain a different plasmid—the GBP control, the untagged target protein control, and the
target–GBP fusion, from left to right. The three highlighted strains (red boxes) show three strains that are
affected in the experiment but not controls. (b) An example of a SPI screen plate, now with 96 GFP strains each
with 16 replicates. Two strains that result in a SPI are highlighted with rectangles (orange and green). The
Mean Log Growth Ratios (LGR) of the two controls versus experiment are then plotted to show the strongest
SPIs (the orange and green dots represent the same SPIs indicated with the rectangles and are the strongest
SPIs on this plate)
 Rewiring the Budding Yeast Proteome using Synthetic Physical Interactions 607

Fig. 5 A spatial defect caused by cells growing differently across a plate is

highlighted by plotting the LGR or z-scores by their row and column (a); a linear
trend line is added to clarify the spatial defect. The bias is also visualized by a
color-coded map of the plate, showing high z-scores in yellow (light) and low
z-scores in blue (dark) (shown below). A simple smoothing algorithm can correct
this defect by normalizing scores in each row and column (b)

experiment (YR) is YR ¼ e1.5LGR. We typically will retest all associa-

tions that have high LGRs with at least 16 replicates to confirm
their reproducibility (for example see Fig. 4b).

3.3 Microscopy It is possible to validate that the GFP- and GBP-tagged proteins are
interacting using fluorescence imaging, since the GBP tag also
includes an RFP sequence (Fig. 7).
1. From the final GAL Leu 5-FOA plate, colonies are picked and
placed into 5 ml of GAL Leu þ Ade (see Note 10; see Note 6
for diploid cells) and grown for ~24 h at 23  C in a shaking
incubator.
2. Collect the cells by centrifugation at 5000  g for 5 min,
resuspend in 5 the pellet volume (GAL Leu þ Ade) (see
Note 11; see Note 6 for diploid cells).
3. Place 1.25 μl of cells and 1.25 μl agar (Leu þ Ade with 1.4%
low melting point agarose—maintained as a liquid at 42  C)
onto a glass microscope slide (see Note 12) cover with a cover-
slip and use fluorescence imaging to detect fluorophores.

4 Notes

1. It is very important that the plates are poured on a level surface

to ensure that the agar surface is flat and level. This will ensure
even colony growth across the plate. If the surface of the agar is
608 Guðjón Ólafsson and Peter H. Thorpe

Fig. 6 The assessment of growth relies upon measuring colony area as a

surrogate for the number of yeast on the plate. (a) We theorize that colonies
approximate to hemispheres and consequently the relationship between area (a)
and volume (v) is v ¼ 0.38a1.5. (b) We collected and counted the number of cells
in colonies to confirm the relationship between the ratio of yeast on control
versus experiment (RY) and the LGR and find that broadly our data fit the model.
We note that a cutoff of LGR ¼ 0.4, the minimum LGR observable by eye,
equates with growth ratio of 1.8, a 45% drop in the number of cells on the
experiment versus the control (inset)

not level it can affect yeast colony pinning, i.e., if the agar level/
height is low on one side compared to another it can cause the
pins to not reach the low side. Additionally, if one side of the
plate has higher/more agar then the colonies on that side will
have more nutrients and thus could grow larger as a result.
2. Yeast nitrogen base is often sold as a mix with ammonium
sulfate, in which case use 6.7 g/L of this mix for synthetic
media. It is possible to buy commercially mixtures of amino
acids and also “homemade” recipes vary slightly from lab
to lab.
Fig. 7 Examples fluorescence microscopy images of GBP–RFP being recruited to GFP-tagged proteins, the
columns from left to right show GFP signal, RFP signal, GFP–RFP merge, and differential interference contrast
(DIC) signal. The scale bar is 5 μm in all images. (a) A GFP-tagged kinase can be recruited to different cellular
compartments via interaction with GBP-target proteins. (b) A GFP-tagged kinase is recruited to a GBP-tagged
kinetochore and a GBP-tagged kinase is recruited to a GFP-tagged kinetochore
610 Guðjón Ólafsson and Peter H. Thorpe

3. Each plate of GFP strains will need to be copied onto several

YPD plates (i.e., make multiple copies). The number of copies
will depend upon the number of separate plasmids (or plasmid
combinations) that you are using in the experiment. For exam-
ple, in a simple experiment you may have three UDS lawns each
containing a different plasmid (encoding GBP alone, protein of
interest alone, and protein of interest fused with GBP). How-
ever, in more complex experiments you may have more than
three conditions and consequently you will need more copies
of the GFP strains.
4. The options for the Singer ROTOR pinning robot settings are
quite extensive, for example the pinning pressure, diameter,
and frequency can be customized. For 1536 colony density
format we use the following settings for source and target
plates: Pinning; Pin Pressure: 25%, Repeat Pin: two times.
Dry Mix: Diameter: 0.1 mm, Cycles: two rotations. For 384
colony density format we use the following settings for source
and target plates: Pinning; Pin Pressure: 10%, Repeat Pin: two
times. Dry Mix: Diameter: 0.3 mm for source plates and
0.1 mm for target plates, Cycles: two rotations. For “lawn
plates” we change The ROTOR robot pinning settings of the
“Dry Mix: Diameter” to 1 mm for source plates. This is to
ensure that cells are picked up in cases where there might be
gaps or uneven cell density on lawn plates, thus pinning with an
increased diameter further ensures cells are picked up.
5. The “mating” step can be done overnight without affecting
results, but prolonged time on YPD can result in “overgrowth”
of the colonies to the extent that separate colonies grow
together.
6. Alternatively to, or in parallel with the haploid assay, the cells
from the mating step can be kept as diploid cells by selecting on
Histidine Leucine media with glucose as a carbon source. A
diploid screen will determine dominant effects of the
GBP–GFP association, since these diploids have an extra set
of chromosomes encoding an untagged version of the GFP-
tagged query protein. For a diploid screen, steps 8–10 should
to be replicated using the following protocol: Copy the mated
colonies by pinning onto glucose His Leu rectangular agar
plates. Grow for ~24 h in a 30  C incubator. Copy the colonies
by pinning again onto fresh glucose His Leu rectangular
agar plates. Grow for ~24 h in a 30  C incubator before
imaging/analyzing. Note that diploid cells in glucose media
typically grow faster than haploid cells in galactose and 5-FOA,
thus the time needed to grow cells for the diploid assay is
reduced compared with the haploid assay. Media used for
microscopy has to be adjusted when imaging diploid cells;
Use glucose instead of galactose in the Leu þ Ade media
when growing cultures overnight for imaging.
 Rewiring the Budding Yeast Proteome using Synthetic Physical Interactions 611

7. We scan the plates using a desktop flatbed scanner (Epson

V750 Pro, Seiko Epson Corporation, Japan) at 300 dpi resolu-
tion in transmission mode. The plate lids are removed and the
plates are inserted upside down in the scanner. We save the
images in “tiff” format to use with the ScreenMill software.
8. We use the following default settings for the ScreenMill CM
engine ImageJ plugin: Fine-crop Mode: Automatic; Colony
measurement method: Standard; Plate Density: 384 or 1536;
Running Mode: Standard.
9. We run the DR Engine in default mode and use the following
Statistical Test settings: Normal Distribution; Do not Bonfer-
roni correct p-values. We sometimes have non-GFP control
colonies on the plates, for example when retesting phenotypes
associated with specific interactions after a proteome-wide
screen. In the these circumstances we use the “Designated
Controls” options under Data Normalization Method, how-
ever in cases when we do not use the non-GFP controls we use
the “Plate Median” option. When using the DR Engine you
can choose to exclude any colonies from the analysis, for exam-
ple because of a pinning issue or contamination on the plates.
ScreenMill has many other options and features that might be
of use to some users, but are not covered in this protocol.
10. Leu þ Ade media used for microscopy contains 100 mg/L
adenine, which helps to minimize the amount of autofluores-
cent purine precursors within cells.
11. The cells collected for imaging should be growing in log phase
for best results.
12. A total volume of 2.5 μl liquid is appropriate for 22 mm cover-
slips, assuming the liquid spreads out over the whole coverslip
area (4.83 cm2). This allows for 5.1 μm distance between slide
and coverslip—bear in mind that haploid yeast cells are ~5 μm
in diameter. Use increased liquid volume for diploid cells, since
they are larger.

Acknowledgments

This work was supported by a Medical Research Council (MRC)

UK centenary award and grant to P.T. (MC_UP_A252_1027). The
Francis Crick Institute is funded by the MRC UK, Cancer Research
UK, the Wellcome Trust, Imperial College London, University
College London and Kings College London. We thank Lisa Berry
for comments on the chapter.
612 Guðjón Ólafsson and Peter H. Thorpe

References
1. Coudreuse D, Nurse P (2010) Driving the cell MPF feedback loop to promote mitotic com-
cycle with a minimal CDK control network. mitment in S. pombe. Curr Biol 23:213–222
Nature 468:1074–1079 13. Olafsson G, Thorpe PH (2015) Synthetic
2. Lau DT, Murray AW (2012) Mad2 and Mad3 physical interactions map kinetochore regula-
cooperate to arrest budding yeast in mitosis. tors and regions sensitive to constitutive Cdc14
Curr Biol 22:180–190 localization. Proc Natl Acad Sci U S A
3. Hagan IM, Grallert A (2013) Spatial control of 112:10413–10418
mitotic commitment in fission yeast. Biochem 14. Huh WK, Falvo JV, Gerke LC, Carroll AS,
Soc Trans 41:1766–1771 Howson RW, Weissman JS, O’Shea EK
4. Scott JD, Pawson T (2009) Cell signaling in (2003) Global analysis of protein localization
space and time: where proteins come together in budding yeast. Nature 425:686–691
and when they’re apart. Science 326:1220–1224 15. Reid RJ, Gonzalez-Barrera S, Sunjevaric I,
5. Rothbauer U, Zolghadr K, Muyldermans S, Alvaro D, Ciccone S, Wagner M, Rothstein R
Schepers A, Cardoso MC, Leonhardt H (2011) Selective ploidy ablation, a high-
(2008) A versatile nanotrap for biochemical throughput plasmid transfer protocol, identi-
and functional studies with fluorescent fusion fies new genes affecting topoisomerase I-
proteins. Mol Cell Proteomics 7:282–289 induced DNA damage. Genome Res
6. Rothbauer U, Zolghadr K, Tillib S, Nowak D, 21:477–486
Schermelleh L, Gahl A, Backmann N, Conrath 16. Butt TR, Sternberg EJ, Gorman JA, Clark P,
K, Muyldermans S, Cardoso MC, Leonhardt H Hamer D, Rosenberg M, Crooke ST (1984)
(2006) Targeting and tracing antigens in live Copper metallothionein of yeast, structure of
cells with fluorescent nanobodies. Nat Meth- the gene, and regulation of expression. Proc
ods 3:887–889 Natl Acad Sci U S A 81:3332–3336
7. Kubala MH, Kovtun O, Alexandrov K, Collins 17. Brachmann CB, Davies A, Cost GJ, Caputo E,
BM (2010) Structural and thermodynamic Li J, Hieter P, Boeke JD (1998) Designer dele-
analysis of the GFP:GFP-nanobody complex. tion strains derived from Saccharomyces cerevi-
Protein Sci 19:2389–2401 siae S288C: a useful set of strains and plasmids
8. Helma J, Cardoso MC, Muyldermans S, Leon- for PCR-mediated gene disruption and other
hardt H (2015) Nanobodies and recombinant applications. Yeast 14:115–132
binders in cell biology. J Cell Biol 18. Hill A, Bloom K (1989) Acquisition and pro-
209:633–644 cessing of a conditional dicentric chromosome
9. Fridy PC, Li Y, Keegan S, Thompson MK, in Saccharomyces cerevisiae. Mol Cell Biol
Nudelman I, Scheid JF, Oeffinger M, Nussenz- 9:1368–1370
weig MC, Fenyo D, Chait BT, Rout MP 19. Dittmar JC, Reid RJ, Rothstein R (2010)
(2014) A robust pipeline for rapid production ScreenMill: a freely available software suite for
of versatile nanobody repertoires. Nat Meth- growth measurement, analysis and visualiza-
ods 11:1253–1260 tion of high-throughput screen data. BMC
10. Hamers-Casterman C, Atarhouch T, Muylder- Bioinformatics 11:353
mans S, Robinson G, Hamers C, Songa EB, 20. Baryshnikova A, Costanzo M, Kim Y, Ding H,
Bendahman N, Hamers R (1993) Naturally Koh J, Toufighi K, Youn JY, Ou J, San Luis BJ,
occurring antibodies devoid of light chains. Bandyopadhyay S, Hibbs M, Hess D, Gingras
Nature 363:446–448 AC, Bader GD, Troyanskaya OG, Brown GW,
11. Muyldermans S (2013) Nanobodies: natural Andrews B, Boone C, Myers CL (2010) Quan-
single-domain antibodies. Annu Rev Biochem titative analysis of fitness and genetic interac-
82:775–797 tions in yeast on a genome scale. Nat Methods
7:1017–1024
12. Grallert A, Chan KY, Alonso-Nunez ML,
Madrid M, Biswas A, Alvarez-Tabares I, Con- 21. Collins SR, Schuldiner M, Krogan NJ, Weiss-
nolly Y, Tanaka K, Robertson A, Ortiz JM, man JS (2006) A strategy for extracting and
Smith DL, Hagan IM (2013) Removal of cen- analyzing large-scale quantitative epistatic
trosomal PP1 by NIMA kinase unlocks the interaction data. Genome Biol 7:R63
 Chapter 40

Reporter-Based Synthetic Genetic Array Analysis: A

Functional Genomics Approach for Investigating Transcript
or Protein Abundance Using Fluorescent Proteins in
Saccharomyces cerevisiae
Hendrikje Göttert, Mojca Mattiazzi Usaj, Adam P. Rosebrock,
and Brenda J. Andrews

Abstract
Fluorescent reporter genes have long been used to quantify various cell features such as transcript and
protein abundance. Here, we describe a method, reporter synthetic genetic array (R-SGA) analysis, which
allows for the simultaneous quantification of any fluorescent protein readout in thousands of yeast strains
using an automated pipeline. R-SGA combines a fluorescent reporter system with standard SGA analysis
and can be used to examine any array-based strain collection available to the yeast community. This protocol
describes the R-SGA methodology for screening different arrays of yeast mutants including the deletion
collection, a collection of temperature-sensitive strains for the assessment of essential yeast genes and a
collection of inducible overexpression strains. We also present an alternative pipeline for the analysis of R-
SGA output strains using flow cytometry of cells in liquid culture. Data normalization for both pipelines is
discussed.

Key words Yeast, Synthetic genetic array, SGA, Fluorescent protein, Reporter gene

1 Introduction

Reporter genes are useful tools for the screening of new regulators
of transcript or protein abundance. For example, gene regulation
can be indirectly monitored by fusing a promoter of interest to a
fluorescent reporter gene such as Green Fluorescent Protein
(GFP). Likewise, whole proteins can be tagged by fusing the GFP
open reading frame to the gene encoding the protein of interest.
Changes in gene expression or protein abundance can then be
detected by assaying levels of fluorescence under various genetic
and environmental conditions. A system called reporter synthetic
genetic array analysis, or R-SGA, takes advantage of this concept
[1]. R-SGA combines standard SGA technology [2] which allows

Marco Muzi-Falconi and Grant W. Brown (eds.), Genome Instability: Methods and Protocols, Methods in Molecular Biology,
vol. 1672, DOI 10.1007/978-1-4939-7306-4_40, © Springer Science+Business Media LLC 2018

613
614 Hendrikje Göttert et al.

for the introduction of any kind of marked gene into any set of
arrayed yeast strains with a dual-color reporter system. The R-SGA
method presented here employs a wild-type query strain, which has
the promoter of a gene of interest fused to a GFP gene carried on a
plasmid, and a control promoter fused to an RFP gene integrated in
the genome (Fig. 1). The query strain is crossed to an array of yeast
mutants and, since all genes of interest are linked to selectable
markers, a series of pinning steps on different selective media
produces an output array in which each strain carries the fluorescent
reporter genes and one mutation from the array strain. Fluores-
cence can then be imaged directly from colonies on a plate using a
fluorescence plate scanner or strains can be transferred into liquid
culture and assessed by flow cytometry to detect upregulation or
downregulation of the fluorescent protein of interest in response to
the tested perturbation.
The R-SGA method was first used to screen histone gene
promoter–GFP fusion reporters in the KanMX deletion array [3],
identifying a roster of loss-of-function mutants corresponding to

HTB1 promoter HTA1

promoter RPL39 endogenous loci

promoter GFP LEU2

reporter constructs CEN plasmid

promoter RFP HYGR

HO locus

Fig. 1 Reporter constructs to assess promoter activation. The query strain employed in the Reporter-Synthetic
Genetic Array analysis (R-SGA) carries two reporter constructs with promoters driving fluorescent reporters.
The example shows the reporter strain used to assess HTA1 expression with RPL39 promoter as control.
Arrows indicate direction of transcription
 Reporter-Based Synthetic Genetic Array Analysis: A Functional Genomics. . . 615

putative activators and repressors of histone gene expression [4].

These screens revealed cell cycle-dependent regulation of active and
repressive chromatin at histone gene promoters, controlled by a cell
cycle oscillator that serves as a master regulator of histone gene
activation [5]. A more recent study used the R-SGA methodology
to assay for increased abundance of the strongly DNA damage-
inducible protein Rnr3, to identify putative mutants exhibiting
genome instability [6]. In addition to screening for genome insta-
bility mutants in the KanMX deletion array, which includes ~4200
nonessential mutant yeast genes, a set of ~1000 essential yeast
alleles was assessed in this study, both in standard conditions and
in the presence of the DNA-damaging agent methylmethane sulfo-
nate [7, 8]. In theory the R-SGA approach is extensible to any
colony-based assay for which the readout is abundance of a fluores-
cent protein, for example plasmid or chromosome loss readouts or
assays for specific types of recombination.

2 Materials

2.1 Array and Query 1. The R-SGA pipeline can easily be adapted for use with different
Strains and Plasmids yeast collections. Three examples are presented here: a haploid
deletion collection [3], a collection of temperature sensitive (ts)
mutants [7], and a collection of yeast strains with plasmids
expressing genes with galactose inducible overexpression [9,
10]. Exemplar array and query strain and plasmid genotypes are
summarized in Table 1 (see Note 1).
2. Promoters described here are defined as the intergenic
sequence beginning at the first base pair (bp) upstream of a
given open reading frame (ORF) and extending to the adjacent
gene, up to a maximum of 1000 bp. This sequence is amplified
by polymerase chain reaction (PCR) from a wild-type strain
using custom-made primers and recombined into a CEN/ARS
shuttle plasmid upstream of a promoterless GFP construct via
transformation and homologous recombination in a wild-type
yeast strain [11]. The plasmid is then extracted by plasmid
rescue [12] and amplified in E. coli cells, then mini-prepped,
sequenced, and transformed into a query strain that contains an
integrated control promoter RFP reporter (see Notes 2–4).

2.2 SGA 1. Pinning robot or a hand pinning tool that can pin in 384 and
768 colony/plate format (see Note 5).
2. Rectangular media trays for solid media plates by Nunc or
Singer (depending on the pinning robot/tool used).
3. SGA media
616 Hendrikje Göttert et al.

Table 1
Strains and plasmids used when screening the deletion, ts or FLEX array using R-SGA

Plasmid Genotype Source

R
DELETION/TS CEN ARS Amp Promoter of interest-GFP LEU2 Kainth et al. [1]
query plasmid
FLEX array CEN ARS AmpR URA3 GAL1pr-ORF Hu et al. [9]
plasmid
pBY011
FLEX query CEN ARS AmpR Promoter of interest-GFP NatMX Andrews Lab,
plasmid University of
Toronto
Strain Genotype Source
DELETION/TS MATa his3Δ1 leu2Δ0 ura3Δ0 met15Δ0 ORF::KanMX Giaever et al. [3]
array strain
BY4741
DELETION/TS MATα ho:: Control promoter-tdTomato::HphMX can1Δ:: Kainth et al. [1]
query strain STE2pr-S.p.His5 lyp1Δ his3Δ1 leu2Δ0 ura3Δ0 met15Δ0
FLEX query strain MATa ho:: Control promoter -tdTomato::HphMX his3Δ1 Andrews Lab,
leu2Δ0 ura3Δ0 met15Δ0 University of
Toronto
The deletion and the ts array strains are isogenic and thus the same query strains can be crossed to both arrays. The strain
containing the FLEX overexpression collection has a different mating type and carries different selectable markers,
therefore a different query strain needs to be employed. The control promoter of choice is integrated into the genome,
while the promoter of choice to be tested is located on a plasmid.

Deletion/TS array FLEX array

Array plates YPD þ G418 SD uracil
Query culture SD leucine YPD þ clonNat
Mating YPD YPD
Diploid SD leucine þ G418 SD uracil þ clonNat
selection (2)
Sporulation Spo þ ¼ G418
Haploid SD leucine arginine
selection (2) histidine lysine
þ canavanine þ thialysine
þ G418
Final selection/ SD leucine arginine SGal uracil þ clonNat
induction histidine lysine þ Hygromycin B
þ canavanine þ thialysine
þ G418 þ Hygromycin B
YPD: 1% yeast extract, 2% peptone, 2% glucose with 40 mg/L
adenine and 40 mg/L tryptophan.
 Reporter-Based Synthetic Genetic Array Analysis: A Functional Genomics. . . 617

Synthetic dextrose medium (SD) without antibiotics: 0.67% yeast

nitrogen base without amino acids, 2% glucose plus complete
amino acid mix (see below).
SD medium containing G418, clonNat, or Hygromycin B: 0.1%
monosodium glutamic acid, 0.17% yeast nitrogen base without
amino acids and without ammonium sulfate, 2% glucose plus com-
plete amino acid mix (see below) (see Note 6).
SGal and SRaf: same as SD medium but with 2% galactose or
raffinose as carbon source.
Amino acid dropout mixes: To make complete amino acid mix
powder, grind or shake together for 15 min: 2 g each of alanine,
arginine, asparagine, aspartic acid, cysteine, glutamine, glutamic
acid, glycine, histidine, inositol, isoleucine, leucine, lysine, phenyl-
alanine, proline, serine, threonine, tryptophan, uracil, valine plus
3 g adenine and 0.2 g para-aminobenzic acid. To make 10 com-
plete amino acid mix for 1 L of SD, add 2 g of powder to 100 mL
water and heat to dissolve. To make dropout, leave out the relevant
amino acids or bases.
Final antibiotic concentrations: 200 μg/mL G418, 100 μg/mL
clonNat, and 300 μg/mL Hygromycin B.
Final amino acid analog concentrations: 50 μg/mL canavanine and
50 μg/mL thialysine (S-(2-aminoethyl)-L-cysteine hydrochloride).
Sporulation medium: 1% potassium acetate, 0.1% yeast extract,
0.05% glucose, supplemented with 0.014 g/L uracil and histidine
and 0.071 g/L leucine and antibiotic at one quarter standard
concentration to retard the growth of mold.
Media composition is summarized in the table above and is the
same for liquid culture as for solid plates but plates contain 2% agar
(see Note 7).

2.3 Fluorescence 1. Scanning fluorescence imager suitable for GFP and RFP image
Scanning acquisition, e.g. Typhoon Trio Variable Mode Imager (GE
Healthcare), Typhoon 7000 (GE Healthcare), or equivalent.
2. Software for image analysis such as GenePix Pro or SpotFinder.

2.4 Liquid Culture 1. 96-well round bottom plates.

2. Breathable membrane covers, e.g. Breath-Easy by E&K
Scientific.
3. HiGro shakers from Digilab (see Note 8).
4. Hand pinning tool, e.g. pin pad from Singer Instruments.

2.5 Flow Cytometry 1. 384-well flat bottom polystyrene plates.

2. Multiwell plate mixer (Eppendorf MixMate).
3. Multiwell plate sonicator (QSonica MPH).
4. Flow cytometer suitable for GFP and RFP analysis, including
high-throughput sampler and acquisition software, e.g. LSRII
by BD Biosciences which is configured with a 488 nm and 561
nm laser for GFP and RFP analysis.
618 Hendrikje Göttert et al.

3 Methods

3.1 R-SGA with a A schematic diagram of the screening pipeline is depicted in Fig. 2a.
Haploid Deletion Array
1. Prepare query strain for lawns. Set up an overnight culture of a
single colony of the query strain containing the promoter of
interest in 10 mL SD leucine (Leu). Incubate at 30  C for one
night.
2. Make lawns and assemble array in 768 format. Spread 3 mL of
the overnight culture on each of three YPD square agar plates.
Incubate at 30  C for two nights to grow confluent lawns of
cells. Also re-pin the 14 plates of the deletion array, which are
maintained in 384 colonies per plate format, into 768 format
by pinning the same 384 colonies twice onto one plate diago-
nally to each other. Incubate at 30  C for two nights.
3. Mate. Use a pinning robot (BioMatrix robot by S&P Robotics,
Singer Rotor) or other pinning tool in 768 format to pin from
the lawns onto fresh YPD square plates, up to six per lawn,
without washing in between pinnings. Mate strains by pinning
the 14 plates of the freshly grown deletion array in 768 format
on top of the query strain, including a washing and sterilization
step after each pinning action. Incubate at 30  C overnight.
4. Select for diploid cells that carry both the LEU2-marked
promoter-GFP plasmid and a KanMX-marked gene deletion.
Pin from the mating plates onto SD-Leu þ G418 and incubate
at 30  C for two nights.
5. Repeat diploid selection. Pin from diploid selection plates onto
fresh SD-Leu þ G418 and incubate at 30  C for another night.
6. Sporulate. Pin from second diploid selection plates onto sporu-
lation medium and incubate at 25  C (room temperature) for
five nights to allow for sufficient tetrad formation.
7. Select haploid MATa strains carrying the GFP reporter plasmid
and a gene deletion. Pin from sporulation plates onto SD Leu
arginine(Arg) histidine(His) lysine (Lys) þ canava-
nine þ thialysine þ G418 and incubate at 30  C for two nights.
8. Repeat. Pin from first haploid selection plates onto fresh SD
Leu Arg His Lys þ canavanine þ thialysine þ G418 and
incubate at 30  C for another two nights.
9. Select for haploid MATa strains carrying the GFP reporter plasmid,
the integrated control promoter (HphMX-marked) and a gene
deletion. Pin from second haploid selection plates onto SD
Leu Arg His Lys þ canavanine þ thialy-
sine þ G418 þ Hygromycin (Hyg) and incubate at 30  C for
two nights (see Note 9).
 Reporter-Based Synthetic Genetic Array Analysis: A Functional Genomics. . . 619

Fig. 2 The R-SGA pipeline. Reporter-Synthetic Genetic Array analysis (R-SGA) uses multiple pinning and
culture steps on selective plates to introduce two reporter cassettes into an array of yeast mutants. The
pipeline can easily be adapted to screen reporter genes against different yeast arrays and with different
means of assessing output strains. (a) The original protocol was designed to screen reporter genes against the
620 Hendrikje Göttert et al.

3.2 R-SGA with a A schematic diagram of the screening pipeline is depicted in Fig. 2b.
Haploid Array Carrying
1. Prepare query strain for lawn and assemble array in 768 format.
Temperature-Sensitive
Set up an overnight culture of a single colony of the query
Alleles of Essential strain containing the promoter of interest in 10 mL SD Leu.
Genes Incubate at 30 C for one night. Re-pin the four plates of the ts
array, which are maintained in 384 colonies per plate format,
into 768 format by pinning the same 384 colonies twice onto
one plate diagonally to each other. Incubate at 25 C for three
nights.
2. Make lawn. Spread 3 mL of the overnight culture on a YPD
square agar plate. Incubate at 30 C for two nights to grow a
confluent lawn of cells.
3. Mate. Use a pinning robot in 768 format to pin from the lawn
onto fresh YPD square plates, four per lawn, without washing
in between pinnings. Then pin the four plates of the freshly
grown ts array in 768 format on top of the pinned query strain,
including a washing step after each pinning. Incubate at 25 C
for one night to allow query and array strains to mate.
4. Select for diploid cells that carry both the LEU2-marked
promoter-GFP plasmid and a KanMX-marked ts allele. Pin
from the mating plates onto SD Leu þ G418 and incubate
at 25 C for two nights.
5. Repeat diploid selection. Pin from first diploid selection plates
onto fresh SD-Leu þ G418 and incubate at 25 C for two
nights.
6. Sporulate. Pin from second diploid selection plates onto sporu-
lation medium and incubate at 25 C (room temperature) for
five nights to allow for sufficient tetrad formation.
7. Select for haploid MATa strains carrying the GFP reporter plas-
mid and a ts allele. Pin from sporulation plates onto SD Leu
Arg His Lys þ canavanine þ thialysine þ G418 and
incubate at 25 C for three nights.
8. Repeat. Pin from first haploid selection plates onto fresh SD
Leu Arg His Lys þ canavanine þ thialysine þ G418 and
incubate at 25 C for three nights.

Fig. 2 (continued) yeast KanMX deletion array. (b) Adaptation for use with the temperature sensitive (ts)
collection involves growth at permissive temperature (25 C) and incubation at semipermissive temperature
(30 C) prior to scanning. (c) A shorter pipeline applies when using an overexpression collection as it can be
assessed in diploid strains. Prior to scanning, overexpression is induced by growth on galactose-containing
medium. (d) To assess output strains by flow cytometry instead of colony scanning, cells are cultured in liquid
medium, subcultured and placed at higher temperature or shifted to galactose-containing medium (as
necessary for ts or overexpression) prior to scanning
 Reporter-Based Synthetic Genetic Array Analysis: A Functional Genomics. . . 621

9. Select for haploid MATa strains carrying the GFP reporter plas-
mid, the integrated control promoter (HphMX-marked) and at
the same time induce loss/reduction of function of ts alleles. Pin
from second haploid selection plates onto SD Leu Arg
His Lys þ canavanine þ thialysine þ G418 þ Hyg and
incubate at semipermissive temperature 30  C for two nights
(see Note 9).

3.3 R-SGA with a A schematic diagram of the screening pipeline is depicted in Fig. 2c.
Galactose-Inducible
1. Prepare query strain for lawns. Set up an overnight culture of a
Overexpression Array
single colony of the query strain containing the promoter of
interest in 10 mL YPD þ clonNat. Incubate at 30 C for one
night.
2. Make lawns and assemble array in 768 format. Spread 3 mL of
the overnight culture on each of three YPD square agar plates.
Incubate at 30 C for two nights to grow confluent lawns of
cells. Also re-pin the 18 plates of the FLEX array, which are
maintained in 384 colonies per plate format, into 768 format
by pinning the same 384 colonies twice onto one plate diago-
nally to each other. Incubate at 30 C for two nights as well.
3. Mate. Use a pinning robot in 768 format to pin from the lawns
onto fresh YPD square plates, six per lawn, without washing in
between pinnings. Then pin the 18 plates of the freshly grown
FLEX array in 768 format on top of the pinned query strain,
including a washing step after each pinning. Incubate at 30 C
for one night.
4. Select for diploid cells which carry both the NatMX-marked
promoter-GFP plasmid and the URA3-marked FLEX overex-
pression plasmid. Pin from the mating plates onto SD
Ura þ clonNat and incubate at 30 C for two nights.
5. Repeat diploid selection and simultaneously induce overexpres-
sion by growth on galactose-containing medium. Pin from dip-
loid selection plates onto synthetic galactose medium (SGal)
Ura þ clonNat þ Hyg and incubate at 30 C for two nights
(see Notes 10 and 11).

3.4 Analyzing Colony 1. Remove plates from incubator and cool to room temperature
Fluorescence by for at least 2 h before scanning to prevent condensation on the
Typhoon Scanning scanner surface (see Note 12).
2. Place plates upside down, without lids onto the cleaned scan-
ning surface and tape edges to the Typhoon platen with clear
tape to prevent movement due to vibration of the instrument
(see Notes 13 and 14).
3. Use the Typhoon Scanner Control version 5.0 software with
the following settings: acquisition mode ¼ fluorescence, laser
622 Hendrikje Göttert et al.

Fig. 3 Colony fluorescence detection using a Typhoon scanner. A plate with 768 yeast mutant colonies which
carry a GFP and an RFP reporter were scanned using a Typhoon scanner. The GFP and the RFP channel are
shown separately and as overlay. Two colonies of the same mutant strain display decreased GFP expression.
The same colonies show unchanged expression levels of RFP and thus appear red in the overlay picture

1: 488 nm and emission filter 520 BP40, laser 2: 532 nm and

emission filter: 580 BP30, pixel size: 100 μm, focal plane:
þ3 mm. Photomultiplier tube (PMT) voltages are chosen
empirically so that GFP and RFP intensities are below satura-
tion with signals in a roughly similar range (see Note 15). An
example of a scanned plate is displayed in Fig. 3.

3.5 Quantifying 1. Open both the GFP and RFP.gel Typhoon files in GenePix Pro
Colony Fluorescence version 6.0 software.
with GenePix Software 2. Assign appropriate wavelengths (488 nm for GFP and 532 nm
for RFP).
3. To draw a template of circles for each colony on the plate,
create “New Blocks” with the following features for 768 colo-
nies/plate and replicate for all plates in the picture:
Number of columns: 24
Column spacing (μm): 462
Number of rows: 32
Row spacing (μm): 225
Feature diameter (μm): 180
Feature Layout: Orange packing #3
4. Position grid by dragging it over colonies in image, aligning
circles approximately with colonies. Click “Align Features in all
Blocks” to automatically adjust circle size to actual colony size
(see Note 16).
5. Click “Analyse” and find the results in the Results tab. Save
results.
 Reporter-Based Synthetic Genetic Array Analysis: A Functional Genomics. . . 623

6. Use the median colony fluorescence intensity corrected for

local background fluorescence for further normalization and
analysis.

3.6 Quantifying 1. Upload in parallel .gel files of the GFP and RFP channel
Colony Fluorescence (ignore error messages).
Using SpotFinder 2. To determine the location of colonies on each plate, a prede-
Software (See Note 17) fined grid is applied to the image and aligned manually for best
fit. The two biological replicates present on the 768-colony R-
SGA plate can be analyzed separately by applying a 384-spot
grid.
For a typical R-SGA plate, grid parameters for one 384-
colony replicate are:
Row number: 16
Column number: 24
X spacing, pix: 45
Y spacing, pix: 45
Meta row, Meta column, Pin X, and Pin Y values are adjusted
depending on the number of plates present on the image and
their positioning.
3. In the Gridding and Processing dialog, enter the Segmentation
Method settings. In our hands, the Otsu Segmentation
method produces good results. Min spot size and max spot
size depend on the amount of colony growth and typically
range between 5 and 40 pix. After adjusting the settings, the
button “Process All” launches the segmentation and proces-
sing loop.
4. On the Data page, the user can review the segmentation out-
put. The displayed data include information on the shape of the
detected colonies and extracted numerical values.
5. After processing, save the data to a MEV file. Use the median
colony fluorescence intensity corrected for local background
fluorescence for further normalization and analysis.
6. Quality control filters (QC, QCA, and QCB) can be used to
discard empty and badly processed colonies. Typically, a good
colony will have a QC value of 0.4 or higher.
(See Note 18.)

3.7 Normalization 1. Assign GFP and RFP intensities obtained from GenePix or
and Analysis of Solid SpotFinder, as described above, to their proper gene or allele
Plate Data name according to their plate position, obtaining two values for
each position when the screen was carried out in 768 format.
624 Hendrikje Göttert et al.

2. Remove border colonies from further analysis as these repre-

sent control strains to maintain similar colony sizes across the
plate. Also remove negative intensity values indicating empty
spots (see Note 19).
3. Divide each GFP value by the median GFP intensity of its plate
to correct for plate specific variations (e.g., caused by different
medium batches/thickness). Do the same for RFP.
4. Calculate log2(GFP/RFP) ratios to correct for colony effects
(e.g., size).
5. Subtract each value by the median log2(GFP/RFP) ratio to
center the whole screen at 0.
6. Average duplicate values for the two colonies next to each
other on a plate.
7. Calculate z-scores by subtracting the screen median and divid-
ing by the screen’s median absolute deviation (MAD) from
each averaged value (see Notes 20 and 21).

3.8 Liquid Culture To assess the final output array by flow cytometry, transfer the
and Assessment by arrayed cells into liquid medium using a hand pinning tool. Do
Flow Cytometry this before induction in case of overexpression and before shift to
restrictive temperature in case of ts mutants. Proceed as follows (see
Fig. 2d):

Day 1
1. Split each 768-format R-SGA output plate into two 384-
format plates using a pinning robot to facilitate manual pinning
into liquid medium.

Day 2 Evening
2. Transfer cells from freshly grown colonies on a solid plate into a
96-well round bottom plate with liquid medium, four per 384-
format plate. Use a medium composition of SD His Leu
Arg Lys þ canavanine þ thialysine þ G418 for deletion and
ts strains and synthetic raffinose medium (SR) Ura þ clonNat
for overexpression strains, 200 μL per well.
3. Cover plates with sterile breathable seals and incubate at
200 rpm overnight at 30  C for deletion and overexpression
strains and at 25 C for ts strains.

Day 3 Morning
4. Subculture plates by transferring 5 μL into 200 μL fresh
medium. Use the same medium as in the first step for deletion
and ts strains. Induce overexpression strains in SGal
Ura þ clonNat.
 Reporter-Based Synthetic Genetic Array Analysis: A Functional Genomics. . . 625

5. Culture all strains shaking at 30 C (semipermissive tempera-

ture for ts strains) for 5 h.

Day 3 Afternoon
6. Shake the 96-well plates well using a plate mixer (Eppendorf
MixMate or vortex).
Use a liquid handling robot to combine the four 96-well plates
into one 384 flat bottom plate.
7. Sonicate 384 plates using a plate sonicator. We use 50% power
for 60 seconds on a Qsonica MPH (This step should be empir-
ically optimized by examining cells under a microscope to
ensure that mother-daughter pairs have been separated).
8. Collect events from the 384 format plate using a flow cyt-
ometer equipped with an HTS by introducing 10 μL of culture
per strain into the cytometer and collecting data for up to
50,000 cells per well:
l Collect GFP, forward scatter (FSC), and side scatter (SSC)
parameters from a 50 mW 488 nm laser through 510/
20 nm and a pair of 488/10 nm filters, respectively.
l Collect tdTomato levels from a 50 mW 561 nm laser
through a 610/20 nm bandpass filter.
l Collect both pulse area and width parameters for scatter
parameters; collect pulse area only for fluorescence
parameters.

3.9 Normalization 1. Open ungated cell population as forward scatter area (FSC-A,
and Analysis of Flow roughly indicative of cell size) vs. side scatter area (SSC-A,
Cytometry Data correlated with yeast size and budding status) scatterplot.
Gate out debris and aggregates to select yeast-like events
(Fig. 4a).
2. Open this gated cell population as an FSC-width (FSC-W) vs.
FSC-A scatterplot. Gate out doublet events that are visible as a
second mode on the FSC-W axis. These represent physically
attached cells or cells that traversed the laser intercept nearly
simultaneously (Fig. 4b).
3. Open gated cell population as SSC-W vs. SSC-A scatterplot. As
in (step 2) above, gate out doublet events (Fig. 4c).
4. Open gated cell population as GFP-area (FITC-A) vs. RFP-
area (tdTomato-A) scatterplot. Gate out events that do not
express fluors properly (Fig. 4d) (see Note 21).
5. For the remaining events of interest (Fig. 4e) generate the
median intensity for GFP and the median intensity for RFP
for further analysis.
6. Remove wells with fewer than 2000 cells from further analysis.
7. Calculate log2(GFP/RFP) ratios to correct for strain-specific
effects (e.g. cell size).
626 Hendrikje Göttert et al.

a Yeast Gate
b 1st Singles Gate

(x1000)
200
(x1000)
100 200

FSC-W
50 100
SSC-A
0

FSC-A (x1000) FSC-A (x1000)

c 2nd Singles Gate
d Fluor Gate
(x1000)

tdTomato-A (x1000)
200

20 40
50 100
SSC-W

0
SSC-A (x1000) FITC-A (x1000)
e Final Events of Interest Final Events of Interest - Fluors
tdTomato-A (x1000)
(x1000)

60
200

40
100
SSC-A

20
0

FSC-A (x1000) FITC-A (x1000)

Fig. 4 Gating of flow cytometry data after acquisition. Gating allows for the exclusion of undesired data points
or events prior to analysis. (a) Forward scatter area (FSC-A) is related to cell size. Side scatter area (SSC-A)
reflects size and bud status. Excluding the extremes of FSC-A and SSC-A help to remove undesired events,
e.g., debris, bacterial contamination, dead cells, or aggregates. (b) A gate using forward scatter width (FSC-W)
excludes cases where two or more cells are stuck together or traverse the laser intercept simultaneously. (c) A
gate using side scatter width (FSC-W) excludes additional cases where multiple cells were scanned together.
(d) A gate excluding events with negative GFP (FITC-A) or tdTomato values removes cells which do not express
the fluors properly. (e) Final events of interest after gating are used for further analysis

8. Determine the mean and standard deviation of all wild-type

strain log2(GFP/RFP) ratios.
9. We define two standard deviations above the mean wild-type
log2(GFP/RFP) as the cut-off to identify genes whose pertur-
bation significantly upregulates expression of the tested pro-
mote and two standard deviations below the mean wild-type
log2(GFP/RFP) to identify genes whose perturbation signifi-
cantly downregulates expression of the tested promoter (see
Notes 22 and 23).

4 Notes

1. Colonies at the border of the plate grow faster than colonies

inside the array due to decreased competition from neighbors.
Reporter-Based Synthetic Genetic Array Analysis: A Functional Genomics. . . 627

Arrays should be built with a border (e.g. rows A and P,

columns 1 and 24) consisting of an isogenic control strain
that will be carried through all SGA steps. These colonies are
excluded from analysis after data acquisition.
2. The two promoters should not be related. For example, if the
goal is to explore cell cycle-regulated transcription, the pro-
moter of interest would be cell cycle-regulated and the control
promoter should not be cell cycle-regulated but rather consti-
tutively expressed.
3. Having the GFP reporter on a plasmid facilitates query strain
construction, especially when a large number of different pro-
moters are to be screened. As the plasmid used here is a CEN
plasmid, copy number variation is of negligible relevance when
screening genome-wide. However, some mutations may cause
increased plasmid loss, confounding results. To avoid this, the
GFP reporter could be integrated into the genome.
4. The method can easily be adapted to assay readouts other than
promoter activation.
5. Protocols are optimized for arrays pinned and grown in 768
colony format. Other formats can be used; however, incuba-
tion times need to be adapted as too small, too large or too
variable colony sizes can make data normalization and analysis
difficult. We have found, for example, that the colony size with
1534-colony format can be too small for accurate
quantification.
6. Ammonium sulfate interferes with the sensitivity of cells to
antibiotics [13]. We therefore use MSG as the nitrogen source
in SD plates containing G418, clonNat, or hygromycin.
7. A plate pouring robot such as the Serial Filler from Singer, a
calibrated pump or simply a graduated cylinder should be used
to pour final selection plates for Typhoon scanning to ensure
optimal and equal thickness of medium. Sixty milliliters is ideal
when using rectangular plates from Nunc.
8. Other shakers can be used to culture cells in microwell plates
but growth efficiencies might vary. Glass beads can be added to
improve mixing and aeration but the additional stress on the
cells has to be considered. We have found more dead cells are
produced when glass beads are used.
9. An additional step could be added here, where colonies are
subjected to different conditions, e.g. drug treatment.
10. Overexpression can be assessed in diploid cells as described
here, saving time and materials. However, for certain studies
it may be preferable to assess effects in haploid cells, which
could be achieved by modifying the SGA pipeline.
628 Hendrikje Göttert et al.

11. Maximal induction of overexpression is achieved in <1 day;

however, 2 days of growth are necessary to obtain colonies big
enough to easily scan and analyze.
12. Preventing condensation is essential as imaging artifacts from
moisture considerably affect results.
13. The Typhoon scanner can scan 12 plates at a time, thus scan-
ning is done in two steps for large arrays such as the deletion
and FLEX arrays. When scanning fewer than 12 plates, arrange
them starting on the vertical (fast) axis on the left side of the
scanner surface to minimize scanning time.
14. Remember that the scanner views the plates from below, thus the
output picture will be inverted. When scanning multiple plates, it
makes sense to place plate 1 in the top left corner and the last
plate in the bottom right. Then choose the vertically mirrored R
in the software to correctly display the output picture.
15. Fluorescence varies by strength of promoter, colony sizes, and
even slight differences in media thickness; PMT voltages thus
have to be optimized for each individual screen. To eliminate
potential artefacts caused by photobleaching of the reporter
proteins during the initial scan PMT should be determined on
a “throwaway” plate.
16. Make sure all existing colonies are detected. If blocks and plates
are poorly aligned, whole plates can be missed. Empty or
undetected spots are indicated by Ø.
17. SpotFinder is an open-source multichannel image analysis tool
which is part of the TM4 microarray software suite [14]. In our
hands SpotFinder produces results comparable to GenePix.
18. All parameters and values are experiment-dependent and need
to be adjusted accordingly for best performance. Detailed user
instructions for grid positioning and other components of
SpotFinder can be found in [15] and in the online manual
(http://www.tm4.org/spotfinder.html).
19. In parallel colony sizes could be measured by photographing
plates and running the pictures through an image analysis
software like SGA Tools [16]. Very small colonies can then be
eliminated from further analysis assuming that they give unre-
liable results.
20. For whole array screens, p-values can be calculated from z-
scores and a statistical cut-off such as a 10% false discovery
rate (FDR) can be used to determine hits. For small scale
assays, for which the distribution is likely not normal, log2
ratios should be compared between mutants and wild-type
colonies on the same plate (not border colonies).
21. In any population of cells expressing fluorescent proteins, we
find a small fraction of cells that have no GFP or RFP fluores-
cent signal.
 Reporter-Based Synthetic Genetic Array Analysis: A Functional Genomics. . . 629

22. We recommend confirming hits identified in a large scale

screen by retesting them on mini-arrays using a strain carrying
a different control promoter and comparing log2 ratios
between mutants and wild-type colonies on the same plate.
We observe confirmation rates ranging from 40% to 70%
depending on the promoter and yeast array used.
23. Assessment by flow cytometry can be used as a small scale
orthogonal confirmation assay for hits identified on a large
scale with the solid plate Typhoon method. We have found
the two methods to agree with Pearson Correlation Coeffi-
cients around R ¼ 0.7.

References
1. Kainth P, Sassi HE, Pena-Castillo L, Chua G, 9. Hu Y, Rolfs A, Bhullar B, Murthy TV, Zhu C,
Hughes TR, Andrews BJ (2009) Comprehen- Berger MF et al (2007) Approaching a com-
sive genetic analysis of transcription factor plete repository of sequence-verified protein-
pathways using a dual reporter gene system in encoding clones for Saccharomyces cerevisiae.
budding yeast. Methods 48:258–264 Genome Res 17:536–543
2. Tong AH, Boone C (2006) Synthetic genetic 10. Douglas AC, Smith AM, Sharifpoor S, Yan Z,
array analysis in Saccharomyces cerevisiae. Durbic T, Heisler LE et al (2012) Functional
Methods Mol Biol 313:171–192 analysis with a barcoder yeast gene overexpres-
3. Giaever G, Chu AM, Ni L, Connelly C, Riles L, sion system. G3 2:1279–1289
Vronneau S et al (2002) Functional profiling of 11. van Leeuwen J, Andrews BJ, Boone C, Tan G
the Saccharomyces cerevisiae genome. Nature (2015) Construction of multifragment plas-
418:387–391 mids by homologous recombination in yeast.
4. Fillingham J, Kainth P, Lambert JP, van Bakel Cold Spring Harb Protoc 9:pdb.top084111.
H, Tsui K, Pena-Castillo L et al (2009) Two- prot085100
color cell array screen reveals interdependent 12. Robzyk K, Kassir Y (1992) A simple and highly
roles for histone chaperones and a chromatin efficient procedure for rescuing autonomous
boundary regulator in histone gene repression. plasmids from yeast. Nucleic Acids Res
Mol Cell 35:340–351 20:3790
5. Kurat CF, Lambert JP, van Dyk D, Tsui K, van 13. Cheng T-H, Chang C-R, Joy P, Yablok S, Gar-
Bakel H, Kaluarachchi S et al (2011) Restric- tenberg MR (2000) Controlling gene expres-
tion of histone gene transcription to S phase by sion in yeast by inducible site-specific
phosphorylation of a chromatin boundary pro- recombination. Nucleic Acids Res 28:e108
tein. Genes Dev 25:2489–2501 14. Saeed AI, Sharov V, White J, Li J, Liang W,
6. Hendry JA, Tan G, Ou J, Boone C, Brown GW Bhagabati N et al (2003) TM4: a free, open-
(2015) Leveraging DNA damage response sig- source system for microarray data management
naling to identify yeast genes controlling and analysis. Biotechniques 34:374–378
genome stability. G3 5:997–1006 15. Saeed AI, Bhagabati NK, Braisted JC, Liang W,
7. Li Z, Vizeacoumar FJ, Bahr S, Warringer J, Li Sharov V, Howe EA et al (2006) TM4 micro-
J, Vizeacoumar FS et al (2011) Systematic array software suite. Methods Enzymol
exploration of essential yeast gene function 411:134–193
with temperature-sensitive mutants. Nat Bio- 16. Wagih O, Usaj M, Baryshnikova A, VanderSluis
technol 29:361–367 B, Kuzmin E, Costanzo M et al (2013) SGA-
8. Ben-Aroya S, Coombes C, Kwok T, O’Donnell tools: one-stop analysis and visualization of
KA, Boeke JD, Hieter P (2008) Toward a com- array-based genetic interaction screens. Nucleic
prehensive temperature-sensitive mutant Acids Res 41:W591–W596
repository of the essential genes of Saccharo-
myces cerevisiae. Mol Cell 30:248–258
 Chapter 41

Statistical Analysis and Quality Assessment of ChIP-seq

Data with DROMPA
Ryuichiro Nakato and Katsuhiko Shirahige

Abstract
Chromatin immunoprecipitation followed by sequencing (ChIP-seq) analysis can detect protein/DNA-
binding and histone-modification sites across an entire genome. As there are various factors during sample
preparation that affect the obtained results, multilateral quality assessments are essential. Here, we describe
a step-by-step protocol using DROMPA, a program for user-friendly ChIP-seq pipelining. DROMPA can
be used for quality assessment, data normalization, visualization, peak calling, and multiple statistical
analyses.

Key words Chromatin immunoprecipitation, ChIP-seq, High-throughput sequencing, Quality man-

agement, Visualization, Statistical analysis, Normalization

1 Introduction

Chromatin immunoprecipitation followed by high-throughput

sequencing (ChIP-seq) analysis was developed to identify
protein/DNA-binding and histone-modification sites across an
entire genome [1]. This assay is a mainstream method in genomics
and epigenomics, and has led to important discoveries for various
species [2–4]. A computational ChIP-seq analysis can be divided
into several steps: (1) Mapping sequenced reads onto a reference
genome; (2) Making data bins with normalized reads filtered for
PCR biases; (3) Identifying significantly enriched regions (peaks);
(4) Statistical analysis based on the obtained peak lists (e.g., aggre-
gation plot). In this chapter, we describe a step-by-step protocol
using DROMPA (DRaw and Observe Multiple enrichment Profiles
and Annotation), a program for user-friendly ChIP-seq pipelining
[5]. The main features of DROMPA are:
l Any species whose genomic sequence is available can be used;
l Multiple input/output file formats (SAM, BAM, Bowtie, WIG,
BED, TagAlign(.gz), bigWig, and bedGraph) are available;

Marco Muzi-Falconi and Grant W. Brown (eds.), Genome Instability: Methods and Protocols, Methods in Molecular Biology,
vol. 1672, DOI 10.1007/978-1-4939-7306-4_41, © Springer Science+Business Media LLC 2018

631
632 Ryuichiro Nakato and Katsuhiko Shirahige

Fig. 1 Workflow of DROMPA pipeline. Red box indicates three internal programs of DROMPA and blue one does
the reference annotations. The outputs of DROMPA include peak lists, pdf figures, and bin data of normalized
read profiles

l Normalization using mappability and GC content biases that

arise in ChIP-seq data;
l Output (in PDF or PNG format) of the read distribution,
ChIP/input enrichment and p-values;
l In addition to typical peak calling, various types of ChIP-seq
analysis are available.
Figure 1 shows the workflow of DROMPA (see Notes 1 in
Section 4 for degtails on annotation data). From version 3.0.0
onward, DROMPA has three internal programs: parse2wig, drom-
pa_peakcall, and drompa_draw. parse2wig preprocesses an input
map file into bin data (the number of mapped reads per bin with
fixed length). Generated bin data are used as input for both drom-
pa_peakcall and drompa_draw. drompa_peakcall calls peaks (peak
calling) and drompa_draw executes various types of visualization
and quantitative analyses. DROMPA is available as an open-source
C package with a detailed tutorial at https://github.com/rna-
kato/DROMPA3.

2 Materials

The examples are presented here as command lines to be used in

UNIX shell prompts (e.g., bash). Commands are prefixed with “$”.
Comments are prefixed with “#” and can be ignored. Examples
provided herein uses DROMPA version 3.2.3. Check the website
for the latest version.
 ChIP-seq Analysis with DROMPA 633

2.1 Installation 1. Installation of DROMPA requires several libraries and pro-

grams. On Ubuntu OS, these programs can be installed using
the following commands:

$ sudo apt-get install git gcc libgtk2.0-dev libgsl-dev

samtools r-base
# download cpdf
$ wget http://github.com/coherentgraphics/cpdf-binaries/ar-
chive/master.zip
$ unzip master.zip

2. Install DROMPA through git:

$ git clone https://github.com/rnakato/DROMPA3.git

$ cd DROMPA3
$ make

If you get an installation error, make sure that all required libraries
are installed.
3. Next, add the software directory to your PATH environment
variable. For example, if you downloaded DROMPA and cpdf
into the $HOME/my_chipseq_exp directory, type:

$ export PATH ¼ $PATH:$HOME/my_chipseq_exp/DROMPA3

$ export PATH ¼ $PATH:$HOME/my_chipseq_exp/cpdf-binaries-mas-
ter/Linux-Intel-**bit

2.2 Required Data: DROMPA requires a genome table file, a tab-delimited file describ-
Genome Table File ing the name and length of each chromosome. Genome-table files
can be generated by makegenometable.pl in the “scripts” directory
as follows:

$ makegenometable.pl genome.fa > genometable.txt

Chromosome names in the genome table file and the reference

genome should be identical. Hereafter, “genometable.txt” indi-
cates this genome table file.

2.3 Read Mapping ChIP-seq analysis with DROMPA starts with map files. For read
mapping, we adopted the program Bowtie [6], which can distin-
guish unique and multiple mapped reads. To map reads (ChIP.
fastq) onto the human genome build hg38 allowing uniquely
mapped reads only, type:

$ bowtie index-UCSChg38 ChIP.fastq -m1 -S > ChIP.sam

634 Ryuichiro Nakato and Katsuhiko Shirahige

To include multiple mapped reads, remove “-m1” option. See

the manual for more information (http://bowtie-bio.sourceforge.
net/manual.shtml). If the users want to treat insertions and dele-
tions in reads, consider using other mapping tools such as Bowtie2
[7] and BWA [8].

3 Methods

3.1 Read parse2wig preprocesses an input map file into bin data (the number
Normalization and Bin of mapped read per bin). The fragment length of each read is
Data Generation calculated automatically. For single-end mode (default), mapped
reads are extended to the expected DNA-fragment length. In
paired-end mode (with “-pair” option), each fragment length is
obtained from the map file, and inter-chromosomal read-pairs and
read-pairs longer than the maximum fragment length (specified by
the “-maxins” option) are ignored.
The command:

$ parse2wig –f SAM -i ChIP.sam -o ChIP -gt genometable.txt -

binsize 100 –n GR -mp mappability/map_fragL150

creates a new directory “parse2wigdir”, and outputs bin files

generated from the map file (ChIP.sam) with a bin size of 100 bp
into it. A bin file is outputted for each chromosome (in the case of
binary and WIG outputs) or for the whole-genome (in the case of
bedGraph and bigWig outputs). In default, parse2wig filters
“redundant reads” (reads starting exactly at the same 50 ends) as
“PCR bias [1]”. When studying highly repetitive regions (e.g.,
rDNA regions in Saccharomyces cerevisiae), this filtering step should
be omitted by supplying “-nofilter” option.
Read normalization is required for comparative ChIP-seq ana-
lyses that evaluate peak similarities and differences among samples
[1]. Using “–n GR” option, parse2wig scales bin data so that the
total number of mapped reads onto the whole genome is 20 million
(by default, see Notes 2 for a point for read normalization). “-mp”
option specifies the mappability files for normalizing each bin data
based on the mappability. When “-mp” is not supplied, all bases are
considered as mappable. These normalized bin data can also be
visualized with other visualization tools such as IGV (http://soft-
ware.broadinstitute.org/software/igv/) and UCSC genome
browser (https://genome.ucsc.edu/).

3.2 Peak Calling drompa_peakcall identifies peaks by assuming a negative-binomial

distribution as a background model and a binomial distribution for
comparison between ChIP and input samples. Both ChIP and
input samples should be specified with a single “-i” option using a
comma as follows:
 ChIP-seq Analysis with DROMPA 635

$ drompa_peakcall PC_SHARP -i parse2wigdir/ChIP,parse2wigdir/

Input -p ChIPpeak -gt genometable.txt

Since the bin files are chromosome-separated, use the prefix

(before the underscore) to specify a sample file as an input file. The
peak files “ChIPpeak.xls” and “ChIPpeak.bed” are outputted in a
tab-delimited text file, which describe same peak sets in different
formats. By default, chromosomes Y and mitochondria are ignored
from peak calling. When mappability file is supplied (“-mp”
option), the low mappable bins (<0.3 as default) are ignored.
See Notes 3 for making bin data by drompa_peakcall.

3.3 Quality In addition to the bin files, parse2wig also outputs the statistics of
Assessment the input file into the output directory (the command in Subhead-
ing 3.1 produce the statistics file “ChIP.100.xls”). The statistics file
describes various quality scores (e.g., read number, fragment length
distribution, library complexity, GC contents, and FRiP score) [1].
The important issues are the number of “nonredundant reads” and
“library complexity”. A low library complexity often occurs when a
small amount of starting material is PCR over-amplified, resulting
in a large amount of unreliable peaks. In other cases, if there are too
few peaks obtained despite good quality scores, the chromatin
immunoprecipitation (ChIP) step might not have worked. Check
the antibody quality and the amount of ChIPed proteins.

3.4 Visualization of 1. drompa_draw can execute various types of visualization including

Read Profile taking multiple ChIP-input pairs as input. Each pair should be
specified with the option “-i”. For example, the command

$ s1¼"-i parse2wigdir/Pol2_b,parse2wigdir/Control,Pol2_b"
$ s2¼"-i parse2wigdir/H3K4me3,parse2wigdir/Control,H3K4me3"
$ s3¼"-i parse2wigdir/H3K27me3,parse2wigdir/Control,H3K27me3"
$ s4¼"-i parse2wigdir/H3K36me3,parse2wigdir/Control,H3K36me3"
$ drompa_draw PC_SHARP -gt genometable.txt -gene refFlat.txt
$s1 $s2 $s3 $s4 -p HeLaS3 -ls 1000 -show_itag 2

generates the PDF files “HeLaS3.pdf” as shown in Fig. 2a (see

Notes 4 for full details on "-i" option). “-ls 1000” indicates
the length (kbp) per one line. By default, drompa_draw visua-
lizes ChIP-read lines only. The “-show itag 2” option displays
the input-read line at the bottom. Similarly, the users can display
the lines of p-value (“-showpinter” and “-showpenrich” options)
and ChIP/input enrichment line (“-showratio” option).
2. The GV mode visualizes a chromosome-wide overview (Fig. 2b)
as follows:

$ drompa_draw GV -gt genometable.txt $s1 $s2 $s3 $s4 -p HeLaS3-

GV -GC GCcontents -gcsize 500000 -GD genedensity -gdsize
500000
636 Ryuichiro Nakato and Katsuhiko Shirahige

Fig. 2 Visualization of ChIP-seq data for human HeLa-S3 cells. (a) PC_SHARP mode: The normalized read
distribution for four ChIP samples and one control sample for 100-bp bins, with a RefSeq gene annotation
(chromosome 1, 35.0–36.0 Mb). Significantly enriched regions (peaks) are highlighted in red. (b) GV mode:
Visualization of the ChIP/Control enrichment distribution for 100-kb bins (chromosome 2). The GC contents
and gene numbers for 500 kb windows are also plotted

where “-GC GCcontents” and “-GD genedensity” specify the GC

content files and the gene-density files, respectively. “-gcsize
500000” and “-gdsize 500000” specify the window sizes of
them. The GV mode does not perform the significance test but
simply highlights the bins containing ChIP/Input enrichments
above the middle of the y axis (the value specified with the
 ChIP-seq Analysis with DROMPA 637

Fig. 3 Schematic representation of replication profile for S. cerevisiae (chromosome 1). Bins containing ChIP/
Input enrichments above the threshold (defined by “-ethre” option) are highlighted in red

option “-scale ratio”) in orange. The bin size is 100k-bp and

the ChIP/Input enrichment lines are shown by default (Fig. 2b
upper). To make y-axis log scale, specify “-showratio 2” as
follows:

$ drompa_draw GV -gt genometable.txt $s1 $s2 $s3 $s4 -p HeLaS3-

GV -GC GCcontents -gcsize 500000 -GD genedensity -gdsize
500000 -showratio 2

See Appendix 1 for the full commands to generate Fig. 2.

3. For a small genome, such as the yeast’s, the sequencing depth is

generally enough (>10-fold). In such cases, the genome-wide
ChIP/Input enrichment distribution is informative, which
minimizes the technical and biological bias in high throughput
sequencing. To make a PDF file of the enrichment distribution
for S. cerevisiae (Fig. 3), type:

$ drompa_draw PC_ENRICH -p SRP009385 -gt genometable.txt -ars

ARS-oriDB_scer.txt -gene SGD_features.tab -gftype 3 $s1 $s2
$s3 $s4 $s5 $s6 $s7 $s8 -ls 250 -lpp 2 -scale_ratio 1 -bn 3 -
ystep 14 -ethre 1.5

where “-lpp 2” specifies the number of lines for each page, and “-bn
3” and “-ystep 14” indicate the number and the height of
638 Ryuichiro Nakato and Katsuhiko Shirahige

separations for y-axis. “-ethre 1.5” specifies the threshold for

enrichment (red regions). When “-nosig” option is specified,
all regions are colored in orange. See Appendix 2 for the full
commands to generate Fig. 3.

3.5 Other Types of drompa_draw executes various types of visualization and quantita-
Visualization tive analyses.
1. For example, the command PROFILE:

$ drompa_draw PROFILE -p aroundBED -gt genometable.txt -i

$IP,,ChIPname -i $IP2,,ChIPname2 -ptype 4 -bed peaklist.bed

outputs a PDF file (aroundgene.pdf) and an R script (aroundgene.

R) of an aggregation plot around the specified bed regions
(Fig. 4a).

2. The CI mode is for quantitative comparison of ChIP-seq

samples.

$ drompa_draw CI -p CIoutput -gt genometable.txt -bed sample.

bed -i $IP,,ChIPname -i $IP2,,ChIPname2

This mode takes two ChIP samples to be compared, and the output
file contains the accumulated read number, the average log2
read density (A), the log2 ratio of read density between two
samples (M) and the significance (log10(p)) of the difference
based on a binomial test, for each peak specified by “-bed”.
This output file can be used for making a MA plot of over-
lapped peak regions (Fig. 4b).
3. The HEATMAP mode outputs a heatmap of ChIP reads or
ChIP/input enrichment profiles around target sites. The
command:

$ drompa_draw HEATMAP -p heatmap -gt genometable.txt $s1 $s2

$s3 -stype 1 -ptype 4 -hmsort 2 -scale_ratio 5 -bed site1.bed,
region1 -bed site2.bed,region2 -png

generates the heatmap of ChIP/input enrichment (-stype 1)

around the sites (-ptype 4) in two peak lists specified “-bed”
option (Fig. 4c). The option “-hmsort 2” sorts the order of
sites (rows) using the second ChIP sample. When specifying “-
hmsort 0” (default), the sites are not sorted. With “-png”
option, drompa_draw outputs the figure in PNG format.
Fig. 4 (a) PROFILE mode: Aggregation plot around peak regions. Shaded regions indicate a 95% confidence
interval. (b) MA plot between M and A columns in the output file of CI mode. This MA plot is generated by plot
function of R. The peaks in which log10(p) is over 10 are colored in red. (c) HEATMAP mode: Heatmap for
ChIP/input enrichment. Sample2 was used for sorting input sites. Each BED file is sorted individually
640 Ryuichiro Nakato and Katsuhiko Shirahige

4 Notes

1. DROMPA accepts the following annotation data: Gene annota-

tion (refFlat, gtf, and SGD_features.tab format); Replication
origin (for yeast); Mappability data; Interaction data (Hi-C,
ChIA-PET); Repeat data (RepBase); GC contents; Bed12 anno-
tations (e.g., peak regions, and chromatin state model generated
by ChromHMM [9]). Some of these data are available on the
DROMPA website (http://www.iam.u-tokyo.ac.jp/chromo-
someinformatics/rnakato/drompa/).
2. It is noted that scaling a small number of reads up to a larger
number (e.g., 1 million ! 10 million) is not recommended
because that will result in plenty of background noise.
3. For drompa_draw, the option “-i” can take the following
comma-separated multiple fields: (1) ChIP sample (required);
(2) Input control sample; (3) Sample name to be shown in
figure; (4) peak list to be highlighted; (5) binsize; (6) Scale of
read; (7) Scale of ratio; (8) Scale of p-value. Except for the
“ChIP sample”, all the other fields can be omitted. For
example,

-i $ChIP,$Input,ChIPname,ChIPpeak.bed,1000,60

specifies 1000-bp bin size, 60 for y-scale of the read line, and the
bed file “ChIPpeak.bed” as highlighted peak regions for this
ChIP-input sample pair.

4. In addition to a peak file, drompa_peakcall can also output bin

data of the ChIP-read profile normalized the input sample with
“-outputwig” option ((1) ChIP/Input ratio; (2) ChIP-internal
p-value; (3) ChIP/Input enrichment p-value) for visualization
with other tools. The bin data is outputted into the directory
‘drompadir’.

Appendix

1. Script for the visualization of human (data obtained from the

Sequence Read Archive (SRA) under accession SRP006944)

# convert map files to bin data

$ for sample in Control H3K27me3 H3K36me3 H3K4me3 H3K9me3
Pol2_b; do
$ for bin in 100 1000 100000;do parse2wig -f BAM -i HeLa-S3_
$sample-n2-m1-hg19.sort.bam -o $sample -gt genome_table -bin-
 ChIP-seq Analysis with DROMPA 641

size $bin; done

$ done
# make pdf files
$ IP1¼"parse2wigdir/Pol2_b"
$ IP2¼"parse2wigdir/H3K4me3"
$ IP3¼"parse2wigdir/H3K27me3"
$ IP4¼"parse2wigdir/H3K36me3"
$ IP5¼"parse2wigdir/H3K9me3"
$ Input¼"parse2wigdir/Control"
# Figure 2a
$ s1¼"-i $IP1,$Input,Pol2_b,,,100"
$ s2¼"-i $IP2,$Input,H3K4me3,,,80"
$ s3¼"-i $IP3,$Input,H3K27me3,,1000,60"
$ s4¼"-i $IP4,$Input,H3K36me3,,1000,60"
$ drompa_draw PC_SHARP -gt genome_table -gene refFlat.txt $s1
$s2 $s3 $s4 -p Fig 2a -lpp 1 -chr 1 -ls 1000 -rmchr -show_itag 2
# Figure 2b
$ s1¼"-i $IP2,$Input,H3K4me3"
$ s2¼"-i $IP3,$Input,H3K27me3"
$ s3¼"-i $IP4,$Input,H3K36me3"
$ s4¼"-i $IP5,$Input,H3K9me3"
$ drompa_draw GV -gt genome_table $s1 $s2 $s3 $s4 -p Fig
2b_liner -GC GCcontents -gcsize 500000 -GD genedensity -gdsize
500000
$ drompa_draw GV -gt genome_table $s1 $s2 $s3 $s4 -p Fig 2b_log
-GC GCcontents -gcsize 500000 -GD genedensity -gdsize 500000 -
showratio 2

2. Script for the visualization of S. cerevisiae (data obtained from

SRA under accession SRP009385)

$ gt¼genome_table_sacCer3
$ index¼UCSC-sacCer3-cs # bowtie index for colorspace data
# read mapping and parse2wig
$ for num in $(seq 398609 398624); do
$ prefix¼SRR$num
$ bowtie -C $index $prefix.fastq -p8 -S > $prefix.sam
$ parse2wig -f SAM -i $prefix.sam -o $prefix -gt $gt
$done
# generate pdf files
$ dir¼parse2wigdir
$ IP1_60¼"$dir/SRR398612$postfix" # YST1019 Gal 60min
$ IP1_0¼"$dir/SRR398611$postfix" # YST1019 Gal 0min
$ IP2_60¼"$dir/SRR398610$postfix" # YST1019 Raf 60min
$ IP2_0¼"$dir/SRR398609$postfix" # YST1019 Raf 0min
$ IP3_60¼"$dir/SRR398616$postfix" # YST1053 Gal 60min
$ IP3_0¼"$dir/SRR398615$postfix" # YST1053 Gal 0min
$ IP4_60¼"$dir/SRR398614$postfix" # YST1053 Raf 60min
642 Ryuichiro Nakato and Katsuhiko Shirahige

$ IP4_0¼"$dir/SRR398613$postfix" # YST1053 Raf 0min

$ IP5_60¼"$dir/SRR398620$postfix" # YST1076 Gal 60min
$ IP5_0¼"$dir/SRR398618$postfix" # YST1076 Gal 0min
$ IP6_60¼"$dir/SRR398619$postfix" # YST1076 Raf 60min
$ IP6_0¼"$dir/SRR398617$postfix" # YST1076 Raf 0min
$ IP7_60¼"$dir/SRR398624$postfix" # YST1287 Gal 60min
$ IP7_0¼"$dir/SRR398623$postfix" # YST1287 Gal 0min
$ IP8_60¼"$dir/SRR398622$postfix" # YST1287 Raf 60min
$ IP8_0¼"$dir/SRR398621$postfix" # YST1287 Raf 0min
$ s1¼"-i $IP1_60,$IP1_0,YST1019_Gal"
$ s2¼"-i $IP2_60,$IP2_0,YST1019_Raf"
$ s3¼"-i $IP3_60,$IP3_0,YST1053_Gal"
$ s4¼"-i $IP4_60,$IP4_0,YST1053_Raf"
$ s5¼"-i $IP5_60,$IP5_0,YST1076_Gal"
$ s6¼"-i $IP6_60,$IP6_0,YST1076_Raf"
$ s7¼"-i $IP7_60,$IP7_0,YST1287_Gal"
$ s8¼"-i $IP8_60,$IP8_0,YST1287_Raf"
$ drompa_draw PC_ENRICH -p SRP009385 -gt $gt -ars ARS-oriDB_s-
cer.txt -gene SGD_features.tab -gftype 3 $s1 $s2 $s3 $s4 $s5
$s6 $s7 $s8 -ls 250 -lpp 2 -scale_ratio 1 -bn 3 -ystep 14 -
ethre 1.5

References
1. Nakato R, Shirahige K (2017) Recent advances species. Nature 512(7515):445–448. doi:10.
in ChIP-seq analysis: from quality management 1038/nature13424
to whole-genome annotation. Brief Bioinform 3. Sutani T, Sakata T, Nakato R, Masuda K, Ishiba-
18:279. doi:10.1093/bib/bbw023 shi M, Yamashita D, Suzuki Y, Hirano T, Bando
2. Gerstein MB, Rozowsky J, Yan KK, Wang D, M, Shirahige K (2015) Condensin targets and
Cheng C, Brown JB, Davis CA, Hillier L, Sisu reduces unwound DNA structures associated
C, Li JJ, Pei B, Harmanci AO, Duff MO, Djebali with transcription in mitotic chromosome con-
S, Alexander RP, Alver BH, Auerbach R, Bell K, densation. Nat Commun 6:7815. doi:10.1038/
Bickel PJ, Boeck ME, Boley NP, Booth BW, ncomms8815
Cherbas L, Cherbas P, Di C, Dobin A, Drenkow 4. Consortium RE, Kundaje A, Meuleman W,
J, Ewing B, Fang G, Fastuca M, Feingold EA, Ernst J, Bilenky M, Yen A, Heravi-Moussavi A,
Frankish A, Gao G, Good PJ, Guigo R, Ham- Kheradpour P, Zhang Z, Wang J, Ziller MJ,
monds A, Harrow J, Hoskins RA, Howald C, Amin V, Whitaker JW, Schultz MD, Ward LD,
Hu L, Huang H, Hubbard TJ, Huynh C, Jha S, Sarkar A, Quon G, Sandstrom RS, Eaton ML,
Kasper D, Kato M, Kaufman TC, Kitchen RR, Wu YC, Pfenning AR, Wang X, Claussnitzer M,
Ladewig E, Lagarde J, Lai E, Leng J, Lu Z, Liu Y, Coarfa C, Harris RA, Shoresh N, Epstein
MacCoss M, May G, McWhirter R, Merrihew CB, Gjoneska E, Leung D, Xie W, Hawkins RD,
G, Miller DM, Mortazavi A, Murad R, Oliver B, Lister R, Hong C, Gascard P, Mungall AJ,
Olson S, Park PJ, Pazin MJ, Perrimon N, Per- Moore R, Chuah E, Tam A, Canfield TK, Han-
vouchine D, Reinke V, Reymond A, Robinson sen RS, Kaul R, Sabo PJ, Bansal MS, Carles A,
G, Samsonova A, Saunders GI, Schlesinger F, Dixon JR, Farh KH, Feizi S, Karlic R, Kim AR,
Sethi A, Slack FJ, Spencer WC, Stoiber MH, Kulkarni A, Li D, Lowdon R, Elliott G, Mercer
Strasbourger P, Tanzer A, Thompson OA, Wan TR, Neph SJ, Onuchic V, Polak P, Rajagopal N,
KH, Wang G, Wang H, Watkins KL, Wen J, Wen Ray P, Sallari RC, Siebenthall KT, Sinnott-
K, Xue C, Yang L, Yip K, Zaleski C, Zhang Y, Armstrong NA, Stevens M, Thurman RE, Wu
Zheng H, Brenner SE, Graveley BR, Celniker J, Zhang B, Zhou X, Beaudet AE, Boyer LA, De
SE, Gingeras TR, Waterston R (2014) Compar- Jager PL, Farnham PJ, Fisher SJ, Haussler D,
ative analysis of the transcriptome across distant Jones SJ, Li W, Marra MA, McManus MT,
 ChIP-seq Analysis with DROMPA 643

Sunyaev S, Thomson JA, Tlsty TD, Tsai LH, alignment of short DNA sequences to the
Wang W, Waterland RA, Zhang MQ, Chadwick human genome. Genome Biol 10(3):R25.
LH, Bernstein BE, Costello JF, Ecker JR, Hirst doi:10.1186/gb-2009-10-3-r25
M, Meissner A, Milosavljevic A, Ren B, Stama- 7. Langmead B, Salzberg SL (2012) Fast gapped-
toyannopoulos JA, Wang T, Kellis M (2015) read alignment with Bowtie 2. Nat Methods 9
Integrative analysis of 111 reference human epi- (4):357–359. doi:10.1038/nmeth.1923
genomes. Nature 518(7539):317–330. doi:10. 8. Li H, Durbin R (2009) Fast and accurate short
1038/nature14248 read alignment with Burrows-Wheeler trans-
5. Nakato R, Itoh T, Shirahige K (2013) form. Bioinformatics 25(14):1754–1760.
DROMPA: easy-to-handle peak calling and visu- doi:10.1093/bioinformatics/btp324
alization software for the computational analysis 9. Ernst J, Kellis M (2012) ChromHMM: auto-
and validation of ChIP-seq data. Genes Cells 18 mating chromatin-state discovery and character-
(7):589–601. doi:10.1111/gtc.12058 ization. Nat Methods 9(3):215–216. doi:10.
6. Langmead B, Trapnell C, Pop M, Salzberg SL 1038/nmeth.1906
(2009) Ultrafast and memory-efficient
 Chapter 42

Quantitative Analysis of DNA Damage Signaling Responses

to Chemical and Genetic Perturbations
Francisco M. Bastos de Oliveira, Dongsung Kim, Michael Lanz,
and Marcus B. Smolka

Abstract
Phosphorylation-mediated signaling is essential for maintenance of the eukaryotic genome. The evolution-
arily conserved kinases ATR and ATM sense specific DNA structures generated upon DNA damage or
replication stress and mediate an extensive signaling network that impinges upon most nuclear processes.
ATR/ATM signaling is highly regulated and can function in a context-dependent manner. Thus, the ability
to quantitatively monitor most, if not all, signaling events in this network is essential to investigate the
mechanisms by which kinases maintain genome integrity. Here we describe a method for the Quantitative
Mass-Spectrometry Analysis of Phospho-Substrates (QMAPS) to monitor in vivo DNA damage signaling in
a systematic, unbiased, and quantitative manner. Using the model organism Saccharomyces cerevisiae, we
provide an example for how QMAPS can be applied to define the effect of genotoxins, illustrating the
importance of quantitatively monitoring multiple kinase substrates to comprehensively understanding
kinase action. QMAPS can be easily extended to other organisms or signaling pathways where kinases can
be deleted or inhibited.

Key words DNA damage checkpoint, DNA damage signaling, Phosphorylation, Quantitative mass
spectrometry, Saccharomyces cerevisiae

1 Introduction

The integrity of eukaryotic genomes relies on the action of the

evolutionary conserved PI3K-like kinases ATR and ATM, which
detect aberrant DNA structures and mediate complex signaling
responses [1]. In Saccharomyces cerevisiae, DNA damage signaling
is initiated by Mec1 and Tel1 (orthologs of human ATR and ATM)
and further transduced to the downstream effector kinase Rad53
(ortholog of human CHK2) [2, 3]. These kinases phosphorylate a
diverse array of proteins and recent proteomic studies have identi-
fied over 100 phosphorylation events that depend on Mec1 and/or
Tel1 [4–7]. Despite the complexity of the DNA damage signaling
network, most studies typically rely on a few selected substrates as

Marco Muzi-Falconi and Grant W. Brown (eds.), Genome Instability: Methods and Protocols, Methods in Molecular Biology,
vol. 1672, DOI 10.1007/978-1-4939-7306-4_42, © Springer Science+Business Media LLC 2018

645
646 Francisco M. Bastos de Oliveira et al.

readouts of the overall activity of the kinases involved. Since Mec1

and Tel1 can target different substrates under different conditions
[5], probing one or a few substrates will not capture all aspects of
Mec1/Tel1 regulation and action.
To comprehensively assess DNA damage signaling, our labora-
tory has recently developed a method for the Quantitative Mass-
Spectrometry Analysis of Phospho-Substrates (QMAPS). This pro-
teomic approach allows us to monitor the dynamics of DNA dam-
age signaling by looking at multiple in vivo kinase substrates in a
systematic, unbiased, and quantitative manner [5]. Our method
employs Stable Isotope Labeling of Amino acids in Cell culture
(SILAC), followed by Immobilized Metal Affinity Chromatogra-
phy (IMAC) and LC-MS/MS to compare the relative abundance of
specific kinase substrates in different conditions [8, 9].
Importantly, QMAPS analysis implicitly relies on a predefined
list of kinase-dependent phosphorylation events. In order to define
a set of kinase-dependent phosphorylation events, the phosphopro-
teomes of wild type cells (WT) are compared against the phospho-
proteomes of mutants lacking kinases involved in DNA damage
signaling (see Fig. 1). Phosphorylation events dependent on these
kinases can be quantitatively monitored upon a range of different
conditions, such as treatment with different genotoxins or muta-
tions/deletions of specific genes. As an example for this protocol,
here, we use the Mec1/Tel1 and Rad53 phospho-substrates to
monitor differences in signaling events induced by two different
genotoxic drugs (see Figs. 2 and 3).

2 Materials

2.1 Stable Isotope 1. Saccharomyces cerevisiae: Strains MBS164 (MATa, ura3-52,

Labeling of Amino leu2Δ1, trp1Δ63, his3Δ200, lys2ΔBgl, hom3-10, ade2Δ1, ade8,
Acids in Cell Culture arg4Δ, sml1::ΔTRP1, bar1::ΔHIS3), MBS2042 (MATa, ura3-
(SILAC) 52, leu2Δ1, trp1Δ63, his3Δ200, lys2ΔBgl, hom3-10, ade2Δ1,
ade8, arg4Δ, sml1::ΔTRP1, bar1::ΔHIS3, mec1::ΔURA3,
tel1::ΔkanMX6) and MB188 (MATa, ura3-52, leu2Δ1,
trp1Δ63, his3Δ200, lys2ΔBgl, hom3-10, ade2Δ1, ade8, arg4Δ,
sml1::ΔTRP1, bar1::ΔHIS3, rad53::ΔURA3) (see Note 1).
2. Culture media: Synthetic Defined (SD)
arginine
lysine
media for SILAC: 0.67 g of Complete Supplement Mixture
(CSM)
arg
his
lys, 6.7 g of YNB, 80 mg of L-proline,
40 mg of L-histidine, and 960 ml of distilled water. After
autoclaving, add 40 ml of a 50% dextrose solution to a final
concentration of 2% (see Note 2).
3. “Light” arginine/lysine stock solution (1000): 30 mg/ml of
L-arginine (0.2 M), 20 mg/ml of L-lysine (0.115 M). Dissolve
amino acids in distilled water and filter using a 0.22 μm dispos-
able syringe filter.
Quantitative Mass-Spectrometry Analysis of Phospho-Substrates (QMAPS) 647

Step 1

Defining kinase-dependet substrates

during DNA damage signaling

WT mec1Δ tel1Δ or rad53Δ

+10 +8
“Light” Arg; Lys “Heavy” Arg ; Lys

HU HU

IMAC

HILIC fractionation

LC-MS/MS

Quantitative Analysis
Intensity
Intensity

m/z m/z

Ratio L/H = 1 Ratio L/H > 1

Kinase independent Kinase dependent
phosphorylation phosphorylation

Fig. 1 Overview of the methodology for the identification of Mec1/Tel1 or Rad53-

dependent substrates during DNA damage response. Wild type and kinase null
mutant strains were grown in “Light” and “Heavy” media, respectively. Cells
were arrested in G1 with alpha-factor and released in media containing 0.04% of
methyl methanesulfonate (MMS) or 100 mM of hydroxyurea (HU). After
treatment, cultures were combined, proteins were extracted, digested with
trypsin, desalted, and subjected to phosphopeptide enrichment using IMAC.
Phosphopeptide elutions were fractionated by HILIC and subjected to
LC–quantitative mass spectrometry analysis. Phosphopeptide ratio equal or
lower than 1 suggests a checkpoint kinase-independent or dependent
phosphorylation, respectively. Dashed line highlights the quantitation profile
expected for the checkpoint kinase-dependent event
648 Francisco M. Bastos de Oliveira et al.

Step 2
Defining chemichal perturbations
to kinase-dependent DNA damage signaling

WT WT

+10 +8
“Light” Arg; Lys “Heavy” Arg ; Lys

MMS Hydroxyurea

IMAC

HILIC fractionation

LC/MS

Quantitative Analysis

Consider ONLY
kinase-dependet
substrates selected on STEP 1

QMAPS generation

Fold Change in Phosphorylation Fold Change in Phosphorylation Fold Change in Phosphorylation

10 8 6 4 2 2 4 6 8 10 10 8 6 4 2 2 4 6 8 10 10 8 6 4 2 2 4 6 8 10
MMS treated

HU treated

Substrate is more phosphorylated Substrate is equally phosphorylated Substrate is more phosphorylated

in cells treated with MMS then HU in cells treated with MMS or HU in cells treated with HU then MMS

Fig. 2 Overview of the methodology for the QMAPS in cells treated with methyl methanesulfonate or
hydroxyurea. Wild type cells were arrested in G1 with alpha-factor and released in “Heavy” or “Light”
media containing 0.001% of methyl methanesulfonate (MMS) or 100 mM of hydroxyurea (HU), respectively.
After incubation, cultures were combined, proteins were extracted, digested with trypsin, desalted, and
subjected to phosphopeptide enrichment using IMAC. Phosphopeptide elutions were fractionated by HILIC
and subjected to LC–quantitative mass spectrometry analysis. To generate the QMAPS, results of the
 Quantitative Mass-Spectrometry Analysis of Phospho-Substrates (QMAPS) 649

4. “Heavy” arginine/lysine stock solution (1000): 38 mg/ml of

13
L-arginine C6, 15N4 (0.2 M), 25 mg/ml of L-lysine 13C6,
15
N2 (0.115 M). Dissolve amino acids in distilled water and
filter using a 0.22 μm disposable syringe filter.
5. To prepare 1 L of SD “Light” or SD “Heavy” add 1 ml of
1000  “Light” arg/lys or 1000  “Heavy” arg/lys stock
solution to 1 L of SD
arg
lys.

a b
Fold Change in Phosphorylation
10 < 8 6 4 2 2 4 6 8 > 10 HU
MMS
(0.001%) (100 mM)
Mec1S38
Mec1/Tel1 Mec1/Tel1
0.001% MMS

100 mM HU

Rfa1
Rfa1S178
Substrates Substrates
Rad53 Rad53

Substrates Substrates
Mec1/Tel1-dependent Rad53-independent
Rad53-dependent

Fig. 3 QMAPS reveals uncoupling of Mec1/Tel1 and Rad53 signaling. (a) QMAPS depicting the fold change in
phosphopeptide abundance between cells treated with MMS or HU. Blue dots correspond to Mec1/Tel1-
dependent and Rad53-independent substrates and red dots correspond to Rad53-dependent substrates. Red
dashed line highlights the uncoupling between Mec1/Tel1 and Rad53-dependent signaling in cells treated
with MMS or HU. (b) Schematic representation of Mec1/Tel1 and Rad53 signaling in cells treated with MMS or
HU. Global analysis of DNA damage signaling allowed us to identify an increase in Mec1 autophosphorylation
and Mec1-dependent Rfa1 phosphorylation in cells treated with 0.001% MMS compared to HU-treated cells.
We also observed an uncoupling between Mec1/Tel1 and Rad53 signaling in cells treated with 0.001% MMS.
Blue arrows represent Mec1/Tel1-dependent and Rad53-independent phosphosignaling and red arrows
represent Rad53-dependent phosphosignaling. Large, medium, and small arrows represent the variations
on phosphosignal intensity

Fig. 2 (continued) phosphoproteome analysis were filtered using the list of kinase-dependent phosphopep-
tides identified in step 1 (see Fig. 1). In the QMAPS, we depict the fold change in phosphopeptide abundance
converted from “Light”versus“Heavy” relative abundances. Each dot corresponds to a different checkpoint
kinase-dependent phosphopeptide identified at least three times in two independent biological replicates.
Phosphopeptides deviated to the left or to the right side of the plot indicate an increase in their abundance in
cells treated with MMS or HU, respectively. Dot placed in the middle of the plot indicates same level of
phosphorylation in both conditions
650 Francisco M. Bastos de Oliveira et al.

2.2 Cell Cycle Arrest, 1. Alpha-factor mating pheromone.

Genotoxin Treatment, 2. Protease from Streptomyces griseus.
and Cell Harvesting
3. Hydroxyurea.
4. Methyl methanesulfonate.
5. TE buffer: 10 mM Tris–HCl pH 8.0 and 5 mM EDTA.

2.3 Cell Lysis 1. Lysis buffer: 50 mM Tris–HCl pH 8.0, 0.2% Tergitol, 150 mM
NaCl, 5 mM EDTA, 1 mM phenylmethylsulfonyl fluoride
(PMSF), 1  complete EDTA-free protease inhibitor cocktail
(Roche), and 1  phosphatase inhibitor (see Note 3).
2. 100  phosphatase inhibitor: 100 mM Na3VO4, 500 mM
NaF, and 1 M β-glycerophosphate.
3. Glass beads, 0.5 mm dia.
4. Protein assay dye reagent.
5. Bead Beater type cell disruptor.

2.4 Protein 1. Alkylation solution: 1 M Tris–HCl pH 8.0 and 0.5 M iodoa-

Precipitation and cetamide (see Note 4).
Tryptic Digestion 2. Protein precipitation (PPT) solution: 50% acetone, 49.9% eth-
anol and 0.1% biochemistry grade acetic acid.
3. Water bath sonicator.
4. Urea–Tris solution: 100 mM Tris pH 8.0 and 8 M urea
(see Note 5).
5. NaCl–Tris solution: 50 mM Tris–HCl pH 8.0 and 150 mM
NaCl.
6. TPCK treated trypsin 10 mg/ml in 0.1% biochemistry grade
acetic acid (see Note 6).

2.5 Protein Sample 1. 10% LC-MS grade formic acid (see Note 7).
Clean-Up 2. 10% trifluoroacetic acid (see Note 7).
3. C18 cartridges Vac 1 cm3, 1 G (Sep-Pak).
4. C18 buffer A: 0.1% trifluoroacetic acid (see Note 7).
5. C18 buffer B: 0.1% biochemistry grade acetic acid (see Note 7).
6. C18 buffer C: 80% HPLC grade acetonitrile and 0.1% biochem-
istry grade acetic acid (see Note 7).
7. Silanized glass vial, 300 μl.
8. Silanized glass vial, 2 ml.
 Quantitative Mass-Spectrometry Analysis of Phospho-Substrates (QMAPS) 651

2.6 Preparation of 1. Ni-NTA (silica) spin columns (Qiagen).

Immobilized Metal 2. Stripping solution: 50 mM EDTA and 1 M NaCl.
Affinity
3. Iron (III) chloride solution: 100 mM FeCl3 in 0.3% biochem-
Chromatography
istry grade acetic acid (see Note 8).
(IMAC) Resin
4. IMAC washing solution 1: 25% acetonitrile, 0.1 M NaCl, and
0.1% biochemistry grade acetic acid.

2.7 Phosphopeptide 1. Glass wool.

Enrichment Using 2. Gel loading tips, 200 μl.
IMAC
3. 1 ml slip-tip disposable syringe w/o needle.
4. IMAC washing solution 1: 25% acetonitrile, 0.1 M NaCl, and
0.1% biochemistry grade acetic acid.
5. IMAC washing solution 2: 1% biochemistry grade acetic acid.
6. IMAC eluting solution: 12% ammonia and 10% HPLC grade
acetonitrile.

2.8 Hydrophilic 1. This protocol assumes access to high performance liquid chro-
Interaction Liquid matography (HPLC) equipment including an elution gradient
Chromatography programmer with data processing software, UV absorbance
(HILIC) detector, and automated fraction collector.
2. 99.9% HPLC grade acetonitrile (see Note 7).
3. Column: 2.0  150 mm TSK gel Amide-80 5 μm particle.
4. Buffer D: 90% HPLC grade acetonitrile (see Note 7).
5. Buffer E: 80% HPLC grade acetonitrile with 0.005% trifluor-
oacetic acid (see Note 7).
6. Buffer F: 0.025% trifluoroacetic acid (see Note 7).

2.9 Reverse Phase 1. The protocol assumes access to an online nano LC system
Liquid interfaced with a mass spectrometer capable of performing
Chromatography– tandem MS/MS. The present study was performed using an
Tandem Mass on-line Nano LC-Ultra® system (Eksigent) coupled with a Q
Spectrometry Exactive mass spectrometer (Thermo Fisher Scientific).
2. Reverse phase analytical column: 125 μm ID  20 cm in-house
packed with 3 μm C18 resin (see Note 9).
3. Buffer I: 0.1% LC-MS grade formic acid (see Note 7).
4. Buffer II: 0.1% LC-MS grade formic acid with 80% HPLC
grade acetonitrile (see Note 7).

2.10 Database 1. For database search: SEQUEST software on a SORCERER

Searching, system (Sage-N Research).
Phosphopeptide 2. For quantification of relative protein abundances: XPRESS
Identification, software, part of the Trans-Proteomic Pipeline (Seattle Prote-
Quantification, and ome Center-SPC).
QMAPS Generation
652 Francisco M. Bastos de Oliveira et al.

3. To increase the confidence of phosphopeptide identification,

we performed a parallel search on Proteome Discoverer 1.4
software (Thermo Fisher Scientific) running SEQUEST and
Percolator.
4. For phospho-site localization probabilities we used PhosphoRS
within Proteome Discoverer (version 1.4.1.14, Thermo Fisher
Scientific).

3 Methods

3.1 Stable Isotope The basic principle of SILAC for quantitative phosphoproteomic
Labeling of Amino analysis consists of growing two cell cultures: one in a medium
Acids in Cell Culture complemented with normal (“Light”) amino acids and the other
(SILAC) in a medium complemented with stable-isotope labeled (“Heavy”)
amino acids. When “Heavy” and “Light” samples are combined
and analyzed by mass spectrometry, each phosphopeptide is
detected as a pair in the mass spectra, with a predictable mass shift
between the “Heavy” and “Light” forms of the phosphopeptide. In
this case, because the “Light” and “Heavy” amino acids are chemi-
cally identical, the ratio of peak intensities in the mass spectrometer
directly represents the ratio of protein phosphorylation abundance
between the two cultures.
1. Inoculate fresh colonies from either strain MBS164 and
MBS2042 (or MBS188) into 100 ml of SD “Light” or SD
“Heavy” media, respectively. Then grow cultures for at least
12 h, at 30  C with constant shaking. For step 2 of the analysis,
inoculate 100 ml of SD “Light” and SD “Heavy” media with
fresh colonies from MBS164 (see Figs. 1 and 2 and Note 10).
2. On the next day the cultures should have an optical density at
600 nm (OD600) around 0.5. Dilute cultures in their respective
SD media to an OD600 ¼ 0.1 in a final volume of 200 ml (see
Note 11).

3.2 Cell Cycle Arrest, Hydroxyurea (HU) is an inhibitor of ribonucleotide reductase. By

Drug Induced-DNA depleting pools of dNTPs, HU stalls the progression of replication
Damage, and Cell forks, leading to single-stranded DNA exposure and activation of
Harvesting DNA damage signaling [10]. Also, the DNA alkylating agent
methyl methanesulfonate (MMS) can efficiently activate the DNA
damage response during S-phase but not during G1 [11]. For the
purpose of this protocol, we use 0.04% of MMS and 100 mM of
HU to assess Mec1/Tel1 and Rad53-dependent signaling during
S-phase.
1. When cells reach an OD600 ¼ 0.2 add alpha-factor to final
concentration of 50 ng/μl for MBS164 and 50 μg/μl for
MBS2042/MBS188 and keep cultures growing for 3 h.
 Quantitative Mass-Spectrometry Analysis of Phospho-Substrates (QMAPS) 653

2. After 3 h, cell cultures should be around OD600 ¼ 0.4. Centri-

fuge the cultures for 2500  g for 3 min at RT. Discard super-
natant and resuspend cells in prewarmed SD media containing
pronase at 100 μg/ml and HU or MMS to final concentration of
0.1 M and 0.04%, respectively. Keep cultures growing (see Note
12).
3. After drug treatment, normalize the amount of cells in each
culture based on their respective OD600.
4. Mix equal amounts of cells from both, light and heavy cultures
and divide the final volume into two separated 500 ml conical
centrifugation bottles.
5. Centrifuge the cultures for 1000  g for 5 min at 4  C. Discard
supernatant.
6. Add 10 ml of ice-cold TE buffer to each centrifugation bottle,
resuspend cell pellets and transfer each to a separate 15 ml
Falcon tubes
7. Repeat step 5.
8. Resuspend each cell pellet in 4 ml of ice-cold TE buffer and
divide the solution in 2 ml conical screw cap tube (total of
4  2 ml conical screw cap tubes, each one containing 1 ml
of cells).
9. Spin down cells for 1 min in a bench top microcentrifuge for
10,000  g at room temperature (RT) and remove superna-
tant. Cell pellets can be stored at
80  C for up to 2 weeks.

3.3 Cell Lysis 1. Add 600 μl of ice-cold glass beads to each tube.
2. Add 1 ml of lysis buffer to each tube and break cells at 4  C for
30 min, with a 1 min pause at each 10 min interval, using the
bead beater cell disruptor.
3. Transfer the lysate together with glass beads to a falcon tube
and leave it on ice until glass beads decant.
4. Transfer the lysate to a 50 ml polycarbonate centrifuge tube
and clear lysate by centrifugation at 45,000  g at 4  C for
30 min.
5. Collect supernatant in a new conical tube and determine the
protein concentration by using protein assay dye reagent. Final
protein concentration should be around 2 mg/ml with a total
of 20 mg of protein.

3.4 Protein 1. To denature and reduce disulfide bonds prior to alkylation, add
Precipitation and SDS and DTT to a final concentration of 1% and 5 mM,
Tryptic Digestion respectively. Incubate samples at 60  C for 10 min.
2. For alkylation of cysteines, add iodoacetamide to a final con-
centration of 25 mM and incubate samples for 15 min at RT.
654 Francisco M. Bastos de Oliveira et al.

Transfer the solution (approximately 12 ml) to a 200 ml conical

bottle.
3. Add 36 ml of PPT solution. Mix it and keep on ice for 10 min
(see Note 13).
4. Centrifuge sample for at 4700  g for 10 min at RT.
5. Pour supernatant out and keep the bottles upside down for
5 min to drain the remaining PPT solution.
6. Add 4.5 ml of distilled water and 45 μl of urea–Tris solution
and resuspend protein pellet using a pipet-aid (see Note 14).
7. Transfer the solution to a 50 ml polycarbonate centrifuge tube
and centrifuge at 45,000  g for 4 min at RT.
8. Carefully remove supernatant with a tip connected to a vacuum
line.
9. Wash the walls twice with 1 ml of distilled water without
disturbing the pellet and repeat step 8.
10. Add 6 ml of urea–Tris solution and solubilize pellet by pipet-
ting up and down.
11. Add 18 ml of Tris–NaCl solution (see Note 15).
12. Add 400 μg of TPCK treated trypsin (40 μl of a 10 mg/ml
stock) and incubate sample overnight at 37  C under constant
agitation.

3.5 Protein Sample 1. On the next day, using a glass micro-syringes, add 500 μl of
Clean-Up 10% formic acid and 500 μl of 10% TFA.
2. To remove particulates prior to sample cleanup, centrifuge
tubes at 2500  g for 5 min at RT.
3. By applying air pressure with a pipette bulb, condition 1 g of
C18 column by adding 2 ml of C18 buffer D (see Note 16).
4. Add 4 ml of C18 buffer A to equilibrate column.
5. Apply the sample through the column and let it flow by gravity.
6. Add 4 ml of C18 buffer B to wash the column. Let it flow by
gravity.
7. Repeat step 6.
8. Wipe residual volume of C18 buffer C off column tip using a
paper wipe.
9. By applying air pressure with a bulb, elute sample in 400 μl of
C18 buffer C in a 2 ml silanized glass vial.
10. Using a gel loader tip, mix sample by pipetting up and down.

11. Dry sample completely at 45 C using a speed-vac
concentrator.
12. Add 400 μl of 1% acetic acid and resuspend sample by pipetting
up and down.
 Quantitative Mass-Spectrometry Analysis of Phospho-Substrates (QMAPS) 655

13. Transfer the vial to 2 ml microcentrifuge tube and spin it at

5000  g for 1 min at room temperature to remove particulates
prior to IMAC chromatography.

3.6 Preparation of To increase coverage of the phosphoproteome we use Immobilized

Immobilized Metal Metal Affinity Chromatography (IMAC) as a method to enrich for
Affinity low abundance phosphopeptides prior to mass spectrometry analy-
Chromatography sis. The IMAC was performed using a homemade Fe3+ resin [8].
(IMAC) Resin 1. With the help of a paper clip, disassemble one Ni-NTA spin
column and add resin (approximately 60 mg) to a 50 ml falcon
tube containing 25 ml of stripping solution.
2. To strip Ni out from the resin, rotate for 40 min at RT on a
nutator.
3. Spin resin for at 1000  g for 1 min at RT.
4. Remove most of the solution using a glass pipette coupled to a
vacuum line.
5. Add 25 ml of water.
6. Repeat steps 3 and 4.
7. Add 25 ml 0.6% acetic acid.
8. Repeat steps 3 and 4.
9. To bind iron to the resin, add 20 ml of iron solution and rotate
for 40 min at RT.
10. Repeat steps 3 and 4.
11. Add 20 ml of 0.6% acetic acid.
12. Repeat steps 3 and 4.
13. Add 20 ml of 0.1% acetic acid.
14. Repeat steps 3 and 4.
15. Resuspend the resin in 1 ml of 0.1% acetic acid and transfer into
a microcentrifuge tube. Wait until the resin is settled and then
remove the supernatant leaving only 100 μl (see Note 18).

3.7 Phosphopeptide 1. Place glass wool fiber at the end of a loading tip and clip to
Enrichment Using stabilize.
IMAC 2. Using a micropipette, add 100 μl of IMAC slurry from Sub-
heading 3.6, step 15 (30 μl of IMAC resin) to the loading tip.
3. Using a slip-tip disposable syringe, apply constant air pressure
to the column and pack the resin (see Note 16).
4. Add trypsinized protein sample to the fresh packed IMAC
resin. Run the sample through the resin by applying constant
air pressure with a 1 ml slip-tip disposable syringe and collect
the flow through as well (see Note 17)
656 Francisco M. Bastos de Oliveira et al.

5. Add 30 μl of IMAC washing solution 1 and run through

column.
6. Add 60 μl of IMAC washing solution 2 and run through
column.
7. Add 30 μl of water and run through column.
8. Add 90 μl of IMAC eluting solution and elute phosphopep-
tides into a 300 μl silanized conical base insert vials.

9. Dry the sample completely at 45 C using a speed-vac
concentrator.
10. Add 15 μl of water and mix by pipetting up and down.
11. Using a glass microsyringes, add 10 μl of 10% formic acid.
12. Using a glass microsyringes, add 60 μl of acetonitrile and mix
well.
13. Spin it at 5000  g for 1 min at room temperature to remove
any precipitates prior to HILIC chromatography.

3.8 Hydrophilic The unique separation ability of hydrophilic interaction liquid

Interaction Liquid chromatography (HILIC) and is orthogonality toward other
Chromatography reverse phase separation techniques make it an ideal method for
(HILIC) multidimensional fractionation of phosphopeptides prior to MS
analysis [12]. Furthermore, because the HILIC mobile phase con-
tains only acetonitrile, water and trifluoroacetic acid, fractions do
not require an extra purification step for desalting. The present
study was performed using a Dionex Ultimate 3000 HPLC con-
trolled by Chromeleon 6.8 software. A TSK gel Amide-80 column
(2 mm  150 mm, 5 μm; Tosoh Bioscience) was used for the
HILIC fractionation experiment. Three buffers were used for the
gradient: buffer D (90% acetonitrile); buffer E (80% acetonitrile
and 0.005% trifluoroacetic acid); and buffer F (0.025% trifluoroa-
cetic acid).
1. Load samples into HILIC TSK gel Amide-80 column via a 100
μl loop and set the fraction collector to time-based mode,
collecting 1 min fractions between 10 and 22 min of the
gradient in 300 μl silanized conical base insert vials. The gradi-
ent used consists on a 100 % buffer D at time ¼ 0 min, 88 % of
buffer E and 12 % of buffer F at time ¼ 5 min, 60 % of buffer E
and 40 % of buffer F at time ¼ 30 min and 5 % of buffer E and
95 % of buffer F from time ¼ 35 to 40 min in a flow rate of 150
μl/min.
2. Once the samples are fractionated, dry the sample fractions
completely at 45  C using a speed-vac concentrator.
3. Resuspend each fraction in 7 μl of 0.1% trifluoroacetic acid.
 Quantitative Mass-Spectrometry Analysis of Phospho-Substrates (QMAPS) 657

3.9 Reverse Phase As part of our multidimensional separation platform, we perform a

Liquid C18 reverse phase fractionation prior to MS analysis. C18 reverse
Chromatography- phase chromatography is a high resolution fractionation technique
Tandem Mass that presents a selectivity orthogonal to HILIC [12]. The present
Spectrometry study was performed by using an on-line Nano LC-Ultra system
coupled with a Q Exactive mass spectrometer.
1. Load sample fractions into a 20-cm column with 125 μm inner
diameter, packed in-house with 3 μm C18 particles. Reverse
phase chromatography is performed with a binary buffer sys-
tem consisting of buffer I (0.1% formic acid) and buffer II
(0.1% formic acid with 80% acetonitrile). The peptides are
separated by a linear gradient of buffer II up to 95% for
80 min with a flow rate of 200 nl/min.
2. Xcalibur 2.2 software (Thermo Fischer Scientific) was used for
the data acquisition and the Q Exactive [13] was operated in
the data-dependent mode. Survey scans were acquired in the
Orbitrap mass analyzer over the range of 380–2000 m/z with a
resolution of 70,000. The maximum ion injection time for the
survey scan was 80 ms with a 3e8 automatic gain-control
target. MS/MS spectra was performed selecting up to the ten
most abundant precursor ions with a charge state 2 and <5
within an isolation window of 2.0 m/z. Selected precursor ions
were fragmented by higher energy collisional dissociation with
normalized collision energy of 27 and analyzed in the Orbitrap
mass analyzer with a mass resolution of 17,500.

3.10 Database Raw MS/MS spectra were searched in a SORCERER (Sage N

Searching, Research, Inc.) system using SEQUEST software and a composite
Phosphopeptide yeast protein database, consisting of both the normal yeast protein
Identification, sequences and their reversed protein sequences as a decoy to esti-
Quantification, and mate the false discovery rate (FDR) in the search results [14]. To
QMAPS Generation increase the confidence of phosphopeptide identification, we per-
formed a parallel search on Proteome Discoverer 1.4 software
(Thermo Fisher Scientific) running SEQUEST and Percolator.
Only phosphopeptides identified with high confidence in both
SORCERER and Proteome discoverer/percolator searches were
considered, which resulted in extremely low false discovery rates
(FDR < 0.02%) [15].
1. The following parameters were used in the database search:
Semitryptic requirement, a mass accuracy of 15 ppm for the
precursor ions, differential modification of 8.0142 Da for
lysine, 10.00827 Da for arginine, 79.966331 Da for phosphor-
ylation of serine, threonine and tyrosine and a static mass
modification of 57.021465 Da for alkylated cysteine residues.
658 Francisco M. Bastos de Oliveira et al.

2. XPRESS software, part of the Trans-Proteomic Pipeline (Seat-

tle Proteome Center), was used to quantify all the identified
peptides [16].
3. The phosphorylation localization probabilities were deter-
mined by using PhosphoRS within Proteome Discoverer (ver-
sion 1.4.1.14, Thermo Fisher Scientific) [17].
4. In the first step of our analysis we established the following
criteria to assign Mec1/Tel1-dependent phosphopeptides: (1)
At least sixfold increase in phosphopeptide abundance in WT
relative to mec1Δtel1Δ cells. (2) For Rad53-dependent phos-
phopeptides, we established a threshold of twofold increase in
abundance in WT relative to rad53Δ cells. (3) Moreover, to be
considered a Mec1/Tel1-dependent substrate, each phospho-
peptide must be Mec1/Tel1-dependent but not Rad53-
dependent and to be considered a Rad53-dependent substrate,
each phosphopeptide has to be Mec1/Tel1-dependent and
Rad53-dependent substrate. (4) Finally, we considered only
phosphopeptides identified in at least two biological replicates.
5. In the QMAPS, we depict the fold change in normalized
phosphopeptide abundance between cells treated with MMS
or HU. Ratios of phosphopeptide abundance were converted
to fold change by using MATLAB software (MathWorks®) as
part of a custom-designed web tool that allowed upload of data
files from SORCERER. For the QMAPS generation, results of
the phosphoproteome analysis were filtered using the list of
preassigned kinase-dependent phosphopeptides identified in
step 1 (see Figs. 1 and 2). In the QMAPS plot, each dot
corresponds to the average fold change of one Mec1/Tel1
and/or Rad53-dependent phosphopeptide identified at least
three times in two independent biological replicates [5]. Phos-
phopeptides deviated to the left or right side of the plot indi-
cate, respectively, a decrease or increase in their
phosphorylation levels between cells treated with MMS or
HU (see Figs. 2 and 3). To assess the probability of having
each phosphopeptide with a ratio significantly distinct from
the ratios of the total set of phosphopeptides identified in the
analysis, we applied the Mann–Whitney–Wilcoxon Test (U-
test). Finally, all plotted phosphopeptides were manually
inspected for phosphor-site assignment and quantitation.

4 Notes

1. Strains MBS164, MBS2042 and MBS188 are SILAC strains

auxotrophic for lysine and arginine. These strains carry a dele-
tion of the ribonucleotide reductase inhibitor SML1 that sup-
presses the lethality of mec1Δ or rad53Δ allele [18].
Quantitative Mass-Spectrometry Analysis of Phospho-Substrates (QMAPS) 659

2. Proline is added to prevent the metabolic conversion of heavy

arginine to heavy proline.
3. Add PMSF, EDTA-free protease inhibitor cocktail and phos-
phatase inhibitor to cell lysis buffer right before proceeding to
cell lysis.
4. Prepare a fresh alkylation solution for every experiment.
5. Prepare a fresh urea–Tris solution for every experiment.
6. After the trypsin is diluted in 0.1% acetic acid, keep small
aliquots frozen at
80  C.
7. To avoid sample contamination with polymers always use a
glass syringe/glass pipette/glass graduated cylinder to dilute
strong acids and high concentrated organic solvents.
8. Prepare a 1 M stock FeCl3 solution in 0.3% acetic acid and let it
decant for 2 weeks prior to use. Prepare the iron chloride
solution by adding 2 ml of 1 M stock solution to 18 ml of
0.3% of acetic acid.
9. The analytical column was generated by pulling capillary to
5 μm-ID tip. Reverse-phase particles were packed directly
into the pulled column at 1000 psi until 20 cm long. The
column was further packed, washed, and equilibrated at 1000
psi in buffer II followed by buffer I.
10. Use fresh yeast culture colonies.
11. Avoid culture saturation in SD media (OD600 > 0.5). Once
saturated, SILAC strains will not grow well. If culture is
saturated, redilute and let it grow for at least four generations
before proceeding with the experiment.
12. Incubation time should be tested empirically depending on the
kinase mutant. For the WT and rad53Δ cells, 45 min incuba-
tion should be enough to allow cell cycle progression through
S-phase. For mec1Δtel1Δ double mutant 3–4 h incubation may
be necessary for proper release from G1 and progression
through S phase. To check cell cycle progression, FACS analy-
sis should be performed at different time points.
13. Proteins will precipitate and solution will get cloudy.
14. This is a critical step to remove hydrophilic contaminants that
interfere with HILIC.
15. This step is required to dilute urea prior to protein digestion
with trypsin.
16. Avoid letting the column dry in any step. Keep resin moist by
leaving a fine layer of liquid over it.
17. Apply constant and slow air pressure in all steps during phos-
phopeptide enrichment using IMAC.
660 Francisco M. Bastos de Oliveira et al.

Acknowledgments

We thank Beatriz S. Almeida for technical support. M.B.S. is sup-

ported by grants from the National Institutes of Health (R01-
GM097272), F.M.B.d.O. is supported by grants from FAPERJ
No E-26/010.002831/2014 and No E-26/010.003001/2014
and from CNPq No 446143/2014 and D.K. is supported by
Cornell Vertebrate Genomic Scholarship.

References
1. Awasthi P, Foiani M, Kumar A (2015) ATM 10. Alvino GM et al (2007) Replication in
and ATR signaling at a glance. J Cell Sci 128 hydroxyurea: it’s a matter of time. Mol Cell
(23):4255–4262 Biol 27(18):6396–6406
2. Cimprich KA et al (1996) cDNA cloning and 11. Tercero JA, Longhese MP, Diffley JF (2003) A
gene mapping of a candidate human cell cycle central role for DNA replication forks in check-
checkpoint protein. Proc Natl Acad Sci U S A point activation and response. Mol Cell 11
93(7):2850–2855 (5):1323–1336
3. Sanchez Y et al (1999) Control of the DNA 12. Horvatovich P et al (2010) Multidimensional
damage checkpoint by chk1 and rad53 protein chromatography coupled to mass spectrometry
kinases through distinct mechanisms. Science in analysing complex proteomics samples. J Sep
286(5442):1166–1171 Sci 33(10):1421–1437
4. Chen SH et al (2010) A proteome-wide analy- 13. Michalski A et al (2011) Mass spectrometry-
sis of kinase-substrate network in the DNA based proteomics using Q Exactive, a high-
damage response. J Biol Chem 285 performance benchtop quadrupole Orbitrap
(17):12803–12812 mass spectrometer. Mol Cell Proteomics 10
5. Bastos de Oliveira FM et al (2015) Phospho- (9):M111.011015
proteomics reveals distinct modes of Mec1/ 14. Eng JK, McCormack AL, Yates JR (1994) An
ATR signaling during DNA replication. Mol approach to correlate tandem mass spectral
Cell 57(6):1124–1132 data of peptides with amino acid sequences in
6. Hustedt N et al (2015) Yeast PP4 interacts a protein database. J Am Soc Mass Spectrom 5
with ATR homolog Ddc2-Mec1 and regulates (11):976–989
checkpoint signaling. Mol Cell 57:273 15. Kall L et al (2007) Semi-supervised learning for
7. Zhou C et al (2016) Profiling DNA damage- peptide identification from shotgun proteo-
induced phosphorylation in budding yeast mics datasets. Nat Methods 4(11):923–925
reveals diverse signaling networks. Proc Natl 16. Han DK et al (2001) Quantitative profiling of
Acad Sci U S A 113(26):E3667–E3675 differentiation-induced microsomal proteins
8. Albuquerque CP et al (2008) A multidimen- using isotope-coded affinity tags and mass
sional chromatography technology for in- spectrometry. Nat Biotechnol 19(10):946–951
depth phosphoproteome analysis. Mol Cell 17. Taus T et al (2011) Universal and confident
Proteomics 7(7):1389–1396 phosphorylation site localization using phos-
9. Ong SE et al (2002) Stable isotope labeling by phoRS. J Proteome Res 10(12):5354–5362
amino acids in cell culture, SILAC, as a simple 18. Zhao X, Muller EG, Rothstein R (1998) A
and accurate approach to expression proteo- suppressor of two essential checkpoint genes
mics. Mol Cell Proteomics 1(5):376–386 identifies a novel protein that negatively affects
dNTP pools. Mol Cell 2(3):329–340
 INDEX

A sequencing ................................................ 64, 162–163

Data visualization
Adaptor ligation .............................................64, 217, 252 ChIP-seq.................................................................. 638
A-like faker (ALF) ......................................................... 1–8 replication profiles.......................................... 196, 204
Anaphase bridges ................................ 481–491, 495–507
Deep sequencing .................................................. 195–206
Atomic force microscopy (AFM) ........................ 557–571 Deletion ................................................ 1–4, 8, 13, 44, 45,
57, 78, 133, 143, 167, 247, 301, 302, 325, 329,
B
330, 404, 422–424, 430, 431, 436, 575, 576,
Base excision repair ...................................................63, 64 580, 581, 585, 592, 594, 618–620, 624,
Bromodeoxyuridine (BrdU)................................ 209–224 6614–616
immunoprecipitation ........... 214, 219, 220, 229–231 DNA
adducts..................................................................... 107
C barcoding....................................... 213–214, 217–218
Canavanine ....................... 26, 28, 29, 36, 38–41, 44, 46, combing................................................................... 149
576–579, 584, 585, 616, 617, 618, 620, 621, 624 damage........................................35, 77–99, 101, 102,
Cell labeling.......................................................... 156, 158 149, 155, 156, 167, 195, 282, 312, 320, 477,
480, 527–539, 546–549, 615, 645–659
Checkpoint ........................ 35, 143, 156, 178, 180, 181,
183, 211, 485, 486, 537, 575, 599, 647, 649 polymerase ............................................ 47, 52, 53, 66,
Chromatin 68–72, 74, 80, 92, 93, 157, 158, 162, 214, 215,
218, 221, 227, 228, 230, 235, 240, 248, 250,
fibers................................................................ 509–525
preparation .............................................83, 85–89, 98 300, 319, 320, 322, 324, 326, 330, 331, 334,
Chromatin immunoprecipitation (ChiP) ..................... 80, 337, 473
repair ................................ 78–80, 101–105, 375, 376,
83–86, 90, 92, 93, 229–231, 631
Chromosome instability (CIN)............................. 16, 478 381–383, 475, 478, 529, 537
Chromosome segregation .............................................. 11 replication ....................................... 11, 102, 195–206,
Chromosome transmission fidelity (CTF)...............11–18 209, 211, 227–237, 245, 261–293, 295–308,
329, 347, 439, 472, 474, 477, 496, 510, 518, 520
Circular ligation ........................................................72–73
Comet assay ....................................................63, 312–314 DNA purification
Confocal microscopy .................................................... 528 genomic .................................................... 52, 330–332
high molecular weight ................................... 159, 160
Copy number variation (CNV) ........................... 196, 627
Coulter counter............................................................. 552 Xenopus egg extracts ...................................... 267, 272
Crosslinking DNA purification (extraction, isolation) from human
cells.................................................................52, 82
formaldehyde.......................................................86, 98
psoralen................................. 262–266, 283, 286, 287 DNA purification (extraction, isolation) from
yeast................................................................52, 82
D DNA-RNA hybrids .............................................. 347–358
DNA-RNA Immunoprecipitation...................... 349–350,
Data analysis 352–356
atomic force microscopy images ................... 557, 558 Double-strand breaks (DSBs) ......................65, 119–128,
ChIP-seq......................................................... 631, 633 131–144, 147, 148, 155–159, 162, 163, 165,
colony fluorescence ........................................ 621–622 167–193, 364, 375–384, 404, 424, 529, 547
colony size ............................................................... 605 mapping and localization........................................ 170
DNA fibers .............................................196, 520–524 Double-stranded DNA (dsDNA) .................... 33–36, 38,
mass spectrometry.......................................... 651, 657 48, 55, 120, 121, 138, 143, 169, 170, 174, 197,
microarray............................................................91, 93 200, 243, 251, 252, 258, 262, 282, 288, 291,
normalization ......................................................94, 95 292, 337, 368

Marco Muzi-Falconi and Grant W. Brown (eds.), Genome Instability: Methods and Protocols, Methods in Molecular Biology,
vol. 1672, DOI 10.1007/978-1-4939-7306-4, © Springer Science+Business Media LLC 2018

661
662 G ENOME INSTABILITY: METHODS
Index
AND PROTOCOLS

E Hydroxyurea (HU) ........................... 211, 212, 215, 224,

229, 232, 237, 289, 486, 487, 489, 511, 514,
E. coli ........................................................ 64, 67–69, 107, 524, 547–549, 647–649, 650, 652, 658
108, 110, 158, 160, 231, 234, 312, 315, 316,
390, 395, 397, 615 I
Electron microscopy ............................................ 261–292
Electrophoresis Image analysis............................................. 163, 293, 513,
alkaline .................................. 135, 139, 144, 247, 249 525, 580, 581, 592, 597, 617, 628
native.......................................................367–368, 370 Immunofluorescence ................................. 103, 104, 148,
single cell ........................................................ 311–317 154, 209, 351, 356–358, 388, 399, 400, 495,
two dimensional ...................................................... 298 496, 510, 512–513, 516–517
Electroporation ...................................109, 111, 116, 118 Immunoprecipitation (IP) ................................77, 81, 82,
End resection..................... 121, 131, 132, 140, 148, 172 89, 90–93, 99, 148, 196, 209–224, 227–231,
349–350, 352–356, 631, 635
F
L
Fiber spreads......................................................... 509, 517
Fitness ................................................................... 575–597 Libraries for sequencing ...........................................55, 74
Fixation ......................................171, 172, 174, 176, 184,
M
394–395, 400, 496–500, 502–507
Flow cytometry ................................................... 196, 198, Mapping sequencing reads ............................................. 55
200, 203–205, 224, 302, 307, 514, 614, 617, Marker loss ................................................................8, 404
620, 624–626, 629 Mass spectrometry (MS)...................................... 645–660
Fluctuation test ......................................51, 59, 421, 422, Microarrays ...............................45, 77–99, 196, 228, 628
424, 428, 434 Microfluidics......................................................... 537–554
Fluorescence in situ hybridization (FISH) .................387, Microsatellite instability ....................................... 422, 439
388, 389, 391, 396, 397, 399, 474, 476, 510 Mutagenesis...................................................... 33–41, 107
Fluorescence scanning .................................................. 617 Mutation rate .....................................8, 21–30, 195, 418,
Fluorescent in situ hybridization......................... 387–401 422, 424–426, 432–436, 449
Fluorescent protein ........................... 485, 486, 491, 528,
537, 546, 547, 613–629 N
Formaldehyde crosslinking.......................................86, 98
Next-generation sequencing .............................. 155–165,
Fragile site
234–236, 240, 241, 243, 244
common.............. 405, 472, 473, 475, 476, 484, 496
Non-homologous end joining ............................ 147, 167
rare .................................................472, 473, 478, 480
Nucleotide excision repair (NER)........................ 78, 101,
Fragility......................403–437, 472–474, 477, 478, 480
102, 104, 311
G P
Gene conversion................................................... 2, 3, 133
PCR
Gross chromosomal rearrangement ............................... 43
colony ................................................... 381, 383, 404,
Growth curve .............................449, 466, 576, 589, 592
405, 410–413, 416, 417, 426–427, 446,
450–452, 454–459, 461, 467, 468
H
quantitative real-time ................................... 81, 91–92
Hemocytometer ..................................28, 37, 38, 85, 149 Peak calling ................................. 193, 223, 632, 634–635
High throughput sequencing............................. 163, 197, Photo-toxicity................................................................ 529
200, 203, 209–224 Protein
HO endonuclease ............................................... 119, 122, phosphorylation ...................................................... 652
133, 364–366 recruitment .............................................................. 102
Homologous recombination (HR)................2, 117, 119, Protein-protein interactions ......................................... 599
131, 147, 172, 186, 193, 240, 376–378, 444, Pulsed-field gel electrophoresis ...................................... 45
445, 615
Hybridization R
microarray............................................................ 92–93
Recombination
Southern ......................................................... 249–250
foci ........................................................................... 529
 GENOME INSTABILITY: METHODS AND PROTOCOLS
Index
663
homology-directed.................................................. 119 Southern blot .................................... 120, 121, 124, 125,
site-specific............................................................... 111 131–154, 239–242, 246–247, 249, 304, 305,
Repeat contractions ............................................. 404, 407 363–373, 444–446, 460–464
Repeat expansions ................................................ 404, 423 Spheroplasting ................................... 134, 136, 137, 142,
Replication 266, 273, 349, 354, 366, 369, 371, 394, 395,
fork barrier............................................................... 295 400, 515
intermediates ................................261, 263, 267–268, Synchronization ......................................... 196, 198–200,
275–276, 279, 283, 285, 286, 305, 477 211–212, 215–216, 224, 232, 270, 271, 289,
origins ......................... 111, 195, 196, 209, 228, 230, 298, 302, 490, 503
234–237, 240, 299, 341–343, 423, 475 Synthetic genetic array (SGA) ...........539–597, 576, 577,
stress............................ 168, 169, 192, 193, 262, 329, 580–583, 585–588, 593, 613–629
480, 489, 496
Reporter gene..........................33–36, 38, 298, 376, 422, T
444, 445, 613, 614, 619 Telomerase.................................363, 364, 372, 375, 376,
Ribonucleic acid (RNA) localization .................. 387, 388
387–401, 406, 474, 480
Ribonucleotide Telomere addition .................................. 44, 57, 363–373,
excision repair.......................................................... 311 406, 407
incorporation................................. 239–258, 329, 330
Time-lapse imaging.............................................. 537–554
R loop ................................................................... 347–359 Transformation
Robotics..............................................580–584, 585, 587, E. coli ....................................................................... 615
588, 602, 603, 618 yeast ............................................... 297, 298, 300–301
Translesion DNA synthesis (TLS).......................... 34, 35,
S
107, 111, 116
Saccharomyces cerevisiae.......................... 1, 11, 43–60, 64, Trinucleotide repeats (TNR).......................404, 439–469
73, 78, 119–121, 126, 142, 196, 200, 211, 222,
239, 248, 264–266, 270, 273–274, 275, 289, U
339, 341, 363, 365, 369, 370, 377–379, 381,
Ultrafine bridges (UFBs)........................... 477, 483–491,
387, 403–437, 439–468, 485, 528, 540, 600, 495–497, 503–507
613–629, 637, 645, 646 Uracil incorporation........................................................ 63
Schizosaccharomyces pombe.......................... 239, 240, 242, UV damage..................................... 64, 71, 101, 103, 104
248, 255, 257, 445, 509, 511, 540
UV photodimers .......................................................65, 74
Single molecule analysis ...............................147–154, 561
Single-strand breaks (SSBs) ................................ 156, 157, W
165, 227, 312
Single-strand DNA (ssDNA)............................ 34–36, 41, Whole chromosomal loss.................................................. 3
119–121, 124, 126, 127, 131, 132, 138, 140, Whole-genome sequencing ..................45, 169, 170, 184
142–144, 147–149, 156, 228, 229, 234–236,
Y
243, 247, 251, 257, 258, 262–264, 281–283,
286, 288–290, 293, 390, 391, 397 Yeast artificial chromosome (YAC) ...405–407, 409, 410,
Sonication of DNA ....................................................... 188 413, 415, 416–418