21 October 2011

|
$10

2
0
1
1
P
e
r
k
i
n
E
l
m
e
r
,
I
n
c
.
4
0
0
2
2
7
_
0
1
.
A
l
l
t
r
a
d
e
m
a
r
k
s
o
r
r
e
g
i
s
t
e
r
e
d
t
r
a
d
e
m
a
r
k
s
a
r
e
t
h
e
p
r
o
p
e
r
t
y
o
f
P
e
r
k
i
n
E
l
m
e
r
,
I
n
c
.
a
n
d
/
o
r
i
t
s
s
u
b
s
i
d
i
a
r
i
e
s
.
When atrition rates, regulations and recall concerns are rising, geting beter information faster
is critical. PerkinElmers multimode detectioninstruments, imaging and analysis systems, versatile
assay platforms and informatics solutions enable you to identify and characterize viable drug
candidates sooner. And along with our multivendor lab services, we help you succeed faster.
See us in action at www.perkinelmer.com/biomedstories1
GeneArt

gene synthesis, build your

next breakthrough
2011 Life Technologies Corporation. All rights reserved. The trademarks mentioned herein are the property of Life Technologies Corporation or their
respective owners. For research use only. Not intended for any animal or human therapeutic or diagnostic use. CO18747
Go to www.invitrogen.com/geneart
Experience the newgeneration of gene synthesis and assembly tools
GeneArt

solutions give you predictability, adaptability, and condence. Whether

increasing protein expression rates as much as 100X or assembling genetic constructs
with unprecedented precision, GeneArt

gene synthesis and assembly tools enable you

to break through your most complex challenges.
GE Healthcare
Life Sciences
For every success theres a successor. Its called
progress. So while we knew we had something
special in Amersham

ECL Plus, our Western

blotting team was quietly working on the next
generation of detection reagent. The result: a new
substrate that operates with superior levels of
sensitivity, signal intensity and stability than even
its famous predecessor, making it an excellent
choice for CCD imagers. Welcome to ECL Prime.
Amersham and ECL are trademarks of GE Healthcare companies.
2010 General Electric Company All rights reserved. First published November 2010.
GE Healthcare Bio-Sciences AB, Bjrkgatan 30, 751 84 Uppsala, Sweden.
Find out more about Amersham ECL Prime at
www.gelifesciences.com/eclprime
Comparison of AmershamECL Prime and AmershamECL Plus
detection of -catenin in NIH 3T3 whole cell lysates using different
dilutions of rabbit anti--catenin and HRP-conjugated anti-rabbit IgG.
Thank you ECL

Plus,
youve been great.
Signal intensity sustained at higher antibody dilutions
Ab dilution Amersham ECL Prime Amersham ECL Plus
Primary Secondary 10 g 156 ng 10 g 156 ng
1:3000 1:300 000
1:10 000 1:500 000
CONTENTS Volume 334 Issue 6054
pages 315 & 389
page 304
EDITORIAL
289 The Cost of Doing Nothing
Kathy Caldwell
NEWS OF THE WEEK
294 A roundup of the weeks top stories
NEWS & ANALYSIS
298 Vaccine Trial Meets Modest
Expectations, Buoys Hopes
299 Drug-Screening Program
Looking for a Home
300 Navy Dives Into Program Offering
Cash for Good Scores
301 China Looks to Purge Academia
of Trash Journals
302 Pre-Clovis Mastodon Hunters
Make a Point
>> Report p. 351
303 In Northern Aral Sea, Rebound Comes
With a Big Catch
NEWS FOCUS
304 The Sterile Neutrino:
Fertile Concept or Dead End?
>> Science Podcast
307 Social Science for Pennies
308 Open-Source Ecology Takes Root
Across the World
LETTERS
310 Editorial Expression of Concern
B. Alberts
Martial Arts Research: Prudent Skepticism
J. M. Strayhorn and J. C. Strayhorn
Martial Arts Research: Weak Evidence
J. Mercer
Response
A. Diamond and K. Lee
311 CORRECTIONS AND CLARIFICATIONS
311 TECHNICAL COMMENT ABSTRACTS
BOOKS ET AL.
312 Sybil Exposed
D. Nathan, reviewed by B. Harris
EDUCATION FORUM
313 Rethink Summer Student Research
F. A. Carrero-Martnez
PERSPECTIVES
315 The Strength of Electrical Synapses
S. Hestrin
>> Report p. 389
316 Watery Disks
R. Akeson
>> Report p. 338
317 Antenna-Guided Light
N. Engheta
>> Research Article p. 333
318 Eddies Masquerade as Planetary Waves
D. J. McGillicuddy Jr.
>> Research Article p. 328
320 Up Close with Membrane
Lipid-Protein Complexes
J. Whitelegge
>> Report p. 380
321 Every Bit Counts
P. J. Thomas
>> Report p. 354
REVIEW
323 Globalization, Land Use, and the
Invasion of West Nile Virus
A. M. Kilpatrick
>> Science Podcast
CONTENTS continued >>
www.sciencemag.org SCIENCE VOL 334 21 OCTOBER 2011 281
COVER
Illustration of a generalized form of Snells Law that fully accounts
for light interaction with subwavelength structured materials,
which can produce phase jumps at the interface between media.
To illustrate this effect, origami ribbons are folded into tapered
cylinders (phase jumps at interfaces), and the lines (light rays)
form angles that depend on the degree of taper. Initially diverging
rays converge after passing through two interfaces. See page 333.
Image: Nanfang Yu/Harvard University; Yi Tan and Jinhua Tan
DEPARTMENTS
285 This Week in Science
290 Editors Choice
292 Science Staff
395 New Products
396 Science Careers
CONTENTS
pages 320 & 380
page 373
page 328
RESEARCH ARTICLES
328 The Inuence of Nonlinear Mesoscale
Eddies on Near-Surface Oceanic
Chlorophyll
D. B. Chelton et al.
Large ocean eddies are the cause of some
sea-surface height and chlorophyll anomalies
previously ascribed to Rossby waves.
>> Perspective p. 318
333 Light Propagation with Phase
Discontinuities: Generalized Laws
of Reection and Refraction
N. Yu et al.
Light propagation can be controlled with
plasmonic interfaces that introduce abrupt
phase shifts along the optical path.
>> Perspective p. 317
REPORTS
338 Detection of the Water Reservoir
in a Forming Planetary System
M. R. Hogerheijde et al.
The detection of cold water vapor in a
nearby planet-forming disk suggests that
water ice exists in its outer regions.
>> Perspective p. 316
340 Supramolecular Linear Heterojunction
Composed of Graphite-Like Semiconducting
Nanotubular Segments
W. Zhang et al.
A supramolecular self-assembly approach is
used to make a nanotubular heterojunction.
343 Dynamics of the Reaction of Methane
with Chlorine Atom on an Accurate
Potential Energy Surface
G. Czak and J. M. Bowman
Theory helps explain the counterintuitive
impacts of vibrational excitation in a widely
studied reaction.
347 800,000 Years of Abrupt Climate
Variability
S. Barker et al.
Greenland climate variability for the
past 800,000 years was inferred from
the Antarctic ice-core temperature record.
351 Pre-Clovis Mastodon Hunting
13,800 Years Ago at the
Manis Site, Washington
M. R. Waters et al.
Further dating of the Manis site shows
that people were hunting mastodons in
North America by 14,000 years ago.
>> News story p. 302; Science Podcast
354 Information Transduction Capacity of
Noisy Biochemical Signaling Networks
R. Cheong et al.
Noise limits information transfer through
a single signaling pathway in a single cell
to just one bit.
>> Perspective p. 321
358 ER Tubules Mark Sites of
Mitochondrial Division
J. R. Friedman et al.
Mitochondrial division occurs at positions
where endoplasmic reticulum tubules contact
mitochondria and mediate constriction.
362 Antimicrobial Peptides Keep Insect
Endosymbionts Under Control
F. H. Login et al.
A beetle species synthesizes an antimicrobial
peptide to constrain a bacterial symbiont in
specialized organs.
366 Stochastic Pulse Regulation in
Bacterial Stress Response
J. C. W. Locke et al.
Energy stress activates an alternative sigma
factor in stochastic pulses and modulates
pulse frequency to control activity.
369 Transgenerational Epigenetic Instability
Is a Source of Novel Methylation Variants
R. J. Schmitz et al.
Spontaneous methylation rates that may
affect phenotype in the plant Arabidopsis
are higher than the mutation rate.
373 Computation-Guided Backbone Grafting
of a Discontinuous Motif onto a Protein
Scaffold
M. L. Azoitei et al.
A two-segment HIV epitope grafted into
a scaffold protein maintains high afnity
for a broadly neutralizing antibody.
376 Antagonists Induce a Conformational
Change in cIAP1 That Promotes
Autoubiquitination
E. C. Dueber et al.
Antagonist binding to an apoptosis inhibitor
releases inhibition by promoting dimerization
required for autoubiquitination.
380 Mass Spectrometry of Intact V-Type
ATPases Reveals Bound Lipids and the
Effects of Nucleotide Binding
M. Zhou et al.
The effect of lipids and nucleotides on
the soluble head domain and membrane
base domain is examined in an intact
adenosine triphosphatase.
>> Perspective p. 320
385 Cerebellum Shapes Hippocampal
Spatial Code
C. Rochefort et al.
Cerebellar protein kinase Cdependent
mechanisms process self-motion information
needed for spatial representation and
accurate navigation.
389 Activity-Dependent Long-Term Depression
of Electrical Synapses
J. S. Haas et al.
Paired bursting in coupled neurons depresses
electrical synapses while their asymmetry
increases after unidirectional use.
>> Perspective p. 315
21 OCTOBER 2011 VOL 334 SCIENCE www.sciencemag.org 282
www.sciencemag.org SCIENCE VOL 334 21 OCTOBER 2011 283
CONTENTS
SCIENCEXPRESS
www.sciencexpress.org
Structural Dynamics of a Catalytic Monolayer
Probed by Ultrafast 2D IR Vibrational Echoes
D. E. Rosenfeld et al.
A method to track fast vibrational motion in
solution has been extended to catalytically
important solid/liquid interfaces.
10.1126/science.1211350
Wolbachia Enhance Drosophila Stem Cell
Proliferation and Target the Germline
Stem Cell Niche
E. M. Fast et al.
A bacterial endosymbiont up-regulates mitosis
of Drosophila germline stem cells and blocks
programmed cell death.
10.1126/science.1209609
Endocannabinoid Hydrolysis Generates Brain
Prostaglandins That Promote Neuroinammation
D. K. Nomura et al.
A new tissue-specic pathway for the synthesis
of proinammatory prostaglandins is described.
10.1126/science.1209200
Recent Synchronous Radiation of a Living Fossil
N. S. Nagalingum et al.
Despite their ancient origin, the majority of
extant cycad species radiated within the past
10 million years.
10.1126/science.1209926
Polarization of PAR Proteins by Advective
Triggering of a Pattern-Forming System
N. W. Goehring et al.
Patterning of Caenorhabditis elegans zygotes
involves passive as well as active mechanisms.
10.1126/science.1208619
TECHNICALCOMMENTS
Comment on How Cats Lap: Water Uptake
by Felis catus
M. Nauenberg
Full text at www.sciencemag.org/cgi/content/
full/334/6054/311-b
Response to Comment on How Cats Lap:
Water Uptake by Felis catus
R. Stocker et al.
Full text at www.sciencemag.org/cgi/content/
full/334/6054/311-c
SCIENCENOW
www.sciencenow.org
Highlights From Our Daily News Coverage
Cute TV Chimps May Harm Their Wild Brethren
Seeing dressed-up chimpanzees doesnt make
viewers care more.
http://scim.ag/chimpads
Seaweed With a Deadly Touch
Toxins on the surfaces of common South Pacic
algae are deadly to corals.
http://scim.ag/CSlamphora
Winged Robots Hint at the Origins of Flight
What were primitive wings used for before
bird ancestors could y?
http://scim.ag/robot-ight
SCIENCEONLINE
SCIENCE (ISSN 0036-8075) is published weekly on Friday, except the last
week in December, by the American Association for the Advancement of
Science, 1200 New York Avenue, NW, Washington, DC 20005. Periodicals Mail
postage (publication No. 484460) paid at Washington, DC, and additional mailing
ofces. Copyright 2011 by the American Association for the Advancement of
Science. The title SCIENCE is a registered trademark of the AAAS. Domestic individual
membership and subscription (51 issues): $149 ($74 allocated to subscription).
Domestic institutional subscription (51 issues): $990; Foreign postage extra: Mexico,
Caribbean (surface mail) $55; other countries (air assist delivery) $85. First class,
airmail, student, and emeritus rates on request. Canadian rates with GST available
upon request, GST #1254 88122. Publications Mail Agreement Number 1069624.
Printed in the U.S.A.
Change of address: Allow 4 weeks, giving old and new addresses and 8-digit account
number. Postmaster: Send change of address to AAAS, P.O. Box 96178, Washington,
DC 200906178. Single-copy sales: $10.00 current issue, $15.00 back issue prepaid
includes surface postage; bulk rates on request. Authorization to photocopy
material for internal or personal use under circumstances not falling within the fair
use provisions of the Copyright Act is granted by AAAS to libraries and other users
registered with the Copyright Clearance Center (CCC) Transactional Reporting Service,
provided that $25.00 per article is paid directly to CCC, 222 Rosewood Drive, Danvers,
MA 01923. The identication code for Science is 0036-8075. Science is indexed in the
Readers Guide to Periodical Literature and in several specialized indexes.
SCIENCESIGNALING
www.sciencesignaling.org
The Signal Transduction Knowledge Environment
18 October issue: http://scim.ag/ss101811
RESEARCH ARTICLE: Agrobacterium
Counteracts Host-Induced Degradation
of Its Effector F-Box Protein
S. Magori and V. Citovsky
A plant pathogen prevents degradation of its key
virulence factor in infected host cells.
REVIEW: Structural Basis for Activation and
Inhibition of Class I Phosphoinositide 3-Kinases
O. Vadas et al.
The regulatory interactions between PI3Ks and their
binding partners could be exploited for therapies.
JOURNAL CLUB: -Synuclein Promotes
Neuroprotection Through NF-BMediated
Transcriptional Regulation of Protein Kinase C
R. Aoki and Y. R. Li
-Synuclein may protect dopaminergic neurons
from apoptosis by reducing the abundance
of a kinase involved in cell death.
FUNDING SOURCES
Check out the latest update in opportunities
for funding signaling research and training.
SCIENCETRANSLATIONAL MEDICINE
www.sciencetranslationalmedicine.org
Integrating Medicine and Science
19 October issue: http://scim.ag/stm101911
FOCUS: Containing the Contagion
Treating the Virus That Inspired the Film
B. Lee
RESEARCH ARTICLE: A Neutralizing Human
Monoclonal Antibody Protects African Green
Monkeys from Hendra Virus Challenge
K. N. Bossart et al.
A neutralizing human monoclonal antibody
can fully protect nonhuman primates from
disease after a lethal Hendra virus challenge.
PERSPECTIVE: Hedgehog Rushes to the
Rescue of the Developing Cerebellum
O. Baud and P. Gressens
RESEARCH ARTICLE: A Small-Molecule
Smoothened Agonist Prevents Glucocorticoid-
Induced Neonatal Cerebellar Injury
V. M. Heine et al.
RESEARCH ARTICLE: Preterm Cerebellar Growth
Impairment After Postnatal Exposure to
Glucocorticoids
E. W. Y. Tam et al.
A small molecule that targets the sonic hedgehog
pathway protects against cerebellar injury caused
by glucocorticoids, which may be given to preterm
infants for treating lung disease.
SCIENCECAREERS
www.sciencecareers.org/career_magazine
Free Career Resources for Scientists
Tooling Up: Views on an Interview, Part 1
D. Jensen
Scott Jacksons interview at ABC technologies
ranges from mundane to terrifying.
http://scim.ag/TU_Views_1
Perspective: An Auxiliary (Research) Engine
D. Albert
To attain your goals, it is sometimes necessary
to employ indirect means.
http://scim.ag/Aux_Engine
Content Collection: Careers at Nonprots
E. Pain
Scientists interested in a social or humanitarian cause
can nd diverse research or alternative careers within
nonprot organizations.
http://scim.ag/CC_Nonprots
SCIENCEPODCAST
www.sciencemag.org/multimedia/podcast
Free Weekly Show
On the 21 October Science Podcast: mastodon
hunting before Clovis, globalization and West Nile
virus, the search for the sterile neutrino, and more.
SCIENCEINSIDER
news.sciencemag.org/scienceinsider
Science Policy News and Analysis
Built for
performance.
DNA Ligases from
New England Biolabs.
Why compromise any aspect of performance in your
cloning experiments? DNA ligases from NEB set the
industry standard for function and purity. Continuous
product development and extensive quality controls
ensure success in your ligation reactions.
DNA ligases from NEB building a strong foundation
for your cloning experiments.
SPECIAL OFFER
Purchase NEB 5-alpha Competent Cells
with NEBs Quick Ligation Kit and
receive the item of lesser value for FREE
For more information, visit
www.neb.com/DNAligases
SilverXpress

is a trademark of Invitrogen.
Monarch caterpillar
Maryellen Baker/Botanica
Equivalent amounts of protein were loaded and analyzed by
Coomassie gel and silver stained using SilverXpress

. Marker M
is NEBs Broad Range Protein Marker (NEB #P7702).
Experience extreme purity with
NEBs T4 DNA Ligase
NEB NEB
Lot Lot
M 101 102 A B C D
Supplier
285
EDITED BY STELLA HURTLEY
C
R
E
D
I
T
S

(
T
O
P

T
O

B
O
T
T
O
M
)
:

A
.

H
E
D
D
I

A
N
D

S
.

B
A
L
M
A
N
D
/
B
F
2
I

L
A
B
O
R
A
T
O
R
Y
;

Z
H
A
N
G

E
T

A
L
.
Continued on page 287
www.sciencemag.org SCIENCE VOL 334 21 OCTOBER 2011
Virus Invasion
West Nile virus is spread through mosquitoes to
birds, wildlife, and humans and has established
itself at an astonishingly rapid rate since it was
introduced to North America in 1999. How did
the West Nile virus establish itself so successfully
to the detriment of human and wildlife popula-
tions? Kilpatrick (p. 323) reviews the scenarios
and dynamics that point to the key bird hosts
and the relative predilections of the associated
mosquito vectors to feed on a variety of animals,
including humans.
Making Waves
Patterns of ocean chlorophyll variability from
satellite observations have been attributed to oce-
anic Rossby wavesslow-moving features with
wavelengths of hundreds of kilometers but with
sea surface heights of only centimetersthat
take months or years to cross ocean basins from
the west to the east. Chelton et al. (p. 328, pub-
lished online 15 September; see the Perspective
by McGillicuddy) report that the cause of these
chlorophyll anomalies has been misidentied.
Analysis of 10 years of remotely observed sea sur-
face height elds and concurrent observations of
upper-ocean chlorophyll concentrations suggests
that the dominant mechanism controlling the
development of these anomalies is the horizontal
advection of chlorophyll-rich surface water caused
by the rotational motions of eddies.
Nano-Heterojunction
Self-Assembly
Nanoscale materials can now be syn-
thesized by a wide range of methods,
including self-assembly techniques.
The junction regions between dissimilar materials
often have unusual and desirable properties.
Zhang et al. (p. 340) were able to extend the
self-assembly toolbox to make heterojunctions of
semiconducting nanotubular segments. The con-
joined segments could transport electrical charge
and also increase the lifetime of photogenerated
charge carriers.
Controlling Light
The behavior of light as it propagates through a
material and from one material to another is very
well understood in terms of classical optics. Yu et
al. (p. 333, published online 1 September; see the
cover; see the Perspective by Engheta) now dem-
onstrate a powerful new method to control light
propagation, based on introducing abrupt phase
shifts along the optical path. These phase disconti-
nuities are constructed using plasmonic interfaces
that consist of an optically thin two-dimensional
matrix of optical antennas with subwavelength
separation. The exibility of the technique should
prove useful for developing a wide variety of
small-footprint planar optical components.
Whence the Water Vapor?
Water vapor has been detected in the inner regions
of planet-forming diskswhere terrestrial planets
are created. Using the Heterodyne Instrument
for the Far-Infrared on board the Herschel Space
Observatory, Hogerheijde et al. (p. 338; see the
Perspective by Akeson) now report the detection of
water vapor over the full extent of the disk around
the young star TW Hydrae. In the outer regions of
this planet-forming disk, water vapor could only
originate from icy grains. Thus, the result suggests
the presence of a large reservoir of water ice in the
region where comets and giant planets form.
When Cl and CH
4
Collide
The simplest class of two-body chemical reaction
is the formation of a diatomic molecule. Two
atoms come together and, generally speaking,
the only variable is their relative velocity. Things
get considerably more complicated if you add
another atom to the mix and consider its reaction
with a preformed diatomic. Now, there are rela-
tive spatial orientations, and the diatomic might
be vibrating or rotating. Nonetheless, over the
past half-century or so, chemists have developed
a rm grasp of how these reactions work in detail.
The next frontier will be to understand how an
atom reacts with a polyatomic, which has many
different ways of vibrating. Czak and Bowman
(p. 343) simulated the reactivity of methane with
a chlorine atom, providing a theoretical basis for
a multitude of pivotal experiments on this system.
Polar Connections
The climate records extracted from ice cores re-
covered from the Greenland Ice Sheet are detailed
but relatively short in durationaround 120,000
years. Ice cores from Antarctica, on the other
hand, have lower temporal resolution but extend
back more than 800,000 years. In order to infer
how Greenlands climate may have varied over a
longer interval, Barker et al. (p. 347, published
online 8 September) used the Antarctic tempera-
ture record, data from Chinese speleothems, and
the concept of the bipolar seesaw to produce a
well-dated reconstruction of inferred Greenland
temperature variability. Abrupt shifts in Northern
Hemisphere climate appear to have occurred
throughout the Late Pleistocene, and glacial
terminations may have been linked to oscillations
of the bipolar seesaw.
Control and Cooperation
How do hosts regulate internal symbionts to prevent them from taking over their bodies without
compromising the advantages of the relationship? Login et al. (p. 362) explored the balance
between host innate immune responses and bacterial replication of an endosymbiont in weevils,
an important beetle pest of wheat. A single peptide, coleoptericin-A (ColA), synthesized by the
beetle constrained the bacteria within bacteriocytes and blocked bacterial replication. When the
weevils ColA expression was silenced, the bacteria were able to replicate normally, escape the
bacteriocyte, and spread throughout the insect.
Building upon your voice and our vision,
Roche scientists have developed new
reagents that position reliability and
comprehensive efciency across
common and hard-to-transfect cell lines.
Created to meet the needs of researchers who require
high efciency, ease of use and reproducibility,
X-tremeGENE reagents provide more viable cells with
more physiologically relevant results. Our newest
transfection reagents deliver results-driven performance,
allowing you to focus your efforts on discovery, and
helping you to shape the future. With Roche, you can
have complete condence in the technology behind you.
Transfect with condence.
For life science research only.
Not for use in diagnostic procedures.
X-TREMEGENE is a trademark of Roche.
2011 Roche Diagnostics. All rights reserved.
Roche Diagnostics Corporation
Roche Applied Science
Indianapolis, Indiana
Introducing X-tremeGENE DNA Transfection Reagents
Request your free sample today at
x-tremegene.roche.com
This Week in Science
C
R
E
D
I
T
:

F
R
I
E
D
M
A
N

E
T

A
L
.
Hunting Mastodons
A mastodon skeleton containing an embedded projectile tip was discovered in the late 1970s near
Manis, Washington. It was initially dated as about 14,000 years ago but the age, and whether the bone
containing the projectile was directly associated with the rest of the skeleton, has been questioned.
Waters et al. (p. 351) provide new dates on the fossils that conrm an age of about 14,000 years ago.
The data, together with genetic analyses, show that the skeletal elements are related and also that the
projectile was fashioned from a mastodon bone. This age predates the Clovis culture in North America
and, along with other sites, shows early exploitation of megafauna.
Tumor Necrosis Factor Response
Engineers use information theory to analyze how noise inuences information transfer, for example,
in telephone systems. Cheong et al. (p. 354, published online 15 September; see the Perspective
by Thomas) have now applied such analysis in biological experiments by monitoring the response
of thousands of single mouse broblasts to stimulation with various doses of tumor necrosis factor
(TNF). Signal transmission was surprisingly noisy, meaning that the cells could only really differ-
entiate whether TNF was present or not. Such limitations of a single signaling pathway appear to
be overcome by the cooperation of multiple signaling pathways in networks, or by groups of cells
collectively averaging their response to the same signal.
Mitochondrial Division
Mitochondrial division regulates both the shape and the distribution of the mitochondrial
network, which is important in maintaining cellular health. Friedman et al. (p. 358, pub-
lished online 1 September; see the 14 October Perspective by Rambold and Lippincott-
Schwartz) demonstrate that mitochondria in both yeast and mammalian cells are constricted
and divide at positions where they form stable contact sites with the endoplasmic reticulum.
Not in the Genes
The mechanism, distribution, and function of DNA methylation in plant genomes have been char-
acterized, but the stability of DNA methylation over multiple generations and the rate of change
are less well understood. Schmitz et al. (p. 369, published online 15 September) determined the
methylation status of several previously sequenced Arabidopsis lines, including three ancestral and
ve descendant lines separated by 30 generations. The frequency of DNA methylation changes was
5 orders of magnitude higher than genetic changes. Also, unlike genetic changes that were mostly
random, DNA methylation changes occurred in hotspots.
Improving the Fit
Designing proteins for specic functions often relies on grafting functional groups onto existing
protein scaffolds. Success has been limited because backbone remodeling, which might allow more
complex grafting, has been computationally challenging. Azoitei et al. (p. 373) integrated com-
putational design and directed evolution to enable the manipulation of protein backbone structure
required for transplantation of the backbone and side chains of discontinous functional motifs. They
grafted a two-segment HIV gp120 epitope that is targeted by the cross-neutralizing antibody b12,
onto an unrelated scaffold. The nal design showed high afnity and specicity for b12 and the
complex structurally mimicked the gp120-b12 interaction.
Navigating with the Cerebellum
To navigate in space, animals use two strategies: landmark-assisted or map-navigation, which requires
self-motion cues. Although the cerebellum is known to assist in the coordination of self-motion informa-
tion, its role in spatial navigation is unclear. Rochefort et al. (p. 385) examined whether impairment
of the cerebellum affects the spatial code in the hippocampus, using transgenic mice with a selective
disruption of protein kinase Cdependent plasticity at parallel berPurkinje cells synapses. Although
landmark-assisted navigation was robust, when the mice had to rely on motion-generated cuesfor
example, in the darknavigation suffered.
Continued from page 285
membercentral.aaas.org
105
things you didnt
(and 3 you probably
shouldnt) know
about some of
your most
respected
colleagues.
One more data point on why
you should spend more time
at membercentral.aaas.org.
There you can enjoy a feast
of blogs, videos, webinars,
discounts, and downloads
created by and for the most
insatiable brains around.
www.sciencemag.org SCIENCE VOL 334 21 OCTOBER 2011
Sign Up At:
www.sciencemag.org/webinar
REGISTER NOW!
Few, if any, proteins carry out their function in isolation. More likely is that they interact with one or more other proteins, during
which conformational changes and even physical modications such as phosphorylation may occur. The ability to detect both
the interaction and modication of proteins is critical to elucidating function and dysfunction in cells and tissues, leading
to a deeper understanding of normal as well as diseased states. This webinar will focus on current methodologies available
to study protein-protein interactions, as well as looking at new technologies for characterizing protein binding events in a
noninvasive manner that more closely resembles a cells natural environment, producing more physiologically relevant data.
During the webinar, our expert panelists will:
Explain how protein structure and function can be affected by
posttranslational modications and interactions with diverse partners
Discuss current state-of-the-art methods and technologies for
characterizing protein-protein interactions and modications
Describe how multiplexing can be used to distinguish similar
biomolecular events, such as multiphosphorylated molecules or
multiple interacting partners
Answer your questions live and in real time!
Webinar sponsored by
Brought to you by the
Science/AAAS Custom
Publishing Ofce
Participating Experts
Anton Simeonov, Ph.D.
National Institutes of Health
Bethesda, MD
James Bradner, M.D.
Dana-Farber Cancer Institute
Boston, MA
Mathieu Arcand, Ph.D.
PerkinElmer
Montral, Canada
Detecting Native
Protein Interaction
and Phosphorylation
New Approaches and
Novel Technologies
WEBINAR
November 9, 2011
12 noon Eastern, 9 a.m. Pacic, 5 p.m. GMT
The Cost of Doing Nothing
AS THE U.S. CONGRESS GRAPPLES FOR SOLUTIONS TO THE ECONOMIC CRISIS, IT IS CRITICAL
to recognize that rebuilding and modernizing infrastructure will be a key driver of economic
growth. Recent reports issued by the American Society of Civil Engineers (ASCE),* the
Urban Land Institute, and Building Americas Future have described the deterioration
of the nations energy, water, and transportation infrastructure. This week, ASCE convenes
its annual meeting, gathering international scientists, engineers, policy-makers, and educa-
tors to share sustainable solutions across a broad spectrum of concerns in the natural and
built environment. Key among the discussions will be the role of scientists and engineers
in developing effective public policy, helping to produce an infrastructure that incorporates
new materials, technologies, and strategies to improve environmental and social well-being.
When the ASCE issued its 2009 Report Card for Americas Infrastructure, it gave the
cumulative grade of D to the condition and performance of 15 of
the countrys infrastructure systems.* Among the worst were roads
and drinking water. The United States not only loses about seven bil-
lion gallons of clean drinking water every day due to leaking water
systems, but pipe failures and resulting oods have collapsed roads,
destroyed homes, and endangered people. It would require an esti-
mated $2.2 trillion over 5 years to raise the grade for all 15 infra-
structure systems to an acceptable level. Sadly, the situation has not
changed since the report was published. Earlier this year, the ASCEs
report Failure to Act, The Economic Impact of Current Investment
Trends in Surface Transportation Infrastructure determined that the
decient surface transportation infrastructure alone will cost U.S.
businesses an added $430 billion (cumulative to 2020) in transpor-
tation costs. By 2020, it is projected that exports will be $28 billion
lower, 70,000 jobs will be lost, households will lose more than $7000
in personal income, and the countrys gross domestic product will take a hit of $897 billion.
Businesses will need to divert increasing portions of income to pay for transportation delays,
wasting money that could instead be invested in innovation. Nearly all sectors will suffer, but
those associated with technology and innovation would probably be the hardest hit.
To meet the many infrastructure challenges, more nancing is needed. Now that the
American Jobs Act has failed to pass Congress, there is discussion of breaking the bill into
pieces that should be easier to pass. The proposed act includes $50 billion to modernize road,
rail, and air transportation systems, and it would establish a National Infrastructure Bank to
leverage public and private capital toward these endeavors. This level of priority and invest-
ment is needed, or the United States will continue its downward slide. Indeed, this years
report from the Urban Land Institute warns that the United States has fallen behind Brazil,
China, and India in bolstering transportation, water, and sewage infrastructure.
Promoting a more sustainable and resilient infrastructure must also be part of this con-
versation. Improved design and construction standards to withstand extreme conditions will
require further R&D. Climate change and environmental preservation also require innova-
tive infrastructure designs. Research is needed to determine the best ways to expand power
generation and transmission. And the growing demands for information technology mean
that underground utilities must be carefully planned.
Infrastructure investments provide an opportunity to improve the economy in the short term
by creating jobs, while also driving the long-term growth needed to compete in the global mar-
ketplace. Although repairing and modernizing the countrys infrastructure may seem daunting
in lean times, the cost of doing nothing will be exponentially greater.
10.1126/science.1214039
Kathy Caldwell
289
EDITORIAL
C
R
E
D
I
T
S
:

(
L
E
F
T
)

A
S
C
E
;

(
R
I
G
H
T
)

F
O
T
O
S
E
A
R
C
H
Kathy Caldwell is
president of the
American Society of
Civil Engineers. E-mail:
kcaldwell@asce.org.
www.sciencemag.org SCIENCE VOL 334 21 OCTOBER 2011
*www.asce.org/reportcard. www.uli.org/~/media/Documents/ResearchAndPublications/Reports/Infrastructure/
Infrastructure2011.ashx. www.bafuture.org/report. www.asce.org/uploadedFiles/Infrastructure/Report_Card/
ASCE-FailureToActFinal.pdf.
290
EDITED BY KRISTEN MUELLER AND JAKE YESTON
EDITORSCHOICE
C
R
E
D
I
T
S

(
T
O
P

T
O

B
O
T
T
O
M
)
:

J
A
M
E
S

K
I
N
G
-
H
O
L
M
E
S
/
S
C
I
E
N
C
E

P
H
O
T
O

L
I
B
R
A
R
Y
;

F
O
T
O
S
E
A
R
C
H
21 OCTOBER 2011 VOL 334 SCIENCE www.sciencemag.org
O C E A N S C I E N C E
What Keeps the Storms Away?
The number of hurricanes that develop over the
Atlantic Ocean each year and the number that
make landfall in North and Central America are
two distinct quantities. The difference between
the two has great practical consequence, as
hurricanes that remain offshore cause few
deaths and do little damage to human prop-
erty or infrastructure. What controls hurricane
tracks, then? Wang et al. looked at sea surface
temperature data from 1970 to 2009 and found
that the size and location of the Atlantic warm
pool help to steer hurricanes by inuencing
both where over the Atlantic they form and to
what extent ensuing atmospheric circulation
patterns push the storms away from the eastern
seaboard of the United States. When the Atlantic
warm pool is large, storms form more to the
east, further from potential landfall, and the
winds along their paths blow more strongly
toward the northeast, also reducing the chance
that the storms ultimately reach a vulnerable
coast. Although these are not the only factors
that control hurricane tracks, consideration of
the sea surface temperature elds of the North
Atlantic Ocean may help improve forecasts of
potential hurricane dangers. HJS
Geophys. Res. Lett. 38, L19702 (2011).
P L A N E T A R Y S C I E N C E
Why No Clay Up North?
The surface of Mars can be divided into two
major regions: the northern lowlands and the
southern highlands. The lowlands, covering
around 1/3 of the planet, are thought to have
once been the site of an ancient, great northern
ocean. However, this hypothesis
is at odds with the record
of the presence of
clay deposits. These
sediments, whose
formation requires
the presence of
liquid water,
are widespread
in the southern
highlands but very
scarcely distributed
in the northern low-
lands. Using a climate
model, Fairn et al. deter-
mined the surface temperatures
on early Mars, assuming a southern super-
continent and a northern ocean. The model tem-
peratures imply that the northern ocean would
H Y D R O L O G Y
Rolling Down the River
All rivers naturally move loads of sediments,
from coarse sand grains rolling along river-
beds to tiny clay and silt particles carried in
suspension. When sediment load gets too
high though, either naturally or from hu-
man activities, biodiversity suffers and water
quality deteriorates. Identifying the sources of
increased sediment loads, which can vary with
such regional factors as land use and precipita-
have had to be a glacial ocean similar to the
seas in Earths polar regions. Calculations of the
rate of clay formation at subzero temperatures
support the lack of clays in the northern low-
lands, because their formation would have been
inhibited at those cold temperatures. Moreover,
the presence of glaciers surrounding the north-
ern ocean would have limited the transport of
continental sediments into the ocean, as is the
case in the Arctic and Antarctic coastal regions
of Earth. MJC
Nat. Geosci. 4, 667 (2011).
E C O N O MI C S
Resource Investment
As Newton famously noted, researchers stand on the shoulders of giants, building on accu-
mulated knowledge. But the mere production of knowledge does not ensure its use by others;
societal benet depends also on mechanisms for storing and accessing knowledge. Research-
ers have sought to understand how different institutions and policies can promote knowledge
use and impact. To explore impacts of institutional resources in the life sciences, Furman and
Stern studied the American Type Culture Collection (ATCC). Among the worlds largest bio re-
source centers, ATCC maintains and distributes a vast collection of cell lines and microbiology
cultures. Because each specimen deposited in ATCC is accompanied by an initial character-
ization in a journal article, bibliometric analyses of article citations provided tools to assess
impacts of ATCC. Besides comparing articles that did and did not link to ATCC specimens, the
authors also analyzed the timing of citation boosts, because the deposition of specimens at
ATCC often did not occur until sometime after the initial journal article describing the speci-
men. The ATCC-deposit citation boost ranged from 57 to 135%, was higher for articles in less
prestigious journals, and was concentrated on follow-up research into more complex subject
matter. A rough approximation of cost per citation suggested that funders might consider
increased investments in ensuring access to existing research rather than focusing so much
on new research. BW
Am. Econ. Rev. 101, 1933 (2011).
EDITORSCHOICE
C
R
E
D
I
T
:

N
I
K
I
T
A
S

A
N
D

L
E
C
U
I
T
www.sciencemag.org SCIENCE VOL 334 21 OCTOBER 2011
tion, is critical for implementing remediation
strategies. As a case study for determining the
mechanism of sediment transport on the scale of
several watersheds, Belmont et al. integrated a
number of data sourcesincluding geochemi-
cal tracers, hydrologic eld measurements, and
remote sensingfrom tributaries or lakes in
Minnesota along North Americas largest river,
the Mississippi. Over the past 150 years, not
only has the amount of ne-grained sediment
increased by a factor of 10, but the sources have
also changed. Historically, upland erosion of soil
contributed most of the sediment load; however,
the new data suggest that up to 70% of the
sediment comes from the erosion of riverbanks
and ravines themselves. The shift is probably
a function of both natural and anthropogenic
activity, including increased precipitation and
extensive modication of drainage networks for
agricultural purposes. NW
Environ. Sci. Technol. 45, 10.1021/es2019109
(2011).
MI C R O B I O L O G Y
Rapid Transport
E-cadherin is a species-specic receptor for the
foodborne pathogen Listeria monocytogenes
but it is located out of reach beneath the tight
junctions formed between gut epithelial cells.
Does the dynamic nature of the intestinal epi-
thelium, which is being remodeled constantly,
with cells being shed and mucus secreted,
allow for E-cadherin to be accessed? Taking a
step back from molecu-
lar studies of pathogen
cell invasion, Nikitas et
al. watched how Listeria
invades the body, using
humanized mice and
two-photon and confocal
microscopy. They found
that E-cadherin is not
perpetually out of sight
but becomes exposed
to the intestinal lumen
when cells are extruded
and cell junctions are disrupted by contracting
goblet cells or folds in the villi. Once inside the
cell, the bacteria have no need for any other
virulence factors, neither listeriolysin-O nor ActA
(which polymerizes the cells actin). All that is
required is the bacterial surface protein InlA
for rapid apico-basal translocation mediated by
the cells microtubules and exocytosis into the
lamina propria. Thirty minutes after invasion,
and Listeria had entered the spleen undetected
by immune surveillance and a systemic infection
was established. CA
J. Exp.Med. 208, 10.1084/jem.20110560 (2011).
P S Y C H O L O G Y
Thats Not Yours!
Much ink has been spilled in arguments about
what it is that children have learned when
they begin to grasp the possibility that other
peoples beliefs can differ from their own. But
what do children comprehend of other peoples
rights, such as the ownership of property?
Rossano et al. describe experimental results
indicating that 3-year-old children exhibit a
more sophisticated understanding of the rights
conferred by ownershipin this instance, the
disposal of a cap or scarfthan 2-year-olds.
Children of both ages complained when their
own hat was thrown away by a puppet, and they
did not protest when the puppet threw away his
own article of clothing; the key distinction was
that older kids registered a normative objection
when the puppet discarded a hat belonging
to a third party (the experimenter). In their
introduction to an edited collection, Friedman
and Ross enumerate the reasons why research
on the developmental origins of ownership will
yield ndings of interest. GJC
Cognition 121, 219 (2011); New Dir. Child Ado-
lesc. Dev. 132, 1 (2011).
C E L L S I G N A L I N G
Enlightening the Load
If you have driven a car with a manual transmis-
sion, you are aware that the response to the
throttle is quite different when the drive train is
connected to a load (when the
clutch is engaged and the engine
drives the wheels) than when it is
not (when the engine spins freely
with the clutch disengaged).
Jiang et al. explored whether a
similar concept of load applies
to biochemical signaling systems;
that is, whether the dynamic
properties of a signaling mecha-
nism were altered in the presence
or absence of substrate molecules
that are targets of the system.
Combined experiments and mathematical
modeling showed that the presence of substrate
could alter the response time of the system,
increasing it when one of the enzymes in the
signaling system was operating at a maximal
rate (saturated) but decreasing it when the
enzymes were operating in a linear manner. The
authors discuss how such effects of downstream
targets on the responsiveness of signaling
systems might be used to design appropriate
responses when modifying biological systems or
designing synthetic ones. LBR
Sci. Sig. 4, ra67 (2011).
AAAS is here
bringing educational
infrastructure to the
developing world.
AAAS is helping the Rwandan
government rebuild its educa-
tional infrastructure as a way
to help drive economic growth
and development. By providing
materials such as the Project
2061 Atlas of Science Literacy,
lesson plans from Science
NetLinks, and access to Science
digital libraries, AAAS is helping
the people of Rwanda work
toward a future built around
science and technology. As a
AAAS member your dues
support these efforts. If youre
not yet a AAAS member, join us.
Together we can make a difference.
To learn more, visit
aaas.org/plusyou/rwanda
292 21 OCTOBER 2011 VOL 334 SCIENCE www.sciencemag.org
www.sciencemag.org
SENIOR EDITORIAL BOARD
A. Paul Alivisatos, Lawrence Berkeley Nat'l. Laboratory
Cori Bargmann, The Rockefeller Univ.
Ernst Fehr, Univ. of Zurich
Richard Losick, Harvard Univ.
Michael S. Turner, University of Chicago
BOARD OF REVIEWING EDITORS
Adriano Aguzzi, Univ. Hospital Zrich
Takuzo Aida, Univ. of Tokyo
Sonia Altizer, Univ. of Georgia
Sebastian Amigorena, Institut Curie
Angelika Amon, MIT
Kathryn Anderson, Memorial Sloan-Kettering Cancer Center
Siv G. E. Andersson, Uppsala Univ.
Peter Andolfatto, Princeton Univ.
Meinrat O. Andreae, Max Planck Inst., Mainz
John A. Bargh, Yale Univ.
Ben Barres, Stanford Medical School
Jordi Bascompte, Estacin Biolgica de Doana, CSIC
Facundo Batista, London Research Inst.
Ray H. Baughman, Univ. of Texas, Dallas
David Baum, Univ. of Wisconsin
Yasmine Belkaid, NIAID, NIH
Philip Benfey, Duke Univ.
Stephen J. Benkovic, Penn State Univ.
Gregory C. Beroza, Stanford Univ.
Peer Bork, EMBL
Bernard Bourdon, Ecole Normale Superieure de Lyon
Ian Boyd, Univ. of St. Andrews
Paul M. Brakeeld, Univ. of Cambridge
Christian Bchel, Universittsklinikum Hamburg-Eppendorf
Joseph A. Burns, Cornell Univ.
William P. Butz, Population Reference Bureau
Gyorgy Buzsaki, Rutgers Univ.
Mats Carlsson, Univ. of Oslo
Mildred Cho, Stanford Univ.
David Clapham, Childrens Hospital, Boston
David Clary, Univ. of Oxford
J. M. Claverie, CNRS, Marseille
Jonathan D. Cohen, Princeton Univ.
Robert Cook-Deegan, Duke Univ.
Alan Cowman, Walter & Eliza Hall Inst.
Robert H. Crabtree, Yale Univ.
Wolfgang Cramer, Medit. Inst. for Ecology & Paleoecology
F. Fleming Crim, Univ. of Wisconsin
Jeff L. Dangl, Univ. of North Carolina
Tom Daniel, Univ. of Washington
Stanislas Dehaene, Collge de France
Emmanouil T. Dermitzakis, Univ. of Geneva Medical School
Robert Desimone, MIT
Claude Desplan, New York Univ.
Ap Dijksterhuis, Radboud Univ. of Nijmegen
Dennis Discher, Univ. of Pennsylvania
Jennifer A. Doudna, Univ. of California, Berkeley
Julian Downward, Cancer Research UK
Bruce Dunn, Univ. of California, Los Angeles
Christopher Dye, WHO
David Ehrhardt, Carnegie Inst. of Washington
Michael B. Elowitz, Calif. Inst. of Technology
Tim Elston, Univ. of North Carolina at Chapel Hill
Gerhard Ertl, Fritz-Haber-Institut, Berlin
Barry Everitt, Univ. of Cambridge
Paul G. Falkowski, Rutgers Univ.
Ernst Fehr, Univ. of Zurich
Tom Fenchel, Univ. of Copenhagen
Alain Fischer, INSERM
Wulfram Gerstner, EPFL Lausanne
Karl-Heinz Glassmeier, TU Braunschweig
Diane Grifn, Johns Hopkins Bloomberg School of
Public Health
Elizabeth Grove, Univ. of Chicago
Taekjip Ha, Univ. of Illinois at Urbana-Champaign
Christian Haass, Ludwig Maximilians Univ.
Steven Hahn, Fred Hutchinson Cancer Research Center
Gregory J. Hannon, Cold Spring Harbor Lab.
Martin Heimann, Max Planck Inst., Jena
Isaac Held, NOAA
James A. Hendler, Rensselaer Polytechnic Inst.
Janet G. Hering, Swiss Fed. Inst. of Aquatic
Science & Technology
Ray Hilborn, Univ. of Washington
Michael E. Himmel, National Renewable Energy Lab.
Kai-Uwe Hinrichs, Univ. of Bremen
Kei Hirose, Tokyo Inst. of Technology
David Hodell, Univ. of Cambridge
Ove Hoegh-Guldberg, Univ. of Queensland
David Holden, Imperial College
Lora Hooper, UT Southwestern Medical Ctr at Dallas
Jeffrey A. Hubbell, EPFL Lausanne
Steven Jacobsen, Univ. of California, Los Angeles
Kai Johnsson, EPFL Lausanne
Peter Jonas, Universitt Freiburg
William Kaelin, Dana-Farber Cancer Inst.
Barbara B. Kahn, Harvard Medical School
Daniel Kahne, Harvard Univ.
Bernhard Keimer, Max Planck Inst., Stuttgart
Joel Kingsolver, Univ. of North Carolina at Chapel Hill
Robert Kingston, Harvard Medical School
Alberto R. Kornblihtt, Univ. of Buenos Aires
Leonid Kruglyak, Princeton Univ.
Mitchell A. Lazar, Univ. of Pennsylvania
David Lazer, Harvard Univ.
Virginia Lee, Univ. of Pennsylvania
Ottoline Leyser, Cambridge Univ.
Olle Lindvall, Univ. Hospital, Lund
Marcia C. Linn, Univ. of California, Berkeley
John Lis, Cornell Univ.
Jianguo Liu, Michigan State Univ.
Richard Losick, Harvard Univ.
Jonathan Losos, Harvard Univ.
Ke Lu, Chinese Acad. of Sciences
Laura Machesky, CRUK Beatson Inst. for Cancer Research
Andrew P. MacKenzie, Univ. of St Andrews
Anne Magurran, Univ. of St Andrews
Oscar Marin, CSIC & Univ. Miguel Hernndez
Charles Marshall, Univ. of California, Berkeley
Martin M. Matzuk, Baylor College of Medicine
Graham Medley, Univ. of Warwick
Yasushi Miyashita, Univ. of Tokyo
Richard Morris, Univ. of Edinburgh
Edvard Moser, Norwegian Univ. of Science and Technology
Sean Munro, MRC Lab. of Molecular Biology
Thomas Murray, The Hastings Center
Naoto Nagaosa, Univ. of Tokyo
James Nelson, Stanford Univ. School of Med.
Timothy W. Nilsen, Case Western Reserve Univ.
Pr Nordlund, Karolinska Inst.
Helga Nowotny, European Research Advisory Board
Luke O'Neill, Trinity College, Dublin
Stuart H. Orkin, Dana-Farber Cancer Inst.
Christine Ortiz, MIT
Elinor Ostrom, Indiana Univ.
Andrew Oswald, Univ. of Warwick
Jane Parker, Max-Planck Inst. of Plant Breeding Research
Donald R. Paul, Univ. of Texas at Austin
P. David Pearson, Univ. of California, Berkeley
Reginald M. Penner, Univ. of California, Irvine
John H. J. Petrini, Memorial Sloan-Kettering Cancer Center
Simon Phillpot, Univ. of Florida
Philippe Poulin, CNRS
Colin Renfrew, Univ. of Cambridge
Trevor Robbins, Univ. of Cambridge
Barbara A. Romanowicz, Univ. of California, Berkeley
Jens Rostrup-Nielsen, Haldor Topsoe
Edward M. Rubin, Lawrence Berkeley National Lab
Mike Ryan, Univ. of Texas, Austin
Shimon Sakaguchi, Kyoto Univ.
Miquel Salmeron, Lawrence Berkeley National Lab
Jrgen Sandkhler, Medical Univ. of Vienna
Randy Seeley, Univ. of Cincinnati
Vladimir Shalaev, Purdue Univ.
Joseph Silk, Univ. of Oxford
Denis Simon, Univ. of Oregon
Alison Smith, John Innes Centre
Davor Solter, Inst. of Medical Biology, Singapore
John Speakman, Univ. of Aberdeen
Allan C. Spradling, Carnegie Institution of Washington
Jonathan Sprent, Garvan Inst. of Medical Research
Elsbeth Stern, ETH Zrich
Ira Tabas, Columbia Univ.
Yoshiko Takahashi, Nara Inst. of Science and Technology
John Thomas, Duke Univ.
Herbert Virgin, Washington Univ.
Bert Vogelstein, Johns Hopkins Univ.
Cynthia Volkert, Univ. of Gottingen
Bruce D. Walker, Harvard Medical School
Douglas Wallace, Leibniz Inst. of Marine Sciences
Ian Walmsley, Univ. of Oxford
David A. Wardle, Swedish Univ. of Agric Sciences
David Waxman, Fudan Univ.
Detlef Weigel, Max Planck Inst., Tbingen
Jonathan Weissman, Univ. of California, San Francisco
Sue Wessler, Univ. of California, Riverside
Ian A. Wilson, The Scripps Res. Inst.
Timothy D. Wilson, Univ. of Virginia
Jan Zaanen, Leiden Univ.
Kenneth Zaret, Univ. of Penn. School of Medicine
Mayana Zatz, University of Sao Paolo
Jonathan Zehr, Univ. of California, Santa Cruz
Huda Zoghbi, Baylor College of Medicine
Maria Zuber, MIT
BOOK REVIEW BOARD
John Aldrich, Duke Univ.
David Bloom, Harvard Univ.
Angela Creager, Princeton Univ.
Richard Shweder, Univ. of Chicago
Ed Wasserman, DuPont
Lewis Wolpert, Univ. College London
1200 New York Avenue, NW
Washington, DC 20005
Editorial: 202-326-6550, FAX 202-289-7562
News: 202-326-6581, FAX 202-371-9227
Bateman House, 82-88 Hills Road
Cambridge, UK CB2 1LQ
+44 (0) 1223 326500, FAX +44 (0) 1223 326501
SUBSCRIPTION SERVICES For change of address, missing issues, new
orders and renewals, and payment questions: 866-434-AAAS (2227)
or 202-326-6417, FAX 202-842-1065. Mailing addresses: AAAS, P.O.
Box 96178, Washington, DC 20090-6178 or AAAS Member Services,
1200 New York Avenue, NW, Washington, DC 20005
INSTITUTIONAL SITE LICENSES please call 202-326-6755 for any
questions or information
REPRINTS: Author Inquiries 800-635-7181
Commercial Inquiries 803-359-4578
PERMISSIONS 202-326-7074, FAX 202-682-0816
MEMBER BENEFITS AAAS/Barnes&Noble.com bookstore www.aaas.org/bn;
AAAS Online Store www.apisource.com/aaas/ code MKB6; AAAS
Travels: Betchart Expeditions 800-252-4910; Apple Store www.
apple.com/eppstore/aaas; Bank of America MasterCard 1-800-833-6262
priority code FAA3YU; Cold Spring Harbor Laboratory Press
Publications www.cshlpress.com/affiliates/aaas.htm; GEICO Auto
Insurance www.geico.com/landingpage/go51.htm?logo=17624;
Hertz 800-654-2200 CDP#343457; Office Depot https://bsd.
ofcedepot.com/portalLogin.do; Seabury & Smith Life Insurance 800-
424-9883; Subaru VIP Program 202-326-6417; VIP Moving Services
www.vipmayower.com/domestic/index.html; Other Benets: AAAS
Member Services 202-326-6417 or www.aaasmember.org.
science_editors@aaas.org (for general editorial queries)
science_letters@aaas.org (for queries about letters)
science_reviews@aaas.org (for returning manuscript reviews)
science_bookrevs@aaas.org (for book review queries)
Published by the American Association for the Advancement of Science
(AAAS), Science serves its readers as a forum for the presentation and
discussion of important issues related to the advancement of science,
including the presentation of minority or conflicting points of view,
rather than by publishing only material on which a consensus has been
reached. Accordingly, all articles published in Scienceincluding
editorials, news and comment, and book reviewsare signed and reect
the individual views of the authors and not ofcial points of view adopted
by AAAS or the institutions with which the authors are afliated.
AAAS was founded in 1848 and incorporated in 1874. Its mission is to
advance science, engineering, and innovation throughout the world for
the benet of all people. The goals of the association are to: enhance
communication among scientists, engineers, and the public; promote and
defend the integrity of science and its use; strengthen support for the
science and technology enterprise; provide a voice for science on societal
issues; promote the responsible use of science in public policy; strengthen
and diversify the science and technology workforce; foster education in
science and technology for everyone; increase public engagement with
science and technology; and advance international cooperation in science.
INFORMATION FOR AUTHORS
See pages 784 and 785 of the 11 February 2011 issue or
access www.sciencemag.org/about/authors
EDITOR-IN-CHIEF Bruce Alberts
EXECUTIVE EDITOR NEWS EDITOR
Monica M. Bradford Colin Norman
MANAGING EDITOR, RESEARCH JOURNALS Katrina L. Kelner
DEPUTY EDITORS R. Brooks Hanson, Barbara R. Jasny, Andrew
M. Sugden
EDITORIAL SENIOR EDITORS/COMMENTARY Lisa D. Chong, Brad Wible; SENIOR
EDITORS Gilbert J. Chin, Pamela J. Hines, Paula A. Kiberstis (Boston), Marc
S. Lavine (Toronto), Beverly A. Purnell, L. Bryan Ray, Guy Riddihough,
H. Jesse Smith, Phillip D. Szuromi (Tennessee), Valda Vinson, Jake
S. Yeston, Laura M. Zahn (San Diego); ASSOCIATE EDITORS Kristen L.
Mueller, Jelena Stajic, Sacha Vignieri, Nicholas S. Wigginton; BOOK REVIEW
EDITOR Sherman J. Suter; ASSOCIATE LETTERS EDITOR Jennifer Sills; EDITORIAL
MANAGER Cara Tate; SENIOR COPY EDITORS Jeffrey E. Cook, Cynthia Howe,
Harry Jach, Lauren Kmec, Barbara P. Ordway, Trista Wagoner; COPY EDITOR
Chris Filiatreau; SENIOR EDITORIAL COORDINATORS Carolyn Kyle, Beverly
Shields; EDITORIAL COORDINATORS Joi S. Granger, Anita Wynn; PUBLICATIONS
ASSISTANTS Ramatoulaye Diop, Aneera Dobbins, Jeffrey Hearn,
Michael Hicks, Lisa Johnson, Scott Miller, Jerry Richardson,
Teresa R. Sakon, Brian White; EDITORIAL ASSISTANT Patricia M. Moore; EXECUTIVE
EDITORIAL ASSISTANT Yolanda O'Bannon (San Francisco); EXECUTIVE ASSISTANT
Alison Crawford; ADMINISTRATIVE SUPPORT Maryrose Madrid; EDITORIAL FELLOW
Melissa R. McCartney
EDITORIAL DIRECTOR, WEB AND NEW MEDIA Stewart Wills; RESEARCH
ASSOCIATE Corinna Cohn; WEB DEVELOPMENT MANAGER Martyn Green;
WEB DEVELOPER Andrew Whitesell; INTERNS Andrew Green, Kerry Klein
NEWS DEPUTY NEWS EDITORS Robert Coontz, David Grimm (Online), Eliot
Marshall, Jeffrey Mervis, Leslie Roberts, John Travis; CONTRIBUTING EDITORS
Elizabeth Culotta, Polly Shulman; NEWS WRITERS Yudhijit Bhattacharjee,
Adrian Cho, Jennifer Couzin-Frankel, Carolyn Gramling, Jocelyn Kaiser,
Richard A. Kerr, David Malakoff, Greg Miller, Elizabeth Pennisi,
Robert F. Service (Pacific NW), Erik Stokstad; WEB DEVELOPER Daniel
Berger; INTERN Daniel Strain; CONTRIBUTING CORRESPONDENTS Jon Cohen (San
Diego, CA), Daniel Ferber, Ann Gibbons, Sam Kean, Andrew Lawler,
Mitch Leslie, Charles C. Mann, Virginia Morell, Gary Taubes; COPY EDITORS
Linda B. Felaco, Melvin Gatling, Melissa Raimondi; ADMINISTRATIVE SUPPORT
Scherraine Mack; BUREAUS San Diego, CA: 760-942-3252, FAX 760-942-
4979; Pacic Northwest: 503-963-1940
PRODUCTION DIRECTOR Wendy K. Shank; ASSISTANT MANAGER Rebecca
Doshi; SENIOR SPECIALISTS Steve Forrester, Chris Redwood, Anthony
Rosen; PREFLIGHT DIRECTOR David M. Tompkins; MANAGER Marcus Spiegler;
SPECIALISTS Jason Hillman, Tara Kelly
ART DIRECTOR Yael Fitzpatrick; ASSOCIATE ART DIRECTOR Laura Creveling;
SENIOR ILLUSTRATORS Chris Bickel, Katharine Sutliff; ILLUSTRATOR Yana
Hammond; SENIOR ART ASSOCIATES Holly Bishop, Preston Huey,
Nayomi Kevitiyagala, Matthew Twombly; ART ASSOCIATE Kay Engman;
PHOTO EDITOR Leslie Blizard
SCIENCE INTERNATIONAL
EUROPE (science@science-int.co.uk) EDITORIAL: INTERNATIONAL MANAGING
EDITOR Andrew M. Sugden; SENIOR EDITOR/COMMENTARY Julia Fahrenkamp-
Uppenbrink; SENIOR EDITORS Caroline Ash, Stella M. Hurtley, Ian S. Osborne,
Peter Stern; ASSOCIATE EDITOR Maria Cruz; LOCUM EDITOR Helen Pickersgill;
EDITORIAL SUPPORT Samantha Hogg, Alice Whaley; ADMINISTRATIVE
SUPPORT John Cannell, Janet Clements, Louise Hartwell; NEWS: DEPUTY
NEWS EDITOR, U.K. Daniel Clery; NEWSWRITER Sara Reardon; CONTRIBUTING
EDITOR, EUROPE Martin Enserink; CONTRIBUTING CORRESPONDENTS Michael Balter
(Paris), John Bohannon (Vienna), Gretchen Vogel (Berlin)
ASIA Japan Ofce: Asca Corporation, Tomoko Furusawa, Rustic Bldg. 7F,
77 Tenjin-cho, Shinjuku-ku, Tokyo 162-0808, Japan; +81 3
6802 4616, FAX +81 3 6802 4615, inquiry@sciencemag.jp;
ASIA NEWS EDITOR Richard Stone (Beijing: rstone@aaas.org); CONTRIBUTING
CORRESPONDENTS Dennis Normile [Japan: +81 (0) 3 3391 0630, FAX
+81 (0) 3 5936 3531; dnormile@gol.com]; Hao Xin [China:
cindyhao@gmail.com]; Mara Hvistendahl [China: mara@
marahvistendahl.com]; Pallava Bagla [South Asia: +91 (0) 11 2271
2896; pbagla@vsnl.com]
EXECUTIVE PUBLISHER Alan I. Leshner
PUBLISHER Beth Rosner
FULFILLMENT SYSTEMS AND OPERATIONS (membership@aaas.org); CUSTOMER
SERVICE SUPERVISOR Pat Butler; SPECIALISTS LaVonda Crawford, Vicki
Linton, April Marshall; DATA ENTRY SPECIALISTS Shirlene Hall, Tarrika Hill,
William Jones
BUSINESS OPERATIONS AND ADMINISTRATION DIRECTOR Deborah Rivera-
Wienhold; BUSINESS SYSTEMS AND FINANCIAL ANALYSIS DIRECTOR Randy Yi;
MANAGER, FULFILLMENT SYSTEMS Frits Buningh; SYSTEMS ANALYST Nicole
Mehmedovich; MANAGER, BUSINESS ANALYSIS Eric Knott; MANAGER,
BUSINESS OPERATIONS Jessica Tierney; BUSINESS ANALYSTS Priti Pamnani,
Celeste Troxler; Christine Wehrli; RIGHTS AND PERMISSIONS:
ADMI NI STRATOR Emilie David; ASSOCI ATE Elizabeth Sandler;
MARKETING DIRECTOR Ian King; MARKETING MANAGERS Allison Pritchard, Alison
Chandler, Julianne Wielga, Samantha Smith; MARKETING ASSOCIATES
Aimee Aponte, Mary Ellen Crowley; SENIOR MARKETING EXECUTIVE Jennifer
Reeves; DIRECTOR, SITE LICENSING Tom Ryan; DIRECTOR, CORPORATE RELATIONS
Eileen Bernadette Moran; SENIOR PUBLISHER RELATIONS SPECIALIST
Kiki Forsythe; PUBLISHER RELATIONS MANAGER Catherine Holland; PUBLISHER
RELATIONS, EASTERN REGION Phillip Smith; PUBLISHER RELATIONS, WESTERN REGION
Ryan Rexroth; CUSTOMER RELATIONS MANAGER Iquo Edim; MARKETING MAN-
AGER Christina Schlecht; MARKETING ASSOCIATE Chad Johnson; ELEC-
TRONIC MEDIA: DIRECTOR Lizabeth Harman; ASSISTANT MANAGER Lisa
Stanford; SENIOR PRODUCTION SPECIALIST Ryan Atkins; PRODUCTION SPE-
CIALISTS Antoinette Hodal, Nichele Johnston, Kimberly Oster; DIREC-
TOR, WEB AND NEW MEDIA Will Collins; PROJECT MANAGER Trista Snyder;
SENIOR PRODUCTION SPECIALIST Christopher Coleman; COMPUTER SPECIALISTS
Walter Jones, Kai Zhang
ADVERTISING DIRECTOR, WORLDWIDE AD SALES Bill Moran
COMMERCIAL EDITOR Sean Sanders: 202-326-6430
ASSISTANT COMMERCIAL EDITOR Tianna Hicklin 202-326-6463
PRODUCT ( sci ence_advert i si ng@aaas. org) ; MI DWEST Ri ck
Bongiovanni: 330-405-7080, FAX 330-405-7081; EAST COAST/
E. CANADA Laurie Faraday: 508-747-9395, FAX 617-507-8189;
WEST COAST/W. CANADA Lynne Stickrod: 415-931-9782, FAX 415-520-
6940; UK/EUROPE/ASIA Roger Goncalves: TEL/FAX +41 43 243 1358;
JAPAN ASCA Corporation, Makiko Hara: +81 (0) 3 6802 4616, FAX
+81 (0) 3 6802 4615; ads@sciencemag.jp; CHINA/TAIWAN Ruolei Wu:
+86 1367 1015 294 rwu@aaas.org
WORLDWIDE ASSOCIATE DIRECTOR OF SCIENCE CAREERS Tracy Holmes: +44 (0)
1223 326525, FAX +44 (0) 1223 326532
CLASSIFIED (advertise@sciencecareers.org); U.S.: MIDWEST/WEST COAST/
SOUTH CENTRAL/CANADA Tina Burks: 202-326-6577; EAST COAST/INDUSTRY
Elizabeth Early: 202-326-6578; SALES ADMINISTRATOR Marci Gallun;
EUROPE/ROW SALES Simone Jux, Alex Palmer; SALES ASSI S-
TANT Lisa Patterson; JAPAN Makiko Hara +81 (0) 90 9853-
9982, FAX +81 (0)3-6369-4491; careerads@sciencemag.jp;
CHINA/TAIWAN Ruolei Wu: +86 1367 1015 294 rwu@aaas.org;
ADVERTISING SUPPORT MANAGER Karen Foote: 202-326-6740; ADVERTISING
PRODUCTION OPERATIONS MANAGER Deborah Tompkins; SENIOR PRODUCTION
SPECIALIST/GRAPHIC DESIGNER Amy Hardcastle; PRODUCTION SPECIALIST
Yuse Lajiminmuhip; SENIOR TRAFFIC ASSOCIATE Christine Hall; SALES
COORDINATOR Shirley Young
AAAS BOARD OF DIRECTORS RETIRING PRESIDENT, CHAIR Alice Huang;
PRESIDENT Nina Fedoroff; PRESIDENT-ELECT William Press; TREASURER
David E. Shaw; CHIEF EXECUTIVE OFFICER Alan I. Leshner; BOARD Nancy
Knowlton, Stephen Mayo, Raymond Orbach, Julia M. Phillips, Sue V.
Rosser, David D. Sabatini, Inder Verma, Thomas A. Woolsey
I help higher education
professionals with their
RETIREMENT
SAVINGS
every day.
Im
READY
TO HELP
you, too.
Keep in mind that investing involves risk. The value of your investment will uctuate over time and you may gain or lose money.
Although consultations are one on one, guidance provided by Fidelity is educational in nature, is not individualized, and is not intended to serve as
the primary or sole basis for your investment or tax-planning decisions.
Fidelity Brokerage Services LLC, Member NYSE, SIPC, 900 Salem Street, Smitheld, RI 02917. 2011 FMR LLC. All rights reserved. 586605.1.0
866.715.6111
Fidelity.com/meetdelity
Im dedicated to helping higher education professionals feel condent
about retirement. Its what I love to do and Im here for you, too.
When you wonr ro rol| wirh someone o6our how ro reoch your gools
When you wonr ro 6e sure you're mo|ing rhe righr choices lor your needs
When you wonr on ongoing diologue o6our how ro sroy on rroc|
As a leading retirement plan provider for colleges and universities
across the country, were with you at every step. Start with a free,
no-obligation consultation with a Fidelity Investment Professional
even if you arent working with us today.
Diane Bolden
Fidelity Investment Education Consultant
Meet us today.
21 OCTOBER 2011 VOL 334 SCIENCE www.sciencemag.org 294
NEWS OF THE WEEK
C
R
E
D
I
T
S

(
T
O
P

T
O

B
O
T
T
O
M
)
:

R
E
U
T
E
R
S
;

C
N
E
S
Sriharikota on the Bay of Bengal.
The $125 million mission will study the
dynamics of cloud formation over the trop-
ics, and how climate change could be affect-
ing the monsoon. As it circles Earth near the
equator, Megha Tropiques will revisit the
same regions more than a dozen times each
day, simultaneously measuring water vapor,
clouds, precipitation, and radiation. These
multiple measurements, mission scientists
say, will give unique insights into the minu-
tiae of how clouds are born
and die during the
monsoon.
India and France have agreed to make
scientic data from the satellite freely avail-
ablea welcome prospect for atmospheric
scientists, says Christian Kummerow at
Colorado State University, Fort Collins.
It is a very exciting mission and we do
look forward to receiving the data from
its instruments.
Indian Ocean 3
Tsunami Warning System Passes
Critical Test
Twenty-three Indian Ocean nations came
together on 12 October to test a new warn-
ing communications network that might save
lives the next time the region is pummeled
by a tsunami. The $100 million Indian Ocean
Tsunami Warning and Mitigation System
performed well, though not awlessly, dur-
ing a simulation modeled on the devastating
tsunami of 26 December 2004 that killed
over 230,000 people in 14 countries. Sev-
eral countries also conducted dry runs of
their own emergency response plans. India,
Kenya, and Malaysia conducted evacuation
drills. An actual warning will depend on rap-
idly analyzing data from numerous seismic
stations, instrumented buoys, and sea-oor
pressure sensors deployed over the past
6 years, and spreading the word through
the networks tested last week.
The Intergovernmental Oceanographic
Commission, a part of the United Nations
Educational, Scientic and Cultural Orga-
nization, which coordinated development
of the system, declared the test a success.
Australia, India, and Indonesia will now take
responsibility for issuing warnings to the
region. The Japan Meteorological Agency
and the United States Pacic Tsunami
Warning Center have issued regional
warnings since 2005.
Moscow 4
Russian Scientists Rally to Protest
Funding Freeze
Hundreds of researchers, many in lab
coats, rallied in Moscows Pushkin Square
13 October to protest a funding freeze at
Russias two grant organizations and on
procurement regulations that they call
major obstacles to research. The rally was
organized by the trade union of the Russian
Academy of Sciences (RAS) and the Young
Scientists Council, together with associa-
tions of Moscow State University students,
and young scientists.
The Russian government recently froze
the budgets for Russias two funding agen-
cies, the Russian Foundation for Basic
Research (RFBR) and the Russian Founda-
tion for Humanities (RFH), leaving only
$200 million for both agencies. Protestors
urged the government to restore the old
rule, under which RFBR received 6% of the
overall budget for civilian science and RFH
1%. The protestors also demanded radical
reform of laws governing public procure-
ment, which severely limit grantees freedom
London 1
Royal Society: Plan Ahead
For Nuclear Power
Despite projections of low nuclear power
growth in Europe and the United States,
a renaissance of nuclear power construc-
tion in China, Southeast Asia, and Russia
is likely, Britains Royal Society notes in a
report released 12 October. As a result, the
report says, governments and international
bodies need to develop long-term policies
to account not only for safety but also for
security, proliferation risk, and fuel cycle
management.
Spent fuel can no longer be an after-
thought and governments worldwide need to
face up to this issue, Roger Cashmore, head
of the U.K. Atomic Energy Authority and
chair of the Royal Society working group
that drafted the report, said in a statement.
The panel recommends that countries
place their civil nuclear programs under
international safeguards run by the Inter-
national Atomic Energy Agency (IAEA), so
that spent fuel cannot be diverted for weap-
ons use. Countries that already have nuclear
weapons should separate their civil and
military nuclear programs. It also suggests
setting up a World Nuclear Forum, made up
of CEOs and government leaders, to discuss
nuclear developments and responsibilities.
http://scim.ag/nuclearUK
Sriharikota, India 2
Monsoon Satellite Promises
Data Deluge
The Indo-French satellite Megha
Tropiques, tasked with helping scientists
understand the water and energy balance
that controls monsoons, launched success-
fully 12 October from the Indian space port
1
4
5
6
2
3
AROUND THE WORLD
Deluged. Damage in Phuket, Thailand,
following the 2004 tsunami.
www.sciencemag.org SCIENCE VOL 334 21 OCTOBER 2011 295
NEWS
C
R
E
D
I
T
:

N
A
N
C
Y

K
E
D
E
R
S
H
A

Making the Invisible Visible
For her colorful microscopy images that reveal the once-hidden secrets of cells, biochem-
ist Nancy Kedersha has won the Lennart Nilsson Award for scientic and medical photog-
raphy, presented annually in honor of Swedish photographer Lennart Nilsson. The award
comes with a cash prize of SEK 100,000 (about $15,500). Kedersha will receive the award
at a ceremony 8 November at the Berwald Hall in Stockholm, Sweden.
Nancy Kedershas colour images open our eyes to the smallest components of life,
the award selection panel stated in a press release 14 October. With the aid of a confocal
microscope, she has turned biological data into an artistic experience.
Kedersha is a researcher at Harvard Medical School and director of the confocal
microscopy core at Brigham and Womens Hospital in Boston. While working in the lab of
biochemist and cell biologist Leonard Rome at the University of California, Los Angeles,
in the 1980s, Kedersha developed a technique to stain and photograph cells to reveal
their inner workings. Using this technique, she co-discovered a mysterious organelle
called a vault that exists in everything from humans to slime molds. She has continued
to develop techniques to identify different cell functions, distinguish healthy cells from
cancerous ones, and observe cells dividing.
THEY SAID IT
If Im going to take money
from a citizen to put into
education then Im going
to take that money to create
jobs. So I want that money
to go to degrees where
people can get jobs in this
state. Is it a vital interest
of the state to have more
anthropologists? I dont
think so.
Florida Governor Rick Scott (R) to the
Sarasota Herald-Tribune on 10 October.
to spend the money as they see t.
This rally is a warning, says Evgeny
Onishchenko of the RAS Institute of Phys-
ics, one of the organizers. We want to make
it clear that if nothing is done to meet our
demands, there will be much more serious
rallies of researchers all over the country.
Klong Luang, Pathum Thani, Thailand 5
Thai Floods Spare Research Park
Flood waters creeping toward Thailands big-
gest research park forced the evacuation of
dozens of public and private labs last week.
The Thailand Science Park, 30 kilometers
north of Bangkok, is home to 2700 employ-
ees working in four national research insti-
tutes under the National Science and Tech-
nology Development Agency as well as in
the labs of 60 private companies.
The campus closed 13 October; although
the science park was still dry on Monday,
it remained closed through 19 October due
to high water in surrounding areas. The
ooding, the worst in 50 years, has already
claimed more than 300 lives and, according
to a Businessweek report, caused over
$5.1 billion in damage.
Washington, D.C. 6
House Panel Lays Out
Spending Preferences
A climate-science satellite, some tech-
nology commercialization efforts, and a
chemical risk assessment program are all
among the federal R&D programs that
Republican leaders of the House of Repre-
sentatives Committee on Science, Space,
and Technology would cut to rein in the
U.S. budget decit. The ideas, which also
include protecting the core budgets of the
National Science Foundation (NSF) and
the Department of Energys (DOEs) Ofce
of Science, were highlighted in an unusu-
ally detailed 14-page letter that the law-
makers sent on 14 October to Congresss
bipartisan Joint Select Committee on De-
cit Reduction, which must devise a plan to
trim at least $1.2 trillion from the decit
over 10 years.
In general, the Republican lawmakers
took a back-to-basics approach, arguing
for protecting traditional science pro-
grams while trimming many newer efforts
championed by the Obama Administra-
tion. They took an especially dim view
of climate-related research; taxpayers
could save $149 million over 5 years,
for instance, by axing NASAs Orbiting
Carbon Observatory-2, designed to map
greenhouse gas emissions. They would
phase out DOEs Advanced Research Proj-
ects AgencyEnergy (ARPA-E), saving
$180 million. All told, they proposed cuts
totaling $1.5 billion.
The panels ranking Democrat, mean-
while, penned a less specic plea for spar-
ing the knife and fattening the federal purse.
When it comes to funding science, it is
critically important for the committee to
include serious revenue enhancements in its
set of recommendations, wrote Representa-
tive Eddie Bernice Johnson (DTX). Neither
letter is likely to have a major impact on the
decit committee, which faces a 23 Novem-
ber deadline for delivering its plan.
21 OCTOBER 2011 VOL 334 SCIENCE www.sciencemag.org 296
NEWS OF THE WEEK
entry, Microstructure-Property rela-
tionships in Ti2448 components pro-
duced by Selective Laser Melting,
Miller, 32, ies in silvery spandex
and a cape as he dances with women
representing titaniums alpha and
beta crystalline forms. He receives
$1000 and a trip to Belgium to be
crowned the winner on 22 November
at TEDxBrussels.
Category winners include X-ray
Crystal Structure of Human Protein
Phosphatase, by FoSheng Hsu of
Cornell University (Chemistry); Smell-
Mediated Response to Relatedness of
Potential Mates, by Cedric Tan of the Uni-
versity of Oxford in the United Kingdom
(Biology); and A Study of Social Inter-
activity Using Pigeon Courtship,
by Emma Ware of Queens University
in Canada (Social Science). Videos of
this years 55 Ph.D. dances are at
www.gonzolabs.org/dance.
C
R
E
D
I
T
S

(
T
O
P

T
O

B
O
T
T
O
M
)
:

J
O
E
L

M
I
L
L
E
R
;

M
I
N
E
R
A
L
I
E
N
T
A
G
E

M

N
C
H
E
N
NEWSMAKERS
CSI: Amphorae?
What did amphorae, the ubiquitous ceramic
jugs of the ancient Mediterranean maritime
world, actually contain? To nd out, mari-
time archaeologist Brendan Foley of the
Woods Hole Oceanographic Institution in
Massachusetts and colleagues turned to a
CSI-like method: swiping the insides of the
amphorae with a swab. They swabbed nine
5th to 3rd century B.C.E. amphorae for DNA
and compared it with snippets of DNA from
various plants.
They identied DNA from a range of
commodities, including olive oil, olives,
and wine, as well as traces of DNA from
oregano, thyme, mint, and juniper, the team
reports in a study published online this
month in the Journal of Archaeological
Science. Eight of the nine amphorae bore
DNA from a complex mixture of foods,
suggesting that amphorae were reused,
Foley says.
Not everyone is convinced that the tech-
nique works, however. It is remarkable
that an amphora should release endog-
enous DNA by simply swabbing the sur-
face, Oliver Craig of the University of York
in the United Kingdom said via e-mail.
Craig, who specializes in recovering DNA
and other molecules from ancient artifacts,
says he would need to see more control
tests to be convinced.
http://scim.ag/CSIamphora
Nearly Intact Dino Fossil Found in Germany
An exceptionally well-preserved baby dinosaur, with traces of skin and protofeathers, will
be the main attraction at a fossil and gem show next week in Munich. The juvenile theropod,
which lived between 145 million and 150 million years ago and was probably less than a year
old when it died, is 98% intact. That makes it the most nearly complete dinosaur ever found
in Europe, says Oliver Rauhut, curator at the Bavarian State Collection for Palaeontology and
Geology in Munich, who led the rst examinations of the fossil.
The fossils hairlike protofeathers may help researchers understand how and when feath-
ers evolved. The roughly 70-cm specimen, not yet fully classied or named, was unearthed
near the Bavarian town of Kelheim and has been registered as a German cultural artifact,
which means that it cant leave the country. The mineral show will display the fossil for 4
days starting 27 October; a spokesperson for the show has said the unnamed owner plans
to lend the fossil to a museum.
FINDINGS
Dances With Titanium
Fifty-ve scientists around the world moon-
lighted as choreographers for Sciences
fourth annual Dance Your Ph.D. contest
and the results are in. This years winning
dances were based on protein x-ray crystal-
lography, fruit-y sex, and pigeon courtship,
each scooping $500 prizes.
The grand winner, announced 20 Octo-
ber, is Joel Miller, a biomedical engineer
at the University of Western Australia in
Perth. Millers dance, which won the phys-
ics category, depicts his work with lasers
to create titanium alloys strong and ex-
ible enough for long-lasting hip replace-
ments. We didnt have a video camera,
says Miller. So he and his friends converted
2200 still photographs of the dance into
stop-motion animation. In his winning
www.sciencemag.org SCIENCE VOL 334 21 OCTOBER 2011 297
NEWS
C
R
E
D
I
T
S

(
T
O
P

T
O

B
O
T
T
O
M
)
:

C
O
U
R
T
E
S
Y

J
O
N
A
T
H
O
N

K
E
A
T
S
;

M
U
S
E
U
M

O
F

L
O
N
D
O
N

A
R
C
H
A
E
O
L
O
G
Y
;

H
O
M
E
R

S
Y
K
E
S

A
R
C
H
I
V
E
/
A
L
A
M
Y

BY THE NUMBERS
1 billion tons Amount of extra
food that that could be grown on
agricultural lands now devoted to
animal feed and biofuel produc-
tion, according to an analysis in this
weeks Nature.
200,000 amps Maximum
power generated by a new airplane
industrysponsored lab at Cardiff
University that studies the effects of
lightning on materials. An average
lightning strike generates 10,000 to
30,000 amps.
Random Sample
The Story Is Dead. Long Live the Story.
Artist and self-styled experimental philosopher
Jonathon Keats is hoping to persuade the art world to
join scientists in the Copernican Revolutionnearly
5 centuries late. In 1543, Nicolaus Copernicus made the
humbling observation that the Earth revolves around
the sun. Modern physicists often cite the Copernican
principle that, as natures rules are the same every-
where, the human viewpoint isnt unique.
But the art world, Keats says, is still stubbornly
Ptolemaic, in that it emphasizes the exceptionalism
of humans and centers on stories about ourselves. So,
in The First Copernican Art Manifesto, an exhibit that
opened Thursday at the Modernism gallery in San Fran-
cisco, California, Keats will feature art that reects banal, average truths about the universe.
The pieces dont assume a human audience or viewpointand they dont aim to appeal
to us, either. One canvas is painted a bland tan, the average color of the starlight of all stars
measured by astronomers. Hydrogen gas released from glassware suspended above otherwise
empty pedestals assumes a form invisible to human eyes. A quarter of the notes in a once-
orderly Bach composition are rearrangedreecting the increasing entropy of the universe
since its tidy, prebig bang singularity.
Although not for humans, the exhibition is aimed at a particular demographic, in a way.
Were the aliens to land and see our show, they wouldnt say, Now I understand humanity,
Keats says. Theyd say, Now I have a better understanding of the universe. The exhibit runs
through the end of November.
Cute TV Chimps May Harm
Wild Brethren
Some entertainment industry moguls claim
that chimpanzees dressed in clothes and
clowning around fosters sympathy for the
species. But a study published 12 October
in PLoS ONE suggests the opposite: Peo-
ple who watch such shows or ads decide
chimpanzees are abundant in the wild and
dont need further protection.
Evolutionary anthropologist Brian Hare
at Duke University in Durham, North Caro-
lina, and colleagues asked 165 people to
answer a questionnaire about the status of
chimpanzees in the wild after watching tele-
vision ads for products such as toothpaste
and soft drinks. Mixed in with the ads was
one of three short lms about chimpanzees.
One showed Jane Goodall urging for their
protection; another showed footage of chim-
panzees in the wild; and the third showed
chimpanzees acting in ads.
The results suggested absolutely no
support for the familiarity hypothesis,

t
-
Hare says. More than 35% of
those who watched the humor-
ous ads thought individuals
should have the right to own a
chimpanzee as a pet, compared
with only 10% of those who
watched the two other lms.
Those who watched the enter-
tainment chimps were also least
likely to donate to a conserva-
tion charity.
http://scim.ag/chimpads
Black Death Spawned
Modern Plague
These skeletonsexcavated in the
1980s from a 14th century graveyard in
Londonbelonged to six of the estimated
30 million people who died from the Black
Death, the plague epidemic that swept
Europe between 1347 and 1351. Research-
ers used teeth from the same graveyard
home to 2500 plague victimsto recon-
struct 99% of the genome of Yersinia
pestis, the bacterium that causes plague.
An analysis of that microbial DNA pub-
lished online 12 October in Nature sug-
gests that Y. pestis strains currently circu-
lating around the world are all descendents
of the medieval strain believed to have
killed 30% to 60% of Europes population.
The 14th century genome closely resem-
bled those of modern strains and did not
have any obvious unique mutations that
might explain its unprecedented virulence.
Other factorssuch as the populations
susceptibility or the ecology of rodents
and eas, which help spread the disease
were probably responsible for the medi-
eval calamity, the team concludes.
21 OCTOBER 2011 VOL 334 SCIENCE www.sciencemag.org 298
NEWS & ANALYSIS
C
R
E
D
I
T
:

J
O
S
E
P
H

O
K
A
N
G
A
/
R
E
U
T
E
R
S
Sometimes just meeting expectations is a
major achievement. Initial, eagerly awaited
results from the worlds first large-scale
trial of a malaria vaccine, carried out at 11
sites in seven African countries, show that
it reduced episodes of the disease by about
half in babies and toddlers. That conrms
the efcacy seen in earlier, much smaller
trials of the experimental vaccine, so far the
only one to show signicant benet against
malaria in real-world settings.
The new ndings keep the candidate on
track to become the rst licensed vaccine
against the disease, says Christopher Plowe,
a malaria vaccine expert at the University of
Maryland School of Medicine in Baltimore,
who is not involved in the trial. Even a par-
tially effective vaccine, used in combination
with other tools like bed nets, could curtail
malarias massive death toll signicantly,
experts say. But the vaccine will be expen-
sive by developing-world standards, and its
cost-effectiveness is yet to be determined.
I am thrilled, says Joe Cohen, one
of the vaccines original developers and
leader of the malaria vaccine project at
GlaxoSmithKline (GSK) Biologicals in
Rixensart, Belgium. The fact that the huge
trial confirms results from smaller pre-
decessors is fabulous, he says. Robert
Newman, head of the World Health Orga-
nizations (WHOs) malaria program,
agrees. The results are in line with what
we expected. But one fears they wont hold
up, so in line is very encouraging.
The rst round of results was published
online on 18 October by The New Eng-
land Journal of Medicine; they were also
announced by Bill Gateswho called them
phenomenalat a Seattle meeting hosted
by the Bill & Melinda Gates Foundation,
which has given more than $200 million to
support trials of the vaccine. The data show
that in 6000 children aged 5 to 17 months,
three doses of the vaccine cut the risk of any
episode of malaria by 56% and the risk of
severe disease by 47%. Thats far from the
90% efcacy that most vaccines against viral
and bacterial disease achieve. But the Plas-
modium falciparum parasite, with its multi-
ple life stages, is a much more difcult target,
and no one expected a rst-generation vac-
cine to be more than partially effective. This
vaccine will not be a magic bullet against
what is a very, very difcult disease, Cohen
says. It is one weapon to be added to an arse-
nal of other interventions.
The vaccine, called RTS,S, was developed
in 1987 by researchers working for a prede-
cessor to GSK Biologicals. It contains an
engineered protein that combines a protein
fragment from P. falciparum and a protein
from the hepatitis B virus that helps trig-
ger a strong immune response. The vaccine
is designed to block the parasites ability to
infect the liver and mature there.
After early human trials in 1997 showed
promising resultsprotecting six of seven
adult volunteersGSK entered a public-
private partnership with the PATH Malaria
Vaccine Initiative (MVI) to further develop the
vaccine. The rst eld trials in 2000 children
in Mozambique, launched in 2003, showed
that the vaccine lowered the risk of devel-
oping malaria symptoms by 30%, with no
severe side effects (Science, 22 October 2004,
p. 587). Since then, phase II trials in Mozam-
bique, Kenya, and Tanzania have consistently
shown that the vaccine can cut the number of
malaria episodes by between 35% and 53%
(Science, 12 December 2008, p. 1622).
The phase III trialthe final test
enrolled more than 15,000 babies aged
6 to 12 weeks and toddlers between 5 and
17 months across sub-Saharan Africa. All
were scheduled to receive three doses, each
1 month apart; a subgroup will receive a
booster dose 18 months later. The results
announced this week are for the toddlers
and cover the 12 months after their first
shot. (Infants were enrolled slightly later, so
results from that group wont be available
until the end of 2012.) Children who missed
one or two of the doses were almost as well
protected as those who received all three
shots, the researchers report.
The vaccine also looks fairly safe. Chil-
dren who received the vaccine had a slightly
higher rate of seizures than those who
received the control injection, a rabies vac-
cine. But the independent safety board that
keeps watch over the trial has not raised
any concerns, says MVI Director Christian
Loucq. Children enrolled in the trial had a
very low risk of dying from malariaeven
if they received the control injections
mainly because clinics put in place proce-
dures to detect and treat cases as soon as
possible. There were only 10 malaria deaths
in the rst 2 years of the study, Cohen says.
In a separate analysis, the researchers
looked at the rate of severe malaria to date
in all 15,460 children enrolled in the study;
they found that the vaccine reduced the rate of
severe, life-threatening disease by 35%. That
hints that effectiveness might be lower in the
babies, but Loucq and Cohen caution that the
number is very preliminary.
The babies received their doses at the
same time they receive the standard infant
vaccinations recommended by WHO. Add-
ing the malaria vaccine to existing vaccina-
Vaccine Trial Meets Modest
Expectations, Buoys Hopes
MALARI A
Promising jab. A baby receives a dose of the experi-
mental malaria vaccine at a trial site in Kili, Kenya.
www.sciencemag.org SCIENCE VOL 334 21 OCTOBER 2011 299
NEWS
C
R
E
D
I
T
S

(
T
O
P

T
O

B
O
T
T
O
M
)
:

T
H
I
E
R
R
Y

R
O
G
E
/
R
E
U
T
E
R
S
;

M
L
P
/
N
I
H
tion schedules would be the most practical
approach if the vaccine is to be widely used,
and smaller-scale trials have suggested that it
is effective and safe when given along with
the other shots.
The trial will continue until 2014 and will
follow the children until 30 months after their
third dose. Once the full results are unblinded,
researchers will know more about how long
protection lasts and will also be able to com-
pare the vaccines performance at different
study sites, which have different rates and
seasonal patterns of malaria trans mission.
That could help governments and public
health experts decide where RTS,S might
have the most impact. The complex vaccine
is expensive to make, and its cost-effective-
ness is a major issue, says Scott Filler of the
Global Fund to Fight AIDS, Tuberculosis and
Malaria. The key question is going to be
cost, he says, given the limited funds avail-
able for ghting malaria.
GSK, which has invested more than
$300 million in RTS,S to date, has pledged
to keep the price as low as possiblejust
manufacturing costs plus a small return to
be reinvested in development of second-
generation malaria vaccines or vaccines
against other neglected tropical diseases.
Even so, a full vaccine course is likely to cost
more than other prevention methods, such as
insecticide-treated bed nets, which also
offer partial protection. Many countries
have already rolled out massive net distribu-
tion projects, and 75% of the children in the
trial slept under a net, Cohen says; the results
show that the vaccine can provide an extra
layer of protection on top of the nets, he says.
WHO is expected to take all such issues
into account when it drafts policy recom-
mendations for use of the vaccine after the
trials nal results come in. The vaccine has
incredible potential to reduce suffering,
Filler says, but deciding how and where to
use it will take much more work. These are
going to be incredibly challenging questions
for which wethe community as a whole
dont have answers yet.
GRETCHEN VOGEL AND LESLIE ROBERTS
A $70-million-a-year program launched
7 years ago at the National Institutes of
Health (NIH) to help academic research-
ers move into industry-style drug discovery
may soon be forced to scale back sharply.
NIH Director Francis Collins has been
one of its biggest champions. But the NIH
Molecular Libraries, according to plan,
must be weaned starting next year from the
NIH directors office Common Fund and
nd support at other NIH institutes. In a time
of tight budgets, nobody wants it.
The fate of the Molecular Libraries pro-
gram became an extremely sensitive politi-
cal issue earlier this year when NIH realized
it would not be easy to nd a new home for
the program, said one NIH ofcial speaking
on background. It illustrates the complexity
of moving projects out of the Common Fund,
says the funds overseer, James Anderson,
director of NIHs Division of Program Coor-
dination, Planning, and Strategic Initiatives.
Obviously the process would be easier if
NIHs budget were growing, he says.
The Molecular Libraries began as a large
piece of the NIH Roadmap, a set of cross-
cutting initiatives announced in 2003 by
thenNIH Director Elias Zerhouni. (The
Roadmap later became the Common Fund.)
As described by Collins, who was then
the genome institute director, academic
researchers would submit protein or cell
assays to a set of academic screening cen-
ters, which would be paid by NIH to look
for biological interactions with thousands
of chemicals. Chemists would then rene
these hits into research probes, some
of which could become drug candidates.
Although some industry scientists were
skeptical of government-led drug research,
NIH scaled up the program from a pilot
phase in 2008 (Science, 8 August 2008,
p. 764). The program now includes four
large screening centersone part of
NIHs in-house research program, the rest
extramuraland five supporting centers.
They have produced more than 240 probes,
one of which is now in clinical trials as a mul-
tiple sclerosis (MS) drug. Hugh Rosen, direc-
tor of a center at Scripps Research Institute in
San Diego, California, said papers in top jour-
nalsincluding those on the MS drugshow
that the program has transformed academic
chemical discovery and target validation.
The centers were told, however, that their
Common Fund money would end in 2014 and
that they would have to nd other sponsors.
Only the intramural center has found a poten-
tial home so far: It will move to NIHs planned
National Center for Advancing Translational
Sciences, if Congress approves.
The extramural centers arent so lucky.
Their Common Fund
awards will drop by
33% next June and
another 33% in June
2013. Other institutes
are reluctant to pick
them up because they
need the money for research grants, accord-
ing to Molecular Libraries program director
Carson Loomis. Loomis says NIH intends to
replace some of the declining money with
new grants to users of the extramural cen-
ters focused on drug discovery. Still, he says,
there will be some belt-tightening.
John Reed, head of the Sanford-Burnham
Medical Research Institute screening center
in San Diego, which receives about $16 mil-
lion a year from the Common Fund, says his
center has so far attracted only modest fund-
ing from drug companies. He expressed frus-
tration with the Common Fund process. NIH
has put a huge investment into [the Molecular
Libraries], and its running very well, he says.
If theres not a long-term commitment to
keep it available to the academic community,
why did we make this hundreds of millions of
dollars investment? JOCELYN KAISER
Drug-Screening Program Looking for a Home
NATI ONAL I NSTI TUTES OF HEALTH
2
a
th
Screen test. NIHs
Molecular Libraries has
yielded scores of probes.
Long-term investment. Joe Cohen of GSK Biologicals
has been working on the RTS,S vaccine for 24 years.
21 OCTOBER 2011 VOL 334 SCIENCE www.sciencemag.org 300
NEWS&ANALYSIS
S
O
U
R
C
E
:

N
A
T
I
O
N
A
L

M
A
T
H

A
N
D

S
C
I
E
N
C
E

I
N
I
T
I
A
T
I
V
E
,

2
0
1
1
The U.S. Ofce of Naval Research (ONR) has
joined forces with the Texas-based National
Math and Science Initiative (NMSI) to
increase the number of high school students
taking rigorous science and math courses. Its
$1.1 million grant, announced last month,
could prompt a closer look at one controver-
sial aspect of the initiatives approach: paying
students to do well on standardized tests.
Launched in 2007 with funding from
major corporations and foundations, NMSI
gives money to school districts that agree to
follow its tightly scripted program. The goal
is to boost participation in Advanced Place-
ment (AP) classes, a curriculum designed
by the nonprot College Board to be on par
with entry-level college courses.
The NMSI program, which targets
low-achieving and low-income high
schools with large minority popu-
lations, assumes that these students
can handle more challenging mate-
rial if given the chance. NMSI of-
cials say that is already happening:
In the midst of a nationwide surge
in AP test-taking, the number of
minority and female students pass-
ing AP tests at 228 NMSI-sponsored
schools this year increased four to 10
times faster than for the country as a
whole (see graphic).
Most of NMSIs approach rep-
resents education orthodoxy: extra
class time for students, additional
resources, and special training
for teachers. But there is less evi-
dence behind another core element,
namely, cash incentives. Students
receive from $100 to $500 for pass-
ing an end-of-the-year test, and their
teachers also get bonuses for each stu-
dent who succeeds and for teaching
AP classes. NMSI pays for success in only
English, math, and science courses (tests
are offered in 34 subjects). But students
may take half a dozen or more AP courses
in those subjects during their high school
careers, so the money can add up.
ONRs investment puts the Navy ahead
of the curve on national education policy.
The Obama Administration has no offi-
cial position on the use of cash incentives,
and Congress has never addressed the topic
in legislation. At the same time, Education
Secretary Arne Duncan has voiced support
for the concept as a tool for raising student
achievement, and a small departmental pro-
gram to increase participation in AP courses
allows ofcials to give money directly to stu-
dents. However, only two of the 55 current
grantees are doing so.
The Navys STEM (science, technology,
engineering, and mathematics) education
budget is expected to double in 5 years, to
more than $100 million, reecting its con-
cern about lling science- and technology-
based Navy jobs. Michael Kassner, head of
the Navys STEM ofce, says NMSI could
become a big part of that investment if
the 3-year pilot, which supports three pub-
lic schools in Virginia and Hawaii that enroll
large numbers of students from military
families, is able to help prime the pump.
We know that students whose parents are
in the military are more likely to go into the
military, he adds.
Kirabo Jackson, a labor economist at
Northwestern University in Evanston, Illi-
nois, has looked at the Texas program that
spawned NMSI. His 2008 study, perhaps the
only one to examine the role of incentives in
the population that NMSI is targeting, found
that the strategy increased AP participation
rates and boosted the number of students with
high scores on national college entrance tests.
At the same time, the program didnt increase
high school graduation rates or the number
of students taking college entrance exams.
That nding suggests its more likely to help
high achievers already headed to college than
to raise the aspirations of those who hadnt
planned to continue their education.
Still, his overall assessment is posi-
tive: Its one of the few programs that does
something good for these students, Jackson
says. Most programs havent been evalu-
ated rigorously. And I dont know if it can be
expanded to other settings, with other popu-
lations. But if a school district had $1 mil-
lion to spend, I think a program like this is a
good investment.
A 2010 study by economist Roland Fryer
Jr. of Harvard University delivers a much
more sobering message about the value of
cash incentives for younger students. Fryer
conducted a randomized trial of experimen-
tal programs in four large urban school dis-
tricts involving 38,000 children from grades
two through nine. Although the program ele-
ments varied greatly from one district to the
next, he found that paying for outputs, such
as test results, didnt work and in some cases
resulted in lower scores. On the other hand,
paying students for inputsshowing up for
class, staying on task, reading a certain num-
ber of bookshad a positive effect, he writes
in a paper posted by the National Bureau of
Economic Research.
Teacher unions have generally been very
leery of incentives or bonuses. In addition to
clashing with most labor agreements, they
are regarded as undermining the learning
process. And almost none of these incentive
programs have worked, says a spokesperson
for the United Federation of Teachers, which
represents New York City schools.
A program serving 31 New York City
schools that is otherwise modeled after
the Texas and NMSI efforts omitted the
teacher incentives after the union declined
to participate. And this month ofcials for
the program, called REACH (Rewarding
Achievement), dropped the student incen-
tives in response to a nancial squeeze. We
found that it prompted more students to take
AP courses, but we didnt see the magnitude
effect that we had hoped, says Kathrine
Mott, REACHs executive director. REACH
will continue to offer professional devel-
opment for teachers, Saturday classes, and
classroom grants, she adds.
JEFFREY MERVIS
Navy Dives Into Program Offering Cash for Good Scores
U. S. SCI ENCE EDUCATI ON
Students taking AP tests
P
e
r
c
e
n
t
a
g
e

i
n
c
r
e
a
s
e

o
f

p
a
s
s
i
n
g

s
c
o
r
e
s
FEMALES*
AFRICAN-
AMERICAN
AND
HISPANIC**
OVERALL**
0
50
100
150
200
250
23%
124%
50%
216%
20%
144%
*Math, Science
**Math, Science,
and English
U.S.
Total
NMSI
Schools
Advancing AP. NMSI ofcials say that test results demon-
strate the value of cash incentives in boosting the number of
high school students taking and passing AP courses.
www.sciencemag.org SCIENCE VOL 334 21 OCTOBER 2011 301
NEWS&ANALYSIS
C
R
E
D
I
T
:

H
O
N
G
M
I
N

J
I
N
In Chinas booming economy, there are
many ways to get rich. For a husband-and-
wife team on Hainan, an island off Chinas
southern coast, scientific publishing was
their cash cow. For 7 years, Guo Hong and
Fu Li operated 20-some journals, collecting
a reported $1.5 million in publication fees
from thousands of contributors. They solic-
ited papers through elaborate Web sites,
offering a discount on the publication fees
common in China. But the journals were
fake, provincial authorities allege. Upon
receiving submissions, the couple would
print up only a few copiesjournal titles
included Chinese Applied Nursing and
Chinese Medicine Forumto send to the
author. Guo and Fu were detained in March;
prosecutors have not yet led charges.
The highly publicized takedown is one
of several recent efforts to clean up Chinas
academic publishing industry. In a country
where low publication standards abound and
every university or institute, it seems, has its
own journal, the Chinese government is get-
ting serious about raising standards. Although
the Hainan journals fraud is an outlier, its
symptomatic of a larger problem: slapdash
and irrelevant publications read by next to no
one. At most Chinese journals, the academic
level is not high, Li Dongdong, vice director
of the General Administration of Press and
Publication (GAPP), which regulates publi-
cations in China, noted in a speech in Decem-
ber. She estimated that two-thirds of journals
are not market-oriented.
As Chinese science barrels ahead, a few
of its journals are getting international atten-
tion, and leading Western publishers have set
up shop on the mainland in response. But the
countrys 4700 scientic periodicals include a
hefty number of what the Chinese press refers
to as trash journals. Despite being second
only to the United States in total papers pub-
lished from 2006 to 2010, China ranked at the
bottom of the top 20 countries for citations
per article over the same period, with just 1.47
citations on average, according to Elseviers
SciVerse Scopus database and SciVal Spot-
light country matrix, compared with 5.16 for
the United States. It doesnt help that many
institutions in China offer fat rewards for pub-
lishing in overseas journals with high impact
factors. For example, according to its Web
site, Guangzhou Medical University doles
out 300,000 yuan ($47,000) to lead authors
on papers in journals with impact factors of
at least 15a level no Chinese journal has
attained. (In Thomson Reuterss Journal
Citation Reports, top-ranked journals in cat-
egories such as cell biology and biochemistry
can show impact factors of over 30.)
As an antidote, GAPP has begun rolling
out a series of reforms aimed at boosting the
prestige of Chinese publishing. GAPP has
been given heaps of money to spend, says
Torsten Weise, a Berlin-based consultant who
advises foreign publishers on operating in
China. In the past 2 years, GAPP has secured
billions of dollars in loans from state banks,
with the major goal, Weise says, of inter-
nationalization: building journals and pub-
lishers capable of becoming multinational.
Chinas 12th 5-year plan, in effect since
March, sets a heady goal for journals. It calls
for making cultural productionincluding
media and publishinga pillar industry.
GAPP has moved swiftly. Earlier this year,
the agency closed six obscure publications
and reprimanded two others for violations
that included indiscriminate printing of up
to 200 papers per issue, over the limits set by
publishing licenses. Then last summer, of-
cials unveiled China Science and Technology
Media Group, one of a handful of agship
publishers due to be rolled out over the next
few years to compete with foreign rivals such
as Wiley, Elsevier, and Springer. Li has talked
about GAPP supporting a group of select
academic journals; editors are unclear when
funding might materialize.
The malaise has deep roots. The prolifera-
tion of journals is tied to the danwei, or work
unit, system put in place after 1949. As the
government brought institutions under cen-
tral control, academic work unitsoften uni-
versity departments or instituteslaunched
journals to publish their scholars work.
Fast-forward to the 1990s. With Chi-
nese science opening up, academic centers
shifted course and began rewarding scientists
for publishing in journals listed in indices
that track citation rates. Pressure to publish
piled up, and although only a few thousand
Chinese papers a year then made it into jour-
nals indexed by Thomson Reuters, authors
seeking to get into print in both Chinese and
English-language outlets proliferated. Medi-
ocre danwei-linked journals gladly solicited
papers from outside scientists and began
charging steep publication fees. (These can
now top $1000.) And so the early journals
persisted, constituting what Cong Cao, a
scholar of Chinese science at the University
of Nottingham, U.K., calls a phenomenon
with Chinese characteristics.
Today, the few stars that have emerged
are published in English and focus on areas
in which China is strong, such as cell biol-
ogy, nanoscience, and materials science.
They have risen quickly: In 1999, the high-
est impact factor of any Chinese journal was
0.5 (Science, 26 November 1999, p. 1683).
Today, Chinas top indexed journal, Cell
Research, has an impact factor of 9.4.
Chinas leading journals have made their
mark by bringing on international editorial
boards, wooing editors from top-shelf West-
ern publicationsCell Research poached
deputy editor in chief Li Dangsheng from
Cell in 2006and taking stabs at branding,
such as shedding China-specific names in
favor of more international monikers. Oth-
ers, including Chinese Medical Journal and
Science China Life Sciences, are experiment-
ing with open-access platforms. We have to
change the way journals are run, says Gang
Pei, editor-in-chief of Cell Research. He has
cultivated relationships with societies and
accelerated response time for submissions
from leading researchers.
The vast majority of journals have little
hope of following that recipe. Their day of
reckoning is not long off. There is no need
to keep poor-quality journals around, says
Meng Zhao, development editor at Neural
Regeneration Research. By administrative
measures or by market measures, Pei says,
there will be some kind of cleanup.
MARA HVISTENDAHL
China Looks to Purge Academia
Of Trash Journals
SCI ENTI FI C PUBLI SHI NG
Raising the bar. Gang Peis Cell Research is Chinas
top indexed journal.
21 OCTOBER 2011 VOL 334 SCIENCE www.sciencemag.org 302
NEWS&ANALYSIS
C
R
E
D
I
T
S
:

C
E
N
T
E
R

F
O
R

T
H
E

S
T
U
D
Y

O
F

T
H
E

F
I
R
S
T

A
M
E
R
I
C
A
N
S
,

T
E
X
A
S

A
&
M

U
N
I
V
E
R
S
I
T
Y
The experienced hunter hurled or thrust his
spear at the left side of the 3-ton adult male
mastodon. The bone point passed through
25 to 30 centimeters of hide and muscle,
then pierced a rib bone. That wound alone
likely did not kill the massive animal, but
under the onslaught of a group of hunters,
the mastodon eventually fell on its left side.
The victorious hunters then retrieved most
of their weapons and butchered the animals
right side. But the valuable spear point
remained inaccessible, buried under
the giant carcass.
That ancient hunters loss is a
gain for researchers on a different
kind of hunt: the search for clues
to the peopling of the Americas.
Back in the 1970s, archaeologists
found the mastodons remains,
complete with rib bone and embed-
ded point, at the Manis Mastodon site
on Washingtons Olympic Peninsula,
near the Juan de Fuca Strait. Now on
page 351 of this issue, researchers led
by Michael Waters of Texas A&M Uni-
versity in College Station use DNA
and radiocarbon dating to demonstrate
that the point came from a mastodon
bone shaped into a weapon by humans
and used a startling 13,800 years ago.
Thats nearly 1000 years before the
Clovis culture, known for its distinctive
stone spear points and long considered
to be the rst culture in the New World.
The nd adds to the wave of recent com-
pelling evidence demonstrating an earlier,
pre-Clovis settling of the Americas (Science,
25 March, p. 1512). Although a few Clovis-
rst holdouts remain unconvinced, the early
bone point also suggests that the extinction
of large mammals such as mastodons and
mammoths may have begun long before the
Clovis people came on the scene. This is
signicant because we have so few widely
accepted pre-Clovis sites, says anthropolo-
gist Daniel Sandweiss of the University of
Maine, Orono. The solid dating combined
with the strong evidence for pre-Clovis hunt-
ing on a site near the coast make the results
particularly important, he adds.
Waterss team subjected the rib and bone
barb to a battery of tests, from DNA sequenc-
ing and protein analyses to radiocarbon dating
and a CT scan. They determined that the barb
comes from another mastodon and appears to
resemble the sharpened bone points used to
kill mammoths, mastodons, and other large
animals in Beringia, the land now partly
submerged in the region around the Bering
Strait. The first Americans likely migrated
from Beringia as the last ice age gave way to
warmer temperatures and glaciers retreated.
For many archaeologists, the debate over
whether pre-Clovis peoples roamed the
Americas is over. Manis is another pre- or
non-Clovis site on the map, says anthropol-
ogist Tom Dillehay of Vanderbilt University
in Nashville, who announced his
own pre-Clovis site in Chile
back in 1997.
Megafauna like the mast-
odon and its relative the mam-
moth disappeared quickly after
the arrival of Clovis points
13,000 years ago, prompting
some scientists to speak of a
blitzkrieg: a rapid hunting of these giant
animals to extinction. But combined with
evidence of pre-Clovis mammoth hunting
at two other North American sites, Waters
argues that human hunters were already at
work on killing megafauna before the debut
of Clovis-style weapons. Other researchers
agree: The notion of the blitzkrieg move-
ment died with the Clovis-rst paradigm,
Dillehay asserts.
Although a well-dated point embedded in
a mastodon rib seems like a smoking gun
or spearfor the pre-Clovis case, a handful
of Clovis-rst advocates remain skeptical.
They say that so far all the evidence, includ-
ing that from Manis, has problems. There
may have been a period of pre-Clovis
human presence in North America, but I
wish I could see a solid demonstration of
that presence somewhere that doesnt have
nagging problems, says archaeologist Gary
Haynes of the University of Nevada, Reno.
He questions whether the bone point was
really a human-shaped weapon, saying that
it might have been a piece of bone acci-
dentally driven into the rib. Archaeologist
Stuart Fiedel of the Louis Berger Group Inc.
in Richmond, Virginia, praises the teams
sophisticated techniques but questions the
dating. He notes that the mastodons envi-
ronment had sources of old carbon,
including the ocean and geothermal pools,
that could give a falsely ancient date
if the mastodon ingested food or
water from those sources. Waters
calls this a red herring, however,
because the surrounding sediment
age closely matches the bone dating.
Haynes adds that the oldest Clovis
sites are only 8 centuries younger
than Manis. Thus the rib may actu-
ally indicate the earliest beginning
of the Clovis era, or an immediately
proto-Clovis stage of human disper-
sal, he says. Proto-Clovis peoples
in small numbers may have ltered
south from Beringia as early as
14,000 years ago, he says, although
their impact was negligible until
the arrival of Clovis technology. To
Waters, such talk of proto-Clovis
amounts to grasping at straws.
Sandweiss says the implications
of the paper in fact go beyond the
Clovispre-Clovis wrangling. Many
scientists argue that pre-Clovis peo-
ple moved south along the Pacific
Coast, possibly by boat (Science, 4 March,
p. 1122). In his view, the location of Manis
near the ocean is an intriguing hint favor-
ing this idea. The Manis site supports early
occupation of the coastal zone, he says.
Waters, who has reported pre-Clovis inte-
rior sites as well, is cautious. We cant say
these were coastal folks, he says of the
Manis hunters. But the papers findings,
Sandweiss says, point to a more wide-
spread and complex early settlement system
than some might have suspected.
ANDREW LAWLER
Pre-Clovis Mastodon Hunters Make a Point
ARCHAEOLOGY
Big game. Archaeologists say that pre-Clovis hunters in
Washington state used a bone spear point (seen above
embedded in a rib) to pierce a mastodon.
www.sciencemag.org SCIENCE VOL 334 21 OCTOBER 2011 303
NEWS&ANALYSIS
C
R
E
D
I
T
S
:

(
L
E
F
T
)

C
H
R
I
S
T
O
P
H
E
R

P
A
L
A
;

(
R
I
G
H
T
)

N
A
S
A

S

M
O
D
I
S

R
A
P
I
D

R
E
S
P
O
N
S
E

S
Y
S
T
E
M

(
2
)
TASTUBEK, KAZAKHSTANAt a shing camp
near this village on the Northern Aral Sea,
a dozen small boats recently returned to
shore bathed in a soft morning light. Their
nets bulged with carp, pike, ounder, perch,
and a half-dozen other species, all edible
and ready for sale. It was a scene unimagi-
nable just 6 years ago, when the Aralonce
the worlds fourth-largest lakehad shrunk
to one-tenth of its original size. Soviet-era
planners had diverted most of the water that
owed into the lake from two rivers to irri-
gate cotton, creating three smaller lakes that
became too salty for most sh. Catches that
once totaled more than 50,000 tons a year
plummeted to just 52 tons in 2004.
Now, since the 2005 construction of a
$65 million dike, the northern part of the
Aral has become a remarkably healthy
shery. The biomass, or weight of all the
sh in the Northern Aral, has soared from
an estimated 3500 tons in 2005 to 18,000
tons today, says Zaualkhan Yermakhanov,
the Kazakh governments regional sheries
director. And its still growing.
Its been an amazingly fast recovery,
says Philip Micklin, a retired geographer
from Western Michigan University who has
been studying the Aral Sea since the 1970s.
During an expedition last month around the
Rhode Islandsized lake, Micklin found
that salinity levels have dropped and oxy-
gen levels increased since the Kazakh gov-
ernment, backed by the World Bank and
other donors, built the 13-kilometer earthen
dike along the Northern Arals southeastern
edge. The dike traps water from one feeder
river, the Syr Darya, and has raised the
lakes water level by 2 meters and expanded
its surface area by some 900 square kilo-
meters (Science, 14 April 2006, p. 183).
During his expedition, Micklin found that
the fresh water had reduced salinity from
12.3 grams of salt per liter in 2005 to
8 grams and increased water clarity. As a
result, not only are fish becoming more
abundant, but aquatic plants and reeds are
spreading fast.
Prior to construction of the dike, only a
hardy ounder introduced from the Black
Sea was able to survive in the northern part
of the lake. Now, about two dozen species of
native freshwater sh that high salinity had
driven into the Syr Daryas delta and adja-
cent lakes have returned to deeper waters
and are reproducing at a rapid clip. And the
rebound has been even bigger than World
Bank planners expected: In the early 2000s,
says the banks Masood Ahmad, a feasi-
bility study had concluded that a revived
Northern Aral would ultimately produce
about 2000 tons of sh a yearbut shing
eets are already catching nearly twice that
amount. To make sure the stocks continue
to ourish, Yermakhanov says he is commit-
ted to limiting yearly catches to about one-
third of the total biomassmuch less than
what is allowed in many managed sheries.
I know we could sh more, he says, but I
want to make sure we can grow the biomass
to at least 40,000 tons.
Kazakhstan, flush with income from
oil and minerals, is now considering tak-
ing the rescue effort a step further. One
project, backed by Yermakhanov and local
fishers, would raise the dike enough to
increase water levels by more than 6 meters,
expanding the lake surface by about 50%, to
5000 square kilometers. That plan would
be the best for the ecosystem, Micklin
says, but it isnt as politically or economi-
cally attractive as an alternative. That would
involve digging a canal to bring water back
to the historic port of Aralsk, which was left
high and dry by the desiccation. The canal
plan would benet more people, Micklin
says. For the moment, however, there isnt
enough water ow for both. The govern-
ment announced last month that it would
appoint an expert panel to decide between
the two options.
CHRISTOPHER PALA
Christopher Pala is a writer based in Almaty, Kazakhstan.
In Northern Aral Sea, Rebound
Comes With a Big Catch
ECOLOGI CAL RESTORATI ON
Aral redux. More freshwater in the Northern Aral
(above, at top) has enabled shers to once again
pick a rich catch out of their gill nets (left).
2004
Tastubek
North Aral Sea
2011
21 OCTOBER 2011 VOL 334 SCIENCE www.sciencemag.org 304
C
R
E
D
I
T
:

L
O
S

A
L
A
M
O
S

N
A
T
I
O
N
A
L

L
A
B
O
R
A
T
O
R
Y
BLACKSBURG, VIRGINIAUnlike old sol-
diers, some scientic concepts seem never to
fade away. Take the hypothetical subatomic
particle called the sterile neutrino, which
would be about the oddest bit of matter imag-
inable. For 15 years, researchers have accu-
mulated hints from particle physics, nuclear
physics, astrophysics, and cosmology that the
particlea more-elusive cousin of the nearly
undetectable neutrinosmight be out there.
But most physicists have found the evidence
unconvincing, as most of the results pointing
toward sterile neutrinos are of marginal statis-
tical signicance.
Recently, however, the case for sterile neu-
trinos has grown stronger, bolstered by a new
analysis of data from nuclear reactors. So last
month 60 physicists from around the world
gathered here
*
to hash out the arguments for
and against the existence of sterile neutrinos
and to try to decide whether its worth stag-
ing a dedicated experiment to settle the matter.
Performing such an experiment wont be
easy. The hypothetical neutrinos are called
sterile because they do not interact at all with
known particles. Youre trying to
prove the existence of something
with no interactions, says Patrick
Huber, a theorist here at Virginia
Polytechnic Institute and State Uni-
versity (Virginia Tech). Its like try-
ing to prove the existence of God.
Still, he says, its time to gure out
what it will take to discover or rule out sterile
neutrinos once and for all. Im afraid well
have the same workshop 15 years from now
and will just have more [inconclusive] results
that dont make the situation any clearer.
Some researchers say the case for a sterile
neutrino is still half-baked. Im quite skepti-
cal, says Yves Dclais, a neutrino physicist
at the University of Lyon in France. Each
piece of evidence itself is not completely
self-consistent, he says. So Im really con-
cerned that there should be more work to
understand each anomaly itself instead of
trying to put together a dedicated experiment
to look for sterile neutrinos.
Abundant in theory
Ordinary neutrinos are already weird. Nearly
massless and hardly interacting with other
matter, they are born in weak nuclear decays
and interactions. For example, a neutron
decays into a proton by emitting an electron
and an antineutrino. A neutrino can emerge
when a nucleus of the isotope beryllium-7
turns into lithium-7 by capturing an electron
and releasing a neutrino. Trillions of neutrinos
stream through each of us every second.
Weirder still, neutrinos come in three
avorselectron neutrinos, muon neutri-
nos, and tau neutrinosthat can morph into
one another. For example, when cosmic rays
strike the atmosphere, they create particles
called muons that decay much as neutrons do,
to produce muon neutrinos. The muon neutri-
nos can then oscillate or mix into other
avors before reaching Earth, as observed in
1998 by physicists using a giant subterranean
detector called Super-Kamiokande in Japan.
Electron neutrinos from the sun also change
avor, as physicists at the Sudbury Neutrino
Observatory in Canada showed in 2001.
A sterile neutrino would be even more elu-
sive than an ordinary neutrino. It would not
participate in weak interactions and would
arise only from ordinary neutrinos oscillating
into a sterile form. As sterile neutrinos would
not interact themselves, physicists could
detect them only indirectly, by observing
ordinary neutrinos disappearing or appearing
where they are not expected.
Theorists have been think-
ing about sterile neutrinos since
the late 1960s, when they rst
suspected that neutrinos from
the sun oscillated. The morph-
ing meant that neutrinos were
not massless, and sterile neutri-
nos would help explain how the
wispy particles put on weight.
Flavor-changing oscillations prove that
neutrinos have mass because a massless par-
ticle must travel at light speed, and, accord-
ing to Einsteins theory of relativity, at light
speed time stands still, making change
impossible. The standard model of particle
Mystery machine. The guts of the LSND detector,
which may have seen sterile neutrinos.
Online
sciencemag.org
Podcast interview
with author
Adrian Cho.
The Sterile Neutrino:
Fertile Concept or
Dead End?
Dozens of physicists gathered recently to debate
whether the phantom particle exists
and if its worth hunting it
NEWSFOCUS
*
Sterile Neutrinos at the Crossroads, 2628 September.
www.sciencemag.org SCIENCE VOL 334 21 OCTOBER 2011 305
NEWSFOCUS
C
R
E
D
I
T
S

(
T
O
P

T
O

B
O
T
T
O
M
)
:

N
A
S
A
/
W
M
A
P

S
C
I
E
N
C
E

T
E
A
M
;

M
.

B
L
A
N
T
O
N

A
N
D

T
H
E

S
L
O
A
N

D
I
G
I
T
A
L

S
K
Y

S
U
R
V
E
Y

(
S
D
S
S
)

C
O
L
L
A
B
O
R
A
T
I
O
N
,

H
T
T
P
:
/
/
W
W
W
.
S
D
S
S
.
O
R
G
physics assumes that neutrinos are massless,
but most extensions of the theory that fix
that problem include sterile neutrinos, says
Paul Langacker, a theorist at the Institute for
Advanced Study in Princeton, New Jersey.
The sterile neutrino is not something bizarre
or exotic, he says.
The details involve another key fact: As
far as physicists know, all neutrinos spiral to
the left, like footballs thrown by left-handed
quarterbacks, and all antineutrinos spiral
to the right. That would be ne if neutrinos
traveled at unobtainable light speed. But as
neutrinos have mass and travel slower, its
possible in principle for an observer to over-
take a left-handed neutrino. The neu-
trino would then appear to travel
the opposite way, as a right-
handed neutrinoa particle
not found in nature.
Theorists have found two
ways around this problem.
First, when overtaken, an ordi-
nary left-handed neutrino could
appear instead as a heavy right-
handed sterile neutrino. Or second, the
overtaken neutrino could appear as a right-
handed antineutrino. Even then, a theoreti-
cal seesaw mechanism would require heavy
sterile neutrinos to explain why ordinary neu-
trinos are so light.
Thats plenty of reason to think sterile neu-
trinos are out there. However, most theories
assume that sterile neutrinos are far heavier
than ordinary neutrinos. For that reason and
others, theory doesnt generally allow an ordi-
nary neutrino to just morph into a sterile neu-
trino in the way some experiments indicate,
Langacker says. So its not clear that theorists
and experimenters are stalking the same beast.
Evidence of all sorts
The strongest experimental evidence for ster-
ile neutrinos comes from the Liquid Scintil-
lator Neutrino Detector (LSND), which ran
at Los Alamos National Laboratory in New
Mexico from 1993 through 1998. Using a par-
ticle accelerator, physicists generated muon
antineutrinos that streamed through a detector
lled with 167 tons of mineral oil.
Those low-energy muon antineutrinos
should have passed right through. However,
an electron antineutrino could interact with
the detector by merging with a proton to cre-
ate a positron and a neutronessentially, the
weak decay of the neutron run backward.
Thus, physicists could spot electron anti-
neutrinos appearing in a beam of muon anti-
neutrinos. And they spotted 88 of them, give
or take 23. Lo and behold, we saw an excess
of events, says Los Alamoss William Louis.
Reported in 1996 and 2001, the LSND
results might seem to show muon anti-
neutrinos mixing into electron antineutrinos.
But it couldnt be that simple, Louis says. Dif-
ferent avors of neutrinos mix at a rate that
depends on the difference in their masses:
The bigger the mass difference, the faster
the mixing. Studies of atmospheric and solar
neutrinos had placed limits on the mass dif-
ferences among the three neutrino avors, and
the values were too low to explain the lickety-
split mixing that LSND saw as the particles
ew just 30 meters, Louis says.
LSND researchers could explain their
results, however, if muon antineutrinos oscil-
lated rst into sterile antineutrinos and then
into electron antineutrinos. The sterile neu-
trinos would have to be heavier than ordinary
neutrinos by 1 electron voltabout 100 times
the differences among ordinary neutrinos.
Hints of extra neutrinos also come from
the heavens. For example, cosmologists
think the universe burst into existence in the
big bang as an ultrahot, ultradense soup of
particles. Tiny uctuations in the density of
the soup then stretched to immense propor-
tions during a faster-than-light growth spurt
known as ination and seeded the formation
of galaxies. The uctuations also limit the
number of neutrino types, theorist Kevork
Abazajian of the University of California,
Irvine, said at the meeting.
The density uctuations can be thought
of as waves of various wavelengths randomly
piled on one another. The uctuations inten-
sity and gravitational pull grow stronger
as their wavelengths decreases; then the
intensity peaks at a certain wavelength and
starts to fall again. The position of that peak
in a graph of intensity versus wavelength
depends on the relative amounts of radiation
and matter in the early universe. And because
lightweight neutrinos acted like an addi-
tional form of radiation, the position
of the peak also reveals the number
of types of light neutrinos.
To deduce the distribution
and the peak in it, scientists
measure tiny variations in the
afterglow of the big bangthe
cosmic microwave background
radiationacross the sky, as
NASAs space-borne Wilkinson
Microwave Anisotropy Probe did from
2001 to 2010. Scientists also measure the dis-
tribution of the galaxies, as the Sloan Digi-
tal Sky Survey has done using a telescope at
the Apache Point Observatory in New Mex-
ico. The results suggest a fourth neutrino,
Abazajian says. Theres a 1-in-20 chance that
its a statistical uctuation, and those sorts of
things go away all the time, he says. Still,
its intriguing that the datas converging to that
value of four types of neutrinos.
The newest bit of evidence comes from
nuclear reactors. Earlier this year, a team of
theorists argued that reactors are putting out
more electron antineutrinos than detectors
tens of meters away show. The result suggests
that some antineutrinos escape detection by
morphing into sterile antineutrinos.
Within a nuclear reactor, nuclei of
the isotopes uranium-235, uranium-238,
plutonium-239, and plutonium-241 split
randomly to make myriad smaller nuclei
that release copious antineutrinos. For
example, a uranium-235 nucleus can split to
make a nucleus of krypton-89. Krypton-89
then changes identity to rubidium-89,
strontium-89, and yttrium-89, as one neu-
tron after another in the nucleus spits out
an electron and an antineutrino and turns
into a proton. Thousands of other chains or
branches of decays also occur.
David Lhuillier of Frances Alternative
Energies and Atomic Energy Commission
in Saclay and colleagues kept track of all
those branches in a new calculation. Previ-
ous calculations showed that a score of reactor
Far out! Cosmic microwave background (top) and
the map of the galaxies hint at an extra neutrino.
21 OCTOBER 2011 VOL 334 SCIENCE www.sciencemag.org 306
NEWSFOCUS
C
R
E
D
I
T
:

A
R
M
E
L

B
R
U
C
E
L
L
E
/
S
Y
G
M
A
/
C
O
R
B
I
S
measurements taken over decades observed
97.6%, give or take 2.4%, of the expected
antineutrino uxne agreement with the
prediction. With the new calculation, the
measurements average 94.3%, plus or minus
2.3%a signicant difference that suggests
neutrinos are disappearing. Before, all the
experiments were in agreement with the pre-
diction, Lhuillier says. Now everybody is
below the prediction. Physicists say this sin-
gle result triggered the workshop.
The disparate hints are tantalizing, says
Joseph Formaggio of the Massachusetts Insti-
tute of Technology in Cambridge. Whats
nice is that these anomalies come from differ-
ent directions, he says.
Signs of discord
Each clue comes with caveats, however.
For example, starting in
2002, physicists tested
the LSND result with the
Mini-Booster Neutrino
Experiment (MiniBooNE)
at Fermi National Accel-
erator Laboratory (Fermi-
lab) in Batavia, Illinois.
First, they red muon neu-
trinosinstead of LSNDs
muon antineutrinos450
meters into a detector
filled with 800 tons of
mineral oil. In 2007, they
saw signs of electron neu-
trinos appearing in the
muon neutrino beam, but
with the wrong energy to
mirror the process seen in LSND with anti-
neutrinos. The result dampened enthusiasm
for sterile neutrinos.
But last year, the researchers reported that
using muon antineutrinos, they see electron
antineutrinos appearing as LSND did, albeit
at lower statistical signicance. The excess
in the MiniBooNE antineutrino data agrees
beautifully with what you would expect from
LSND, says Louis, who also works on Mini-
BooNE. But it also makes matters more com-
plicated. To explain why the effect appears
only for antineutrinos, physicists need to add
two sterile neutrinos to their theory.
The cosmological evidence for sterile
neutrinos also comes with qualications, says
Yvonne Wong of RWTH Aachen University
in Germany. The unknown particles scien-
tists might be glimpsing in cosmic radiation
are signicantly lighter than the sterile neu-
trinos hinted at by LSND and MiniBooNE.
They arent even necessarily true neutrinos,
Wong says, but could be any feebly inter-
acting particle.
Even the newfound reactor anomaly has
not bowled skeptics over. Petr Vogel, a theo-
rist at the California Institute of Technology in
Pasadena, who worked on the original reactor
calculations 30 years ago, says the new cal-
culations are undoubtedly more thorough and
realistic than the old ones. However, they still
leave out important details that might make
the falloff in neutrinos less impressive, Vogel
says. I think what has been done is state of the
art, and the shift [in the prediction] looks rea-
sonable to me, Vogel says. But whether the
error is really 2.3% remains to be seen.
Finally, the signs of sterile neutrinos may
not agree with one another, says Thomas
Schwetz-Mangold of the Max Planck Institute
for Nuclear Physics in Heidelberg, Germany,
who presented a global t to all the data. In
particular, if ordinary neutrinos quickly oscil-
late into sterile neutrinos, then experiments
that send muon neutrinos to distant detectors
should see a decrease in the total number of
neutrinos reaching their detectors. But experi-
ments such as Fermilabs Main Injector Neu-
trino Oscillation Search, which res neutrinos
735 kilometers to a detector in Minnesota, see
no such loss. If I take everything at face value,
then the probability is less than a percent that it
all ts together, Schwetz-Mangold says.
The killer experiment
In spite of the odds, some experimenters are
still eager to hunt sterile neutrinos. Plans vary
widely, but physicists generally agree on what
a killer experiment must do. If ordinary neu-
trinos morph into sterile neutrinos and back,
then the number of ordinary neutrinos in a
beam should go up and down as the neutri-
nos y away from their source. So scientists
would have to spot that spatial oscillation over
tens of meters.
The easiest way would be to add a sec-
ond detector to the MiniBooNE experiment
closer to the neutrino source or to move the
existing detector. The rate at which electron
antineutrinos appear should then change.
Building a second detector 200 meters from
the source would cost $10 million, Geoffrey
Mills, a MiniBooNE team member from
Los Alamos, said at the conference. Alter-
natively, researchers could move the current
detector for about $5 million, he reported.
Adding the second detector to MiniBooNE
is a must-do, some researchers say. But
Roxanne Guenette of Yale University warned
that a denitive measurement would likely
take two more-expensive new detectors.
Others want to look for the oscillation of
electron neutrinos by putting an intense radio-
active source inside a jumbo detector. The
number of electron neutrino detections should
go up and down as the distance within the
detector from the source
increases. Virginia Techs
Jonathan Link proposes
placing a chromium-51
source in the center of the
Sudbury Neutrino Obser-
vatory, which is a sphere
filled with 1000 tons of
heavy water.
The source should cost
less than $3 million, Link
says. I do believe that this
is the cheapest option that
has some chance of mak-
ing some sort of statement
about the LSND-type ster-
ile neutrino, he says. Gio-
acchino Ranucci of Italys
National Institute of Nuclear Physics in Milan
presented a proposal to place a source under
the 270-ton Borexino detector in Italys sub-
terranean Gran Sasso National Laboratory.
Hanging over all of this is the question of
money, as the United States particle phys-
ics budget has been stuck at $800 million for
years. Virginia Techs Ramaswamy Raghavan
is developing the Low Energy Solar Neutrino
Spectrometer detector, which would study
solar neutrinos and, with a radioactive source,
could look for sterile neutrinos. It would cost
$50 million to $75 million. Can you predict
in the current scal situation in the U.S. that
this is going to happen? Raghavan says.
To help make the case for funding, con-
ference attendees plan to write a white paper
laying out the options. Theres some urgency,
says Huber, the Virginia Tech theorist who
helped organize the meeting. I dont want to
do sterile neutrinos my whole career, he says.
He doesnt say whether the ephemeral beast
will continue to entice him if no denitive
answer is quick in coming. ADRIAN CHO
MIA. New calculations suggest that nuclear reactors
put out more neutrinos than are observed.
www.sciencemag.org SCIENCE VOL 334 21 OCTOBER 2011 307
NEWSFOCUS
C
R
E
D
I
T
:

P
A
N
O
S

I
P
E
I
R
O
T
I
S

&

D
A
H
N

T
A
M
I
R
,

N
E
W

Y
O
R
K

U
N
I
V
E
R
S
I
T
Y
/
G
O
O
G
L
E

F
U
S
I
O
N

T
A
B
L
E
S
Its a problem that all social scientists face.
You have a brilliant idea for a study. You have
the experimental design all worked out, and
your universitys review board has approved
it. But you still have to recruit hundreds of
people as subjects for the experiment.
Gabriel Lenz, a political scientist at the
University of California, Berkeley, faced this
problem last year when he and collaborators
wanted to follow up on another groups study
of voting behavior (Science, 10 June 2005,
p. 1623). For that study, Americans were
shown photographs of past U.S. congressional
candidates and asked to rate the politicians on
various characteristics, such as competence
and attractiveness. Even though the study
subjects had no information beyond an image
of the candidates faces, their snap judgments
were a signicant predictor of who actually
won the races. Lenz wanted to see if that sur-
prising result collapsed when those evaluating
the photos come from cultures different from
those of the candidates. But how to recruit
people living in multiple countries?
Lenz and his research assistant Michael
Myers had an idea: Why not order research
subjects through Amazon.com? The company
runs an online marketplace called Mechani-
cal Turk for people across the world avail-
able to do work on computers. (The name is
a reference to an 18th century chess-playing
machine that actually worked by virtue of
a man hidden inside.) For tiny sums, anyone
can hire people to perform almost any kind of
simple task, such as tagging items in images.
Lenzs experiment required people to look at
photographs of Brazilian political candidates
and ll in a data sheet.
But first, he and his colleagues had to
decide on how much they would pay each
participant. Those offering a job through
MTurk, known as requestors, compete with
each other to recruit Turkers, the 500,000
people currently registered with the MTurk
site as available for work. The task of rating
the political candidate photos required about
4 minutes. We played around with various
payment rates, Lenz says. For Turkers based
in India, the researchers started low, offering
15 cents. In just 4 days, they received data
from 100 people. Then for a control group,
they recruited more than 300 Americans for
between 20 and 50 cents each. The total cost?
About $160, and that includes the 10% fee
Amazon charges.
In just a few weeks, Lenz had all the data
his group needed. In spite of the cultural dif-
ferences, the snap-judgment effect persisted:
American and Indian subjects predicted the
winners of Brazilian political races based on
nothing more than a mug shot, the research-
ers reported last year in the social science
journal World Politics.
As others follow Lenzs lead, many more
social science papers using MTurk will appear
in the coming years, predicts Adam Berinsky,
a political scientist at the Massachusetts Insti-
tute of Technology in Cambridge. Everyone I
know is using it, he says. For example, social
scientists used 10,000 Turkers to create a tool
for tracking the emotional content of Twitter
messages (Science, 30 September, p. 1814).
For now, most researchers are using MTurk
for pilot studies, quickly and cheaply testing
online versions of experiments that they then
perform with subjects face to face. But the
use of MTurk subjects will eventually become
mainstream, Berinsky says. The obvious
advantage is the speed and cost. Generally,
we pay $8 for a 15- to 20-minute experiment
in a lab. We can run the same study on MTurk
for 75 cents to a dollar.
There are other advantages. Turkers
are amazingly focused research subjects,
Berinsky says. Unlike the typical univer-
sity undergraduates used for social science
studies, Turkers get paid only if they
generate usable data. This is neces-
sary to eliminate not only people who
dont understand the task but also
spammers, people who try to exploit
MTurk by skimming through the jobs
and giving random responses wherever
possible to accelerate the process.
For example, Lenz had to reject
about 20% of his American and 50%
of his Indian Turkers for those rea-
sons. But that is a manageable problem,
Berinsky says. A counterintuitive solution is
to keep the price low. If you offer more than a
dollar, you attract the spammers who sort jobs
by level of pay, he says. You have to nd the
sweet spot where the payment is not too high
but still attractive enough for most Turkers.
So far, that sweet spot seems to be between
15 and 50 cents for a 10-minute job.
Even if MTurk is cheap and fast, doubts
will linger about interpreting data from
research subjects whom you never meet. To
address those concerns, Berinsky and Lenz
are teaming up with Gregory Huber, a politi-
cal scientist at Yale University, to study the
Turker population. And of course, they are
using MTurk to do so. They recently repli-
cated two classic survey experiments and a
political science experiment. In each case,
the data obtained with MTurk were consis-
tent with published studies that tested people
in laboratories.
The scientists have found some differ-
ences, too. Turkers are younger and more
ideologically liberal than the U.S. public,
Berinsky says. However, they are more repre-
sentative of the U.S. population than a typical
cohort of university undergraduates.
There is one long-term concern: the
super-Turkers, people who are essentially
professional workers on MTurk, some of them
logging more than 20 hours per week. Many
social science experiments rely on the sub-
jects not knowing the researchers intentions.
Berinsky says super-Turkers could potentially
skew experiments if they try too hard to please
researchers. There is incentive to do that
because MTurk uses a reputation system. If a
Turker does not have at least a 95% positive
approval rating from their requestors, theyll
often go unhired.
Mechanical Turk seems like the prover-
bial goose that lays the golden eggs, Berinsky
says. But I worry that in the rush for cheap
research subjects, were going to trample the
goose to death. JOHN BOHANNON
Global pool. This map shows a 10% sample
of workers (red) available on Amazon.coms
Mechanical Turk.
Social scientists are turning to online retail giant Amazon.com to cheaply recruit
people around the world for research studies
Social Science for Pennies
HUMAN SUBJECT RESEARCH
21 OCTOBER 2011 VOL 334 SCIENCE www.sciencemag.org 308
C
R
E
D
I
T
S
:

(
T
O
P

L
E
F
T

T
O

R
I
G
H
T
)

J
O
S
L
I
N

M
O
O
R
E
;

C
H
E
N
G
J
I
N

C
H
U
;

(
B
O
T
T
O
M
)

E
L
I
Z
A
B
E
T
H

B
O
R
E
R
In 2005, a handful of young researchers in
Santa Barbara, California, were fed up with
their inability to answer a major ecological
question by reviewing the literature. So they
decided to take matters into their own hands
and created a network of small experiments.
In the past 6 years, the network has spread
to six continents and is now poised to make
substantial contributions to ecology. Were
on the edge of something big, says John
Orrock of the University of Wisconsin, Mad-
ison, a network co-founder.
The half-dozen Ph.D. students and post-
docs were part of a workshop at the National
Center for Ecological Analysis and Synthesis
(NCEAS) in Santa Barbara. The group was
investigating fundamental inuences on the
structure of grasslands,
such as herbivory and
nutrients. Trying to ana-
lyze data from far-ung
places, the group was
stymied by a common
obstacle. Its really
fr ustrating because
everyone does their
studies differently, says
Elizabeth Borer, who is
now at the University of
Minnesota, Twin Cities.
During a coffee
break at NCEAS, Borer
and a few others hatched a plan: They would
each set up a small research plot, use the same
methods, then pool their data. The vision was
a network of sites that would be quick and
cheap to set up without the need for major
grants, enabling simple experiments around
the world. Its like big science on a shoe-
string, says Scott Collins of the University
of New Mexico, Albuquerque, who later
joined the network.
The collaboration, called the Nutrient
Networknow known as NutNethas
grown far beyond initial expectations, with
scientists volunteering at 68 sites in 12 coun-
tries. In part, its popular because the simple
experiments are designed to answer a broad
set of questions about how grasslands respond
to global changewithout disproportionate
effort by any one individual. Its not a brand-
new idea, but its novel that theyve pulled it
off, says Alan Townsend of the University of
Colorado, Boulder, who is not involved. The
network also provides an easy way for young
faculty members, postdocs, and grad students
to get involved in a large collaboration and
contribute to high-prole papers.
So far, the effort has been funded with
just a single $322,000 grant from the U.S.
National Science Foundation (NSF) for coor-
dinating data and analysis, yet already the
rst few papers have been published over the
past year. The most recent, which appeared in
Science last month (23 September, p. 1750),
challenged a long-standing idea in ecology
about plant diversity and productivity. Doz-
ens more papers are in the works, and ecol-
ogists enthuse about the networks potential
for cost-effective, rapid results. NutNet has
tremendously improved on the way weve
done things, says Alan Knapp of Colorado
State University, Fort Collins, another ecolo-
gist who is not involved. Ive been incred-
ibly impressed.
Keep it simple
Research networks arent new to ecology, of
course. The Long Term Ecological Research
(LTER) network, for example, is composed
of 26 research sites and stations, almost all in
the United States, that have been collecting
data for 30 years. And construction began
this fall on some of the 20 U.S. observatories
that will make up the $434 million National
Ecological Observatory
Network. These hefty
networks require a fair
amount of money to
operate, because staff
members collect hun-
dreds of types of data,
often year-round.
During the NCEAS
workshop, NutNets
f ounde r s qui c kl y
sketched an alternative
vision: Each researcher
would conduct the
same few experiments
Open-Source Ecology Takes Root
Across the World
A new collaboration of volunteer research sites is running simple yet powerful
experiments to shed light on global change in grasslands
NETWORK SCI ENCE
Experimental
Observational
only
Diversity. NutNet sites include 1747 plant taxa in many ecosystems, such as (see photos, left
to right) subalpine grassland, alpine meadow, desert, pasture, sagebrush steppe, and savanna.
Standardized. Researchers worldwide add nutri-
ents and measure plots the same way.
China
Australia
www.sciencemag.org SCIENCE VOL 334 21 OCTOBER 2011 309
NEWSFOCUS
C
R
E
D
I
T
S
:

(
T
O
P

L
E
F
T

T
O

R
I
G
H
T
)

S
C
O
T
T

C
O
L
L
I
N
S
;

Y
A
N
N

H
A
U
T
I
E
R
;

P
E
T
E
R

A
D
L
E
R
;

T
.

M
I
C
H
A
E
L

A
N
D
E
R
S
O
N
;

(
M
I
D
D
L
E
)

T
I
M

R
U
M
M
E
L
H
O
F
F
;

(
B
O
T
T
O
M
)

C
O
U
R
T
E
S
Y

O
F

S
T
A
N

H
A
R
P
O
L
E
in several plots of 25 square meters. They
would add combinations of three crucial
plant nutrientsnitrogen, phosphorus, and
potassiumand they would fence part of
the plots to exclude deer, zebras, kangaroos,
and other herbivores.
By measuring changes in biomass and spe-
cies composition, they would try to tease apart
the relative impact of herbivores and nutrients
on the structure of the com-
munity. Ecologists have been
fascinated by this question
for a long time, Orrock says.
Moreover, the experiments
simulate the impacts of anthro-
pogenic global change. Nutri-
ent levels have been boosted
dramatically by fertilizers and
pollution from fossil fuels. At
the same time, humans have
altered the density of herbi-
vores in many places through
farming or indirectly by hunt-
ing of predators.
Several attendees at the
NCEAS workshop immedi-
ately volunteered to partici-
pate. One of the rst was Helmut Hillebrand
of the Carl von Ossietzky University of Old-
enburg in Germany, who set up a NutNet
site, even though hes a plankton ecologist.
I think its the next generation of ecologi-
cal experiments, he says. The site he started
is located in an old eld 5 minutes from his
parents house, so he drops by to collect data
while visiting.
Borer and the others also invited a few
colleagues to join, and the idea began to
spread by word of mouth. Sensing potential,
the group sent an e-mail in November 2006
to just about every grassland ecologist they
knew. By the time data started arriving the
next year, there were 51 sites.
Members of the network agree to submit
data immediately to a central database. All
participantsnow about 100, including a
dozen or so graduate studentshave access
to the data. Simply by contributing data, they
can be an author on high-prole papers that
address the projects big questions. The net-
work is already making a mark: Last months
paper in Science showed that a textbook
idea about the relationship between plant
productivity and species richness in fact
occurs rarely. Other key papers, based on the
experimental results of adding nutrients and
excluding herbivores, are still being written.
NutNet participants must propose papers
on additional ideas to the whole group.
The goal is to avoid duplication and allow
other members to contribute to analysis or
writing the manuscript. Jennifer Firn of
the Queensland University of Technology
in Brisbane, Australia, for
example, wanted to look at
invasive species in the plots.
The process of turning this
idea into a paper was the best
learning experience I have
ever had, says Firn, who
became an assistant professor
in February. I had more than
30 authors and co-authors,
so it meant so much advice
and expertise were available.
Published in Ecology Letters
in March, the paper showed
that non-native plants, some
invasive, dont all spread
like the worst weeds. Instead,
most species in the NutNet
plots were about as common in their new
environment as in their native range. That
suggests that regulators of plant imports
might want to focus on screen-
ing out plants that are highly
abundant overseas.
Network members decide
among themselves what kinds
of additional data to gather.
This is like an indie garage
band, a cooperative without
all the top-down headaches,
says co-founder W. Stanley
Harpole, an assistant pro-
fessor at Iowa State Univer-
sity in Ames. (Others make
analogies to the development
of open-source software or
start-up companies.) Eighteen
members are analyzing regular
deliveries from other participants, who col-
lect everything from soil microbes to arthro-
pods and leaf litter. It is simple, mail-order
sampling, says co-founder Eric Seabloom
of the University of Minnesota, Twin Cities.
The person in the eld doesnt have to do
that much.
Facing the future
An all-volunteer approach may have its limi-
tations, however. So far, the majority of sites
are in the United States. Peter Adler of Utah
State University in Logan, a co-founder, says
the group tried to recruit scientists in South
America without much success. Maybe its
just [bad] luck, he says. Townsend expects
that more researchers in less developed coun-
tries will eventually sign up, as word spreads
about the network and its publications. Ear-
lier this month, several sites in India agreed
to provide observational data, and a few
more will also conduct experiments.
A larger question is how long a volun-
teer effort can be sustained. In absence of
external funding, I fear that the good will of
those individuals and their institutions may
not persist, says Michael Willig of the Uni-
versity of Connecticut, Storrs, who is not a
participant in the network. But co-founder
Melinda Smith of Yale University predicts
that interest will remain high as long as
the network produces high-impact papers.
Harpole points out that each plot has space
reserved for experiments not yet planned.
Were banking for the future, he says.
The looming danger is the expiration
of the NSF grant in January 2013. These
funds pay for collaboration meetings and
for a postdoc, Eric Lind of
the University of Minnesota,
Twin Cities, who runs the cen-
tral database. The death of
the Nutrient Network will be
when the funding for that post-
doc position runs out, Adler
says. The steering committee
hopes to cover those expenses
with future research grants for
more ambitious analyses.
Even if the NutNet peters
out, the founders hope it will
be a model. To Borer, the suc-
cess so far shows that individ-
ual scientists at any stage of
their career can help answer
big questions even if they havent landed a
major grant. Were out to change the cul-
ture, she says. The success of this model
could empower other groups to address
equally important ecological problems at a
global scale.
ERIK STOKSTAD
Were out to
change the culture.
ELIZABETH BORER,
UNIVERSITY OF MINNESOTA,
TWIN CITIES
This is like an indie
garage band.
STAN HARPOLE,
IOWA STATE UNIVERSITY
USA USA Switzerland Tanzania
21 OCTOBER 2011 VOL 334 SCIENCE www.sciencemag.org 310
LETTERS
edited by Jennifer Sills
LETTERS I BOOKS I POLICY FORUM I EDUCATION FORUM I PERSPECTIVES
312
A multiple personality fable
Bending light with
antenna arrays
317
COMMENTARY
Martial Arts Research:
Prudent Skepticism
A. DI AMOND AND K. LEE S REVI EW
Interventions shown to aid executive func-
tion development in children 4 to 12 years
old (special section on Investing Early in
Education, 19 August, p. 959) leaves the
impression that martial arts training as usu-
ally delivered enhances executive functions.
This is far from established. Martial arts
training is a heterogeneous independent vari-
able with average effects that may be negli-
gible or even negative.
Diamond and Lee cite two studies in sup-
port of martial arts. In the Trulson study (which
was based on 34 students and 1 instructor),
the only outcome measures are the self-report
personality inventories completed by the
delinquent students (1). Trulson concluded
that the meditation, contemplation of goals,
and other noncombat components of martial
arts are helpful, but pure competitive ght
training is harmful. The Lake and Hoyt study
(207 students and 1 instructor) found the most
positive effects on a measure of behavior dur-
ing completion of an obstacle course (2). With
teacher ratings, however, insignicant effects
were reported for four out of ve variables,
including self-control.
Longitudinal studies observing the results
of many instructors lead to skepticism about
the effects of martial arts training. Endresen
and Olweus (3), using a longitudinal design,
reported that participation in power sports
Editorial Expression of Concern
IN THE 4 JUNE 2010 ISSUE, SCIENCE PUBLISHED THE REPORT SPHK1 REGULATES PROINFLAM-
matory responses associated with endotoxin and polymicrobial sepsis by P. Puneet et al. (1).
After the receipt of an anonymous e-mail on 22 March 2011, Science learned that authori-
ties at the authors principal institutions at the time of publication (University of Glasgow
and National University of Singapore) and the University of Liverpool (corresponding author
A.J.M.s more recent afliation) were investigating allegations of gure manipulation in the
Science Report and in a paper published in Nature Immunology [Nature Immunology 12, 344
(2011)] also by P. Puneet et al. The Nature Immunology paper was subsequently retracted after
an investigation by the University of Liverpool, but we have been informed that the investiga-
tion into the Science Report has not yet reached a conclusion, despite indications that it was
near completion.
On 14 January 2011, Science published a Correction to two of the gures in the Puneet et
al. Report, after correspondence with A.J.M. In light of the continuing investigation, we can no
longer be condent in the reliability of the corrected record. Pending the results of the investiga-
tions, Science is publishing this Editorial Expression of Concern to alert our readers to the fact
that serious questions have been raised about the validity of ndings in the Puneet et al. paper.
BRUCE ALBERTS
Editor-in-Chief
Reference
1. P. Puneet, C. T. Yap, L. Wong, L. Yulin, D. R. Koh, S. Moochhala, J. Pfeilschifter, A. Huwiler, A. J. Melendez, Science 328, 1290
(2010).
Published online 3 October 2011; 10.1126/science.1214735
[including martial arts] actually leads to an
increase or enhancement of antisocial involve-
ment in the form of elevated levels of violent
as well as non-violent antisocial behavior out-
side sports. We analyzed data from a large,
nationally representative sample (4). The out-
come variable was teacher-rated behavior,
including self-control and attention. In each of
our two main outcome analyses, we found that
martial arts had no effect on behavior.
In a world beset by violence, there is irony
and pathos in hoping that our children will be
improved by teaching punching, kicking, and
tripping. Unless the evidence for benet is
robust, it is prudent to be skeptical.
JOSEPH M. STRAYHORN
1
* AND
JILLIAN C. STRAYHORN
2
1
Department of Psychiatry, Drexel University College of
Medicine, Philadelphia, PA 19129, USA.
2
Undergraduate,
Department of Psychology, Cornell University, Ithaca, NY
14853, USA.
*To whom correspondence should be addressed. E-mail:
joestrayhorn@gmail.com
References
1. M. E. Trulson, Hum. Relat. 39, 1131 (1986).
2. K. D. Lakes, W. T. Hoyt, Appl. Dev. Psychol. 25, 283 (2004).
3. I. M. Endresen, D. Olweus, J. Child Psychol. Psych. 46,
468 (2005).
4. J. M. Strayhorn, J. C. Strayhorn, J. Child Adolesc. Psychiatr.
Ment. Health 3, 32 (2009).
Martial Arts Research:
Weak Evidence
THE REVIEW INTERVENTIONS SHOWN TO
aid executive function development in chil-
dren 4 to 12 years old by A. Diamond and
K. Lee (special section on Investing Early in
Education, 19 August, p. 959) cited work that
close examination shows to be weak. Some of
the studies (1, 2) were randomized, but they
failed to meet other criteria such as blind-
ing of teachers and parents to pupils treat-
ment groups. Studies involving martial arts
and physical exercise were particularly weak
on isolation of variables. One study on mar-
www.sciencemag.org SCIENCE VOL 334 21 OCTOBER 2011 311
318 321
Signs of ocean
eddies
Information capacity
in cell signaling
C
R
E
D
I
T
:

S
T
U
A
R
T

M
O
N
K
/
I
S
T
O
C
K
P
H
O
T
O
.
C
O
M
tial arts training for
children (1) com-
pared a treatment
group who wore spe-
cial uniforms, medi-
tated, bowed to their
instructor, and were
reminded of self-
awareness and self-
control, to a control
group who continued
with their ordinary
physical education
activities; these authors concluded that when
some improvement on some scales occurred
for the treatment group, the change was
caused by the self-awareness and self-control
messages, rather than by other ways the two
groups differed. Another study (2) compared
children who did sport stacking, a bimanual
physical task, with a control group that did not
experience any exciting new activity, and con-
cluded that improvement on one of two read-
ing measures was caused by the stacking task.
A relevant volume dealing with treatments
for developmental disabilities (3) stressed the
weakness of evidence for special education
interventions and described some such con-
ditions as fad magnets. Unfortunately, early
educational interventions seem to be similarly
weak in evidence. The stakes are high and the
resources scarce in both cases. JEAN MERCER
Richard Stockton College, Pomona, NJ 08240, USA. E-mail:
jean.mercer@stockton.edu
References
1. K. D. Lakes, W. T. Hoyt, Appl. Dev. Psychol. 25, 283 (2004).
2. T. A. Uhrich, R. L. Swalm, Percept. Mot. Skills 104, 1935
(2007).
3. J. W. Jacobson, R. M. Foxx, J. A. Mulick, Controversial
Therapies for Developmental Disabilities (Erlbaum,
Mahwah, NJ, 2005).
Response
WE AGREE WITH STRAYHORN AND STRAYHORN
that modern and traditional versions of mar-
tial arts differ. We tried to emphasize that
modern American martial arts (which empha-
size punching and kicking and competi-
tion) have been found to make unproductive
behaviors worse, whereas evidence indicates
that traditional martial arts [which emphasize
self-control, self-defense, patience, waiting
for the other person to make an error, con-
centration, respect, and humility (1)] improve
executive functions.
We agree with Mercer about weaknesses
in many studies thus far published on execu-
tive function interventions. We reviewed only
peer-reviewed studies and provided detailed
information about them (see tables S1 to S3 in
the supporting online material) to give read-
ers an opportunity to judge the evidence for
themselves. We disagree with Mercer about
the martial arts study being particularly weak.
First studies are designed to determine whether
there is an overall difference. Follow-up stud-
ies can then try to dissect which aspect(s) of
a program had the most effect. That said, the
martial arts study by Lakes and Hoyt (1) is to
be commended. It used random assignment,
pre- and post-testing, an intervention imple-
mented during regular school hours (making
it feasible to reach many children), an active
control group that also engaged in physical
activity, and incrementally increasing levels
of difculty in the martial arts condition, and
it provided evidence that executive-function
improvements generalized to multiple con-
texts. Unlike many studies that have targeted
disadvantaged children and/or those behind
on executive function, children in this study
were socioeconomically advantaged, making
the ndings especially impressive.
ADELE DIAMOND* AND KATHLEEN LEE
Department of Developmental Cognitive Neuroscience,
University of British Columbia, Vancouver, BC V6N 3L6, BC,
Canada.
*To whom correspondence should be addressed. E-mail:
adele.diamond@ubc.ca
Reference
1. K. D. Lakes, W. T. Hoyt, Appl. Dev. Psychol. 25, 283
(2004).
Letters to the Editor
Letters (~300 words) discuss material published in
Science in the past 3 months or matters of gen-
eral interest. Letters are not acknowledged upon
receipt. Whether published in full or in part, Let-
ters are subject to editing for clarity and space.
Letters submitted, published, or posted elsewhere,
in print or online, will be disqualied. To submit a
Letter, go to www.submit2science.org.
CORRECTIONS AND CLARIFICATIONS
Review: Interventions shown to aid executive
function development in children 4 to 12 years
old by A. Diamond and K. Lee (special section
on Investing Early in Education, 19 August, p.
959). The journal cited in reference 28 should
have been Appl. Dev. Psychol.
Education Forum: Mathematics teachers
subtle, complex disciplinary knowledge by
B. Davis (24 June, p. 1506). The number line
on the right in part C of the gure was mis-
numbered. The correct panel is shown here.
TECHNICAL COMMENT ABSTRACTS
Comment on How Cats Lap: Water Uptake by Felis catus
Michael Nauenberg
Reis et al. (Reports, 26 November 2010, p. 1231) reported on the mechanism by which cats lap and gave a theo-
retical and experimental analysis of their observations. Their explanation for the cats lapping frequency, however,
is based on an incorrect application of the principles of uid dynamics. The revised analysis given here agrees with
their observations and predicts a similar lapping frequency for cats and dogs.
Full text at www.sciencemag.org/cgi/content/full/334/6054/311-b
Response to Comment on How Cats Lap: Water Uptake by Felis catus
Roman Stocker, Jeffrey M. Aristoff, Sunghwan Jung, Pedro M. Reis
We return to the physics of cat lapping to show that our proposed scaling analysis predicts the functional depen-
dencies revealed by the experimental data more accurately than a recently proposed alternative description by
Nauenberg. Experimental verication of functional dependencies, rather than single numerical values, represents
the appropriate test for any scaling argument.
Full text at www.sciencemag.org/cgi/content/full/334/6054/311-c
C
+4
+2
+1
0
1
2
3
4
+3
+8
+4
+2
0
2
8
6
4
+6
Factor a (multiplier):
adjustment ratio
Factor b (multiplicand):
starting position
Product c:
corresponding position
on the adjusted number
line
BOOKS ET AL.
21 OCTOBER 2011 VOL 334 SCIENCE www.sciencemag.org 312
C
R
E
D
I
T
:

(
1
)
/
I
M
A
G
E

C
O
U
R
T
E
S
Y

B
E
N

H
A
R
R
I
S
I
n 1972, incorporation papers for Sybil,
Incorporated, were drawn up by a patient,
her therapist, and a journalist who had
just nished writing a book about her case.
Flora Rheta Schreibers Sybil
( 1) told the story of a young
woman who was cured of
having her consciousness
split into 16 separate person-
alities. The corporations pur-
pose was to share the prots
from the book, a planned lm,
and spin-offs that included
T-shirts, dolls, a jigsaw
puzzle, and a board game.
Although the corporation
did not last and the ancillary
products did not materialize, the book sold
millions of copies. It was dramatized twice
in made-for-television movies. In the rst,
Sybils psychiatrist was played by Joanne
Woodward, who had previously won the
Academy Award for Best Actress for her per-
formance in the role of a patient with mul-
tiple personality disorder in Nunnally John-
sons 1957 lm The Three Faces of Eve.
The most notable product of Sybil, Inc.s
principals was the epidemic of multiple
personality disorder that swept the United
States in the 1970s and 1980s. Unlike most
previous cases, these late-20th-century
variants featured personalities stuck in
various stages of childhood. The patients
alters revealed memories of childhood
sexual abuseoften horricelicited by
hypnosis. Today, researchers and clinicians
are skeptical about most cases of multiple
personality disorder, which has been recon-
ceptualized and renamed dissociative iden-
tity disorder.
Over the past two decades, revelations
have emerged to raise doubts about the
validity of the case of Sybil. First, a clinician
who had met the patient disputed the diag-
nosis promoted in the book. Then, tapes left
by the books author suggested that she col-
luded with the therapist to create symptoms
to t their diagnosis.
In Sybil Exposed, journalist Debbie
Nathan chronicles the rise and fall of Sybil
as the paradigm-setting case of multiple per-
sonality disorder. She does so in three inter-
twined biographies, beginning with that of
the patient, Shirley Mason. Masons upbring-
ing was strict (Seventh Day
Adventist), and although her
mother was odd and subject
to mood swings, she was not
the sexual sadist depicted
in Sybil. Starting at age 22,
Mason was treated by Corne-
lia Wilbur, an ambitious psy-
chiatrist who progressed from
treating traumatized World
War II soldiers with hypno-
sis and sodium pentothal to
inducing traumatic memories
in Mason with the same tools. This she did
on and off for more than two decades, rst in
Nebraska and then in New York City. In New
York, Wilbur often saw Mason daily, making
house calls in the evening with a satchel of
drugs and a portable electroconvulsive ther-
apy machine. Beginning in 1963, the patient
and her therapist collaborated with Schreiber,
a journalist with a tendency to make her mag-
azine stories more salable by massaging bio-
graphical details.
Nathan offers a compelling account of
the creation, packaging, and selling of this
case of medical and journalistic malpractice.
Her sources include transcripts of therapy
sessions, letters by the patient to childhood
friends and former roommates, and inter-
views with acquaintances and colleagues of
all three women. Nathans credentials suit
this topic well, as she coauthored an ear-
lier book ( 2) that helped reverse the ood
of false memories implanted in children by
prosecutors and therapists inspired by Sybil.
As a feminist, she was dismayed that a seg-
ment of the womens movement channeled
its social concerns into a hunt for psycho-
sexual demons that unjustly targeted teach-
ers and day care workers.
Analyzing the significance of Sybil,
Nathan shows how the dilemmas faced by
postWorld War II women helped shape that
case and gave it cultural resonance. Wilbur,
she explains, saw herself as a psychiatric
Betty Friedan, encouraging female patients
to try out new social roles and, indeed, even
new personalities. Similarly, journalist
Schreiber saw Mason as a country girl from
a stiing background who found a new iden-
tity in the big city. Both Wilbur and Schreiber
had struggled to find acceptance in male-
dominated professions and reveled in the sta-
tus and remuneration that Sybil brought them.
Looking beyond the three women who
created Sybil, Nathan explores the institu-
tional and professional context of both that
case and the epidemic that followed. With
its shifting boundaries and history of diag-
nostic uncertainty, psychiatry was a medi-
cal specialty with little resistance to the fad-
dishness and yearning for breakthroughs
that fueled the multiple personality disor-
der fervor. Adding to clinicians enthusiasm
was the embrace of multiple personality
disorder by celebrities, journalists, and tele-
vision producers.
In concluding her cautionary tale of this
great American multiplicity, Nathan regrets
that desire for personal change went awry at a
fractured moment in history. The result was
that women and their social struggles were
reduced to a bizarre illness. The cure was not
critical inquiry or protest marches or efforts
at the polls. Instead the cure was drugs [and]
hypnosis. Reading Sybil Exposed, that con-
clusion seems warranted.
References
1. F. R. Schreiber, Sybil (Regnery, Chicago, 1973).
2. D. Nathan, M. R. Snedeker, Satans Silence: Ritual Abuse
and the Making of a Modern American Witch Hunt (Basic,
New York, 1995).
Sybil, Inc.
PSYCHIATRY
Ben Harris
The reviewer is at the Department of Psychology, Univer-
sity of New Hampshire, Durham, NH 03824, USA. E- mail:
bh5@unh.edu.
Sybil Exposed
The Extraordinary Story
Behind the Famous
Multiple Personality Case
by Debbie Nathan
Free Press (Simon and
Schuster), New York, 2011.
328 pp. $26, C$29.99, 16.95.
ISBN 9781439168271.
10.1126/science.1212843
The account that made the story famous. The
cover of the rst edition of Sybil.
www.sciencemag.org SCIENCE VOL 334 21 OCTOBER 2011 313
EDUCATIONFORUM
F
or academic researchers and educators,
summer months are often among the
most productive times of year. Unfortu-
nately, at many minority-serving institutions
(MSIs) that serve a high proportion of under-
represented minorities (URMs), the summer
often brings a brain drain that threatens to
erode capacity for conducting research. Sum-
mer research programs (SRPs) recruit stu-
dents from MSIs to spend a summer working
at a research-intensive university. SRPs can
expose students to resources and experiences
they may not as easily access at their home
MSI. Among benefits to host institutions,
SRPs help fulfill some broader impacts
(i.e., nonresearch, such as education and out-
reach) that are required by many funders.
MSI faculty may benet from skills the SRP
students bring back. But this brain drain can
undermine the ability of MSI faculty to effec-
tively teach research skills and to develop tal-
ent among a diverse pool of students.
In 2008, ~25% of U.S. science, technol-
ogy, engineering, and math (STEM) bacca-
laureate degrees awarded to URM students
were from MSIs ( 1). These institutions enroll
~30% of all URM students ( 2). Although the
percentage of URMs earning baccalaureate
degrees has declined ( 1), those who complete
their degrees show Ph.D. graduation rates
similar to non-Hispanic white counterparts
( 1). Many MSIs are among the leading STEM
bachelors degreegranting institutions for
Hispanic and African American students who
complete Ph.D. degrees ( 3). But whereas
interest in diversity underlies undergradu-
ate SRPs, questions remain about intended
recruitment and retention of URM students in
graduate programs. Despite a steady increase
in graduate program graduation rates over the
past decades ( 4), URM participation in STEM
graduate programs is lower than expected on
the basis of U.S. demographics ( 5).
Erosion of the Capacity to Develop Talent
Despite SRPs benefits, they may hinder
the ability of talented individuals to access
the skill set required to succeed in research
careers. Institutionalized one-way bridges,
like SRPs, can siphon well-trained talented
students from MSIs. As a result, research
goals of MSI faculty may not advance as well
as they could; well-trained personnel are key
to producing good-quality data that become
the basis for publications and grant applica-
tions. In turn, reviews of research grant appli-
cations from MSIs may criticize productivity
levels, the amount of preliminary data, and
the adequacy of academic environments to
implement research programs. Such percep-
tions of MSIs may erode funding for facili-
ties, equipment, and supplies.
A vicious cycle may unfold. Adequately
supported MSI faculty can develop talent
among URM students with high potential, but
lacking research experience. Early exposure
to research ( 6) and long-term experiences ( 6,
7) have been correlated with outcomes such
as persistence in research careers. Applicant
training is a key element in successful grant
applications. Early and consistent access to
research mentors and adequate resources may
provide cumulative advantages ( 8). Although
it is not clear what factors contribute to differ-
ential success rates for URM faculty in obtain-
ing research funds ( 8, 9), lack of funding in
turn affects the overall capacity to develop and
nurture local students, leading to a shortage of
well-qualied, diverse talent.
Ensuring Long-Term Sustainability
Although some promising programs exist
( 10), we must invest in sustainable develop-
ment of talent across a wider spectrum of aca-
demic institutions. This requires a shift away
from the prevailing system at research-inten-
sive institutions ( 11) and MSIs ( 12) to provide
research-intensive experiences for all stu-
dents. Collaborations between research-inten-
sive institutions and MSIs could be rebalanced
for more two-way bridge partnerships.
For example, there are well-equipped MSI
research facilities supported by initiatives
from funders such as the National Institutes
of Health and National Science Foundation
(e.g., Spelman Colleges Center for Health
Disparities Research and Education, and the
Ponce School of MedicineMoffitt Cancer
Center Partnership). Thoughtful integration
of research and teaching training activities,
guided by aligned interests, may allow a post-
doctoral fellow to receive training in teaching
at a MSI partner school. Graduate students
may gain valuable experience in communicat-
ing science ( 13) and acting as mentors to URM
colleagues at partner MSIs. Students who act
as role models demonstrate increased learn-
ing and tend to make stronger commitments
to their studies ( 14). Students from MSIs will
share experiences with groups at SRP host
institutions. This may increase their sense of
belonging, shown to help students overcome
the uncertainty ( 15) that discourages many
from pursuing a STEM career ( 16).
The benets of diversity at all stages of
the academic pipeline ( 17) are well docu-
mented and recognized by funding agen-
cies ( 10). Amid calls for supporting MSIs to
expand their effective recruiting and retention
rates ( 18), while establishing basic indicators
of student outcomes to enable institutions
to assess their effectiveness ( 12), now is the
time to rethink our approach to developing a
diverse, talented STEM workforce.
References and Notes
1. National Science Foundation (NSF), Women, Minorities,
and Persons with Disabilities in Science and Engineering:
2011 (NSF, Arlington, VA, 2011).
2. X. Li, Characteristics of Minority-Serving Institutions and
Minority Undergraduates Enrolled in These Institutions
(NCES 2008-156, U.S. Department of Education, Wash-
ington, DC, 2007).
3. National Academies and Institute of Medicine, Expanding
Underrepresented Minority Participation: Americas Science
and Technology Talent at the Crossroads (National Acad-
emies Press, Washington, DC, 2010).
4. NSF, Doctorate Recipients from U.S. Universities: 2009
(NSF, Arlington, VA, 2010); www.nsf.gov/statistics/
nsf11306.
5. P. Einaudi, Science Resources Statistics InfoBrief
(NSF11-319, NSF, Arlington, VA, 2011).
6. S. H. Russell et al., Science 316, 548 (2007).
7. A. L. Zydney et al., J. Eng. Educ. 91, 151 (2002).
8. D. K. Ginther et al., Science 333, 1015 (2011).
9. L. A. Tabak, F. S. Collins, Science 333, 940 (2011).
10. E. M. August et al., J. Cancer Educ., (2011); 10.1007/
s13187-011-0265-4.
11. W. A. Anderson et al., Science 331, 152 (2011).
12. L. E. Malcom et al., Tapping HSI-STEM Funds to Improve
Latina and Latino Access to the STEM Professions (Univ.
of Southern California, Los Angeles, CA, 2010).
13. D. F. Feldon et al., Science 333, 1037 (2011).
14. J. M. Good et al., J. Negro Educ. 69, 375 (2000).
15. G. Cohen, J. Garca, Curr. Dir. Psychol. Sci. 17, 365 (2008).
16. G. M. Walton, G. L. Cohen, J. Pers. Soc. Psychol. 92, 82
(2007).
17. P. Gurin et al., Harv. Educ. Rev. 72, 330 (2002).
18. E. M. Bensimon, L. E. Malcom, B. Dvila, (Re)constructing
Hispanic-serving institutions: Moving beyond numbers
toward student success (EP3: Education Policy and Prac-
tice Perspectives 6, Iowa State Univ., Ames, IA, 2010).
19. The author thanks undergraduate researchers (supported
by Department of Defense grant W911NF-09-1-0219);
particularly, D. Kiehart, T. Littleton, and C. Doe, who
participated in the two-way collaboration. The author was
supported by NSF.
Rethink Summer Student Research
EDUCATION
Franklin A. Carrero-Martnez
Research and training at institutions serving
minority students may suffer as top students
leave for other schools each summer.
Department of Biology, University of Puerto Rico,
Mayagez, Mayagez, PR 00681, USA. E-mail: franklin.
carrero@upr.edu 10.1126/science.1209555
AJournal withImpact fromAAAS, the publisher of Science
Science Translational Medicine
Integrating Medicine and Science
A recent journal article features the sequencing of fetal DNA from plasma
of a pregnant woman to permit prenatal, noninvasive genome-wide
screening to diagnose fetal genetic disorders.
Sci Transl Med 8 December 2010:
Vol. 2, Issue 61, p. 61ra91
DOI: 10.1126/scitranslmed.3001720
Recommend an
institutional subscription
to your library today!
ScienceOnline.org/recommend
ScienceTranslationalMedicine.org
Indexed in
MEDLINE/PubMed
Academic Opportunities in
European Science
In This Issue
Scientists who wish to pursue academic careers in Europe
have much with which to contend, especially now amidst an
uncertain scal landscape. However, buttressed by the European
Commission, several member states, including Germany, the
United Kingdom, and France, as well as Scandinavia, have plans
in place to bolster scientic research and innovation, and make
those regions attractive destinations for early and mid-career
professionals looking for academic positions.
See full story on page 397.
Upcoming Features:
Neuroscience: Emerging FieldsNovember 4
Focus on ChinaDecember 9
BS/MS Scientists (online only)January 13
FOCUS ON EUROPE Brought to you by the Science/AAAS Custom Publishing Ofce
www.sciencemag.org SCIENCE VOL 334 21 OCTOBER 2011
315
PERSPECTIVES
E
lectrical synapses have a rich and
sometimes controversial history.
During the early 20th century, the
question of whether chemical or electrical
synapses underlie the main mode of signal-
ing in the mammalian central nervous sys-
tem (CNS) was hotly debated ( 1). However,
after the discovery of neuronal inhibition in
the early 1950s, it was accepted that trans-
mission of information via neurotransmit-
ter molecules (chemical synapses) repre-
sents the major form of signaling among
CNS neurons. Although a consensus devel-
oped that electrical synapses are present
in a subset of neuronal connections, it was
unknown whether these synapses exhibit
selectivity and plasticity as shown for most
chemical synapses, or whether neuronal
activity can result in long-term change of
synaptic strength of electrical synapses ( 2)
in the mammalian CNS. On page 389 of this
issue, Haas et al. ( 3) demonstrate that the
strength of electrical synapses among spe-
cic neurons in the thalamus of the mam-
malian brain affects long-term depression
(LTD), a process important for learning and
memory. Given the proposed role that elec-
trical synapses play in synchronizing neuro-
nal activity, these ndings suggest a mech-
anism for controlling the coordination of
neuronal activity.
Chemical synapses selectively connect
different types of neurons to provide a spe-
cific path of communication among CNS
neurons ( 4, 5). It was discovered more than
10 years ago that electrical synapses are
formed with exquisite selectivity among
specic classes of inhibitory neurons ( 6, 7).
Moreover, another hallmark of synapses
modulation by neurotransmittershas been
also demonstrated in both the thalamus and
hippocampus ( 8, 9).
To investigate the effects of neuronal
activity on the strength of electrical syn-
apses in the mammalian CNS, Haas et al.
recorded electrical activity from pairs of
neurons in the rat thalamic reticular nucleus
(TRN). All sensory messages are conveyed
to the cortex via the thalamus (see the gure)
( 10). The TRN is located between the thala-
mus and the cortex and, unlike other groups
of thalamic neurons (thalamic nuclei), is
composed of inhibitory neurons that are
connected by electrical synapses ( 11). These
electrical synapses, together with inhibi-
tory chemical synapses (those that transmit
-aminobutyric acid), can synchronize r-
ing of TRN neurons. Because TRN neurons
project to other thalamic nuclei, synchro-
nized activity of TRN neurons can entrain
the activity of other thalamic regions ( 10).
To measure the strength of electrical cou-
pling, Haas et al. injected depolarizing cur-
rent into one TRN neuron and measured the
response in both the injected cell and the
noninjected neuron. The ratio of the two
electrical responses is dened as the cou-
pling coefcient (cc). The coupling coef-
cient among TRN neurons is large (indicat-
ing strong electrical coupling), and often a
burst of electrical activity (action potential)
in one neuron will produce action potential
in the coupled neuron. The coupling coef-
ficient can be estimated in two ways: by
injecting current in cell 1 (i.e., cc
12
) or by
injecting current in cell 2 (i.e., cc
21
). Inter-
estingly, Haas et al. found that a substantial
number of neuron pairs exhibited asymmet-
rical coupling. That is, the distribution of
the ratio cc
1
/cc
2
was skewed. In vivo, TRN
neurons often re bursts of electrical activ-
ity (spikes). The authors found that bursts
of spikes in pairs of coupled TRN neurons
resulted in LTD of the strength of electri-
cal synapses. Burst activity in both neurons
produced symmetrical LTD such that cc
12
=
cc
21
. However, when cell 1 was induced to
burst while the coupled cell (cell 2) was pre-
vented from bursting (by current injection),
the resulting LTD was asymmetrical. Under
these conditions, the coupling measure from
cell 1 to cell 2 (cc
12
) decreased much less than
the coupling from the inactive cell (cc
21
) that
was prevented from bursting. Although Hass
et al. observed a modest decrease of synap-
tic strength, they show that LTD of electrical
coupling is sufcient to prevent propagation
of bursting activity among neurons.
What are the mechanisms underlying
activity-dependent LTD of electrical syn-
apses? The results of Haas et al. suggest
that sodium-dependent action potentials
may play a role (in response to a stimulus,
sodium channels open and allow sodium
into the neuron, which triggers the fir-
NEUROSCIENCE
After burst
activity
Inactive
Brain
Depression of electrical synapses. (Enlarged view, left side) Inhibitory neurons (circles) in the mamma-
lian thalamic reticular nucleus (TRN) are connected via electrical (small black arrows) and inhibitory chemi-
cal synapses (not illustrated). External excitatory inputs impinge on TRN neurons and can activate them.
(Enlarged view, right side) When TRN neurons generate bursts of action potentials, electrical synapse
strength decreases. The decrease of strength can be asymmetrical between cells that do not generate bursts
(white arrows) and those that do (gray arrows).
C
R
E
D
I
T
:

B
.

S
T
R
A
U
C
H
/
S
C
I
E
N
C
E
Department of Comparative Medicine, Stanford University,
Stanford, CA 94305, USA. E-mail: shaul.hestrin@stanford.edu
The Strength of Electrical Synapses
Shaul Hestrin
The strength of electrical synapses in the
mammalian brain can be modulated by
neuronal ativity.
21 OCTOBER 2011 VOL 334 SCIENCE www.sciencemag.org 316
PERSPECTIVES
Watery Disks
PLANETARY SCIENCE
Rachel Akeson
The Herschel Space Observatory has detected
a large reservoir of water stored as ice in the
disk surrounding a nearby star.
ing of an action potential). In addition, an
increase in cytoplasmic calcium concen-
tration could activate a protein kinase that
modies (by phosphorylation) gap junction
channels,which bridge electrically coupled
neurons. Alternatively, partial replacement
of nonrectifying gap junctions with rectify-
ing gap junctions (thus allowing electrical
current to pass preferentially in one direc-
tion) is a possible mechanism.
The functional consequences of LTD
produced by electrical synapses are of great
interest. By producing asymmetrical changes
of electrical coupling, the outbound connec-
tion of a group of bursting neurons will be
stronger than the inbound connection from
coupled neurons that did not burst. The func-
tional consequences resulting from the gen-
eration of asymmetry of electrical synapses
within the TRN remains to be explored. The
impact of reducing the strength of electrical
synapses at the circuit level and their poten-
tial effect on rhythmic activity in the TRN is
also of interest. In vivo, TRN neurons can
switch between burst and tonic ring modes
( 10). If these activities can induce LTD, one
may expect to nd changes in the strength of
electrical coupling in the TRN that correlate
with different behavioral states during sleep
and wakefulness.
References
1. W. M. Cowan, E. R. Kandel, in Synapses, W. M. Cowan,
T. C. Sdhof, C. F. Stevens, Eds. (Johns Hopkins Univ.
Press, Baltimore, 2001), pp. 187.

2. A. E. Pereda et al., Proc. Natl. Acad. Sci. U.S.A. 95,
13272 (1998).
3. J. S. Haas, B. Zavala, C. E. Landisman, Science 334, 389
(2011).
4. S. P. Brown, S. Hestrin, Curr. Opin. Neurobiol. 14, 415
(2009).
5. H. Ko et al., Nature 473, 87 (2011).
6. M. Galarreta, S. Hestrin, Nature 402, 72 (1999).
7. J. R. Gibson, M. Beierlein, B. W. Connors, Nature 402, 75
(1999).
8. C. E. Landisman, B. W. Connors, Science 310, 1809
(2005).
9. V. Zsiros, G. Maccaferri, J. Neurosci. 28, 1804
(2008).
10. S. M. Sherman, R. W. Guillery, in The Synaptic Orgnaiza-
tion of the Brain, G. M. Shepherd, Ed. (Oxford Univ.
Press, Oxford, ed. 5, 2004), pp. 311359.
11. C. E. Landisman et al., J. Neurosci. 22, 1002 (2002).
10.1126/science.1213894
T
he paradigm for star formation is
understood to center around the for-
mation of a rotating disk from a cloud
of gas and dust. The circumstellar disk fun-
nels material onto the newly formed central
star and also serves as a reservoir of mate-
rial from which a planetary system may arise.
Determining the physical and chemical com-
position of these disks is necessary to under-
stand the formation and evolution of planets.
Previous observations have detected the pres-
ence of molecules within the disk, thereby
demonstrating an active chemical network.
However, this chemistry is harder to trace in
the majority of the disk where low tempera-
tures result in the molecules freezing out onto
grains. On page 338 of this issue, Hoger heijde
et al. ( 1) use the Herschel Space Observatory
to detect cold water vapor in one of the closest
young stars, TW Hydrae. The source of that
water vapor is likely to be a large reservoir of
ice grains.
In the past few decades, the basic physical
information we have learned about circum-
stellar diskstheir mass, size, and lifetimes
come from observations of micrometer-sized
dust grains within them. However, the vast
majority of the material is gaseous, primarily
molecular hydrogen. Transitions of molecular
hydrogen are very difcult to observe, and the
only H
2
detections in disks are of warm gas
in the center. Attention has therefore turned to
other molecules such as CO, HCO
+
, and CN
to trace the gaseous component.
Water is the main constituent of the
mantles on grains found in the low-density
medium that fills the space between stars
and can serve as an oxygen reservoir in the
gas phase. Thus, water may be one of the
key components in the chemistry and ther-
mal balance of both the parent cloud and
the circumstellar disk. Spectroscopic obser-
vations of water vapor from ground-based
telescopes are hindered by water vapor fea-
tures in Earths atmosphere. In the inner disk,
where water is present as hot vapor and can
be observed from the ground, observations
have shown water vapor at abundances above
the levels measured in molecular clouds ( 2).
These observations include the so-called
habitable zone where terrestrial planets are
believed to have formed in our solar system.
However, most of the mass within the disk is
NASA Exoplanet Science Institute, California Institute of
Technology, Pasadena, CA 91125, USA. E-mail: rla@ipac.
caltech.edu
Rocky planetesimals Icy planetesimals
Snow line
A frozen line of division. Interior to the snow line, planetesimals forming in the disk are rocky, while out-
side they are icy. In our solar system, this line corresponds to the divide between the inner terrestrial planets
and the outer gas giants. C
R
E
D
I
T
:

P
.

H
U
E
Y
/
S
C
I
E
N
C
E
www.sciencemag.org SCIENCE VOL 334 21 OCTOBER 2011 317
PERSPECTIVES
below the 150 K condensation temperature
for water, and thus most water exists as ice
on grains. Spectroscopic features from water
ice have been seen ( 3, 4), but these techniques
have only been applied to a few objects with
favorable viewing geometries. Observations
of the water isotope HDO (D is deuterium)
in the gas phase of the outer disk led to mod-
els of ultraviolet photons from the central
star desorbing water molecules from the icy
grains back into the gas phase ( 5). The obser-
vations of Hogerheijde et al. are consistent
with this cycle. Their calculations show that
a large population of icy grains, equivalent to
several thousand Earth oceans, is necessary
to maintain the observed level of water vapor
on the surface of the disk.
Water can also play a critical role in the
formation and nal surface composition of
planets. Ice enhances the solid material in
the cold outer part of a protoplanetary disk,
which promotes the formation of cores of
gaseous planets ( 6). The disk radius where
ice can condense is often termed the snow
line, and the location of this line is a property
of the stellar mass and disk evolutionary state
(see the gure). The location and evolution of
this snow line may affect the formation rate of
large planets ( 7).
The distribution of water ice in the cir-
cumstellar disk can also help address the
issue of where Earths water originated.
While forming, Earth is believed to have
been too hot to have liquid water and would
have retained little water vapor from the gas-
eous component of the disk. Thus, the water
we have now arrived later, most likely from
ice-covered comets or asteroids from the
outer parts of the solar system. In addition
to the water abundance, the spectra obtained
by Hogerheijde et al. allow determination of
the spin isomer ratio, where the spin refers
to the alignment of the hydrogen proton spin
vectors (that is, the ratio of the amount of
para- to ortho-hydrogen in the water mole-
cules). They found a ratio much lower than
that measured for solar system comets, sug-
gesting that material from multiple locations
in the TW Hydrae disk is mixed before incor-
poration into larger bodies. Evidence for
such radial transport in the early solar sys-
tem includes results from the Stardust mis-
sion that returned comet samples containing
material formed at high temperatures ( 8).
As the number of planets discovered
around other stars expands to include many
systems with multiple planets, it is clear that
the universe includes many planetary sys-
tem architectures very different from that of
our own solar system. To constrain models
of planet formation, including the chemi-
cal composition, we need to understand the
distribution and evolution of molecules in
the disk, including water, a key catalyst for
life on Earth. The next several years will pro-
vide many opportunities to progress in this
study as Herschel and other observatories
will make spectroscopic observations of a
much larger sample of disks covering a range
of stellar age and mass. Also, new facilities,
such as the Atacama Large Millimeter Array,
will greatly expand on the current sensitivity
levels to allow spatially resolved observations
of molecules in the disk.
References
1. M. R. Hogerheijde et al., Science 334, 338 (2011).
2. J. S. Carr, J. R. Najita, Science 319, 1504 (2008).
3. K. M. Pontoppidan et al., Astrophys. J. 622, 463 (2005).
4. M. Honda et al., Astrophys. J. 690, L110 (2009).
5. C. Ceccarelli, C. Dominik, E. Caux, B. Leoch, P. Caselli,
Astrophys. J. 631, L81 (2005).
6. C. Hayashi, K. Nakazawa, Y. Nakagawa, Protostars and
Planets II (University of Arizona Press, Tucson, 2005),
pp. 11001153.
7. G. M. Kennedy, S. J. Kenyon, Astrophys. J. 673, 502
(2008).
8. T. Nakamura et al., Science 321, 1664 (2008).
10.1126/science.1213752
Antenna-Guided Light
APPLIED PHYSICS
Nader Engheta
Compact arrays of gold nanoantennas can
be used to create optical structures that bend
the path of light in unusual ways.
T
he bent appearance of a stick half-sub-
merged in water is caused by the dif-
ference in refractive indices of air and
waterlight travels more slowly in water than
in air (see the gure, panel A) and refracts and
reects off the air-water interface. Snells law
( 1) lets us calculate the bending angle if we
know the geometry and the refractive indices.
In complex optical instruments, where sev-
eral lenses, mirrors, and other components
may be present, designers control the bending
by keeping track of the phase shifts imposed
along the wavefront of the light; for exam-
ple, a light beam can be focused by different
phase shifts that occur along a curved lens.
These optical components are much larger
than the wavelength of light, which limits the
minimum size of devices. On page 333 of this
issue, Yu et al. ( 2) show how arrays of struc-
tures smaller than the wavelength of light,
V-shaped nanoantennas made of gold, bend
light by creating abrupt phase shifts through
the excitation of resonances. The authors
show that these compact metasurfaces fol-
low a more general version of Snells law that
accounts for the bending of a light beam in
unconventional but potentially useful ways.
Conventionally, the bending of light may
occur at an abrupt interface of two media
(e.g., air and water), or through a gradual
change of refractive index (e.g., air above
the hot desert roads causes mirage; see the
figure, panel B). However, it is possible
to obtain the desired phase shift along the
wavefront by tailoring planar interfaces.
One of the early examples is the Fresnel
lens, in which a set of concentric lenses are
cut to different curvatures and impose dif-
ferent phase shifts. Although a Fresnel lens
is much thinner than an equivalent conven-
tional lens, its thickness is still far greater
than the wavelength of light.
Light does not always simply pass
through a medium; it can also excite reso-
nances that can lead to absorption and emis-
sion. For the much longer wavelengths of
light used in radio and microwave com-
munications, antennas called reectarrays
( 3) and transmitarrays ( 4) contain multiple
antenna elements that act as resonators to
control the direction in which signals are
received or broadcast. However, the reso-
nant elements responsible for the required
phase shift and their arrangement in peri-
odic arrays are still comparable in size to
the wavelength of operation ( 3 5). These
devices often operate over only a narrow
range of frequencies.
For shorter-wavelength light, such as
infrared and visible light, plasmonic phenom-
enathe excitation of collective oscillations
of electrons in materials such as gold and
silvercan allow subwavelength objects to
undergo resonance responses in the scatter-
ing process. Yu et al. designed subwavelength
University of Pennsylvania, Department of Electrical and
Systems Engineering, Philadelphia, PA 19104, USA. E-mail:
engheta@ee.upenn.edu
21 OCTOBER 2011 VOL 334 SCIENCE www.sciencemag.org 318
PERSPECTIVES
gold antennas with a V shape; they varied the
scattering of light by changing the length of
the arm and the angle and the orientation of
these Vs. The phase difference between the
scattered and incident elds is tailored over a
small distance along the lights path, that is,
the structures are optically thin.
Yu et al. printed planar arrays of such
V-shaped nanoantennas in suitably designed
patterns on a silicon wafer and demonstrated
several intriguing light-bending scenarios at
these metasurfaces, including unconventional
reection and refraction angles, total internal
reflections with two critical angles (rather
than only one), and reected light becoming
evanescent (diminishing in amplitude with
distance away from the interface, rather than
propagating) at certain angles. None of these
effects are predicted from the conventional
Snells law, but they do follow a generalized
version derived by the authors that allows for
desired variations of the change of phase on
the interface.
These arrays of nanoantennas, which
could include movable sections, could be
used to design photonic components such as
lenses and mirrors that are ultrathin, confor-
mal (angle-preserving), and even deform-
able. Reconfigurable couplers and wave-
guides, which could be driven by electric,
magnetic, or optical stimuli, may be envi-
sioned that could guide and mix light beams
through almost arbitrary paths chosen along
a surface. Yu et al. have also created optical
vortices with orbital angular momentum ( 6)
by impinging a beam at normal incidence on
the specially designed planar metasurface of
these V-shaped nanoantennas. Such vortices
could nd use in applications such as opti-
cal tweezers.
Metasurfaces ( 7) are the planar version
of metamaterials that are engineered to con-
trol and tailor the light interaction in uncon-
ventional ways (for example, creating mate-
rials with optical band gaps that completely
reect light over a given frequency range).
In the three-dimensional metamaterials, it
can be difcult to engineer a structure that
maintains its designed performance and
avoids performing like a bulk material. Meta-
surfaces may offer advantages in this regard
because their constituent resonant elements
are all distributed in a planar surface and
more readily assembled. This type of two-
dimensional structure will add another tool
to the eld of transformation optics ( 8, 9), in
which a prescribed change (such as a phase
shift or amplitude variation) is designed into
the light path for applications such as cloak-
ing, or where metasurfaces are used to creat-
ing highly conned cavity modes ( 10, 11) of
potential interest in quantum optics.
References
1. M. Born, E. Wolf, Principles of Optics (Pergamon,
Oxford, 1980).
2. N. Yu et al., Science 334, 333 (2011).
3. D. M. Pozar, S. D. Targonski, H. D. Syrigos, IEEE Trans.
Antenn. Propag. 45, 287 (1997).
4. C. G. M. Ryan et al., IEEE Trans. Antenn. Propag. 58,
1486 (2010).
5. N. Bliznyuk, N. Engheta, Mic. Opt. Tech. Lett. 40, 361
(2004).
6. M. Padgett, J. Courtial, L. Allen, Phys. Today 57, 35
(2004).
7. E. F. Kuester, M. A. Mohamed, M. Piket-May, C. L. Hol-
loway, IEEE Trans. Antenn. Propag. 51, 2641 (2003).
8. J. B. Pendry, D. Schurig, D. R. Smith, Science 312, 1780
(2006).
9. U. Leonhardt, Science 312, 1777 (2006).
10. M. Caiazzo, S. Maci,, N. Engheta, IEEE Antenn. Wirel.
Propag. Lett. 3, 261 (2004).
11. C. L. Holloway, D. C. Love, E. F. Kuester, A. Salandrino, N.
Engheta, IET Microwave Antenn. Propag. 2, 120 (2008).
A
C D B
L >> L >>
d
5 m
Bending light, big and small. Several mechanisms for bending light are depicted. The optical structures
shown in (A) and (B) are much larger than the wavelength of light. In (A), an interface between two media
with two different indices of refraction bends light. In (B), light is bent by a material that gradually changes
refractive index with distance. Yu et al. caused the bending of light in unusual ways (C) with thin metasur-
faces. These metasurfaces contain distributed arrays of gold nanoantennas (D) that are smaller than the
wavelength of light. In such arrays, the proper patterns of phase changes created by resonant nanostructures
lead to bending effects not anticipated by conventional laws of reection and refraction in optics.
10.1126/science.1213278
T
he advent of satellite-based remote
sensing of ocean color in the late
1970s ( 1) provided the first large-
scale views of chlorophyll distributions in
the upper ocean. These distributions are
a proxy for the biomass of phytoplankton,
which drive oceanic productivity. More
recently, ocean color measurements have
been combined with satellite data on sea-
surface height (SSH) and other physical
properties of the ocean to elucidate the pro-
cesses that regulate primary production in
the sea. On page 328 of this issue, Chelton et
al. ( 2) further advance this eld by showing
that ocean eddies exert a strong inuence on
near-surface chlorophyll.
Initial comparisons ( 3, 4) of satellite
ocean color measurements and SSH data
showed that some of the variability in ocean
color was associated with large-scale SSH
patterns that propagate westward in extra-
tropical latitudes. The authors attributed
these patterns to planetary or Rossby waves,
which are freely propagating modes of large-
scale variability in the ocean. Four basic
processes have been proposed to explain
the observed relations, including lateral
Eddies Masquerade as
Planetary Waves
OCEANS
Dennis J. McGillicuddy Jr.
Variabilities in sea-level and upper-ocean chlorophyll reveal the systematic inuence of
nonlinear eddies.
Woods Hole Oceanographic Institution, Woods Hole, MA
025431541, USA. E-mail: dmcgillicuddy@whoi.edu
www.sciencemag.org SCIENCE VOL 334 21 OCTOBER 2011 319
PERSPECTIVES
advection of the mean chlorophyll gradi-
ent, uplift of the deep chlorophyll maximum
into the surface layer, enhancement of phy-
toplankton biomass stimulated by upwelling
of nutrients, and accumulation of material
in convergence zones within the planetary
wave eld ( 5 7).
These early studies focused on large-
scale signals characteristic of Rossby waves
by processing the satellite measurements
with scale-selective lters. This processing
was intended to remove seasonal variability
as well as the effects of mesoscale (tens to
hundreds of kilometers) eddiesubiquitous
features of ocean circulation (sometimes
referred to as the internal weather of the sea)
that result from both direct forcing and inter-
nal instability processes.
Merging data from multiple satellite mis-
sions led to altimetric data sets ( 8 10) with
higher resolution than used previously, but
it remains difcult to differentiate between
Rossby waves and eddies in the merged data
sets. Single Rossby waves (see the gure,
panel A) are rarely if ever observed in the
ocean, but superposition of multiple Rossby
waves can result in eddy-like features (panel
A, inset) that are similar to the patterns seen
in altimeter data (panel B). However, plane-
tary waves and eddies have different degrees
of nonlinearity: Nonlinear eddies trap uid
inside them, whereas linearly propagating
wavelike disturbances do not. The degree
of nonlinearity can be estimated as the ratio
between an eddys swirl velocity and its
translation speed.
Previously, Chelton and co-workers used
this insight to show that mid-latitude SSH
variability is dominated by westward-prop-
agating nonlinear eddies, and developed
automated tracking algorithms to compile a
global synthesis of eddy trajectories ( 9, 10).
Chelton et al. have now overlaid those eddy
tracks on the westward-propagating signals
previously attributed to Rossby waves in
the ltered SSH and ocean color data. The
results strongly suggest that eddies are driv-
ing these signals ( 2).
How might mesoscale eddies masquer-
ade as larger-scale Rossby waves? Due to
the latitudinal dependence of the effects of
Earths rotation, both types of features move
westward at roughly the same speed. More-
over, Chelton et al. show that the statisti-
cal properties of a patchwork of westward-
propagating eddies are qualitatively similar
to those expected for Rossby waves. This
observation explains why eddies can pass
through the filters intended to eliminate
them in earlier studies.
Chelton et al.s ndings require reassess-
ment of the underlying mechanisms used to
explain satellite observations of variability in
SSH and upper-ocean chlorophyll. Although
the same four basic processes of biomass
modulation mentioned above for Rossby
waves remain valid for eddies, lateral advec-
tion of the mean chlorophyll gradient is the
dominant mechanism revealed in Chelton et
al.s analysis. However, the relative impor-
tance of each of the four processes can vary
with oceanographic regime and scale, rang-
ing from the mesoscale down to the submeso-
scale ( 11 13). At present, submesoscale fea-
tures are not resolved by operational remote-
sensing technology for SSH.
Although higher-resolution data are
expected in the future for both SSH and
ocean color, in situ observations will con-
tinue to be critical for those variables that
cannot be measured from space. Moreover,
because near-surface waters are depleted in
nutrients over large areas of the mid-lati-
tudes, key aspects of the biological response
to physical perturbations take place too deep
to be detected by satellite ocean color imag-
ery ( 14). Although Chelton et al.s results
must be interpreted with that caveat, their
ndings constitute a key step forward in our
understanding of physical-biological inter-
actions in the ocean, with important rami-
cations for both ecosystem dynamics and
biogeochemical cycling.
References and Notes
1. J. F. R. Gower, K. L. Denman, R. J. Holyer, Nature 288,
157 (1980).
2. D. B. Chelton, P. Gaube, M. G. Schlax, J. J. Early, R. M.
Samelson, Science 334, 328 (2011); 10.1126/
science.1208897.
3. B. M. Uz, J. A. Yoder, V. Osychny, Nature 409, 597
(2001).
4. P. Cipollini, D. Cromwell, P. G. Challenor, S. Raffaglio,
Geophys. Res. Lett. 28, 323 (2001).
5. P. D. Killworth, P. Cipollini, B. M. Uz, J. R. Blundell,
J. Geophys. Res. 109, C07002 (2004).
6. Y. Dandonneau, A. Vega, H. Loisel, Y. du Penhoat,
C. Menkes, Science 302, 1548 (2003).
7. G. Charria, F. Mlin, I. Dadou, M.-H. Radenac, V. Garon,
Geophys. Res. Lett. 30, 1125 (2003).
8. A. Pascual, Y. Faugre, G. Larnicol, P.-Y. Le Traon,
Geophys. Res. Lett. 33, L02611 (2006).
9. D. B. Chelton, M. G. Schlax, R. M. Samelson, Prog.
Oceanogr. 91, 167 (2011).
10. D. B. Chelton, M. G. Schlax, R. M. Samelson, R. A. de
Szoeke, Geophys. Res. Lett. 34, L15606 (2007).
11. M. Lvy, P. Klein, A.-M. Treguier, J. Mar. Res. 59, 535
(2001).
12. E. R. Abraham, Nature 391, 577 (1998).
13. D. A. Siegel, P. Peterson, D. J. McGillicuddy Jr., S. Mari-
torena, N. B. Nelson, Geophys. Res. Lett. 38, L13608
(2011).
14. D. J. McGillicuddy Jr. et al., Science 316, 1021 (2007).
15. J. C. McWilliams, G. R. Flierl, Deep-Sea Res. 23, 285
(1976).
16. Altimeter data were produced and distributed by AVISO
(www.aviso.oceanobs.com) as part of the SSALTO ground-
processing segment.
17. I thank NSF and NASA for support and L. Anderson for
preparing the gure.
Rossby waves Satellite altimeter
Single wave
Superposition of 4 waves
Longitude (W)
L
a
t
i
t
u
d
e

(

N
)
A
70 65 60 55 70 65 60 55
29
32
35
Longitude (W)

S
e
a
-
l
e
v
e
l

a
n
o
m
a
l
y

(
c
m
)
B
29
32
35
20
10
0
10
20
Rossby waves and eddies. (A) A model of a single Rossby wave propagating
through a still ocean leads to a highly regular sea-level anomaly pattern. If sev-
eral such waves are superposed, an eddy-like pattern results (inset) ( 15). (B)
Mapped observations ( 16) of sea-level anomaly for a region of the western North
Atlantic on 17 June 2005. Satellite ground tracks are shown as dotted lines.
Although the patterns seen in the altimeter data resemble those of the super-
posed Rossby waves, Chelton et al. show that they are in fact caused by nonlinear
mesoscale eddies.
10.1126/science.1208892
21 OCTOBER 2011 VOL 334 SCIENCE www.sciencemag.org 320
PERSPECTIVES
C
ells, and the organelles within them,
are surrounded by lipid-bilayer
membranes that compartmentalize
biochemical reactions and pathways. Mem-
brane-embedded proteins control the ux
of molecules, energy, and information such
that the segregated compartments func-
tion as a unied living cell. Traditionally,
membrane proteins were pictured as oat-
ing around quite independently of the sur-
rounding lipids, yet when this uid-mosaic
model was described, Singer and Nicolson
qualied that a small fraction of the lipid
may interact specically with the membrane
proteins ( 1). It has taken 40 years to fully
appreciate the importance of this assertion.
On page 380 of this issue, Zhou et al. ( 2)
report mass spectrometry of intact integral
membrane protein complexes solubilized
from bilayers. The results show that specic
structural lipids remain bound in the gas
phase and can be counted.
Despite some technical hurdles, integral
membrane proteins and their complexes
have proven amenable to the soft ionization
technologies of biological mass spectrom-
etry. Intact complexes are isolated by gen-
tle solubilization of membranes with mild
nonionic detergent, followed by biochemi-
cal isolation within a detergent micelle for
solubility. Earlier studies next used organic
solvents to break noncovalent associations
with lipids and between subunits. Chroma-
tography in aqueous/organic mixtures cou-
pled to mass spectrometry provided intact
molecular mass proles of individual sub-
units and their covalent modifications on
low-resolution analyzers ( 3, 4). Top-down
dissociation experiments on high-resolution
Fourier-transform instruments yielded fully
assigned primary structures with high statis-
tical condence ( 5, 6).
However, if the complexes could be ana-
lyzed intact, subunit stoichiometry and the
presence of intimate cofactors such as spe-
cific lipid molecules could also be deter-
mined. In 2004, Robinsons group realized
that electrospray ionization could be used
to transfer membrane-protein micelles from
aqueous solutions into the gas phase, with
subsequent depletion of detergent in vacuo
( 7). Development of this native mass spec-
trometry approach has enabled the measure-
ment of bound lipid stoichiometry in mem-
brane lipid-protein complexes described by
Zhou et al.
The detection of lipids within integral
membrane protein structures requires care-
ful solubilization to preserve native asso-
ciations. This is important because micelle-
forming detergents tend to displace lipids,
even while preserving protein subunit inter-
actions. For example, successful crystalliza-
tion of the cytochrome b
6
f complex required
addition of lipids to isolated complexes, even
though preparations retained some native
lipids ( 8, 9).
As early as 1973, Santiago et al. argued,
based upon lipid-peroxidation/protection
data, for the specic requirement of the lipid
cardiolipin for rat mitochondrial adenosine
triphosphatase (ATPase) activity (see the
gure, left), concluding that specic lipids
were critical to the function of this integral
membrane protein complex ( 10). Apprecia-
tion of the importance of specically bound
lipids to the structure and function of inte-
gral membrane proteins has grown steadily
as x-ray crystallography has provided ever
more examples from numerous membrane
proteins. Cardiolipin was rst resolved in the
structure of the purple bacterial reaction cen-
ter ( 11) and later shown to provide thermal
stability to the structure ( 12). Although the
structure of bovine mitochondrial rotary ATP
synthase has been solved, cardiolipin was not
seen, presumably due to its displacement by
detergent ( 13).
The rotary ATP synthases couple the
electrochemical energy of a transmembrane
proton gradient to synthesis of ATP; proton
movement through an integral membrane
ring of subunits causes rotation of this rotor
relative to a stationary integral stator and
associated extrinsic subunits, driving confor-
mational changes that facilitate ATP synthe-
sis ( 14). The number of rotor subunits varies
across living organisms, with those having
the least subunits synthesizing the most ATP
per proton translocated. Current research
explores the role of bound lipids in the rotor
structure, and measurements of the number
of bound lipids and their binding sites by
mass spectrometry should thus help resolve
some of the outstanding details of the mecha-
nism of rotary coupling.
Zhou et al. use native electrospray-ion-
ization mass spectrometry to measure rotor
subunit and bound lipid stoichiometry in
two intact rotary ATPases from the bacteria
Thermus thermophilus and Enterococcus
hirae. The results show that the solubilized
T. thermophilus complex has a ring with 12
subunits, with 6 phosphatidylethanolamine
lipids bound to it, whereas the E. hirae rotor
has 10 subunits with 10 cardiolipins. The
Up Close with Membrane
Lipid-Protein Complexes
STRUCTURAL BIOLOGY
Julian Whitelegge
State-of-the-art mass spectrometry reveals
how many specic lipids are bound to
membrane proteins.
O
O
P O
O
O
O
OH
O
O
O
O
P
O
O
O
O
O
O
Cardiolipin
Along for the ride. Cardiolipin (left) is an
unusual phospholipid with four fatty acyl
chains. In the integral membrane rotor of
the eight-membered vertebrate mitochon-
drial ATP synthase (right), each of four
cardiolipin molecules binds the four trans-
membrane domains of two ring subunits.
C
R
E
D
I
T
:

A
D
A
P
T
E
D

B
Y

P
.

H
U
E
Y
/
S
C
I
E
N
C
E
NPI-Semel Institute for Neuroscience and Human Behav-
ior, University of California at Los Angeles, Los Angeles, CA
900241759, USA. E-mail: jpw@chem.ucla.edu
www.sciencemag.org SCIENCE VOL 334 21 OCTOBER 2011 321
PERSPECTIVES
Every Bit Counts
CELL SIGNALING
Peter J. Thomas
Information theory is applied to cell biology
to determine the processing properties of
signaling pathways and the consequences
of noise.
authors picture the T. thermophilus complex
as six dimers each with a single lipid, con-
sistent with electron microscopy images that
support a hexameric structure ( 15). The E.
hirae complex has 10 monomers, each with a
bound cardiolipin, correcting a crystal struc-
ture with misassigned lipids ( 16).
These measurements cover a range of 0.5
to 2 lipid fatty acyl chains per transmembrane
domain within the rotor, consistent with the
value of one inferred from the measurement
of four cardiolipin per eight-subunit verte-
brate mitochondrial ATPase rotor using
31
P
nuclear magnetic resonance (NMR) ( 17). The
16 transmembrane domains that, along with
four intimately associated cardiolipins, make
up this eight-membered ring (see the gure,
right) function as the smallest natural ATPase
rotor, accounting for the maximal efciency
of ATP synthase activity observed for this
complex (13).
Zhou et al. go on to show that gas-phase
dissociation patterns are sensitive to pH and
the presence of nucleotides, suggesting con-
formational changes within the complexes.
This conclusion is further supported by
gas-phase ion mobility mass spectrometry
experiments that elegantly demonstrate the
technological sophistication of the instru-
mentation used ( 2). The authors further dis-
cuss how rearrangement of lipid-protein
interactions might serve to modulate ATP
synthase/ATPase activities.
The ability to profile intact membrane
lipid-protein complexes and detect organiza-
tional changes related to dynamic processes
by mass spectrometry, as reported by Zhou
et al., will empower our understanding of
the functioning of biological membranes. A
new class of neopentyl glycol detergents that
appear to solubilize membrane lipid-protein
complexes with minimal structural pertur-
bation bodes well for native mass spectrom-
etry ( 18). Future research should also aim to
elucidate the role of bilayers in general, as
well as the specic lipid-protein associations
discussed above, in the insertion, folding,
assembly, function, and regulation of integral
membrane proteins.
References
1. S. J. Singer, G. L. Nicolson, Science 175, 720 (1972).
2. M. Zhou et al., Science 334, 380 (2011).
3. I. M. Fearnley, J. E. Walker, Biochem. Soc. Trans. 24, 912
(1996).
4. J. P. Whitelegge, H. Zhang, R. Aguilera, R. M. Taylor,
W. A. Cramer, Mol. Cell. Proteomics 1, 816 (2002).
5. B. Thangaraj et al., Proteomics 10, 3644 (2010).
6. C. M. Ryan et al., Mol. Cell. Proteomics 9, 791 (2010).
7. L. L. Ilag, I. Ubarretxena-Belandia, C. G. Tate, C. V.
Robinson, J. Am. Chem. Soc. 126, 14362 (2004).
8. H. Zhang, G. Kurisu, J. L. Smith, W. A. Cramer, Proc.
Natl. Acad. Sci. U.S.A. 100, 5160 (2003).
9. S. S. Hasan et al., J. Mol. Biol. (2011), 10.1016/
j.jmb.2011.09.023.
10. E. Santiago, N. Lpez-Moratalla, J. F. Segovia, Biochem.
Biophys. Res. Commun. 53, 439 (1973).
11. K. E. McAuley et al., Proc. Natl. Acad. Sci. U.S.A. 96,
14706 (1999).
12. P. K. Fyfe, N. W. Isaacs, R. J. Cogdell, M. R. Jones,
Biochim. Biophys. Acta 1608, 11 (2004).
13. I. N. Watt, M. G. Montgomery, M. J. Runswick, A. G.
Leslie, J. E. Walker, Proc. Natl. Acad. Sci. U.S.A. 107,
16823 (2010).
14. D. Stock, A. G. Leslie, J. E. Walker, Science 286, 1700
(1999).
15. R. A. Bernal, D. Stock, Structure 12, 1789 (2004).
16. T. Murata, I. Yamato, Y. Kakinuma, A. G. Leslie, J. E.
Walker, Science 308, 654 (2005).
17. K. S. Eble, W. B. Coleman, R. R. Hantgan, C. C.
Cunningham, J. Biol. Chem. 265, 19434 (1990).
18. P. S. Chae et al., Nat. Meth. 7, 1003 (2010).
10.1126/science.1214084
T
he term signal transduction has been
in use for over 40 years, originating in
biological studies of cellular photo-
receptors and chemotaxis, and elaborated
through discoveries of signaling systems
such as the protein kinases ( 1). In engineer-
ing, signaling has a precise meaning that
denes the notion of information ( 2), and
quanties the capabilities of a signaling sys-
tem in terms of its channel capacity, measured
in bits ( 3). Despite the attractive analogy, a
number of obstacles have thwarted attempts
to apply information theory quantitatively
to cellular signal transduction networks. On
page 354 of this issue, Cheong et al. ( 4) suc-
ceed in doing so, making rigorous and quan-
titative measurements of information capaci-
ties in a biochemical signaling system.
A rigorous analysis of signaling requires
a well-defined communications channel,
including an ensemble of channel inputs as
well as outputs that are conditional on each
possible input (see the gure). In a biological
setting, it is not always clear what denes the
appropriate ensembles. Cheong et al. use the
signaling pathway engaged by tumor necro-
sis factor (TNF), which stimulates a host of
intracellular responses mediated by two tran-
scription factors, nuclear factor kappa B (NF-
B) and activating transcription factor2
(ATF-2). The TNF concentration serves as
the input to the channel, and the NF-B and
ATF-2 responses serve as outputs.
Estimating channel capacities requires
estimating the entropy of probability distri-
butions, which is a notoriously data-inten-
sive undertaking. Channel capacities and
quantities such as information per spike
have been estimated for neural networks
in systems with large ring (activity) rates
( 5). Previous attempts to accomplish a sim-
ilar analysis in a biochemical signaling sys-
tem have not been convincing. Using quan-
titative immunocytochemistry, Cheong et
al. obtained four-dimensional arrays of data
(time, dosage, genotype, and pathway) for
thousands of individually resolved cells.
With this much data, it becomes possible to
estimate the entropies required to establish
the channel capacity, accounting for known
biases in entropy measurements due to nite
(but large) sample sizes.
Biochemical systems often form net-
works with multiple interactions (cross
talk) that make the analysis of statistical
dependencies, and hence channel capaci-
ties, problematic. Here, Cheong et al. mea-
sured two readouts in the same system. The
amount of information carried by the com-
bination of the ATF-2 and the NF-B path-
way depends, in principle, on the topology
of interactions between the pathways. If both
pathways get their information directly from
the incoming signal, the mutual informa-
tion of the two responses taken together can
potentially be substantially increased. This
situation is described as a bush topology,
because the signals propagate directly along
multiple branches from the root of the net-
work. Alternatively, the network could have
a tree topology, in which the input signal
passes through a common intermediate (the
trunk of the tree) before going to independent
readout branches. Remarkably, the differ-
Departments of Mathematics, Biology, and Cognitive
Science, Case Western Reserve University, Cleveland, OH
44106, USA. E-mail: pjthomas@case.edu
21 OCTOBER 2011 VOL 334 SCIENCE www.sciencemag.org 322
PERSPECTIVES
ence in the mutual information predicted in
these two scenarios is large enough, and the
data set collected by the authors substantial
enough, to unambiguously identify the TNF
NF-BATF-2 network as having a treelike,
rather than bushlike, topology. Each pathway
alone carries just short of one bit of informa-
tion. [In information theory, one bit, or one
binary digit, is the information needed to dis-
tinguish exactly two equally likely outcomes
of an experiment (e.g., heads versus tails in a
fair coin toss). Two bits is enough informa-
tion to distinguish four outcomes, and N bits
is enough information to distinguish 2
N
out-
comes.] However, the joint response of both
pathways carries just over one bit, enough to
reliably distinguish a single yes-no decision,
such as presence or absence of TNF. Indeed,
the treelike topology likely reects the fact
that before branching out to stimulate multi-
ple pathways, the TNF response is mediated
by a single receptor signaling (and pathway
activation) complex. As the authors show, this
arrangement limits the capacity of the net-
work. Even if an unlimited number of inde-
pendent pathways separately read the output
of the TNF receptor complex, the information
obtained by a single cell about the amount of
stimulatory TNF cannot exceed about 1.25
bits, which is only slightly more than from the
NF-B and ATF-2 pathways alone.
The capacity of any channel is intimately
related to the nature of the noise affecting the
input-output relationship, and it can be dif-
cult to establish quantitative characteristics
of the noise sources present in a biochemical
channel. In engineering, one usually consid-
ers the properties of a communications chan-
nel (such as a telegraph line or ber optic
cable) as fixed, and the engineering prob-
lem amounts to choosing the best encoding
scheme given the (noisy) signal transmission
equipment. A decade ago, it was proposed
that biological information channels should
be sculpted by evolutionary pressures so as
to match to the information sources that feed
into them ( 6). In sensory neural systems,
source-channel matching takes the form of
redundancy reduction and efcient coding
( 7 9). The pioneering work by Cheong et al.
opens the door to similar analyses for bio-
chemical signaling systems.
The biggest surprise in the analysis of
Cheong et al. is how small the informa-
tion capacity of the TNF signaling system
appears to be. Noise suppression in bio-
chemical systems is notoriously challeng-
ing for cells to implement (and for investi-
gators to analyze) ( 10). Possible strategies
for increasing the information about a signal
such as TNF concentration include reduc-
ing noise by introducing a negative-feedback
loop, averaging over noise by integrating
received signals over time, or pooling infor-
mation across multiple cells. The authors
explore each possibility. Negative feedback
proves to be a double-edged sword: It can
reduce the amount of noise in transducing a
signal, which tends to boost capacity, but it
can also restrict the dynamic range of the sig-
nal input, reducing capacity. In a mutant cell
line lacking a certain negative-feedback loop
that normally inhibits the TNF receptor com-
plex, the authors found increased informa-
tion capacity in the short term (output at 30
min after stimulation) but decreased capac-
ity in the long term (4 hours after stimula-
tion). In both cases, however, the information
capacity was still approximately
one bit. [In a signaling system
comprising a rate-modulated Pois-
son processfor example, com-
municating by directly detecting
occurrence of chemical events
it can be proven mathematically
that feedback cannot increase the
channel capacity ( 11).] Integration
over time (or averaging over time)
does not fare much better, because
once activated, NF-B activity
remains approximately constant in
any given cell. In other biochemi-
cal systems, time integration could
still play a role in noise reduction.
Of the strategies explored, pooling
information across multiple cells
that each independently process
the same input signal (an example
of the bush topology) holds the
most promise. The authors show
that readouts from local pools of 14 cells give
twice the information of single cells alone.
We can look forward to more contribu-
tions from information theory to cell biol-
ogy (and vice versa), particularly in gra-
dient sensing the process by which motile
eukaryotic cells orient themselves during
chemotaxis ( 12). In this process, the ensem-
ble of inputs is all the possible directions of
the chemoattractant gradient ( 13 15).
In 2002, Berger asserted that biology had
become positioned to prot meaningfully
from an invasion by information theorists
( 6). Nearly a decade later, the invasion is
well under way and shows no sign of slow-
ing down.
References
1. T. Hunter, B. M. Sefton, Proc. Natl. Acad. Sci. U.S.A. 77,
1311 (1980).
2. C. Shannon, Bell Syst. Tech. J. 27, 379 and 623 (1948).
3. T. M. Cover, J. A. Thomas, Elements of Information Theory
(Wiley, New York, ed. 2, 2006).
4. R. Cheong, A. Rhee, C. J. Wang, I. Nemenman,
A. Levchenko, Science 334, 354 (2011); 10.1126/
science.1204553.
5. F. Rieke, D. Warland, R. de Ruyter van Steveninck,
W. Bialek, Spikes: Exploring the Neural Code (MIT Press,
Cambridge, MA, 1999).
6. T. Berger, IEEE Inf. Theory Soc. Newsl. 53 (March 2003).
7. H. Barlow, in Sensory Communication, W. A. Rosenblith,
Ed. (MIT Press, Boston, MA, 1961), pp. 217234.
8. B. A. Olshausen, D. J. Field, Nature 381, 607 (1996).
9. E. C. Smith, M. S. Lewicki, Nature 439, 978 (2006).
10. I. Lestas et al., Nature 467, 174 (2010).
11. Y. M. Kabanov, Theory Probab. Appl. 23, 143 (1978).
12. C. A. Parent, P. N. Devreotes, Science 284, 765 (1999).
13. B. W. Andrews, P. A. Iglesias, PLOS Comput. Biol. 3, e153
(2007).
14. J. M. Kimmel, R. M. Salter, P. J. Thomas, Adv. Neural Inf.
Process. Syst. 19, 705 (2007).
15. D. Fuller et al., Proc. Natl. Acad. Sci. U.S.A. 107, 9656
(2010).
Receiver Destination
Noise
source
Signal Received
signal
Message Message
Information
source
Transmitter
Information flow. Schematic diagram of a general communication system ( 2). The information capacity of a
signaling system quanties its ability to transmit information (measured in bits), whether it is an engineered cable or
a living cell.
10.1126/science.1213834 C
R
E
D
I
T
:

Y
.

H
A
M
M
O
N
D
/
S
C
I
E
N
C
E
Globalization, Land Use, and the
Invasion of West Nile Virus
A. Marm Kilpatrick
Many invasive species that have been spread through the globalization of trade and travel are
pathogens. A paradigmatic case is the introduction of West Nile virus (WNV) into North America
in 1999. A decade of research on the ecology and evolution of WNV includes three findings
that provide insight into the outcome of future pathogen introductions. First, WNV transmission
in North America is highest in urbanized and agricultural habitats, in part because the hosts
and vectors of WNV are abundant in human-modified areas. Second, after its introduction,
the virus quickly adapted to infect local mosquito vectors more efficiently than the originally
introduced strain. Third, highly focused feeding patterns of the mosquito vectors of WNV result
in unexpected host species being important for transmission. This research provides a framework
for predicting and preventing the emergence of foreign vector-borne pathogens.
T
he growth of human populations and the
development of rapid transportation sys-
tems have made the worlds biota more
connected than at any time in Earths history. The
result has been a breakdown in biogeographic
barriers and the introduction of species into nov-
el habitats. Globally, introduced invasive spe-
cies are estimated to cause >$120 billion in
damage annually (1) and include several patho-
gens that have direct impacts on the health of
humans, livestock, and wildlife. Pathogens spread
by trade and travel in the past 500 years include
those causing the human diseases malaria, dengue,
and HIV/AIDS; wildlife and livestock pathogens,
such as anthrax, rinderpest, rabies, and avian ma-
laria; and numerous diseases of crops and wild
plants, including chestnut blight, potato blight,
and sudden oak death (2, 3). Introductions have
continued with invasions by novel strains of in-
fluenza, severe acute respiratory syndrome, and
West Nile virus (WNV), among many others. The
factors that determine the outcome and impact of
invasions are frequently complex and poorly un-
derstood (4, 5); however, extensive research on
WNVover the past decade has enabled a detailed
exploration of its invasion, including pathways of
introduction, interactions with the biotic and abi-
otic environment in the new region (Fig. 1), and
impacts on ecosystem health.
WNV is a single-stranded RNA virus in the
family Flaviviridae that includes several impor-
tant humans pathogens: dengue, Japanese enceph-
alitis, and yellow fever viruses (6). WNV was
first isolated in 1937 from a febrile patient in
Uganda, and subsequent studies showed that WNV
transmission was endemic and widespread across
tropical parts of Africa, southern Asia, and north-
ern Australia, and episodic in more temperate
parts of Europe (7). As with other vector-borne
diseases, the warmer temperatures in the tropics
facilitate longer transmission seasons and some-
times increased transmission intensity through
faster mosquito and virus development and in-
creased biting rates. In some populations in Af-
rica, >80% of people over 15 years old have
antibodies to WNV (8); however, WNV was
previously considered nearly asymptomatic and
in the 1950s was even tried as an anticancer
therapy (9).
In 1999, WNV was introduced into North
America, where it spread rapidly with major ec-
onomic and public health consequences (7). The
virus reached the west coast in only 4 years
(Fig. 2), with regional epidemics in 2002 and
2003 and more localized epidemics occurring in
other years. Between 1999 and 2010, ~1.8 mil-
lion people were infected, with ~360,000 ill-
nesses, 12,852 reported cases of encephalitis/
meningitis, and 1308 deaths. The threat of WNV
infection has led to the costly implementation of
national blood donor screening, as well as vac-
cine and drug development (10). Public out-
reach campaigns have altered human behavior,
including the time spent outdoors, especially
by older people, who are at high risk for WNV
disease.
The impacts of WNVon wildlife have been
yet more severe than those on humans. Millions
of birds have died from WNV infection, and
regional-scale population declines of >50% have
been observed for several species (11). The range
of taxa that have suffered declines is surprisingly
large and includes corvids, chickadees and tit-
mice, wrens, and thrushes (Fig. 1) and probably
others. Some populations have recovered after
initial declines, whereas others have not (11). The
REVIEW
Department of Ecology and Evolutionary Biology, Univer-
sity of California, Santa Cruz, CA 95064, USA. E-mail:
akilpatr@ucsc.edu
Fig. 1. An American robin (T. migratorius) and its nestlings. Robins flourish in human-altered landscapes
and appear to play a key role in WNV amplification across many regions of North America. [Photo credit:
Bruce Lyon]
www.sciencemag.org SCIENCE VOL 334 21 OCTOBER 2011 323
ecological and economic consequences of these
regional declines in bird populations have not yet
been elucidated and need further study.
Globalization and the Introduction of Pathogens
The probability of the introduction and establish-
ment of introduced species has been shown to
increase with the propagule pressure, or the
rate at which individual organisms are introduced
to a newregion (Fig. 3) (5). The yearly propagule
pressure and the pathway by which WNVreached
North America in 1999 remain unknown, but
several possibilities have been proposed, including
mosquitoes being transported by shipping, air-
planes, or wind; migratory birds or birds in trade;
and humans traveling (12). The large and increas-
ing volume of air traffic into NewYork City over
the past five decades makes the transport of in-
fected mosquitoes on an airplane a likely path-
way. A close phylogenetic relationship between
viruses isolated in New York in 1999 and those
circulating in Israel in the previous year sug-
gests a possible Middle East origin (6). Trade
and travel have also previously introduced key
mosquito vectors of WNV, Culex pipiens and
C. quinquefasciatus, as well as vectors for den-
gue, yellow fever, and chikungunya viruses, such
as Aedes albopictus and A. aegypti (13).
What predictions could have been made in
1999 about the outcome of the introduction of
WNVinto New York City that summer? An an-
swer comes fromcomparing the ecology of WNV
transmission in the Americas with that in Africa
and Europe (14, 15).
WNV Ecology in Its Native Range
Studies of endemic WNV transmission in Egypt,
Sudan, and South Africa and of Kunjin virus, a
subtype of WNV, in Australia showthat the virus
was most frequently isolated fromCulex mosqui-
toes. In Australia, most isolations come from
C. annulirostris, which is a competent laboratory
vector (16). In Africa, C. univittatus makes up the
largest fraction of WNV-infected mosquitoes
(8, 17). Interestingly, there is little evidence of
WNVinfection in C. pipiens in South Africa, de-
spite frequent feeding on avian hosts. C. pipiens is
an important WNV vector in Europe and North
America (15, 18). It is possible that the lower
WNVinfection prevalence observed in C. pipiens
than in C. univittatus can be attributed to its being
less susceptible to infection (17).
Accurate quantification of the contribution of
different host species to viral amplification re-
quires data on mosquito feeding patterns and host
abundance from the same place and time, com-
bined with information on the duration and in-
tensity of host infectiousness (19). Host abundance
and mosquito feeding data have never been col-
lected simultaneously for WNV hosts and vec-
tors in Africa, Asia, Australia, or Europe and have
only rarely been collected in North America. As a
result, only tentative conclusions can be drawn
about the relative importance of host species for
WNV outside North America and these largely
come from studies of seroprevalence and infec-
tiousness based on viremia (concentration of virus
in the blood) observed after experimental infec-
tions. In Egypt, hooded crows (Corvus cornix)
and house sparrows (Passer domesticus) had
high antibody prevalence and infectiousness (8).
In South Africa, waterbirds (ducks and rails) and
passerine birds in the family Ploceidae (weavers
and Old World sparrows, including house spar-
rows) were most infectious and most frequently
had antibodies to WNV (17).
The Vectors, Hosts, and Transmission
of WNV in the Americas
The most important vectors in North America
share some similarities with those in Africa, Eu-
rope, and Australia. Although C. univittatus is not
present in the Americas, C. pipiens, C. quinque-
fasciatus, and several other species that take the
majority of their blood meals from birds are
found in North America, including C. restuans,
C. tarsalis, and C. nigripalpus. Based on their
feeding ecology and their vector competence, all
these species would be expected to be important
in enzootic (bird-to-bird) transmission (18, 20).
In addition, their abundance in anthropogenically
modified areas points to a significant role in hu-
man WNVepidemics.
An important insight was gained in the course
of determining the vector species responsible for
transmitting WNV from nonhuman animals to
humans (i.e., bridge vectors). Bridge vectors
were initially thought to be mosquito species that
fed frequently on mammals (such as Aedes mos-
quitoes), but an integrated analysis of the abun-
dance, infection prevalence, feeding patterns, and
vector competence of a wide range of mosquitoes
indicated that C. pipiens and C. restuans mos-
quitoes, which frequently take <15%of their blood
meals from humans, may nonetheless be respon-
sible for the majority of human infections in sev-
eral regions (18, 21). Their importance results from
their higher relative abundance and WNV infec-
tion prevalence than the more anthropophilic mos-
quito species. Applying this integrated approach
to other pathogens may simplify targets for vec-
tor control, especially when the same species serves
as both the bridge and the enzootic vector.
0
47
0.0
0.2
0.4
0.6
0.8
1.0
Not detected yet
1999
2000
2001
2002
2003
2004
2005
2007
2009
1 2 3 4 5 6 7 8 9 10 11
Years since WNV detection in that state
Years of WNV
detection
R
e
l
a
t
i
v
e

n
u
m
b
e
r

o
f

c
a
s
e
s
47 47 47 47 47 47 46 43 27 12 4
Fig. 2. Spread of WNV throughout the Americas. The map shows the year that WNV was first detected in a
state, province, or country. The box plot shows the temporal pattern of WNV incidence at the state level
after WNV arrival. The y axis shows the relative number of WNV neuroinvasive cases (the fraction of the
maximum observed in that state) that occurred in each state in each year, starting with the year WNV was
first detected in birds, mosquitoes, humans, or horses. The number of states included in each column is
shown above the box.
21 OCTOBER 2011 VOL 334 SCIENCE www.sciencemag.org 324
REVIEW
Predicting which avian hosts would be im-
portant for WNV transmission in the Americas
based on data from Africa or Europe would have
failed, because inferences based solely on abun-
dance, infectiousness, or serological (antibody)
prevalence can be misleading (19). Analyses of
data from the mid-Atlantic to Colorado that have
combined host abundance and mosquito feeding
data with host infectiousness suggested that al-
though introduced house sparrows (P. domesticus)
and crows (Corvus spp.) are abundant and/or
highly infectious, they appear to be of minor im-
portance in WNV transmission (2123). Crows
make up a small fraction of mosquito blood meals,
and house sparrows are rarely fed on by mos-
quitoes relative to their abundance, resulting in
few bites per individual and inefficient trans-
mission. Instead, a species of thrush, the Amer-
ican robin (Turdus migratorius), appears to be
more important in WNV transmission (Fig. 4)
(2123). This is primarily because 30 to 80% of
mosquito feedings by the dominant WNV vec-
tors (C. pipiens, C. restuans, and C. tarsalis) are on
robins, despite robins making up only 1 to 20%
of the avian communities studied. Questions
that arise are why are robins so frequently fed
on by mosquitoes, and do robins share a trait
with other thrushes that makes them generally
important for avian arboviruses? For example,
serological studies of the avian Sindbis virus in
Sweden indicated higher exposure of thrushes
than of any other group (24).
Research has also shown that focused feeding
on robins amplifies WNV transmission intensity
(22). This raises the following question: If Amer-
ican robins, which have increased 50 to 100% in
abundance over the past 25 years with the ur-
banization of the North American landscape (11),
were less abundant, would WNVtransmission be
lower? Its uncertain, because if mosquito abun-
dance and feeding patterns remain constant, de-
creasing host abundance increases the vector:host,
ratio which increases transmission intensity. In
addition, seasonal decreases in robin abundance
have been correlated with a shift in mosquito feed-
ing from birds to humans, which increases hu-
man WNVinfections (25). However, in the western
and southern United States, where robins are less
abundant, they provide only a small fraction of
mosquito blood meals, and yet mosquito feeding
by another species, C. quinquefasciatus, on hu-
mans is no greater than in the east (26). As a
result, the impact of reducing robin abundance on
WNVtransmission is unknown and would proba-
bly depend on the identity, abundance, and infec-
tiousness of alternate sources of mosquito blood
meals.
In summary, three important insights have
been gained in determining the amplification hosts
of WNVin North America. First, abundant hosts
Human land use
Urbanization, Agriculture
Globalization of trade & travel
Greenhouse gases
Increases in human
commensal vectors
and hosts
Travel
Trade in animals
Animal migration
Altered CO
2
,
temperatures,
and precipitation
Fig. 3. Anthropogenic processes that facilitate the introduction and es-
tablishment of novel pathogens and increase their transmission. Trade,
travel, and animal movement introduce new pathogens. Climate, hosts,
and the abundance and feeding ecology of vectors determine establish-
ment and transmission intensity. Land use modifies animal communities
that serve as hosts and vectors for pathogens, and climate change alters
pathogen and vector demographic rates. [Image credits: Google and Tele
Atlas (aerial photos); Edward Canda (rice paddy); Photos8.com (corn-
field); L. Hufnagel (air traffic map); Dori (dori@merr.info) (smokestacks);
Joe Hoyt (left mosquito); Andrew Flemming (right mosquito); Richard Kuhn,
Purdue Department of Biological Sciences (virus); NASA (clouds); Marm
Kilpatrick (others)]
www.sciencemag.org SCIENCE VOL 334 21 OCTOBER 2011 325
REVIEW
may be fed on infrequently by vectors, making
them less important in transmission. Second, the
importance of hosts may be determined more by
how frequently mosquitoes feed on them than by
variation in their infectiousness. Finally, variation
in the abundance of key avian hosts can have
unpredictable impacts on transmission, especially
to humans.
Hosts and Vectors as Ecological Niches
for Pathogens
Studies of WNV amplification hosts show how
the feeding patterns and competence of insect
vectors and vertebrate hosts create ecological
niches for introduced vector-borne pathogens. Char-
acterization of these niches can informpredictions
of establishment probability for pathogen intro-
ductions (19) and augment projections that are
frequently based on climatic conditions alone.
Such predictions can be used to guide manage-
ment decisions in allocating resources toward
prevention of pathogen introductions (such as
vaccine development and testing and quarantine
of imported animals). One example of the insight
gained from host and vector competence studies
comes from an elegant comparison of WNVand
St. Louis encephalitis virus (SLEV, a flavivirus
native to the Americas) by Reisen et al. (27), which
showed that WNV is more infectious in hosts to
biting vectors than SLEV and explained why
WNVepidemics are more severe than those caused
by SLEV.
Land Use and WNV Transmission
Recent evidence has suggested that at the county
scale in eastern and western North America, hu-
man WNVincidence increases with urbanization
and agriculture, respectively (28) (Figs. 3,4). This
may result from the habitats used and human-
commensal nature of three important WNVmos-
quito vectors, C. pipiens, C. quinquefasciatus, and
C. tarsalis, although the exact mechanisms acting
at local scales are not yet known. Nevertheless,
the distribution of WNVindicates that it is similar
to other pathogens whose transmission is linked
with anthropogenic land use and increased abun-
dance of domesticated animals and human-tolerant
wildlife species (Figs. 3 and 4). For example,
H5N1 avian influenza emerged from poultry in-
tensification; rabies transmission in the Serengeti
is maintained by domestic dogs; Lyme disease
increases with the fragmentation of forests in
eastern North America; and yellow fever, den-
gue, and chikungunya viruses are all transmitted
by the anthropophilic mosquitoes A. aegypti and
A. albopictus (2933). Perhaps ecologically based
land-use planning, combined with improved de-
velopment and sanitation, could reduce contact
with and the abundance of human-commensal spe-
cies and hence transmission of their pathogens.
Coevolution of Hosts, Vectors, and Pathogens
Rapid coevolution between WNV and its hosts,
vectors, and other pathogens is expected based
on reciprocal fitness impacts and in many cases, the
lack of shared evolutionary history (11, 27, 34, 35).
Still, it was somewhat surprising that by 2005,
the strain of WNVthat was introduced into North
America in 1999 (NY99) had been displaced
continent-wide by a locally evolved genotype,
WN02 (36). WN02 was first detected in 2001
and spread continent-wide between 2002 and
2004. Viruses in the WN02 clade consistently
differ from NY99 viruses by only three nucle-
otides that result in one amino acid change.
Nonetheless, WN02 viruses are more efficient-
ly transmitted by both C. pipiens and C. tarsalis
mosquitoes, and the difference was found to in-
crease with temperature in the laboratory, as would
be expected if the WN02 viruses replicate at a
higher rate (37, 38).
Fig. 4. WNV ecology across an urbanization gradient in the northeastern and
midwestern United States. WNV is transmitted primarily by C. pipiens mosqui-
toes among a wide range of birds, but American robins (outlined) are a key
amplification host. The diversity of avian hosts decreases with urbanization,
whereas C. pipiens abundance appears to increase. [Image credit: U.S. Geo-
logical Survey (mosquito); Marm Kilpatrick and Ryan Peters (others)]
21 OCTOBER 2011 VOL 334 SCIENCE www.sciencemag.org 326
REVIEW
The difference between NY99 and WN02 vi-
ruses in competence (that is, in magnitude and
duration of infectiousness) in avian hosts has not
yet been determined. However, another single-
stranded RNAvirus, Venezuelan equine enceph-
alitis virus, has repeatedly evolved to be able to
infect novel hosts and mosquito vectors efficient-
ly, and this shows that host adaption is also
possible (33).
There may be evolutionary selective pressure
for WNV to kill its avian hosts. Individual birds
that die from WNV infection have higher vi-
remia, and thus infectiousness to biting mosqui-
toes, than individuals that survive (27, 34), and
host illness from infection decreases host de-
fenses against biting mosquitoes, which would
increase vector-host contact rates. Both of these
mechanisms increase pathogen fitness by in-
creasing host-to-vector transmission. In addition,
in contrast to an assumption made in many mod-
els of the evolution of virulence, host death from
WNV does not appear to reduce the length of
the infectious period of the host: Most avian
hosts that survive WNV infection clear virus
from their blood between days 4 and 6 after in-
fection, and most individuals that die from WNV
infection do so at approximately the same time
after infection (27, 34). Akey question is whether
viral evolution that increased replication and vi-
rulence in hosts would have deleterious effects
in the vector.
It is also unknown whether North American
birds have evolved increased resistance to WNV.
This could be determined by repeating WNVlab-
oratory challenge experiments using individuals
from the same populations in which resistance
was previously measured early in the WNVepi-
demic (27, 34). Ideally, such studies would in-
clude a range of host species or populations that
have experienced different selective pressures
exerted on them by WNV; for example, in terms
of WNV transmission intensity or initial suscep-
tibility to WNVmortality (for example, doves are
more resistant than corvids) (27, 34).
Outlook: Unanswered Questions
A key question is will WNV follow the boom-
and-bust pattern seen in some plant and animal
species invasions (5)are the worst WNV epi-
demics behind us? WNV epidemics peaked in
many states the year after it arrived, with fewer
human cases having been observed subsequent-
ly (Fig. 2). This reduction in WNV disease has
led to reduced research focus and less funding
from public health agencies for WNV, and, more
recently, less testing for WNV by health care
providers.
Reduced transmission may be a product of
several factors, including elevated immunity in
birds or humans, especially the subset of people
most at risk: the homeless and those spending
more time outdoors at dusk (7). However, annual
recruitment of young-of-the-year birds apparent-
ly fuels WNV(39), which reduces the importance
of avian host immunity in suppressing transmis-
sion. Instead, its possible that WNVtransmission
is modulated primarily by rainfall and temper-
ature, and if so, climatic conditions in 2002 and
2003 were especially suitable. Climate is known
to influence many aspects of WNV transmission,
including mosquito abundance, biting rate, and
survival as well as viral replication within the
mosquito (37). If WNV transmission is regulated
by climate, then severe epidemics could recur,
especially if they are facilitated by climate change
(Fig. 3). It is notable that the largest number of
neuroinvasive WNVcases observed in NewYork
State was in 2010, 11 years after the virus was
introduced. Clearly, determining the relative roles
of climate versus other factors in year-to-year
variation in transmission is important and neces-
sary to predict the long-term trajectory of WNV
in North America.
Asecond unanswered question is why havent
more morbidity and mortality been reported in
horses, humans, and birds in tropical regions
(7, 40)? Less surveillance is undertaken in these
less-developed countries than in North Amer-
ica, and the presence of other diseases, such as
dengue, malaria, and Chagas, whose public health
impact dwarfs that of WNV, could account for
lower reporting, despite similar WNVincidence
and illness. Alternatively, cross-protection by anti-
bodies or evolved resistance to illness from other
flaviviruses (such as SLEV, dengue, or yellow
fever viruses) in humans and horses may decrease
illness, without reducing bird-mosquito trans-
mission. In addition, enzootic transmission may
be lower in the tropics than in North America,
owing to cross-protecting flavivirus antibodies in
birds, a mismatch between periods of peak mos-
quito abundance and susceptible young-of-the-
year birds, or lower infectiousness of tropical
avian hosts. These mechanisms may be operating
simultaneously.
Continual introduction of pathogens to new
regions is inevitable in our globally connected
planet. It is unclear which vector-borne pathogen
will be the next to cross hemispheres, but many
viruses of public health concern exist in Africa,
Asia, and Europe, including other arthropod-borne
viruses such as Japanese encephalitis, Rift Valley
fever, tick-borne encephalitis, and chikungunya
(31). Conversely, there are several pathogens from
the Americas that could be introduced into the
Old World, including Venezuelan equine enceph-
alitis and SLEV. Insights gained from studying
the invasion of WNV can be used to help predict
which are the highest-risk pathogens for estab-
lishment after cross-hemispheric introduction (19).
Gaining an understanding of the ecology of zoo-
notic viruses, combined with fast-developing re-
combinant vaccine technologies that have already
been applied to wildlife (41), could formthe basis
of a strategy to prevent the emergence of newly
introduced pathogens.
References and Notes
1. D. Pimentel, R. Zuniga, D. Morrison, Ecol. Econ. 52,
273 (2005).
2. P. K. Anderson et al., Trends Ecol. Evol. 19, 535 (2004).
3. P. Daszak, A. A. Cunningham, A. D. Hyatt, Science 287,
443 (2000).
4. R. N. Mack et al., Ecol. Appl. 10, 689 (2000).
5. A. K. Sakai et al., Annu. Rev. Ecol. Syst. 32, 305
(2001).
6. R. S. Lanciotti et al., Science 286, 2333 (1999).
7. L. R. Petersen, E. B. Hayes, Med. Clin. North Am. 92,
1307, ix (2008).
8. H. S. Hurlbut, F. Rizk, R. M. Taylor, T. H. Work, Am. J.
Trop. Med. Hyg. 5, 579 (1956).
9. C. M. Southam, A. E. Moore, Am. J. Trop. Med. Hyg.
3, 19 (1954).
10. D. W. C. Beasley, Immunotherapy 3, 269 (2011).
11. S. L. LaDeau, A. M. Kilpatrick, P. P. Marra, Nature 447,
710 (2007).
12. A. M. Kilpatrick et al., Conserv. Biol. 20, 1224 (2006).
13. L. P. Lounibos, Annu. Rev. Entomol. 47, 233 (2002).
14. Z. Hublek, J. Halouzka, Emerg. Infect. Dis. 5, 643
(1999).
15. M. V. Fyodorova et al., J. Med. Entomol. 43, 552 (2006).
16. R. A. Hall, J. H. Scherret, J. S. Mackenzie, Ann. N.Y. Acad.
Sci. 951, 153 (2001).
17. B. M. McIntosh, P. G. Jupp, D. B. Dickinson, G. M. McGillivray,
J. Sweetnam, S. Afr. J. Med. Sci. 32, 1 (1967).
18. A. M. Kilpatrick et al., Emerg. Infect. Dis. 11, 425 (2005).
19. Supporting material is available on Science Online.
20. M. J. Turell, M. R. Sardelis, M. L. O'Guinn, D. J. Dohm,
in Japanese Encephalitis and West Nile Viruses Vol. 267
Current Topics in Microbiology and Immunology,
J. Mackenzie, A. Barrett, V. Deubel, Eds. (Springer-Verlag,
Berlin, 2002), pp. 241252.
21. G. L. Hamer et al., Am. J. Trop. Med. Hyg. 80, 268
(2009).
22. A. M. Kilpatrick, P. Daszak, M. J. Jones, P. P. Marra,
L. D. Kramer, Proc. Biol. Sci. 273, 2327 (2006).
23. R. Kent, L. Juliusson, M. Weissmann, S. Evans, N. Komar,
J. Med. Entomol. 46, 380 (2009).
24. J. O. Lundstrm, K. M. Lindstrm, B. Olsen, R. Dufva,
D. S. Krakower, J. Med. Entomol. 38, 289 (2001).
25. A. M. Kilpatrick, L. D. Kramer, M. J. Jones, P. P. Marra,
P. Daszak, PLoS Biol. 4, e82 (2006).
26. G. Molaei et al., Am. J. Trop. Med. Hyg. 83, 1269 (2010).
27. W. K. Reisen, Y. Fang, V. M. Martinez, J. Med. Entomol.
42, 367 (2005).
28. S. E. Bowden, K. Magori, J. M. Drake, Am. J. Trop.
Med. Hyg. 84, 234 (2011).
29. J. L. Kyle, E. Harris, Annu. Rev. Microbiol. 62, 71
(2008).
30. B. F. Allan, F. Keesing, R. S. Ostfeld, Conserv. Biol. 17,
267 (2003).
31. S. C. Weaver, W. K. Reisen, Antiviral Res. 85, 328 (2010).
32. T. Lembo et al., J. Appl. Ecol. 45, 1246 (2008).
33. S. C. Weaver, A. D. T. Barrett, Nat. Rev. Microbiol. 2,
789 (2004).
34. N. Komar et al., Emerg. Infect. Dis. 9, 311 (2003).
35. J. N. Thompson, Am. Nat. 153, S1 (1999).
36. C. T. Davis et al., Virology 342, 252 (2005).
37. A. M. Kilpatrick, M. A. Meola, R. M. Moudy, L. D. Kramer,
PLoS Pathog. 4, e1000092 (2008).
38. R. M. Moudy, M. A. Meola, L. L. Morin, G. D. Ebel,
L. D. Kramer, Am. J. Trop. Med. Hyg. 77, 365 (2007).
39. G. L. Hamer et al., Vector Borne Zoonotic Dis. 8,
57 (2008).
40. N. Komar, G. G. Clark, Pan Am. J. Public Health 19,
112 (2006).
41. C. A. Russell, D. L. Smith, J. E. Childs, L. A. Real,
PLoS Biol. 3, e88 (2005).
Acknowledgments: I thank T. Bogich, K. Koelle, I. Parker,
N. Komar, and the Kilpatrick lab for input. NSF grants
EF-0914866 and BCS-0826779 and NIH grant
1R01AI090159-01 provided funding.
Supporting Online Material
www.sciencemag.org/cgi/content/full/334/6054/323/DC1
SOM Text
Fig. S1
References
10.1126/science.1201010
www.sciencemag.org SCIENCE VOL 334 21 OCTOBER 2011 327
REVIEW
The Influence of Nonlinear
Mesoscale Eddies on Near-Surface
Oceanic Chlorophyll
Dudley B. Chelton,
1
* Peter Gaube,
1
Michael G. Schlax,
1
Jeffrey J. Early,
2
Roger M. Samelson
1
Oceanic Rossby waves have been widely invoked as a mechanism for large-scale variability of
chlorophyll (CHL) observed from satellites. High-resolution satellite altimeter measurements have
recently revealed that sea-surface height (SSH) features previously interpreted as linear Rossby
waves are nonlinear mesoscale coherent structures (referred to here as eddies). We analyze 10 years
of measurements of these SSH fields and concurrent satellite measurements of upper-ocean CHL
to show that these eddies exert a strong influence on the CHL field, thus requiring reassessment
of the mechanism for the observed covariability of SSH and CHL. On time scales longer than 2
to 3 weeks, the dominant mechanism is shown to be eddy-induced horizontal advection of CHL by
the rotational velocities of the eddies.
A
decade of concurrent satellite measure-
ments of sea surface height (SSH) and
upper-ocean chlorophyll (CHL) is en-
abling studies of physical-biological interaction
that are not feasible from ship-based observations.
Although satellites provide only near-surface in-
formation about ocean physics and biology, they
are the only practical means of obtaining dense,
global observations. Altimetric measurements of
SSH reveal that westward propagation is ubiqui-
tous (1) with characteristics similar to the linear
Rossby waves by which the ocean adjusts to wind
and thermal forcing (2). Westward propagation
is also evident in CHL estimates derived from sat-
ellite measurements of ocean color. The wide-
spread interpretation of the westward-propagating
SSH variations as Rossby waves led naturally to
interpretations that the CHL variations are also
induced by Rossby waves (35).
The mechanism for Rossby wave influence
on CHL has been debated (49), in part because
of inconsistency in the lag between variations of
SSH and CHL. The most widely accepted view
is that the covariability between SSH and CHL
arises from cyclical advection of CHL by the
horizontal velocity field associated with passing
Rossby waves (79).
The prevailing view before the recent focus
on Rossby wave influence was that CHL concen-
tration is influenced by nonlinear eddies (1015).
Investigations of this eddy influence have con-
tinued in parallel with Rossby wave studies.
Here, we show that the copropagation of CHL
and SSH previously interpreted as having been
caused by Rossby waves is in fact attributable
to eddies.
Nonlinearity of SSH variability. High-
resolution SSH fields produced by merging the
measurements from two simultaneously operat-
ing satellite altimeters (16) reveal that westward-
propagating features previously believed to be
linear Rossby waves are actually nonlinear rotat-
ing coherent structures (eddies) with radii of
~100 km (17, 18). Because such mesoscale fea-
tures propagate westward with approximately
the speed of long Rossby waves (1719), they can
masquerade as Rossby waves in low-resolution
SSH fields constructed from measurements by a
single altimeter.
The degree of nonlinearity of a mesoscale fea-
ture is characterized by the ratio of the rotational
fluid speed U to the translation speed c of the
feature. When U/c > 1, the feature is nonlinear,
which allows it to maintain a coherent structure
as it propagates (20). This requires that all of
the wavelength components of the feature prop-
agate at the same speed, i.e., nondispersively. With
linear Rossby wave dynamics, features that are
initially spatially compact quickly lose their co-
herent structure through dispersion (21).
At latitudes higher than 25, 98% of the fea-
tures tracked for 10 weeks have U/c > 1 (fig.
S2). The degree of nonlinearity is slightly less
at lower latitudes where the propagation speeds
c are faster (17, 18). But even in the latitude range
15 to 25, 95% of the tracked features have
U/c > 1 (fig. S2).
Westward copropagation of SSH and CHL.
The revised interpretation of westward-propagating
SSH as nonlinear eddies mandates a reassessment
of past conclusions that the westward copropa-
gation of CHL and SSH is indicative of Rossby
wave influence on CHL. The alternative hy-
pothesis that CHL variability is eddy-induced is
examined here from 10 years of concurrent mea-
surements of SSH and CHL in the southeastern
Pacific (SEP) near 20S that has been a focus of
past studies (6, 7).
The trajectories of mesoscale eddies (18) in
the SEP are shown in Fig. 1A. Compared with
eddies observed globally in the latitude range
15 to 25, their mean amplitude is smaller (3.2 cm
versus 6.2 cm) but their mean radius is the same
(110 km). Because U is approximately propor-
tional to eddy amplitude, eddies in the SEP are
less nonlinear (fig. S2); 87% have U/c > 1.
The mean CHL distribution has a generally
northward gradient over most of the SEP (Fig.
1B). The influence of eddies is evident from the
sinuous character of the CHL field at any par-
ticular time (Fig. 1, C and D). The distortions
of an otherwise smoothly varying CHL field
are most apparent in regions of strong CHL
gradient.
Eddy influence on the CHL field becomes
clearer after filtering to remove the large-scale
and seasonally varying CHL and SSH (22). West-
ward copropagation of CHL and SSH is apparent
from time-longitude plots of the resulting anom-
aly fields (Fig. 2, A and B). The trajectories of
the centroids of clockwise (CW) and counter-
clockwise (CCW) rotating eddies in the SEP co-
incide, respectively, with negative and positive
extrema of westward-propagating SSH (Fig. 2A).
The positive lag of maximum positive correla-
tion in Fig. 2C indicates that the SSH extrema at
the eddy centroids lag the extrema of westward-
propagating CHL by ~1 month in the eastern
SEP, decreasing to ~0.5 month in the west. There
is a weaker negative correlation at negative lags
of 1 to 1.5 months.
Eddy influence on CHL. To interpret the lagged
correlations in Fig. 2C, anomaly CHL was com-
posite averaged within eddy interiors in a trans-
lating and rotated coordinate system in which the
large-scale CHL gradient vector is oriented at a
polar angle of 90 (22). The CHL anomaly com-
posites consist of dipoles with opposing signs
and with different orientations in CWand CCW
rotating eddies (Fig. 3A). As indicated by the
ratio r in Fig. 3A, the dipoles are asymmetric in
both cases with larger magnitudes in the left half
of each composite, corresponding to the leading
half of these westward-propagating eddies. The
displacements from the eddy centroid are smaller
for these primary poles than for the secondary
poles of opposite sign in the trailing (right) halves
of the eddies (see also fig. S4).
The negative extremum of SSH at the cen-
troids of CW rotating eddies in the SEP is strad-
dled by negative and positive poles of CHL to the
west and east, respectively. The opposite occurs
in CCW rotating eddies, for which the positive
extremum of SSH at the centroids is straddled
by positive and negative poles of CHL to the
west and east, respectively. The parallel bands
of positive and negative lagged correlations in
Fig. 2C thus arise from a combination of west-
RESEARCHARTICLES
1
College of Oceanic and Atmospheric Sciences, 104 COAS Ad-
ministration Building, Oregon State University, Corvallis, OR
973315503, USA.
2
Northwest Research Associates, Post Of-
fice Box 3027, Bellevue, WA 98009, USA.
*To whom correspondence should be addressed. E-mail:
chelton@coas.oregonstate.edu
21 OCTOBER 2011 VOL 334 SCIENCE www.sciencemag.org 328
ward propagation and the zonal alignments of
the monopole extrema of SSH and dipole ex-
trema of CHL.
The geographical patterns of anomalous CHL
within the eddy interiors in this region of gen-
erally northward CHL gradient are indicative of
horizontal advection of CHL by the rotational
velocities of the eddies. For CW rotating eddies
(Figs. 3A and 4A, top panels), the northward
velocity in the western half of each eddy advects
low CHL from south to north, resulting in anom-
alously low CHL in the northwest quadrant. The
southward velocity in the eastern half of each
such eddy advects high CHL from north to south,
resulting in anomalously high CHL in the south-
east quadrant. The opposite rotational sense of
CCW rotating eddies (Figs. 3A and 4A, bottom
panels) results in anomalously high and low
CHL in the southwest and northeast quadrants,
respectively.
The importance of composite averaging
the CHL in a rotated coordinate system is clear
from Fig. 4, A and B. The dipole patterns of
anomaly CHL result from a combination of
the rotational sense of the eddies and the di-
rection of the CHL gradient. These dipoles are
manifest as distortions of the total CHL field
(Fig. 1, C and D).
The CHL anomaly within the trailing half
of the eddy is generally weaker and noisier
than within the leading half because the trailing
half encounters a CHL field that has been
distorted by the leading half. The noisiness
of the secondary poles accounts for the some-
what weaker negative correlations in Fig. 2C,
as well as their smaller composite average mag-
nitudes compared with the primary poles (Fig. 3A).
Within an individual eddy, the structure of
the dipole of anomaly CHL from eddy-driven
advection varies, depending on the strength
and orientation of the geographically and tem-
porally varying gradient of CHL, the degree
of eddy nonlinearity U/c, and the influence of
other eddies that have recently perturbed the
CHL field. Past confusion about geographi-
cal and temporal variations of the lag relation-
ships between SSH and CHL is therefore not
surprising.
Rotational advection of CHL by eddies is
reproduced in numerical model simulations of
random westward-propagating eddies in a tracer
field with a northward gradient (22). The dipoles
of anomalous tracer concentration for weakly
nonlinear eddies (Fig. 3B) are very similar to the
dipoles of observed CHL in Fig. 3A. The model
reproduces the asymmetry of the magnitudes of
the dipoles, as well as the smaller offset between
the eddy centroid and the primary pole in the
leading halves of the eddies compared with the
secondary pole in the trailing halves. The higher
rotational velocities within strongly nonlinear ed-
dies result in tracer dipoles with larger magni-
tudes and with centers advected farther around
the eddy interiors (Fig. 3C).
Composite averaging separately for the SEP
eddies east and west of 108W (figs. S3 and S5)
reveals that the longitudinal variations of the
couplet of positive and negative lagged correla-
tions in Fig. 2C are attributable to longitudinal
variations of the structures of the CHL dipoles.
The shorter lag of maximum positive correlation
in the west is due to smaller displacements
between the eddy centroids and the primary
poles of CHL anomaly in the leading halves
of the eddies. The near-symmetry of the lags
of positive and negative correlation bands in the
east is consistent with the near-symmetric dis-
placements of the dipole centers from the eddy
centroids.
The geographical patterns of eddy-induced
CHL anomalies in the SEP are similar to the
patterns found in other regions of northward
gradient of CHL. Composite averages of anomaly
CHL computed globally between latitudes of
15 and 45 are shown in Fig. 3D for the
tracked eddies within regions of northward CHL
gradient (22). The telltale asymmetric dipole
patterns from opposing meridional advection
in opposite halves of the eddies are readily
apparent.
Anomaly CHL was composite averaged for
regions of generally southward CHL gradient
by rotating the translating coordinate system
Fig. 1. Geographical characteristics of observed
SSH and CHL in the SEP. (A) The trajectories from
a 16-year data record of eddies that rotate clock-
wise (CW, blue lines) and counterclockwise (CCW,
red lines), with the starting locations shown by
solid circles. (B) The 10-year average log
10
(CHL)
for CHL in units of mg m
3
, with a contour in-
terval of log
10
(CHL) = 0.1, increasing northward,
and with the thick line corresponding to log
10
(CHL) =
1.3. (C) An example map for 7 March 2001
showing log
10
(CHL) in color with contours of
positive and negative anomaly SSH (solid and
dashed lines, respectively) at intervals of 2 cm,
excluding the zero contour. (D) The same as (C),
except showing anomaly SSH in color with con-
tours of log
10
(CHL) at the same interval as in (B).
The horizontal lines in each panel are the section
along which the time-longitude plots in Fig. 2 and
the spectra in Fig. 5, A to C, were computed.
A
B
C
D
www.sciencemag.org SCIENCE VOL 334 21 OCTOBER 2011 329
RESEARCH ARTICLES
so that the large-scale CHL gradient vector is
oriented at a polar angle of 90. The axes of
the CHL dipoles in these composite averages
(Fig. 3E) are rotated ~180 relative to those
in regions of northward CHL gradient. These
mirrored dipoles still have larger magnitude in
the leading (left) half of the westward-propagating
eddies.
When U/c > 1, fluid is trapped within the
eddy interior and transported along the eddy tra-
jectory (20, 21). This nonlinear signal of sustained
transport of fluid trapped in the eddy core is not
A B C D E
Fig. 3. Composite averages of filtered fields in a rotated and normalized
coordinate system (22) within the interiors of CW (top panels) and CCW
rotating eddies (bottom panels). (A) log
10
(CHL) in the region 18S to 22S,
130W to 80W. (B) A tracer field in a model simulation seeded with weakly
nonlinear Gaussian eddies. (C) A tracer field in the model seeded with strongly
nonlinear Gaussian eddies; (D) log
10
(CHL) globally between 15 and 45
latitude in regions of northward CHL gradient. (E) log
10
(CHL) globally be-
tween 15 and 45 latitude in regions of southward CHL gradient. The outer
perimeter of each circle corresponds to twice the eddy radius scale L
s
(22).
The vectors in each panel are the gradient of the composite average SSH,
which is proportional to the geostrophic velocity. The number N of eddy
realizations in the composite average and the magnitude r of the ratio of the
primary pole in the leading (left) half of each composite to the secondary
pole in the trailing (right) half are labeled on each panel. The estimated
95% confidence intervals along profiles connecting the dipole extrema in
each of these composite averages are shown in fig. S4.
Fig. 2. Spatial and tem-
poral variability of filtered
SSH and log
10
(CHL) obser-
vations (22) along 20S be-
tween 130W and 80W.
Time-longitude sections of
westward-only propagation
over a 3-year portion of the
10-year period analyzed
here are shown for (A) SSH
with eddy tracks within T2
of 20S overlaid (dashed and
solid lines for CW and CCW
rotating eddies, respectively);
(B) log
10
(CHL) with the same
eddy tracks overlaid; and (C)
the lagged cross-correlation
between log
10
(CHL) at time
t and SSH at time t + lag,
calculated over the full 10-year data record; the white areas correspond to
correlations smaller than the estimated 95% significance level of 0.083 (22).
Positive lags correspond to log
10
(CHL) leading SSH, and the contour interval
is 0.2 with the zero contour omitted for clarity. Analogous time-longitude
sections and lagged cross correlations are shown for the SSH and tracer
fields from a quasi-geostrophic model in fig. S6.
A B C
21 OCTOBER 2011 VOL 334 SCIENCE www.sciencemag.org 330
RESEARCH ARTICLES
the signature revealed in the CHL observations
described here. The extrema of the dipole CHL
features occur near the radius of maximum rota-
tional velocity that defines the outer edge of the
eddy core (21). As the westward-translating eddies
impinge on preexisting CHL gradients, ambient
fluid is advected part way around the portion of
the eddy outside of the core region of closed
streamlines in the translating reference frame
(20, 21).
Spectral analysis of SSH and CHL. Wave
numberfrequency spectra of SSH and CHL
provide further evidence that the westward propa-
gation of CHL is induced by nonlinear eddies.
The nondispersive character of nonlinear eddies
implies that the spectral variance should fall ap-
proximately along a straight line in wave number
frequency space. In contrast, the spectral variance
for linear Rossby waves would be constrained to
frequencies defined by their dispersion relation.
The wave numberfrequency spectrum of
SSH in the SEP (Fig. 5A) is clearly inconsistent
with the theoretical Rossby wave dispersion
relation, regardless of whether the classical the-
ory is modified to account for effects of mean
currents or rough bottom topography. Where-
as the dispersion relations from these theories
all flatten with increasingly negative wave num-
ber, the spectral variance of observed SSH
extends to higher frequencies, straddling the
Fig. 5. Zonal (east-west) wave numberfrequency
spectra. (A) Filtered SSH for 130W to 80W along
20S (see fig. S1 for SSH spectra along four other
zonal sections). (B) SSH for Gaussian approxima-
tions of each tracked eddy for 130W to 80W along
20S [different color bars are used to accommodate
the smaller variance of the Gaussian approximations
(23)]. (C) Filtered log
10
(CHL) for 130W to 80W
along 20S. (D) SSH from days 3000 to 15,000 of a
model simulation with purely linear dynamics. (E)
SSH from days 3000 to 15,000 of a model sim-
ulation with nonlinear quasi-geostrophic dynam-
ics (see also fig. S7, A to C). (F) A tracer field in
the same model simulation as in (E) (see also fig.
S7, D to F). The negative wave numbers correspond
to westward propagation. The units are cm for SSH,
log
10
of mg m
3
for CHL, and arbitrary for the
tracer field. The straight lines are the mean eddy
speeds from observed SSH [4.9 cm s
1
, (A) to (C)]
and from SSH in the model [4.3 cm s
1
, (D) to
(F)]. The curved lines correspond to the dispersion
relations for linear Rossby waves from the classical
theory for a flat bottom, no mean currents, and
zero meridional (north-south) wave number (2)
(solid), a theory that accounts for mean currents
(25) (dashed), and a theory for rough bottom
topography and no mean currents (26) (dotted).
The latter two are irrelevant to the SSH fields from
the flat-bottom model with no mean currents in (D)
to (F).
A B C
D E F
Fig. 4. Schematic diagram of eddy-driven horizontal advection of
CHL for CW and CCW rotating eddies (top and bottom, respectively)
propagating westward in regions where the CHL gradient is (A)
northward and (B) northeastward. An otherwise smooth contour of
CHL (dashed lines) is distorted by the rotational velocity field within
the eddy, as shown by the solid lines. Advection of CHL within the
large-scale background CHL gradient results in the positive and
negative CHL anomalies shown by the red and purple regions, re-
spectively. The dependence of the locations of these CHL anomalies
on the direction of the large-scale background CHL gradient that is
evident from comparison of (A) and (B) was accounted for by com-
posite averaging in a coordinate system rotated for each eddy (22).
www.sciencemag.org SCIENCE VOL 334 21 OCTOBER 2011 331
RESEARCH ARTICLES
straight line of constant (nondispersive) prop-
agation speed.
We tested our hypothesis that the spectral
characteristics of SSH in Fig. 5A are attributable
to eddies with compact structures by construct-
ing SSH fields consisting only of the interiors
of the tracked eddies and zero outside of the ed-
dies. Each eddy was approximated as an axi-
symmetric Gaussian feature with amplitude and
scale estimated from the automated tracking
procedure (18). The spectral variance of the re-
sulting eddy-only SSH fields (Fig. 5B) strad-
dles the same straight line as the spectrum of
total SSH (23, 24). The discrepancies between
the wave numberfrequency spectra of SSH and
the Rossby wave dispersion relations are thus
consistent with SSH variability being attributa-
ble to a field of propagating eddies.
The importance of nonlinearity is clarified
from model simulations. The spectral variance of
model SSH with linear dynamics is restricted to
frequencies below the dispersion relation for
Rossby waves with zero meridional wave num-
ber (Fig. 5D). The spread of variance to frequen-
cies below this dispersion relation is from Fourier
components with finite meridional wave num-
bers (2).
With nonlinear quasi-geostrophic dynamics,
the spectral variance of model SSH primarily rep-
resents long-lived, coherent eddy features, in which
the small-scale spectral components are phase-
locked to the large-scale components (21). The
resulting spectrum (Fig. 5E) consists of a non-
dispersive band along a straight line that extends
to frequencies higher than are allowed by linear
Rossby wave dynamics, very similar to the spec-
trum of observed SSH in Fig. 5A.
A short spur of spectral variance centered
at about 0.007 cycles per day (cpd) extends along
the dispersion relation in the model SSH spec-
trum (Fig. 5E). This spur arises from small-
amplitude SSH variability outside the nonlinear
cores of the eddies. The extent of the spur is
found to decrease with increasing nonlinearity
of the eddies (fig. S7). Although not apparent in
the spectrum of observed SSH in Fig. 5A, sug-
gestions of similar spurs exist in SSH spectra in
other regions (fig. S1).
Determination of the spectrum of CHL var-
iability is more challenging than for SSH (22).
The spectrum of CHL in the SEP (Fig. 5C) is
nonetheless similar to the spectrum of SSH. In
particular, the spectral variance is concentrated
along the same straight line of nondispersive
variability to higher frequencies than are allowed
by linear Rossby waves theories. Moreover, the
spectral variance is restricted to smaller negative
wave numbers than would be the case if the
CHL variability were induced by linear Rossby
waves. A short spur of spectral variance straddles
the dispersion relation at about 0.008 cpd for
the same reason discussed above for the model
SSH spectra. Similar spurs are found in spectra
of tracer fields in quasi-geostrophic model sim-
ulations (Fig. 5F and fig. S7, D to F). The re-
stricted wave number extents of these spurs are
an important distinction from the spectral char-
acteristics in Fig. 5D that would be found if the
CHL or tracer variability were attributable to
linear Rossby waves.
Conclusions. Westward copropagation of CHL
and SSH that has previously been attributed to
linear Rossby waves is actually caused by non-
linear eddies that were not resolvable in the SSH
fields analyzed in past studies. This eddy influ-
ence is manifest as distortions of the CHL field
and is most evident in regions of strong gradi-
ents of the CHL field. Eddy influence on CHL
becomes clearer after filtering. The distinctly
different dipole structures of the resulting anom-
alous CHL distribution within the interiors of
CW and CCW rotating eddies (Fig. 3, A, D,
and E) and their similarity to the dipoles of a
tracer field in model simulations (Fig. 3, B and C)
are evidence that the dominant mechanism for
eddy-induced CHL variability on the time scales
>2 to 3 weeks considered here (22) is horizontal
advection of CHL by the rotational velocity with-
in the interior of each eddy.
The dominance of horizontal advection as
the mechanism for the observed westward prop-
agation of CHL variability has been suggested
previously (79) but has been attributed to hori-
zontal advection by linear Rossby waves, rather
than to nonlinear eddies. [The role of eddies in
defining the dipole structure of CHL anomalies
has been suggested for the central North At-
lantic (14, 15).] The distinction between linear
Rossby waves and eddies is important because
the dynamics of nonlinear eddies differ fun-
damentally from those of linear Rossby waves.
While distinct from the rotational advection
identified here, eddies can trap and transport
fluid parcels and their associated water proper-
ties (20, 21), including nutrients, CHL, and zoo-
plankton. They can also upwell nutrient-rich water
by various mechanisms (1015), thus stimulat-
ing the growth of phytoplankton and increasing
CHL in the eddy core. Such eddy-induced en-
hancements often occur deep in the euphotic zone
where they can be undetectable in the satellite-
based measurements of near-surface ocean color
(1215).
Because of the unique trapping of fluid in
the cores of nonlinear eddies, it is perhaps sur-
prising that the CHL distribution associated with
the eddies consists of dipoles with extrema out-
side of the eddy cores, rather than monopoles of
positive or negative CHL anomalies trapped at
the eddy centers. Monopole structures with very
active physical-biological interaction are some-
times observed within eddy cores. In contrast to
the ubiquitous presence of rotational advec-
tion around the outer portion of the eddies iden-
tified here, however, such monopole structures
are usually episodic, often with time scales short-
er than the 2- to 3-week filtering applied here.
Although it is unclear whether the rotational
advection that dominates the variability on longer
time scales has important biological consequences,
the results presented here clarify the dynamical
importance of eddies to the observed CHL
distribution.
References and Notes
1. D. B. Chelton, M. G. Schlax, Science 272, 234 (1996).
2. A. E. Gill, Atmosphere-Ocean Dynamics (Academic Press,
Cambridge, 1982).
3. P. Cipollini, D. Cromwell, P. G. Challenor, S. Raffaglio,
Geophys. Res. Lett. 28, 323 (2001).
4. B. M. Uz, J. A. Yoder, V. Osychny, Nature 409, 597 (2001).
5. D. A. Siegel, Nature 409, 576 (2001).
6. Y. Dandonneau, A. Vega, H. Loisel, Y. du Penhoat,
C. Menkes, Science 302, 1548 (2003).
7. P. D. Killworth, P. Cipollini, B. M. Uz, J. R. Blundell,
J. Geophys. Res. 109, C07002 (2004).
8. G. Charria et al., J. Mar. Res. 64, 43 (2006).
9. E. Gutknecht, I. Dadou, G. Charria,
P. Cipollini, V. Garon, J. Geophys. Res. 115,
C05004 (2010).
10. P. G. Falkowski, D. Ziemann, Z. Kolber, P. K. Bienfang,
Nature 352, 55 (1991).
11. D. J. McGillicuddy Jr. et al., J. Geophys. Res. 104, 13381
(1999).
12. D. A. Siegel, D. J. McGillicuddy Jr., E. A. Fields,
J. Geophys. Res. 104, 13359 (1999).
13. D. J. McGillicuddy Jr. et al., Science 316, 1021 (2007).
14. D. A. Siegel et al., Deep Sea Res. Part II Top. Stud.
Oceanogr. 55, 1218 (2008).
15. D. A. Siegel, P. Peterson, D. J. McGillicuddy Jr.,
S. Maritorena, N. B. Nelson, Geophys. Res. Lett. 38,
L13608 (2011).
16. N. Ducet, P.-Y. Le Traon, G. Reverdin, J. Geophys. Res.
105, 19477 (2000).
17. D. B. Chelton, M. G. Schlax, R. M. Samelson,
R. A. de Szoeke, Geophys. Res. Lett. 34, L15606 (2007).
18. D. B. Chelton, M. G. Schlax, R. M. Samelson, Prog.
Oceanogr. 91, 167 (2011).
19. J. C. McWilliams, G. R. Flierl, J. Phys. Oceanogr. 9, 1155
(1979).
20. G. R. Flierl, Geophys. Astrophys. Fluid Dyn. 18, 39 (1981).
21. J. J. Early, R. M. Samelson, D. B. Chelton, J. Phys. Oceanogr.
41, 1535 (2011).
22. Information on methods is available as supporting online
material on Science Online.
23. The somewhat smaller variance is due to unavoidable
biases in the amplitudes of compact mesoscale
features as estimated by the automated procedure (18).
24. The large spectral variance at small wave numbers
(long wavelengths) in the spectrum of eddy-only SSH
fields is easily understood qualitatively from consideration
of a single Gaussian eddy. Because the Fourier
transform of a Gaussian in space is a Gaussian in wave
number, the associated spectrum is dominated by
variance at small wave numbers. It is evident from
Fig. 5B that this dominance of spectral variance at
small wave numbers for a single Gaussian feature is
maintained in the spectrum of a superposition of
many Gaussian eddies.
25. P. D. Killworth, D. B. Chelton, R. A. de Szoeke, J. Phys.
Oceanogr. 27, 1946 (1997).
26. R. Tailleux, J. C. McWilliams, J. Phys. Oceanogr. 31, 1461
(2001).
Acknowledgments: We thank D. McGillicuddy and T. Farrar
for discussions throughout the course of this study.
We also thank L.-L. Fu, V. Combes, D. Siegel, P. Cipollini,
E. Shulenberger, and two anonymous reviewers for
helpful comments on the manuscript. This research was
supported by NASA grants NNX08AI80G and
NNX08AR37G and NSF Award 0621134.
Supporting Online Material
www.sciencemag.org/cgi/content/full/science.1208897/DC1
Materials and Methods
Figs. S1 to S7
References
25 May 2011; accepted 26 August 2011
Published online 15 September 2011;
10.1126/science.1208897
21 OCTOBER 2011 VOL 334 SCIENCE www.sciencemag.org 332
RESEARCH ARTICLES
Light Propagation with Phase
Discontinuities: Generalized Laws of
Reflection and Refraction
Nanfang Yu,
1
Patrice Genevet,
1,2
Mikhail A. Kats,
1
Francesco Aieta,
1,3
Jean-Philippe Tetienne,
1,4
Federico Capasso,
1
* Zeno Gaburro
1,5
*
Conventional optical components rely on gradual phase shifts accumulated during light
propagation to shape light beams. New degrees of freedom are attained by introducing abrupt
phase changes over the scale of the wavelength. A two-dimensional array of optical resonators
with spatially varying phase response and subwavelength separation can imprint such phase
discontinuities on propagating light as it traverses the interface between two media. Anomalous
reflection and refraction phenomena are observed in this regime in optically thin arrays of metallic
antennas on silicon with a linear phase variation along the interface, which are in excellent
agreement with generalized laws derived from Fermats principle. Phase discontinuities provide
great flexibility in the design of light beams, as illustrated by the generation of optical vortices
through use of planar designer metallic interfaces.
T
he shaping of the wavefront of light with
optical components such as lenses and
prisms, as well as diffractive elements such
as gratings and holograms, relies on gradual phase
changes accumulated along the optical path. This
approach is generalized in transformation optics
(1, 2), which uses metamaterials to bend light
in unusual ways, achieving such phenomena as
negative refraction, subwavelength-focusing, and
cloaking (3, 4) and even to explore unusual ge-
ometries of space-time in the early universe (5).
A new degree of freedom of controlling wave-
fronts can be attained by introducing abrupt phase
shifts over the scale of the wavelength along the
optical path, with the propagation of light gov-
erned by Fermats principle. The latter states that
the trajectory taken between two points A and B
by a ray of light is that of the least optical path,

B
A
n(

r )dr, where n(

r ) is the local index of re-

fraction, and readily gives the laws of reflection
and refraction between two media. In its most
general form, Fermats principle can be stated as
the principle of stationary phase (68); that is,
the derivative of the phase
B
A
d(

r ) accumu-
lated along the actual light path will be zero with
respect to infinitesimal variations of the path. We
show that an abrupt phase shift F(

r
s
) over the
scale of the wavelength can be introduced in the
optical path by suitably engineering the interface
between two media; F(

r
s
) depends on the co-
ordinate

r
s
along the interface. Then, the total
phase shift F(

r
s
)
B
A

k d

r will be stationary
for the actual path that light takes;

k is the wave
vector of the propagating light. This provides a
generalization of the laws of reflection and re-
fraction, which is applicable to a wide range of
subwavelength structured interfaces between two
media throughout the optical spectrum.
Generalized laws of reflection and refraction.
The introduction of an abrupt phase shift, de-
noted as phase discontinuity, at the interface be-
tween two media allows us to revisit the laws of
reflection and refraction by applying Fermats
principle. Consider an incident plane wave at an
angle q
i
. Assuming that the two paths are infi-
nitesimally close to the actual light path (Fig. 1),
then the phase difference between them is zero
k
o
n
i
sin(q
i
)dx (F dF)
k
o
n
t
sin(q
t
)dx F 0 1
where q
t
is the angle of refraction; F and F+dF
are, respectively, the phase discontinuities at the
locations where the two paths cross the interface;
dx is the distance between the crossing points; n
i
and n
t
are the refractive indices of the two media;
and k
o
= 2p/l
o
, where l
o
is the vacuum wave-
length. If the phase gradient along the interface is
designed to be constant, the previous equation
leads to the generalized Snells law of refraction
sin(q
t
)n
t
sin(q
i
)n
i

l
o
2p
dF
dx
2
Equation 2 implies that the refracted beam can
have an arbitrary direction, provided that a suit-
able constant gradient of phase discontinuity along
the interface (dF/dx) is introduced. Because of
the nonzero phase gradient in this modified Snells
law, the two angles of incidence Tq
i
lead to dif-
ferent values for the angle of refraction. As a
consequence, there are two possible critical an-
gles for total internal reflection, provided that
n
t
< n
i
:
q
c
arcsin

n
t
n
i

l
o
2pn
i
dF
dx

3
Similarly, for reflection we have
sin(q
r
) sin(q
i
)
l
o
2pn
i
dF
dx
4
where q
r
is the angle of reflection. There is a
nonlinear relation between q
r
and q
i
, which is
markedly different from conventional specular re-
flection. Equation 4 predicts that there is always a
critical angle of incidence
q

c
arcsin 1
l
o
2pn
i

dF
dx

5
above which the reflected beam becomes
evanescent.
In the above derivation, we have assumed that
F is a continuous function of the position along
the interface; thus, all the incident energy is trans-
ferred into the anomalous reflection and refraction.
However, because experimentally we use an array
of optically thin resonators with subwavelength
separation to achieve the phase change along
the interface, this discreteness implies that there
are also regularly reflected and refracted beams,
which follow conventional laws of reflection
and refraction (dF/dx = 0 in Eqs. 2 and 4). The
separation between the resonators controls
the amount of energy in the anomalously re-
flected and refracted beams. We have also
assumed that the amplitudes of the scattered
radiation by each resonator are identical, so that
the reflected and refracted beams are plane waves.
In the next section, we will showwith simulations
which represent numerical solutions of Maxwells
1
School of Engineering and Applied Sciences, Harvard Uni-
versity, Cambridge, MA 02138, USA.
2
Institute for Quantum
Studies and Department of Physics, Texas A&M University,
College Station, TX 77843, USA.
3
Dipartimento di Fisica e
Ingegneria dei Materiali e del Territorio, UniversitPolitecnica
delle Marche, via Brecce Bianche, 60131 Ancona, Italy.
4
Lab-
oratoire de Photonique Quantique et Molculaire, Ecole Nor-
male Suprieure de Cachan and CNRS, 94235 Cachan, France.
5
Dipartimento di Fisica, Universit degli Studi di Trento, via
Sommarive 14, 38100 Trento, Italy.
*To whom correspondence should be addressed. E-mail:
capasso@seas.harvard.edu (F.C.); gaburro@seas.harvard.
edu (Z.G.)
Fig. 1. Schematics used to derive the generalized
Snells law of refraction. The interface between the
two media is artificially structured in order to in-
troduce an abrupt phase shift in the light path,
which is a function of the position along the in-
terface. F and F + dF are the phase shifts where
the two paths (blue and red) cross the boundary.
www.sciencemag.org SCIENCE VOL 334 21 OCTOBER 2011 333
RESEARCH ARTICLES
equationshow, indeed, one can achieve the
equal-amplitude condition and the constant phase
gradient along the interface through suitable de-
sign of the resonators.
There is a fundamental difference between the
anomalous refraction phenomena caused by phase
discontinuities and those found in bulk designer
metamaterials, which are caused by either negative
dielectric permittivity and negative magnetic
permeability or anisotropic dielectric permittivity
with different signs of permittivity tensor com-
ponents along and transverse to the surface (3, 4).
Phase response of optical antennas. The phase
shift between the emitted and the incident radia-
tion of an optical resonator changes appreciably
across a resonance. By spatially tailoring the geom-
etry of the resonators in an array and hence their
frequency response, one can design the phase
discontinuity along the interface and mold the
wavefront of the reflected and refracted beams in
nearly arbitrary ways. The choice of the reso-
nators is potentially wide-ranging, from electro-
magnetic cavities (9, 10) to nanoparticle clusters
(11, 12) and plasmonic antennas (13, 14). We
concentrated on the latter because of their widely
tailorable optical properties (1519) and the ease
of fabricating planar antennas of nanoscale thick-
ness. The resonant nature of a rod antenna made
of a perfect electric conductor is shown in Fig.
2A (20).
Phase shifts covering the 0-to-2p range are
needed to provide full control of the wavefront.
To achieve the required phase coverage while
maintaining large scattering amplitudes, we used
the double-resonance properties of V-shaped an-
tennas, which consist of two arms of equal length
h connected at one end at an angle D (Fig. 2B).
We define two unit vectors to describe the ori-
entation of a V-antenna: along the symmetry axis
of the antenna and perpendicular to (Fig. 2B).
V-antennas support symmetric and antisym-
metric modes (Fig. 2B, middle and right), which
are excited by electric-field components along
and axes, respectively. In the symmetric mode,
the current distribution in each arm approximates
Fig. 2. (A) Calculated phase and amplitude of
scattered light froma straight rod antenna made of
a perfect electric conductor (20). The vertical dashed
line indicates the first-order dipolar resonance of
the antenna. (B) A V-antenna supports symmetric
and antisymmetric modes, which are excited, re-
spectively, by components of the incident field along
and axes. The angle between the incident po-
larization and the antenna symmetry axis is 45.
The schematic current distribution is represented
by colors on the antenna (blue for symmetric and
red for antisymmetric mode), with brighter color
representing larger currents. The direction of cur-
rent flow is indicated by arrows with color gradient.
(C) V-antennas corresponding to mirror images of
those in (B). The components of the scattered elec-
tric field perpendicular to the incident field in (B)
and (C) have a p phase difference. (D and E) An-
alytically calculated amplitude and phase shift of
the cross-polarized scattered light for V-antennas
consisting of gold rods with a circular cross section
and with various length h and angle between the
rods D at l
o
= 8 mm(20). The four circles in (D) and
(E) indicate the values of h and D used in exper-
iments. The rod geometry enables analytical cal-
culations of the phase and amplitude of the scattered
light, without requiring the extensive numerical
simulations needed to compute the same quan-
tities for flat antennas with a rectangular cross-
section, as used in the experiments. The optical
properties of a rod and flat antenna of the same
length are quantitatively very similar, when the
rod antenna diameter and the flat antenna
width and thickness are much smaller than the
length (20). (F) Schematic unit cell of the plasmonic
interface for demonstrating the generalized laws of
reflection and refraction. The sample shown in Fig. 3A
is created by periodically translating in the x-y plane
the unit cell. The antennas are designed to have
equal scattering amplitudes and constant phase
difference DF =p/4 between neighbors. (G) Finite-
difference time-domain (FDTD) simulations of the
scattered electric field for the individual antennas
composing the array in (F). Plots show the scat-
tered electric field polarized in the x direction for
y-polarized plane wave excitation at normal in-
cidence from the silicon substrate. The silicon
substrate is located at z 0. The antennas are equally spaced at a sub-
wavelength separation G/8, where G is the unit cell length. The tilted red
straight line in (G) is the envelope of the projections of the spherical waves
scattered by the antennas onto the x-z plane. On account of Huygenss
principle, the anomalously refracted beam resulting from the superposi-
tion of these spherical waves is then a plane wave that satisfies the
generalized Snells law (Eq. 2) with a phase gradient |dF/dx| = 2p/G along
the interface.
21 OCTOBER 2011 VOL 334 SCIENCE www.sciencemag.org 334
RESEARCH ARTICLES
that of an individual straight antenna of length
h (Fig. 2B, middle), and therefore the first-order
antenna resonance occurs at h l
eff
/2, where
l
eff
is the effective wavelength (14). In the anti-
symmetric mode, the current distribution in each
arm approximates that of one half of a straight
antenna of length 2h (Fig. 2B, right), and the
condition for the first-order resonance of this
mode is 2h l
eff
/2.
The polarization of the scattered radiation
is the same as that of the incident light when
the latter is polarized along or . For an ar-
bitrary incident polarization, both antenna modes
are excited but with substantially different am-
plitude and phase because of their distinctive reso-
nance conditions. As a result, the scattered light
can have a polarization different from that of the
incident light. These modal properties of the
V-antennas allow one to design the amplitude,
phase, and polarization state of the scattered light.
We chose the incident polarization to be at 45
with respect to and so that both the symmetric
and antisymmetric modes can be excited and
the scattered light has a substantial component
polarized orthogonal to that of the incident light.
Experimentally, this allows us to use a polarizer
to decouple the scattered light fromthe excitation.
As a result of the modal properties of the
V-antennas and the degrees of freedomin choosing
antenna geometry (h and D), the cross-polarized
scattered light can have a large range of phases
and amplitudes for a given wavelength l
o
; ana-
lytical calculations of the amplitude and phase
response of V-antennas assumed to be made of
gold rods are shown in Fig. 2, D and E. In Fig.
2D, the blue and red dashed curves correspond to
the resonance peaks of the symmetric and anti-
symmetric modes, respectively. We chose four
antennas detuned from the resonance peaks, as
indicated by circles in Fig. 2, D and E, which
provide an incremental phase of p/4 from left to
right for the cross-polarized scattered light. By
simply taking the mirror structure (Fig. 2C) of an
existing V-antenna (Fig. 2B), one creates a new
antenna whose cross-polarized radiation has an ad-
ditional p phase shift. This is evident by observing
that the currents leading to cross-polarized radia-
tion are p out of phase in Fig. 2, Band C. Aset of
eight antennas were thus created from the initial
four antennas, as shown in Fig. 2F. Full-wave sim-
ulations confirm that the amplitudes of the cross-
polarizedradiationscatteredbythe eight antennas are
nearlyequal, withphases inp/4increments (Fig. 2G).
A large phase coverage (~300) can also be
achieved by using arrays of straight antennas (fig.
S3). However, to obtain the same range of phase
shift their scattering amplitudes will be substan-
tially smaller than those of V-antennas (fig. S3).
As a consequence of its double resonances, the
V-antenna instead allows one to design an array
with phase coverage of 2p and equal, yet high,
scattering amplitudes for all of the array elements,
leading to anomalously reflected and refracted
beams of substantially higher intensities.
Experiments on anomalous reflection and
refraction. We demonstrated experimentally the
generalized laws of reflection and refraction
using plasmonic interfaces constructed by peri-
odically arranging the eight constituent antennas
as explained in the caption of Fig. 2F. The spacing
between the antennas should be subwavelength
so as to provide efficient scattering and to prevent
the occurrence of grating diffraction. However, it
should not be too small; otherwise, the strong near-
field coupling between neighboring antennas
would perturb the designed scattering amplitudes
and phases. A representative sample with the
densest packing of antennas, G = 11 mm, is shown
in Fig. 3A, where G is the lateral period of the
antenna array. In the schematic of the experimen-
tal setup (Fig. 3B), we assume that the cross-
polarized scattered light from the antennas on the
left side is phase-delayed as compared with the
ones on the right. By substituting into Eq. 2 2p/G
for dF/dx and the refractive indices of silicon
and air (n
Si
and 1) for n
i
and n
t
, we obtain the
angle of refraction for the cross-polarized beam
q
t,
= arcsin[n
Si
sin(q
i
) l
o
/G]
Figure 3C summarizes the experimental results
of the ordinary and the anomalous refraction for
six samples with different G at normal incidence.
The incident polarization is along the y axis in
Fig. 3A. The sample with the smallest G corre-
sponds to the largest phase gradient and the most
Fig. 3. (A) Scanning electron microscope (SEM)
image of a representative antenna array fabricated
on a silicon wafer. The unit cell of the plasmonic
interface (yellow) comprises eight gold V-antennas
of width ~220 nm and thickness ~50 nm, and it
repeats with a periodicity of G = 11 mm in the x
direction and 1.5 mmin the y direction. (B) Schematic
experimental setup for y-polarized excitation (electric
field normal to the plane of incidence). (C and D)
Measured far-field intensity profiles of the refracted
beams for y- and x-polarized excitations, respective-
ly. The refraction angle is counted from the normal
to the surface. The red and black curves are mea-
sured with and without a polarizer, respectively, for
six samples with different G. The polarizer is used to
select the anomalously refracted beams that are
cross-polarized with respect to the excitation. The
amplitude of the red curves is magnified by a factor
of two for clarity. The gray arrows indicate the
calculated angles of anomalous refraction according
to Eq. 6.
6
www.sciencemag.org SCIENCE VOL 334 21 OCTOBER 2011 335
RESEARCH ARTICLES
efficient light scattering into the cross-polarized
beams. We observed that the angles of anoma-
lous refraction agree well with theoretical pre-
dictions of Eq. 6 (Fig. 3C). The same peak
positions were observed for normal incidence,
with polarization along the x axis in Fig. 3A(Fig.
3D). To a good approximation, we expect that the
V-antennas were operating independently at the
packing density used in experiments (20). The
purpose of using a large antenna array (~230 mm
by 230 mm) is solely to accommodate the size
of the plane-wavelike excitation (beam radius
~ 100 mm). The periodic antenna arrangement
is used here for convenience but is not necessary
to satisfy the generalized laws of reflection and
refraction. It is only necessary that the phase
gradient is constant along the plasmonic interface
and that the scattering amplitudes of the antennas
are all equal. The phase increments between
nearest neighbors do not need to be constant, if
one relaxes the unnecessary constraint of equal
spacing between nearest antennas.
The angles of refraction and reflection are
shown in Fig. 4, A and B, respectively, as a
function of q
i
for both the silicon-air interface
(black curves and symbols) and the plasmonic
interface (red curves and symbols) (20). In the
range of q
i
= 0 to 9, the plasmonic interface
Fig. 4. (A) Angle of refraction versus angle of incidence for the ordinary (black curve and triangles)
and anomalous refraction (red curve and dots) for the sample with G = 15 mm. The curves are
theoretical calculations made by using the generalized Snells law for refraction (Eq. 2), and the
symbols are experimental data extracted from refraction measurements as a function of the angle of
incidence (20). The shaded region represents negative refraction for the cross-polarized light, as
illustrated in the inset. The blue arrows indicate the modified critical angles for total internal reflection.
(B) Angle of reflection versus angle of incidence for the ordinary (black curve) and anomalous (red
curve and dots) reflection for the sample with G = 15 mm. The top left inset is the zoom-in view. The
curves are theoretical calculations made by using Eq. 4, and the symbols are experimental data
extracted from reflection measurements as a function of the angle of incidence (20). The shaded
region represents negative reflection for the cross-polarized light, as illustrated in the bottom right
inset. The blue arrow indicates the critical angle of incidence above which the anomalously reflected
beam becomes evanescent. Experiments with lasers emitting at different wavelengths show that the
plasmonic interfaces are broadband, anomalously reflecting and refracting light from l 5 mm to l
10 mm.
Fig. 5. (A) SEM image of a plasmonic interface that
creates an optical vortex. The plasmonic pattern
consists of eight regions, each occupied by one
constituent antenna of the eight-element set of Fig.
2F. The antennas are arranged so as to generate a
phase shift that varies azimuthally from 0 to 2p, thus
producing a helicoidal scattered wavefront. (B) Zoom-in
view of the center part of (A). (C and D) Respectively,
measured and calculated far-field intensity distributions
of an optical vortex with topological charge one. The
constant background in (C) is due to the thermal ra-
diation. (E and F) Respectively, measured and calcu-
lated spiral patterns created by the interference of the
vortex beam and a co-propagating Gaussian beam. (G
and H) Respectively, measured and calculated interfer-
ence patterns with a dislocated fringe created by the
interference of the vortex beam and a Gaussian beam
when the two are tilted with respect to each other. The
circular border of the interference pattern in (G) arises
from the finite aperture of the beam splitter used to
combine the vortex and the Gaussian beams (20). The
size of (C) and (D) is 60 mmby 60 mm, and that of (E)
to (H) is 30 mm by 30 mm.
21 OCTOBER 2011 VOL 334 SCIENCE www.sciencemag.org 336
RESEARCH ARTICLES
exhibits negative refraction and reflection for
the cross-polarized scattered light (schematics are
shown in the bottom right insets of Fig. 4, A and
B). The critical angle for total internal reflection
is modified to ~8 and +27 (Fig. 4A, blue ar-
rows) for the plasmonic interface in accordance
with Eq. 3, compared with T17 for the silicon-air
interface; the anomalous reflection does not exist
beyond q
i
= 57 (Fig. 4B, blue arrow).
At normal incidence, the ratio of intensity R
between the anomalously and ordinarily refracted
beams is ~0.32 for the sample with G = 15 mm
(Fig. 3C). R rises for increasing antenna packing
densities (Fig. 3, C and D) and increasing angles
of incidence [up to R 0.97 at q
i
= 14 (fig.
S1B)]. Because of the experimental configuration,
we are not able to determine the ratio of intensity
between the reflected beams (20), but we expect
comparable values.
Vortex beams created by plasmonic interfaces.
To demonstrate the versatility of the concept of
interfacial phase discontinuities, we fabricated a
plasmonic interface that is able to create a vortex
beam (21, 22) upon illumination by normally in-
cident linearly polarized light. Avortex beamhas
a helicoidal (or corkscrew-shaped) equal-phase
wavefront. Specifically, the beam has an azi-
muthal phase dependence exp(il) and carries an
orbital angular momentum of L l per photon
(23). Here, the topological charge l is an integer,
indicating the number of twists of the wavefront
within one wavelength; is the azimuthal angle
with respect to the beam axis; and is the
reduced Planck constant. These peculiar states
of light are commonly generated by using a spiral
phase plate (24) or a computer-generated holo-
gram (25) and can be used to rotate particles (26)
or to encode information in optical communica-
tion systems (27).
The plasmonic interface was created by arrang-
ing the eight constituent antennas as shown in
Fig. 5, Aand B. The interface introduces a spiral-
like phase shift with respect to the planar wave-
front of the incident light, creating a vortex beam
with l = 1. The vortex beam has an annular in-
tensity distribution in the cross section, as viewed
in a mid-infrared camera (Fig. 5C); the dark re-
gion at the center corresponds to a phase singu-
larity (22). The spiral wavefront of the vortex
beam can be revealed by interfering the beam
with a co-propagating Gaussian beam (25), pro-
ducing a spiral interference pattern (Fig. 5E). The
latter rotates when the path length of the Gaussian
beam was changed continuously relative to that
of the vortex beam (movie S1). Alternatively, the
topological charge l = 1 can be identified by a
dislocated interference fringe when the vortex and
Gaussian beams interfere with a small angle (Fig.
5G) (25). The annular intensity distribution and
the interference patterns were well reproduced in
simulations (Fig. 5, D, F, and H) by using the
calculated amplitude and phase responses of the
V-antennas (Fig. 2, D and E).
Concluding remarks. Our plasmonic inter-
faces, consisting of an array of V-antennas, im-
part abrupt phase shifts in the optical path, thus
providing great flexibility in molding of the op-
tical wavefront. This breaks the constraint of stan-
dard optical components, which rely on gradual
phase accumulation along the optical path to
change the wavefront of propagating light. We
have derived and experimentally confirmed gen-
eralized reflection and refraction laws and studied
a series of intriguing anomalous reflection and
refraction phenomena that descend from the
latter: arbitrary reflection and refraction angles
that depend on the phase gradient along the
interface, two different critical angles for total
internal reflection that depend on the relative
direction of the incident light with respect to the
phase gradient, and critical angle for the reflected
beam to be evanescent. We have also used a
plasmonic interface to generate optical vortices
that have a helicoidal wavefront and carry orbit-
al angular momentum, thus demonstrating the
power of phase discontinuities as a design tool
of complex beams. The design strategies presented
in this article allow one to tailor in an almost
arbitrary way the phase and amplitude of an
optical wavefront, which should have major im-
plications for transformation optics and integrated
optics. We expect that a variety of novel planar
optical components such as phased antenna arrays
in the optical domain, planar lenses, polarization
converters, perfect absorbers, and spatial phase
modulators will emerge from this approach.
Antenna arrays in the microwave and
millimeter-wave regions have been used for the
shaping of reflected and transmitted beams in
the so-called reflectarrays and transmitarrays
(2831). These typically consist of a double-layer
structure comprising a planar array of antennas
and a ground plane (in the case of reflectarrays)
or another array (in the case of transmitarrays),
separated by a dielectric spacer of finite thick-
ness. Reflectarrays and transmitarrays cannot be
treated as a single interface for which one can
write down the generalized laws because they
rely on both antenna resonances and the propa-
gation of waves in the spacer to achieve the
desired phase control. The generalization of the
laws of reflection and refraction we present is
made possible by the deeply subwavelength thick-
ness of our optical antenna arrays and their asso-
ciated abrupt phase changes, with no contribution
from propagation effects. These generalized laws
apply to the whole optical spectrum for suitable
designer interfaces and can be a guide for the
design of new photonic devices.
References and Notes
1. J. B. Pendry, D. Schurig, D. R. Smith, Science 312, 1780
(2006).
2. U. Leonhardt, Science 312, 1777 (2006).
3. W. Cai, V. Shalaev, Optical Metamaterials: Fundamentals
and Applications (Springer, New York, 2009).
4. N. Engheta, R. W. Ziolkowski, Metamaterials: Physics and
Engineering Explorations (Wiley-IEEE Press, New York,
2006).
5. I. I. Smolyaninov, E. E. Narimanov, Phys. Rev. Lett. 105,
067402 (2010).
6. S. D. Brorson, H. A. Haus, J. Opt. Soc. Am. B 5, 247 (1988).
7. R. P. Feynman, A. R. Hibbs, Quantum Mechanics and
Path Integrals (McGraw-Hill, New York, 1965).
8. E. Hecht, Optics (Addison Wesley, Boston, ed. 3,
1997).
9. H. T. Miyazaki, Y. Kurokawa, Appl. Phys. Lett. 89, 211126
(2006).
10. D. Fattal, J. Li, Z. Peng, M. Fiorentino, R. G. Beausoleil,
Nat. Photonics 4, 466 (2010).
11. J. A. Fan et al., Science 328, 1135 (2010).
12. B. Lukyanchuk et al., Nat. Mater. 9, 707 (2010).
13. R. D. Grober, R. J. Schoelkopf, D. E. Prober, Appl. Phys.
Lett. 70, 1354 (1997).
14. L. Novotny, N. van Hulst, Nat. Photonics 5, 83
(2011).
15. Q. Xu et al., Nano Lett. 7, 2800 (2007).
16. M. Sukharev, J. Sung, K. G. Spears, T. Seideman,
Phys. Rev. B 76, 184302 (2007).
17. P. Biagioni, J. S. Huang, L. Du, M. Finazzi, B. Hecht,
Phys. Rev. Lett. 102, 256801 (2009).
18. S. Liu et al., Opt. Lett. 34, 1255 (2009).
19. J. Ginn, D. Shelton, P. Krenz, B. Lail, G. Boreman,
Opt. Express 18, 4557 (2010).
20. Materials and methods are available as supporting
material on Science Online.
21. J. F. Nye, M. V. Berry, Proc. R. Soc. London A Math.
Phys. Sci. 336, 165 (1974).
22. M. Padgett, J. Courtial, L. Allen, Phys. Today 57, 35
(2004).
23. L. Allen, M. W. Beijersbergen, R. J. C. Spreeuw,
J. P. Woerdman, Phys. Rev. A 45, 8185 (1992).
24. M. W. Beijersbergen, R. P. C. Coerwinkel, M. Kristensen,
J. P. Woerdman, Opt. Commun. 112, 321 (1994).
25. N. R. Heckenberg, R. McDuff, C. P. Smith, A. G. White,
Opt. Lett. 17, 221 (1992).
26. H. He, M. E. J. Friese, N. R. Heckenberg, H. Rubinsztein-
Dunlop, Phys. Rev. Lett. 75, 826 (1995).
27. G. Gibson et al., Opt. Express 12, 5448 (2004).
28. D. M. Pozar, S. D. Targonski, H. D. Syrigos, IEEE Trans.
Antenn. Propag. 45, 287 (1997).
29. J. A. Encinar, IEEE Trans. Antenn. Propag. 49, 1403
(2001).
30. C. G. M. Ryan et al., IEEE Trans. Antenn. Propag. 58,
1486 (2010).
31. P. Padilla, A. Muoz-Acevedo, M. Sierra-Castaer,
M. Sierra-Prez, IEEE Trans. Antenn. Propag. 58, 2571
(2010).
Acknowledgments: The authors acknowledge helpful
discussion with J. Lin, R. Blanchard, and A. Belyanin.
The authors acknowledge support from the National
Science Foundation (NSF), Harvard Nanoscale Science
and Engineering Center (NSEC) under contract
NSF/PHY 06-46094, and the Center for Nanoscale
Systems (CNS) at Harvard University. This work was
supported in part by the Defense Advanced Research
Projects Agency (DARPA) N/MEMS S&T Fundamentals
program under grant N66001-10-1-4008 issued by the
Space and Naval Warfare Systems Center Pacific
(SPAWAR). Z.G. acknowledges funding from the
European Communities Seventh Framework Programme
(FP7/2007-2013) under grant agreement PIOF-GA-
2009-235860. M.A.K. is supported by NSF through a
Graduate Research Fellowship. Harvard CNS is a
member of the National Nanotechnology Infrastructure
Network. The Lumerical (Vancouver, BC, Canada)
FDTD simulations in this Research Article were run
on the Odyssey cluster supported by the Harvard
Faculty of Arts and Sciences Sciences Division Research
Computing Group.
Supporting Online Material
www.sciencemag.org/cgi/content/full/science.1210713/DC1
Materials and Methods
SOM Text
Figs. S1 to S6
References (3239)
Movie S1
5 July 2011; accepted 19 August 2011
Published online 1 September 2011;
10.1126/science.1210713
www.sciencemag.org SCIENCE VOL 334 21 OCTOBER 2011 337
RESEARCH ARTICLES
Detection of the Water Reservoir in a
Forming Planetary System
Michiel R. Hogerheijde,
1
* Edwin A. Bergin,
2
Christian Brinch,
1
L. Ilsedore Cleeves,
2
Jeffrey K. J. Fogel,
2
Geoffrey A. Blake,
3
Carsten Dominik,
4
Dariusz C. Lis,
5
Gary Melnick,
6
David Neufeld,
7
Olja Pani,
8
John C. Pearson,
9
Lars Kristensen,
1
Umut A. Yldz,
1
Ewine F. van Dishoeck
1,10
Icy bodies may have delivered the oceans to the early Earth, yet little is known about water in
the ice-dominated regions of extrasolar planet-forming disks. The Heterodyne Instrument for the
Far-Infrared on board the Herschel Space Observatory has detected emission lines from both spin
isomers of cold water vapor from the disk around the young star TW Hydrae. This water vapor
likely originates from ice-coated solids near the disk surface, hinting at a water ice reservoir
equivalent to several thousand Earth oceans in mass. The waters ortho-to-para ratio falls well
below that of solar system comets, suggesting that comets contain heterogeneous ice mixtures
collected across the entire solar nebula during the early stages of planetary birth.
W
ater in the solar nebula is thought to
have been frozen out onto dust grains
outside 3 astronomical units (AU)
(1, 2). Stored in icy bodies, this water provided a
reservoir for impact delivery of oceans to the
Earth (3). In planet-forming disks, water vapor is
thought to be abundant only in the hot (>250 K)
inner regions, where ice sublimates and gas-phase
chemistry locks up all oxygen in H
2
O. Emission
from hot (>250 K) water has been detected from
several disks around young stars (4, 5). In the
cold (20 K) outer disk, water vapor freezes out,
evidenced by spectral features of water ice in a
few disks (6, 7). However, (inter)stellar ultravi-
olet radiation penetrating the upper disk layers
desorbs a small fraction of water ice molecules
back into the gas phase (8), suggesting that cold
(<100 K) water vapor exists throughout the radial
extent of the disk. The detection of this water
vapor would signal the presence of a hidden ice
reservoir.
We report detection of ground-state rotation-
al emission lines of both spin isomers of water
(J
KAKC
1
10
-1
01
from ortho-H
2
O and 1
11
-0
00
from
para-H
2
O) from the disk around the premain-
sequence star TW Hydrae (TW Hya) using the
Heterodyne Instrument for the Far-Infrared (HIFI)
spectrometer (9) on board the Herschel Space
Observatory (10) (Fig. 1) (11, 12). TW Hya is
a 0.6 M

(solar mass), 10-million-year-old T

Tauri star (13) 53.7 T 6.2 pc away from Earth. Its
196-AU-radius disk is the closest protoplanetary
disk to Earth with strong gas emission lines. The
disks mass is estimated at 2 10
4
to 6 10
4
M

in dust and, using different tracers and assump-

tions, between 4 10
5
and 0.06M

in gas (1416).
The velocity widths of the H
2
O lines (0.96 to
1.17 kms
1
) (table S1) exceed by 40%those of
cold CO (14). These correspond to CO emission
fromthe full 196-AU-radius rotating disk inclined
at 7 with only little (<65 ms
1
) turbulence (17).
The wider H
2
O lines suggest that the water emis-
sion extends to 115 AU, where the gas orbits
the star at higher velocities compared with
196 AU.
To quantify the amount of water vapor traced
by the detected lines, we performed detailed sim-
ulations of the water chemistry and line for-
mation using a realistic disk model matching
previous observations (12, 18). We adopted a
conservatively low dust mass of 1.9 10
4
M

and, using a standard gas-to-dust mass ratio of

100, a gas mass of 1.9 10
2
M

. We explored
the effects of much lower gas-to-dust ratios. We
followed the penetration of the stellar ultraviolet
and x-ray radiation into the disk; calculated the
resulting photodesorption of water and ensuing
gas-phase chemistry, including photodissociation;
and solved the statistical-equilibrium excitation
and line formation. The balance of photodesorption
of water ice and photodissociation of water vapor
results in an equilibrium column of water H
2
O
vapor throughout the disk (Fig. 2). Consistent
with other studies (19), we find a layer of max-
imum water vapor abundance of 0.5 10
7
to
2 10
7
relative to H
2
at an intermediate height in
the disk. Above this layer, water is photodis-
sociated; below it, little photodesorption occurs
and water is frozen out, with an ice abundance, set
by available oxygen, of 10
4
relative to H
2
.
In our model, the 100- to 196-AU region
dominates the line emission, which exceeds ob-
servations in strength by factors of 5.3 T 0.2 for
H
2
O 1
10
-1
01
and 3.3 T 0.2 for H
2
O 1
11
-0
00
. A
lower gas mass does not decrease the line in-
tensities, if we assume that the water ice, from
1
Leiden Observatory, Leiden University, Post Office Box 9513,
2300 RA Leiden, Netherlands.
2
Department of Astronomy, Uni-
versity of Michigan, Ann Arbor, MI 48109, USA.
3
Division of
Geological and Planetary Sciences, California Institute of Tech-
nology, Pasadena, CA 91125, USA.
4
Astronomical Institute
Anton Pannekoek, University of Amsterdam, 1098 XH Am-
sterdam, Netherlands.
5
Division of Physics, Mathematics, and
Astronomy, California Institute of Technology, Pasadena, CA
91125, USA.
6
Harvard-Smithsonian Center for Astrophysics,
Cambridge, MA 02138, USA.
7
Department of Physics and As-
tronomy, Johns Hopkins University, Baltimore, MD 21218,
USA.
8
European Southern Observatory, 85748 Garching, Ger-
many.
9
Jet Propulsion Laboratory, California Institute of Tech-
nology, Pasadena, CA 91109, USA.
10
Max-Planck-Institut fr
Extraterrestrische Physik, 85748 Garching, Germany.
*To whom correspondence should be addressed. E-mail:
michiel@strw.leidenuniv.nl
Fig. 1. Spectra of para-H
2
O
1
11
-0
00
(A) and ortho-H
2
O 1
10
-
1
01
(B) obtained with HIFI on
the Herschel Space Observatory
toward the protoplanetary disk
around TWHya after subtraction
of the continuum emission. The
vertical dotted lines show the
systems velocity of +2.8 km s
1
relative to the Suns local en-
vironment (local standard of
rest).
REPORTS
21 OCTOBER 2011 VOL 334 SCIENCE www.sciencemag.org 338
which the water vapor derives, formed early in
the disks evolution, before substantial gas loss
occurred, and remains frozen on grains. The most
plausible explanation involves a difference in the
relative location of small, bare grains regulating
the ultraviolet radiative transport and larger, ice-
carrying grains. Differential settling of large grains
relative to small grains moves much of the ice
reservoir below the reach of the ultraviolet ra-
diation, resulting in less water vapor and weaker
lines. Our model matches the observations if only
12% of the original ice content remains above
this line (20). A radially increasing degree of
settling of icy grains explains the observed H
2
O
line widths.
The detected water vapor, resulting from
photodesorption, implies an ice reservoir in the
giant planet formation zone and beyond. In our
simulations, the 7.3 10
24
g of detected water
vapor (equivalent to 0.005 times the mass of
Earths oceans) originate from a total ice reser-
voir of 9 10
27
g (or several thousands of Earths
oceans) throughout the disk. The size of this res-
ervoir is tied to the dust mass contained in the
disk, for which we adopt a conservatively low
value. Although the ice reservoir is only observed
indirectly, no known mechanism can remove it
from the regions probed by Herschel. Any small-
er ice reservoir implies the corresponding ab-
sence of elemental oxygen that efficiently reforms
water ice on the grains.
The detection of both spin isomers of water
vapor allows its ortho-to-para ratio (OPR) to be
derived, because our simulations indicate that the
lines are optically thin. An OPR of 0.77 T 0.07
matches our observations (12). This value is
much lower than the OPR range of 2 to 3 ob-
served for solar system comets (21). It is com-
mon practice to associate the OPR with a spin
temperature T
spin
at which a Boltzmann distri-
bution reproduces the ratio of spin isomers. Our
derived OPR corresponds to T
spin
= 13.5 T 0.5 K,
whereas the range for solar system comets yields
a T
spin
of >20 K.
Radiative conversion between spin isomers is
not allowed in the gas phase, preserving the OPR
for long time scales. Gas-phase formation of wa-
ter occurs through exothermic reactions leading
to an OPR of 3. On grains, water forms and sur-
vives at low temperatures, and it is tempting to
equate T
spin
with the grain temperature. However,
the energetics of water formation and ortho-to-
para exchange on grains are poorly understood
(22), and the water OPR may be changed by
photodesorption. This process starts by dissociat-
ing water to H and OH in the ice and continues
with the energetic H kicking out a neighboring
H
2
O molecule from the ice matrix or with the H
and OHrecombining in the ice to formH
2
Owith
sufficient internal energy to sublimate (23). The
latter route drives the OPR to at least unity, im-
plying an even lower original ice OPR, to yield a
resulting OPR of 0.77. Cometary volatiles are
released through thermal sublimation, and their
measured OPRs are interpreted to reflect the OPR
of their ice constituents. Equating T
spin
with the
physical temperature of the grain on which the
ice formed is supported by the similarity of mea-
sured T
spin
of NH
3
and H
2
O in several individual
solar system comets (24).
Solar system comets consist of a heteroge-
neous mixture of ices and solids, likely assem-
bled in the giant planet formation zone by mixing
local material with material that drifted in from
larger radii (25). Our water vapor observations
probe cold, ice-coated precometary grains resid-
ing beyond >50 AU, representing the bulk of the
latter material. The presence in comets of crys-
talline silicates, requiring formation temperatures
>800 K(26), together with COand H
2
Oices that
condense at 20 to 100 K, argues for transport of
hot material from near the star to the icy outer
regions of the solar nebula (27). Provided that
spin temperatures reflect formation histories, the
different T
spin
inferred for the water ice in TW
Hya (<13 K) and solar system comets (>20 K)
indicates a similar mixing of volatiles throughout
the entire solar nebula, blending water formed at
>50 K and an OPR of 3 with water formed at 10
to 20 K and an OPR < 1 probed by our observa-
tions. In this case, the range of T
spin
values of the
cometary inventory reflects the stochastic nature
of transport and mixing.
Our Herschel detection of cold water vapor in
the outer disk of TW Hya demonstrates the pres-
ence of a considerable reservoir of water ice in
this protoplanetary disk, sufficient to formseveral
thousand Earth oceans worth of icy bodies. Our
observations only directly trace the tip of the ice-
berg of 0.005 Earth oceans in the form of water
vapor.
References and Notes
1. One astronomical unit (AU) is the mean distance between
Earth and the Sun of 1.49598 10
11
m.
2. C. Hayashi, Prog. Theor. Phys. 70 (suppl.), 35
(1981).
3. T. Matsui, Y. Abe, Nature 322, 526 (1986).
4. K. M. Pontoppidan, C. Salyk, G. A. Blake, H. U. Kufl,
Astrophys. J. 722, L173 (2010).
5. J. S. Carr, J. R. Najita, Astrophys. J. 733, 102
(2011).
6. H. Terada et al., Astrophys. J. 667, 303 (2007).
7. M. Honda et al., Astrophys. J. 690, L110
(2009).
8. C. Dominik, C. Ceccarelli, D. Hollenbach, M. Kaufman,
Astrophys. J. 635, L85 (2005).
9. T. de Graauw et al., Astron. Astrophys. 518, L6
(2010).
10. G. L. Pilbratt et al., Astron. Astrophys. 518, L1
(2010).
11. Water in Earths atmosphere obstructs ground-based
detection of cold water vapor in planet-forming disks.
Although Herschel cannot spatially resolve even the
closest disk in water ground-state emission lines, HIFI
spectrally resolves the H
2
O line profiles. Comparison
with previous spectrally and spatially resolved
observations of CO confirms the disk origin of the
H
2
O lines.
12. Materials and methods are available as supporting
material on Science Online.
13. R. A. Webb et al., Astrophys. J. 512, L63 (1999).
14. J. H. Kastner, B. Zuckerman, D. A. Weintraub, T. Forveille,
Science 277, 67 (1997).
15. N. Calvet et al., Astrophys. J. 568, 1008 (2002).
16. U. Gorti, D. Hollenbach, J. Najita, I. Pascucci, Astrophys. J.
735, 90 (2011).
17. A. M. Hughes, D. J. Wilner, S. M. Andrews, C. Qi,
M. R. Hogerheijde, Astrophys. J. 727, 85 (2011).
18. W. Thi et al., Astron. Astrophys. 518, L125 (2010).
19. P. Woitke, W. Thi, I. Kamp, M. R. Hogerheijde,
Astron. Astrophys. 501, L5 (2009).
20. R. Meijerink, K. M. Pontoppidan, G. A. Blake,
D. R. Poelman, C. P. Dullemond, Astrophys. J. 704, 1471
(2009).
A
C D
B
Fig. 2. Adopted model for the TWHya protoplanetary disk. (A) H
2
number density, (B) dust temperature,
(C) the number density of water vapor molecules and contours of volume-averaged water ice abundance
decreasing from white to black as 2 10
4
, 2 10
5
, 2 10
6
, 2 10
7
, and 2 10
8
, relative to H
2
,
and (D) one quadrant of the resulting water emission line intensity fromthe near face-on disk, in arbitrary
units. In (A) and (B), the blue contour delineates the layer of maximum water vapor abundance.
www.sciencemag.org SCIENCE VOL 334 21 OCTOBER 2011 339
REPORTS
21. B. P. Bonev et al., Astrophys. J. 661, L97
(2007).
22. H.-H. Limbach et al., ChemPhysChem 7, 551
(2006).
23. S. Andersson, E. F. van Dishoeck, Astron. Astrophys. 491,
907 (2008).
24. Y. Shinnaka et al., Astrophys. J. 729, 81 (2011).
25. S. J. Weidenschilling, Mon. Not. R. Astron. Soc. 180, 57
(1977).
26. S. A. Sandford et al., Science 314, 1720
(2006).
27. D. H. Wooden, Space Sci. Rev. 138, 75
(2008).
Acknowledgments: Herschel is a European Space Agency
space observatory with science instruments provided by
European-led principal investigator consortia and with
important participation from NASA. This work was
partially supported by Nederlandse Organisatie voor
Wetenschappelijk Onderzoek grant 639.042.404, NSF
grant 0707777, and, as part of the NASA Herschel HIFI
guaranteed time program, NASA. The data presented
here are archived at the Herschel Science Archive,
http://archives.esac.esa.int/hda/ui, under OBSID
1342198337 and 1342201585.
Supporting Online Material
www.sciencemag.org/cgi/content/full/334/6054/338/DC1
Materials and Methods
Table S1
References (2839)
25 May 2011; accepted 20 September 2011
10.1126/science.1208931
Supramolecular Linear Heterojunction
Composed of Graphite-Like
Semiconducting Nanotubular Segments
Wei Zhang,
1,2
Wusong Jin,
3
Takanori Fukushima,
1,4
* Akinori Saeki,
5
Shu Seki,
5
Takuzo Aida
1,2
*
One-dimensionally connected organic nanostructures with dissimilar semiconducting properties are
expected to provide a reliable platform in understanding the behaviors of photocarriers, which are
important for the development of efficient photon-to-electrical energy conversion systems. Although
bottom-up supramolecular approaches are considered promising for the realization of such nanoscale
heterojunctions, the dynamic nature of molecular assembly is problematic. We report a
semiconducting nanoscale organic heterojunction, demonstrated by stepwise nanotubular
coassembly of two strategically designed molecular graphenes. The dissimilar nanotubular segments,
thus connected noncovalently, were electronically communicable with one another over the
heterojunction interface and displayed characteristic excitation energy transfer and charge transport
properties not present in a mixture of the corresponding homotropically assembled nanotubes.
H
eterojunctions, occurring between two
dissimilar semiconducting materials, are
expected to provide peculiar electronic
properties that are hard to realize by homojunc-
tions. Heterojunctions of varying dimensions
are readily fabricated from inorganic semicon-
ductors and lead to many applications, including
solid-state lasers, diodes, solar cells, and tran-
sistors (16). Organic heterojunctions are of im-
portance in the development of organic thin-film
solar cells (7, 8). However, most that have been
studied are so-called bulk heterojunctions, which
are formed only coincidentally fromdonor/acceptor
mixtures upon phase separation (912). Although
bottom-up supramolecular approaches (13, 14)
are a potent tool for the formation of organic het-
erojunctions, such studies have just started with
molecularly engineered donor/acceptor couples
(1519). From a fundamental viewpoint, one chal-
lenge would be to tailor a linear organic hetero-
junction at the nanoscale by joining together
dissimilar semiconducting one-dimensional mo-
lecular objects, because one has to overcome the
essential problem arising from the dynamic
nature of molecular assembly (13, 14, 2024).
We reported that a Gemini-shaped hexa-peri-
hexabenzocoronene (HBC) derivative, bearing
two triethylene glycol-appended phenyl groups
on one side of the HBC core and dodecyl side
chains on the other, self-assembles into a semi-
conducting nanotube with inner and outer diam-
eters of 14 and 20 nm, respectively (25, 26). A
recent structural analysis using a synchrotron
x-ray diffraction technique revealed that the nano-
tube is composed of a graphite-like bilayer wall
consisting of helically twisted columnar arrays
of p-stacked HBC units (26). For the realiza-
tion of a nanotubular heterojunction using this
self-assembling motif, essential requisites are (i)
the formation of a morphologically stabilized
seed nanotube and (ii) the design of a second
graphene monomer capable of tubularly assem-
bling from the extremely thin facets of the seed
nanotube termini. Further issues include how to
cope with a high dispersibility of the seed nano-
tube and a solubility of the second monomer
under assembling conditions.
A keen examination, taking into account all
the above requisites, led us to HBC derivatives
1 and 2 (27) as the monomers for the seed and
second nanotubular segment, respectively (Fig.
1). HBC 1 carries two bipyridine (bpy) units, in
order for the resulting seed nanotube (Fig. 1A)
to be morphologically stabilized by wrapping
with a metal-coordination network (Fig. 1 B)
(28). The charged surface of the resultant seed
also merits its homogeneous dispersion by an
electrostatic repulsion (29). On the other hand,
HBC 2 bears four electron-withdrawing fluorine
substituents, so it can adhere electronically to
the seed termini and self-assemble selectively
from their nanotubular facets. When these HBC
molecules coassemble stepwise (Fig. 1C), the
resultant connecting segments are electronically
dissimilar to one another (Fig. 1D).
As a typical example of the preparation of the
seed nanotube (NT
1Cu
, Fig. 1B), a 5-ml glass
vial containing a tetrahydrofuran (THF) solution
(2.0 ml) of HBC 1 (0.5 mg, 1.5 10
4
M) was
placed in a 50-ml glass vial containing 10 ml of
methanol (MeOH) and allowed to stand at 25C,
whereupon a yellow suspension gradually formed.
Absorption spectroscopy of the suspension after
a 24-hour incubation (fig. S1, A and B, broken
curve) showed red-shifted absorption bands at
426 and 459 nm characteristic of J-aggregated
HBCs (25, 26). Scanning electron microscopy
(SEM, fig. S2A) and transmission electron mi-
croscopy (TEM, fig. S2B) of an air-dried sample
of the suspension allowed for visualizing nano-
tubes (NT
1
) with a uniform diameter of 20 nm,
although they were heavily bundled (Fig. 1A)
just like other HBC nanotubes (25, 26). We in-
vestigated the metal-coordination capability of
NT
1
by using Cu
2+
, because bpy is known to bind
to Cu
2+
, affording a bpy
2
Cu
2+
complex. As soon
as a MeOH solution (1.0 ml) of copper(II) tri-
fluoromethanesulfonate [Cu(OTf )
2
, 0.5 mg, 1.5
10
6
mol; 5.0 equivalents to HBC 1] was added,
the suspension containing bundled NT
1
became
clear, suggesting that Cu
2+
ions are bound to the
surface bpy groups (Fig. 1, A to B) and make
the nanotubes (NT
1Cu
) electrostatically repulsive
from one another (29). When an air-dried sam-
ple of NT
1Cu
, isolated by filtration and washed
with MeOH to remove free Cu(OTf )
2
, was sub-
jected to SEM, highly dispersed nanotubes were
observed (fig. S3A). Complete coordination of
Cu
2+
with bpy-appended HBC 1 was confirmed
by matrix-assisted laser desorption ionization time-
of-flight (MALDI-TOF) mass spectrometry of
isolated NT
1Cu
(fig. S4); no peaks attributable
to metal-free 1 were detected, but those assign-
able to 1Cu, dissociated from NT
1Cu
, were. The
metal coordination of NT
1
did not give rise to
any shift of the J-aggregate absorption bands (426
and 459 nm; fig. S1B, solid curve). Hence, the
p-stacking geometry of the HBC units is intact
1
Functional Soft Matter Research Group, RIKEN Advanced
Science Institute, 2-1 Hirosawa, Wako, Saitama 351-0198,
Japan.
2
Department of Chemistry and Biotechnology, School of
Engineering, The University of Tokyo, 7-3-1 Hongo, Bunkyo-ku,
Tokyo 113-8656, Japan.
3
College of Chemistry, Chemical Engi-
neering and Biotechnology, Donghua University, 2999 North
Renmin Road, Songjiang, Shanghai 201620, Peoples Republic
of China.
4
Chemical Resources Laboratory, Tokyo Institute of
Technology, 4259 Nagatsuta, Midori-ku, Yokohama 226-8503,
Japan.
5
Graduate School of Engineering, Osaka University, 2-1
Yamadaoka, Suita, Osaka 565-0871, Japan.
*To whom correspondence should be addressed. E-mail:
fukushima@riken.jp (T.F.); aida@macro.t.u-tokyo.ac.jp (T.A.)
21 OCTOBER 2011 VOL 334 SCIENCE www.sciencemag.org 340
REPORTS
even with the formation of a metal-coordination
network on the nanotube surface.
We investigated whether HBC 2 alone has the
ability to form a nanotubular assembly. On the
basis of a computational study using density func-
tional theory (DFT), the core of HBC 2 with four
fluorine substituents most likely adopts a slightly
twisted, concave structure (fig. S5). Probably be-
cause of this skeletal distortion, HBC 2 was un-
able to assemble into nanotubes under a variety of
conditions used previously (16, 18, 25, 26, 28, 29).
However, we eventually found that, in acetone,
2 self-assembles into tubules. This observation
was unexpected because none of other reported
HBC derivatives assemble to form nanotubes in
acetone. For the nanotubular assembly of 2, an
acetone suspension of this HBC (1.5 10
4
M)
was heated to 50C, and the resultant clear solu-
tion was allowed to cool to 25C. The solu-
tion gradually became turbid and displayed a
red-shifted absorption spectrum characteristic of
J-aggregated HBCs (fig. S1, C and D). SEM of
an air-dried sample of the suspension, obtained
after a 12-hour incubation, showed the presence
of heavily bundled cylindrical nanostructures (fig.
S6A). TEM revealed that the cylinders are actu-
ally nanotubes (NT
2
) with a uniform diameter
of 20 nm (fig. S6B).
NT
1Cu
dispersed individually in acetone,
where the nanotubular assembly of HBC 2 can
occur. As a typical example of the successful het-
erojunction (Fig. 1C), the acetone dispersion
of NT
1Cu
, used as the seed (Fig. 1B), was pre-
sonicated for 5 to 10 min so that the nanotubes
were cut into short pieces for enhancing the prob-
ability of linear heterojunction (fig. S3C). The
resultant dispersion of NT
1Cu
(1.5 10
4
M,
1.0 ml) was mixed at 50C with an acetone solu-
tion of HBC 2 (1.5 10
4
M, 1.0 ml). When
the mixture was allowed to cool and stand at
25C, the assembly of HBC 2 took place. The
self-assembly of 2 without NT
1Cu
gave rise to a
suspension of bundled NT
2
. However, the co-
assembling mixture in the presence of NT
1Cu
remained clear even after 12 hours. When an air-
dried sample of this clear dispersion was studied
by SEM, micrometer-long nanotubular objects
with bright and dark segments were observed
(Fig. 2A and fig. S7). Most of the nanotubes ap-
peared to consist of two block segments, but
some were composed of three block segments.
Likewise, scanning TEM allowed for visualiz-
ing the presence of segments with different con-
trasts in single nanotubes (Fig. 2C). By means of
element mapping using TEM energy-dispersive
x-ray spectroscopy (TEM-EDX), we confirmed
that copper is localized in the bright segments
(Fig. 2, E and F), whereas carbon populates over
the entire nanotube (Fig. 2, D and F). These ob-
servations demonstrate the occurrence of a linear
heterojunction to give block-NT
1Cu
/NT
2
(Fig.
1C). Considering the electronic characters of the
two HBC molecules, the heterotropic p-stacking
interaction seems stronger than the homotropic
one. This drives the preferential assembly of
HBC 2 on the nanotubular facets of seed NT
1
.
Cu
.
Because multiblock heterojunction nanotubes such
as NT
1Cu
NT
2
NT
1Cu
NT
2
and NT
1Cu
NT
2

NT
1Cu
were not detected, postconnection of NT
2
with NT
1Cu
is unlikely. In tapping-mode atomic
force microscopy (AFM) on a silicon wafer, these
two block segments exhibited different height
profiles, 16 to 18 nm and 8 to 10 nm (Fig. 2B),
Fig. 2. Microscopic imaging of block-NT
1Cu
/NT
2
([2]/[1] = 1.0). The sample was prepared by drop-casting
its acetone dispersion and air dried. (A) SEM (scale bar, 500 nm), (B) tapping-mode AFM (scale bar,
200 nm), and (C) scanning TEM (scale bar, 50 nm) micrographs. TEM-EDX mapping of (D) carbon (scale
bar, 50 nm), (E) copper (scale bar, 50 nm), and (F) carbon and copper elements (scale bar, 50 nm).
Fig. 1. Molecular structures of HBCs 1 and 2 and schematic illustrations of the preparation of (A) NT
1
(bundled) by MeOH vapor diffusion into a THF solution of HBC 1, (B) seed NT
1Cu
(dispersed) by post-
functionalization of NT
1
with Cu(OTf)
2
in MeOH, and (C) block-NT
1Cu
/NT
2
(dispersed) by cooling a hot
acetone solution of HBC 2 in the presence of NT
1Cu
as the seed. (D) Schematic illustration of an
idealized cross section of block-NT
1Cu
/NT
2
at the heterojunction interface.
www.sciencemag.org SCIENCE VOL 334 21 OCTOBER 2011 341
REPORTS
that were the same as those observed for sepa-
rately prepared homotropic NT
1Cu
and NT
2
(fig. S8). The exceptionally small height profile
of NT
2
indicates a low structural robustness of
its graphite-like wall because of the nonplanar
HBC core of 2 (fig. S5).
As shown in fig. S9, NT
1
(black solid curve)
and NT
2
(black broken curve), when photoex-
cited at 365 nm, fluoresced most intensely at 585
and 594 nm, respectively (16, 18, 29). In sharp
contrast, NT
1Cu
luminesced much less intensely
(fig. S9, red) as a consequence of possible flu-
orescence quenching by Cu
2+
in the coordination
network around the nanotube. These contrasting
luminescence features allowed us to investigate
whether the two graphite-like nanotubular seg-
ments in block-NT
1Cu
/NT
2
(Fig. 1C) communi-
cate with one another by excitation energy transfer
over the heterojunction interface. When an ace-
tone dispersion of block-NT
1Cu
/NT
2
prepared
by using the mole ratio [2]/[1] of 0.5 for the step-
wise coassembly was photoexcited at 365 nm,
the observed fluorescence (Fig. 3B, pink) was
as weak as that of homotropically assembled
NT
1Cu
(Fig. 3B, red). Even when the mole ra-
tio [2]/[1] used for the stepwise coassembly was
increased from 0.5 to 1.0 (orange), 1.5 (green),
and 2.0 (blue) (Fig. 3A), the fluorescence in-
tensity of block-NT
1Cu
/NT
2
remained almost
unchanged at a very low level (Fig. 3B). In con-
trast to block-NT
1Cu
/NT
2
(Fig. 3B inset), a mix-
ture of homotropic NT
1Cu
and NT
2
([2]/[1] =
1.0) (Fig. 3C, orange) displayed a rather bright
fluorescence (Fig. 3D inset), where the observed
intensity was the sum of those of NT
1Cu
and
NT
2
(Fig. 3D, orange). When the mixing ratio of
NT
2
to NT
1Cu
was varied ([2]/[1] = 0.5 to 2.0)
(Fig. 3C), the overall fluorescence intensity changed
depending on the amount of NT
2
(Fig. 3D). To-
gether with the results of these control experi-
ments, the fluorescing properties observed for
block-NT
1Cu
/NT
2
indicate that the NT
1Cu
and
NT
2
segments communicate efficiently over the
heterojunction interface (Fig. 1D) by excitation en-
ergy transfer. The results also suggest that the
amount of unconnected NT
2
, if any formed in the
stepwise coassembly (Fig. 1C), is negligibly small.
Because the excitation energy transfers over
the heterojunction interface from one nanotu-
bular segment (NT
2
) to the other (NT
1Cu
) in
block-NT
1Cu
/NT
2
(Fig. 1C), we were motivated
to explore the behaviors of charge carriers, if
generated, in block-NT
1Cu
/NT
2
. For this pur-
pose, we used a flash-photolysis time-resolved
microwave conductivity (FP-TRMC) method (30),
which allows for evaluating the intrinsic prop-
erties of charge carriers without electrodes. At
first, we investigated the FP-TRMC profiles
of homotropically assembled NT
1Cu
and NT
2
.
Upon laser excitation at 355 nm (photon den-
sity, 4.7 10
15
cm
2
), both nanotube samples in
the solid state displayed TRMC signals, indicat-
ing that these nanotubes are photoconductive.
The TRMC signals thus observed for NT
1Cu
and
NT
2
decayed at comparable rates (fig. S10, A
and B) with lifetimes (t
1/e
) of 1.4 10
6
and 2.5
10
6
s, respectively (Fig. 4). Of particular inter-
est to note here, the charge carriers generated in
block-NT
1Cu
/NT
2
([2]/[1] = 1.0) were long-
lived (fig. S10D): The observed lifetime (t
1/e
=
8.8 10
6
s) was roughly five times longer than
those of NT
1Cu
and NT
2
(Fig. 4). Taking into
account the energy levels of 1 (16) and 2 (fig.
S11) estimated from their electrochemical and
spectral data, it is most likely that the NT
1Cu
and NT
2
segments preferentially accommodate
hole and electron, respectively. As a consequence
of such preferential localization of hole and elec-
tron in block-NT
1Cu
/NT
2
, the probability of
charge recombination could be reduced. In sharp
contrast, a mixture of homotropic NT
1Cu
and NT
2
([2]/[1] = 1.0) (fig. S10C) hardly showed suppres-
sion of charge recombination (t
1/e
= 2.0 10
6
s)
(Fig. 4).
The behaviors of excitons and charge car-
riers thus observed for semiconducting block-
NT
1Cu
/NT
2
are remarkable considering that
they are brought about only by noncovalent con-
nection of two different homotropic blocks with
Fig. 3. (A) Absorption and (B) fluorescence spectra in acetone at 25C of block-NT
1Cu
/NT
2
prepared
by using different molar ratios of 2 to 1 for their stepwise coassembly {[1] = 1.5 10
4
M; [2] = 0
(red), 0.75 10
4
M (pink), 1.5 10
4
M (orange), 2.3 10
4
M (green), and 3.0 10
4
M (blue)}. (B
inset) A photograph of an acetone dispersion of block-NT
1Cu
/NT
2
([2]/[1] = 1.0) upon 365-nm
excitation. (C) Absorption and (D) fluorescence spectra in acetone at 25C of mixtures of NT
1Cu
and
NT
2
at different molar ratios of 2 to 1 {[1] = 1.5 10
4
M; [2] = 0 (red), 0.75 10
4
M (pink), 1.5
10
4
M (orange), 2.3 10
4
M (green), and 3.0 10
4
M (blue)}. (D inset) A photograph of an acetone
suspension of a mixture of NT
1Cu
and NT
2
([2]/[1] = 1.0) upon 365-nm excitation. The baseline uprise
in (C) upon increment of [2] is due to light scattering by bundled NT
2
.
Fig. 4. Lifetimes (t
1/e
) of charge carriers generated
by laser excitation of NT
1Cu
, NT
2
, a mixture of
NT
1Cu
and NT
2
([2]/[1] = 1.0), and block-NT
1Cu
/NT
2
([2]/[1] = 1.0) in the solid state. t
1/e
is defined by the
time when the FP-TRMC transient decays down to 1/e
of its maximum value (fig. S10).
21 OCTOBER 2011 VOL 334 SCIENCE www.sciencemag.org 342
REPORTS
an extremely thin (~3 nm) facet. In particular,
the nearly perfect fluorescence quenching of its
NT
2
block substantiates that electronic effects
of heterojunctions can indeed propagate over a
micrometer-long distance through a great number
of p-stacks in semiconducting organic materials.
References and Notes
1. M. T. Bjrk et al., Nano Lett. 2, 87 (2002).
2. S. Banerjee, S. S. Wong, Nano Lett. 2, 195 (2002).
3. O. Harnack, C. Pacholski, H. Weller, A. Yasuda,
J. M. Wessels, Nano Lett. 3, 1097 (2003).
4. B. Tian et al., Nature 449, 885 (2007).
5. A. I. Hochbaum, P. Yang, Chem. Rev. 110, 527 (2010).
6. F.-S. Tsai et al., Appl. Phys. Express 4, 025002 (2011).
7. S. Gnes, H. Neugebauer, N. S. Sariciftci, Chem. Rev.
107, 1324 (2007).
8. B. C. Thompson, J. M. J. Frchet, Angew. Chem. Int. Ed.
47, 58 (2008).
9. G. Yu, J. Gao, J. C. Hummelen, F. Wudl, A. J. Heeger,
Science 270, 1789 (1995).
10. P. Peumans, S. Uchida, S. R. Forrest, Nature 425, 158 (2003).
11. G. Li et al., Nat. Mater. 4, 864 (2005).
12. J. Y. Kim et al., Science 317, 222 (2007).
13. T. F. A. De Greef et al., Chem. Rev. 109, 5687 (2009).
14. F. J. M. Hoeben, P. Jonkheijm, E. W. Meijer, A. P. H. J.
Schenning, Chem. Rev. 105, 1491 (2005).
15. F. Wrthner et al., J. Am. Chem. Soc. 126, 10611
(2004).
16. Y. Yamamoto et al., Science 314, 1761 (2006).
17. A. L. Sisson et al., Angew. Chem. Int. Ed. 47, 3727 (2008).
18. Y. Yamamoto et al., Proc. Natl. Acad. Sci. U.S.A. 106,
21051 (2009).
19. N. Sakai, R. Bhosale, D. Emery, J. Mareda, S. Matile,
J. Am. Chem. Soc. 132, 6923 (2010).
20. By using crystallizable core units, the formation of rodlike
block co-micelles has been demonstrated (2124).
21. X. Wang et al., Science 317, 644 (2007).
22. T. Gdt, N. S. Ieong, G. Cambridge, M. A. Winnik,
I. Manners, Nat. Mater. 8, 144 (2009).
23. J. B. Gilroy et al., Nat. Chem. 2, 566 (2010).
24. S. K. Patra et al., J. Am. Chem. Soc. 133, 8842
(2011).
25. J. P. Hill et al., Science 304, 1481 (2004).
26. W. Jin et al., J. Am. Chem. Soc. 130, 9434 (2008).
27. Materials and methods are available as supporting
material on Science Online.
28. W. Zhang, W. Jin, T. Fukushima, N. Ishii, T. Aida, Angew.
Chem. Int. Ed. 48, 4747 (2009).
29. G. Zhang et al., J. Am. Chem. Soc. 129, 719
(2007).
30. A. Saeki, T. Fukumatsu, S. Seki, Macromolecules 44,
3416 (2011).
Acknowledgments: We thank E. Ohta (RIKEN) for DFT
calculation of HBC 2. W.Z. thanks the Japan Society for
the Promotion of Science Young Scientist Fellowship
(219925). T.F. thanks Ministry of Education, Culture,
Sports, Science, and Technology, Japan, for financial
support (21350108).
Supporting Online Material
www.sciencemag.org/cgi/content/full/334/6054/340/DC1
Materials and Methods
Figs. S1 to S11
References (3133)
27 June 2011; accepted 14 September 2011
10.1126/science.1210369
Dynamics of the Reaction of Methane
with Chlorine Atom on an Accurate
Potential Energy Surface
Gbor Czak* and Joel M. Bowman
The reaction of the chlorine atom with methane has been the focus of numerous studies that
aim to test, extend, and/or modify our understanding of mode-selective reactivity in polyatomic systems.
To this point, theory has largely been unable to provide accurate results in comparison with experiments.
Here, we report an accurate global potential energy surface for this reaction. Quasi-classical trajectory
calculations using this surface achieve excellent agreement with experiment on the rotational
distributions of the hydrogen chloride (HCl) product. For the Cl + CHD
3
HCl + CD
3
reaction at low
collision energies, we confirm the unexpected experimental finding that CH-stretch excitation is no more
effective in activating this late-barrier reaction than is the translational energy, which is in contradiction
to expectations based on results for many atom-diatom reactions.
D
ecades of experimental and theoretical
studies of atom-diatom reactions led to a
well-validated framework for predicting
the effect of vibrational excitation on the ensuing
dynamics (1, 2). Earlier fundamental research dem-
onstrated the importance of the reaction barrier
location on the efficacy of partitioning the total
energy between internal excitation of the diatomic
molecule and relative translational energy of the
reactants. The careful and correct analysis of these
reactions led to the Polanyi rules (3), which
state that vibrational energy is more efficient than
is translational energy in activating a late-barrier
reaction, whereas the reverse is true for an early-
barrier reaction. Recent studies have investigated
the generality and validity of these paradigms for
polyatomic systems. The X + methane (CH
4
and deuterated isotopologues) reactions (which
replace the diatomic with a five-atom molecule)
have played a central role in this research, in
which, for example, the choice of X as H, O, F,
and Cl has permitted the height and location of
the reaction barrier to vary widely. Recent ex-
periments by Liu and co-workers (48) on the F
and Cl + methane reactions have uncovered sur-
prising departures from expectations that present a
strong challenge to theory, which ultimately pro-
vides detailed understanding of chemical reaction
dynamics. A rigorous theoretical approach to re-
action analysis consists of two components. The
first is to determine the global potential energy
surface (PES) (9), which governs the nuclear mo-
tion, and the second is to perform dynamics
calculations with the PES. We succeeded in carry-
ing out this process recently for the early-low-
barrier F + CHD
3
reaction, for which we were able
to illuminate the surprising experimental result of
the enhancement of the DF + CHD
2
channel by
exciting the CH-stretch (10, 11). In this Report,
we take the same approach to address and inter-
pret experiments on the late-high-barrier Cl +
CHD
3
reaction by Liu and co-workers (5), which
also uncovered a surprising result, namely that at
low collision energies (E
coll
) vibrational excita-
tion of the CH-stretch was no more effective than
was translational energy in promoting the reaction.
This result, as pointed out by these authors, contra-
dicts the rule of thumb of reaction dynamics
(Polanyi rules). Other interesting experimental re-
sults of this reaction are also successfully addressed.
As with the previous accurate PES for F +
CH
4
(10), the Cl + CH
4
PES is a permutationally
invariant fit (12, 9) to roughly 16,000 high-level
ab initio electronic energies. The selection of con-
figurations for the PES is quite similar to proce-
dures used for the F + CH
4
PES (10, 12), and
details are given in (13). A key part of this ap-
proach is the use of an electronic structure method
that gives accurate energies, especially for the
barrier height, reaction enthalpy, and the entrance
and exit channel van der Waals (vdW) wells. These
wells, which result from long-range attractive
interactions, are ubiquitous in chemistry, and as
we show here, the prereactive one has a substan-
tial effect on the Cl + CHD
3
reaction dynamics at
low collision energies. The inclusion of the spin-
orbit (SO) correction is also essential in the present
case because it effectively increases the barrier
height and reaction endoergicity by 0.8 kcal/mol
and has a substantial effect on the entrance chan-
nel vdW well. (The SO correction is a relativistic
effect, which lowers the energy of the halogen
atoms and has about twice as large an energy shift
on the heavier Cl than on F. The widely applied
nonrelativistic electronic structure computations
neglect this effect.) There is also a substantial
basis setsuperposition error in this region, which
has to be corrected. Thus, the goal for the present
PES is to take all of this into account. In order to
achieve this goal, we used a composite electronic
structure method, which provides accurate en-
ergies with affordable computational time. The
general composite approach (14), which is wide-
ly used, combines results from several levels of
ab initio method and basis (15, 16). Second, in
order to account for the SO effect differences be-
tween the SOand non-SOground-state electronic
energies, obtained by means of multireference
configuration interaction (MRCI) with Davidson
correction (MRCI + Q) with basis set aug-cc-
pVTZ (MRCI + Q/aug-cc-pVTZ), were added to
Cherry L. Emerson Center for Scientific Computation and De-
partment of Chemistry, Emory University, Atlanta, GA 30322,
USA.
*To whom correspondence should be addressed. E-mail:
czako@chem.elte.hu
www.sciencemag.org SCIENCE VOL 334 21 OCTOBER 2011 343
REPORTS
the composite non-SO energies at 1598 ClCH
4
configurations in the entrance channel, where the
SO corrections are substantial and nonconstant.
(2000 Cl + CH
4
fragment data were also shifted
by the constant SO correction of the Cl atom.)
Third, counterpoise correction for entrance chan-
nel basis setsuperposition error at the above-
mentioned 1598 configurations was applied. The
present PES is based on electronic energies with
basis set effects up to aug-cc-pCVTZ(correlation-
consistent polarized core-valence triple-zeta basis
augmented with diffuse functions) (17); electron
correlation up to the gold standard CCSD(T)
(coupled-cluster with single and double and per-
turbative triple excitations) method (18); correla-
tion effects of the core-core and core-valence
electrons; and SO and counterpoise corrections
for the entrance channel. As a result, this present
PES corresponds to the SO ground state, ac-
curately describes the vdW regions both in the
entrance and exit channels, and describes both
the abstraction (HCl + CH
3
) and substitution
(H + CH
3
Cl) channels. Thus, it represents a con-
siderable advance over previous PESs for Cl +
CH
4
, including several semiempirical (19, 20)
and ab initiobased surfaces in 2 (21), 3 (22),
and 12(full) (23) dimensions.
Aschematic of the PES, given in Fig. 1, shows
the stationary-point structures and energies of
the abstraction and substitution reactions Cl(
2
P
3/2
,
2
P) + CH
4
HCl + CH
3
and H + CH
3
Cl, where
Cl(
2
P
3/2
) is the SO ground state of the Cl atom
lying below the non-SO energy of Cl(
2
P) by
0.8 kcal/mol. The vdWwells in the entrance chan-
nel for the CHCl and HCCl linear bond ar-
rangements are belowCl(
2
P
3/2
) + CH
4
(eq) by 0.3
and 0.6 kcal/mol, respectively. (That the most
favorable vdWorientation is the HCCl one is both
meaningful and easily understood from a simple
sum-of-pairs interaction because C is more polar-
izable than H.) The abstraction reaction has a late
(product-like) saddle-point of energy 7.6 kcal/mol
relative to Cl(
2
P
3/2
) + CH
4
(eq) and a vdW well
in the product channel with a dissociation energy
(D
e
) of 2.4 kcal/mol. The PES reaction endoer-
gicity for the HCl + CH
3
channel is 5.7 kcal/mol.
These values are in excellent agreement with our
benchmark energies of 7.6, 2.4, and 6.0 kcal/mol,
respectively, which were obtained by using the
focal-point analysis (FPA) approach (14) [see com-
putational details in (24)]. The PES reaction en-
doergicity is also in excellent agreement with the
experimental value, deduced tobe 6.0 kcal/mol (13).
None of the previous PESs are as accurate in
all of these key energies. In particular, focusing
on the important entrance and exit channel vdW
wells, the present PES contains an accurate de-
scription of the latter, (CH
3
HCl), with a sub-
stantial D
e
value of 2.4 kcal/mol. This contrasts
with the conclusion based on a semiempirical PES
with a D
e
of 0.3 kcal/mol that the existence of
this product complex is questionable (20). With
respect to the entrance channel vdW well, there
are even greater contrasts with previous PESs.
This vdW region in particular is affected by SO
2
P 0.8
2
P
3/2
0
H + CH
3
Cl
HCl + CH
3
6.0 (5.7)
3.6 (3.3)
Cl + CH
4
7.6 (7.6)
0.3
25.2 (26.0)
42.1 (43.0)
0.6
(CH
3
---HCl)
vdW
(H--CH
3
--Cl)
SP
(CH
3
--H--Cl)
SP
Reaction coordinate
R
e
l
a
t
i
v
e

e
n
e
r
g
y

(
k
c
a
l
/
m
o
l
)
Fig. 1. Schematic of the global SO and non-SO ground-state PESs of the Cl + CH
4
reaction showing the
accurate benchmark energies and the PES values in parentheses; for example, 7.6 (7.6) shows the
excellent agreement between the benchmark and PES barrier heights. All the energies are relative to
the SO ground state Cl(
2
P
3/2
) + CH
4
(eq). The negative energies correspond to the attractive region of
the vdW well in the entrance channel on the SO surface (Fig. 2).
3.0 3.5 4.0 4.5 5.0 5.5 6.0 6.5 7.0 7.5 8.0
-0.5
0.0
0.5
1.0
1.5
2.0
2.5
E
SO
3
SO
2
SO
1
A
1
MRCI+Q
H
3
CH---Cl (C
3v
)
E
n
e
r
g
y

(
k
c
a
l
/
m
o
l
)
2.5 3.0 3.5 4.0 4.5 5.0 5.5 6.0
-0.5
0.0
0.5
1.0
1.5
2.0
2.5
E
A
1
SO
3
SO
2
SO
1
MRCI+Q
HCH
3
---Cl (C
3v
)
E
n
e
r
g
y

(
k
c
a
l
/
m
o
l
)
R(C---Cl) /
3.0 3.5 4.0 4.5 5.0 5.5 6.0 6.5 7.0 7.5 8.0
A
1
SO
1
H
3
CH---Cl (C
3v
)
PES
2.5 3.0 3.5 4.0 4.5 5.0 5.5 6.0
SO
1
A
1
HCH
3
---Cl (C
3v
)
PES
R(C---Cl) /
Fig. 2. Potential energy curves of CH
4
Cl as a function of the CCl distance along the C
3v
axis with
fixed CH
4
(eq) geometry and (top) CHCl or (bottom) HCCl linear bond arrangements that were (left)
computed at the MRCI+Q/aug-cc-pVTZ level or (right) obtained from the non-SO and SO-corrected
ground-state PESs. A
1
and E denote the ground and excited non-SO electronic states, respectively, and
SO
1
, SO
2
, SO
3
are the three SO states. The energies are relative to Cl(
2
P
3/2
) + CH
4
(eq).
21 OCTOBER 2011 VOL 334 SCIENCE www.sciencemag.org 344
REPORTS
interaction, as shown in Fig. 2, where potential
curves both with and without SO coupling are
shown for the entrance channel. Our multi-
reference configuration interaction results show
that the non-SO electronic ground state potential
has minima with CHCl and HCCl bond ar-
rangements with depths of 0.3 kcal/mol and 0.9
kcal/mol, respectively (25). The SO interaction
has a minor effect on the former, whereas it de-
creases the depth of the latter by 0.3 kcal/mol;
but, the HCCl orientation still remains the deeper
minimum. It is critical that the present PES
describes these vdWregions accurately, as shown
in Fig. 2, because they play an important role in
the low E
coll
dynamics. [Additional comparisons
between the PES and benchmark properties, in-
cluding data for the substitution channel, are
presented in (13).]
Having shown the high accuracy of the PES
by comparison with benchmark ab initio data, we
applied it to simulations of the collision dynamics
of the title reaction. We performed more than 2
million quasi-classical trajectories (QCTs) for the
reactions of Cl(
2
P
3/2
) with CH
4
(v = 0), CHD
3
(v =
0), and CHD
3
(v
k
= 1) [k = 1, 3, 6, 5], where v = 0,
v
1
= 1, and v
k
= 1 [k = 3, 6, 5] denote the vi-
brational ground state, CH-stretch, and three dif-
ferent bend excitations, respectively. Details of
the QCTcalculations are given in (13). Classical
zero-point leak and vibrational energy relaxation
fromv
k
= 1 was investigated thoroughly for CHD
3
and found not to be a serious issue, as already
reported in (8, 11) and discussed further in (13).
All the results presented below correspond to
the SO-corrected PES. Computations on the non-
SO PES show that the inclusion of the SO cor-
rection in the PES decreases the cross sections by
Fig. 3. Computed and experimental HCl
rotational distributions for the Cl + CH
4
re-
action at a collision energy of 3.7 kcal/mol.
Theory uses quasi-classical trajectory calcu-
lations on the SO-corrected PES considering
trajectories in which CH
3
has at least zero-
point vibrational energy. Experimental data
are taken from(28). On the basis of analysis
of two batches of trajectories, the estimated
statistical uncertainty of the computed
results is less than 15%.
Fig. 4. (A) Computed cross sections for the ground
state (v = 0) and reactant CH-stretch (v
1
= 1) and
bend (v
k
= 1) [k = 3, 6, 5] excited Cl + CHD
3
(v)
HCl + CD
3
reactions and (B) their ratios as a func-
tion of E
coll
. (C) Cross section ratios at equivalent
amount of total energy (E
tot
= E
coll
+ E
v
), where the
vibrational energies (E
v
) are 0, 8.6, and 3.0 kcal/mol
(relative to zero-point energy) for the ground state
and stretch- and bend-excited reactions, respec-
tively. (Right) Snapshots of a nonreactive Cl +
CHD
3
(v
1
= 1) trajectory illustrating the stereo-
dynamics in the vdW region causing the unexpected
s
v
/s
g
< 1 ratio at E
tot
= 9.6 kcal/mol, as seen in (C).
Cl + CHD
3
(v
1
=1)
E
coll
= 1.0 kcal/mol
0 2 4 6 8 10 12 14 16 18 20
0
10
20
30
40
50
60
0 2 4 6 8 10 12 14 16 18 20
0
1
2
3
4
v
6
=1
v
3
=1
v
5
=1
v
6
=1
v
3
=1
v
5
=1
v
1
=1
E
coll
(kcal/mol)

v

/

g
0 2 4 6 8 10 12 14 16 18 20
0
1
2
3
4
5
v
3
=1
v
5
=1
v
6
=1
v
1
=1
v=0
Cl + CHD
3
(v) > > > > HCl + CD
3
E
coll
(kcal/mol)

(
b
o
h
r
2
)
6 8 10 12 14 16 18 20
0.0
0.5
1.0
1.5
2.0
2.5
3.0
3.5
v
3
=1
v
6
=1
v
5
=1
v
1
=1

v

/

g
E
tot
(kcal/mol)
A
B
C
www.sciencemag.org SCIENCE VOL 334 21 OCTOBER 2011 345
REPORTS
a factor of 1.5 to 2.5 at low E
coll
because of the
increase in the barrier height, but it has no sub-
stantial effect on the final state distributions.
First, consider the HCl(v = 0, J ) rotational
distributions for the Cl + CH
4
(v = 0) reaction.
These were reported by three experimental groups
showing extremely cold rotational populations
(2628). Theoretical simulations have been strug-
gling to reproduce this rotational distribution for
many years; the previous work overestimates the
rotational temperature of the HCl product en-
semble (19, 23, 29). In Fig. 3, we present QCT
results obtained by using the present PES and, as
seen, the agreement between theory and experi-
ment (27, 28) is excellent. We also computed the
HCl rotational distributions for Cl + CHD
3
(v
1
= 1)
showing cold rotational distribution of the
stretch-excited product HCl(v = 1, J ), similar
to HCl(v = 0, J ) from Cl + CH
4
(v = 0), and sub-
stantially hotter rotational temperature of the
vibrationally ground state HCl in qualitative
agreement with an older experiment (which had
large uncertainties) (30).
Next, we considered the effect of vibrational
excitation in CHD
3
, mentioned already. The Polanyi
rules (3) state that for late-barrier (atom+ diatom)
reactions, the reactant vibrational energy is more
efficient than is the translational energy in sur-
mounting the barrier. However, as noted already
a recent crossed molecular beam experiment (5)
found that this picture could not be simply ex-
tended for the Cl + CHD
3
reaction. To investigate
this finding, we calculated cross sections for the
reactant ground state and bend and CH-stretch
excited Cl + CHD
3
(v
k
) HCl + CD
3
reactions
and show the results in Fig. 4 as a function of
both E
coll
and total energy (E
tot
). At the same
value of E
coll
, we see that all the bending and
especially the CH-stretch excitations enhance the
reaction relative to Cl + CHD
3
(v = 0) (Fig. 4, A
and B). Thus, in this sense there is enhancement
of the reaction by excitation of these modes.
However, as noted in the experimental study, at
the same E
tot
translational energy is more effective
than the excitation of the reactive CH-stretch or
bend at low E
coll
. As the E
coll
increases, the in-
tuitively expected enhancement of reactivity is
seen upon vibrational excitation. This is best seen
by plotting cross section ratios (s
v
/s
g
) as a func-
tion of E
tot
(Fig. 4C). As seen, s
v
/s
g
is less than
1 if the E
tot
is below 11 and (15, 15, 9) kcal/mol
for the CH-stretch and (v
3
, v
6
, v
5
) bending modes,
respectively. Only CD
3
(v =0) products were probed
experimentally, whereas theory shows the total re-
activity. Correlating the QCT cross sections to
CD
3
(v = 0) results in a decrease of the s
CH
/s
ground
ratio, improving the agreement with the measured
data (fig. S1). Furthermore, the experiment applied
thermal bending excitation; thus, the measured
bending cross sections show the average effect
of the three bending modes (5). Theory predicts
that the v
5
(e) (CD
3
deformation) bending mode
is the most efficient to drive the reaction (Fig.
4B), similar to F + CHD
3
. Overall, both theory and
experiment show that the same amount of total
energy distributed among different nuclear motions
has different effects on chemical reactivity.
In order to gain insight into these results, we
examined the trajectories for the ground and stretch-
excited CHD
3
at the same E
tot
of 9.6 kcal/mol, thus
corresponding to a lowE
coll
for CHD
3
(v
1
= 1). We
determined the distributions of the smallest H-Cl
and C-Cl distances for nonreactive trajectories
and, as shown in fig. S2, most of the trajectories
for the CH-stretchexcited reaction do not reach
the transition state. Instead, the Cl atomturns back
at the vdWregion in the r
C-Cl
= 2.9 to 3.5 range.
This occurs because at low E
coll
, the CHD
3
rotates
to the energetically favorable, but nonreactive,
H-CCl orientation (Fig. 4, right). Thus, at lowE
coll
the entrance channel vdWwell orients the reactants
in an unreactive configuration. At higher collision
energies, this effect is diminished, and the expected
enhancement of the reaction for the stretch-excited
CHD
3
(over translational energy) is seen.
Next, we considered the experimental results
on the vibrational distribution of HCl (5). As noted
in (5), vibrationally adiabatic theory predicts that
the ground state and CH-stretchexcited Cl +
CHD
3
reactions produce exclusively HCl(v = 0)
and HCl(v = 1) products, respectively. Experi-
ment found the breakdown of this simple theory
for the excited reaction because the measured
fraction of HCl(v = 1), correlated to CD
3
(v = 0),
was only 45%(5). Our dynamics calculations show
that at higher E
coll
, above the energetic threshold for
HCl(v = 1), the ground-state reaction still produces
mainly HCl(v = 0), and the fraction of HCl(v = 1) is
only 1% with only a slight E
coll
dependence. This
small ratio increases to about 2% if the results are
correlated to CD
3
(v = 0), which is in quantitative
agreement with experiment (5). The reactant CH-
stretch excitation increases the fraction of HCl(v =1)
to 10% and 30 to 50% for all the CD
3
states and
CD
3
(v = 0), respectively, which is again in good
agreement with the above-mentioned correlated
experiment (5). The computations support the
experimental observation: The ground state reaction
is vibrationally adiabatic, whereas the CH-stretch
excited reaction is nonadiabatic.
Last to be considered was the very recent ex-
periment (6) on steric control of Cl +CHD
3
(v
1
=1).
We have performed QCTcalculations with aligned
CHD
3
(v
1
= 1) in which the CH-stretch is parallel
or perpendicular to the initial relative velocity
vector of the reactants. We found that the total
reactivity of H-abstraction is higher at parallel
alignment relative to that at perpendicular ori-
entations, which is in agreement with experiment
(6). The trajectories show that the initial orien-
tation is maintained while the Cl approaches
CHD
3
(v
1
= 1), supporting the recent experiment
(6). However, at the turning point the QCTs show
substantial energy transfer causing rotational ex-
citation of CHD
3
, and the prealignment is not
conserved after the collision (Fig. 4, right).
References and Notes
1. L. Che et al., Science 317, 1061 (2007).
2. D. Skouteris et al., Science 286, 1713 (1999).
3. J. C. Polanyi, Science 236, 680 (1987).
4. J. J. Lin, J. Zhou, W. Shiu, K. Liu, Science 300, 966 (2003).
5. S. Yan, Y.-T. Wu, B. Zhang, X.-F. Yue, K. Liu, Science
316, 1723 (2007).
6. F. Wang, J.-S. Lin, K. Liu, Science 331, 900 (2011).
7. W. Zhang, H. Kawamata, K. Liu, Science 325, 303 (2009).
8. G. Czak, Q. Shuai, K. Liu, J. M. Bowman, J. Chem. Phys.
133, 131101 (2010).
9. J. M. Bowman, G. Czak, B. Fu, Phys. Chem. Chem. Phys.
13, 8094 (2011).
10. G. Czak, B. C. Shepler, B. J. Braams, J. M. Bowman,
J. Chem. Phys. 130, 084301 (2009).
11. G. Czak, J. M. Bowman, J. Am. Chem. Soc. 131, 17534 (2009).
12. B. J. Braams, J. M. Bowman, Int. Rev. Phys. Chem. 28,
577 (2009).
13. Materials and methods are available as supporting
material on Science Online.
14. A. G. Csszr, W. D. Allen, H. F. Schaefer, J. Chem. Phys.
108, 9751 (1998).
15. The composite energies were obtained as E[UCCSD( T)/
aug-cc-pVDZ] + E[AE-UMP2/aug-cc-pCVTZ] E[UMP2/
aug-cc-pVDZ], where AE denotes correlating all the
electrons. For the entrance channel, counterpoise and
spin-orbit corrections were computed at the AE-UMP2/
aug-cc-pCVTZ and MRCI+Q/aug-cc-pVTZ levels of theory,
respectively. The PES was represented by a polynomial
expansion in y
ij
= exp(r
ij
/a) (where a = 2 bohr), including
all terms up to total degree six. 3262 coefficients were
determined by a weighted linear least-squares fit of
roughly 16,000 energy points. The root mean square
(RMS)fitting errors are 0.2, 0.4, and 1.0 kcal/mol for
energy intervals (0, 31), (31, 63), and (63, 143), respectively.
16. For 15 arbitrary geometries with energies in the wide
0 to 36 kcal/mol range, we found that this composite
method gives results comparable with extremely
high-quality all-electron CCSD(T)/aug-cc-pCVQZ
calculations with a RMS of only 0.4 kcal/mol, whereas
the RMS error of high-level CCSD(T)/aug-cc-pVTZ
calculations is 1.1 kcal/mol. Furthermore, the composite
method reduces the computational time by factors of
about 1000 and 5 relative to the above-mentioned
high-level computations, respectively.
17. K. A. Peterson, T. H. Dunning Jr., J. Chem. Phys. 117,
10548 (2002).
18. K. Raghavachari, G. W. Trucks, J. A. Pople, M. Head-Gordon,
Chem. Phys. Lett. 157, 479 (1989).
19. J. C. Corchado, D. G. Truhlar, J. Espinosa-Garca, J. Chem.
Phys. 112, 9375 (2000).
20. C. Rangel, M. Navarrete, J. C. Corchado, J. Espinosa-Garca,
J. Chem. Phys. 124, 124306 (2006).
21. S. T. Banks, D. C. Clary, Phys. Chem. Chem. Phys. 9, 933 (2007).
22. H.-G. Yu, G. Nyman, J. Chem. Phys. 111, 6693 (1999).
23. J. F. Castillo, F. J. Aoiz, L. Baares, J. Chem. Phys. 125,
124316 (2006).
24. The benchmark FPA study considers extrapolation to the
complete basis set limit using aug-cc-pVnZ [n = 5 and 6]
bases, electron correlation beyond CCSD(T), core correlation
effects, scalar relativistic effects, and spin-orbit corrections.
25. These relative energies are benchmarked at the
highly accurate CCSD(T)/aug-cc-pCVQZ level of theory
correlating all the electrons, including corrections
for the basis set superposition error.
26. W. R. Simpson, T. P. Rakitzis, S. A. Kandel, T. Lev-On,
R. N. Zare, J. Phys. Chem. 100, 7938 (1996).
27. D. F. Varley, P. J. Dagdigian, J. Phys. Chem. 99, 9843 (1995).
28. C. Murray, B. Retail, A. J. Orr-Ewing, Chem. Phys. 301,
239 (2004).
29. S. J. Greaves et al., Phys. Chem. Chem. Phys. 13, 11438 (2011).
30. W. R. Simpson, T. P. Rakitzis, S. A. Kandel, A. J. Orr-Ewing,
R. N. Zare, J. Chem. Phys. 103, 7313 (1995).
Acknowledgments: G.C. thanks the National Science
Foundation (grant CHE-0625237), and J.M.B. thanks
the U.S. Department of Energy (grant DE-FG02-97ER14782)
for financial support.
Supporting Online Material
www.sciencemag.org/cgi/content/full/334/6054/343/DC1
Materials and Methods
Figs. S1 to S3
Tables S1 to S3
17 May 2011; accepted 9 September 2011
10.1126/science.1208514
21 OCTOBER 2011 VOL 334 SCIENCE www.sciencemag.org 346
REPORTS
800,000 Years of Abrupt
Climate Variability
Stephen Barker,
1
* Gregor Knorr,
2
R. Lawrence Edwards,
3
Frdric Parrenin,
4,5
Aaron E. Putnam,
6
Luke C. Skinner,
7
Eric Wolff,
8
Martin Ziegler
1
We constructed an 800,000-year synthetic record of Greenland climate variability based on the
thermal bipolar seesaw model. Our Greenland analog reproduces much of the variability seen in
the Greenland ice cores over the past 100,000 years. The synthetic record shows strong similarity
with the absolutely dated speleothem record from China, allowing us to place ice core records
within an absolute timeframe for the past 400,000 years. Hence, it provides both a stratigraphic
reference and a conceptual basis for assessing the long-term evolution of millennial-scale variability
and its potential role in climate change at longer time scales. Indeed, we provide evidence for a
ubiquitous association between bipolar seesaw oscillations and glacial terminations throughout the
Middle to Late Pleistocene.
I
ce core records from Greenland first demon-
strated the existence of repeated, large, abrupt
shifts in Northern Hemisphere climate dur-
ing the last ice age (1, 2). These shifts are one
expression of a global system that is capable of
driving major changes in climate components
such as ocean temperatures (3, 4) and monsoon
rainfall (5). The Greenland records provide an
archetypal view of abrupt climate variability (6)
over the last glacial cycle, which was character-
ized by rapid alternations between cold (stadial)
and warmer (interstadial) conditions [known as
Dansgaard-Oeschger (D-O) oscillations]. But iron-
ically, the very high temporal resolution of these
records makes it difficult to look farther back in
time; the high accumulation rates on the Green-
land ice sheet mean that more than 3000 m of
ice may represent just 100,000 years of climate
history. Fortunately, climate records preserved in
Antarctic ice (7) enable us to address this funda-
mental problem.
The thermal bipolar seesaw model (8, 9) at-
tempts to explain the observed relationship between
millennial-scale temperature variability observed
in Greenland and Antarctica by calling on var-
iations in the strength of the Atlantic meridional
overturning circulation (AMOC). The north-
ward heat transport associated with this circulation
(10) implies that changes in the strength of over-
turning should lead to opposing temperature re-
sponses in either hemisphere. According to the
seesaw model, a transition from weak to strong
AMOCwould cause an abrupt warming across the
North Atlantic region (a D-O warming event)
while temperatures across Antarctica would (in
general) shift fromwarming to cooling. The ocean-
atmosphere climate system is an integrated and
synergistic system, and it is important to note that
the overall concept of the bipolar seesaw we in-
voke here is not restricted to oceanic processes but
also includes atmospheric shifts that may be re-
lated to the variations we are interested in(9, 1114).
According to the thermal bipolar seesaw
model, we should observe an antiphase relation-
ship between the Greenland temperature anom-
aly and the rate of change of Antarctic temperature
(9). This can be illustrated by a lead/lag analysis
of the methane-tuned temperature records after
removal of their orbital time scale variability (6)
(Greenland, GL
T
_hi, and Antarctica, AA
T
_hi)
and the first time derivative of Antarctic temper-
ature, AA
T
_hi (Fig. 1). Comparison of the undif-
ferentiated records illustrates the historical debate
as to whether the two signals are positively cor-
related, with a southern lead of 1000 to 1600
years, or negatively correlated, with the north lead-
ing by 400 to 800 years (15, 16). However, as
implied in (9) and illustrated in Fig. 1, a near
zero-phase anticorrelation is observed between
GL
T
_hi and AA
T
_hi. Uncertainties in the ice
agegas age offset (Dage), which may be up to
hundreds of years (17), mean that an exact anti-
phase relationship is unlikely to be observed (6).
As described by (18) using a similar approach,
the process of differentiating amplifies noise in
the original temperature record. Smoothing the
record of AA
T
_hi before differentiating reduces
this noise but will compromise the ability of
AA
T
_hi to replicate the abrupt nature of D-O
warming events and reduce the predicted ampli-
tude of smaller events. The choice of smoothing
window is therefore a trade-off between these ef-
fects (6) (Fig. 1).
The empirical relations illustrated in Fig. 1 of-
fer the possibility of producing a synthetic record
of Greenland climate using the Antarctic record,
with the purpose of reconstructing the nature
of northern variability beyond the present limit
of the Greenland records. The record of Green-
land temperature (GL
T
) is broken down into its
orbital and millennial time scale components,
GL
T
_lo and GL
T
_hi, respectively (Fig. 2 and fig.
S5), where GL
T
_lo is a 7000-year smoothing of
GL
T
(6) and GL
T
_hi is the difference between
GL
T
and GL
T
_lo. We consider GL
T
_hi as the north-
ern temperature anomaly with respect to mean
background conditions. Building on Fig. 1, we
assume that the rate of Antarctic temperature
change is inversely proportional to the northern
temperature anomaly. We therefore scale the am-
plitude of AA
T
_hi to match that of GL
T
_hi to
produce a synthetic record of northern millennial-
scale temperature variability, GL
T
_syn_hi (Fig. 2C)
(6). It can be seen that a synthetic reconstruction
of GL
T
could be made by combining GL
T
_syn_hi
with an estimate for GL
T
_lo. The orbital time
scale components of the Greenland and Ant-
arctic temperature records (GL
T
_lo and AA
T
_lo,
respectively) are highly correlated, with the south-
ern record leading the north by ~2000 years
(6). We therefore incorporate longer time scale
variations into our reconstruction by substitut-
ing GL
T
_lo with a scaled version of AA
T
_lo,
shifted by 2000 years (which we call GL
T
_syn_lo)
(fig. S5) (6). Our full reconstruction, GL
T
_syn
(Fig. 2D), is then the sum of GL
T
_syn_lo and
GL
T
_syn_hi.
Our formulation of the thermal bipolar see-
saw concept is qualitatively analogous to that of
(9) in that it implies the existence of a heat reser-
voir that convolves the northern signal, produc-
ing a southern signal with a longer characteristic
time scale. Our approach is slightly different in
that we relate the rate of Antarctic temperature
change directly to the northern temperature anom-
aly. Indeed, we note that for some long stadial
events, particularly those associated with gla-
cial terminations, Antarctic temperatures appear to
rise unabated until an abrupt warming event oc-
curs in the north (19). On the other hand, our
formulation does not imply that Antarctic temper-
atures must continue to rise indefinitely when-
ever Greenland is cold, only while it is cold with
respect to background conditions (defined by
the orbital time scale component). We also note
that northern temperature (regardless of back-
ground conditions) is not always constant through-
out stadial events. For example, Greenland warmed
significantly during cold stadial 21 (Fig. 2D).
By our formulation, the rate of Antarctic tem-
perature rise during this event would decrease
correspondingly, in line with observations (Fig.
2A) (20).
We used a thresholding approach for predict-
ing the occurrence of abrupt Greenland warming
events based on minima in the second time dif-
ferential of AA
T
(AA
T
) (Fig. 2, E and F). This
has an advantage over use of the first differential
(decreasing through zero) because it is capable of
distinguishing between events of varying mag-
nitude and incorporates information about con-
ditions before and after an abrupt event (6). Using a
relatively insensitive threshold (blue dashed line in
Fig. 2F), we are able to identify the largest D-O
1
School of Earth and Ocean Sciences, Cardiff University, Cardiff
CF10 3AT, UK.
2
Alfred Wegener Institute, 27570 Bremerhaven,
Germany.
3
Department of Earth Sciences, University of Min-
nesota, Minneapolis, MN 55455, USA.
4
Laboratoire de Glaciologie,
CNRS and Joseph Fourier University, 38400 Grenoble, France.
5
Laboratoire Chrono-Environnement, 25000 Besanon, France.
6
Department of Earth Sciences and Climate Change Institute,
University of Maine, Orono, ME 04469, USA.
7
Godwin Lab-
oratory for Palaeoclimate Research, Department of Earth
Sciences, University of Cambridge, Cambridge CB2 3EQ, UK.
8
British Antarctic Survey, Madingley Road, High Cross,
Cambridge CB3 0ET, UK.
*To whom correspondence should be addressed. E-mail:
barkers3@cf.ac.uk
www.sciencemag.org SCIENCE VOL 334 21 OCTOBER 2011 347
REPORTS
temperature shifts recorded in Greenland, whereas
smaller events, such as D-O 2, require a more
sensitive threshold (red line). Accordingly, we are
able to identify almost all of the canonical D-O
events over the past 90,000 years without intro-
ducing spurious events. We incorporate a cor-
rection, based on the time integrated per 55-cm
sample of ice (7), to account for loss of temporal
resolution in the deeper parts of the ice core (6).
Our synthetic reconstruction of Greenland tem-
perature closely resembles the observed record
in terms of both its timing and the structure of in-
dividual events (Fig. 2). This is despite probable
variability in the relationships between d
18
O [for
the Greenland Ice Sheet Project 2 ice core (GISP2)]
or dD [for the European Project for Ice Coring
in Antarctica Dome C ice core (EDC)] versus
local temperature through time (21) and other
millennial-scale variability that might not be re-
lated to the bipolar seesaw. We find similar results
using alternative Antarctic ice core records (6).
On the basis of the predictive ability of GL
T
_syn
over the past ~100,000 years, we can extend our
reconstruction back to ~800,000 years ago (Fig.
3 and Fig. 4). In doing so we implicitly assume
that the empirical relationships observed over
the last glacial cycle held during earlier periods.
Given the inherent uncertainty in this assump-
tion, we do not claim particular skill at pre-
dicting the absolute amplitude of earlier events;
however, we do suggest that to the extent that
the underlying physical mechanisms did persist
throughout the past 800,000 years, the timing
and overall structure of events will be relatively
robust.
To test this hypothesis, we compared our syn-
thetic reconstructions with real climate records.
The record of Asian monsoon variability derived
from cave deposits (speleothems) in China (5, 22)
is one of the best candidates for this task (Fig. 3).
The Chinese speleothem d
18
O record is thought
to represent changes in the proportion of low
d
18
O (summer monsoon) rainfall within annual
totals (22) and can be considered a measure of
the amount of summer monsoon rainfall or mon-
soon intensity. The combined record from several
deposits taken from a number of caves provides
a continuous, absolutely dated record over the
past ~400,000 years (22). The record is domi-
nated by orbital time scale changes, possibly re-
lated to the influence of boreal summer insolation
on the strength of the Asian monsoon (6). Re-
moval of this variability by normalizing to the
insolation curve (6) reveals distinctive millennial
time scale activity that has been shown to cor-
respond with D-O variability over Greenland
during the last glacial cycle (5, 22). This corre-
spondence is thought to be caused by latitudinal
shifts in the position of the Intertropical Con-
vergence Zone (ITCZ) and related atmospher-
ic phenomena in response to variations in the
AMOC and related changes in North Atlantic
temperature (11).
There is a strong one-to-one correspondence
between inferred weak-monsoon events and our
reconstructed cold events in Greenland (Fig. 3).
Moreover, there are pronounced similarities in
the structure of abrupt events, particularly during
deglacial episodes (terminations); the multiple
weak-monsoon events associated with glacial
terminations of the Late Pleistocene (22) are re-
flected by multiple cold events in our records.
Our reconstruction suggests the occurrence of
large-amplitude D-Otype oscillations between
160,000 and 180,000 years ago [during marine
isotope stage 6 (MIS 6)] (Fig. 3). These may be
compared with similar events in the records of
planktonic d
18
O and tree pollen from a marine
3 2 1 0 1 2 3
0.2
0
0.2
0.4
0.6
0.8
1
Lag (kyr)
North leads South leads
B
Methane
C
o
r
r
e
l
a
t
i
o
n
C
o
r
r
e
l
a
t
i
o
n
North leads South leads
3 2 1 0 1 2 3
0.8
0.4
0.4
0.8
Lag (kyr)
0
1000yr
700yr
200yr
C
Vs AA
T
_hi
-10
0
10
30 40 50 60 70 80
A
A
T
_
h
i

(

)
Age (kyr) (EDC3)
-0.05
0
0.05
-0.01
0
0.01
-0.01
0
0.01
-2
0
2
G
L

T
_
h
i

(

)
-100
0
100
E
D
C

C
H
4
(
p
p
b
v
)

-100
0
100
G
I
S
P

C
H
4
(
p
p
b
v
)
A
Fig. 1. Ice core records from Greenland (GISP2) and Antarctica (EDC) (all records have orbital time
scale variability removed; kyr, thousands of years). (A) Methane (17, 25) with tuning points (crosses)
used to place all records on the EDC3 age model (26), Greenland temperature (GL
T
_hi with 200-year
smoothing) derived from d
18
O of ice (2), and Antarctic temperature (AA
T
_hi, 200-year smoothing)
from dD of ice (7). First derivatives of AA
T
_hi (AA
T
_hi) are shown for various smoothing lengths (in brackets) of the undifferentiated record. (B) Lead/lag
correlation of the methane records suggests successful tuning of the gas records. (C) Lead/lag correlations between GL
T
_hi, AA
T
_hi, and AA
T
_hi reveal the
well-known relation between northern and southern temperature records and the antiphase relationship between GL
T
_hi and AA
T
_hi.
21 OCTOBER 2011 VOL 334 SCIENCE www.sciencemag.org 348
REPORTS
Fig. 2. Reconstructing millennial-scale climate
variability over Greenland using the Antarctic tem-
perature record. (A) The record of dD from the EDC
ice core (AA
T
) (7) with a 7000-year smoothing of
the same record (AA
T
_lo). (B) Removal of orbital
time scale variability and application of a 700-
year smoothing (6) produces AA
T
_hi. (C) AA
T
_hi is
differentiated and then scaled to GL
T
_hi (green
curve) to produce GL
T
_syn_hi (orange curve). (D)
A synthetic reconstruction of Greenland temper-
ature variability (GL
T
_syn; red curve) constructed
by adding GL
T
_syn_hi to GL
T
_syn_lo (6). Green
curve is GISP2 d
18
O placed on EDC3 via methane
tuning (Fig. 1). (E and F) Minima in AA
T
below a
threshold (dashed lines) are used to predict the
occurrence of major warming events in Greenland
(F), identified by the corresponding colored dots in
(E). A threshold value of 1.2 10
5
(red dashed
line and dots) has good success at picking canon-
ical D-O events [green numbers in (D) and (E)]
without introducing spurious events.
1
2
3
4 5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
0
1 2 3 4 5 6 7 8 91011 1213 14151617 18 19 20 21 22 0
10 20 30 40 50 60 70 80 90
Age kyr (EDC3)
Age kyr (EDC3)
10 20 30 40 50 60 70 80 90
GL
T
E
-44
-40
-36
G
L

T

(

1
8
O
,

)
D
GL
T
_syn
-2x10
-5
0
2x10
-5
F
-10
0
10
B
A
A
T
_
h
i
-450
-430
-410
-390
E
D
C

D

(

)
A
AA
T
AA
T
_lo
-4
-2
0
2
4
C
GL
T
_hi
GL
T
_syn_hi
G
L

T

_
h
i
Fig. 3. (A to H) Compari-
son of reconstructed Green-
land climate variability with
other records. The normal-
ized record of monsoon
variability from China (A)
(5, 22), marine records from
the Iberian Margin [(C)
to (E)] (23, 32), and the
record of atmospheric CH
4
(G) (25) share many fea-
tures in common with our
records derived from the
Antarctictemperaturerecord
[(B) and (F)]. Colored dots
in (H) represent the oc-
currence of D-O events
predicted from AA
T
using
a fixed (red) or variable
(blue) threshold (6). All
records are on the EDC3
time scale (26) except the
monsoon record which is
on its own absolute time
scale(22). Thepollenrecord
from MD95-2042 (32) (C)
was placed on EDC3 by tun-
ing the corresponding plank-
tonic d
18
O record (24) to
GL
T
_recon. Gray bars indi-
cate cold conditions and
periods of weak monsoon.
Glacial terminations are indicated by Roman numerals.
400
500
600
700
C
H
4

(
p
p
b
v
) G -40
-35
-30
G
L

T
_
s
y
n

(

)
F
120 140 160 180 200 220 240 260
Age (kyr)
0
20
40
T
r
e
e

p
o
l
l
e
n

(
%
)
E
20
40
60
M
e
d
i
t
e
r
r
a
n
e
a
n
p
o
l
l
e
n

(
%
)
D
1
2
3
p
l
a
n
k
t
o
n
i
c

1
8
O

(

)
C
-4
0
4
G
L
T
_
s
y
n
_
h
i

(

)
B
S
p
e
l
e
o
t
h
e
m

1
8
O
(
n
o
r
m
a
l
i
z
e
d
)
A
H
T-II T-III
www.sciencemag.org SCIENCE VOL 334 21 OCTOBER 2011 349
REPORTS
sediment core taken from the Iberian Margin (23).
On the basis of the findings of Shackleton et al.
(24), the Iberian Margin records were tuned to
the EDC dD record via the record of benthic
d
18
O from the same core (23). The tuning exer-
cise did not involve the surface records, which
therefore provide a quasi-objective target for
comparing with our reconstruction [which should
be aligned with the surface-ocean records accord-
ing to (24)], and there is good agreement in the
timing and structure of the abrupt events during
MIS 6. We also note good agreement between
our reconstructions and the record of atmospheric
CH
4
(25). Our predicted D-O warming events are
generally aligned with sharp increases in CH
4
(similar to the observed relationship during MIS
3). This relationship holds for the entire 800,000-
year record (Fig. 4) and provides critical ground-
truthing for our reconstruction.
Building on previous studies (22), we used
the precise and absolutely dated Chinese speleo-
them record to place our reconstruction on an
absolute time scale for the past 400,000 years.
We did this by aligning the cold events in our
reconstruction with the inferred weak-monsoon
events in the speleothem record (Fig. 4) (6). The
EDC3 age scale (which remains the fundamen-
tal basis for our model) was derived through a
combination of ice flow modeling and various
age markers, including orbital tuning constraints
(26). By tuning the millennial-scale features of
GL
T
_syn to the speleothem record, we provide
a refinement of the age scale that provides an
Fig. 4. (A and B) 800,000
years of abrupt climate varia-
bility. Records of North Atlantic
IRD (4), monsoon rainfall (5, 22)
(normalized) and SST from
the Iberian Margin (27) all
show strong similarities with
our reconstruction of Greenland
climate variability (GL
T
_syn_hi
and GL
T
_syn). Glacial termi-
nations (identified by Roman
numerals) are characterized by
cold conditions across Green-
land and the North Atlantic
and weakened monsoon rain-
fall, with a corresponding rise
in atmospheric CO
2
(33), fol-
lowed by an abrupt warming
over Greenland, strengthening
of the monsoon, and sharp rise
in atmospheric CH
4
(25). Pink
boxes indicate terminal North-
ern Hemisphere cold periods.
Red and blue dots are predicted
D-O warming events using a
fixed or variable threshold, re-
spectively. Lowermost curves in
each panel are moving win-
dows of the standard devia-
tion of AA
T
_hi, our bipolar
seesaw activity index (note
that orbital time scale varia-
tions have been removed; blue
is 5000-year window; green is
10,000-year window). Increased
millennial-scale activity is gen-
erally observed during transi-
tions between climate states,
with minimal activity during
interglacials and glacial max-
ima. All records are on the new
Speleo-Age (A) or the EDC3
(B) time scale except the ben-
thic d
18
O stack (LR04) of (34),
which is on its own time scale.
TI TII TIIIA TIII TIV
TV TVI TVII
400
600
800
C
H
4

(
p
p
b
v
)
-45
-40
-35
-30
G
L

T
_
s
y
n

(

)
-4
0
4
G
L
T
_
s
y
n
_
h
i
(

)
A
A
T
_
h
i

S
D

(
x
1
0
-
3
)
0
5
10
3
4
5

1
8
O

(
L
R
0
4
)
400 450 500 550 600 650 700 750 800
Age (kyr) (EDC3)
B
200
300
C
O
2

(
p
p
m
v
)
400
600
800
C
H
4

(
p
p
b
v
)
5
10
15
20
I
b
e
r
i
a
n

M
a
r
g
i
n
S
S
T

(

C
)
-45
-40
-35
-30
G
L

T
_
s
y
n

(

)
-4
0
4
G
L
T
_
s
y
n
_
h
i
(

)
S
p
e
l
e
o
t
h
e
m

1
8
O
(
n
o
r
m
a
l
i
z
e
d
)
0
50
100
N
o
r
t
h

A
t
l
a
n
t
i
c
I
R
D

(
%
)
Age (kyr) Speleo-Age
0 50 100 150 200 250 300 350 400
A
3
4
5

1
8
O

(
L
R
0
4
)
A
A
T
_
h
i

S
D

(
x
1
0
-
3
)
0
5
10
200
300
C
O
2

(
p
p
m
v
)
21 OCTOBER 2011 VOL 334 SCIENCE www.sciencemag.org 350
REPORTS
alternative to the ultimate dependence on orbital
tuning. In addition to providing an absolute time
scale for the ice and gas records from Antarctica,
we can also use our absolutely dated Green-
land reconstruction as a tuning target for other
high-resolution paleo-records, such as records
of ice-rafted debris (IRD) from a North Atlantic
sediment core (4) and a record of sea surface
temperature (SST) from a core off the Iberian
Margin (27) (Fig. 4). Each of these records has
been tuned to our reconstruction on its absolute
time scale (6).
Our synthetic records confirm that millennial
time scale variability and abrupt climate oscilla-
tions occurred in Greenland throughout the past
800,000 years, and more specifically they sug-
gest that the underlying physical mechanisms rep-
resented by the conceptual thermal bipolar seesaw
were relatively invariant throughout this period.
In line with observations for the last glacial pe-
riod (28), our reconstructions suggest that higher-
amplitude variability and more frequent D-Olike
warming events occurred when climate was in
an intermediate state or during the transitions
between states (Fig. 4). Extending the observa-
tions of (22), we find that glacial terminations
of the Middle to Late Pleistocene in general
were characterized by oscillations of the bipolar
seesaw.
This apparently ubiquitous association of
millennial-scale climate variability with glacial
terminations raises an important question: Is this
mode of variability a necessary component of de-
glacial climate change, or merely a complicating
factor? Previous studies (28, 29) have suggested
that D-Otype variability might represent an in-
herent resonance of the climate system, attaining
a high amplitude only within certain windows
of opportunity (i.e., intermediate climate states).
Given that global climate must pass through such
a window during deglaciation, one could argue
that terminal oscillations of the bipolar seesaw are
merely a symptom of deglacial climate change
(29). However, the precise correspondence observed
between bipolar seesaw oscillations and changes
in atmospheric CO
2
during glacial terminations
(Fig. 4) suggests that the bipolar seesaw may play
more than just a passive role in the mechanism
of deglaciation (i.e., through the positive feedbacks
associated with increasing CO
2
) (14, 19, 22). With
the supercritical size of continental ice sheets as
a possible precondition (30), and in combination
with the right insolation forcing (31) and ice al-
bedo feedbacks, the CO
2
rise associated with an
oscillation of the bipolar seesaw could provide
the necessary additional forcing to promote de-
glaciation. In this sense, the overall mechanism
of glacial termination during the Middle to Late
Pleistocene might be viewed as the timely and
necessary interaction between millennial and or-
bital time scale variations.
References and Notes
1. W. Dansgaard et al., Science 218, 1273 (1982).
2. M. Stuiver, P. M. Grootes, Quat. Res. 53, 277
(2000).
3. G. Bond et al., Nature 365, 143 (1993).
4. J. F. McManus, D. W. Oppo, J. L. Cullen, Science 283, 971
(1999).
5. Y. J. Wang et al., Nature 451, 1090 (2008).
6. See supporting material on Science Online.
7. J. Jouzel et al., Science 317, 793 (2007).
8. W. S. Broecker, Paleoceanography 13, 119 (1998).
9. T. F. Stocker, S. J. Johnsen, Paleoceanography 18, 1087
(2003).
10. K. E. Trenberth, J. M. Caron, J. Clim. 14, 3433
(2001).
11. M. Vellinga, R. A. Wood, Clim. Change 54, 251
(2002).
12. J. C. H. Chiang, M. Biasutti, D. S. Battisti,
Paleoceanography 18, 1094 (2003).
13. R. F. Anderson et al., Science 323, 1443 (2009).
14. S. Barker et al., Nature 457, 1097 (2009).
15. A. Schmittner, O. A. Saenko, A. J. Weaver, Quat. Sci. Rev.
22, 659 (2003).
16. E. J. Steig, R. B. Alley, Ann. Glaciol. 35, 451
(2002).
17. T. Blunier, E. J. Brook, Science 291, 109 (2001).
18. M. Siddall et al., Quat. Sci. Rev. 25, 3185 (2006).
19. E. W. Wolff, H. Fischer, R. Rothlisberger, Nat. Geosci. 2,
206 (2009).
20. E. Capron et al., Quat. Sci. Rev. 29, 222 (2010).
21. E. Capron et al., Clim. Past 6, 345 (2010).
22. H. Cheng et al., Science 326, 248 (2009).
23. V. Margari et al., Nat. Geosci. 3, 127 (2010).
24. N. J. Shackleton, M. A. Hall, E. Vincent,
Paleoceanography 15, 565 (2000).
25. L. Loulergue et al., Nature 453, 383 (2008).
26. F. Parrenin et al., Clim. Past 3, 485 (2007).
27. B. Martrat et al., Science 317, 502 (2007).
28. M. Schulz, W. H. Berger, M. Sarnthein, P. M. Grootes,
Geophys. Res. Lett. 26, 3385 (1999).
29. A. Sima, A. Paul, M. Schulz, Earth Planet. Sci. Lett. 222,
741 (2004).
30. M. E. Raymo, Paleoceanography 12, 577 (1997).
31. J. D. Hays, J. Imbrie, N. J. Shackleton, Science 194, 1121
(1976).
32. M. F. Snchez Goi, F. Eynaud, J. L. Turon, N. J. Shackleton,
Earth Planet. Sci. Lett. 171, 123 (1999).
33. D. Lthi et al., Nature 453, 379 (2008).
34. L. E. Lisiecki, M. E. Raymo, Paleoceanography 20,
PA1003 (2005).
Acknowledgments: We thank the authors of all of the
studies cited here for making their results available for
this work. Supported by a Philip Leverhulme Prize (S.B.),
Natural Environment Research Council (UK) awards
NE/F002734/1 and NE/G004021/1 (S.B.), and NSF grants
0502535 and 1103403 (R.L.E.). This study is also part
of the British Antarctic Survey Polar Science for Planet
Earth Programme, funded by the Natural Environment
Research Council (UK).
Supporting Online Material
www.sciencemag.org/cgi/content/full/science.1203580/DC1
Materials and Methods
Figs. S1 to S14
Tables S1 to S3
References
31 January 2011; accepted 26 August 2011
Published online 8 September 2011;
10.1126/science.1203580
Pre-Clovis Mastodon Hunting 13,800
Years Ago at the Manis Site, Washington
Michael R. Waters,
1
* Thomas W. Stafford Jr.,
2,5
H. Gregory McDonald,
3
Carl Gustafson,
4
Morten Rasmussen,
5
Enrico Cappellini,
5
Jesper V. Olsen,
6
Damian Szklarczyk,
6
Lars Juhl Jensen,
6
M. Thomas P. Gilbert,
5
Eske Willerslev
5
The tip of a projectile point made of mastodon bone is embedded in a rib of a single disarticulated
mastodon at the Manis site in the state of Washington. Radiocarbon dating and DNA analysis
show that the rib is associated with the other remains and dates to 13,800 years ago. Thus, osseous
projectile points, common to the Beringian Upper Paleolithic and Clovis, were made and used
during pre-Clovis times in North America. The Manis site, combined with evidence of mammoth
hunting at sites in Wisconsin, provides evidence that people were hunting proboscideans at
least two millennia before Clovis.
R
ecent studies have strengthened the case
that the makers of Clovis projectile points
were not the first people to occupy the
Americas (15). If hunting by humans was re-
sponsible for the megafauna extinction at the
end of the Pleistocene (6), hunting pressures
must have begun millennia before Clovis (7).
Here we reexamine the evidence from the Manis
site in the state of Washington (8), an early mas-
todon kill that dates to 800 years before Clovis.
Between 1977 and 1979, a single male mas-
todon (Mammut americanum) was excavated from
sediments at the base of a kettle pond at the
Manis site (figs. S1 to S3) (810). Some bones
were spirally fractured, multiple flakes were re-
moved from one long bone fragment, and other
bones showed cut marks (8, 11, 12). The only
documented artifact associated with the masto-
don was a foreign osseous fragment, interpreted
as the tip of a bone or antler projectile point,
1
Center for the Study of the First Americans, Departments of
Anthropology and Geography, Texas A&M University, 4352
TAMU, College Station, TX 778434352, USA.
2
Stafford Re-
search, 200 Acadia Avenue, Lafayette, CO 800261845, USA.
3
Park Museum Management Program, National Park Service,
1201 Oakridge Drive, Suite 150, Fort Collins, CO 80525, USA.
4
245 Southeast Derby Street, Pullman, WA 991632217, USA.
5
Centre for GeoGenetics, University of Copenhagen, ster
Voldgade 5-7, 1350 Copenhagen, Denmark,
6
Novo Nordisk
Foundation Center for Protein Research, Faculty of Health Sci-
ences, University of Copenhagen, Blegdamsvej 3b, 2200 Co-
penhagen, Denmark.
*To whom correspondence should be addressed. E-mail:
mwaters@tamu.edu
www.sciencemag.org SCIENCE VOL 334 21 OCTOBER 2011 351
REPORTS
embedded in a rib fragment that was recovered
ex situ from sediments excavated when a back-
hoe uncovered the bone bed (Fig. 1 and fig. S4)
(8). Organic matter associated with the masto-
don yielded calibrated radiocarbon ages of ~14
thousand years ago (ka) (8, 10) (table S1). Over
the past 35 years, the age and evidence for hu-
man involvement with the Manis mastodon have
been challenged (13).
We obtained 13 accelerator mass spectrom-
etry (AMS)
14
C dates from purified bone collagen
(4) extracted from the mastodon rib containing
the embedded osseous object and from both
tusks (table S2). All dates were statistically iden-
tical at 1 SD and establish an age of 11,960 T
17
14
C years before the present (yr B.P.) for the
Manis mastodon (Table 1; average of four XAD
fractions; 13,860 to 13,765 calendar yr B.P.) (14).
These dates show that the ex situ mastodon rib
and in situ skeleton are contemporaneous.
High-resolution x-ray computed tomography
(CT) scanning (15) revealed that the osseous
object embedded in the rib is dense bone shaped
to a point (Fig. 1 and movies S1 and S2). The
point penetrated 2.15 cm into the rib; the tip
broke after entering the rib and separated from
the main shaft. The combined length of the point
fragment (tip length plus the length of the em-
bedded and external shaft piece) is 3.5 cm.
The rib with the embedded projectile point is
a right 12th, 13th, or 14th rib in a series of 19,
but most likely the 14th rib (Fig. 2). The projec-
tile point entered the dorsal surface of the prox-
imal end of the rib immediately distal to the lateral
margins of the two articular facets at approxi-
mately a 45 angle relative to the axis of the head
of the rib. The point would have penetrated the
hair and skin and about 25 to 30 cm of super-
ficial epaxial muscles (Fig. 2 and fig. S5). Thus it
was at least 27 to 32 cm long, comparable with
the known length of later, Clovis-age thrown and
thrust bone points (16). There is no evidence of
bone growth around the point, indicating that the
mastodon died soon after it was attacked.
DNA and protein sequencing were under-
taken on the rib and bone point (supporting on-
line material text 4 and 5). Attempts to amplify
a 140base pair (bp) fragment of the 16S mito-
chondrial DNA (mtDNA) from the rib using
universal vertebrate primers (17) produced only
modern (human) contamination. However, re-
designing primers for a 69-bp fragment (including
primers, table S8) of D-loop mtDNA produced
sequences from both the rib and bone point that
were identical to mastodon and distinct from
other proboscideans (mammoth or elephant) by
nine substitutions.
We also obtained high-resolution tandem mass
spectrometry (MS/MS)based protein sequences
from the projectile point and rib, and used another
mastodon sample as a second reference (tables
S3 to S6). The MS/MS spectra from the bone
point matched the reconstructed mastodon col-
lagen sequences, with the highest scores being
within a reference set of collagen sequences (table
S7 and supporting table of bone point marker
peptides). These results and controls show that
the point was fashioned from mastodon bone.
The Manis site provides further evidence of
a human presence in the New World 800 years
before Clovis [13 ka (4)] and shows that people
were hunting with mastodon bone weapons
made from earlier kills. Evidence for pre-Clovis
hunting also comes from the 14.2-ka Schaefer
site and 14.8-ka Hebior site, Wisconsin (18, 19),
where stone artifacts, but no projectile points,
were found with the remains of mammoth (Mam-
muthus primigenius). Additional evidence of mega-
fauna hunting comes from sites where artifacts
are absent, but taphonomic evidence suggests hu-
man butchering, such as at the 13.8-ka Ayer Pond
site (45SJ454), Orcas Island, Washington (20).
Studies of the dung fungal spore Sporormiella
from lakes in Indiana and New York imply that
megafauna populations collapsed there between
14.8 and 13.7 ka (7). Thus, the impact of human
hunters on the North American megafauna was
more prolonged than previously hypothesized and
was not a Clovis blitzkrieg (21). The absence
of stone projectile points at Manis, Hebior,
Schaefer, and Orcas Island and the presence of
an osseous projectile point at Manis suggest that
osseous projectile points may have been the pre-
dominant hunting weapon during the pre-Clovis
period. Bone and ivory points and other tools
are common in the Upper Paleolithic of Siberia
and in late Pleistocene sites in Beringia (2224).
They are durable and lethal hunting weapons
that continued to be used during and after Clovis
(16, 23, 25). The invention and spread of a new
hunting weapon at 13 kathe Clovis lithic point
may have accelerated the demise of or doomed
the last megafaunal species.
Table 1. AMS
14
C ages used to date the Manis Mastodon.
Specimen dated Date (
14
C yr B.P. T 1 SD) Lab number Material dated
Mastodon tusk ivory
sample no. 1
11,975 T 35 UCIAMS-11350 XAD-gelatin
(KOH collagen)
Mastodon tusk ivory
sample no. 1
11,975 T 35 UCIAMS-12046 XAD-gelatin
(KOH collagen)
Mastodon tusk ivory
sample no. 2
11,890 T 35 UCIAMS-11677 XAD-gelatin
(KOH collagen)
Mastodon rib with
embedded bone
projectile point
11,990 T 30 UCIAMS-29113
XAD-gelatin
(KOH collagen)
Average of four
radiocarbon
measurements
11,960 T 17
14
C yr B.P.
(13,860 to 13,763 calendar yr B.P.)
n = 4
XAD-gelatin
(KOH collagen)
0 1cm 0.5
A
C
B
0 1 2 cm
0 1 2 cm
D
Fig. 1. Mastodon rib with the embedded bone projectile point. (A) Closeup view. (B) Reconstruction
showing the bone point with the broken tip. The thin layer represents the exterior of the rib. (C) CT
x-ray showing the long shaft of the point from the exterior to the interior of the rib. (D) The entire rib
fragment with the embedded bone projectile point.
21 OCTOBER 2011 VOL 334 SCIENCE www.sciencemag.org 352
REPORTS
References and Notes
1. T. Goebel, M. R. Waters, D. H. ORourke, Science 319,
1497 (2008).
2. M. T. P. Gilbert et al., Science 320, 786 (2008).
3. M. R. Waters et al., Science 331, 1599 (2011).
4. M. R. Waters, T. W. Stafford Jr., Science 315, 1122 (2007).
5. T. D. Dillehay et al., Science 320, 784 (2008).
6. P. S. Martin, in Quaternary Extinctions, a Prehistoric
Revolution, P. S. Martin, R. G. Klein, Eds. (Univ. of
Arizona Press, Tucson, AZ, 1984), pp. 354403.
7. J. L. Gill, J. W. Williams, S. T. Jackson, K. B. Lininger,
G. S. Robinson, Science 326, 1100 (2009).
8. C. E. Gustafson, D. Gilbow, R. Daugherty, Can. J. Archaeol.
3, 157 (1979).
9. K. L. Petersen, P. J. Mehringer Jr., C. E. Gustafson,
Quat. Res. 20, 215 (1983).
10. V. E. Morgan, thesis, Washington State University,
Pullman, WA (1985).
11. D. W. Gilbow, thesis, Washington State University,
Pullman, WA (1981).
12. A. L. Runnings, thesis, Washington State University,
Pullman, WA (1984).
13. G. Haynes, The Early Settlement of North America:
The Clovis Era (Cambridge Univ. Press, Cambridge, 2002).
14. P. J. Reimer et al., Radiocarbon 51, 1111 (2009).
15. T. M. Ryan, G. R. Milner, J. Archaeol. Sci. 33, 871 (2006).
16. B. A. Bradley, M. B. Collins, C. A. Hemmings,
Clovis Technology (International Monographs in
Prehistory, no. 17, Ann Arbor, MI, 2010).
17. P. G. Taylor, Mol. Biol. Evol. 13, 283 (1996).
18. D. F. Overstreet, in Paleoamerican Origins: Beyond
Clovis, R. Bonnichsen, B. T. Lepper, D. Stanford,
M. R. Waters, Eds. (Center for the Study of the First
Americans, Texas A&M University, College Station, TX,
2005), pp 183195.
19. D. J. Joyce, Quat. Int. 142-143, 44 (2006).
20. S. M. Kenady, M. C. Wilson, R. F. Schalk, R. R. Mierendorf,
Quat. Int. 233, 130 (2011).
21. D. K. Grayson, D. J. Meltzer, J. Archaeol. Sci. 30, 585
(2003).
22. T. Goebel, Evol. Anthropol. 8, 208 (1999).
23. R. D. Guthrie, in Animals and Archaeology: Hunters and
Their Prey, J. Clutton-Brock, C. Grigson, Eds. (British
Archaeological Reports International Series 163, Oxford,
1983), pp. 273294.
24. C. E. Holmes, Arctic Anthropol. 38, 154 (2001).
25. H. Knecht, in Projectile Point Technology, H. Knecht, Ed.
(Plenum, New York, 1997), pp. 191212.
Acknowledgments: We thank the North Star Archaeological
Research Program established by J. Cramer and R. Cramer
and the Chair in First Americans Studies for funding.
We thank J. Southon for providing the ultrafiltration
14
C measurements. Work conducted at the Center for
GeoGenetics was supported by the Danish National
Research Foundation. E.C. is supported by the
European Union with a Marie Curie Intra European
Fellowship (grant number 237227). J.O., D.S., and
L.J. are supported by the Novo Nordisk Foundation
Center for Protein Research. CT scanning was
performed at the High Resolution X-ray CT Facility at
the University of Texas, Austin. J. Halligan prepared
the illustrations. T. Jennings, J. Halligan, T. Goebel,
S. Fiedel, and two anonymous individuals reviewed
the manuscript.
Supporting Online Material
www.sciencemag.org/cgi/content/full/334/6054/351/DC1
SOM Text
Figs. S1 to S5
Tables S1 to S8
References
Table of bone point marker peptides
29 April 2011; accepted 8 September 2011
10.1126/science.1207663
Fig. 2. Anatomical position of the
Manis rib. (A) Two vertebrae with
the Manis rib inserted into its cor-
rect anatomical position. The blue
arrow points to the embedded point
fragment. (B) Side view of mastodon
vertebrae with the Manis rib inserted
into its correct anatomical position,
with the trajectory of the point indi-
cated. (C) Mastodon skeleton show-
ing the location of ribs 12 to 14.
0 4 8 cm
transverse
processes
anterior
posterior
p
o
i
n
t

t
r
a
j
e
c
t
o
r
y
point tip
A B
C
www.sciencemag.org SCIENCE VOL 334 21 OCTOBER 2011 353
REPORTS
Information Transduction Capacity of
Noisy Biochemical Signaling Networks
Raymond Cheong,
1
Alex Rhee,
1
Chiaochun Joanne Wang,
1
Ilya Nemenman,
2
Andre Levchenko
1
*
Molecular noise restricts the ability of an individual cell to resolve input signals of different
strengths and gather information about the external environment. Transmitting information
through complex signaling networks with redundancies can overcome this limitation. We developed
an integrative theoretical and experimental framework, based on the formalism of information
theory, to quantitatively predict and measure the amount of information transduced by
molecular and cellular networks. Analyzing tumor necrosis factor (TNF) signaling revealed that
individual TNF signaling pathways transduce information sufficient for accurate binary decisions,
and an upstream bottleneck limits the information gained via multiple integrated pathways.
Negative feedback to this bottleneck could both alleviate and enhance its limiting effect,
despite decreasing noise. Bottlenecks likewise constrain information attained by networks
signaling through multiple genes or cells.
S
ignaling networks are biochemical systems
dedicated to processing information about
the environment provided by extracellular
stimuli. Large populations of cells can accurately
sense signaling inputs, such as the concentration
of growth factors or other receptor ligands, but
this task can be challenging for an individual cell
affected by biochemical noise (13). Noise maps
an input signal to a distribution of possible output
responses, which can cause loss of information
about the input. For example, a cell cannot re-
liably distinguish different inputs that, because of
noise, can generate the same output (Fig. 1A).
Conventional metrics related to the standard
deviation or variance of the response distribution
measure noise magnitude (48), but fail to elu-
cidate how noise quantitatively affects the ac-
curacy of information processing in single cells.
By contrast, an information theoretic approach
(Fig. 1B), and the metric of mutual information in
particular, can quantify signaling fidelity in terms
of the maximum number of input values that a
cell can resolve in the presence of noise. Such
methods have been commonly used to evaluate
man-made telecommunication systems (9) and
more recently in computational neuroscience and
in analyses of transcriptional regulatory systems
(1014), but have not been applied to biochem-
ical signaling networks. We developed a general
integrative theoretical and experimental frame-
work to predict and measure the mutual infor-
mation transduced by one or more signaling
pathways. Applying this framework to analyze a
four-dimensional compendium of single-cell re-
sponses to tumor necrosis factor (TNF) (Fig. 1C,
see also SOM section 1), an inflammatory cyto-
kine that initiates stochastic signaling at physio-
logic concentrations spanning about four orders
of magnitude (1521), shows that signaling via a
network rather than a single pathway can abate
the information lost to noise. Furthermore, an in-
formation bottleneck can restrict the maximum
information a network can capture, and negative
feedback potentially but not always relieves this
limitation.
The mutual information, I(R;S), measured in
bits, is the binary logarithmof the maximumnum-
ber of input signal values (S), such as ligand con-
centrations, that a signaling system can perfectly
resolve on the basis of its noisy output responses
(R) (9). One bit of information can resolve two
different signal values, 2 bits resolves four val-
ues, etc. More generally,
I(R;S) =
S

R
P(R,S)log
2
P(R,S)
P(R)P(S)

dRdS
1
1
Department of Biomedical Engineering, Johns Hopkins Uni-
versity, 3400 North Charles Street, Baltimore, MD 21218, USA.
2
Departments of Physics and Biology, Emory University, 400
Dowman Drive, Atlanta, GA 30322, USA.
*To whom correspondence should be addressed. E-mail:
alev@jhu.edu
A
B
C D
Fig. 1. Information theoretic analysis of cell signaling fidelity. (A) Schematic
showing information loss due to overlapping noisy response distributions. (B)
Diagram of the TNFNF-kB signaling pathway represented in biochemical
form (left) and as a noisy communication channel (right). (C) Experimental
flowchart for using immunocytochemistry to sample the conditional response
distribution at single-cell resolution and resulting four-dimensional compen-
dium of multiple responses in cells of multiple genetic backgrounds to
multiple TNF concentrations, at multiple time points. The data were collected
in a single experiment, allowing controlled, quantitative comparisons along
each dimension. (D) Distributions of noisy NF-kB nuclear translocation re-
sponses to 30-min TNF exposure (examples shown at top) used to compute the
channel capacity of the TNFNF-kB pathway. Scale bars, 20 mm.
21 OCTOBER 2011 VOL 334 SCIENCE www.sciencemag.org 354
REPORTS
The joint distribution P(R,S) determines the
marginal distributions P(R) and P(S), and hence
also the mutual information, and can be decom-
posed as P(R,S) = P(S) P(R|S). The response dis-
tribution, P(R|S), is experimentally accessible by
sampling responses of individual isogenic cells to
various signal levels (Fig. 1C), and its spread re-
flects the noise magnitude given any specific
input. The signal distribution, P(S), reflects po-
tentially context-specific frequencies at which a
cell experiences different signal values. Although
the amount of information might thus vary from
case to case, one can also determine the maximal
amount of transducible information, given the ob-
served noise (see SOM section 2). This quantity,
known as the channel capacity (9), is a general
characteristic of the signaling system and the
signal-response pair of interest and can thereby be
experimentally measured without making as-
sumptions about the (possibly nonlinear) relation
between Rand S, signal power, or noise properties.
Using immunocytochemistry, we assayed nu-
clear concentrations of the transcription factor
nuclear factor kB (NF-kB) in thousands of in-
dividual mouse fibroblasts 30 min after exposure
to various TNFconcentrations (Fig. 1D). We chose
this time point because NF-kBtranslocation peaks
at 30 min regardless of the concentration used,
initiating expression of early-response inflamma-
tory genes (1922). The NF-kB response value
in a single cell could yield at most 0.92 T 0.01 bits
of information, which is equivalent to resolving
2
0.92
= 1.9, or about 2, concentrations of the TNF
signal, thus essentially only reliably indicating
whether TNF is present or not. (See SOM sec-
tions 2.2 and 3 regarding the low experimental
uncertainty.) A bimodal input signal distribution,
P(S), with peaks at lowand high TNF concentra-
tions, maximizes the information (fig. S1), support-
ing the notion of essentially binary (digital) sensing
capabilities of this pathway (18), although we did
not observe bimodal output responses, P(R|S).
Noise also limits other canonical pathways,
including signaling by platelet-derived growth
factor (PDGF), epidermal growth factor (23), and
G proteincoupled receptors (24), to ~1 bit (fig.
S2, A to C, and table S1). Even the most reliable
system we examined, morphogen gradient sig-
naling through the receptor Torso in Drosophila
embryos (25), was limited to 1.61 bits (fig. S2D
and table S1), corresponding to about three dis-
tinguishable signal levels.
The pathways examined above are examples
of individual biochemical communication chan-
nels (Fig. 1B) that capture relatively lowamounts
of information about signal intensity, which would
allow only limited reliable decision making by a
cell. However, information in biological systems
is typically processed by networks comprising
multiple communication channels, each transduc-
ing information about the signal. For instance, a
transcription factor often regulates many genes,
a receptor many transcription factors, and a dif-
fusible ligand many cells. The integrated outputs
of such multiple channels can provide more in-
formation about the signal than the output of any
one channel (see SOM section 4). Subsequently,
downstream signaling processes that converge
to co-regulate common effectors, biological pro-
cesses, or physiologic functions can provide the
point needed to integrate the multiple outputs to
realize the benefit of increased aggregate infor-
mation (fig. S3). To provide a unified framework
for analyzing such various networks, we first the-
oretically investigated the information gained by
network signaling in general, then experimentally
tested the predictions made by the theory when
applied to a specific system.
We considered two information theoretic mod-
els, similar to models of population coding in
neural systems (2628), for transmitting a signal
S through multiple channels to the responses R
1
,
R
2
, , R
n
, under the assumption of Gaussian
variables (see SOM section 5). The bush model
uses independent channels (topologically resem-
bling an upside-down shrub) (Fig. 2A), whereas
the tree model signals through a common chan-
nel (trunk) to the intermediate, C, before di-
verging into independent branches (Fig. 2B). The
information resulting from the bush model is
I
bush
(R
1
, :::, R
n
; S) =
1
2
log
2
1 n
s
2
S
s
2
SR

2
wheres
2
S
is the variance of the signal distribution,
and s
2
SR
is the noise (variance) introduced in
each branch. Thus, the information can grow
logarithmically with the number of branches
without an upper bound. In contrast, the informa-
tion resulting from the tree model is
I
tree
(R
1
, :::, R
n
; S) =
1
2
log
2
1
ns
2
S
=s
2
CR
1 ns
2
SC
=s
2
CR

3
where s
2
SC
and s
2
CR
are the trunk and branch
noises, respectively (see SOMsection 3.3). As the
number of branches increases, the information
asymptotically approaches an upper limit equal
to the mutual information between the input
signal and the common intermediate. Thus, the
information lost to noise in the trunk determines
the maximum throughput of a tree network.
The key difference between bush and tree
networks is the absence or presence of this trunk-
based information bottleneck. The biochemical
structure of a network can resemble a tree, but if
there is little loss of information upstream, the
A
B
C D
Fig. 2. Information gained by signaling through a network comprising mul-
tiple communication channels. (A) Schematic of a bush network with inde-
pendent channels lacking an information bottleneck. (B) Schematic of a tree
network with channels sharing a common trunk that forms an information
bottleneck. Circles represent noise introduced in the indicated portions of the
signaling network; see text for definition of symbols. (C) Comparison of bush
and tree model predictions for the capacity of the TNF network to experimental
values. At 30 min, the NF-kB and ATF-2 pathways together capture more
information about TNF concentration than either pathway alone (bars 1 to 3),
and the tree rather than bush model accurately predicts this increase (bars 3 to
5). The tree model further predicts a receptor-level bottleneck of 1.26 T 0.13
bits (bar 6). (D) Joint distribution of NF-kB and ATF-2 responses to 30-min
stimulation of TNF. Each data point represents a single cell, and each
concentration of TNF examined is shown with a distinct color.
www.sciencemag.org SCIENCE VOL 334 21 OCTOBER 2011 355
REPORTS
bush model lacking a bottleneck might best es-
timate the capacity of the network. Additionally,
the bush and tree models make various semi-
quantitative predictions (see SOM section 6),
such as the information captured by a network
based on the capacities of its component path-
ways. For example, for a bush network compris-
ing two pathways each with 1-bit responses,
Eq. 2 implies s
2
S
=s
2
SR
3 and that together
they should yield
1
2
log
2
1 23 1:4 bits.
TNF activates the NF-kB and c-Jun N-
terminal kinase (JNK) pathways, stimulating nu-
clear localization of NF-kB and phosphorylated
activating transcription factor2 (ATF-2) (fig. S4),
respectively (29). To determine if the TNF sig-
naling network contains an appreciable upstream
information bottleneck limiting the information
captured by these pathways, we examined wheth-
er the bush (bottleneck absent) or tree (bottleneck
present) network model better approximates the
network (fig. S5). The models are applicable be-
cause the NF-kB (Fig. 1D) and ATF-2 (fig. S6)
response distributions are approximately Gaussian
at all TNF concentrations. We found that NF-kB
alone yielded at most 0.92 bits of information
about TNF concentration, and ATF-2 alone
yielded at most 0.85 T 0.02 bits (fig. S1B and
table S1). Together, the bush model predicts that
these pathways jointly yield 1.27 T 0.01 bits (Fig.
2C), and a similar model assuming indepen-
dent pathway responses that are not necessarily
Gaussian likewise predicts an increase to 1.13 T
0.01 bits. The actual information determined by
dual-staining immunocytochemistry (Fig. 2D) was
1.05 T 0.02 bits, much lower than both predic-
tions (Fig. 2C), demonstrating that the bush mod-
el does not approximate the TNF network well.
In contrast, the tree model predicts 1.03 T 0.01
bits, matching the experimental value within er-
ror (Fig. 2C), and also correctly predicts the sta-
tistical dependency between the responses given
the signal (fig. S7).
The correspondence between the tree mod-
el predictions and experimental measurements
strongly indicates that the network contains an
information bottleneck. The tree model predicts
that the maximum information that can pass
through the bottleneck is 1.26 T 0.13 bits (Fig. 2C),
A B C D
E
Fig. 3. Effect of negative feedback to the bottleneck on information transfer. (A) TNF signaling network
diagram showing A20-mediated negative feedback to the information bottleneck. (B) Comparison of infor-
mation about TNF concentration captured with and without A20 negative feedback. The information is larger at
30 min but smaller at 4 hours in wild-type cells as compared to A20
/
cells. (C and D) Schematic of NF-kB
dynamics in wild-type and A20
/
mouse fibroblasts exposed to saturating concentrations of TNF. Average
dynamics (black) and the expected magnitudes of the dynamic range (double arrow) and noise (single arrow)
are shown. See fig. S9 for experimental support. (E) Comparison of NF-kB responses to zero (basal) or sat-
urating concentrations of TNF. Differences in the means with and without TNF indicate the dynamic range, and
error bars (SD) indicate the noise.
A B C
Fig. 4. Information gained by signaling through networks of multiple genes. (A)
Plot shows the unique curve (solid black) determined by the tree model (inset),
passing through the experimentally determined values (circles), for information as
a function of the number of copies of a NF-kB reporter gene. The upper limit,
corresponding to the maximum information captured by integrating NF-kB ac-
tivity over time, is 1.64 T 0.36 bits (blue dashed line). (B) Expression-level dis-
tributions of clonal cell lines containing different numbers of copies of an NF-kB
reporter gene in response to ~10 hours of TNF exposure. (C) Time courses cor-
responding to individual cells showing cell-to-cell differences in the onset and rate
of NF-kB reporter gene expression (left). In each cell, expression is nearly linear
and deterministic in time, as quantified by the correlation coefficient (right) of the
time course after onset of expression (shown schematically in inset on left).
21 OCTOBER 2011 VOL 334 SCIENCE www.sciencemag.org 356
REPORTS
corresponding to just 2
1.26
= 2.3 distinguishable
TNF concentrations. The known biochemistry of
TNF signaling implies that the bottleneck (trunk)
comprises the steps of TNF receptor complex ac-
tivation common to both pathways, including lig-
and binding, receptor trimerization, and complex
formation and activation. Because all TNF sig-
naling passes through the receptor complex, mul-
tiple pathways in the TNF signaling network,
activated at the 30-min time point, only modestly
increase the information about TNF concentra-
tion regardless of the number of pathways or their
fidelity.
We next explored whether negative feedback,
which can reduce noise (12, 30, 31), might alle-
viate the receptor-level signaling bottleneck. The
information captured by a single channel (Eq. 2,
n = 1) can be written as
1
2
log
2
s
2
R
=s
2
SR
. Thus,
negative feedback can have equivocal effects on
information, depending on the balance of the tend-
encies for negative feedback to reduce both the
dynamic range of the signaling response (32), rep-
resented by the response variance s
2
R
, and noise,
represented bys
2
SR
. Indeed, comparison of wild-
type cells and cells lacking A20 (fig. S8), an
inhibitor of TNF receptor complexes whose ex-
pression is up-regulated by NF-kB(33) (Fig. 3A),
showed that A20-mediated negative feedback in-
creases information at the 30-min time point, but
decreases it at 4 hours (Fig. 3B).
To understand these different outcomes, we
examined how A20 affects the dynamic range
and noise at either time point. At the early time
point, constitutively expressed A20 inhibits basal
NF-kB activity, but TNF does not induce A20
expression rapidly enough to affect saturating
levels of NF-kB at 30 min (Fig. 3, C and D, and
fig. S9) (17, 34). Hence, A20 negative feedback
decreases noise, primarily at low TNF concen-
trations, and also increases the dynamic range by
lowering basal NF-kB levels (Fig. 3E and fig.
S10A), explaining why information at 30 min is
higher for wild-type than for A20
/
cells (Fig. 3B).
In contrast, at the late time point, A20 is in-
creased in wild-type cells (17, 34). The negative
feedback decreases noise at all TNF concentra-
tions but also decreases the dynamic range by
strongly suppressing the maximum inducible
NF-kB activity (Fig. 3E and fig. S10A). The net
effect is lower information for wild-type versus
A20
/
cells at 4 hours (Fig. 3B).
We observed that A20 negative feedback sim-
ilarly both improves and limits information at the
early and late time points, respectively, for ATF-2
alone, or together with NF-kB (Fig. 3B and fig.
S10B), consistent with A20 affecting the portion
of the network common to both pathways. Nev-
ertheless, the maximal information about TNF con-
centration acquired with or without A20-mediated
negative feedback was still ~1 bit, suggesting
limited advantages for mitigating the information
bottleneck in this pathway by using negative
feedback.
We next considered whether networks com-
prising multiple target genes can capture sub-
stantial amounts of information through time
integration. If the target gene product lifetime is
long compared to its transcription and translation
time scales, the accumulated protein concentra-
tion is approximately proportional to the time in-
tegral of signaling activity, thereby averaging out
temporal fluctuations (35, 36). However, the bio-
chemical readout of protein synthesis can intro-
duce extra noise, confounding determination of
the information contained in the time integral.
Fortunately, the maximum information captured
by a tree network, in which the time integral of
transcription factor activity is the intermediate sig-
nal activating multiple independent target genes
(Fig. 4A, inset), is determined by the trunk (time
integration) rather than branch noise (readout mech-
anism). We measured the information captured
by such tree networks in cells stably transfected
with different copy numbers (1.8-fold difference,
as determined by polymerase chain reaction) of
the gene coding for a stable green fluorescent
protein (GFP) (37) reporting on NF-kB activity
(Fig. 4B). Using the tree model to extrapolate the
extent of the bottleneck, under the assumption
that ~10 hours of TNF exposure induces similar
expression level and noise for each gene, in-
dicates that 1.64 T 0.36 bits is the maximum in-
formation that integrating NF-kB activity over
the experimental time period can yield about
TNF concentration (Fig. 4A), regardless of the
readout mechanism.
To understand why information was only mod-
erately higher compared to a single time point
(1.64 versus 0.92 bits), we monitored GFP re-
porter gene expression in individual cells, finding
that, for any given cell, GFP accumulated linearly
in time in a nearly deterministic fashion, although
its onset and accumulation rate varied fromcell to
cell (Fig. 4C). This is consistent with observa-
tions made with live cell probes (1820) showing
NF-kB dynamics to be essentially deterministic
over the experimental time scale within each cell,
but distinct across cells. We thus conclude that
the ability of time integration to increase the in-
formation about TNF concentration is limited by
the lack of rapid temporal fluctuations that would
otherwise be suppressed by integration over the
10-hour response.
Finally, we considered signaling via multiple
cells, each considered as separate information
channels within a network (Fig. 5A, inset). An
ensemble of cells resembles a bush network if
each cell directly and independently accesses the
same signal, and because bush networks do not
contain trunk-based bottlenecks, substantial in-
creases in information might be obtained. To test
this hypothesis, we analyzed the collective TNF
response of different numbers of cells, as mea-
sured by immunocytochemistry. We varied cell
number by considering cells within nonoverlap-
ping circular regions of variable diameter (Fig.
5B) and used the average NF-kB response with-
in each region to simulate cells contributing to a
collective response in proportion to their NF-kB
activity. The bush model predicts (Eq. 2), and
the data confirm (Fig. 5A), that the information
should increase logarithmically with the number
of independently signaling cells functioning
collectively.
Moreover, we found that networks of just 14
cells can yield up to 1.8 bits of information, far
greater than the other network types analyzed
above. Because ensembles of this size can plau-
sibly experience a similar concentration of a dif-
fusing signal such as TNF and function collectively
(21, 38) [e.g., TNF-activated blood vessel en-
dothelial cells (39)], collective cell behavior can
effectively increase the information gained and
produce responses that can discriminate be-
tween many TNF concentrations. Nonetheless,
networks relying on cell-cell communication can
still contain bottlenecks. For instance, TNF can
be secreted by macrophages stimulated by lipo-
polysaccharide (LPS) from invading bacteria,
with the information about the initial LPS dose
lost within the macrophage signaling networks
before secretion of TNF.
By treating biochemical signaling systems as
information theoretic communication channels,
we have rigorously and quantitatively shown that,
Fig. 5. Information gained by signaling through networks of multiple cells. (A) Comparison of ex-
perimentally measured information obtained by collective cell responses (circles) versus logarithmic trend
(solid black line) predicted the bush model (inset). (B) Schematic of methodology used to measure
collective cell responses.
www.sciencemag.org SCIENCE VOL 334 21 OCTOBER 2011 357
REPORTS
in a single cell, noise can substantially restrict the
amount of information transduced about input
intensity, particularly within individual signaling
pathways. The bush and tree network models,
which provide a unified theoretical framework for
analyzing branched motifs widespread in natural
and synthetic signaling networks, further dem-
onstrated that signaling networks can be more
effective in information transfer, although bot-
tlenecks can also severely limit the information
gained. Receptor-level bottlenecks restrict the TNF
and also PDGF signaling networks (fig. S11) and
may be prevalent in other signaling systems.
We explored several strategies that a cell
might use to overcome restrictions due to noise.
We found that negative feedback can suppress
bottleneck noise, which can be offset by concom-
itantly reduced dynamic range of the response.
Time integration can increase the information
transferred, to the extent that the response under-
goes substantial dynamic fluctuations in a single
cell over the physiologically relevant time course.
The advantage of collective cell responses can
also be substantial, but limited by the number of
cells exposed to the same signal or by the in-
formation present in the initiating signal itself.
Responses incorporating the signaling history
of the cell might also increase the information
(40, 41). For instance, responses relative to the
basal state (fold-change response) might be less
susceptible to noise arising from diverse initial
states (23), although this does not necessarily
translate into large amounts of transferred infor-
mation (table S1). Similarly, for the reporter gene
system described here (fig. S12), ~0.5 bits of ad-
ditional information can be obtained if a cell can
determine expression levels at both early and late
time points. However, noise in the biochemical
networks that a cell uses to record earlier output
levels and to later compute the final response
may nullify the information gain potentially pro-
vided by this strategy. Overall, we anticipate that
the information theory paradigm can extend to
the analysis of noise-mitigation strategies and
information-transfer mechanisms beyond those
explored here, in order to determine what specific
signaling systems can do reliably despite noise.
References and Notes
1. J. G. Albeck, J. M. Burke, S. L. Spencer, D. A. Lauffenburger,
P. K. Sorger, PLoS Biol. 6, e299 (2008).
2. N. Rosenfeld, J. W. Young, U. Alon, P. S. Swain,
M. B. Elowitz, Science 307, 1962 (2005).
3. T. J. Perkins, P. S. Swain, Mol. Syst. Biol. 5, 326
(2009).
4. W. J. Blake, M. Kaern, C. R. Cantor, J. J. Collins, Nature
422, 633 (2003).
5. M. B. Elowitz, A. J. Levine, E. D. Siggia, P. S. Swain,
Science 297, 1183 (2002).
6. J. Paulsson, Nature 427, 415 (2004).
7. J. M. Pedraza, A. van Oudenaarden, Science 307,
1965 (2005).
8. J. M. Raser, E. K. OShea, Science 304, 1811
(2004).
9. T. M. Cover, J. A. Thomas, Elements of Information Theory
(Wiley, New York, 1991).
10. R. R. de Ruyter van Steveninck, G. D. Lewen, S. P. Strong,
R. Koberle, W. Bialek, Science 275, 1805 (1997).
11. D. Fuller et al., Proc. Natl. Acad. Sci. U.S.A. 107, 9656
(2010).
12. E. Ziv, I. Nemenman, C. H. Wiggins, PLoS ONE 2,
e1077 (2007).
13. G. Tkacik, C. G. Callan Jr., W. Bialek, Proc. Natl. Acad.
Sci. U.S.A. 105, 12265 (2008).
14. P. Mehta, S. Goyal, T. Long, B. L. Bassler, N. S. Wingreen,
Mol. Syst. Biol. 5, 325 (2009).
15. R. Cheong, A. Hoffmann, A. Levchenko, Mol. Syst. Biol. 4,
192 (2008).
16. R. Cheong, C. J. Wang, A. Levchenko, Mol. Cell. Proteomics
8, 433 (2009).
17. S. L. Werner et al., Genes Dev. 22, 2093 (2008).
18. S. Tay et al., Nature 466, 267 (2010).
19. L. Ashall et al., Science 324, 242 (2009).
20. D. E. Nelson et al., Science 306, 704 (2004).
21. R. Cheong et al., J. Biol. Chem. 281, 2945 (2006).
22. A. Hoffmann, A. Levchenko, M. L. Scott, D. Baltimore,
Science 298, 1241 (2002).
23. C. Cohen-Saidon, A. A. Cohen, A. Sigal, Y. Liron, U. Alon,
Mol. Cell 36, 885 (2009).
24. X. R. Bao, I. D. Fraser, E. A. Wall, S. R. Quake,
M. I. Simon, Biophys. J. 99, 2414 (2010).
25. M. Coppey, A. N. Boettiger, A. M. Berezhkovskii,
S. Y. Shvartsman, Curr. Biol. 18, 915 (2008).
26. B. B. Averbeck, P. E. Latham, A. Pouget,
Nat. Rev. Neurosci. 7, 358 (2006).
27. J. W. Pillow et al., Nature 454, 995 (2008).
28. E. Schneidman, W. Bialek, M. J. Berry II, J. Neurosci. 23,
11539 (2003).
29. H. Wajant, K. Pfizenmaier, P. Scheurich, Cell Death Differ.
10, 45 (2003).
30. A. Becskei, L. Serrano, Nature 405, 590 (2000).
31. I. Lestas, G. Vinnicombe, J. Paulsson, Nature 467,
174 (2010).
32. R. C. Yu et al., Nature 456, 755 (2008).
33. I. E. Wertz et al., Nature 430, 694 (2004).
34. E. G. Lee et al., Science 289, 2350 (2000).
35. V. Shahrezaei, P. S. Swain, Proc. Natl. Acad. Sci. U.S.A.
105, 17256 (2008).
36. S. Krishna, M. H. Jensen, K. Sneppen, Proc. Natl. Acad.
Sci. U.S.A. 103, 10840 (2006).
37. S. Thierfelder, K. Ostermann, A. Gbel, G. Rdel,
Appl. Biochem. Biotechnol. 163, 954 (2011).
38. K. Francis, B. O. Palsson, Proc. Natl. Acad. Sci. U.S.A. 94,
12258 (1997).
39. J. Parkin, B. Cohen, Lancet 357, 1777 (2001).
40. I. Nemenman, G. D. Lewen, W. Bialek, R. R. de Ruyter
van Steveninck, PLOS Comput. Biol. 4, e1000025
(2008).
41. S. P. Strong, R. Koberle, R. R. de Ruyter van Steveninck,
W. Bialek, Phys. Rev. Lett. 80, 197 (1998).
Acknowledgments: We thank A. Hoffmann, M. Simon,
S. Shvartsman, C. Cohen-Saidon, and U. Alon for sharing
data and materials; A. Ganesan and H. Chang for
experimental assistance; and P. Iglesias, Y. Qi, and
A. Feinberg for insightful discussions and reviewing
drafts of the manuscript. This work was supported by
NIH grants GM072024 and RR020839 (R.C., A.R., C.J.W.,
and A.L.) and CA132629 (I.N.), the Medical Scientist
Training Program at the Johns Hopkins University (R.C.),
and, in early stages of the work, the Los Alamos
National Laboratory Directed Research and Development
program (I.N.).
Supporting Online Material
www.sciencemag.org/cgi/content/full/science.1204553/DC1
Materials and Methods
SOM Text
Figs. S1 to S12
Table S1
References (4255)
18 February 2011; accepted 7 September 2011
Published online 15 September 2011;
10.1126/science.1204553
ER Tubules Mark Sites of
Mitochondrial Division
Jonathan R. Friedman,
1
Laura L. Lackner,
2
Matthew West,
1
Jared R. DiBenedetto,
1
Jodi Nunnari,
2
Gia K. Voeltz
1
*
Mitochondrial structure and distribution are regulated by division and fusion events.
Mitochondrial division is regulated by Dnm1/Drp1, a dynamin-related protein that forms
helices around mitochondria to mediate fission. Little is known about what determines sites
of mitochondrial fission within the mitochondrial network. The endoplasmic reticulum (ER)
and mitochondria exhibit tightly coupled dynamics and have extensive contacts. We tested
whether ER plays a role in mitochondrial division. We found that mitochondrial division occurred
at positions where ER tubules contacted mitochondria and mediated constriction before Drp1
recruitment. Thus, ER tubules may play an active role in defining the position of mitochondrial
division sites.
R
egulation of mitochondrial division is crit-
ical to normal cellular function; excess
division is linked to numerous diseases,
including neurodegeneration and diabetes (1, 2).
The central player in mitochondrial division is
the highly conserved dynamin-related protein
(Drp1 in mammals, Dnm1 in yeast), which belongs
to a family of large guanosine triphosphatases
(GTPases) that self-assemble to regulate mem-
brane structure (3). Division dynamins are likely
to work by oligomerizing in a GTP-dependent
manner into helices that wrap around mitochon-
dria; locally controlled assembly-stimulated GTP
hydrolysis is thought to provide the mechano-
chemical force that completes fission of the out-
er and inner membranes (4). There are additional
proteins required for mitochondrial division, such
as the outer membrane protein Mff (mitochondrial
fission factor), which is present only in mam-
mals (5). Although general mechanisms exist for
1
Department of Molecular, Cellular, and Developmental Biol-
ogy, University of Colorado, Boulder, CO 80309, USA.
2
De-
partment of Molecular and Cellular Biology, University of
California, Davis, CA 95616, USA.
*To whom correspondence should be addressed. E-mail:
gia.voeltz@colorado.edu
21 OCTOBER 2011 VOL 334 SCIENCE www.sciencemag.org 358
REPORTS
Fig. 1. Mitochondrial constriction and division oc-
curs at ER-mitochondrial contacts in yeast. (A) The
3D models (left images) of ER (green) and mito-
chondria (purple) at contact domains were imaged
by EM and tomography of high-pressure frozen yeast
cells. Middle images are 2D tomographs of contact
sites (second column, ER drawn in green) and the
corresponding 3D models of each (third column).
Contact, marked in red, is defined as regions where
the ER membrane comes within 30 nm of the mito-
chondrial membrane, and ribosomes are excluded
(third column). Right schematics demonstrate the
percentage of the mitochondrial circumference that
makes contact with the ER membrane [red is contact,
white is not (19)]. The diameter of each mitochon-
drion at positions of ER contact is shown. Regions
where the mitochondria are constricted (models a
and c) have a high percent of ER wrapping. Addi-
tional EM tomographs and analysis of constrictions
are shown in fig. S1, A and B. (B) Time-lapse images
of yeast cells expressing mito-dsRed and GFP-HDEL
(ER). A single focal plane is shown. Arrows and arrow-
heads indicate sites of mitochondrial division. A cor-
responding z-series is shown in fig. S1C. Scale bars
indicate, in (A), 200 nm; (B), 2 mm.
A
0s 40s 80s 120s 160s 200s 240s 280s
B
G
F
P
-
H
D
E
L
m
i
t
o
-
d
s
R
E
D
m
i
t
o
-
d
s
R
E
D
74%
138 nm
43%
215 nm
91%
146 nm
11%
193 nm
a
b
a
b
a
b
c
d
c
d
c
d
a
b
c
d
GFP-Sec61
mito-dsRed
mito-dsRed
A B
GFP-Sec61
mito-dsRed
0s 10s 20s 30s 0s 10s 20s 30s
0s 10s 20s 40s
C
0s 10s 20s 30s
D
mito-dsRed
Fig. 2. Mitochondrial division occurs at ER-mitochondrial contact sites in
mammalian cells. (A to D) Four examples of mitochondrial division over time
courses shown in Cos-7 cells expressing GFP-Sec61b (ER) and mito-dsRed.
The site of mitochondrial division (white arrows) and the position of the newly
formed mitochondrial ends (yellow arrows) are shown. Additional examples are
included in fig. S2A. Scale bars, 1 mm.
www.sciencemag.org SCIENCE VOL 334 21 OCTOBER 2011 359
REPORTS
recruiting Dnm1 or Drp1 to mitochondria, it is not
known whether there are specific sites on mito-
chondria that are marked for division (6). Fur-
thermore, both Dnm1 and Drp1 oligomerize into
helices that are much smaller than the diameter of
mitochondria (Dnm1 helices have reported mean
diameters of 109 nmin yeast and 129 nmin vitro),
suggesting that Dnm1 (Drp1)independent mito-
chondrial constriction may be needed to facilitate
mitochondrial division (4, 69).
Contact sites exist between mitochondria and
the endoplasmic reticulum (ER) and are impor-
tant for phospholipid synthesis and calciumsignal-
ing [for review, see (10)]. Based on recent data,
there are likely several types of molecular bridges
that mediate these contacts, such as the ERMES
complex identified in yeast and the mitochondrial
fusion protein mitofusin 2 (Mfn2) in mammalian
cells (11, 12). These physical contacts are per-
sistent and maintained under dynamic conditions
(13), suggesting that the ER-mitochondrial inter-
face is vital for function. We have used electron
microscopy (EM) and tomography to analyze the
three-dimensional (3D) structure of contacts be-
tween the ER and mitochondria in the yeast
Saccharomyces cerevisiae. We observed the high-
resolution (~4 nm) structure and 3D models of
four ER-mitochondrial contacts taken from two
cells (Fig. 1A). In these examples, the ER was
wrapped around mitochondria to varying degrees.
In two of the four examples, the ER almost com-
pletely circumscribed the mitochondrial outer
membrane, and mitochondria were constricted at
the point of contact (mitochondrial diameter 138
nm and 146 nm circumscribed versus 215 nm
and 193 nmuncircumscribed at ERcontact) (Fig.
1A; fig. S1, A and B; and movies S1 and S2).
These data suggest that ER tubules associate
with and may mediate mitochondrial constric-
tion sites.
We thus examined the role of ER in mito-
chondrial division by using fluorescence micros-
copy in live yeast cells transformed with an ER
marker (GFP-HDEL) and mito-dsRed to image
the behavior of ER and mitochondria simulta-
neously over time. The vast majority of mito-
chondrial division events were spatially linked to
sites of ER-mitochondrial contact (87%, n = 112
Linescan BFP-KDEL
mito-EGFP mito-EGFP
mCherry-Drp1
BFP-KDEL
mito-EGFP
mCherry-Drp1
mito-EGFP
E
Distance (pixels)
R
e
l
a
t
i
v
e

f
l
u
o
r
e
s
c
e
n
c
e
i
n
t
e
n
s
i
t
y
B
D
mCherry-Drp1
GFP-Sec61
mito-BFP
mito-BFP / mCherry-Drp1 / GFP-Sec61
0s 40s 80s 120s
0s 40s 80s 120s
Always
Part of time
Never
P
e
r
c
e
n
t

o
f

p
u
n
c
t
a
e
n=101
0s 40s 80s 120s
Distance (pixels)
R
e
l
a
t
i
v
e

f
l
u
o
r
e
s
c
e
n
c
e
i
n
t
e
n
s
i
t
y
Distance (pixels) R
e
l
a
t
i
v
e

f
l
u
o
r
e
s
c
e
n
c
e
i
n
t
e
n
s
i
t
y
Drp1 localization to
ER-mitochondrial contacts
GFP-HDEL mito-CFP
Dnm1-
mCherry Merge
A
0s
10s
20s
30s
C
Fig. 3. Dnm1- and Drp1-mediated mitochondrial division occurs at ER con-
tact sites. (A) Time-lapse images of wild-type yeast cells expressing mito-CFP,
GFP-HDEL (ER), and Dnm1-mCherry. A single focal plane is shown. Arrows indi-
cate the site of mitochondrial division, which is marked by both ER-mitochondria
contact and Dnm1. (B) Merged image of a live Cos-7 cell expressing GFP-Sec61b
(ER), mito-BFP, and mCherry-Drp1. (C) Examples of cells as in (B) that show that
Drp1 punctae maintain colocalization with positions of ER-mitochondrial contact
over time. White arrows indicate Drp1 punctae that maintain contact with both
the ER and mitochondria. Yellow arrows indicate a rare example of Drp1 that
does not contact the ER. (D) The percentage of mitochondrial Drp1 punctae that
colocalize with the ER membrane over a 2-min time course. (E) Examples of
mitochondrial constrictions at ER contact sites marked by Drp1. Left-hand im-
ages showCos-7 cells expressing mito-EGFP, BFP-KDEL (ER), and mCherry-Drp1,
merged as indicated. Right graphs are line scans drawn through the mitochon-
dria and show the relative fluorescence intensity of mitochondria (green), ER
(blue), and Drp1 (red) along its length. White arrows at constrictions on images
correspond to black arrows shown on the line scan. Additional examples are
shown in fig. S4. Scale bars for (A), (C), and (E), 1 mm; (B), 5 mm.
21 OCTOBER 2011 VOL 334 SCIENCE www.sciencemag.org 360
REPORTS
from 281 cells) (Fig. 1B). ER tubules crossed
over (Fig. 1B, yellow arrows) and wrapped around
mitochondria (Fig. 1B, white arrows, and fig.
S1C). At ER-mitochondrial contact sites, mito-
chondrial constriction followed by mitochondrial
division was observed (Fig. 1B).
We next tested whether ER plays a similar
role in mammalian mitochondrial division by using
fluorescence microscopy of live Cos-7 cells tran-
siently transfected with fluorescent markers for
ER(GFP-Sec61b) and mitochondria (mito-dsRed).
We imaged regions of the cell periphery where
contacts between the mitochondria and ER were
well resolved and observed that mitochondrial
division events predominantly occurred at sites
of contact between ERand mitochondria (94%,
n = 32 from23 cells) (Fig. 2, fig. S2A, and movies
S3 and S4). Furthermore, the majority of events
(88%) were sites of ER tubules crossing over the
mitochondria, suggesting that the structural con-
text of the interaction is important. The frequency
of ER-associated mitochondrial division is much
higher than would be predicted on the basis of
the area of mitochondria covered by crossing ER
tubules as determined by colocalization of mito-
chondrial and ER markers (fig. S2B).
Thus, in both yeast and mammalian cells, ER
tubules are at mitochondrial division sites and
may be involved in mitochondrial constriction dur-
ing this process. Next, we asked whether mito-
chondrial division occurs in yeast cells that have
substantially reduced levels of tubules because of
the absence of the membrane shaping proteins
Rtns and Yop1 (14, 15). By using both EM and
fluorescence microscopic analyses, we observed
that, in regions of mutant cells in which ER tu-
bules were dramatically reduced, short ER tu-
bules extended out of the massive ER cisternae
and associated with mitochondrial constrictions
and division events (fig. S3). Thus, ER tubules
are a consistent feature of ER contact at mito-
chondrial constrictions, even under conditions
where most tubules are depleted. Furthermore,
Rtns and Yop1 are dispensable for the biogenesis
of the ER tubules that associate with mitochon-
drial division events.
To ask whether ER-associated division events
are spatially linked to the mitochondrial division
machinery, we determined the relationship of ER-
mitochondrial contacts to the division dynamins
Dnm1 and Drp1. Dnm1 and Drp1 assemble into
punctate structures at steady state, and a subset
of these structures are found on mitochondria
and at mitochondrial division sites (6, 16, 17).
We imaged live yeast transformed with Dnm1-
mCherry, mitocyan fluorescent protein (CFP),
and GFP-HDEL (ER) and observed that a large
percentage of Dnm1 punctae were at sites of
mitochondrial-ER contact (46%, n = 225). These
Dnm1 punctae could be observed at sites where
ER tubule crossover and mitochondrial division
occurred (Fig. 3A). In Cos-7 cells transiently trans-
fected with GFP-Sec61b (ER), mitoblue flu-
orescent protein (BFP), and mCherry-Drp1, we
observed that the majority of Drp1 punctae sta-
bly associated with mitochondria and localized
to ER-mitochondrial contacts over time (Fig. 3,
B to D, and movie S5). Furthermore, a subset
of Drp1 at these contacts was associated with a
mitochondrial constriction site (78%, excluding
punctae localized to mitochondrial tips, n = 50).
The mitochondrial constrictions marked by Drp1
punctae were always either at ER tubule cross-
overs (81%) or adjacent to them (19%) (Fig. 3E
and fig. S4). Together, the localization of the
mitochondrial division dynamins in yeast and
mammalian cells to regions of ER-mitochondrial
contacts and the observations that these regions
are associated with constricted mitochondria and
subsequent division indicate a direct role of the
ER in the process of mitochondrial division.
Mff is a mammalian-specific mitochondrial
outer membrane protein required for mitochon-
drial localization of Drp1 and division (5, 18).
Drp1 and Mff colocalize in punctate structures on
mitochondria, and Mff punctae persist in cells
where Drp1 expression is reduced by RNAi (18).
Thus, Mff punctae may mark the future sites of
mitochondrial division before Drp1 recruitment
(18). In Cos-7 cells transiently transfected with
C
BFP-KDEL
mito-dsRed
GFP-Mff BFP-KDEL
mito-dsRed mito-dsRed mito-dsRed
GFP-Mff
Linescan
Distance (pixels)
R
e
l
a
t
i
v
e

f
l
u
o
r
e
s
c
e
n
c
e
i
n
t
e
n
s
i
t
y
Distance (pixels)
R
e
l
a
t
i
v
e

f
l
u
o
r
e
s
c
e
n
c
e
i
n
t
e
n
s
i
t
y
BFP-KDEL

mito-dsRed
mito-dsRed
D
GFP-Sec61
mito-dsRed
M
f
f

R
N
A
i
Linescan
R
e
l
a
t
i
v
e

f
l
u
o
r
e
s
c
e
n
c
e
i
n
t
e
n
s
i
t
y
Distance (pixels)
Distance (pixels)
R
e
l
a
t
i
v
e

f
l
u
o
r
e
s
c
e
n
c
e
i
n
t
e
n
s
i
t
y
A
BFP-KDEL
mito-dsRed
GFP-Mff
D
r
p
1

R
N
A
i
GAPDH
Drp1
siRNA
c
o
n
t
r
o
l
D
r
p
1
IB:
B
GAPDH
Mff
c
o
n
t
r
o
l
M
f
f
IB:
At ER tubule crossing 16 (64%)
Adjacent to ER tubule crossing 4 (16%)
Alongside ER tubule 3 (12%)
Not at ER 2 (8%)
Constrictions marked
by Mff (Drp1 RNAi)
n = 25
GFP-Sec61
Fig. 4. The ER localizes to mitochondrial constrictions before Drp1 and
Mff recruitment. (A) Examples of mitochondrial constrictions at ER contacts
marked by Mff in Cos-7 cells depleted of Drp1. Left and center images show
these cells expressing mito-dsRed, BFP-KDEL (ER), and GFP-Mff, merged as
indicated. Right graphs are line scans drawn through the mitochondria and
show the relative fluorescence intensity of mitochondria (red), ER (blue), and
Mff (green) along their length. White arrow positions at constrictions corre-
spond to black arrows on the line scan. Additional examples are shown in fig.
S6. (B) Western blots with antibody against Drp1 (top) or Mff (bottom) and
GAPDH demonstrate depletion of Drp1 in lysates from cells transfected with
siRNA against Drp1 [as in (A)] or Mff [as in (D)] compared with control RNAi
cells. (C) The number of Mff-localized mitochondrial constrictions in Drp1-
depleted cells that colocalize with ER tubules, from 23 cells. (D) As in (A), for
cells depleted of Mff and expressing GFP-Sec61b (ER; green on line scan) and
mito-dsRed (red on line scan). Scale bars for (A) and (D) large left images, 5 mm;
(A) and (D) smaller center images, 1 mm.
www.sciencemag.org SCIENCE VOL 334 21 OCTOBER 2011 361
REPORTS
GFP-Mff, mCherry-Drp1, and mito-BFP, we ob-
served that Mff circumscribed and localized to
punctae on mitochondria, the majority of which
colocalized with Drp1 (fig. S5, A to C). To test
whether Mff punctae localize to ER contacts
independently of Drp1, we depleted Drp1 from
Cos-7 cells with small interfering RNA (siRNA)
and cotransfected these cells with GFP-Mff, mito-
dsRed, and BFP-KDEL (ER). Drp1 was sub-
stantially depleted in Drp1 RNA interference
(RNAi) cells in comparison with the control
cells (Fig. 4B). Selective depletion of Drp1 was
further supported by the aberrant and elongated
mitochondrial morphology in Drp1 RNAi cells
(Fig. 4A and fig. S5D). As expected (18), in Drp1-
depleted cells, Mff punctae localized to mitochon-
dria (Fig. 4A). We asked whether mitochondria
were constricted at Mff punctae in the absence of
Drp1, and if so, whether these sites localized to
ERcontacts. Of the 25 constrictions we resolved,
16 were at an ER crossover (64%), and another
4 were adjacent to an ER tubule crossing (16%)
(Fig. 4, Aand C, and fig. S6). Thus, Mff localizes
in a Drp1-independent manner to mitochondrial
constrictions at sites of ERcontact. We next asked
whether the ER localizes to regions of mitochon-
drial constriction in the absence of Mff. Cos-7
cells were depleted of Mff by siRNAand cotrans-
fected with GFP-Sec61b (ER) and mito-dsRed.
As expected, mitochondrial morphology was
elongated in these cells (Fig. 4, B and D, and
fig. S5E). In cells depleted of Mff, we observed
mitochondrial constriction at sites of ER contact,
indicating that ER-mitochondrial contacts form
and mark positions of mitochondrial constriction
independently of both Mff and Drp1 recruitment
(Fig. 4D).
Here, we have shown that ER-mitochondrial
contacts are a conserved feature of mitochondrial
division. We envision two ways that ER contact
might directly regulate mitochondrial division: (i)
ERproteins intimately participate in division, and/or
(ii) ER tubules physically wrap around and con-
strict mitochondria to a diameter comparable to
Dnm1 and Drp1 helices to facilitate their recruit-
ment and assembly to complete fission (fig. S9).
The latter is attractive given that the diameter of
Dnm1 helices (~110 to 130 nm) is considerably
narrower than that of mitochondria and is quite
similar to the diameter of constricted mitochon-
dria at ER tubule contacts (138 nm and 146 nm)
(4, 69). Regardless of the exact mechanism, the
ER appears to mark the division site and is likely
to be an active participant in this process, because
it remains in contact with the mitochondria through
the entire fission event. Many human diseases are
associated with excessive mitochondrial division,
raising the intriguing possibility that these diseases
could involve an alteration of ER-mitochondrial
contacts.
References and Notes
1. D. H. Cho, T. Nakamura, S. A. Lipton, Cell. Mol. Life Sci.
67, 3435 (2010).
2. Y. Yoon, C. A. Galloway, B. S. Jhun, T. Yu, Antioxid. Redox
Signal. 14, 439 (2011).
3. L. L. Lackner, J. M. Nunnari, Biochim. Biophys. Acta
1792, 1138 (2009).
4. E. Ingerman et al., J. Cell Biol. 170, 1021 (2005).
5. S. Gandre-Babbe, A. M. van der Bliek, Mol. Biol. Cell 19,
2402 (2008).
6. A. Legesse-Miller, R. H. Massol, T. Kirchhausen, Mol. Biol.
Cell 14, 1953 (2003).
7. A. M. Labrousse, M. D. Zappaterra, D. A. Rube,
A. M. van der Bliek, Mol. Cell 4, 815 (1999).
8. Y. Yoon, K. R. Pitts, M. A. McNiven, Mol. Biol. Cell 12,
2894 (2001).
9. J. A. Mears et al., Nat. Struct. Mol. Biol. 18, 20
(2011).
10. O. M. de Brito, L. Scorrano, EMBO J. 29, 2715
(2010).
11. O. M. de Brito, L. Scorrano, Nature 456, 605
(2008).
12. B. Kornmann et al., Science 325, 477 (2009); 10.1126/
science.1175088.
13. J. R. Friedman, B. M. Webster, D. N. Mastronarde,
K. J. Verhey, G. K. Voeltz, J. Cell Biol. 190, 363
(2010).
14. M. West, N. Zurek, A. Hoenger, G. K. Voeltz, J. Cell Biol.
193, 333 (2011).
15. G. K. Voeltz, W. A. Prinz, Y. Shibata, J. M. Rist,
T. A. Rapoport, Cell 124, 573 (2006).
16. H. Sesaki, R. E. Jensen, J. Cell Biol. 147, 699
(1999).
17. E. Smirnova, L. Griparic, D. L. Shurland,
A. M. van der Bliek, Mol. Biol. Cell 12, 2245
(2001).
18. H. Otera et al., J. Cell Biol. 191, 1141 (2010).
19. Materials and methods are available as supporting
material on Science Online.
Acknowledgments: This work is supported by NIH grant
R01 GM083977 and a Searle Scholar award (to G.K.V.),
NIH training grant GM08759 (to J.R.F.), NIH grant
R01 GM062942 and an American Heart Innovative
Research Grant (to J.N.), and grants from the Biological
Sciences Initiative (BURST grant) and the Undergraduate
Research Opportunity Program at the University of
Colorado (to J.R.D.). We thank the Boulder 3D Electron
Microscopy facility for shared equipment and helpful
suggestions.
Supporting Online Material
www.sciencemag.org/cgi/content/full/science.1207385/DC1
Materials and Methods
SOM Text
Figs. S1 to S9
References (2029)
Movies S1 to S5
22 April 2011; accepted 17 August 2011
Published online 1 September 2011;
10.1126/science.1207385
Antimicrobial Peptides Keep Insect
Endosymbionts Under Control
Frdric H. Login,
1,2
Sverine Balmand,
1,2
Agns Vallier,
1,2
Carole Vincent-Mongat,
1,2
Aurlien Vigneron,
1,2
Michle Weiss-Gayet,
2,3
Didier Rochat,
4
Abdelaziz Heddi
1,2
*
Vertically transmitted endosymbionts persist for millions of years in invertebrates and play an
important role in animal evolution. However, the functional basis underlying the maintenance
of these long-term resident bacteria is unknown. We report that the weevil coleoptericin-A (ColA)
antimicrobial peptide selectively targets endosymbionts within the bacteriocytes and regulates their
growth through the inhibition of cell division. Silencing the colA gene with RNA interference
resulted in a decrease in size of the giant filamentous endosymbionts, which escaped from the
bacteriocytes and spread into insect tissues. Although this family of peptides is commonly
linked with microbe clearance, this work shows that endosymbiosis benefits from ColA,
suggesting that long-term host-symbiont coevolution might have shaped immune effectors
for symbiont maintenance.
C
ooperative associations between animals
and symbiotic bacteria are widespread in
nature and common in insects that exploit
unusually restricted nutritional resources (1). In
manyinsects, intracellular bacteria (endosymbionts)
are transmitted vertically and provide nutrient
supplementation to their hosts, thereby im-
proving their adaptive traits and their invasive
power (24).
However, maintaining the beneficial nature of
this long-term relationship requires both the host
and the symbiont to constrain adaptive interac-
tions. Genomic and evolutionary data have shown
that major deletions and mutations of genes occur
in endosymbionts, some of which are involved in
bacterial virulence and host tolerance (57). Data
on how host immune systems have evolved to
tolerate cooperative bacteria remain scarce and
are mainly limited to extracellular associations
with environmental and/or horizontal symbiont
transmission (8, 9).
To protect permanent endosymbionts from
the hosts systemic immune response, and prevent
competition with opportunistic invaders, sym-
bionts are sequestered in bacteria-bearing host
cells, called the bacteriocytes, which, in some spe-
cies, group together to forma bacteriome (10). To
investigate the immune specificities of bacterio-
cytes, we have studied associations with Sitophilus
1
INSA-Lyon, INRA, UMR203 BF2I, Biologie Fonctionnelle Insectes
et Interactions, F-69621Villeurbanne, France.
2
Universitde Lyon,
F-69003 Lyon, France.
3
Universit Lyon 1, CNRS UMR5534,
Centre de Gntique et de Physiologie Molculaire et Cellulaire,
F-69622 Villeurbanne, France.
4
INRA, Universit Pierre et Marie
Curie, UMR1272 Physiologie de lInsecte Signalisation et Com-
munication, F-78026 Versailles, France.
*To whom correspondence should be addressed. E-mail:
abdelaziz.heddi@insa-lyon.fr
21 OCTOBER 2011 VOL 334 SCIENCE www.sciencemag.org 362
REPORTS
genus. This genus of beetle includes three cereal
pest species (Sitophilus oryzae, S. zeamais, and S.
granarius), all of which share intimate intracellular
symbiosis with a Gram-negative g-Proteobacterium,
called Sitophilus primary endosymbiont (SPE)
(11, 12). In contrast to Buchnera in aphids, the
Sitophilus association with SPE is relatively re-
cent [dating to 20 million years ago (Ma)] (13, 14).
The SPE genome has not suffered appreciable
genome shrinkage (15) and retains secretion sys-
tems and bacterial wall elements that are impli-
cated in bacterial recognition by the host immune
receptors (6). Transcriptomic data revealed that
SPE induces a strong systemic response when in-
jected into weevil hemolymph (16), whereas per-
manent infection of bacteriocytes with SPE leads
to the up-regulation of only one antimicrobial
peptideencoding gene coleoptericin-A (colA,
fig. S1) (16).
In immunohistochemistry experiments with
antibody against ColA (14), we found that, in apo-
symbiotic insects, colA is expressed constitutively
in epithelial cells surrounding the intestine and in
the fat body, with a high concentration under the
cuticle (Fig. 1Band fig. S2). In symbiotic insects,
colAis further expressed in all the tissues housing
endosymbionts (Fig. 1B). Similarly, follicular cells
also expressed ColA as a thin layer around the
oocytes, and ColA signals were stronger in bacte-
riocytes surrounding bacteriomes (Fig. 1B). Thus,
a relatively high expression of colAin tissues facing
the external environment, and at the boundary of
tissues housing endosymbionts, supports the idea
that ColA may either prevent pathogen intrusion
or retain endosymbionts within the bacteriomes
and oocytes. Moreover, ColA appeared to colocal-
ize with endosymbionts in bacteriocytes (Fig. 1B).
Using immunogold electron microscopy, we found
ColA located inside SPE cytoplasm (Fig. 1C), and
some ColAspots were also attached to the bacterial
membrane surface. Overall, microscopic observa-
tions indicate that ColA expression targets endo-
symbionts in both somatic and germ cell lines.
We tested the antimicrobial activity of ColA
against microbes. The weevil paralog ColB was
used for comparison because the colB gene,
unlike colA, is down-regulated in bacteriocytes
(16) and ColBshows important sequence identity
with ColA (fig. S3). ColA and ColB showed
similar bactericidal activity against the Gram-
positive Micrococcus luteus and the Gram-
Fig. 1. ColA peptide distribution in weevil tis-
sues. (A) Schemes of larva and adult weevils
showing bacteriome localizations (in red). (B) (Up-
per panel) Tissues from aposymbiotic S. zeamais:
Left and middle images are cuticle sections stained
with a preimmune serum [negative controls can
be found in SOM (14)] and with antibody against
ColA (anti-ColA), respectively; the right image
is a gut section stained with anti-ColA. ColA
signals are detected in the fat body, with relatively high intensity under the cuticle and within gut epithelial cells. (Middle and lower panels) Tissues from
symbiotic insects stained with anti-ColA. Middle panel: ColA signals can be seen in oocytes and follicular cells (left), in apical bacteriomes of ovaries
(middle), and in adult mesenteric caeca (right); lower panel: ColA signals in larval bacteriome (left), within bacteriocytes (middle), and in bacteriome
squashes (right) (14). Arrows indicate high ColA signals at the periphery of tissues and show ColA colocalizing with SPE in bacteriocytes and bacteriome
squashes. (C). Immunogold staining of SPE with anti-ColA. Bacteriocyte sections are shown with ColA spots inside symbiont cytoplasm and attached to
bacterial membranes.
www.sciencemag.org SCIENCE VOL 334 21 OCTOBER 2011 363
REPORTS
negative Escherichia coli (Fig. 2), whereas the
yeast Saccharomyces cerevisiae tolerated both
peptides (table S1). Both peptides exhibited a
wide range of bacteriostatic activity against Gram-
negative bacteria. SPE persists in the bacteriocyte
because colA expression may lead either to bac-
terial growth inhibition or dose-dependent bacte-
rial clearance. At bacteriostatic concentrations,
both coleoptericins halt E. coli growth, but only
ColA inhibited cell division or caused bacterial
gigantism (Fig. 2B). The shape of M. luteus was
not affected by ColA or by ColB (fig. S4). Thus,
ColA and ColB have distinct functions in weevil
immunity and symbiosis in regard to their general
effect on Gram-negative bacteria.
Since the discovery of coleoptericins (17, 18),
their function in the immune system and their
role in symbiosis have not yet been explored.
One notable observation is that all the coleop-
teran endosymbionts observed exhibit a similar
elongated morphology (Fig. 2C), resembling
symbionts in other associations. For example, in
Rhizobium-legume symbiosis, Rhizobium elon-
gation has been interpreted to be the result of
repeated chromosome DNA replication without
cell cytokinesis (19). We measured the relation
between bacterial size and genome amplifica-
tion in E. coli, SPE, and Nardonella, the ances-
tral endosymbiont of weevils (125 Ma) (13). All
bacteria were polyploid, and bacterial size was
highly correlated with chromosome number. The
highest scores were seen in Nardonella, with 120
chromosomes observed in a giant cell of 200 mm
(Fig. 2D). Bacterial division in plants is inhibited
by nodule-specific cysteine-rich peptides (20) that
induce irreversible elongation of bacteria and ren-
der them incapable of multiplying in vitro. By
using phylogenetically unrelated molecules, plants
and animals target bacterial cytokinesis while
preserving DNA replication, hence domesticat-
ing the bacteria as symbionts.
To elucidate the mechanism by which ColA
reaches the bacterial cytoplasm and elicits cell
elongation, we used far-Western blotting to iden-
tify bacterial molecules targeted by ColA and
ColB peptides. ColA specifically interacted with
OmpA, OmpC, rp-L2, EF-Ts, and GroEL (fig.
S5 and table S2). No interaction was detected
with Hsp60, the eukaryotic cytosolic homolog
of GroEL. ColB interacted with OmpC and pro-
teins involved in translation, but not with OmpA
or GroEL.
As with colicins and phages (21), it is likely
that Omps are receptors that allow ColA to enter
the cell. Tight attachments of ColA to the SPE
membrane (Fig. 1C) support this assumption, and
SPE genome sequence analysis showed that SPE
encodes a functional ompC gene (table S3). Al-
though the genome sequence is not available, we
found that ColA targets Nardonella of the palm
weevil Rhynchophorus ferrugineus (fig. S6), sug-
gesting that ColA may have a broad impact on
weevil symbioses. Whether Nardonella has re-
tained omps or whether ColAenters the bacterium
by other mechanisms remains to be determined.
We propose that after entering the cytoplasm,
ColA elicits cell elongation through interaction
with GroEL, because ColA, but not ColB, inter-
acts with GroEL protein, and because groEL mu-
tations in E. coli also trigger cell gigantism (22).
The absence of any interaction between ColB and
GroELalso indicates that ColAmay have evolved
a specific interaction with GroEL. In this con-
text, it is notable that GroEL is the most abundant
protein in insect endosymbionts (23); however,
selective up-regulation of groEL has often been
interpreted as an adaptive mechanism for protein
folding in endosymbionts with a high A+T bias
in the genome (24).
We used RNA interference (RNAi) to inhibit
colA transcription in the larval weevils. Injection
of dsRNAcolA resulted in a significant reduc-
tion in the number of bacteriocyte colA transcripts
and abundance of ColA peptide for more than 2
weeks (fig. S7). In contrast to the plant-Rhizobium
interaction, colA inhibition resulted in the SPE
population declining by half (Fig. 3A). However,
whether this was due to resumption of cytokinesis
or multiplication of small bacteria is unclear. We
Fig. 2. ColA and ColB activities against bacteria. (A) ColA (solid line) and ColB (dashed line) activities
against E. coli (triangles) and M. luteus (squares). ColA and ColB have a similar inhibitory effect
(analysis of variance, P =0.76). They show bactericidal activities against M. luteus at low concentrations.
For E. coli, low concentrations of ColA and ColB have bacteriostatic activity, and higher concentrations
kill this bacterium. (B) Effect of low concentrations of ColA and ColB on E. coli morphology. Bacteria
were incubated in LB broth (left, control), in LB with 8 mM ColA (middle), or in LB with 8 mM ColB (right).
Cell gigantism is observed with ColA peptide only. (C) Gram staining of endosymbionts from S. oryzae
(rod-shaped, left), S. zeamais (spiral, middle), and R. ferrugineus (filamentous, right). (D) Chromosome
visualization of E. coli treated with 8 mM ColA (left), SPE (middle), and Nardonella (right). SPE and
Nardonella were isolated from larval bacteriomes of S. oryzae and R. ferrugineus, respectively. Chro-
mosome number was highest in Nardonella.
21 OCTOBER 2011 VOL 334 SCIENCE www.sciencemag.org 364
REPORTS
quantified SPE DNA by quantitative polymerase
chain reaction (qPCR) and showed that colA in-
hibition did not affect bacterial chromosome
replication (fig. S8). This, and the decreased num-
ber of large-sized cells after RNAi treatment (Fig.
3A), supports the former hypothesis (i.e., resump-
tion of cytokinesis and loss of the elongated form)
but does not exclude the latter (multiplication of
the small-sized cells) if we consider bacterial turn-
over and de novo synthesis processes.
To understand ColA function in symbiosis,
we monitored SPE in insects by fluorescence
in situ hybridization (FISH) after dsRNA-colA
injections. Unexpectedly, several SPE cells exited
the bacteriome on the ninth day after treatment
(Fig. 3B). This phenomenon increased at day 14,
when bacteria were found spread throughout the
larval tissues. Nevertheless, symbiont escape
from the bacteriome did not affect insect mortal-
ity under laboratory conditions (table S4), although
these data indicate that ColA acts to prevent
bacterial tissue invasion.
The weevil ColA peptide demonstrates sev-
eral properties important to immunity and sym-
biosis. It appears to act as a first line of defense in
insects against microbial intrusion, and the range
of bacteriostatic and bactericidal activities of
ColA suggests its precise regulation of endosym-
biont number and location. The interaction of
GroEL with ColA (but not with ColB) supports
the idea that long-term coevolution may have se-
lected ColA for this symbiotic function.
References and Notes
1. N. A. Moran, Curr. Biol. 16, R866 (2006).
2. A. E. Douglas, Annu. Rev. Entomol. 43, 17 (1998).
3. A. Heddi, A. M. Grenier, C. Khatchadourian, H. Charles,
P. Nardon, Proc. Natl. Acad. Sci. U.S.A. 96, 6814 (1999).
4. A. G. Himler et al., Science 332, 254 (2011).
5. S. Shigenobu, H. Watanabe, M. Hattori, Y. Sakaki,
H. Ishikawa, Nature 407, 81 (2000).
6. C. Dale, G. R. Plague, B. Wang, H. Ochman, N. A. Moran,
Proc. Natl. Acad. Sci. U.S.A. 99, 12397 (2002).
7. B. L. Weiss, Y. Wu, J. J. Schwank, N. S. Tolwinski,
S. Aksoy, Proc. Natl. Acad. Sci. U.S.A. 105, 15088 (2008).
8. J. L. Round, S. K. Mazmanian, Nat. Rev. Immunol. 9,
313 (2009).
9. J. V. Troll et al., Environ. Microbiol. 12, 2190 (2009).
10. C. Anselme, A. Vallier, S. Balmand, M. O. Fauvarque,
A. Heddi, Appl. Environ. Microbiol. 72, 6766 (2006).
11. A. Heddi, H. Charles, C. Khatchadourian, G. Bonnot,
P. Nardon, J. Mol. Evol. 47, 52 (1998).
12. H. Charles, A. Heddi, Y. Rahb, C. R. Acad. Sci. 324,
489 (2001).
13. C. Conord et al., Mol. Biol. Evol. 25, 859 (2008).
14. Supporting material is available on Science Online.
15. H. Charles, G. Condemine, C. Nardon, P. Nardon,
Insect Biochem. Mol. Biol. 27, 345 (1997).
16. C. Anselme et al., BMC Biol. 6, 43 (2008).
17. P. Bulet et al., J. Biol. Chem. 266, 24520 (1991).
18. A. Sagisaka, A. Miyanoshita, J. Ishibashi, M. Yamakawa,
Insect Mol. Biol. 10, 293 (2001).
19. P. Mergaert et al., Proc. Natl. Acad. Sci. U.S.A. 103,
5230 (2006).
20. W. Van de Velde et al., Science 327, 1122 (2010).
21. D. Fourel, C. Hikita, J. M. Bolla, S. Mizushima,
J. M. Pags, J. Bacteriol. 172, 3675 (1990).
22. E. Chapman et al., Proc. Natl. Acad. Sci. U.S.A. 103,
15800 (2006).
23. H. Charles, A. Heddi, J. Guillaud, C. Nardon, P. Nardon,
Biochem. Biophys. Res. Commun. 239, 769 (1997).
24. N. A. Moran, Proc. Natl. Acad. Sci. U.S.A. 93, 2873 (1996).
Acknowledgments: This work was supported by INRA, INSA
de Lyon, the French ANR-06-BLAN-0316 (EndoSymArt)
and ANR-2010-BLAN-170101 (ImmunSymbArt), a
grant from Rgion Rhne-Alpes (cluster infectiologie),
and the COST action FA0701 (Arthropod Symbioses).
The data reported in this paper are posted in the
supporting online material (14). colA (EY122872) and
colB (EY122826) sequences are published in GenBank.
SPE omp sequences (GenBank JN575265, JN575266,
JN575267) were provided by C. Dale and R. B. Weiss
(University of Utah) from an ongoing SPE genome
sequencing and annotation project supported by NSF
grant EF-0523818 and the Ministerio de Educacin y
Ciencia project BFU2006-06003/BMC to A. Moya and
A. Latorre (University of Valencia). We thank R. Gil and
K. Oakeson for omp sequence analysis; V. E. Shevchik,
G. Condemine, and M. Lemaire for supplying E. coli
and S. cerevisiae strains; W. J. Miller and B. Loppin for
critical reading of the manuscript; and V. James for
correction of English. Electron microscopy was carried
out in Centre Technologique des Microstructures (UCBL).
This paper is an homage to the work of Paul Nardon.
Supporting Online Material
www.sciencemag.org/cgi/content/full/334/6054/362/DC1
Materials and Methods
Figs. S1 to S9
Tables S1 to S6
References (25, 26)
13 June 2011; accepted 7 September 2011
10.1126/science.1209728
Fig. 3. Effects of colA inhibition with RNAi on SPE size and location. (A) Typical forward scatterarea/
side scatterarea plots showing size ( x axis) and granularity (y axis) of SPE isolated from larvae injected
with dsRNA-gfp (left) and with dsRNA-colA (right). Three SPE populations were defined arbitrarily.
Small-sized cells (P1) and intermediate-sized cells (P2) significantly increased with dsRNA-colA treatment,
whereas the population of large-sized cells (P3) decreased (c
2
-test, P < 0.0001; see percentage values).
The mean size of P2 and P3 significantly decreased (Mann-Whitney test, P < 0.0001), whereas the mean
size of P1 was equal in dsRNA-gfp and dsRNA-colAinjected larvae (Mann-Whitney test, P = 0.37). (B)
FISH visualization of SPE 9 days (upper panels) and 14 days (lower panels) after larvae were injected with
dsRNA-gfp (left) and dsRNA-colA (right). colA inhibition resulted in SPE escaping the bacteriome (see
arrows). See table S4 for experimental details.
www.sciencemag.org SCIENCE VOL 334 21 OCTOBER 2011 365
REPORTS
Stochastic Pulse Regulation in
Bacterial Stress Response
James C. W. Locke,* Jonathan W. Young,* Michelle Fontes,
Mara Jess Hernndez Jimnez, Michael B. Elowitz
Gene regulatory circuits can use dynamic, and even stochastic, strategies to respond to environmental
conditions. We examined activation of the general stress response mediated by the alternative sigma
factor, s
B
, in individual Bacillus subtilis cells. We observed that energy stress activates s
B
in discrete
stochastic pulses, with increasing levels of stress leading to higher pulse frequencies. By perturbing
and rewiring the endogenous system, we found that this behavior results from three key features of the
s
B
circuit: an ultrasensitive phosphorylation switch; stochasticity (noise), which activates that switch;
and a mixed (positive and negative) transcriptional feedback, which can both amplify a pulse and switch
it off. Together, these results show how prokaryotes encode signals using stochastic pulse frequency
modulation through a compact regulatory architecture.
G
ene circuits exhibit fluctuations (noise)
in the concentrations of key components
such as transcriptional regulatory pro-
teins (1, 2). Increasingly, noise appears to play
functional roles in some systems (35). For ex-
ample, noise could enable a subpopulation of
cells to enter a transient antibiotic-resistant state,
enhancing their survival (6). However, it remains
unclear how genetic circuits use noise to control
cellular behaviors. To address this issue, we an-
alyzed s
B
, the transcriptional regulator of general
stress response in Bacillus subtilis, at the single-
cell level (79). Here, we show how s
B
controls
its target genes through sustained pulsing, rather
than continuous activation; how noise enables
this behavior; and how stress levels modulate the
frequency of these pulses (Fig. 1A).
In prokaryotes, alternative sigma factors form
a part of the RNA polymerase holoenzyme, di-
recting it to regulons that control distinct regu-
latory programs (10). s
B
is found in Gram-positive
bacteria and impacts pathogenicity in Listeria
monocytogenes and Staphyloccous aureus (11, 12).
In B. subtilis, s
B
activates more than 150 target
genes in response to diverse stresses (7, 13). s
B
is kept inactive by its anti-sigma factor RsbW
and is activated by the anti-anti-sigma factor
RsbV, which can be reversibly phosphorylated
(see Fig. 1B for regulatory interactions). To
analyze s
B
activation dynamics, we constructed
reporter strains incorporating a yellow fluores-
cent reporter (yfp) for s
B
activity (Fig. 1B) and
used quantitative time-lapse microscopy to follow
s
B
activation in individual cells (14). To quantify
s
B
activity in movies, we computed the P
sigB
promoter activity, defined as the rate of produc-
tion of yellowfluorescent protein (YFP) [fig. S1A
and supporting online material (SOM)].
We first measured the response of s
B
to my-
cophenolic acid (MPA), an energy stress trans-
duced by RsbQP (fig. S2) (15). Constant MPA
led to pulses of s
B
activation in individual cells
(Fig. 1, C and D, and movie S1). These pulses
were unsynchronized across the population, spo-
radic in time, and sustained, continuing through-
out the movie (about six generations) (fig. S3).
Similar behavior was observed with other energy
stresses and during growth in liquid culture (figs.
S1 and S4). Pulses reflected changes in s
B
ac-
tivity and not intrinsic variability of the P
sigB
-YFP
promoter (fig. S5). Increasing MPA concentra-
tion caused a strong increase in pulse frequency,
with weaker increases in mean pulse amplitude
and duration (Fig. 1E and fig. S6), showing that
s
B
is regulated predominantly by frequency mod-
ulation (FM) in response to energy stress (Fig. 1A).
Pulse amplitudes exhibited broad and monoton-
ically increasing variability with increasing MPA,
Howard Hughes Medical Institute, Division of Biology and Bio-
engineering, Broad Center, California Institute of Technology,
1200 East California Boulevard, Pasadena, CA 91125, USA.
*These authors contributed equally to this work.
To whom correspondence should be addressed. E-mail:
melowitz@caltech.edu
Fig. 1. Energy stress modulates the frequency of stochastic pulses of s
B
ac-
tivation. (A) Schematic of FM pulse regulation. The input signal (black line)
controls the frequency of stochastic pulses (blue line, schematic). (B) Sche-
matic diagram of s
B
regulatory interactions and states (7). When RsbV (V) is
phosphorylated (OFF state), s
B
is sequestered by RsbW (W) and inactive (28).
Under energy stresses such as MPA, RsbV is dephosphorylated by the RsbQP
phosphatase complex (QP) (29). Other stress inputs are mediated by the RsbTU
phosphatase complex (not shown; see SOM text for discussion). Dephosphoryl-
ated RsbV can bind to RsbW, releasing s
B
to activate target genes, including
its own operon (30), and the yfp reporter (yellow). (C) Promoter activity of the
P
sigB
-YFP reporter pulses in individual lineages (colored solid lines), and its
mean and standard deviation across all lineages in four data sets (dashed line
and shaded area, respectively). (D) Filmstrip of s
B
activation at 60 mg/mL MPA.
Heterogeneous expression levels of P
sigB
-YFP reflect pulsing activity. (E) MPA
concentration strongly modulates the mean frequency, while more weakly mod-
ulating the mean amplitude and duration, of pulses. Error bars, mean T SEM. (F)
Pulse amplitude histograms for varying levels of MPA. In (E) and (F), each data
point represents data from four microcolonies, acquired on two different days.
21 OCTOBER 2011 VOL 334 SCIENCE www.sciencemag.org 366
REPORTS
with coefficients of variation ranging from 0.16
to 0.70 (Fig. 1F). Together, these results provoke
the question of how FM pulse regulation is im-
plemented by the s
B
regulation circuit.
In principle, pulses could be generated in
two qualitatively distinct ways: They could arise
through amplification of an inherently stochastic
underlying process (16, 17). Alternatively, they
could result from a limit cycle oscillator whose
dynamics become erratic due to noise (18). Sys-
tematically reducing cellular noise would distin-
guish between these possibilities, eliminating pulses
in the first case but making them more regular in
the second case.
To modulate the amplitude of noise in cells,
we created strains that could be induced to grow
into long multinucleoid filaments by controlling
expression of FtsW, a cell division protein neces-
sary for septation. These cells exhibit similar mean
expression levels of cellular components (fig. S7)
but reduced fluctuations (19), allowing us to in-
vestigate how reduced noise affects pulse fre-
quency. In time-lapse movies, we observed a
systematic decline in pulse frequency in noise-
reduced (long) cells across a range of energy stress
levels (Fig. 2A and fig. S8), which was also con-
sistent with reduced cell-cell variability in s
B
ac-
tivity in liquid conditions (Fig. 2B). This reduction
in pulse frequency did not reflect reduced sensi-
tivity to MPA, which had a similar effect on growth
rate in long and short cells (fig. S8). Together, these
results rule out limit cycle models and suggest a
noise-dependent mechanism for pulse generation.
How, then, does the s
B
circuit amplify noise
to initiate discrete pulses of s
B
activity? To ad-
dress this question, we analyzed the response of
s
B
to increased expression of each circuit com-
ponent. Up-regulation of kinase (RsbW) and phos-
phatase (RsbQP) expression had much stronger
(opposite) effects on s
B
activity compared with
up-regulation of RsbV (fig. S9). This result is in-
teresting because opposing kinase and phos-
phatase activities can generate sharp, switchlike
responses in the phosphorylation of their sub-
strate. An extreme example is zero-order ultrasen-
sitivity, where the phosphatase and kinase operate
at saturation (2022). We found that s
B
activity
exhibited an ultrasensitive response to inducible
phosphatase concentration (Fig. 2Cand fig. S10),
with an effective Hill coefficient of 2.12 [95%
confidence interval (CI), n
H
= 2.09 to 2.15]. Sim-
ilar results were also observed with the RsbTU
phosphatase (fig. S11). Moreover, this ultrasen-
sitivity was not due to the transcriptional feedback
loop. It could be observed in an open-loop
strain, in which operon expression was inducible
and independent of s
B
(fig. S12). In this strain,
increasing operon expression led to increasing
ultrasensitivity to phosphatase level. These ef-
fects could be explained by a minimal mathemat-
ical model of the phosphoswitch that does not
include the detailed dynamics of the network (see
SOM). Finally, consistent with this model, ectopic
expression of the kinase RsbWshifted the switch-
ing point to higher phosphatase expression levels
(fig. S13). This ultrasensitive phosphoswitch could
Fig. 2. Pulsing is noise-dependent and involves an ultrasensitive phospho-
switch. (A) Pulse frequency in long cells (green) is strongly reduced com-
pared with short cells (gray; data replotted fromFig. 1E). Error bars, mean T SEM.
(B) Variability in P
sigB
-YFP expression decreases with increasing cell length
(see fig. S8). Equal numbers of cells (represented by dots) are plotted in each
log-spaced bin (delimited by gray vertical lines). (Inset) Overlay of phase
contrast and P
sigB
-YFP expression (green) at different cell lengths. Note
greater s
B
variability in short cells. (C) s
B
expression is ultrasensitive to
RsbQP phosphatase levels. Each dot represents the mean RsbQP-YFP level and
P
sigB
-CFP level of one cell, using the strain shown schematically (table S1). The
red line is a Hill function with Hill coefficient n
H
= 2.12 (95% CI, n
H
= 2.09 to
2.15).
Fig. 3. A mixed transcriptional feedback loop amplifies and terminates
pulses. (A) Schematic diagram of supra- and subthreshold protocols. Before
time-lapse acquisition (gray region), phosphatase is induced to a constant
level by addition of xylose. After the start of acquisition, isopropyl-b-D-
thiogalactopyranoside (IPTG) is added to induce rsbVWB to levels greater than
(solid red line) or less than (dashed red line) the level of phosphatase. This
results in pulsed (solid green line) or sustained (dashed green line) s
B
activity
dynamics. (B) s
B
promoter activity exhibits a transition between sub- and
suprathreshold behaviors. Each trace shows the mean P
sigB
promoter activity
averaged over four colonies. The promoter activity of the IPTG-inducible s
B
operon ( x axis) was estimated using a separate strain containing a similar
IPTG-inducible yfp reporter. Two repeat movies showed similar behaviors.
(Inset) Schematic diagram of strain used for this experiment (table S1). (C) A
minimal mathematical model of the open-loop s
B
network reproduces the main
features of the experimental data. (Inset) In this model, the unphosphorylated
activator, A, directly activates target genes (see SOM).
www.sciencemag.org SCIENCE VOL 334 21 OCTOBER 2011 367
REPORTS
activate s
B
in response to fluctuations in the phos-
phatase/kinase ratio and thereby initiate pulses.
How are s
B
pulses further amplified and sub-
sequently terminated? s
B
activates its own ope-
ron (rsbV-rsbW-sigB). This feedback loop could
increase s
B
activity, due to the activating effects
of RsbV and s
B
, or it could repress s
B
activity,
due to increased production of RsbW. We hy-
pothesized that the phosphoswitch sets a thresh-
old between activating (phosphatase dominant)
and repressing (kinase dominant) feedback re-
gimes. As long as phosphatase activity exceeds
kinase activity, activation of the operon increases
free s
B
(positive feedback). However, this also in-
creases production of RsbW kinase. When kinase
activity approaches that of the phosphatase, in-
creased operon expression will cause RsbWlevels
to cross the threshold, shutting off activation
(negative feedback). Thus, autoregulation could
result in a mixed (positive and negative) feed-
back loop, providing a compact mechanism to
first amplify and then terminate a pulse (23).
To test this hypothesis, we constructed an
open-loop strain and quantified the change in s
B
activity in response to a step increase in operon
expression. In these experiments, we first estab-
lished a basal level of phosphatase activity in
cells and subsequently induced a step, of varying
size, in s
B
operon expression (Fig. 3A). We ob-
served a striking transition between two qualita-
tively different responses: At lower operon induction
levels, the system produced a sustained response,
whereas at higher induction levels, it exhibited a
pulse (Fig. 3B). These results are consistent with
the mixed feedback model: Initially, increased
operon induction produces more s
B
, which is
active due to the high levels of phosphatase, en-
gaging the positive feedback loop. For lower
(subthreshold) operon induction levels, RsbWlev-
els never exceed phosphatase levels, so the sys-
temremains on indefinitely (Fig. 3A, dashed lines).
In contrast, at higher induction levels (superthresh-
old), RsbWactivity eventually crosses the thresh-
old set by the phosphatase and thereby shuts the
system off, resulting in a pulse (Fig. 3A, solid
lines). Indeed, RsbW dominated other operon
components at steady state, suppressing s
B
ac-
tivity in a dose-dependent manner (fig. S9). In
this mechanism, pulse amplitude should be rough-
ly proportional to the difference between the phos-
phatase level and the kinase level immediately
after the initiating fluctuation, as confirmed ex-
perimentally (fig. S14). A minimal mathematical
model of the circuit exhibited qualitatively simi-
lar behavior (Fig. 3C). This mechanism for pulse
initiation, amplification, and termination is sum-
marized in Fig. 4, A and B.
These results provoke a final question: How
can the cell modulate the frequency of pulses in
this system? Systematic changes in the activity of
either kinase or phosphatase could modulate the
likelihood of threshold-crossing events and there-
by control pulse frequency. We first examined the
distribution of RsbQP expression levels, using a
P
rsbQP
-rsbQP-YFP protein fusion that complements
an RsbQP null mutant (fig. S15). In response to
40 mg/ml MPA, we observed a ~3-fold increase
in mean RsbQP-YFP levels, and a ~6-fold in-
crease in mean s
B
activity (fig. S16, A and B).
At the single-cell level, RsbQP-YFP expression
mapped to s
B
activity (fig. S16C), closely fol-
lowing the (independently determined) ultrasen-
sitive response function (Fig. 2C). These results
suggest that stress increases s
B
pulse frequency
by increasing the distribution of RsbQP expres-
sion levels and thereby increasing the frequency
with which RsbQP fluctuations cross the thresh-
old set by RsbW (Fig. 4C). These results do not
rule out the complementary possibility, suggested
previously (24), that some energy stresses may
activate s
B
by reducing kinase activity.
To show that this mechanism is indeed suf-
ficient to enable frequency modulation, we re-
wired the endogenous circuit, replacing RsbQP
with an inducible, constitutively active RsbTU
phosphatase complex that was unaffected by en-
ergy stress (Fig. 4D, inset). The rewired system
exhibited stochastic pulsing in response to RsbTU
expression (figs. S17 and S18 and movie S2).
Furthermore, we observed an increased frequen-
cy of pulsing in response to increased RsbTU
phosphatase expression (Fig. 4D), with weaker
effects on pulse amplitude and duration (fig. S20).
These results, qualitatively similar to those ob-
served in wild-type cells under energy stress, also
match an extended mathematical model that in-
cludes the wild-type transcriptional feedback
(gray dashed lines in Fig. 4D, figs. S19 and S20)
(see SOM). Thus, modulation of phosphatase ex-
pression is sufficient to recapitulate FM pulsing,
and no special property of the RsbQP phospha-
tase is required.
Fig. 4. Mechanism of FM pulse con-
trol. (A) Schematic time course of phos-
phatase RsbQP (denoted P, purple),
free s
B
(s
B
, green), and kinase (W,
red) during a pulse cycle. Circled num-
bers indicate specific steps in (B). (B)
Schematic diagram of pulse control.
The relative concentration of each com-
ponent is indicated by size. (1) Initial
state: System components are at low
levels, kinase activities exceed phos-
phatase activities, and therefore RsbV
is mostly phosphorylated. A threshold-
crossing upward fluctuation in RsbQP
level dephosphorylates V
P
, leading to
(2) Pulse Initiation. Activation of s
B
(indicated by glowing halo) leads to
up-regulation of operon components
(operon feedback). (3) Pulse peak: s
B
activity peaks just before RsbWkinase
activity exceeds phosphatase activity.
(4) Termination: Rephosphorylation
of RsbV shuts the system off. (5) Di-
lution: Component levels reset to the
original state. (C) Mechanismof frequen-
cy modulation. Fluctuations in phos-
phatase level (purple arrow from state 1 to 2) can cross the kinase threshold
(red line) to initiate a pulse, with amplitude determined by the size of fluc-
tuation (dashed line). Increased stress shifts the distribution of phosphatase
levels from lower to higher values (dark and light gray, respectively), in-
creasing the frequency of threshold-crossing events and thereby increasing
pulse frequency (inset). (D) Tuning of phosphatase expression by IPTG (strain
indicated schematically in inset) can regulate pulse frequency. Gray dashed
lines show a similar behavior for the mathematical model (fit to data). Each
data point represents statistics fromtwo colonies. Two repeat data sets showed
similar trends.
21 OCTOBER 2011 VOL 334 SCIENCE www.sciencemag.org 368
REPORTS
FM pulsing can be implemented by a simple
circuit of three genes (rsbW, rsbV, and sigB), with
input from a phosphatase complex. This system
provides a fundamental signal-processing capabil-
ity to bacterial cells, enabling them to convert
steady DC inputs into pulsatile, predominantly
AC outputs. Noise plays a key functional role
in this signal processing system (3). The s
B
cir-
cuit conserves its core architecture in diverse
bacteria (7), and other alternative sigma factors
similarly feature both posttranslational regulation
by anti-sigma factors and autoregulatory feed-
back. Thus, related stochastic pulse modulation
schemes are likely employed more generally in
bacteria (10). The relatively slowtime scale of s
B
pulses (Fig. 1E) could confer advantages in re-
sponding to unpredictable environments and main-
taining a broad, but dynamic, distribution of states
in the population through bet-hedging (25, 26).
Given the negative effect of s
B
activation on
growth rate in some conditions, even under energy
stress (27), these results suggest that cells balance
the benefits and costs of s
B
activation dynami-
cally. It will be interesting to see whether other
dynamic encoding schemes are similarly imple-
mented by relatively simple circuit modules.
References and Notes
1. A. Raj, A. van Oudenaarden, Cell 135, 216 (2008).
2. N. Rosenfeld, J. W. Young, U. Alon, P. S. Swain,
M. B. Elowitz, Science 307, 1962 (2005).
3. A. Eldar, M. B. Elowitz, Nature 467, 167 (2010).
4. R. Losick, C. Desplan, Science 320, 65 (2008).
5. L. Cai, C. K. Dalal, M. B. Elowitz, Nature 455, 485
(2008).
6. E. Rotem et al., Proc. Natl. Acad. Sci. U.S.A. 107,
12541 (2010).
7. M. Hecker, J. Pan-Farr, U. Vlker, Annu. Rev. Microbiol.
61, 215 (2007).
8. W. G. Haldenwang, R. Losick, Nature 282, 256
(1979).
9. O. A. Igoshin, M. S. Brody, C. W. Price, M. A. Savageau,
J. Mol. Biol. 369, 1333 (2007).
10. T. M. Gruber, C. A. Gross, Annu. Rev. Microbiol. 57,
441 (2003).
11. M. J. Kazmierczak, S. C. Mithoe, K. J. Boor, M. Wiedmann,
J. Bacteriol. 185, 5722 (2003).
12. U. Lorenz et al., Microbes Infect. 10, 217 (2008).
13. M. Hecker, U. Vlker, Mol. Microbiol. 29, 1129
(1998).
14. J. C. Locke, M. B. Elowitz, Nat. Rev. Microbiol. 7, 383
(2009).
15. S. Zhang, W. G. Haldenwang, J. Bacteriol. 187, 7554
(2005).
16. A. L. Hodgkin, A. F. Huxley, J. Physiol. 117, 500 (1952).
17. G. M. Sel, J. Garcia-Ojalvo, L. M. Liberman,
M. B. Elowitz, Nature 440, 545 (2006).
18. M. B. Elowitz, S. Leibler, Nature 403, 335 (2000).
19. G. M. Sel, R. P. Kulkarni, J. Dworkin, J. Garcia-Ojalvo,
M. B. Elowitz, Science 315, 1716 (2007).
20. A. Goldbeter, D. E. Koshland Jr., Proc. Natl. Acad.
Sci. U.S.A. 78, 6840 (1981).
21. G. J. Melen, S. Levy, N. Barkai, B. Z. Shilo, Mol. Syst. Biol.
1, 2005.0028 (2005).
22. Z. Cheng, F. Liu, X. P. Zhang, W. Wang, Biophys. J.
97, 2867 (2009).
23. J. C. Ray, O. A. Igoshin, PLOS Comput. Biol. 6, e1000676
(2010).
24. S. Alper, L. Duncan, R. Losick, Cell 77, 195
(1994).
25. M. Acar, A. Becskei, A. van Oudenaarden, Nature 435,
228 (2005).
26. E. Kussell, S. Leibler, Science 309, 2075 (2005).
27. T. Schweder, A. Kolyschkow, U. Vlker, M. Hecker,
Arch. Microbiol. 171, 439 (1999).
28. A. Dufour, W. G. Haldenwang, J. Bacteriol. 176, 1813
(1994).
29. M. S. Brody, K. Vijay, C. W. Price, J. Bacteriol. 183,
6422 (2001).
30. A. A. Wise, C. W. Price, J. Bacteriol. 177, 123
(1995).
Acknowledgments: We thank C. Price and D. Rudner for
providing strains. We thank A. Eldar, R. Kishony, C. Price,
N. Wingreen, J. Levine, and other members of M.B.Es
laboratory for helpful discussions. Work in M.B.Es
laboratory was supported by NIH grants R01GM079771
and P50 GM068763, U.S. National Science Foundation
CAREER Award 0644463, and the Packard Foundation.
J.C.W.L was supported by the International Human
Frontier Science Program Organization and the European
Molecular Biology Organization.
Supporting Online Material
www.sciencemag.org/cgi/content/full/science.1208144/DC1
Materials and Methods
SOM Text
Figs. S1 to S20
Table S1
References
Movies S1 and S2
10 May 2011; accepted 1 September 2011
Published online 6 October 2011;
10.1126/science.1208144
Transgenerational Epigenetic
Instability Is a Source of
Novel Methylation Variants
Robert J. Schmitz,
1,2
Matthew D. Schultz,
1,2,3
Mathew G. Lewsey,
1,2
Ronan C. OMalley,
2
Mark A. Urich,
1,2
Ondrej Libiger,
4
Nicholas J. Schork,
4
Joseph R. Ecker
1,2,5
*
Epigenetic information, which may affect an organisms phenotype, can be stored and stably
inherited in the form of cytosine DNA methylation. Changes in DNA methylation can produce
meiotically stable epialleles that affect transcription and morphology, but the rates of spontaneous
gain or loss of DNA methylation are unknown. We examined spontaneously occurring variation
in DNA methylation in Arabidopsis thaliana plants propagated by single-seed descent for 30
generations. We identified 114,287 CG single methylation polymorphisms and 2485 CG differentially
methylated regions (DMRs), both of which show patterns of divergence compared with the ancestral
state. Thus, transgenerational epigenetic variation in DNA methylation may generate new allelic
states that alter transcription, providing a mechanism for phenotypic diversity in the absence of
genetic mutation.
C
ytosine methylation is a DNA base mod-
ification with roles in development and
disease in animals as well as in silencing
transposons and repetitive sequences in plants
and fungi (1). In plants, CG methylation is com-
monly found within gene bodies (25), whereas
non-CG methylation, CHG and CHH (where H
is A, C, or T), is enriched in transposons and re-
petitive sequences (1). The RNA-directed DNA
methylation (RdDM) pathway targets both CG
and non-CG sites for methylation and is com-
monly associated with transcriptional silencing
(6). This pathway can also target and silence
protein-coding genes, giving rise to epigenetic al-
leles or so-called epialleles that can be heritable
through mitosis and/or meiosis (7, 8) and can be
dependent on the methylation of a single CG di-
nucleotide (9).
Two meiotically heritable epialleles result-
ing in morphological variation are the peloric
(Linaria vulgaris) and colorless non-ripening
(Solanum lycopersicum) loci (10, 11). Both show
spontaneous epigenetic silencing events within
their respective populations (10, 12). However, the
frequency at which such spontaneous meiotically
heritable epialleles naturally arise in populations
is unknown. Although epiallelic variation has been
identified between genetically diverse populations
within Arabidopsis thaliana (13), it is unclear
whether these identified epialleles are due to
underlying genetic variation. Epialleles have al-
so been artificially generated after mutagenesis
or because of mutations in the cellular com-
ponents required for the maintenance of DNA
methylation (1416).
An A. thaliana (Columbia-0) population, the
MA lines, derived by single-seed descent for 30
generations (17) was used to examine the extent
of naturally occurring variation in DNA methyla-
tion and the frequency at which spontaneous epi-
alleles emerge over time. We used the MethylC-Seq
method (3) to determine the whole-genome base
resolution DNA methylomes for three ancestral
1
Plant Biology Laboratory, The Salk Institute for Biological
Studies, La Jolla, CA 92037, USA.
2
Genomic Analysis Labo-
ratory, The Salk Institute for Biological Studies, La Jolla, CA
92037, USA.
3
Bioinformatics Program, University of California
at San Diego, La Jolla, CA 92093, USA.
4
The Scripps Transla-
tional Science Institute and the Department of Molecular and
Experimental Medicine, The Scripps Research Institute, La
Jolla, CA 92037, USA.
5
Howard Hughes Medical Institute, The
Salk Institute for Biological Studies, 10010 North Torrey Pines
Road, La Jolla, CA 92037, USA.
*To whom the correspondence should be addressed. E-mail:
ecker@salk.edu
www.sciencemag.org SCIENCE VOL 334 21 OCTOBER 2011 369
REPORTS
MA lines (numbers 1, 12, and 19) and five
descendant MA lines (numbers 29, 49, 59, 69,
and 119) (fig. S1). We refer to lines 1, 12, and 19
as ancestors throughout this study, although
they are not direct ancestors because they are
three generations removed from the original
founder line (fig. S1). These specific descendant
lines were selected because their genomes have
been sequenced and they have a known level of
spontaneous mutation (18). Biological replicates
(sibling plants) for each leaf methylome were
sequenced to an average of ~34-fold coverage,
which allowed for an average per line exami-
nation of 39,897,093 (96.35%) uniquely mapped
cytosines and 5,307,077 (98.39%) uniquely mapped
CGs (table S1).
A total of 1,730,761 CGs were methylated
(mCGs) in at least one MA line (Fig. 1A), and
about 91% of the covered mCGs were invar-
iably methylated across all eight lines (19). The
variable mCGs revealed a set of 114,287 high-
confidence CG single methylation polymorphisms
(SMPs) that showed a consensus of the meth-
ylation status of CG dinucleotides between bi-
ological replicates (Fig. 1A). Next, a reference
MA founder DNA methylome was created by
pooling the completely conserved mCG site
calls for all ancestral MA lines and used to de-
termine the frequency of discordant CG-SMP
sites within the descendant population (Fig. 1B).
Within the descendant lines, ~1.62% of the CG
methylome shows susceptibility to dynamic ac-
quisitions and losses of mCGs over time (table
S2). On average, ~66,000 methylated CG-SMPs
(mCG-SMPs) were identified for each ances-
tral and descendant line (fig. S2). Although the
total number of mCG-SMPs was similar be-
tween all lines, the conservation of these poly-
morphisms among and between ancestral and
descendant populations was different (Fig. 1C
and table S3). A pairwise comparison of both
populations for methylation conservation, esti-
mated by global similarity of mCG-SMP sites
(19), revealed that all of the ancestral lines are
highly similar (table S4). Descendant lines showed
greater similarity in CG-SMPs methylation sta-
tus to ancestral lines than to other descendant
lines (table S4).
We calculated an estimate of the epimutation
rate per generation in this population by using
linear regression and TREE PUZZLE, which re-
vealed 704 and 2876 methylation changes each
generation, respectively (19). We estimated a lower
bound of the epimutation rate with the linear
regression results, which revealed 4.46 10
4
methylation polymorphisms per CG site per gen-
eration (P < 0.0000216) (table S5). This finding
contrasts with the previously reported spontane-
ous genetic mutation rate of 7 10
9
base sub-
stitutions per site per generation for these same
MA lines (18). The TREE PUZZLE analysis re-
vealed higher estimated epimutation rates in earlier
generations (19). One possible source of this var-
iation could be due to seed age, storage, and/or
selection for seed survival. Therefore, although
DNA methylation is predominantly static over
relatively long periods of time, changes in cyto-
sine methylation do occur and at a frequency
greater than that of mutation observed at the
DNA sequence level.
By using CG-SMPs derived from both an-
cestral and descendant populations, we carried
out a genome-wide analysis of differentially meth-
ylated regions (DMRs) and identified 2485 CG-
DMRs that ranged in size from 11 to 1110 base
pairs (bp) (Fig. 2A and table S6). Hierarchical
clustering of CG-DMRs in this population, cal-
culated solely on the basis of the methylation
density, revealed that the ancestral lines segregate
as an independent cluster from the descendant
lines (Fig. 2B and fig. S3). Multivariate distance-
based regression (MDMR) (20, 21) confirmed
this finding, indicating a statistically significant
(P < 0.00005) association between ancestor or
descendant status and methylation density of the
CG-DMR profiles. The ancestor or descendant
status explained 47% of the variance in the dis-
similarity in methylation density of CG-DMRs
between pairs of samples, indicating that, over
time, there is a divergence of DNA methylation
patterns in both formation and elimination of CG-
DMRs. Furthermore, the genome-wide locations
of these CG-DMRs were not uniformly distributed
(P < 2.20 10
16
), because 60.5% (1504/2485)
were found in genic regions compared with 3.3%
(82/2485) and 36.2% (899/2485) located in in-
tergenic regions and transposons, respectively
(Fig. 2B).
Next, we performed a genome-wide survey
for nonCG-DMRs and uncovered a total of 284
among all eight lines (table S7). In general, the
nonCG-DMRs were largely localized to inter-
genic regions (141/284) of the genome, because
only 57/284 overlapped with genes and 86/284
overlapped with transposons. The size ranges of
the nonCG-DMRs were similar to those of the
CG-DMRs because the vast majority occurred in
smaller segments of the genome (10 to 682 bp).
Therefore, variation in DNA methylation ap-
pears to occur in all three methylation sequence
contexts.
CG methylation is present within gene bodies
and is enriched toward the 3 end (25), whereas
CG and nonCG methylation is associated with
heterochromatin, transposons, and repetitive se-
quences (1). In agreement with these findings, we
observed that the 3 portion of genes contained
the greatest source of CG-DMRs and that the
majority of nonCG-DMRs were enriched out-
side of the gene bodies (Fig. 2C). Furthermore,
we observed a ~twofold depletion of CG-DMRs
in exons compared with introns (Fig. 2D). The
genome-wide distributions of CG-SMPs, CG-DMRs,
Fig. 1. Epigenetic variation of CG-SMPs. (A) An example of a CG-SMP. Gold lines indicate CG methyl-
ation, maroon rectangle indicates the untranslated regions, and green rectangles indicated exons. (B) A
breakdown of the methylation distribution of CG dinucleotides among all samples. (C) A heatmap indi-
cating the number of CG-SMPs that differ between two samples (table S3).
21 OCTOBER 2011 VOL 334 SCIENCE www.sciencemag.org 370
REPORTS
and nonCG-DMRs were depleted in hetero-
chromatic regions in the genome (Fig. 2, E
and F). These depletions were mostly observed at
the pericentromeres and centromeres (Fig. 2, E
and F, and figs. S4 and S5). CG-DMRs are
enriched in transposons located in euchromatin
but depleted in transposons present near the
centromere. Because the centromeric regions of
the genome contain the highest density of DNA
methylation (Fig. 2, E and F), these observations
combined with the observations that CG-DMRs
are enriched in intron sequences may indicate
that DNA methylation that is associated with nu-
cleosomes (22) (i.e., exons or tightly packaged
chromatin in the pericentromeres and centro-
meres) may be maintained at a higher fidelity
and that DNA methylation not associated with
nucleosomes may undergo greater epigenetic
drift.
A genome-wide screen for DMRs simulta-
neously occurring in all three methylation sequence
contexts (C-DMRs are CG, CHG, and CHH)
was performed to assess the extent of epiallelic
variation that is characteristic of RdDM across
the MA population. In total, 72 C-DMRs were
identified, of which functional categorization
revealed that two-thirds overlapped with trans-
poson and intergenic sequences whereas about
one-third overlapped with gene bodies and pro-
moters (Fig. 3A and table S8). To determine
whether transposition-induced methylation could
potentially give rise to the methylated C-DMRs
(mC-DMRs) (23), genomic DNA encompassing
all C-DMRs was amplified and compared in all
ancestral and descendant lines. In every case,
the observed amplicon size was identical for
all MA lines and was equal to the expected size
of the locus (table S8), indicating that these
C-DMRs are unlinked to cis-genetic variation
located within 500 bp, a distance that would be
expected to reveal methylation induced by trans-
poson insertions at these loci (23). Additionally,
none of the genetic variants identified by genome
resequencing of this population (18) overlapped
with any of these C-DMRs. Lastly, restriction
enzyme digestion and Southern blot analyses
were performed to rule out the possibility that
copy number variants were the cause of spon-
taneous epiallele formation, as is the case for the
PAI epialleles (24). In all cases examined, the ob-
served hybridization pattern and gene copy num-
ber were identical for each of the MA lines
(fig. S6). Therefore, we conclude that the 72
C-DMRs represent a set of spontaneously occur-
ring epialleles within the MA lines, because they
were not associated with any genetic variation.
By using a set of C-DMRs that exhibited
an identical methylation status (fig. S7), we de-
termined the frequency of discordance of the
ancestral state with the descendant lines and found
that 29 of the C-DMRs were highly variable (>1
descendant line was discordant with the ances-
tral state) (Fig. 3B). C-DMRs discordant in only
one of the five descendant lines were the most
frequent class, but there was an unexpectedly
high number of C-DMRs (63%) that were dis-
cordant in more than one descendant (Fig. 3B).
Within the set of 576 C-DMRs identified (eight
lines by 72 C-DMRs), 7 were discordant between
the biological replicates (table S8). These data
suggest that, although many C-DMRs represent
the formation of spontaneous epialleles, a small
subset may reflect the presence of hotspots
(metastable epialleles).
We sequenced small RNA (smRNA) pop-
ulations for all eight lines and found that
smRNAs [represented as RPKCMs (reads per
kilobase of each C-DMR per million reads) in
Fig. 2. CG-DMRs diverge over time and are enriched in
gene bodies. (A) Example CG-DMR present in an unmeth-
ylated state in both replicates of line 69. (B) A heatmap
representation of a two-dimensional hierarchical cluster-
ing based on DMRs. Columns represent samples. Rows
indicate DMRs. The column to the left of the heatmap
indicates the genomic location of the DMR (blue, gene
body; gold, transposon; gray, intergenic; red, transposon
in gene body). (C) The average distribution of CG-DMRs
(red) and nonCG-DMRs (blue) across gene bodies (from the start of the 5 UTR to the end of the 3 UTR, including 500 bp up- and downstream). (D) CG gene-
body DMRs are specifically depleted in exons. (E) Genome-wide distributions of mCG (red), CG-SMPs (green), and CG-DMRs (blue) across chromosome I. (F)
Genome-wide distributions of methylated nonCGs (mnonCG, red) and nonCG-DMRs (green) across chromosome I. The centromere is indicated by the pink
vertical bar for (E) and (F).
www.sciencemag.org SCIENCE VOL 334 21 OCTOBER 2011 371
REPORTS
tables S9 to 12] were associated with an in-
crease in the average methylation density of
C-DMRs (Fig. 3C). Furthermore, this association
resembled a binary switch, because the most
densely methylated C-DMRs contained abun-
dant 24-nucleotide (nt) smRNAs (Fig. 3C).
Of the eight previously documented plant
epialleles resulting in phenotypic variation, all af-
fected transcriptional output of the differentially
methylated locus (911, 2328). mRNA abun-
dance was measured in all eight lines with quan-
titative reverse transcription polymerase chain
reaction (qRT-PCR) at eight C-DMRs that over-
lapped with protein-coding regions. In four of
these genes, the gain or loss of DNA methylation
was correlated with a large decrease or increase
in mRNA abundance, respectively, and with the
presence of 24-nt smRNAs at each silenced epi-
allele (Fig. 3, D to F, and fig. S8). These find-
ings reveal that changes in epiallelic state can
lead to major effects on transcriptional output
(fig. S9).
We also observed that the methylation sta-
tus of one C-DMR resulted in alternative pro-
moter usage of ACTIN RELATED PROTEIN 9
(At5g43500) (fig. S10C). The loss of DNA meth-
ylation within the 5 untranslated region (UTR)
of the At5g43500.1 isoform led to an increase in
mRNA expression, whereas expression of iso-
form At5g43500.2, with a transcriptional start
site located further downstream, was unaffected
(fig. S10, D and E).
Although epialleles can have major impacts
on phenotypic diversity, until now their identi-
fication was not trivial. Even more puzzling is
the origin of pure alleles, which are defined
by their formation in the absence of any genetic
variation in cis or trans (8). One route to epi-
allele formation may be the failure to correctly
maintain the proper methylation status through-
-1 0 1 2 3 4 5 6 7 8
A B
Transposons
n = 27
Intergenic
n = 21
Genes
n = 14
Promoters
n = 7
ncRNAs
n = 2
Pseudogene
n = 1
# of descendant lines
discordant with ancestral state
0
2
4
6
8
10
12
14
16
18
1 2 3 4 5
Methylated Unmethylated
D E
1 rep1
1 rep2
19 rep1
19 rep2
12 rep1
12 rep2
29 rep1
29 rep2
49 rep1
49 rep2
59 rep1
59 rep2
69 rep1
69 rep2
119 rep1
119 rep2
A
n
c
e
s
t
o
r
s
D
e
s
c
e
n
d
a
n
t
s
Log2 fold change in mRNA levels
of At5g24240 (relative to line 1)
At5g24240
At5g24250
1
19
12
29
49
59
69
119
N
u
m
b
e
r

o
f

C
-
D
M
R
s
C
F
24nt
23nt
22nt
21nt
1
19
12
29
49
59
69
119
smRNA levels at At5g24240 C-DMR (RPKCMs)
0 2 4 6 8 10 12 14 16
mC-DMR density quantiles (%)
A
v
e
r
a
g
e

s
m
R
N
A

R
P
K
C
M
s
0
2
4
6
8
10
12
14
10 20 30 40 50 60 70 80 90 100
21nt
22nt
23nt
24nt
Fig. 3. Epiallelic variation at protein-coding loci is associated with transcrip-
tional variation. (A) Classification of C-DMRs and their genomic locations. (B)
The number of descendant lines discordant with the ancestral C-DMR state
and the C-DMR methylation status. The black portions of the bar indicate the
descendant C-DMRs that became methylated, whereas the white portions
indicate regions that became unmethylated, compared with the ancestral pop-
ulation. (C) The 24-nt smRNA levels are associated with increasing methyla-
tion density. The 24-nt smRNA RPKCMs for all 576 C-DMRs (8 MA lines by
72 C-DMRs) were ranked and binned into 10% quantiles, and then the aver-
age mC densities were plotted. (D) A representative C-DMR at At5g24240 in
which both biological replicates of descendant line 59 were unmethylated. (E)
qRT-PCR analysis of At5g24240 reveals >50-fold increase in mRNA abundance
in unmethylated line 59. Error bars indicate SEM. (F) The 24-nt smRNAs are
enriched specifically in the MA lines that are transcriptionally silenced in (E)
for the At5g24240 locus with the exception of line 59, which is abundantly
expressed in (E).
0
10
20
30
40
50
60
met1 ddc
Partially
Methylated
# of mC-DMRs
that become
unmethylated in
# of C-DMRs
that become
re-methylated in
rdd
Not
methylated
in
Col-0
N
u
m
b
e
r

o
f

C
-
D
M
R
s
Fig. 4. Methylation status of all 72 epialleles in methylation and demethylation mutant backgrounds.
Most of the epialleles become unmethylated in met1-3, whereas a smaller number become remeth-
ylated in the DNA demethylase triple mutant rdd.
21 OCTOBER 2011 VOL 334 SCIENCE www.sciencemag.org 372
REPORTS
out epigenetic reprogramming that occurs post-
fertilization (29, 30). It is noteworthy that 63 of
the 72 C-DMRs overlap with regions previously
shown to have altered methylation patterns in
methylation enzyme mutants (Fig. 4) (3). Of the
14 C-DMRs that overlap with genes, 5 become
reexpressed in met1-3 and 1 transcript becomes
silenced in rdd (3). These results suggest that a
failure to faithfully maintain genome-wide meth-
ylation patterns by MET1 and/or RDD is likely
one source of spontaneous epiallele formation.
Regardless of their origin, the majority of epi-
alleles identified in this study are meiotically sta-
ble and heritable across many generations in this
population. Understanding the basis for such trans-
generational instability and the mechanism(s)
that trigger and/or release these epiallelic states
will be of great importance for future studies.
References and Notes
1. J. A. Law, S. E. Jacobsen, Nat. Rev. Genet. 11, 204 (2010).
2. S. J. Cokus et al., Nature 452, 215 (2008).
3. R. Lister et al., Cell 133, 523 (2008).
4. X. Zhang et al., Cell 126, 1189 (2006).
5. D. Zilberman, M. Gehring, R. K. Tran, T. Ballinger,
S. Henikoff, Nat. Genet. 39, 61 (2007).
6. S. W.-L. Chan et al., Science 303, 1336 (2004).
7. J. Paszkowski, U. Grossniklaus, Curr. Opin. Plant Biol. 14,
195 (2011).
8. E. J. Richards, Nat. Rev. Genet. 7, 395 (2006).
9. K. Shibuya, S. Fukushima, H. Takatsuji, Proc. Natl. Acad.
Sci. U.S.A. 106, 1660 (2009).
10. P. Cubas, C. Vincent, E. Coen, Nature 401, 157 (1999).
11. K. Manning et al., Nat. Genet. 38, 948 (2006).
12. A. J. Thompson et al., Plant Physiol. 120, 383 (1999).
13. M. W. Vaughn et al., PLoS Biol. 5, e174 (2007).
14. F. Johannes et al., PLoS Genet. 5, e1000530 (2009).
15. F. K. Teixeira et al., Science 323, 1600 (2009);
10.1126/science.1165313.
16. A. Vongs, T. Kakutani, R. A. Martienssen, E. J. Richards,
Science 260, 1926 (1993).
17. R. G. Shaw, D. L. Byers, E. Darmo, Genetics 155, 369
(2000).
18. S. Ossowski et al., Science 327, 92 (2010).
19. Additional experiments and descriptions of methods used
to support our conclusions are presented as supporting
material on Science Online.
20. C. M. Nievergelt et al., Am. J. Med. Genet. B. Neuropsychiatr.
Genet. 141B, 234 (2006).
21. M. A. Zapala, N. J. Schork, Proc. Natl. Acad. Sci. U.S.A.
103, 19430 (2006).
22. R. K. Chodavarapu et al., Nature 466, 388 (2010).
23. J. Liu, Y. He, R. Amasino, X. Chen, Genes Dev. 18,
2873 (2004).
24. J. Bender, G. R. Fink, Cell 83, 725 (1995).
25. S. Melquist, B. Luff, J. Bender, Genetics 153, 4017
(1999).
26. S. E. Jacobsen, E. M. Meyerowitz, Science 277, 1100
(1997).
27. H. Saze, T. Kakutani, EMBO J. 26, 3641 (2007).
28. W. J. Soppe et al., Mol. Cell 6, 791 (2000).
29. R. A. Mosher et al., Nature 460, 283 (2009).
30. R. K. Slotkin et al., Cell 136, 461 (2009).
Acknowledgments: We thank M. White, R. Lister, M. Galli,
and R. Amasino for discussions; R. Shaw and E. Darmo
for seeds; J. Nery for sequencing operations; and
M. Axtell for Southern blot protocol. R.J.S. was supported
by an NIH National Research Service Award postdoctoral
fellowship (F32-HG004830). M.D.S. was supported by
a NSF Integrative Graduate Education and Research
Traineeship grant (DGE-0504645). M.G.L. was supported
by an European Union Framework Programme 7
Marie Curie International Outgoing Fellowship
(project 252475). O.L. and N.J.S. are supported by
NIH/National Center for Research Resources grant
number UL1 RR025774. This work was supported by
the Mary K. Chapman Foundation, the NSF (grants
MCB-0929402 and MCB1122246), the Howard Hughes
Medical Institute, and the Gordon and Betty Moore
Foundation (GBMF) to J.R.E. J.R.E. is a HHMIGBMF
Investigator. Analyzed data sets can be viewed
at http://neomorph.salk.edu/30_generations/browser.
html. Sequence data can be downloaded from National
Center for Biotechnology Information Sequence Read
Archive (SRA035939). Correspondence and requests for
materials should be addressed to J.R.E. (ecker@salk.edu).
Supporting Online Material
www.sciencemag.org/cgi/content/full/science.1212959/DC1
Materials and Methods
SOM Text
Figs. S1 to S11
Tables S1 to S16
References
22 August 2011; accepted 7 September 2011
Published online 15 September 2011;
10.1126/science.1212959
Computation-Guided Backbone
Grafting of a Discontinuous Motif
onto a Protein Scaffold
Mihai L. Azoitei,
1
* Bruno E. Correia,
1,2
* Yih-En Andrew Ban,
1
Chris Carrico,
1,3
Oleksandr Kalyuzhniy,
1
Lei Chen,
4
Alexandria Schroeter,
1
Po-Ssu Huang,
1
Jason S. McLellan,
4
Peter D. Kwong,
4
David Baker,
1,5
Roland K. Strong,
3
William R. Schief
1,6,7

The manipulation of protein backbone structure to control interaction and function is a

challenge for protein engineering. We integrated computational design with experimental selection
for grafting the backbone and side chains of a two-segment HIV gp120 epitope, targeted by the
cross-neutralizing antibody b12, onto an unrelated scaffold protein. The final scaffolds bound b12 with
high specificity and with affinity similar to that of gp120, and crystallographic analysis of a scaffold
bound to b12 revealed high structural mimicry of the gp120-b12 complex structure. The method
can be generalized to design other functional proteins through backbone grafting.
C
omputational protein design tests our
understanding of protein structure and
folding and provides valuable reagents
for biomedical and biochemical research; long-
term goals include the design of field- or clinic-
ready biosensors (1), enzymes (2), therapeutics (3),
and vaccines (4, 5). A major limitation has been
an inability to manipulate backbone structure;
most computational protein design has involved
sequence design on predetermined backbone struc-
tures or with minor backbone movement (15).
Accurate backbone remodeling presents a sub-
stantial challenge for computational methods
owing to limited conformational sampling and
imperfect energy functions (6).
Novel recognition modules (7), inhibitors (8, 9),
enzymes (2), and immunogens (4, 5, 10, 11) have
been designed by grafting functional constel-
lations of side chains onto protein scaffolds of
predefined backbone structure. In all cases, the
restriction to using predetermined scaffold back-
bone structures limited the complexity of the
functional motifs that could be transplanted. For
example, the de novo enzymes could accommo-
date grafting of only three or four catalytic groups,
whereas many natural enzymes have six or more
(12), and the immunogens were limited to con-
tinuous (single-segment) epitopes even though
most antibody epitopes are discontinuous (involv-
ing two or more antigen segments) (13, 14).
To address the challenge of incorporating back-
bone flexibility modeling into grafting design, we
developed a hybrid computational-experimental
method for grafting the backbone and side chains
of functional motifs onto scaffolds (Fig. 1). We
tested this method by grafting a discontinuous
HIV gp120 epitope, targeted by the broadly neu-
tralizing monoclonal antibody b12 (15), onto
an unrelated scaffold. b12 binds to a conserved
epitope within the CD4-binding site (CD4bs) of
gp120 (16), an area of great interest for vaccine
design. We focused on transplantation of two
segments from gp120: residues 365 to 372, known
as the CD4b (CD4 binding) loop (17), and resi-
dues 472 to 476, known as the ODe (outer domain
exit) loop (16). The b12-gp120 interaction in-
volves six or seven backbone segments on gp120
(16), but 60% of the buried surface area on gp120
lies on the CD4b and ODe loops, and a Rosetta
energy calculation (18) suggested that these two
1
Department of Biochemistry, University of Washington, Seattle,
WA 98195, USA.
2
Ph.D. Program in Computational Biology,
Instituto Gulbenkian de Cincia, Oeiras, Portugal.
3
Divison of
Basic Sciences, Fred Hutchinson Cancer Research Center, Seattle,
WA 98109, USA.
4
Vaccine Research Center, National Institute of
Allergy and Infectious Diseases, Bethesda, MD 20892, USA.
5
Howard Hughes Medical Institute, University of Washington,
Seattle, WA 98195, USA.
6
IAVI Neutralizing Antibody Center,
The Scripps Research Institute, La Jolla, CA 92037, USA.
7
De-
partment of Immunology and Microbial Science, The Scripps
Research Institute, La Jolla, CA 92037, USA.
*These authors contributed equally to this work.
Present address: Arzeda Corporation, Seattle, WA 98102,
USA.
To whom correspondence should be addressed. E-mail:
schief@scripps.edu
www.sciencemag.org SCIENCE VOL 334 21 OCTOBER 2011 373
REPORTS
segments could account for up to 80% of the
binding energy.
The work flow (Fig. 1) has four stages: (i)
scaffold search, in which the Protein Data Bank
(PDB) (19) is searched for scaffolds suitable to
accept the backbone segments comprising the
motif; (ii) scaffold design, in which the motif back-
bone segments replace native scaffold backbone
and new connecting segments and surrounding
side chains are built to support the motif confor-
mation; (iii) computation-guided library design,
in which a small set of mutagenesis libraries for
sequential screening are derived from an ensem-
ble of designs with expanded structural and com-
positional diversity in the connecting segments;
and (iv) in vitro screening, in which computation-
guided libraries are screened to identify clones
with optimal functional activity.
For scaffold search, we developed an algorithm
(Multigraft Match) that exhaustively searched a
culled PDB for suitable scaffolds. For all possi-
ble combinations of four insert positions in
every scaffold, Multigraft Match produced a
low-resolution prediction of whether the epitope
backbone segments could be grafted onto the
scaffold while maintaining backbone continuity
and avoiding steric clash (fig. S1). Eleven
scaffolds satisfied the geometrical and steric
clash requirements and were selected for design
(table S1).
For scaffold design, we developed an algo-
rithm (Multigraft Design) that, given a prelimi-
nary rigid-body orientation for a discontinuous
epitope relative to a scaffold, deleted appropriate
regions of the scaffold, built new segments to
connect the epitope to the scaffold, and designed
side chains neighboring the epitope and connect-
ing segments to support the graft (fig. S2). This
involved aggressive structural manipulations, in-
cluding replacement of ordered secondary structure
motifs by the epitope segments, flexible back-
bone modeling of two or more connecting seg-
ments, and sequence design of 10 or more core
residues. Several design variants of each candi-
date scaffold (fig. S3) were tested for expres-
sion and purification in Escherichia coli. Of 62
Scaffold
search
Scaffold
design
Computation-guided
library design
In vitro
screening
Loop closure
+
Sequence design
Motif
transplantation
Motif
selection
Structural ensemble Sequence Profile
F
r
e
q
u
e
n
c
y

(
%
)
Scaffold display
A
n
t
i
b
o
d
y

b
i
n
d
i
n
g
Round 1 Round 6
ODe Loop CD4b Loop
b12
antibody
gp120
b12
antibody
scaffold
Fig. 1. Combined in silicoin vitro strategy for the transplantation of complex structural motifs to heterologous scaffold proteins. The diagrams illustrate the
stages in the design of a non-HIV scaffold presenting two loops from the b12 epitope on HIV gp120.
Table 1. Affinity and kinetics of the interaction between recombinant 2bodx variants and b12. For
all the reported values, the standard error is T7 of the last significant digit. RL, random library;
L1, library 1; L2, library 2; L3, library 3. k
on
and k
off
represent the kinetic association and disso-
ciation rates, respectively, of the measured interactions.
2bodx variant Origin
b12 interaction parameters (SPR)
k
on
(M
1
s
1
)
k
off
(s
1
)
K
D
(kinetic)
(nM)
K
D
(equilibrium)
(nM)
03 Initial design >300 10
3
Y3 RL ~30 10
3
42 L1 + L2 1.3 10
6
2.3 10
1
177 166.6
43 L1 + L2 + RL 3.0 10
6
1.0 10
1
33.3 33.5
44 L1 + L2 + RL + L3 1.9 10
6
3.6 10
2
18.9 19.5
45 L1 + L2 + RL + L3 3.8 10
6
3.9 10
2
10.3 10.3
21 OCTOBER 2011 VOL 334 SCIENCE www.sciencemag.org 374
REPORTS
candidates tested, 25 could be solubly expressed
and purified (table S1).
Purified designs were tested for b12 binding
by surface plasmon resonance (SPR). One de-
sign, 2bodx_03, which had 39 mutations and 11
deletions relative to the parent protein (fig. S4),
bound to b12 weakly (dissociation constant K
D

300 mM). The binding was specific, because no
binding was detected for the epitope mutant
Asp
114
Arg (D114R) (20) (fig. S5). A high-
resolution (1.3 ) crystal structure of 2bodx_03
showed no discernible electron density for the
epitope or connecting segments (fig. S5 and table
S2), indicating that these regions were flexible
in solution. In an initial attempt to optimize the
b12 affinity of 2bodx_03, a whole-protein ran-
dom mutagenesis library was screened by yeast
display (21). Clone 2bodx_R3 was thereby iso-
lated with two mutations [Ser
177
Gly (S177G)
and Ala
118
Val (A118V)] from 2bodx_03 and
a factor of 10 higher affinity for b12 (K
D
30 mM)
(Table 1, Fig. 2, and fig. S6). This interaction re-
mained three orders of magnitude weaker than
gp120-b12 interaction [K
D
= 20 nM (16)]. The
lowaffinity was likely due to nonoptimal sequences
and conformations in the connecting segments.
Optimization by targeted random mutagenesis
and in vitro screening was not feasible because
allowing 20 amino acids at all 21 positions judged
to be important in the connecting segments would
yield impractical library sizes of 2 10
27
.
In computation-guided library design, we
used a structure-sequence diversification protocol
(fig. S7) to devise relatively small libraries based
on more complete sampling of low-energy struc-
tures and sequences in the connecting segments.
For each connecting segment, 20,000 backbone
conformations were separately generated and
subjected to sequence design while keeping the
rest of the 2bodx_03 structure fixed. Several low-
energy models for each segment were exhaustive-
ly recombined in silico and subjected to further
sequence design to identify 2bodx models with
optimal structures and sequences in all connect-
ing segments. After a final round of conforma-
tional resampling and design (fig. S7), the best
45 models by several Rosetta metrics (18) were
used to generate sequence profiles to identify the
amino acids that occurred at each of the 21 po-
sitions in the connecting segments (fig. S8). The
diversity was reduced by eliminating residues
that occurred at low frequency, that were similar
in size and chemical nature to more frequent res-
idues, or that were judged likely to bury a polar
side chain. The final library allowed mutations at
21 positions and had a theoretical size of 10
12
.
For in vitro screening, we used yeast display.
To overcome the limitations of the library size
supported by yeast display (10
7
), we constructed
two partially overlapping sublibraries and screened
them sequentially (figs. S9 and S10). The first
sublibrary (library 1) contained all (4 10
6
) of
the computationally designed ODe loop con-
necting segments combined with eight design
variants of the CD4b loop connecting segments
present in 23 of the 45 models. After three rounds
of screening, the selected ODe loop variants
(from at least 18 different clones) were combined
with all (2 10
5
) of the computationally designed
CD4b loop variants to create library 2. This sub-
library was screened for three rounds to isolate
clone 2bodx_42, which differed from 2bodx_03
by17 mutations (fig. S11). Recombinant 2bodx_42
bound b12 with a K
D
of 166 nM, an improvement
by a factor of >1800 over 2bodx_03 (Table 1).
Introducing the A118Vmutation from 2bodx_R3
further increased b12 affinity, as the resulting
variant (2bodx_43) bound b12 with a K
D
of 33 nM
(Table 1, Fig. 2, and fig. S6), within a factor of 2
of the b12-gp120 affinity (16). Introducing the
D114R mutation on 2bodx_43 resulted in loss of
detectable b12 binding (fig. S12), demonstrating
that the binding was specific to the epitope. Further,
2bodx_43 was thermally stable (melting point =
75C) and monomeric in solution (fig. S13).
To assess whether the b12 affinity could be im-
proved further and to evaluate if the computation-
Fig. 2. Isolation of scaffold 2bodx variants
with high b12 affinity and specificity. (A)
Screening of the computation-guided libraries
led to rapid enrichment of clones with
high b12 affinity; R1-R3 refer to rounds
1 to 3 of selection. (B) SPR equilibrium anal-
ysis of the initial computational design
(2bodx_03) and the 2bodx variants identi-
fied from the directed libraries (Table 1 and
fig. S9). (C) 2bodx_43 binds to b12, but
not to CD4 or other antibodies that target
the CD4bs on gp120.
Fig. 3. Atomic-level recapitulation of the b12-gp120 interface by the b12-2bodx_43 complex. (A)
Structure of b12 in complex with 2bodx_43. (B) The conformations of the transplanted loops (yellow) in
2bodx_43 (red) mimic their conformations on gp120 (green). (C) Conformations of side chains (sticks)
making important contacts in the b12-gp120 complex are preserved at the 2bodx_43-b12 interface;
H1, H2 and H3 refer to the CDR loops of b12.
www.sciencemag.org SCIENCE VOL 334 21 OCTOBER 2011 375
REPORTS
guided libraries restricted the sequence space
effectively, we screened a third library based on
2bodx_43 with expanded sampling at seven po-
sitions (fig. S10). The highest-affinity clone iso-
lated (2bodx_45) differed from 2bodx_43 by
two mutations (fig. S4) and bound b12 with a
K
D
of 10 nM, a factor of 3 better than 2bodx_43
and as tightly as gp120 (16) (Table 1, Fig. 2,
and fig S6). Another high-affinity variant se-
lected from this library (2bodx_44, K
D
= 19 nM,
fig. S6) was used to investigate the b12 binding
contributions of library-selected mutations. We
measured the b12 binding of 2bodx_44 constructs
in which the evolved residues were individually
reverted to their 2bodx_03 identity. Only 6 of 16
reversions reduced the b12 affinity of 2bodx_44
by a factor of 3 or more, and a 2bodx_03 variant
that contained nine of the 2bodx_44 mutations
had only micromolar affinity for b12 (K
D
=
1.5 mM) (fig. S14). Thus, the selected mutations
made synergistic contributions to the high b12
affinity of 2bodx constructs.
To evaluate the degree to which the 2bodx-
b12 interaction recapitulated the gp120-b12 inter-
action, we solved a crystal structure for 2bodx_43
complexed with b12 at 2.07 resolution (Fig.
3A and table S2). Comparison with the gp120-
b12 complex (16) revealed a high degree of mim-
icry; superposition of the epitope and paratope
of both complexes gave an overall backbone root
mean square deviation (RMSD) of 0.71 (Fig.
3, B and C) (22). Consistent with good backbone
mimicry, important interactions involving b12
heavy-chain residues Tyr
53
, Tyr
98
, Trp
100
, Asn
100g
,
and Tyr
100h
were recapitulated in the 2bodx_43-
b12 complex (Fig. 3C and tables S3 and S4). The
total buried areas in the complexes were also
similar, except for a small additional area on the
scaffold outside the epitope (fig. S15).
The CD4bs is a major antibody target in HIV
infection (23). Reagents are desired that bind b12
but not CD4bs-directed non-neutralizing anti-
bodies (24) such as b13 that engages gp120
similarly to b12 (25). Of eight CD4bs-directed
antibodies tested, 2bodx_43 bound tightly to b12
only (Fig. 2C). Additional SPR analyses showed
that 2bodx_43 binds more tightly to b12 than to
b13 by a factor of >10,000 (fig. S16) (26). These
results indicate that b12 epitope-scaffolds are
promising tools for HIV vaccine research and en-
courage the application of backbone grafting to
engineer antigens, enzymes, and inhibitors.
References and Notes
1. L. L. Looger, M. A. Dwyer, J. J. Smith, H. W. Hellinga,
Nature 423, 185 (2003).
2. J. B. Siegel et al., Science 329, 309 (2010).
3. S. J. Fleishman et al., Science 332, 816 (2011).
4. B. E. Correia et al., Structure 18, 1116 (2010).
5. G. Ofek et al., Proc. Natl. Acad. Sci. U.S.A. 107, 17880 (2010).
6. D. J. Mandell, T. Kortemme, Nat. Chem. Biol. 5, 797 (2009).
7. S. E. Rutledge, H. M. Volkman, A. Schepartz, J. Am.
Chem. Soc. 125, 14336 (2003).
8. S. K. Sia, P. S. Kim, Proc. Natl. Acad. Sci. U.S.A. 100, 9756 (2003).
9. L. Martin et al., Nat. Biotechnol. 21, 71 (2002).
10. E. Drakopoulou et al., J. Biol. Chem. 271, 11979 (1996).
11. S. M. Lu, R. S. Hodges, J. Biol. Chem. 277, 23515 (2002).
12. D. Baker, Protein Sci. 19, 1817 (2010).
13. L. Jin, B. M. Fendly, J. A. Wells, J. Mol. Biol. 226, 851 (1992).
14. N. D. Rubinstein et al., Mol. Immunol. 45, 3477 (2008).
15. D. R. Burton et al., Science 266, 1024 (1994).
16. T. Zhou et al., Nature 445, 732 (2007).
17. B. Chen et al., Nature 433, 834 (2005).
18. See supporting material on Science Online.
19. H. M. Berman et al., Nucleic Acids Res. 28, 235 (2000).
20. Y. Li et al., Nat. Med. 13, 1032 (2007).
21. G. Chao et al., Nat. Protoc. 1, 755 (2006).
22. With the overall interface superposition of the 2bodx_43-b12
and gp120-b12 complexes, RMSD values for individual
elements were low, as follows: CD4b loop, 0.8 ; ODe loop,
1.5 ; b12 CDRH1 residues 25 to 34, 0.28 ; H2 residues
52 to 56, 0.6 ; H3 residues 94 to 101, 0.7 .
23. L. Stamatatos, L. Morris, D. R. Burton, J. R. Mascola,
Nat. Med. 15, 866 (2009).
24. R. Pantophlet et al., J. Virol. 77, 642 (2003).
25. L. Chen et al., Science 326, 1123 (2009).
26. The very low affinity of b13 for 2bodx_43 is likely due to
(i) the different conformations of the CD4b and ODe loops
in the b13-gp120 structure compared to the b12-gp120
and b12-2bodx_43 structures, and (ii) important contacts
for b13 at gp120 residues 419, 421, and 425 in the
b20-b21 region not present on the scaffold.
Acknowledgments: Supported by grants from the International
AIDS Vaccine Initiative Neutralizing Antibody Consortiumand
the Bill and Melinda Gates Foundation Consortium for
AIDS Vaccine Discovery. B.E.C. was supported by a fellowship
from the Portuguese Fundao para a Cincia e a Tecnologia
(SFRH/774 BD/32958/2006). We thank I. Wilson for
comments on the manuscript; D. Burton, A. Hessell, and
the IAVI Neutralizing Antibody Consortium Reagent
Repository for providing b6, b12, and b13; J. Mascola for
VRC01 and VRC03; Progenics Inc. for CD4-IgG2; D. Dimitrov
for m6, m14, and m18; J. Robinson for 15E; and R. Wyatt
for F105 and HxB2 gp120. We thank G. Nabel, M. Kanekiyo,
and Z.-Y. Yang for experiments on early generations of
b12 scaffolds. Coordinates and structure factors were
deposited in the PDB as entries 3RPT and 3RU8. The
University of Washington has filed a patent application
on the b12 scaffolds and the scaffolding method
developed in this study. Materials and information will
be provided to noncommercial users under the Uniform
Biological Materials Transfer Agreement. The Multigraft
software developed in this work is freely available to
noncommercial users through the Rosetta Commons
agreement (www.rosettacommons.org).
Supporting Online Material
www.sciencemag.org/cgi/content/full/334/6054/373/DC1
Materials and Methods
Figs. S1 to S16
Tables S1 to S4
References
6 June 2011; accepted 8 September 2011
10.1126/science.1209368
Antagonists Induce a Conformational
Change in cIAP1 That
Promotes Autoubiquitination
Erin C. Dueber,
1
Allyn J. Schoeffler,
1
Andreas Lingel,
1
J. Michael Elliott,
2
Anna V. Fedorova,
1
Anthony M. Giannetti,
3
Kerry Zobel,
1
Brigitte Maurer,
4
Eugene Varfolomeev,
1
Ping Wu,
4
Heidi J. A. Wallweber,
4
Sarah G. Hymowitz,
4
Kurt Deshayes,
1
Domagoj Vucic,
1
* Wayne J. Fairbrother
1
*
Inhibitor of apoptosis (IAP) proteins are negative regulators of cell death. IAP family members
contain RING domains that impart E3 ubiquitin ligase activity. Binding of endogenous or
small-molecule antagonists to select baculovirus IAP repeat (BIR) domains within cellular IAP
(cIAP) proteins promotes autoubiquitination and proteasomal degradation and so releases
inhibition of apoptosis mediated by cIAP. Although the molecular details of antagonistBIR
domain interactions are well understood, it is not clear how this binding event influences the
activity of the RING domain. Here biochemical and structural studies reveal that the unliganded,
multidomain cIAP1 sequesters the RING domain within a compact, monomeric structure that
prevents RING dimerization. Antagonist binding induces conformational rearrangements that
enable RING dimerization and formation of the active E3 ligase.
I
nhibitor of apoptosis (IAP) proteins are anti-
apoptotic factors important in blocking pro-
grammed cell death, or apoptosis, in response
to a variety of stimuli (1, 2). Whether initiated by
external death signals transduced by specific cell
surface receptors (extrinsic pathway) or by inter-
nal cues of compromised cellular integrity (intrinsic
pathway), apoptotic signaling pathways converge
in the activation of caspases (cysteine-dependent
aspartyl-specific proteases), which effect wide-
spread proteolytic damage and cell death (3). IAPs
hold these cellular executioners in check, either
through direct inhibitory interactions or by imped-
ing upstream caspase activation pathways (4).
Many cancer cells overexpress IAPs, which al-
lows themto resist cytotoxic therapies (2, 5). Thus,
IAPs are potentially important targets for cancer
treatment (2, 5, 6).
IAP-targeting therapeutics designed to mimic
the endogenous IAP antagonist, SMAC (second
mitochondrial activator of caspases)DIABLO
(direct IAP-binding protein with low isoelectric
point) (7, 8), have recently entered phase I clinical
1
Department of Early Discovery Biochemistry, Genentech, 1 DNA
Way, South San Francisco, CA 94080, USA.
2
Department of
Protein Chemistry, Genentech, 1 DNA Way, South San Francisco,
CA 94080, USA.
3
Department of Biochemical Pharmacology,
Genentech, 1 DNA Way, South San Francisco, CA 94080, USA.
4
Department of Structural Biology, Genentech, 1 DNA Way,
South San Francisco, CA 94080, USA.
*To whom correspondence should be addressed. E-mail:
fairbro@gene.com (W.J.F.); domagoj@gene.com (D.V.)
21 OCTOBER 2011 VOL 334 SCIENCE www.sciencemag.org 376
REPORTS
trials (9). The conserved amino-terminal tetra
amino acid peptide (Ala-Val-Pro-Ile or AVPI) of
mature SMAC binds to a surface groove on bacu-
lovirus IAPrepeat (BIR) domains (10, 11). SMAC
mimetics show proapoptotic activity by prevent-
ing interactions between IAPs and active caspases
(2, 4). Moreover, binding of these small-molecule
antagonists to cellular IAP (cIAP) BIR domains
leads to a second mechanism of antagonism
rapid proteasomal degradation of the cIAP pro-
teins induced by autoubiquitination (1214).
Although analyses of antagonistBIR domain
complexes have yielded detailed information re-
garding the molecular determinants of these inter-
actions (1517), how antagonist binding induces
cIAP RING E3 ligase activity is not understood.
We sought to define a minimal cIAP1 construct
that would maintain key full-length cIAP1 proper-
ties and be suitable for biophysical and structural
studies. As IAP antagonists bind preferentially to
the BIR3 domain of cIAP1 (13), we focused on a
truncated version of the protein that lacks the first
two BIR domains (Fig. 1Aand fig. S1A). Remov-
al of the BIR1 and BIR2 domains, which include
the tumor necrosis factor receptorassociated
factor 2 (TRAF2)interacting region of cIAP1
(1820), did not affect antagonist-induced deg-
radation (fig. S1). The resulting cIAP1 construct,
comprising the BIR3 domain followed by the
ubiquitin-associated (UBA) domain, the caspase
activation and recruitment domain (CARD), and
the carboxy-terminal RINGdomain (BIR3-RING),
recapitulates full-length cIAP1 behavior and dem-
onstrates proteasomal degradation that is mediated
by IAP antagonist-induced autoubiquitination ac-
tivity (fig. S1).
RING E3 ligases often function as dimers
(21, 22), and disruption of RING-mediated di-
merization stabilizes IAPs against IAP antagonists
(23). We thus assessed whether IAP antagonists
affect the oligomerization state of the BIR3-RING
construct. Monomeric unliganded BIR3-RING
appears as a single, discrete band by native gel
analysis (Fig. 1B). Addition of a bivalent antag-
onist (BV6) that can cross-link the protein through
simultaneous binding of separate BIR3 domains
(13) is a control for migration of the BIR3-RING
dimer. Analytical size-exclusion chromatography
and multiangle light scattering (SEC-MALS) re-
sults confirm that the native gel bands represent
monomer and dimer BIR3-RING species (fig.
S2D). Monovalent small-molecule (MV1) and
peptide (AVPW) antagonists shift the mobility of
the BIR3-RING band toward the BV6 dimer con-
trol. In contrast, a BIR3-RING construct lacking
RING dimerization residues at the carboxy termi-
nus (M266 to G611; BIR3-RINGDC7) (18, 23)
does not band-shift unless presented with the bi-
valent antagonist. Together, these findings sug-
gest that binding of monovalent antagonists to the
BIR3 domain induces RING dimerization through
an intramolecular, allosteric mechanism.
It intrigued us that the mobility of the ligand-
free apo BIR3-RINGDC7 band is considerably
retarded compared with apo BIR3-RING, despite
modest differences in molecular weight and net
charge for the two proteins (Fig. 1B). MV1- and
AVPW-bound BIR3-RINGDC7 demonstrate iden-
tical band migrations as the apo BIR3-RINGDC7
sample. Further SEC-MALS analysis corrobo-
rates the distinct mobilities seen by native gel (fig.
S2D). These results suggest that the dimerization-
deficient mutant exists in a more extended con-
formation than the monomeric BIR3-RINGprotein
and that the carboxy-terminal residues of the RING
Fig. 1. BIR3-binding IAP antagonists induce
RING dimerization. (A) cIAP1 constructs. (B)
Native gels show the relative mobility of
BIR3-RING and BIR3-RINGDC7 proteins (M,
monomer; D, dimer) in the absence and
presence of IAP antagonists. The differing
effects of monovalent antagonists MV1 and
AVPWon cIAP1 BIR3-RINGprotein mobility
are consistent with their binding half-lives
of 0.9 and 4.2 min, respectively, as deter-
mined by surface plasmon resonance (SPR)
spectroscopy (fig. S2, A and B). For compari-
son, the half-life of BV6 binding to BIR3-
RING is 10.6 min (fig. S2C).
Fig. 2. Crystal structure of the apo cIAP1 monomer. (A) A ribbon diagram of
BIR3-RING
xtal
illustrates the global arrangement of the BIR3 domain, UBA
domain, and CARD around a core RING domain. Inset shows surface views of
the structure. Protein domains are shown in cyan, green, purple, orange, and
yellow as indicated; zinc ions are shown as gray spheres. (B) Structural align-
ment with a dimeric cIAP2 RING protomer (red) illustrates how the RING
dimerization elements are sequestered in the BIR3-RING
xtal
structure. Spe-
cifically, the RING carboxy-terminal strand (C strand), which self-associates in
the RING:RING interface, interacts with the edge strand of the BIR3 b sheet in
the monomer structure; and the RING amino-terminal helix (N helix) is bent
~110 and partially unwound as compared with dimerized, active cIAP2 RING
protomers and so affects the E2-binding surface (fig. S5). cIAP1 domains other
than the RING domain are shown faded for visual clarity. Inset depicts the cIAP2
RING dimer structure (red and gray protomers, PDB code 3EB5).
www.sciencemag.org SCIENCE VOL 334 21 OCTOBER 2011 377
REPORTS
domain are critical for maintaining the more com-
pact apo structure.
Crystallization required design of a variant of
the BIR3-RING protein that eliminated two pro-
tease cleavage sites by deletion of a 10-residue
section within the BIR3 to UBAlinker (DS363 to
D372) and a point mutation in which Tyr replaced
Phe
387
(F387Y) (BIR3-RING
xtal
) (figs. S1 and S3).
The structure of BIR3-RING
xtal
was solved to
1.9 resolution by single-wavelength anoma-
lous dispersion (SAD) by using three bound zinc
ions, and the final model was refined to an R
free
of 23% (fig. S4 and table S1).
Consistent with an allosteric activation hy-
pothesis, the apo monomer adopts a compact con-
formation, with the BIR3 domain, UBA domain,
CARD, and intervening linkers combining to se-
quester a centralized RING domain (Fig. 2A).
Multiple interfaces formed between the RING
domain and the surrounding components together
bury >4400
2
of surface area (table S2). Al-
though portions of the RING domain remain
accessible to solvent, key RING dimerization ele-
ments are buried or displaced in this autoinhibited
state (23) (Fig. 2B). In contrast, the conformations
of the BIR3 domain, UBAdomain, and CARDare
similar to homologous domain structures (fig. S6).
The interdomain linkers also form well-ordered
secondary structures that interact with the RING
domain. Inspection of the interdomain interfaces
reveals specific interactions that appear to con-
tribute to the confinement of the RING domain
and thus to suppression of RING:RING dimer-
ization. The BIR3 domain forms an extensive
interface with the RINGdomain (table S2), which
is buttressed by the UBA domain and a section
of the BIR3-UBA linker (BU-linker). The BIR3
residue W329, located in a hydrophobic pocket
between these four modules, is positioned to
A B
E F
W329
V613
I584
L355
L359
L356
L359
L360
L558
L565
L589
V613
R569
E508 D511
R465
H2O
R300
R314
N301
R600
C D
K610
R600
R606
K601
E440
E444
E447
E451
R614
A
V
P
V613
K610
T612
I
A
G611
G
s
t
d
s
A
p
o
A
V
P
W
B
V
6
A
p
o
A
V
P
W
B
V
6
A
p
o
A
V
P
W
B
V
6
A
p
o
A
V
P
W
B
V
6
A
p
o
A
V
P
W
B
V
6
A
p
o
A
V
P
W
B
V
6
A
p
o
A
V
P
W
B
V
6
A
p
o
A
V
P
W
B
V
6
A
p
o
A
V
P
W
B
V
6
A
p
o
A
V
P
W
B
V
6
WT W329A W329F R300A
N301A
R314A
E317A
S318A
R465A
E508A
D511A
E440K
E444K
E447K
E451K
K610A K610E T612R
A
p
o
A
V
P
W
B
V
6
A
p
o
A
V
P
W
B
V
6
E440-
E452
M545-
L556
L356A
L359A
L360A
A
p
o
A
V
P
W
B
V
6
D
M
H
WT W329A
R300A
N301A
R314A T612R
L356A
L359A
L360A
MG132:
BV6:
cIAP1:
WT W329F
actin
cIAP1 (FL)
v
e
c
t
o
r
-
+
-
-
+
-
-
+
- -
+
- -
+
- -
+
- -
+
- -
+
-
-
+
- -
+
- -
+
- -
+
- -
+
- -
+
-
-
-
K305
R314
Fig. 3. Domain-domain interactions. Detailed views of the (A) BIR3:BU-
linker:UBA:RING, (B) BU-linker:RING, (C) BIR3:RING (D) CARD:RING, and (E)
UC-linker:RING interfaces are shown with key residues highlighted as sticks.
Color scheme is the same as that of Fig. 2A, with the exception of electro-
static surfaces shown for the UC-linker (top) and RING (bottom) in panel (E).
(F) Overlay of the RING C strand (orange sticks) and SMAC-peptide antag-
onist (white sticks, PDB code 3D9U) within the BIR3 (cyan surface) peptide-
binding groove. Unlike the peptide antagonist, which demonstrates good
surface complementarity and makes numerous favorable interactions with
the BIR3 domain, the RING C strand merely crosses over the peptide-binding
groove, making a single hydrogen bond contact with G312 of the BIR3 domain
via residue T612. (G) Native gel analysis reveals mutants of BIR3-RING
xtal
that dimerize in the absence of antagonist (red labels). All gels aligned by
standards shown at far left (stds). Positions of WT monomer (M) and dimer
(D) are indicated. (H) The stability of full-length FLAG-tagged cIAP1 variants
in 293T cells (T10 mM MG132 or 2 mM BV6 for 1 hour) was determined by
evaluating cell lysates for cIAP1 constructs by Western blotting with anti-
Flag antibody. Actin was probed as a loading control.
21 OCTOBER 2011 VOL 334 SCIENCE www.sciencemag.org 378
REPORTS
interact favorably with the helix dipole of an
adjacent UBA a helix (Fig. 3A). In addition to
contributing to this pocket, residues P353 to T362
of the BU-linker form an amphipathic helix
whose hydrophobic surface pairs with the hydro-
phobic face of the RING N helix and a nearby
surface of the RING domain (Fig. 3B). These in-
teractions pin the N helix in a bent conformation
(relative to dimeric RINGprotomers) (Fig. 2B) by
effectively substituting the N helix dimerization
contacts with similar intramolecular interactions.
On the opposite side of the RING domain, back-
bone carbonyl groups and residue R600 interact
with R300 and R314, and N301 from the BIR3
domain, respectively (Fig. 3C). The remaining
RING interactions are also polar, with the large
CARD:RING interface (table S2) characterized
by a hydrogen-bonding network that includes
residues R465, E508, D511, and R569, as well as
ordered water molecules (Fig. 3D). Additionally,
the UBA-CARD linker (UC-linker) acts as a
negatively charged helical strut that extends
from the final UBA a helix. The strut flanks a
positively charged surface of the RING domain
(Figs. 3E and 2A) and thus positions the CARD
for its interactions with the RING domain. Final-
ly, the C strand of the RING associates with the
edge of the BIR3 b sheet. Comparison with a
structure of the BIR3 domain bound to the SMAC
amino-terminal peptide reveals that the RING C
strand partially occludes the antagonist-binding
site (Fig. 3F) (24).
To evaluate the contributions of these and
other specific interactions to the compact mono-
mer state, we introduced mutations designed to
disrupt these contacts into the BIR3-RING
xtal
construct (fig. S7). Evaluating these variants
with native gels revealed that mutations within
the BIR3:RING and BU-linker:RING inter-
faces promote dimerization, whereas muta-
tions targeting other interactions have little to
no effect, compared with the nearwild-type
(WT) BIR3-RING
xtal
control (Fig. 3G). Note
that the cIAP1 monomer was insensitive to per-
turbation of the highly polar CARD:RING and
UC-linker:RING interfaces, consistent with pre-
vious gross deletions of regions encompassing
the UBA domain or CARD that showed no
observable effect on cIAP1 autoubiquitination
activity (18). SEC-MALS analyses of the dimer-
promoting mutants confirm the dimeric state of
these variants in the absence of antagonist (fig.
S8). Accordingly, addition of AVPW did not
significantly alter the apparent molecular weight
of these mutants, whereas WTand W329F mono-
mer samples shifted to dimeric species under sim-
ilar conditions.
We next assessed the effect of these interface
mutations on the stability and, therefore, on the
autoubiquitination activity of full-length cIAP1
expressed in human embryonic kidney 293Tcells.
As predicted, WTcIAP1 was considerably more
stable than variants expected to dimerize on the
basis of biophysical analysis of the BIR3-RING
constructs (Fig. 3H). This was particularly evi-
dent for the W329A and T612R mutants; these
could only be detected after treatment with the
proteasome inhibitor, MG132, which suggests
that their instability results from elevated auto-
ubiquitination. Similar results were observed on
expression of these variants of the shorter BIR3-
RING
xtal
construct (fig. S9).
To visualize activated, dimeric cIAP1, we col-
lected small-angle x-ray scattering (SAXS) data
for the AVPW-bound BIR3-RING construct. An
averaged ab initio model of the dimer reveals an
extended, two-fold symmetric complex that is
compatible with placement of the RING:RING
homodimer structure at its center (Fig. 4, A and
B). Ab initio SAXS models generated for two
other cIAP1 constructs, an N-terminally MBP-
tagged BIR3-RING (MBP-BIR3-RING) dimer
and a shorter UBA-RING (S385 to S618) dimer,
support a model in which BIR3 domains are
located at the periphery of an elongated, nearly
planar BIR3-RINGhomodimer that forms around
a central RING:RINGinterface (Fig. 4, Ato C, and
figs. S10 and S11). In agreement with the mu-
tational analysis, UBA-RING purifies as a con-
stitutive dimer, whereas MBP-BIR3-RING, like
BIR3-RING, requires the addition of antagonist
(AVPW) to dimerize (fig. S12).
Taken together, these data suggest an E3 ligase
activation model in which apo BIR3-RING exists
predominantly in a closed, inactive monomer state.
Antagonist binding to the BIR3 domain is in-
compatible with this conformation and stabilizes
an open state in which the RING domain is ex-
posed and free to dimerize. RING dimerization
enables rapid cIAP1 autoubiquitination and sub-
sequent degradation by the proteasome (23, 25).
Given the high sequence conservation between
cIAP1 and 2, the proposed model likely reflects a
general mechanism of antagonist-induced cIAP
activation, although recent work suggests that apo
cIAP2 may have a greater propensity to dimerize
than cIAP1 (25). In contrast, modeling suggests
that X-linked AIP(XIAP), which lacks the CARD,
could not adopt the same autoinhibited confor-
mation as cIAP1. Nevertheless, the current work
provides a molecular basis for understanding the
mechanism of action of IAP antagonists and of-
fers important insights into the regulation of cIAP
E3 ligase activity.
Fig. 4. SAXS analysis of the cIAP1 dimer. (A) Proposed domain ar-
rangements for MBP-BIR3-RING, BIR3-RING, and UBA-RING dimers
and (B) two views of averaged ab initio SAXS envelopes (P2 symmetry)
for the same constructs. All data were collected in the presence of
1 mM AVPW. (C) Pairwise distance-distribution functions of SAXS data
for each dimer.
www.sciencemag.org SCIENCE VOL 334 21 OCTOBER 2011 379
REPORTS
References and Notes
1. G. S. Salvesen, C. S. Duckett, Nat. Rev. Mol. Cell Biol. 3,
401 (2002).
2. D. Vucic, W. J. Fairbrother, Clin. Cancer Res. 13, 5995
(2007).
3. G. S. Salvesen, J. M. Abrams, Oncogene 23, 2774 (2004).
4. J. N. Dynek, D. Vucic, Cancer Lett. (2010).
5. A. M. Hunter, E. C. LaCasse, R. G. Korneluk, Apoptosis 12,
1543 (2007).
6. S. W. Fesik, Nat. Rev. Cancer 5, 876 (2005).
7. C. Du, M. Fang, Y. Li, L. Li, X. Wang, Cell 102, 33
(2000).
8. A. M. Verhagen et al., Cell 102, 43 (2000).
9. J. A. Flygare, W. J. Fairbrother, Expert Opin. Ther. Pat.
20, 251 (2010).
10. Z. Liu et al., Nature 408, 1004 (2000).
11. G. Wu et al., Nature 408, 1008 (2000).
12. A. Gaither et al., Cancer Res. 67, 11493 (2007).
13. E. Varfolomeev et al., Cell 131, 669 (2007).
14. J. E. Vince et al., Cell 131, 682 (2007).
15. C. Ndubaku et al., ACS Chem. Biol. 4, 557 (2009).
16. T. K. Oost et al., J. Med. Chem. 47, 4417 (2004).
17. K. Zobel et al., ACS Chem. Biol. 1, 525 (2006).
18. J. W. Blankenship et al., Biochem. J. 417, 149 (2009).
19. T. Samuel et al., J. Biol. Chem. 281, 1080 (2006).
20. E. Varfolomeev, S. M. Wayson, V. M. Dixit,
W. J. Fairbrother, D. Vucic, J. Biol. Chem. 281, 29022
(2006).
21. R. J. Deshaies, C. A. Joazeiro, Annu. Rev. Biochem. 78,
399 (2009).
22. P. D. Mace, S. Shirley, C. L. Day, Cell Death Differ. 17, 46
(2010).
23. P. D. Mace et al., J. Biol. Chem. 283, 31633 (2008).
24. R. Kulathila et al., Acta Crystallogr. D Biol. Crystallogr.
65, 58 (2009).
25. R. Feltham et al., J. Biol. Chem. 286, 17015 (2011).
Acknowledgments: We thank C. Yu, P. Lupardus, Shamrock
Structures LLC, and staff at the Advanced Light Source,
the Stanford Synchrotron Radiation Lightsource (SSRL)
and the Advanced Photon Source at Argonne National
Laboratory for their assistance with sample handling and
data collection. The Advanced Light Source is supported
by the Director, Office of Science, Office of Basic
Energy Sciences, of the U.S. Department of Energy
(DOE) under contract no. DE-AC02-05CH11231. SSRL is
a Directorate of SLAC National Accelerator Laboratory
and an Office of Science User Facility operated for the
DOE Office of Science by Stanford University. The SSRL
Structural Molecular Biology Program is supported by
the DOE Office of Biological and Environmental Research
and by the Biomedical Technology Program, National
Center for Research Resources, NIH (P41RR001209). Use
of the Advanced Photon Source, a facility operated for
DOE Office of Science by Argonne National Laboratory,
was supported under contract no. DE-AC02-06CH11357.
We also thank the Microchemistry and Proteomics
Laboratory and the Oligo Synthesis and DNA Sequencing
facilities at Genentech for their technical support.
Atomic coordinates and structure factors have been
deposited in the Protein DataBank with accession code
3T6P. Reagents are available from Genentech subject
to a material transfer agreement.
Supporting Online Material
www.sciencemag.org/cgi/content/full/334/6054/376/DC1
Materials and Methods
Figs. S1 to S12
Tables S1 and S2
References (2637)
3 May 2011; accepted 15 September 2011
10.1126/science.1207862
Mass Spectrometry of Intact V-Type
ATPases Reveals Bound Lipids and the
Effects of Nucleotide Binding
Min Zhou,
1
* Nina Morgner,
1
* Nelson P. Barrera,
2,3
* Argyris Politis,
1
Shoshanna C Isaacson,
1
Dijana Matak-Vinkovi,
2
Takeshi Murata,
4
Ricardo A. Bernal,
5
Daniela Stock,
6,7
Carol V. Robinson
1

The ability of electrospray to propel large viruses into a mass spectrometer is established and is
rationalized by analogy to the atmospheric transmission of the common cold. Much less clear is the
fate of membrane-embedded molecular machines in the gas phase. Here we show that rotary
adenosine triphosphatases (ATPases)/synthases from Thermus thermophilus and Enterococcus hirae
can be maintained intact with membrane and soluble subunit interactions preserved in vacuum.
Mass spectra reveal subunit stoichiometries and the identity of tightly bound lipids within the
membrane rotors. Moreover, subcomplexes formed in solution and gas phases reveal the regulatory
effects of nucleotide binding on both ATP hydrolysis and proton translocation. Consequently, we
can link specific lipid and nucleotide binding with distinct regulatory roles.
R
otary ATPases/synthases are membrane-
associated molecular machines that per-
form biological energy conversion. Both
V-type and F-type complexes consist of two re-
versible motors: the ion pump/turbine in V
O
/F
O
and the chemical motor/generator in V
1
/F
1
. The
mode of operation is influenced by the ratio of
the two fuels (protons:ATP) that drive the two
motors. (13). The membrane-embedded V
O
/F
O
domain mediates the movement of Na
+
or pro-
tons across the membrane, whereas V
1
/F
1
do-
mains interact with nucleotides and inorganic
phosphate either to produce or consume ATP in
the case of the eukaryotic F- and V-type families,
respectively. Eubacteria and archaea typically
have only one type of rotary ATPase/synthase
for both functions. Most bacteria have complexes
of the F-type, but some bacteria and all known
archaea have complexes closely related to eu-
karyotic V-type ATPases [also known as A-type
ATPases/synthases (4)]. Whether of F- or V-type,
the physiological function of most prokaryotic
complexes is ATP synthesis; however, many have
evolved regulatory functions that allow reversal
into ATP-driven proton pumps if required.
F
1
-F
O
, and V
1
-V
O
, are mechanically coupled
by a central rotating shaft and held together by
peripheral stalks (Fig. 1). Structural details derive
from isolated subcomplexes of F
1
and V
1
(57)
and from membrane embedded proteolipid rings
of various species (811). Despite this wealth of
structural information, no high-resolution struc-
tures of any intact rotary ATPases/synthases have
been reported. Thus, regulatory allosteric changes
that involve both the soluble head and the mem-
brane sector are lost. In addition heterogeneous
interactions with lipids and nucleotides are dif-
ficult to observe with existing structural biology
approaches.
We show using electrospray mass spectrom-
etry (MS) that rotary ATPases/synthases from
Thermus thermophilus (TtATPase) and Entero-
coccus hirae (EhATPase) can remain intact in the
gas phase. Previously, composite models were as-
sembled of the intact TtATPase by low-resolution
electron microscopy (EM) data in combination
with high-resolution x-ray structures of subunits
(12), whereas the first cryo-EM data revealed
views of the entire membrane-embedded region
(13). We compared the MS of the two complexes,
the TtATPase with that of the less well charac-
terized EhATPase. Current models suggest that
the EhATPase has only one peripheral stalk (14),
and the stoichiometry of the K subunits in the
membrane ring was determined as seven from
EM (15) and 10 from x-ray analysis (8).
TtATPase was purified as described (16) with
dodecyl maltoside (DDM) for solubilization, be-
cause under these conditions the complex is most
stable and does not formaggregates. The complex
was introduced into a mass spectrometer modified
for high-mass complexes (17). Well-resolved charge
states were assigned to the intact particle consist-
ing of 26 subunits and nine different proteins
(Fig. 1A). An experimentally determined mass of
659,202 (T131) dalton corresponds to that calcu-
lated for the intact complex, on the basis of subunit
masses determined by MS(table S1), plus addition-
al mass due to incomplete desolvation, and lipid
and nucleotide binding (fig. S1). Gas-phase ac-
tivation is necessary to release the complex from
its detergent micelle (18, 19), giving rise to peaks
at higher mass/charge ratio (m/z) than the intact
1
Department of Chemistry, Physical and Theoretical Chem-
istry Laboratory, University of Oxford, Oxford OX1 3QZ, UK.
2
Department of Chemistry, Lensfield Road, University of Cam-
bridge, Cambridge CB2 1EW, UK.
3
Department of Physiol-
ogy, Pontificia Universidad Catlica de Chile, Alameda 340,
Santiago, Chile.
4
Department of Chemistry, Graduate School
of Science, Chiba University, 1-33 Yayoi-cho, Inage, Chiba 263-
8522, Japan.
5
Department of Chemistry, University of Texas
at El Paso, El Paso, TX 79968, USA.
6
The Victor Chang Cardiac
Research Institute, Lowy Packer Building, 405 Liverpool Street,
Darlinghurst NSW 2010, Australia.
7
Faculty of Medicine, Uni-
versity of New South Wales, Sydney 2052, Australia.
*These authors contributed equally to this work.
To whom correspondence should be addressed. E-mail:
carol.robinson@chem.ox.ac.uk
21 OCTOBER 2011 VOL 334 SCIENCE www.sciencemag.org 380
REPORTS
complex, formed by unfolding and dissociation
of highly charged subunits, predominantly sub-
units E, G, and I (20) (fig. S2). At lower m/z,
charge states are assigned to the membranous
subcomplex V
O
(ICL
12
), their bimodal distribu-
tion implying that they are formed both in solu-
tion and gas phases. The corresponding soluble
V
1
complex is also observed, confirming that un-
der these solution conditions a proportion of the
complex dissociates spontaneously (21) (fig. S3).
Using similar experimental parameters, we
recorded a spectrumfor the EhATPase isolated in
DDM, where its functional activity has been es-
tablished previously (22, 23). The membrane-
embedded rotor for this complex is larger because
each Ksubunit contains four transmembrane heli-
ces as compared with two for the corresponding
L subunits from TtATPase. Under conditions in
which the complex emerges from the micelle
surviving intact, the spectrumis not well resolved
(fig. S4). Increases in activation energy lead to
better desolvation and also appearance of sub-
complexes in which either the membrane ring
remains but the peripheral stalks and subunit I
have dissociated, or the stalk subunits are attached
but the membrane region is disrupted (Fig. 1B
and fig. S5). The number of peripheral stalks was
determined as two and the stoichiometry in the
K ring as 10 (175 kD) (fig. S6).
The lipid components in the K
10
ring were
identified as a series of negatively charged cardio-
lipins (figs. S2 and S7 and table S2). The stoi-
chiometry of lipid binding was determined as 10
from the mass of the membrane ring. We iden-
tified six cardiolipin isomers and, from quanti-
tative analysis and measurement of the protein
concentration, deduce specific binding of one
lipid per subunit [Eh Ksubunit 1: 1.2 T 0.1 cardio-
lipins (fig. S8)]. Previously, the lipids were lo-
cated based on the atomic structure of the isolated
K
10
ring, in which peaks of positive density were
attributed to 20 bifurcated phosphatidylglycerol
lipids (8). We docked 10 negatively charged sym-
metric cardiolipins inside the K
10
ring, proximal
to the conserved Lys
32
. The four hydrophobic
chains are positioned with two chains emanating
fromboth sides of the polar head, thus providing a
hydrophobic liningtothe inside of the ring (Fig. 2A).
For the TtATPase L
12
ring, we identified the
bound lipid as phosphatidylethanolamine (PE)
(Fig. 2 and fig. S9). Unexpectedly, dissociated
L subunits were observed either with bound PE
lipid (holo) or without lipid (apo) after tandem
MS of a subcomplex containing the membrane
subunits and peripheral stalks (Fig. 2). Before
disruption, the mass of the subcomplex is con-
sistent with binding of six lipids. This means that
apo and holo subunits coexist within the same
complex. We confirmed this observation by quan-
tifying the protein:lipid ratio as one L subunit:
0.55 T 0.1 PE (i.e., one PE lipid per L subunit
dimer) (fig. S10). Six equivalent binding sites for
lipids in protein dimers in a 12-membered ring
imply a sixfold symmetrical state. Previously, EM
studies of the same preparation of the TtATPase
revealed a sixfold symmetric membrane-embedded
ring (Fig. 2 and fig. S11) (16). Both sets of data
are consistent with close packing of two neigh-
boring subunits, forming six dimers each with
four transmembrane helices (TMHs), thus emu-
lating the arrangement of C subunits in eukary-
otic V-ATPases in which gene duplication has led
to four TMHs per subunit (24, 25). From model-
ing, we find that a rotation of the dimers by
about 60 relative to their original orientation
in the ring effectively locks six Glu
63
residues
in an occluded position, preserving lipid binding
at the dimer interface. Proton transfer can occur at
the remaining six active glutamates [Fig. 2, (i) to
(iii)]; consequently, the proton:ATP ratio for the
rotary enzyme is effectively halved.
A
B
L
K
52+
25+
32+
G G E
E
19+
45+
36+
G G E
E
-C
-F
-F
- I(EF)
2
r
e
l
a
t
i
v
e

i
n
t
e
n
s
i
t
y

(
%
)
m/z 10000 15000 20000 25000
100
0
C
D
B
A
G
C
E
F
D
B
A
G
D
B
A
C
E
D
B
A
G
E
F
D
B
A
G
E
D
B
A
G
m/z
100
26+
34+
-G
-E
6000 9000 12000 15000
57+
45+
r
e
l
a
t
i
v
e

i
n
t
e
n
s
i
t
y

(
%
)
0
26+
L
C
G
D
B
A
F
I
L
C
G
E
D
B
A
F
I
L
C
D
B
A
F
I
L
C
I
B
F
C
D
I
G
E
A
L
C
I
7800 7000
24+
K
E
F
B
A
G
D
K
C
I
Fig. 1. Mass spectra of the intact rotary ATPases from T. thermophilus and E. hirae. (A) Peaks are assigned
to the intact TtATPase complex (stars), loss of the membrane subcomplex (ICL
12
) in solution and gas phases
(dark green and green hexagons, respectively), and dissociation of subunits E and G from the peripheral
stalk (blue circles and squares respectively). (B) For EhATPase, the membrane subcomplex is observed in
contact with the soluble head (green squares). (Inset) Mass spectra of the K ring in aqueous solution.
www.sciencemag.org SCIENCE VOL 334 21 OCTOBER 2011 381
REPORTS
In contrast to E. hirae, which has both F- and
V-type ATPases, T. thermophilus has only one
type of rotary ATPase that operates in both pro-
ton pumping and ATP synthesis modes in vitro
(26). Given that each of the 12 L subunits in the
TtATPase membrane ring contains only two
TMHs, alternating between synthesis and pumping
would involve switching the 12-membered ring
from a high proton:ATP ratio, primed for syn-
thesis (27), into a lower one biased for pumping
(16). This latter scenario thus mimics the eu-
karyotic V-type enzyme where gene duplication
has provided four TMHs each with one active
glutamate. Consequently, this mechanism for
switching between ATP synthesis and ion-pumping
modes is likely assisted by specific lipid binding
for this dual-function rotary ATPase.
These different lipid-binding patterns in the
two rotary ATPases studied here were found to
be invariant between repeat preparations and
different detergent concentrations. Together, they
provide an explanation as to how the disparate
membrane rings L
12
and K
10
(93.6 and 160 kD)
can interact with their respective Csubunits (35.8
and 38.2 kD), which are likely conserved. Lining
with cardiolipin, bound specifically to the inside
of the K
10
ring, reduces the orifice from 54 to
38 (Fig. 2Aand fig. S12). Similarly, converting
a 12- to a 6-membered ring for TtATPase reduces
the orifice from 47 to 39 , creating two very
similar inner diameters (38 and 39 ) (Fig. 2B).
To assess the role of nucleotides in changing
subunit interactions, we compared spectra for the
TtATPase, with and without addition of 50 mM
ATP. In the presence of ATP, the intact complex
and membrane-embedded subcomplex ICL
12
are the predominant species formed in solution
(Fig. 3A and fig. S13). When ATP is depleted,
loss of subunit B leads to extensive dissociation
of the soluble head. In addition, tandem MS of
V
1
in the presence of ATP leads predominantly to
loss of subunit F, with subsequent loss of subunit
A
B
C
D
C H
3
O
O
O
O
P
O
O
NH
3
+
O
-
O H
C H
3
C H
3
O
O
C
O
P
OH
O
O
C H
3
O
O H
C H
O
O
C
O
P
O
O
O H
C H
3
3
O
O H
O H
H
(iii) (ii) (i) (iv)
2000 4000
%
0
100
18000 16000
34+ 35+
36+
8500 8000
3+
4+
2+
3+
4+
5+
6+
2+
12+
11+
12+
10+
C
m/z
Fig. 2. Lipid binding and its effect on the membrane rings. (Upper panel)
Predominant lipid molecules, one of the six cardioplipins (left) and phos-
phatidylethanolamine (PE, right), found in intact EhATPase and TtATPase,
respectively. (Central panel) Tandem MS of a subcomplex from TtATPase
[ICL
12
E
2
G
2
F (282 740 daltons)] leads to disruption of the L
12
ring, releasing
proteolipids L T PE (red/green circle, 8539 daltons; red circles, 7849 daltons)
and a stripped-complex ICE
2
G
2
F (blue squares, 184 242 daltons). Atomic
structure of the K
10
ring (8) of EhATPase with docking of 10 cardioplipins to
show reduction in the inner diameter (A) and after docking subunit C (B).
Models for sixfold symmetry of the L
12
ring with six PE molecules (green)
(C) and with subunit C (42) (blue) docked into the ring (D). (Lower panel)
schematics of the rotor ring with 12 L subunits each having two TMHs (red
cylinders) and one conserved Glu
63
(yellow) as seen in EM of two-dimensional
crystals of isolated L ring (27) (i). Transformation into a sixfold symmetric ring
[(ii) and (iii)]. (iv) Comparison of the sixfold symmetrical model with EM data
reported previously for the intact TtATPase (iv) (16).
21 OCTOBER 2011 VOL 334 SCIENCE www.sciencemag.org 382
REPORTS
D (Fig. 3B). When ATP is depleted, however,
loss of subunit D before subunit F becomes pos-
sible, leaving a subcomplex with F interacting
directly with the A
3
B
3
hexamer. Extension of
F toward the soluble head has been proposed
previously as a braking mechanism to prevent
unregulated consumption of ATP after in vivo
dissociation of the head from the base in yeast
V-ATPase (28). In line with this proposal, an x-ray
structure and cryo-EM reconstruction of the iso-
lated A
3
B
3
DF complex fromTtATPase and yeast
V-ATPase respectively, also reveal interactions
of subunit F with the A
3
B
3
hexamer (7, 29). An
x-ray structure of the auto-inhibited F
1
head of
the Escherichia coli F-ATPase showed interac-
tions between the subunits analogous to F
1
e
and b in the soluble head (30). Subjecting the
E. coli F
1
complex to the same tandem MS pro-
cedure outlined above showed that subunit e
makes direct interactions with the soluble head
(fig. S14). The similar MS dissociation patterns
observed in E. coli F-ATPase, in which the braking
mechanism is well established, and in TtATPase
suggest an analogous mechanism to prevent
ATP hydrolysis in the uncoupled V
1
complex of
TtATPase.
We observed further sensitivity of TtATPase
to low ATP concentrations, notably loss of sub-
unit I fromICL
12
to formCL
12
(Fig. 3A). Expan-
sion of the peaks assigned to ICL
12
is consistent
with binding of up to six lipids and up to two
nucleotides (ATP or ADP) (fig. S15). This agrees
with proposals that a eukaryotic functional equiv-
alent of subunit I senses cellular nucleotide levels
by binding selectively to ADP and undergoing
conformational change (31). To investigate this
conformational change, we applied ion mobility
MS (IM-MS) (32) to the intact TtATPase, ICL
12
,
and CL
12
subcomplexes
.
Because ions with mul-
tiple conformations result in broad arrival-time
distributions (ATDs), we conclude that the rela-
tively compact ATDs for both the intact ATPase
and CL
12
are consistent with one predominant
conformation (fig. S16 and tables S3 and S4). By
contrast, the ATDs for the ICL
12
complex are
much broader than those of the intact TtATPase
and CL
12
, consistent with multiple conformations
of subunit I in the isolated V
O
complex (Fig. 4
and table S4). The lack of conformational het-
erogeneity in the intact complex is rationalized
by the tethering of subunit I by forces exerted
by the peripheral stalks (subunits E and G), as
suggested recently (13). Once released from the
intact complex, subunit I in V
O
is not constrained,
and flexibility of the hinge domain, located
between the soluble and transmembrane domains,
likely leads to its conformational heterogeneity.
I
D
6000 8000 10000 12000
I
n
t
e
n
s
i
t
y

(
%
)
L
C
G
E
A
F
C
I
L
C
B
F
D
2000 10000 14000 18000 22000
37+
38+
26+
31+
30+
17+
17+
19+
7+
15+
L3+
L3+
7+
15+
m/z
ADP
ATP
ATP
ADP
D
B
A
F
D
A
F
B
D
B
A
F
D
A
B
A
F
B
A
B
F
D
D
B
A
F
D
B
A
F
A
B
D
D
14+
13+
19+
7+
12+
7+
D
FF
L
C
L
C
I
ATP
ADP
+
F
56+
56+
26+
25+
19+
L
I
I L
C
G
E
D
A
F
A B
B
Fig. 3. Nucleotide binding and its effects on intact TtATPase and the
membrane-embedded and soluble head complexes. (A) Depletion of ATP
leads to dissociation of the B subunit fromthe head and subunit I fromICL
12
.
(B) Tandem MS of the soluble head (A
3
B
3
DF) reveals sequential loss of sub-
units F and D in the presence of 50 mM ATP. In TtATPase solutions containing
50 mM ADP, two conformations of subunit F are evident from the bimodal
distribution of charge states formed for the V
1
stripped of F (D7+ and D12+)
together with a direct loss of subunit D. The D7+ series is similar to that
formed from the ATP-bound complex. The D12+ series is consistent with an
extended conformation of subunit F. (Inset) Schematic representation of ef-
fects of ATP/ADP on the membrane ICL
12
complex and the movement of an
extended subunit F in the A
3
B
3
DF complex.
www.sciencemag.org SCIENCE VOL 334 21 OCTOBER 2011 383
REPORTS
The soluble domain of I in EM density maps
of the intact TtATPase is at 90 to the proton
channel (12, 13). Our IMdata for the V
O
complex,
based on modeling of the ICL
12
and CL
12
com-
plexes (fig. S17), are consistent with a range of
conformational states with angles from 90 to
135 (fig. S18). We propose, therefore, that the
conformational dynamics demonstrated here, to-
gether with preferential binding of ADP to the
proposed site near the hinge region (31), destabi-
lize interactions between I and CL
12
as evidenced
by the facile loss of subunit I under low-ATP
conditions.
Given that cellular nucleotide levels likely
affect proton translocation in the isolated V
O
complex, it might also be anticipated that changes
in the proton gradient would induce similar reg-
ulatory effects. To test this hypothesis, we in-
creased the pHof the ATPase-containing solution
to mimic reduction of the proton concentration.
Mass spectra of the TtATPase incubated at pH
9.0 led to subcomplexes formed by loss of IGE
(fig. S19). This observation, together with the
dissociation of subunit I from V
O
, suggests a reg-
ulatory role for subunit I, sensing both proton
and ATP concentrations. Previous proposals in-
voked a locking together of the membrane por-
tion of subunit I with the CL
12
membrane ring,
preventing relative movement and hence explain-
ing the absence of passive H
+
translocation in
isolated V
O
(33, 34). Because we observe facile
loss of subunit I under both lowH
+
concentra-
tion ([H
+
]) and low-[ATP] conditions, and given
the lack of extensive interactions between I and
the membrane ring observed in cryo-EM data
(13), our results point to a mechanism in which
subunit I moves away from the ring, and the re-
sulting gap could then be sealed with membrane
lipids (Fig. 4D).
Our results showthat lipids with two and four
hydrophobic chains associate with subunits with
two and four TMHssubunits L and K, respec-
tively. In both ATPases, the lipids identified in
situ are not the most prevalent ones in the cell
(35, 36), implying that lipids are selected from
the available pool for specific structural roles
and metabolic regulation. This further supports
the proposal that membrane proteins possess
specific lipid binding sites (37) and demon-
strates the ability of lipids to fine tune subunit
interactions by defining the conformations and
inner dimensions of the membrane rings. Our
nucleotide-binding experiments show that a de-
crease in cellular ATP concentrations is sensed
by V
1
, not only with the movement of subunit
F but also with changes in interactions at the A:B
interface. Unexpectedly, V
O
is also sensitive to
low [ATP] and [H
+
], both of which promote dis-
placement of subunit I. We suggest that membrane
lipids subsequently seal the proton-conducting
channel. Consequently, both ATP and proton/ion
gradients are conserved when reversible disso-
ciation takes place in vivo.
The existence of intact rotary ATPases/
synthases in vacuum has been demonstrated
previously with LILBID MS (38, 39). At its cur-
rent resolution, however, it is not possible to
identify bound lipids or nucleotides or to probe
their effects on subunit interactions. Moreover,
previous ES-MS experiments produced well-
resolved mass spectra for the V
1
domain (40, 41),
but lack of an intact V
O
domain, or any interactions
between V
1
and V
O
, is attributed to dissociation
of the complex in the absence of the protective
micelle (18, 19). By contrast, the ES approach
used here enables interrogation of subunit inter-
actions within intact rotary ATPase/synthases
and allows us to probe the synergistic effects of
lipid and nucleotide binding.
References and Notes
1. K. C. Jefferies, D. J. Cipriano, M. Forgac, Arch. Biochem.
Biophys. 476, 33 (2008).
2. C. von Ballmoos, A. Wiedenmann, P. Dimroth, Annu. Rev.
Biochem. 78, 649 (2009).
5000 6000 7000 8000 9000
21+
5
20
15
D
r
i
f
t

t
i
m
e

(
m
s
)
10
7
22
17
D
r
i
f
t

t
i
m
e

(
m
s
)
12
m/z
A
25+
ADP
ATP
E
r
H
+
D
no ATP / with ADP with ATP
membrane
lipids
Drift time (ms)
5 20 35
%
0
100
10.62
10.97
11.49
B
Drift time (ms)
21+
22+
23+
%
0
100
16.05
16.91
17.94
19.48
5 20 35
24+
25+
26+
C
L
C
L
C
I
Fig. 4. Conformational heterogeneity and dissociation of subunit I fromICL
12
implies a mechanism for closing the H
+
channel. (A) IM-MS of the trans-
membrane ICL
12
and CL
12
complexes formed in solution from intact TtATPase.
Charge states used for IM measurement are labeled gray and red (B and C).
Broader arrival-time distributions for ICL
12
than for CL
12
are consistent with
conformational heterogenity in subunit I. (D) Possible mechanism to close
the H
+
channel after lateral movement of subunit I within the membrane.
Surrounding lipids effectively block the channel. The direction of proton
pumping (green) and rotation of the ring (yellow) for ATP hydrolysis, as seen
fromthe top (site of interaction with V
1
). (E) Coarse-grained and atomic model
of the ICL
12
complex generated according to IM restraints and homology
modeling (see supplementary methods). The proposed nucleotide binding
area in the hinge region of I (orange) and Glu
63
:Arg
563
from subunits L and I
are shown in yellow and green, respectively.
21 OCTOBER 2011 VOL 334 SCIENCE www.sciencemag.org 384
REPORTS
3. S. P. Muench, J. Trinick, M. A. Harrison, Quat. Rev. Biophys.
44, 311 (2011).
4. V. Mller, G. Grber, Cell. Mol. Life Sci. 60, 474 (2003).
5. J. P. Abrahams, A. G. Leslie, R. Lutter, J. E. Walker, Nature
370, 621 (1994).
6. K. Okazaki, S. Takada, Structure 19, 588 (2011).
7. N. Numoto, Y. Hasegawa, K. Takeda, K. Miki, EMBO Rep.
25, 25 (2009).
8. T. Murata, I. Yamato, Y. Kakinuma, A. G. Leslie, J. E. Walker,
Science 308, 654 (2005).
9. T. Meier, P. Polzer, K. Diederichs, W. Welte, P. Dimroth,
Science 308, 659 (2005).
10. A. Dautant, J. Velours, M. F. Giraud, J. Biol. Chem. 285,
29502 (2010).
11. I. N. Watt, M. G. Montgomery, M. J. Runswick, A. G. Leslie,
J. E. Walker, Proc. Natl. Acad. Sci. U.S.A. 107, 16823
(2010).
12. L. K. Lee, A. G. Stewart, M. Donohoe, R. A. Bernal,
D. Stock, Nat. Struct. Mol. Biol. 17, 373 (2010).
13. W. C. Lau, J. L. Rubinstein, Proc. Natl. Acad. Sci. U.S.A.
107, 1367 (2010).
14. M. Yamamoto et al., J. Biol. Chem. 283, 19422 (2008).
15. T. Murata et al., J. Biol. Chem. 278, 21162 (2003).
16. R. A. Bernal, D. Stock, Structure 12, 1789 (2004).
17. H. Hernndez, C. V. Robinson, Nat. Protoc. 2, 715 (2007).
18. N. P. Barrera, N. Di Bartolo, P. J. Booth, C. V. Robinson,
Science 321, 243 (2008).
19. N. P. Barrera et al., Nat. Methods 6, 585 (2009).
20. J. L. Benesch, C. V. Robinson, Curr. Opin. Struct. Biol. 16,
245 (2006).
21. K. Yokoyama et al., J. Biol. Chem. 278, 42686 (2003).
22. K. Takase et al., J. Biol. Chem. 269, 11037 (1994).
23. T. Murata, K. Takase, I. Yamato, K. Igarashi, Y. Kakinuma,
J. Biochem. 125, 414 (1999).
24. M. Mandel et al., Proc. Natl. Acad. Sci. U.S.A. 85, 5521
(1988).
25. D. K. Clare et al., Structure 14, 1149 (2006).
26. K. Yokoyama et al., J. Biol. Chem. 275, 13955
(2000).
27. M. Toei et al., Proc. Natl. Acad. Sci. U.S.A. 104, 20256
(2007).
28. P. M. Kane, A. M. Smardon, J. Bioenerg. Biomembr. 35,
313 (2003).
29. Z. L. Hildenbrand, S. K. Molugu, D. Stock, R. A. Bernal,
PLoS ONE 5, e12588 (2010).
30. G. Cingolani, T. M. Duncan, Nat. Struct. Mol. Biol. 18,
701 (2011).
31. A. Armbrster et al., FEBS Lett. 579, 1961 (2005).
32. C. Uetrecht, R. J. Rose, E. van Duijn, K. Lorenzen,
A. J. Heck, Chem. Soc. Rev. 39, 1633 (2010).
33. J. Zhang, M. Myers, M. Forgac, J. Biol. Chem. 267, 9773
(1992).
34. C. Beltrn, N. Nelson, Acta Physiol. Scand. Suppl. 607,
41 (1992).
35. D. Carson, R. A. Pieringer, L. Daneo-Moore, Biochim.
Biophys. Acta 575, 225 (1979).
36. P. H. Ray, D. C. White, T. D. Brock, J. Bacteriol. 108, 227 (1971).
37. C. Hunte, S. Richers, Curr. Opin. Struct. Biol. 18, 406
(2008).
38. T. Meier et al., Mol. Microbiol. 65, 1181 (2007).
39. J. Hoffmann et al., Phys. Chem. Chem. Phys. 12, 13375
(2010).
40. O. Esteban et al., J. Biol. Chem. 283, 2595 (2008).
41. N. Kitagawa, H. Mazon, A. J. Heck, S. Wilkens, J. Biol.
Chem. 283, 3329 (2008).
42. M. Iwata et al., Proc. Natl. Acad. Sci. U.S.A. 101, 59
(2004).
Acknowledgments: Funding from the Wellcome Trust and
by the PROSPECTS (HEALTHF4-2008-201648) grant
within the Research Framework of the European Union
together with funding from the Royal Society (C.V.R.),
the Australian National Health and Medical Research
Council grant 573712 (D.S.), and the FONDECYT
1100515 (N.P.B.) is acknowledged.
Supporting Online Material
www.sciencemag.org/cgi/content/full/334/6054/380/DC1
Materials and Methods
Figs. S1 to S19
Tables S1 to S4
References (4358)
22 June 2011; accepted 15 September 2011
10.1126/science.1210148
Cerebellum Shapes Hippocampal
Spatial Code
Christelle Rochefort,
1
* Arnaud Arabo,
1
* Marion Andr,
2
Bruno Poucet,
2
Etienne Save,
2
* Laure Rondi-Reig
1
*
Spatial representation is an active process that requires complex multimodal integration from a
large interacting network of cortical and subcortical structures. We sought to determine the role
of cerebellar protein kinase C (PKC)dependent plasticity in spatial navigation by recording the
activity of hippocampal place cells in transgenic L7PKCI mice with selective disruption of
PKC-dependent plasticity at parallel fiberPurkinje cell synapses. Place cell properties were
exclusively impaired when L7PKCI mice had to rely on self-motion cues. The behavioral
consequence of such a deficit is evidenced here by selectively impaired navigation capabilities
during a path integration task. Together, these results suggest that cerebellar PKC-dependent
mechanisms are involved in processing self-motion signals essential to the shaping of hippocampal
spatial representation.
I
t is well established that rodents build an
internal cognitive map to navigate in their en-
vironment. A key neural substrate enabling
such representation is the hippocampus, which
contains CA1 and CA3 pyramidal cells described
as place cells. Each place cell fires for a restricted
region (the place field) of the environment (1, 2).
Both external cues and self-motion cues (i.e.,
vestibular, proprioceptive, and optic flow cues)
control place cell firing (3, 4), which suggests the
involvement of a large network of cortical and
subcortical structures interacting with the hippo-
campus for navigation. Determining the function-
al architecture of such a network is thus essential
to our understanding of how the hippocampal
place cell code is generated. The medial ento-
rhinal cortex, a key relay structure between neo-
cortical areas and the hippocampus, contains grid
cells with regularly spaced multiple firing fields
(5), which integrate self-motion information and
participate in path integration (4, 6, 7).
The cerebellum has also been shown to be
essential to the processing of self-motion infor-
mation: Cerebellar Purkinje cells respond to vestib-
ular signals by transforming head-centered vestibular
afferent information into Earth-reference self-
motion and spatial orientation signals (8, 9), and
electrophysiological investigations suggest that
the cerebellumand the hippocampus can be func-
tionally connected during eyeblink conditioning
(10, 11). However, it is still unknown whether
such an interaction is functionally relevant in nav-
igation, and a mechanism that might underlie
such a process has not been identified.
In the transgenic mouse strain L7PKCI, the
pseudosubstrate protein kinase Cinhibitor (PKCI)
is selectively expressed in Purkinje cells under
the control of the pcp-2 (L7) gene promoter (12).
This results in an impaired long-term depression
(LTD) at cerebellar parallel fiberPurkinje cell
synapses. Such a plasticity mechanism has been
proposed to work as an error-based (anti-Hebbian)
learning process (13, 14) during conditioning
tasks (15) and in optimization of motor response
during navigation (16).
A total of 506 dorsal CA1 hippocampal cells
were recorded. A subset of 150 place cells was
further analyzed in six L7PKCI mice and five
wild-type littermate control mice. Relative to wild-
type mice, L7PKCI mice had a significantly lower
proportion of place cells [L7PKCI, n = 53/218
(24.3%); wild type, n = 97/288 (33.7%); c
2
= 5.2,
df = 1, P < 0.025]. Neural activity was sampled
as the mice freely explored a circular arena con-
taining a salient cue (a card with a bottle attached
to it), in standard sessions (S1 and S2) and in-
volving cue manipulation in subsequent sessions
(S3 and S4). A last session (S5) similar to ses-
sions S1 and S2 was run to determine whether
we could restore the initial firing pattern irre-
spective of the changes in cell firing observed
during the cue manipulation sessions (Fig. 1A) (17).
After recording in the standard sessions, we
used two distinct environmental manipulations,
cue removal and cue conflict, in which mice are
forced to use self-motion cues. In the cue removal
condition, the arena was in the dark and the cue
1
Neurobiologie des Processus Adaptatifs (UMR 7102), Naviga-
tion, Memory, and Aging (ENMVI) Team, Universit Pierre et
Marie CurieCentre National de la Recherche Scientifique
(CNRS), F-75005 Paris, France.
2
Laboratory of Neurobiology
and Cognition (UMR 6155), Aix Marseille UniversitCNRS, 3
place Victor Hugo, 13331 Marseille, France.
*These authors contributed equally to this work.
Present address: Laboratoire de Psychologie et Neurosciences
de la Cognition et de lAffectivit (EA 4306), Universit de
Rouen, Facult des Sciences, Place Emile Blondel, 76821
Mont-Saint-Aignan Cedex, France.
Present address: Department of Experimental Neurophys-
iology, Ruhr University Bochum, Universittsstrasse 150, MABF
01/551, 44801 Bochum, Germany.
To whom correspondence should be addressed. E-mail:
laure.rondi@snv.jussieu.fr
www.sciencemag.org SCIENCE VOL 334 21 OCTOBER 2011 385
REPORTS
was removed. A control condition was also per-
formed with the cue still present in darkness (Fig.
1B). In the cue conflict condition, we used a
protocol previously developed in rats, in which
the external cue was rotated 180 in the absence
(hidden rotation) or in the presence (visible
rotation) of the animal, therefore producing a
conflict between visual and self-motion infor-
mation (18). During the conflict, rats maintain
place field stability relative to the standard ses-
sion, thus suggesting the dominant use of self-
motion cues (18).
The basic firing properties of place cells in the
light condition were unaffected in L7PKCI mice
(table S1). In addition, place field stability (mea-
sured as a correlation between two similar light
sessions) was higher in L7PKCI mice (0.78 T 0.05)
than in wild-type mice (0.60 T 0.03) (t
146
= 3.0,
Fig. 1. The compulsory use of self-motion cues affects hippocampal place cell
properties in L7PKCI mice. (A and B) Schematic diagram of the protocol used to
assess the effect of self-motion stimulation on place cell properties. After two
consecutive standard sessions (S1 and S2), light was turned off (S3 and S4) and
objects were either removed (A) or maintained (B) in the arena. S5 was similar to
S1 and S2. (C and D) Examples of color-coded rate maps showing firing activity
of wild-type (WT) and L7PKCI single CA1 pyramidal cells over the five con-
secutive sessions; color coding ranges from blue (silent) to red (peak activity).
Peak firing rates are indicated for each rate map. (E to H) Analysis of place cell
characteristics shows that the suppression of external cue inputs significantly
alters both the mean field rate (E) and spatial coherence (F) in L7PKCI mice
specifically, whereas the suppression of the visual cue alone has no effect [(G)
and (H)]. (I to L) Place field stability, as measured within (I) or across (J) sessions,
is affected in L7PKCI mice after suppression of all external cues, but not after
suppression of the visual cue alone [(K) and (L)]. *P < 0.05, **P < 0.01, ***P <
0.001 with a Newman-Keuls post hoc analysis. Error bars represent SEM.
21 OCTOBER 2011 VOL 334 SCIENCE www.sciencemag.org 386
REPORTS
P = 0.003). In sharp contrast, several firing pa-
rameters were strongly affected in the dark ses-
sions after cue removal in L7PKCI mice (Fig. 1,
C, E, and F, and figs. S1A and S2). The mean
field rate, peak firing rate, and overall mean firing
rate declined during the dark sessions in L7PKCI
mice but not in wild-type mice (mean field rate,
F
1,39
= 11.5, P = 0.002; peak firing rate, F
1,39
=
12.1, P = 0.001; overall mean firing rate, F
1,39
=
8.4, P = 0.006) (Fig. 1E and fig. S2). The field
spatial coherence was also found to be decreased
in the dark (F
1,39
= 11.2, P = 0.002) (Fig. 1F).
Finally, place field stability in L7PKCI mice was
markedly affected by the dark condition: Where-
as wild-type mice maintained place field stabil-
ity throughout the dark sessions, place cells of
L7PKCI mice showed a progressive decrease of
within-session stability (F
1,39
= 15.6, P= 0.0003)
as well as between-session stability (F
2,78
= 6.2,
P= 0.003) (Fig. 1, I and J). These results indicate
that place field stability in L7PKCI mice grad-
ually decreased over sessions in the dark (post
hoc analysis; P < 0.01 between S1-S2 and S2-S3
correlations, P< 0.001 between S1-S2 and S2-S4
correlations) (Fig. 1J).
By contrast, place cell firing properties and
place field stability were restored in session S5
(fig. S3). In most instances, all simultaneously
recorded cells behaved homogeneously (i.e., all
fields were either stable or remapped together).
The modification of place cell properties of
L7PKCI mice was not due to an impaired explor-
atory activity in the dark, because wild-type and
L7PKCI mice displayed similar speed (2.09 T
0.08 cm/s versus 1.97 T 0.05 cm/s, t
19
= 1.7, P >
0.05, t test) and similar traveled distance (14.24 T
0.59 cm versus 13.10 T 0.35 cm, t
19
= 1.4, P >
0.05, t test) (Table 1). When the cue was available
in the dark, the firing parameters and place field
stability were not affected in L7PKCI mice
(Fig. 1B and fig. S1B) (P > 0.05 for all param-
eters analyzed).
These results suggest that in the dark and in
the absence of the cue, the place cell system of
L7PKCI mice failed to use self-motion informa-
tion to maintain stable place fields. Consistent
with this finding, the mice were able to maintain
stable place fields when they could update their
position by using the cue. As a consequence,
the number of place fields away fromthe object
(>20 cm) was drastically reduced in L7PKCI mice
relative to wild-type mice [L7PKCI, n = 1/16
cells (6%); wild type, n = 20/37 cells (46%); c
2
=
10.67, df = 1, P = 0.0011]. The relative power
of the hippocampal theta band (5 to 10 Hz) was
similar in L7PKCI and wild-type mice (F1,21 =
1.72, P > 0.05) in both light and dark cue re-
moval conditions (F1,21 = 3.86, P > 0.05), which
suggests that alteration of path integration was
not caused by a modification of theta rhythm
(fig. S4).
To further investigate the respective influence
of self-motion and external information on spatial
firing pattern in L7PKCI mice, we conducted a
conflict condition protocol (Fig. 2A). After two
standard sessions, a 180 hidden rotation of the
cue resulted in similar rotation of the place fields
in both wild-type and L7PKCI mice (Fig. 2, B
Fig. 2. Field locations are not efficiently controlled by self-motion cues in
L7PKCI mice. (A) Schematic diagram illustrating the protocol used to assess
the effect on place cell firing of a 180 rotation of the cue in the absence
(hidden rotation) or presence (visible rotation) of the mouse in the arena. (B)
Color-coded rate maps showing firing activity of WT and L7PKCI single CA1
pyramidal cells over the five consecutive sessions. (C and D) Histograms
showing the intersession similarity coefficient score associated to a 0 or 180
field rotation after a hidden (C) or a visible (D) rotation of the cue. Field
stability significantly decreased in L7PKCI after a visible rotation of the cue
(D). (E and F) Polar distribution of the place field rotation angles after the
visible rotation in WT mice (E) and L7PKCI mice (F). *P < 0.05, Student t test.
Error bars represent SEM.
Table 1. General sensory-motor abilities of WT and L7PKCI mice in the dark (means T SEM). No
significant differences between WT and L7PKCI mice were revealed by the different sensory-motor tasks
(t test, P > 0.05 for all parameters) assessed in the dark (i.e., using primarily the vestibular system).
Task Measure
L7PKCI
(n = 7)
WT
(n = 6)
Mann-Whitney P
Spontaneous
locomotor activity in
the dark
Speed (cm/s) 1.97 T 0.05 2.09 T 0.08 0.10
Distance traveled (cm) 13.10 T 0.35 14.24 T 0.59 0.09
Rearing frequency
(number/min)
4.29 T 0.73 6.20 T 0.73 0.18
Dynamic balance in
the dark
Falling latency (s) 180 180
Distance traveled (cm) 629 T 77 542 T 51 0.45
Static balance in the dark Falling latency (s) 144 T 18 105 T 17 0.10
Motor coordination in
the dark (rotarod)
5 rpm walking time (s) 141 T 21 138 T 12 0.45
10 rpm walking time (s) 141 T 28 157 T 17 0.63
www.sciencemag.org SCIENCE VOL 334 21 OCTOBER 2011 387
REPORTS
and C, and fig. S1C), indicating that the cue ef-
ficiently controlled place cell activity. Visible ro-
tation of the cue was then performed, producing a
conflict between external and self-motion sen-
sory information. During the conflict, 63% of
place cells in wild-type mice maintained their
place field stability relative to the previous ses-
sion (0 T 30 rotation), which suggests that
the mice resolved the conflict by relying on self-
motion cues (18) (Fig. 2, B, D, and E, and fig.
S1C). The distribution of place field rotation
angles after the visible rotation was therefore
concentrated around the same position (Fig. 2E;
Z =13.53, P< 0.001, Rayleigh test). In contrast, a
majority of place cells in L7PKCI mice exhibited
remapping at a different location, leading to a
homogeneous distribution of place field rota-
tion angles (Fig. 2, B, D, and F; Z = 1.55, P= 0.2,
Rayleigh test): 30% of place fields remained
stable, suggesting a control by self-motion cues;
20% exhibited a 180 T 30 rotation, suggesting
a control by the external cue; and the remaining
50% rotated at various angles. As a result, field
stability (as measured by intersession similarity
coefficient at a rotation angle of 0) between
sessions S4 and S5 was significantly lower in
L7PKCI mice than in wild-type mice (Fig. 2D;
t
54
= 2.0, P < 0.05). The inability to maintain
stable place fields in L7PKCI mice strengthens
the idea of a deficit in the use of self-motion cues.
We next examined the ability of L7PKCI
mice to navigate in the dark (i.e., using self-
motion cues). L7PKCI mice were trained to find
an escape platform at a constant location with a
constant departure point in the water maze (Fig.
3A). Path optimization was analyzed in light
and dark conditions (17) (table S3). In the light,
L7PKCI mice learned to reach the platform as
rapidly and accurately as their control littermates
(Fig. 3, B and C). Both groups increased the use
of direct trajectories across training sessions and
decreased other nonoptimal trajectories (Fig. 3, D
and E). In sharp contrast, navigation performance
in the dark was impaired in L7PKCI mice (Fig. 3,
B and C). Escape latencies and heading were
significantly greater in L7PKCI mice than in
wild-type mice (genotype effect, F
1,28
= 4.98,
P = 0.034, and F
1,28
= 9.63, P = 0.004, respec-
tively), even though there was no difference in
swimming speed (F
1,28
= 2.01, P = 0.2), circling
(F
4,112
= 0.56, P = 0.7) (fig. S5), or other be-
havioral parameters that could interfere with nav-
igation (Table 1 and table S2; P > 0.05) (19).
Thus, assessment of navigation abilities in the
dark demonstrates impaired path integration per-
formances in the L7PKCI mice.
The trajectories of the mutant mice were less
efficient than those of their control littermates in
darkness, as highlighted by the differences in the
type of trajectory used (Fig. 3, D and E). The
importance of the dark context on the deficit ex-
hibited here by the L7PKCI mice was reinforced
by the absence of significant genotype effect
observed during a control trial that took place in
the light condition during dark session 3 (D3T1
in Fig. 3, B and C) (table S3) (17). Accordingly,
comparing this trial with the mean of the last
trial (L5) in the light condition revealed no sig-
nificant differences. This indicates that the dis-
turbed trajectories displayed by transgenic mice
in the dark cannot be attributed either to a deficit
in the use of task rules, or to altered motivation.
The fundamental finding of our study is that
mice lacking PKC-dependent cerebellar LTD
showed disrupted hippocampal place cell proper-
ties and impaired goal-directed navigation in con-
ditions in which self-motion information must be
predominantly used. We previously suggested a
role of PKC-dependent mechanisms in the link-
age between the spatial context and the motor
response characterized by the animals trajectory
(16, 20). Here, we demonstrate an additional and
complementary role of PKC-dependent cerebel-
lar LTD in self-motionbased hippocampal rep-
resentation and path integration. Although the
cerebellumis classically viewed as a motor struc-
ture, a growing body of evidence indicates that
cerebellar circuitry is well suited to act as an
adaptive filter of sensory information (2123). In
particular, vestibular information is combined
with proprioceptive inputs in the cerebellar fas-
tigial nucleus to generate appropriate internal
estimates of the animals self-motion (24). In ad-
dition, cerebellar Purkinje cells from lobules IX
and X transform vestibular head-centered signals
into self-motion and spatial orientation signals
relative to the external world (8, 9). It thus ap-
pears that, beyond its role in motor adaptation
during navigation (16, 20), cerebellar LTD con-
tributes to the representation of the relation of
the body to the external world, thereby shaping
hippocampal spatial representation.
Recent data show a clear contribution of the
vestibular systemto hippocampal-dependent spa-
tial memory (25, 26) as well as to spatial firing of
Fig. 3. Inactivation of PKC-dependent cerebellar LTD deteriorates path integration. (A) Design of the
experimental space developed to evaluate navigation abilities using self-motion cues. (B and C) Quan-
tification of escape latencies (B) and heading (C) in WT and L7PKCI mice during both light and dark
conditions. In the light condition, WT and L7PKCI mice improved their performances significantly over
sessions without genotype effect. In the dark condition, both groups improved their performance over
time, but the performance of LKPCI mice was significantly poorer than that of their control littermates. (D
and E) Swim path analyses during both light and dark conditions. The direct trajectory was significantly
impaired in L7PKCI mice during the dark condition (D). L7PKCI mice cannot perform optimal trajectories
during path integration, as highlighted by the differences between WT and L7PKCI mice in the type of
trajectory used in the dark but not in the light condition (E). The P values indicated in (B) to (D) correspond
to the genotype effect. *P < 0.05 with Newman-Keuls post hoc analysis. Error bars represent SEM.
21 OCTOBER 2011 VOL 334 SCIENCE www.sciencemag.org 388
REPORTS
hippocampal neurons (27). However, examina-
tion of the vestibular-associated motor activity of
L7PKCI mice in both light and dark conditions
revealed no deficit. Our data do not suggest a
vestibular implication underlying the observed
alterations of place cell firing and navigation in
L7PCKI mice. Rather, they demonstrate that cer-
ebellar LTD is also involved in processing self-
motion cues. The cerebellum may therefore
contribute to two major circuits crucial for the
representation of space in the hippocampal sys-
tem. The first is the retrosplenial cortex, which
is closely associated with vestibular function (27).
The second is the parietal cortex, which integrates
self-motion and external information and receives
input from the deep cerebellar nuclei (28, 29).
Our study demonstrates the crucial role of PKC-
dependent cerebellar LTD in the preprocessing of
self-motion information required for optimal hip-
pocampal representation. This process appears to
be essential for path integration.
References and Notes
1. H. Eichenbaum, Curr. Biol. 10, R785 (2000).
2. J. OKeefe, L. Nadel, The Hippocampus as a Cognitive
Map (Clarendon, Oxford, 1978).
3. P. J. Best, A. M. White, A. Minai, Annu. Rev. Neurosci. 24,
459 (2001).
4. K. J. Jeffery, Hippocampus 17, 775 (2007).
5. T. Hafting, M. Fyhn, S. Molden, M. B. Moser, E. I. Moser,
Nature 436, 801 (2005).
6. B. L. McNaughton, F. P. Battaglia, O. Jensen, E. I. Moser,
M. B. Moser, Nat. Rev. Neurosci. 7, 663 (2006).
7. F. Sargolini et al., Science 312, 758 (2006).
8. T. A. Yakusheva et al., Neuron 54, 973 (2007).
9. D. E. Angelaki, T. A. Yakusheva, A. M. Green,
J. D. Dickman, P. M. Blazquez, Cerebellum 9, 174 (2010).
10. L. C. Hoffmann, S. D. Berry, Proc. Natl. Acad. Sci. U.S.A.
106, 21371 (2009).
11. J. Wikgren, M. S. Nokia, M. Penttonen, Neuroscience
165, 1538 (2010).
12. C. I. De Zeeuw et al., Neuron 20, 495 (1998).
13. J. S. Albus, Math. Biosci. 10, 25 (1971).
14. D. Marr, J. Physiol. 202, 437 (1969).
15. M. Ito, M. Kano, Neurosci. Lett. 33, 253 (1982).
16. E. Burguire et al., Nat. Neurosci. 8, 1292 (2005).
17. See supporting material on Science Online.
18. A. Rotenberg, R. U. Muller, Philos. Trans. R. Soc.
London Ser. B 352, 1505 (1997).
19. L. Rondi-Reig et al., Neuroscience 104, 207 (2001).
20. E. Burguire, A. Arabo, F. Jarlier, C. I. De Zeeuw,
L. Rondi-Reig, J. Neurosci. 30, 13265 (2010).
21. S. Pasalar, A. V. Roitman, W. K. Durfee, T. J. Ebner,
Nat. Neurosci. 9, 1404 (2006).
22. C. C. Bell, V. Han, N. B. Sawtell, Annu. Rev. Neurosci. 31,
1 (2008).
23. P. Dean, J. Porrill, C. F. Ekerot, H. Jrntell,
Nat. Rev. Neurosci. 11, 30 (2010).
24. J. X. Brooks, K. E. Cullen, J. Neurosci. 29, 10499
(2009).
25. Y. Zheng, M. Goddard, C. L. Darlington, P. F. Smith,
Hippocampus 19, 480 (2009).
26. M. Goddard, Y. Zheng, C. L. Darlington, P. F. Smith,
Behav. Neurosci. 122, 448 (2008).
27. R. W. Stackman, A. S. Clark, J. S. Taube, Hippocampus
12, 291 (2002).
28. S. Giannetti, M. Molinari, Brain Res. Bull. 58, 481
(2002).
29. D. M. Clower, R. A. West, J. C. Lynch, P. L. Strick,
J. Neurosci. 21, 6283 (2001).
Acknowledgments: We thank C. De Zeeuw for providing
L7PKCI mice; S. Quet and F. Jarlier for technical
supports; C. Lamouroux for animal care; G. Dallrac
and K. Benchenane for helpful comments on the
manuscript and programming; and C. Lna for
scientific insights. Supported by grants from Agence
Nationale de la Recherche (ANR Young Researcher
07-JCJC-0108-01 and ANR-09-EMER-005-02)
(L.R.R.) and by Action Concerte Incitative (ACI) grant
NIC0027 (B.P. and E.S.).
Supporting Online Material
www.sciencemag.org/cgi/content/full/334/6054/385/DC1
Materials and Methods
Figs. S1 to S5
Tables S1 to S3
References
22 April 2011; accepted 13 September 2011
10.1126/science.1207403
Activity-Dependent Long-Term
Depression of Electrical Synapses
Julie S. Haas,
1,2
* Baltazar Zavala,
2
Carole E. Landisman
1,2
*
Use-dependent forms of synaptic plasticity have been extensively characterized at chemical
synapses, but a relationship between natural activity and strength at electrical synapses remains
elusive. The thalamic reticular nucleus (TRN), a brain area rich in gap-junctional (electrical)
synapses, regulates cortical attention to the sensory surround and participates in shifts between
arousal states; plasticity of electrical synapses may be a key mechanism underlying these processes.
We observed long-term depression resulting from coordinated burst firing in pairs of coupled
TRN neurons. Changes in gap-junctional communication were asymmetrical, indicating that
regulation of connectivity depends on the direction of use. Modification of electrical synapses
resulting from activity in coupled neurons is likely to be a widespread and powerful mechanism
for dynamic reorganization of electrically coupled neuronal networks.
T
he thalamic reticular nucleus (TRN) is
a shell comprising a homogenous pop-
ulation of parvalbumin (PV)positive g-
aminobutyric acid(GABA)releasing(GABAergic)
neurons surrounding the dorsal thalamus (1, 2).
These cells provide powerful inhibition to thala-
mocortical relay neurons (3) upon integration of
their corticothalamic and thalamocortical inputs.
In addition to its proposed role in focusing the
neural spotlight of attention (4, 5), the TRN is
strongly involved in regulating states of arousal
(6, 7) by means of alternation between burst
and tonic modes of firing. Burst firing in the
TRN is a prominent component of sleep spin-
dles (8, 9) and absence seizures (9, 10), both of
which are marked by dramatic changes in cortical
attention and behavioral responsiveness to sen-
sory input.
In central mammalian neurons, electrical
(gap-junctional) synapses appear all over the
brain (11, 12) and mainly couple GABAergic
neurons of similar subtype (1315). Electrical
synapses contribute to synchrony in coupled net-
works (11, 1621), although computational studies
suggest that the precise role of gap junctions in
synchrony can be complex (2224).
Cells in the TRN are densely and powerfully
connected by electrical synapses (17, 18) that
persist into adulthood (25) and, as in other areas,
participate in its synchronous activity (18). The
experimentally isolated TRN generates spindle
rhythms in the absence of other inputs (26), sug-
gesting that electrical synapses are likely to be
key players in TRN synchrony and in behavioral
switching between firing states.
Activity-dependent forms of plasticity have
been extensively described at excitatory (gluta-
matergic) chemical synapses (27, 28) and, to a
lesser extent, at inhibitory (GABAergic) chemi-
cal synapses (2931). Although the issue has
received far less attention than plasticity of chem-
ical synapses, modifications of electrical synapses
have been documented in a handful of reports
(32, 33). Because electrical synapses are likely to
play a major role in coordinating TRN activity,
we sought to investigate the effects of natural
forms of activity in coupled neurons on the strength
of the electrical synapses between them.
We recorded from pairs of gap junction
coupled TRN neurons (Fig. 1A) within con-
ventional thalamocortical brain slices (34). To
measure electrical synaptic strength, we delivered
hyperpolarizing current injections into one neu-
ron (cell 1) while recording voltage (V) responses
in both neurons, which were maintained at a base-
line V
m
= 65 mV (Fig. 1B). Using these deflec-
tions, we determined the coupling coefficient cc
12
=
DV
cell 2
/DV
cell 1
, and frominjecting current into cell
2, similarly determined cc
21
= DV
cell 1
/DV
cell 2
. We
also calculated coupling conductance G
C
(34) in
each direction. From a total of 313 paired record-
ings of coupled TRN neurons, we found an av-
erage cc of 0.12 T 0.08 and G
C
of 0.80 T 0.63 nS
(mean T SD) (Fig. 1C), which is in line with the
values for previous reports in TRN (17, 18, 33)
1
Childrens Hospital, Department of Neurology, Harvard Uni-
versity, 300 Longwood Avenue, Boston, MA 02115, USA.
2
Center for Brain Science, Harvard University, 52 Oxford Street,
Cambridge, MA 02138, USA.
*To whom correspondence should be addressed. E-mail:
julie.haas@gmail.com (J.S.H.); carole.landisman@hms.
harvard.edu (C.E.L.)
www.sciencemag.org SCIENCE VOL 334 21 OCTOBER 2011 389
REPORTS
and of similar size to cc values reported in the
cortex and other areas (3537). Other parameters,
such as intraneuronal distance and probability of
coupling, were similar to previous reports (18).
Asymmetry of electrical synapses has been
observed experimentally (36, 37). In our data,
coupling was rarely symmetrical, which is shown
by the spread of values when plotting cc
21
against
cc
12
for each pair (Fig. 1D). We quantified asym-
metry by the ratio of directional ccs (cc
21
/cc
12
)
for each pair; for all pairs, the mean ratio of ccs
was 1.6 T 0.6 (n = 313 pairs) (Fig. 1E). Some of
Fig. 1. (A) Magnification 60 infrared image from patch
recordings of a coupled pair of TRN neurons. (B) Current
injection into one cell (I
1
) of a coupled pair drives a direct
response in that cell (V
1
) and a gap junctionrelayed
response in the second cell (V
2
); cc
12
= DV
2
/DV
1
. Scale bars,
5 mV, 0.1 s. (C) Mean electrical synaptic conductance (G
C
)
plotted against mean cc (dots). Open circles are binned
averages, with a slope of 7.9 [bin width, 0.02; coefficient of
determination (r
2
) = 0.77]. (D) Directional cc (purple, scaled
by 10) and G
C
(orange) for each pair; 12 represents
coupling measured by current injection into cell 1, as in (B).
(E) Coupling asymmetry was quantified by distribution of
ratios (cc
12
/cc
21
and G
12
/G
21
, larger value/smaller; bin
width, 0.05). (F) Spikes driven by current injection into one
cell (gray) caused spikes in the unstimulated coupled cell
(black), as shown for three pairs with cc between 0.2 and 0.4
maintained at baseline V
m
65 mV. Scale bar, 25 mV, 0.1 s.
(G) Wide-field image of TRN cells loaded with OGB-Bapta
1AM (Invitrogen, Carlsbad, California). (H) Stimulation of a
patched cell (gray) drove bursting and strong calcium re-
sponses in that cell and in several neighboring cells (scale
bars, 1% DF/F, 50 ms and 25 mV, 50 ms for bottom trace).
Traces are from the cells labeled by color and number in (G).
I
1
V
2
V
1
0 2 4
0
2
4
12
2

1
10cc
G
C
0 0.15 0.3
0
1
2
3
mean cc
m
e
a
n

G
C

(
n
S
)
1 2 3
0
0.1
0.2
ratio
d
e
n
s
i
t
y
cc
G
C
A B C D E
F
G H
1
2
3
4
5
6
7
8
Fig. 2. (A) Paired bursting driven by simultaneous
current injections into both cells of coupled pairs.
Scale bars, 20 mV, 50 ms. (Inset) Close-up of paired
burst event. (B) Mean cc and G
C
before and after
paired bursting (gray bar). (C) Average normalized
input resistance (R
in
) and membrane potential (V
m
) for
the neurons summarized in (B). (D) Example paired
responses before and after activity pairing as in (A).
Scale bars, 100 ms, 2.5 mV (coupled response, in
black), 5 mV (direct response, in gray). (E) Bursting
driven by current injections into one cell of a coupled
pair (gray trace) while the other neuron was quiescent
(black trace). Scale bars, 20 mV, 50 ms. (Inset) Close-
up of burst in cell 1 and burstlet in cell 2. (F) Mean cc
and G
C
before and after single-cell bursting (gray bar).
(G) Average normalized input resistance (R
in
) and
membrane potential (V
m
) for the neurons summarized
in (F). (H) Example paired responses before and after
activity pairing as in (E). Scale bars, 100 ms, 2.5 mV
(coupled response, in black), 5 mV (direct response,
in gray).
10 0 10 20
0.8
1
1.2
Elapsed Time (min)
R
i
n
,

V
m

(
n
o
r
m
)
10 0 10 20
0.8
0.9
1
1.1
G
C
,

c
c

(
n
o
r
m
)
A
B
C
D
10 0 10 20
0.8
1
1.2
Elapsed Time (min)
R
i
n
,

V
m

(
n
o
r
m
)
10 0 10 20
0.8
0.9
1
1.1
G
C
,

c
c

(
n
o
r
m
)
E
F
G
H
control
post
activity
control
post
activity
21 OCTOBER 2011 VOL 334 SCIENCE www.sciencemag.org 390
REPORTS
the observed asymmetry in cc is due to mis-
matches in input resistance; however, ratios of
directionally measured G
C
(G
21
/G
12
), which are
independent of input resistance (34), had a mean
of 1.2 T 0.27 (Fig. 1E).
Like many thalamic neurons, TRN neurons
spike in two modes: conventional fast sodium-
based tonic spikes and slower low-threshold cal-
cium spikes (LTS), known as bursts, that are
crowned by a barrage of fast sodium spikes. In
many pairs in the current study, bursts elicited
by positive current injection into one neuron were
sufficient to drive bursts in its coupled neighbor
(Fig. 1F). Imaging experiments revealed that
bursting activity driven in a patched cell propa-
gated through a network of coupled cells (Fig. 1,
G and H).
To determine the effects of bursting in cou-
pled cells on electrical synaptic strength, we
tested coupling strength before and after 5 min of
synchronous evoked bursting in pairs of coupled
neurons. Bursting was driven by simultaneous cur-
rent injections of 100 to 300 pAfor 50 ms at 2 Hz
through the recording electrodes of both neu-
rons, which were maintained at membrane poten-
tials between 65 and 70 mV by means of
steady-state current injection (Fig. 2A). After paired
bursting, cc was reduced by 12.0 T 3.6%, and G
C
was depressed by 13.2 T 1.8% (P < 0.05, two-
tailed unpaired t test, n = 7 pairs) (Fig. 2B). This
long-term depression (LTD) persisted for the
length of recordings (for at least 30 min after
paired bursting), with no apparent signs of di-
minishing (Fig. 2B). There were no significant
changes in input resistance or membrane resting
potential (Fig. 2C), ruling out the possibility that
the observed changes in electrical synaptic strength
reflected changes in the intrinsic properties of the
neurons at the whole-cell level. Although path-
ological changes in internal calcium concentra-
tion are known to affect gap-junctional strength
(32, 38), our estimates of the calciuminflux from
the slow rate of bursting used here are much
smaller. Bursting rates in vivo are often faster
than those used here (19, 26).
To determine whether bursting in one neuron
alone is sufficient to induce LTD, we repeated the
activity paradigm, this time only stimulating burst-
ing in a single neuron of a pair (Fig. 2E) while
holding the coupled cell at ~70 mV so as to
prevent it frombursting. After single-cell bursting,
cc was reduced by 15.0 T 3.4%, and G
C
was re-
duced by 13.0 T 2.3% (P < 0.05, n = 11 pairs)
(Fig. 2F). The magnitude of LTD was not signif-
icantly different for the single-cell burst paradigm
fromthe paired-bursting paradigm(unpaired t test).
To determine the contribution of sodium
spikes to LTD, we repeated the bursting para-
digm in both cells using a bath application of
1 mM tetrodotoxin (TTX), which completely and
reversibly blocks the quick barrage of sodium-
mediated action potentials crowning the calcium-
mediated bursts (Fig. 3A). After paired bursting
in TTX, cc decreased by 12.3 T 3.2%, and G
C
decreased by 11.7 T 2.6% (P < 0.05, n = 9 pairs)
(Fig. 3B). We also repeated the bursting para-
digm in one cell alone in TTX (Fig. 3D). After
single-cell activity in TTX, cc decreased by 6.5 T
2.3%, and G
C
was reduced by 6.0 T 2.0% (P <
0.05, n = 11 pairs) (Fig. 3E). When depolarized to
rest just below spiking threshold (~40 mV) and
stimulated to spike with 50-ms pulses repeated at
2 Hz, in order to emulate spiking during bursting
without activating the LTS, coupling decreased
by a smaller and delayed amount (DG
C
= 7.2 T
2.0%, Dcc = 7.0 T 2.8%; P = 0.03, n = 8 pairs;
spike frequency during this paradigmwas twice as
slow as during LTS bursts) (fig. S2). Of these
activity paradigms, the amount of depression
fromsingle-cell bursting in TTXwas significantly
smaller than others [P < 0.05, analysis of variance
(ANOVA)] (Fig. 3H).
Activity paradigms in which only one cell
was active allowed us to characterize the time
course of changes in electrical synaptic strength
by measuring the amplitude of the postsynaptic
burstlet in the coupled cell during the 5 min of
Fig. 3. (A) Paired bursting driven by simultaneous
current injections into both cells of coupled pairs, in the
presence of 1 mM TTX. Scale bars, 10 mV, 50 ms. (Inset)
Close-up of paired burst events. (B) Mean cc and G
C
before and after paired bursting in TTX (gray bar). (C)
Average normalized input resistance (R
in
) and mem-
brane potential (V
m
) for the neurons summarized in (B).
(D) Bursting driven by injections of current into one cell
of a coupled pair (gray trace) while the other neuron
was quiescent (black trace), also in TTX. Scale bars, 10
mV, 50 ms. (Inset) Close-up of burst event and burstlet.
(E) Mean cc and G
C
before and after single-cell bursting in
TTX (gray bar). (F) Average normalized input resistance
(R
in
) and membrane potential (V
m
) for the neurons
summarized in (E). (G) Burstlet amplitudes (from Fig.
2E) during single-cell activity plotted against elapsed time
and normalized to final values. (H) Summary of changes
in G
C
for the four paradigms: paired bursting (2B), single-
cell bursting (1B), paired bursting in TTX (2B + T), and
single-cell bursting in TTX (1B + T). Asterisk indicates
significance (P < 0.05, ANOVA).
10 0 10 20
0.8
1
1.2
Elapsed Time (min)
R
i
n
,

V
m

(
n
o
r
m
)
10 0 10 20
0.8
0.9
1
1.1
G
C
,

c
c

(
n
o
r
m
)
A
B
C
G
10 0 10 20
0.8
1
1.2
Elapsed Time (min)
R
i
n
,

V
m

(
n
o
r
m
)
10 0 10 20
0.8
0.9
1
1.1
G
C
,

c
c

(
n
o
r
m
)
D
E
F
H
0 1 2 3 4
1
1.1
1.2
Elapsed time during pairing (min)
B
u
r
s
t
l
e
t
A
m
p
l
i
t
u
d
e
(
n
o
r
m
)
2B 1B 2B+T 1B+T
20
10
0

G
C

(
%
)
www.sciencemag.org SCIENCE VOL 334 21 OCTOBER 2011 391
REPORTS
activity. For both single-cell bursting and single-
cell bursting in TTX, changes in synaptic strength
(burstlet amplitude) reached their steady-state
reduced values within 2 min of activity (Fig. 3G).
In two of our activity paradigms, the activity
of the coupled pair, and thus the use of the syn-
apse, was also asymmetrical (Figs. 2E and 3D)
that is, one neuron was active while the other was
quiescent, resulting in largely unidirectional cur-
rent flow across the gap junction channels during
activity. These asymmetrical stimuli allowed us to
investigate whether the LTD was also expressed
asymmetrically. First, we quantified the effects of
activity on each direction of coupling, with re-
spect to the active cell. Coupling measured with
current injection into cell 1 (the active cell during
pairing), or outbound coupling, we denote as cc
12
,
whereas coupling measured with current injection
into the quiet cell 2 and relayed by the gap
junction back to the active cell 1, or inbound
coupling, is cc
21
(Fig. 4A). For full bursting in
one neuron (Fig. 4B), the inbound coupling cc
21
decreased by 16.0 T 3.4%, whereas outbound
coupling, cc
12
, decreased by 8.6 T 3.7%(P < 0.05
for both directions; two-tailed, paired t test, n = 11
pairs) (Fig. 4, C and D). The change in cc
21
was
significantly larger than in cc
12
(P < 0.05). Di-
rectional conductances decreased similarly; G
21
decreased by 10.8 T 3.2%, and G
12
decreased by
6.8 T 3.2% (P < 0.05). For single-cell LTS burst-
ing in TTX (Fig. 4D), inbound coupling, cc
21
, de-
creased by 10.0 T 3.0% (P < 0.05, n = 10 pairs),
whereas the change in outbound coupling, cc
12
,
was not significant (5.5 T 2.7%, P= 0.07, n = 10
pairs) (Fig. 4G and H). In TTX, outbound G
12
decreased by 7.5 T 2.0% (P = 0.04), and G
21
decreased by 6.6 T 2.5% (P = 0.09).
In principle, asymmetrical use of a gap junc-
tion could potentially act to either decrease, in-
crease, or preserve the pre-activity asymmetry of
coupling in any given pair. To examine the sys-
tematic effects of unidirectional synapse use on
asymmetry, we plotted the ratios of directional
ccs and G
C
s (cc
21
/cc
12
and G
21
/G
12
) for each pair
after unidirectional activity (Fig. 4, E and I, y
axis) against the initial values (Fig. 4, E and I, x
axis). The identity line corresponds to coupling
asymmetry that was unaffected by asymmetrical
use of the synapse. For full bursts in one cell,
ratios of ccs increased on average by 9.1 T 2.4%
after activity (P < 0.01, n = 11 pairs) (Fig. 4E);
this shift represents a systematic trend of greater
change in the coupling of inbound communica-
tion, cc
21
, relative to outbound communication,
cc
12
. Ratios of G
C
also increased, by 5.0 T 2.2%
(P < 0.05). Changes in asymmetry were not due
to coordinated shifts in input resistance; R
1
/R
2
decreased by 3.1 T 2.6% (P = 0.25). For LTS
bursts without sodiumspikes in one cell, ratios of
ccs and G
C
s fell along the identity line after ac-
tivity, with an insignificant change in rectification
from initial values (ratio of ccs: 6.0 T 4.2%, P =
0.6; ratio of G
C
s: 0.6 T 2.0%, P = 0.76; R
1
/R
2
:
5.3 T 3.4%, P = 0.07; n = 10 pairs) (Fig. 4I),
indicating that the changes in rectification may
be due to sodium spikes. As expected, ratios of
coupling coefficients also did not change sig-
nificantly for symmetrical synaptic use (paired
bursting).
Although activity-dependent changes have
been extensively described and characterized at
chemical synapses, long-term modification of
electrical synapses by precise patterns of activity
of coupled cells themselves has not yet been de-
scribed. The changes we measured, ~15%, are
small as compared with some changes measured
at chemical synapses. Neurons receive thousands
of individual chemical synaptic inputs, which are
each very small, often distant from the soma, and
of short, stereotyped time courses. Chemical syn-
aptic inputs are orders of magnitude smaller than
are electrical synaptic inputs and typically in-
effective as single voices in driving a cell to spike.
The average coupling measured here (cc = 0.12)
Fig. 4. (A) For activity in cell 1, cc
12
(blue) represents the outbound
coupling measured with current injection into cell 1, and cc
21
(green) rep-
resents inbound coupling. (B) Single-cell bursting in cell 1 (gray) with
postsynaptic burstlets in cell 2. Scale bars, 15 mV, 25 ms. (C) Inbound cc
21
before and after full bursts in cell 1. (D) Outbound cc
12
before and after full
bursts in cell 1. (E) Ratios of directional cc [black solid circles; division of the
changes in (C) divided by the changes in (D) for each pair] and G
C
(open
circles, P < 0.05 for both cc and G
C
) after full bursts in cell 1, plotted against
initial values. (F) Bursts in cell 1 (gray) in 1 mM TTX. Scale bars, 10 mV, 25 ms.
(G) Inbound cc
21
before and after bursts in cell 1 in TTX. (H) Outbound cc
12
before and after bursts in cell 1 in TTX. (I) Ratios of directional cc (red solid
squares; P = 0.6) and G
C
(open squares; P = 0.76) after bursts in cell 1 in TTX,
plotted against initial values. ( J) Model of an asymmetrical gap junction as
two parallel branches. R
C
represents the minimum conductance (maximum
resistance) common to both sides of the gap junction, and R
D
represents
additional, asymmetrical conductance in one direction.
21 OCTOBER 2011 VOL 334 SCIENCE www.sciencemag.org 392
REPORTS
applied to an average presynaptic burst (~50 mV)
yields a ~6-mV burstlet in a coupled cell, which
persists for the entire ~50 ms of the burst (Figs.
1F, 2E, and 3D); a single burstlet is often large
enough to drive bursts directly in a coupled neigh-
bor (Fig. 1). Areduction by <15%is considerable
for these already strong synapses and is sufficient
to prevent a cells burstlet from driving its neigh-
bor to burst (fig. S1).
By preferentially diminishing coupling inbound
to bursting cells, activity-dependent LTD could
unplug single bursting cells froman overly active
or synchronous neighbor or network or adjust
input preference between intra-TRN electrical
input and input from corticothalamic or thalamo-
cortical fibers. The effects of activity-dependent
changes may be more complex in vivo because
of multiple electrical synapses and/or recurrent
synapses between neurons.
What cellular processes might underlie the ob-
served LTDof electrical synapses? Gap junctions
are plaques comprising hundreds to thousands
of individual channels. Insertion and deletion of
gap junction channels is a normal component of
cellular function and a candidate mechanism for
changing synaptic strength. In addition, con-
nexin36 (Cx36) proteins have multiple phos-
phorylation sites (39, 40). Phosphorylation-related
changes in coupling mediated by either protein
kinase A(41, 42) or CamKII (40) as well as hemi-
channel conductance changes at Cx35 channels
(43) have been described.
Our experiments indicate that electrical syn-
aptic strength is asymmetrical at baseline and is
further adjustable in a use-directional manner. Re-
sults in mice in which Cx36 has been knocked
out indicate that synapses composed of non-
Cx36 proteins are more asymmetrical than those
in wild-type (44); thus, one possible source of
asymmetry is inclusion of non-Cx36 proteins
and/or pores at the synapse. Our results further
indicate that coupling asymmetry can be shifted
by activity; neurons can fine-tune the relative
proportion of signals they send or receive to or
from coupled neighbors, respectively. Increased
expression or activation of non-Cx36 proteins
could account for this increase in asymmetry.
Despite evidence of gap-junctional rectifica-
tion in mammalian systems, the canonical sym-
bol for those electrical synapses has remained the
simple linear resistor (R
C
). Our observations of
baseline asymmetry and activity-dependent shifts
in asymmetry (Fig. 4E) led us to reconsider the
standard model because a linear resistor cannot
account for asymmetry or increases in asym-
metry. Diodes have been used to model heavily
rectifying invertebrate gap junctions (45) but have
not yet been considered for mammalian gap-
junctional synapses. We suggest a model of a
mammalian gap-junctional synapse as two branches
in parallel (Fig. 4J): One branch carries the com-
mon resistance (R
C
) or the maximum resistance
(minimum of conductance G
C
) measured from
both directions. A parallel branch consists of a
resistor (R
D
) in series with a diode, representing
the increase in conductance (or decreased re-
sistance) observed as asymmetry.
References and Notes
1. C. R. Houser, J. E. Vaughn, R. P. Barber, E. Roberts,
Brain Res. 200, 341 (1980).
2. P. T. Ohara, A. R. Lieberman, J. Neurocytol. 14, 365 (1985).
3. D. Pinault, M. Deschnes, J. Comp. Neurol. 391, 180
(1998).
4. F. Crick, Proc. Natl. Acad. Sci. U.S.A. 81, 4586 (1984).
5. K. McAlonan, J. Cavanaugh, R. H. Wurtz, J. Neurosci. 26,
4444 (2006).
6. D. L. Rowe et al., Int. J. Neurosci. 115, 1273 (2005).
7. M. Steriade, D. A. McCormick, T. J. Sejnowski, Science
262, 679 (1993).
8. M. Deschnes, M. Paradis, J. P. Roy, M. Steriade,
J. Neurophysiol. 51, 1196 (1984).
9. M. Steriade, Trends Neurosci. 28, 317 (2005).
10. M. Inoue, J. Duysens, J. M. Vossen, A. M. Coenen,
Brain Res. 612, 35 (1993).
11. M. V. Bennett, R. S. Zukin, Neuron 41, 495 (2004).
12. B. W. Connors, M. A. Long, Annu. Rev. Neurosci. 27,
393 (2004).
13. M. Galarreta, S. Hestrin, Nat. Rev. Neurosci. 2, 425
(2001).
14. J. R. Gibson, M. Beierlein, B. W. Connors, Nature 402,
75 (1999).
15. Z. Chu, M. Galarreta, S. Hestrin, J. Neurosci. 23, 96
(2003).
16. M. Beierlein, J. R. Gibson, B. W. Connors, Nat. Neurosci.
3, 904 (2000).
17. C. E. Landisman et al., J. Neurosci. 22, 1002 (2002).
18. M. A. Long, C. E. Landisman, B. W. Connors, J. Neurosci.
24, 341 (2004).
19. P. Fuentealba, M. Steriade, Prog. Neurobiol. 75, 125
(2005).
20. M. Blatow et al., Neuron 38, 805 (2003).
21. G. Tams, E. H. Buhl, A. Lrincz, P. Somogyi,
Nat. Neurosci. 3, 366 (2000).
22. K. Vervaeke et al., Neuron 67, 435 (2010).
23. C. C. Chow, N. Kopell, Neural Comput. 12, 1643
(2000).
24. T. J. Lewis, J. Rinzel, J. Comput. Neurosci. 14, 283
(2003).
25. K. L. Blethyn, S. W. Hughes, V. Crunelli,
Thalamus Relat. Syst. 4, 13 (2008).
26. M. Steriade, L. Domich, G. Oakson, M. Deschnes,
J. Neurophysiol. 57, 260 (1987).
27. R. C. Malenka, M. F. Bear, Neuron 44, 5 (2004).
28. R. C. Froemke, D. Debanne, G.-Q. Bi, Front. Syn. Neurosci.
2, 19 (2010).
29. K. P. Lamsa, D. M. Kullmann, M. A. Woodin,
Front. Syn. Neurosci. 2, 8 (2010).
30. F. S. Nugent, J. A. Kauer, J. Physiol. 586, 1487 (2008).
31. J. S. Haas, T. Nowotny, H. D. Abarbanel, J. Neurophysiol.
96, 3305 (2006).
32. A. E. Pereda, D. S. Faber, J. Neurosci. 16, 983 (1996).
33. C. E. Landisman, B. W. Connors, Science 310, 1809
(2005).
34. Materials and methods are available as supporting
material on Science Online.
35. J. R. Gibson, M. Beierlein, B. W. Connors, J. Neurophysiol.
93, 467 (2005).
36. A. Devor, Y. Yarom, J. Neurophysiol. 87, 3048 (2002).
37. P. Mann-Metzer, Y. Yarom, J. Neurosci. 19, 3298
(1999).
38. B. Rrig, B. Sutor, Mol. Neurobiol. 12, 225 (1996).
39. W. W. Kothmann, X. Li, G. S. Burr, J. OBrien,
Vis. Neurosci. 24, 363 (2007).
40. C. Alev et al., Proc. Natl. Acad. Sci. U.S.A. 105, 20964
(2008).
41. S. Urschel et al., J. Biol. Chem. 281, 33163 (2006).
42. W. W. Kothmann, S. C. Massey, J. OBrien, J. Neurosci.
29, 14903 (2009).
43. G. Mitropoulou, R. Bruzzone, J. Neurosci. Res. 72, 147
(2003).
44. T. Zolnik, B. W. Connors, presented at the Society for
Neuroscience Conference, Prog. 42.14. San Diego, CA, 2010.
45. E. J. Furshpan, D. D. Potter, J. Physiol. 145, 289
(1959).
Acknowledgments: We thank D. Feldman and K. Bender
for valuable feedback on previous version of the
manuscript. This work was supported by the Milton
Fund.
Supporting Online Material
www.sciencemag.org/cgi/content/full/334/6054/389/DC1
Materials and Methods
Figs. S1 and S2
26 April 2011; accepted 16 August 2011
10.1126/science.1207502
www.sciencemag.org SCIENCE VOL 334 21 OCTOBER 2011 393
REPORTS
Weekly Cell Signaling Journal fromScience
ScienceSignaling
As a AAASmember, add Science Signaling access for over
60%off regular price. Subscribe in any of these ways:
Go to sciencesignaling.org
Call +1 202-326-6417
Mail or fax this formwith your payment to +1 202-842-1065
1200 New York Ave., NW Washington, DC 20005 +1 202-326-6417 sciencesignaling.org
I AAASmember price US$52 online only;
US$225 print and online (add US$100 for non-US delivery)
AAAS membership number required ______________________________
I Nonmember price US$156 online only;
US$465 print and online (add US$100 for non-US delivery)
Name ___________________________________________
Address _________________________________________
_________________________________________________
City _____________________________________________
State/Province____________________________________
Zip/Postal Code ___________________________________
Country __________________________________________
E-mail____________________________________________
Phone____________________________________________
Online resources on cell signaling
Payment
ICheck (payable to AAAS Science Signaling)
Mail check and this formto:
AAAS
Attn: Membership Department
1200 NewYork Avenue, NW
Washington, DC 20005 USA
ICharge my:
I VISA I MasterCard I American Express
Card Number_________________________________________
Expiration Date_______________________________________
Signature____________________________________________
Date ________________________________________________
If paying by credit card, you may FAST FAX your order to
+1 202-842-1065
Subscribe nowfor 1 years access to Science Signaling
(required for subscription activation)
Gain insight into the fast-paced research in cell signaling with
original reviews and perspectives by leading researchers.
Stay abreast of the latest developments with summaries of
the weeks hottest research and the nightly updatedVirtual
Journal, with primary research articles from49 publishers.
Give your own research a boost with detailed protocols that
guide you through the latest techniques.
Learn about the relationships controlling cell behavior from
the Connections Maps pathways, a graphical interface into the
Database of Cell Signaling, with information provided by cell
signaling experts.
Stay ahead in this rapidly advancing, multidisciplinary eld
with customalerts and personalization tools.
Sitewide access available for your institution
today. Contact sciencesignaling@aaas.org
or call +1 866-265-4152
ScienceSignaling.org
Nowavailable
in print edition!
Prices valid until December 31, 2011.
NNo
iinn
395 www.sciencemag.org/products SCIENCE VOL 334 21 OCTOBER 2011
Newly offered instrumentation, apparatus, and laboratory materials of interest to researchers in all disciplines in academic, industrial, and governmental organizations are
featured in this space. Emphasis is given to purpose, chief characteristics, and availability of products and materials. Endorsement by Science or AAAS of any products or
materials mentioned is not implied. Additional information may be obtained from the manufacturer or supplier.
Electronically submit your new product description or product literature information! Go to www.sciencemag.org/products/newproducts.dtl for more information.
LIFE SCIENCE TECHNOLOGIES
NEW PRODUCTS: CELL/TISSUE CULTURE
Produced by the Science/AAAS Custom Publishing Office
AUTOMATED CELL CULTURE SYSTEM
The CompacT SC is an automated T-ask cell culture system designed
to fully control expansion, differentiation, and maintenance of stem
cells. The new CompacT SC, which can be used to aseptically culture
multiple different stem cell lines in parallel using T75 and T175 asks,
has a number of features that makes controlling the growth and
differentiation of stem cells an easy, convenient task. The system
includes automated cell imaging and analysis, allowing scientists to
program the CompacT SC to monitor cell morphology and growth
curves. It also permits researchers to set the system to respond to
any pre-programmed changes by automatically passaging cells and
adding new media, reagents, or vectors at appropriate times. This
means scientists can consistently maintain cell lines in the state they
need them, either as pluripotent cells or during cell differentiation
and transfection. Processes can be performed at any time of the day
or night, without researchers having to be in the laboratory.
TAP Biosystems
For info: +44-(0)-1763 227200 www.tapbiosystems.com
MANUAL COLONY COUNTER
The aCOLade is an affordable, manual colony counter with automatic
result recording, ideal for use in any microbiology laboratory. The
new aCOLade colony counter consists of a compact white light
box with a pressure controlled back plate and a pen attached to an
integrated digital counter. Operating aCOLade is quick and simple
because as soon as microbiologists touch the Petri dishs lid with
the pen to count a colony, a bleeping sound occurs and the count is
automatically added and displayed on the digital counter. To record
results, aCOLade can be connected to any computer and comes
with software, which allows scientists to automatically transfer their
counts to a spreadsheet. This exible system has a well-designed
adjustable plate ring that is used to hold plates as small as 50 mm
diameter and can be easily altered to accommodate 90 mm and
120 mm diameter plates, making it suitable for a range of counting
applications.
Synbiosis
For info: +44-(0)-1223-727125 www.synbiosis.com
ENZYME BLOCKING SOLUTION
In immunohistochemical staining, background staining can often
be traced to the existence of active endogenous enzyme, such as
peroxidase and alkaline phosphatase, in the tissue preparation. This
type of interference can generate false positive results or obscure
the specic staining of the target antigen. To assist the pathologist,
histologists, and other research scientists in obtaining meaningful
immunohistochemical data, BLOXALL Blocking Solution has been
designed as a dual blocking solution for endogenous alkaline phos-
phatase and peroxidase in tissues. BLOXALL inactivates all endog-
enous peroxidase, pseudoperoxidase, and alkaline phosphatase in
formalin-xed, parafn-embedded tissue sections, frozen tissue
sections, and cell preparations. BLOXALL Blocking Solution is a
simple, single step, ready-to-use reagent provided in a convenient
dropper bottle.
Vector Laboratories
For info: 650-697-3600 www.vectorlabs.com
CELL HEALTH/GENE EXPRESSION ASSAY
The ONE-Glo + Tox Assay is designed for improved analysis of re-
porter gene expression in the context of cell health. The assay com-
bines luciferase chemistry with a cell viability marker in a two-step,
addition-only process, to enable measurement of cell viability and
gene expression in a single well of a plate, negating the need to
run parallel assays. By analyzing reporter activity in relation to cell
viability, false results due to toxicity or changes in cell number can
be uncovered. The rst part of the assay is a non-lytic uorescence
assay (CellTiter-Fluor Cell Viability Assay) that measures the relative
number of the live cells in a culture population after experimental
manipulation. The second part of the assay uses the ONE-Glo Lu-
cifease Assay System to quantify rey luciferase reporter gene ex-
pression from cells made to express this reporter enzyme. Ideally
suited for high and ultrahigh throughput applications, the volumes of
each assay component can be scaled to meet throughput needs and
is amenable to automation up to 1,536-well format.
Promega
For info: 800-356-9526 www.promega.com
INCUBATORS
The CB series incubators provide excellent growth conditions for mammalian
cell and tissue cultures, offering precise, reproducible control of oxygen
concentrations as low as 0.2% and quick recovery times, important factors
in conducting research under hypoxic conditions. Models CB 53, 150, and
210 allow sample access via optional small internal doors that, in addition to
maintaining steady CO
2
and O
2
levels and accelerating recovery, minimize
the risk of contamination. The incubators feature as standard a 180C hot-air
cycle that provides true sterilization. Interior walls stay dry through controlled
humidication with a dened condensation point, and chamber preheating
with the VENTAIR air jacket provides homogeneous temperature distribution.
A seamless, deep-drawn interior features an integrated shelf-support system
for easy cleaning. Single-beam infrared sensor measures and controls CO
2
in real-time, independent of other gasses and humidity, and provides stable
pH values for optimum, uniform cell growth throughout the entire chamber.
Binder
For info: 866-885-9794 www.binder-world.com
POSITIONS OPEN
TENURE TRACK FACULTY POSITIONS in
Structural Biology and Dynamics with a focus on
Membrane Proteins
The Departments of Biochemistry & Molecular Bi-
ology (website: http://www.bmb.msu.edu) and
Chemistry (website: http://www.chemistry.msu.
edu) at Michigan State University (MSU) seek out-
standing candidates for two tenure-track ASSIST-
ANT PROFESSORSHIPS in the fields of structural
biology and dynamics with an emphasis on mem-
brane proteins and biomembranes. Successful candi-
dates will develop a vigorous, externally funded research
program using and/or developing modern molecu-
lar biophysical approaches to examine the structure
and dynamics of membrane proteins involved in im-
portant biological functions including but not limited
to, host pathogen interactions, signaling, ion or mo-
lecular transport, bioenergetics, and membrane bio-
genesis. These individuals will have the opportunity to
use state-of-the-art facilities, which include a 900 MHz
NMR, mass spectrometry facility, and dedicated syn-
chrotron beamlines at Argonne National Laboratory.
There are also opportunities to interact with a diverse
group of MSU faculty working in structural biology
and dynamics. In addition, the candidates will contrib-
ute to teaching undergraduate, graduate, and/or med-
ical students in their home department.
Review of application materials will begin on
November 15, 2011 and will continue until suitable
candidates are identified. The Departments are highly
committed to a diverse faculty and encourage applica-
tions from individuals in groups that have been
traditionally underrepresented in science faculties in-
cluding women, persons of color, veterans, and persons
with disabilities. Application materials should include
a single file containing a cover letter, curriculum vitae,
statement of research interests and future directions,
and the names and e-mail addresses of three refer-
ences. All application materials should be uploaded
electronically at website: http://www.jobs.msu.
edu for Position #5297. Questions regarding this
position may be directed to Dr. Shelagh Ferguson-
Miller (e-mail: fergus20@msu.edu) in Biochem-
istry & Molecular Biology and/or Dr. David Weliky
(e-mail: weliky@chemistry.msu.edu) in Chemistry.
ASSISTANT OR ASSOCIATE PROFESSOR
UCSD Department of Neurosciences
Website: http://neurosciences.ucsd.edu/
Faculty Position for Neuroscientist who studies
genomic or epigenomic control of basic neural pro-
cesses, and/or the influence of genomic variation
on disease mechanism or therapy. Candidate should
be familiar with neuroscience, and have significant
expertise with bioinformatics tools, including high-
resolution analyses of genomic variation and expres-
sion. Ability to integrate information across platforms
is highly desirable, together with willingness to in-
teract in a highly collaborative environment while pur-
suing independent research interests. Background in
bioinformatics is desirable. Assistant or Associate lev-
el. The La Jolla neuroscience community is extremely
strong and diverse, providing an outstanding oppor-
tunity to develop an independent research program.
Position will be within newly established Translational
Neuroscience Institute in the School of Medicine.
Competitive salary and startup. Commensurate with
qualifications and based on University of California pay
scales. Review of applications will begin November 30,
2011 and continue until the position is filled.
Application materials should be submitted via UCSD
AP On-Line RECRUIT (website: https://apol-recruit.
ucsd.edu/), an electronic job application system. See
Position Posting 10-165 in RECRUIT. Please be
prepared to provide curriculum vitae, a two-page
statement of research interests, and three letters of
recommendation.
University of California, San Diego is an Affirmative Action/
Equal Opportunity Employer with a strong institutional com-
mitment to excellence through diversity.
POSITIONS OPEN
FACULTY POSITION
Yale University School of Medicine
Department of Neurobiology
New Haven, CT 06520-8001
Website: http://medicine.yale.edu/
neurobiology/index.aspx
The Department of Neurobiology at Yale
University School of Medicine is seeking to hire
a scientist using genetic, molecular, or cellular
approaches to examine nervous system develop-
ment and function. Although the emphasis will
be placed on recruiting an ASSISTANT PRO-
FESSOR, extraordinary applicants at the AS-
SOCIATE PROFESSOR or PROFESSOR
levels will also be considered. We seek excep-
tional candidates with a track record of creativ-
ity and productivity, demonstrated potential for
outstanding future achievements, and a desire
to participate in a dynamic and growing Neuro-
science community at Yale. Candidates are ex-
pected to mount a productive and innovative
research program, to obtain outside funding, and
to participate actively in graduate and medical
education.
Candidates must hold a Ph.D., M.D., or equiv-
alent degree. Please send curriculum vitae, se-
lected reprints, a research plan, and the names of
at least three references. All application mate-
rials should be sent electronically to the following
e-mail: neuro.search@yale.edu. Applications
will be reviewed as they are received, but must
be received before December 1, 2011.
Yale University is an Affirmative Action/Equal Op-
portunity Employer. Yale values diversity in its faculty,
students, and staff and especially welcomes applications
from women and underrepresented minorities.
The Department of Biology at the University of
Minnesota Duluth (UMD) invites applications for a
tenure-track ASSISTANT PROFESSOR position in
Genetics, broadly defined, beginning August 2012.
We seek a person with research experience in genetics,
or closely related field, who will instruct lecture and
laboratory courses in genetics, and develop at least one
advanced course in their area of specialization. The
area of specialization is open; applications from can-
didates with expertise ranging from population ge-
netics to genomics will be equally considered. The
successful candidate will establish an independent, ex-
ternally funded research program involving undergrad-
uates, and M.S. and Ph.D. students. Service to the
department, college, and University is also expected.
Opportunities exist for collaboration with researchers
at UMD_s Natural Resources Research Institute, Large
Lakes Observatory, College of Pharmacy, School of
Medicine, and the EPA Mid-Continent Ecology Di-
vision. State-of-the-art research and instruction facil-
ities and competitive startup funding are available.
Essential qualifications include a Ph.D. or terminal de-
gree in the biological sciences, evidence of potential
for achievement in teaching appropriate for appoint-
ment at the Assistant Professor level, peer-reviewed
publications, and strong oral and written communi-
cation skills. The University of Minnesota requires
that you apply online for this position. For a com-
plete position description and information on how
to apply online, visit website: http://employment.
umn.edu/, and search for Job Requisition 174713.
Complete applications will be reviewed beginning
November 22, 2011 and continue until the position
is filled. The University of Minnesota is an Equal Opportunity
Educator and Employer.
o
n
l
i
n
e

@
s
c
i
e
n
c
e
c
a
r
e
e
r
s
.
o
r
g
21 OCTOBER 2011 VOL 334 SCIENCE www.sciencecareers.org 396
www.sciencecareers.org 397
UPCOMING FEATURES
Neuroscience: Emerging FieldsNovember 4
Focus on ChinaDecember 9
BS/MS Scientists (online only)January 13
T
his summer, the European Commission announced
a 7 billion (US$9.6 billion) investment in science,
its largest funding package ever for research. This
boost in science capital, up 9 percent from last
year, is part of the European Unions (EU) Seventh
Framework Programme for research and technology develop-
ment (FP7), and will fund research that will tackle the biggest
societal challenges facing Europe and the world. It is expect-
ed to create approximately 174,000 jobs in the short-term and
450,000 jobs total, and nearly 80 billion (US$113 billion) in
gross domestic product (GDP) growth over 15 years.
This is good news for scientists looking for work in Europe.
An integral part of excellence is integration or mobility of
researchers from many different nations, says Michael
Jennings, spokesperson for Mire Geoghegan-Quinn,
European Commissioner of Research, Innovation and Science.
We are investing in research and innovation in Europe and
creating the conditions and priorities to better attract and
retain scientists.
For academics who desire employment on this diverse and
captivating continent, it is important to remember that while
many aspects of European science are the same no matter
which country you are in (for example, all EU scholars can
apply for funds from the European Research Council [ERC]),
there are also numerous differences that exist between
the members states and their higher education systems.
The postdoc appointment is one such illustration: Although
the concept of postdoctoral researchers is well understood,
there is considerable variability in job titles and in the practical
organization of this stage in Europe, according to the 2010
European Science Foundation (ESF) Report, Research Careers
in Europe: Landscape and Horizons (http://scim.ag/qrFZHZ).
But one thing is consistent across borders: Opportunities for
employment abound, as long as you know how to navigate the
member states systems.
GERMANY: STRONG AND OPEN FOR BUSINESS
Germany, with a population of approximately 82 million, seems
to be faring better than many of its EU neighbors continued

C
R
E
D
I
T
S
:

(
L
E
F
T
)

I
S
T
O
C
K
P
H
O
T
O
.
C
O
M
/
T
R
O
T
S
C
H
E
;

(
S
E
I
P
K
E
)

C
O
R
I
N
N
E

A
P
P
I
A
-
A
Y
M
E
;

(
K
O
T
A
K
O
S
K
I
)

E
E
R
O

H
O
L
M
S
T
R

M
FOCUS ON CAREERS
FOCUS ON EUROPE
Produced by the Science/AAAS Custom Publishing Office
You need to have something to offersuch as
collaborations, expertise on topics which would
complement theirs, or teaching experiencein
order to be selected for a permanent position.
Jani Kotakoski
Academic Opportunities in
European Science
Norwegian University of
Science and Technology
Bjrg Elisabeth Kilavik Ryan F. Seipke
Scientists who wish to pursue academic careers in Europe have much with
which to contend, especially now amidst an uncertain scal landscape. With
much of the continent still at risk for a recession, there is legitimate reason
to be anxious over the future of science funding in the European Union.
However, buttressed by the European Commission, several member states,
including Germany, the United Kingdom, and France, as well as Scandinavia,
have plans in place to bolster scientic research and innovation, and make
those regions attractive destinations for early and mid-career professionals
looking for academic positions. By Alaina G. Levine
Jani Kotakoski
www.sciencecareers.org 398
during this economic tumult. According to the Economist, it is
in a comfortable scal situation, aided by a strongly growing
economy will help to reduce the decit without tough public
spending cuts or tax rises. Cathleen Fisher, president of the
American Friends of the Alexander Von Humboldt Foundation,
which promotes and supports the activities of the parent Hum-
boldt Foundation, which provides exchanges between German
and American scientists, echoes this sentiment. Germany is
in a relatively good position, having weathered the 2008 nan-
cial crisis quite well, she states.
The nation is advancing in the midst of nancial fragility for
a variety of reasons, says Andreas Pinkwart, former minister
for research in North-Rhine-Westphalia, and currently dean of
HHLLeipzig Graduate School of Management. The German
economy has learned to be very export-oriented and global-
oriented. For example, in the last ve years, the government
developed a ranking system for its universities in an effort to
make them more internationally competitive, he adds.
This impetus to improve the higher learning institutes is one
slice of a pie composed of three federal initiatives, which will
support research and development, student funding, and uni-
versity infrastructure. Each enterprise has its own time frame
and budget, although the total amount exceeds 20.7 billion
(US$28.4 billion), and guarantees annual budget increases of
at least 5 percent for the largest science funding and science
performance organizations, explains Max Vgler, director of
the North America Ofce of the German Research Foundation
(DFG, Deutsche Forschungsgemeinschaft).
Jani Kotakoski, currently an adjunct professor of physics at
the University of Helsinki, completed his postdoc at Technische
Universitt Darmstadt. His three-year contract was funded by
DFG and focused on materials science and high-pressure phys-
ics. I was interested in a permanent position in Germany, he
says, but I was extremely unlikely to obtain one with the expe-
rience I had at the time. By that point, Kotakoski explains, he
had completed only one postdoc, which is not usually enough
experience to be considered a top candidateat least in a eld
like materials science where the competition is very tough.
He returned to his native Finland and rejoined the group from
which he had received his Doctorate. Now Kotakoski is on his
way to a new position at the University of Vienna, which is
likely to greatly improve my research prole, he says. My
view is that after two to three years there, I will be able to get
a good position elsewhere, like Austria, Germany, Finland, or
maybe other Nordic countries.
For scientists interested in relocating here, Kotakoski advises
interested parties to obtain contacts in advance and pursue a
contract position. One thing to also keep in mind is that the
people who make the decision are the other faculty mem-
bers, he afrms. Hence, you need to have something to of-
fersuch as collaborations, expertise on topics which would
complement theirs, or teaching experiencein order to be se-
lected for a permanent position.
With a known deciency in tenure-track positions, and a
distinctive tradition of requiring academics to complete a
second thesis (called Habilitation) to even qualify for tenured
employment, Germany has tried to introduce new paths to-
ward landing these treasured jobs, says Fisher. The addition
of junior professorships as an alternative to the fulllment of
a Habilitation is one such route, depending on the institution
and eld.
Furthermore, it is signicant to note that in Germany, the Ph.D.
is considered to be an acceptable entry point for industrial jobs.
Vgler cites that with 2.4 percent of the German workforce
holding Doctorates, compared to an estimated 1.4 percent in
the United States, there is no general expectation to go into
academia, he says. Pinkwart comments that approximately
half the Doctorates go to work in industry versus academia.
UK: GETTING BETTER ALL THE TIME
The United Kingdom is the third largest economy in Europe,
after Germany and France, and is a scientic powerhouse that
cannot be doubted: With just 1 percent of the worlds popula-
tion, the UK receives over 12 percent of citations to published
papers, and receives 10 percent of internationally recognized
prizes each year, according to its embassys website.
But only a decade ago, the government commissioned
a study in which various problems that negatively affect the
supply chain of scientic academic jobs were identied. It
appears that this is not an attractive career path for many of
the brightest Ph.D. graduates. This is both harming the UKs
research base and causing recruitment and retention difcul-
ties for universities, according to the report led by Sir Gareth
Roberts, a Welsh physicist. Among the challenges concerning
postdoctoral and other contract research staff noted were: Un-
certain career prospects associated with work on a short-term
contractual basis, unsatisfactory training in the skills required
in an academic career, and increasingly uncompetitive salaries.
Today, the UK system still has its troubles. But things seem
to be getting better, and the nation continues to attract sci-
entists from abroad to its noted institutions. Case in point:
Raymond E. Goldstein, the Schlumberger Professor of Com-
plex Physical Systems in the Department of Applied Mathe-
matics and Theoretical Physics at the University of Cambridge.
An American who arrived in the United Kingdom
FOCUS ON CAREERS
FOCUS ON EUROPE
Produced by the Science/AAAS Custom Publishing Office
continued

C
R
E
D
I
T
:

H
H
L
-
L
E
I
P
Z
I
G

G
R
A
D
U
A
T
E

S
C
H
O
O
L

O
F

M
A
N
A
G
E
M
E
N
T
Andreas Pinkwart
The Alexander von Humboldt Professorships are Germanys way of creating a beacon effect
and energising its research landscape. Every year, the Alexander von Humboldt Foundation is
offering ten of the worlds leading researchers up to fve million EUR each to create new or
consolidate existing internationally visible research focus areas at German universities.
Shine your light in Germany
Proft from excellent conditions for research
Academics of all disciplines are eligible for an Alexan-
der von Humboldt Professorship, provided that they
are established abroad and recognised internation-
ally as top-class researchers. They will be nominated
by German universities where appropriate in co-
operation with non-university research institutions.
Each Alexander von Humboldt Professorship will be
sponsored for a period of fve years on the premise
that the university presents a convincing strategy
to sustain the position once the funding period has
come to an end. Accordingly, universities are asked to
submit an implementation plan as part of the fund-
ing application. This will allow new, long-term re-
search groups to be established, conducting cutting-
edge (international) research. The German Ministry of
Education and Research is supporting this programme.
Visit www.humboldt-foundation.de/ahp for more
information. Next closing dates for applications:
15 November 2011 and 15 May 2012
Alexander von Humboldt Foundation
Jean-Paul-Str. 12
53173 Bonn
Germany
E-Mail: info@avh.de
www.humboldt-foundation.de
Exzellenz verbindet
be part of a worldwide network.
o
n
l
i
n
e

@
s
c
i
e
n
c
e
c
a
r
e
e
r
s
.
o
r
g
f
o
c
u
s

o
n

e
u
r
o
p
e
www.sciencecareers.org 400
of higher education and re-
search, says a French of-
cial involved in the manage-
ment of bilateral cooperation
programs. Initiatives include
the creation of a new fund-
ing agency in 2005, the
National Research Agency
(ANR, LAgence nationale de
la recherch), a decentraliza-
tion of universities to give
them more independence,
and nancial incentives to
internationalize laboratories,
which is part of the govern-
ment push to attract faculty
from abroad, he describes.
One element of Frances higher education is seemingly
matchless in the EU: prior to the last ve years, academia
offered only tenured positions explains the French ofcial.
Now we have contract positions also. This evolution, com-
bined with the exibility granted to the institutions to recruit
and negotiate salaries for their research faculty, has helped the
nation become more attractive to top scientic talent. But the
completion of an in-country contract position is still the pre-
ferred route to obtaining a professorship.
Bjrg Elisabeth Kilavik can attest to this manner of hiring.
After completing a postdoc in Marseille, she will start a perma-
nent job as a principal investigator/research scientist at the Na-
tional Center for Scientic Research (CNRS, Centre National de
la Recherche Scientique). As an employee of the state, Kilavik
will be subject to a mandatory retirement age, but she will not
have teaching responsibilities, although this could change. At
these types of institutions, we are very privileged, she notes.
Kilavik realizes how attractive France is for academics seek-
ing tenure. In Germany, where she completed her Ph.D., I
could maybe have gotten a ve year contract as a junior pro-
fessor, but there are few possibilities of getting a tenure-track
position, she says. In France, at least the permanent posi-
tions exist,which gives stability. More people in Europe are
realizing[France] is close to the only place where these posi-
tions exist.
SCANDINAVIA: FAIRING WELL AND GROWING
United by similar cultures and language, the Nordic countries of
Norway, Sweden, and Denmark support internationalism and
an ease of movement of scholars from one nation to another.
Almost all universities are state funded, and grants are gener-
ally bestowed by the individual countries research councils,
either directly to a PI or through the PIs institution, depending
upon the country and type of grant. While many other coun-
tries demonstrate scal distress, Scandinavias economies are
resilient. Norway has a budgetary surplus, and Sweden is debt-
free. All three countries have annually increased spending on
research and innovation for the last few years.
We are so fortunate in Scandinavia, we are not
FOCUS ON CAREERS
FOCUS ON EUROPE
continued

Produced by the Science/AAAS Custom Publishing Office

Germany is in
a relatively good
position, having
weathered the
2008 nancial
crisis quite well.
Cathleen Fisher
ve years ago, Goldstein was
a professor at the University
of Arizona when he was ap-
pointed to the chair.
He admires the process
to hire him. Their interest
in me was not the money I
could bring in, but rather the
science, he says. Goldstein
appreciates that faculty are
paid full-year salaries, as op-
posed to nine month salaries
as seen in the States. This
is recognition that research
is a fundamental part of your
job, he adds. There is a man-
datory retirement age (67, which may be changing), but many
retirees keep their labs, supported by their pension.
For younger scientists eyeing careers in British academia,
Goldstein suggests that the best way to get a job may be to
do an early postdoc here. Indeed, this is something that Ryan
Seipke, a senior research associate (essentially a postdoc) in
the School of Biological Sciences at the University of East An-
glia, is currently undertaking. Originally from the United States,
he chose the United Kingdom for his training in part because he
wanted to gain international experience. When his three-year
contract ends in 2012, he can apply for another.
As he starts his job search process, Seipke notes a number
of ways in which one could potentially join the faculty of an
English university. One is to secure a postdoc-to-professor
transition fellowship. These highly competitive, prestigious
opportunities are offered by several organizations (private and
governmental), including the Royal Society, Biotechnology
and Biological Sciences Research Council, Leverhulme Trust,
and Medical Research Council. They are an unofcial stamp
of approval to the university that you might be a star and
successful in securing more grant money, opines Seipke. At
the conclusion of the fellowship, he says, you are often offered
a permanent faculty job.
FRANCE: PUTTING SCIENCE BACK INTO THE HEART
OF THE NATION
The French Republic is the 20th largest country in the world by
demography, but the fth biggest scientic power with over
210,000 public and private researchers and, in total, almost
800,000 engineers and scientists nationwide, according to a
report issued by its Ministry of Higher Education and Research.
With a modern history of domestic scientic achievement
dating back to the end of World War II, France has capitalized
on its scientic assets with the adoption, in 2009, of a National
Research and Innovation Strategy with a very specic and
worthwhile goal: To put back research and innovation at the
heart of French society and economy.
Indeed, as the country transitions from an economy that
incorporates more governmental control to one that is more
autonomous, France has strived to reshape the landscape
London Research Institute
Research Group Leaders
R
e
g
i
s
t
e
r
e
d
c
h
a
r
i
t
y
n
o
1
0
8
9
4
6
4
The London Research Institute (LRI) is Cancer Research UK's flagship research
Institute, focusing on the analysis of fundamental biological processes involved in cancer.
The Institutes international staff work in 50 research groups, housed in well-supported
laboratories at Lincoln's Inn Fields in central London, and at Clare Hall in Hertfordshire.
LRI encourages pursuit of ambitious and longer-term research programmes at the highest
level. The London Research Institute is core-funded by Cancer Research UK. LRI Group
Leaders receive generous Institute core funding for personnel (research fellows, graduate
students and technical support), laboratory consumables, access to the Institute's
comprehensive core technology facilities, backed by a substantial laboratory space
and equipment package and competitive employment terms.
For 2011 recruitment, we are interested in outstanding Scientists seeking to establish
independent innovative research programmes to address fundamental questions in
Cancer Biology including but not limited to:
Tumour Biology: tumour-host interactions, cancer models, human cancer genomics
Chromosome Biology: DNA damage, Cell Cycle regulation
Computational Biology: Bioinformatics, biological networks, image processing.
Informal enquiries may be made by e-mail to Julian.Downward@cancer.org.uk or
John.Diffley@cancer.org.uk
Applications by Clinically qualified candidates wishing to hold Senior Clinical Research
Fellowships at the Institute are welcomed. For information about the London Research
Institute, its staff, and their research interests visit www.london-research-institute.co.uk
In 2015 the LRI will become part of the new Francis Crick Institute, based in a
state-of-the-art laboratory building at St Pancras in central London. Research at the new
Institute will focus on interdisciplinary approaches to the biology of human health and
disease. The Crick will be home to 1,250 researchers from the LRI, the MRC National
Institute for Medical Research, and the London Universities UCL, KCL and Imperial College.
Read about The Francis Crick Institute at www.crick.ac.uk
To apply, please visit https://lrigroupleader.cancerresearchuk.org and submit a CV, publication list,
past and future research plans (approx 2,000 words) and the contact details of at least three
academic referees in a .pdf format. Please note that referees will be instructed to submit letters
of recommendation online at the time your application is received.
Closing date: 27th November 2011.
o
n
l
i
n
e

@
s
c
i
e
n
c
e
c
a
r
e
e
r
s
.
o
r
g
f
o
c
u
s

o
n

e
u
r
o
p
e
www.sciencecareers.org 402
Kristoffer Meinander
DOI: 10.1126/science.opms.r1100110
Alaina G. Levine is a science writer based in Tucson, Arizona, USA.
FOCUS ON CAREERS
FOCUS ON EUROPE
in an economic crisis such as in central and south Europe,
says Mikael Lindgren, a Swede who is currently a professor
of optical sciences at the Norwegian University of Science and
Technology. There appears to be no lack of money in Norway
for research and development in sectors such as energy,
health, offshore and construction, only a lack of skilled labor.
Permanent academic jobs can be found throughout the region,
although they are scarce in Denmark and Sweden because
there are few universities in those countries. In Norway it
can be difcult to recruit local Ph.D. students and postdocs
because of the competition from the industrial and health
sectors, he remarks.
Ylva Hellsten, a professor of exercise and sports science
at the University of Copenhagen, has been in Denmark for 15
years. She acknowledges the difculty for foreigners in secur-
ing permanent academic positions in this former seat of Viking
power, particularly those from outside of Scandinavian coun-
tries, who may be unfamiliar with the language. Furthermore,
there are only eight universities, and Ph.D. students typically
stay within the Danish system for academic employment, she
says. However, the tide may be changing. In 2007, 10 percent
of researchers at Danish universities were of foreign descent,
states Christian Lundager, assistant to the director general
of the Danish Agency for Science, Technology and Innovation
(Forsknings- og Innovationsstyrelsen), but from 20072009,
one-third of all new appointments at the assistant professor
level or above were from abroad.
Peter Byass, a professor of global health at Ume Centre
for Global Health Research within Ume University in Sweden,
recently considered how the country might become more at-
tractive to international researchers. He cites that while the na-
tion of 9 million people boasts a positive working environment
which emphasizes academic autonomy, public appreciation for
science, and universal healthcare, there are still several hur-
dles to recruiting foreign scholars. For example, the academic
review process for senior posts can seem unbelievably slow
and complex to outsiders, and it can be hard for outsiders to
understand expectations put on researchers, especially given
unfamiliar management styles, he says.
But with world-class research institutions and the ability to
submit grant proposals in English, Scandinavia is an attrac-
tive option for international scientists. Kristoffer Meinander,
a Swedish-speaking Finn who is a postdoc at the University of
Aarhus in Denmark, says that foreigners may be surprised to
learn the extent of scientic knowhow and funding for research
infrastructure that exists here. I wouldnt have expected to
nd this high level in Scandinavia.
Indeed, throughout Europe, international scholars are
discovering its scientic assets are continuing to prosper
even amidst an ambiguous economic landscape. And now
with support from the European Unions Seventh Framework
Programme and related national initiatives, the continent
seems poised to advance even further, creating more attractive
opportunities for foreign scholars looking to contribute to its
research endeavors.
Produced by the Science/AAAS Custom Publishing Office
FEATURED PARTICIPANTS
Norwegian University of
Science and Technology
www.ntnu.edu
Seventh Framework
Programme for Research
and Technology
Development
europa.eu/legislation_
summaries/energy/
european_energy_policy/
i23022_en.htm
Ume Centre for Global
Health Research
www.globalhealthresearch.
net
University of Cambridge
www.cam.ac.uk
University of
Copenhagen
www.ku.dk/english
University of East Anglia
www.uea.ac.uk
University of Helsinki
www.helsinki./university
Aarhus University
www.au.dk/en
American Friends of the
Alexander von Humboldt
Foundation
www.americanfriends-of-avh.
org
Danish Agency for Science,
Technology and Innovation
en..dk
Delegation of the European
Union to the USA
www.eurunion.org/eu
Embassy of France in
the U.S.
ambafrance-us.org
HHL-Leipzig Graduate School
of Management
www.hhl.de
National Center for
Scientic Research (CNRS)
www.cnrs.fr/index.php
North America Ofce of
the German Research
Foundation (DFG)
www.dfg.de/en/dfg_prole/
head_ofce/dfg_abroad/north_
america/
For further information and access to the online application material, please consult:
www.ist.ac.at/professor-applications
Deadline for receiving Assistant Professor applications: January 15, 2012
IST Austria values diversity and is committed to equality. Female researchers are encouraged to apply.
ISTAUSTRIAISLOOKINGFOR
Professors and
Assistant Professors
IST Austria (Institute of Science and Technology Austria) invites applications
for Professors and Assistant Professors in all elds of the natural and mathe-
matical sciences and related disciplines. Outstanding scientists in physics,
chemistry, and mathematics are especially encouraged to apply.
The Institute, which is situated on the outskirts of Vienna, was established by the Austrian
government with a focus on basic research. The campus opened in 2009 and is expected
to grow to 45 research groups and over 500 employees by 2016. IST Austria is entitled
to award PhD degrees and includes an English-language graduate school. It aims to
achieve an international mix of scientists and chooses them solely on the basis of their
individual excellence and potential contribution to research.
The Institute recruits tenured and tenure-track leaders of independent research groups.
The successful candidates will receive a substantial annual research budget but are
expected to also apply for external research grants.
For further information and access to the online application please consult www.ist.ac.at/gradschool.
For inquiries, please contact gradschool@ist.ac.at. For students wishing to enter the program in the
fall of 2012, the deadline for applications is January 15, 2012.
IST Austria values diversity and is committed to equality. Female students are encouraged to apply.
CALL FORPhDSTUDENTS
The Graduate School at IST Austria invites applicants from all countries to its PhD program. IST Austria is
a new institution located on the outskirts of Vienna dedicated to cutting-edge basic research in the natural sciences
and related disciplines. The language at the Institute and the Graduate School is English.
The PhD program combines advanced coursework and research, with a focus on Biology, Computer Science,
Neuroscience, and interdisciplinary areas. IST Austria offers internationally competitive PhD salaries supporting 4-5 years
of study. Applicants must hold either a BS or MS degree or equivalent.
The Institute offers PhD students positions with the following faculty:
Additional faculty members will be announced on the IST website www.ist.ac.at.
Nick Barton Evolutionary and Mathematical Biology
Jonathan P. Bollback Evolutionary Biology
Sylvia Cremer Evolutionary and Behavioral Biology
Caroline Uhler Statistics
Tobias Bollenbach Biophysics and Systems Biology
Ca lin C. Guet Systems and Synthetic Biology
Carl-Philipp Heisenberg Cell and Developmental Biology
Harald Janovjak Molecular and Cellular Biophysics
Daria Siekhaus Cell and Developmental Biology
Michael Sixt Cell Biology and Immunology
Jozsef Csicsvari Systems Neuroscience
Peter Jonas Neuroscience
Gaper Tkac ik Theoretical Biophysics and Neuroscience
Krishnendu Chatterjee Game Theory and Software Systems Theory
Herbert Edelsbrunner Algorithms, Geometry, and Topology
Thomas A. Henzinger Software Systems Theory
Vladimir Kolmogorov Computer Vision and Graph Algorithms
Christoph Lampert Computer Vision and Machine Learning
Krzysztof Pietrzak Cryptography
Chris Wojtan Computer Graphics
CAMPUS
VISIT DAY
November 26,
2011
o
n
l
i
n
e

@
s
c
i
e
n
c
e
c
a
r
e
e
r
s
.
o
r
g
f
o
c
u
s

o
n

e
u
r
o
p
e
Vertex creates new possibilities in medicine. Our team discovers,
develops and commercializes innovative therapies so people with
serious diseases can lead better lives. Vertex scientists and our
collaborators are working on new medicines to cure or signicantly
advance the treatment of hepatitis C, cystic brosis, epilepsy and
other life-threatening diseases. Founded more than 20 years ago in
Cambridge, MA, we now have ongoing worldwide research programs
and sites in the United States, United Kingdom and Canada. Vertex has
consistently been recognized as one of the industrys top workplaces by
leading publications such as the Boston Globe, Boston Business Journal,
San Diego Business Journal and The Scientist, and most recently was
named the top employer in Science magazines 2011 annual survey.
As the UK-based subsidiary, Vertex Pharmaceuticals (Europe) Ltd brings
together scientists from various disciplines and backgrounds to form a
highly creative and productive team that works on nding medicines
for the treatment of cancer and inammatory diseases. Our new
research facility in Oxfordshire is equipped with the latest technology
that enables us to tackle all aspects of early phase small molecule drug
discovery and, with the help of the wider organization, take on some of
the most ambitious goals in the industry.
Structure based drug design, has enabled all our major achievements
so far. It is our intention that this technology will continue to play a
central role in the way we do drug discovery. The close proximity of the
Diamond Light Source makes the UK site particularly important in this
eld of Vertex research. The now vacant job of Head of Crystallography
is therefore a wonderful career opportunity for someone with skills in
structural biology and a desire to make important new drugs.
You are most likely to have a PhD and extensive experience in protein
crystallography. Strong leadership skills are required and excellent
written and oral publication record is a must. Experience with
pharma-biased methodologies such as fragment-based lead generation
and the ability to rapidly assess and introduce novel technologies
would be an advantage. The Head of Crystallography will be expected
to inuence Vertex research on both sides of the Atlantic within the
structural biology groups and also by providing ideas and impetus for
new projects. In order to do this, communication skills and the ability
to work in collaboration must be rst rate. An interest in biological and
chemical science beyond structural biology would be most useful since
this would allow the candidate to take career opportunities in multi
disciplinary project leadership.
If you are innovative and focussed on achieving success within a
collaborative environment, please visit our website www.vrtx.com to
nd out more information and apply.
Closing date for applications is 7th November 2011.
Head of Crystallography, Oxford
Graduate Program in Computational
Engineering Science
The Aachen Institute for Advanced Study in Computational Engineering
Science (AICES) is a graduate school established within the frame of the
German Excellence Initiative. AICES focuses on computational engineer-
ing science including such diverse felds as modeling and simulation;
optimization; inverse problems and high-performance computing.
There is a limited number of openings in the AICES graduate program for
exceptionally qualifed students who hold a bachelors or masters degree
in engineering, natural science, mathematics, or computer science.
AICES offers a fve-year path to the doctorate for students with bachelors
degrees, or a three-year path with masters degree, due to a novel advis-
ing and training concept. Admitted candidates receive tax-free stipends
of 2000 per month for the doctoral phase and 500 per month for the
masters phase.
Graduate School AICES
RWTH Aachen
Germany
www.aices.rwth-aachen.de
www.aices.rwth-aachen.de/admission
Photo:
Peter Winandy
o
n
l
i
n
e

@
s
c
i
e
n
c
e
c
a
r
e
e
r
s
.
o
r
g
f
o
c
u
s

o
n

e
u
r
o
p
e
GRADUATE PROGRAM
These newly established Fellowships in Experimental or Theoretcal Quantum
Science will be awarded both on the PhD- and the postdoc-level on the basis of
an internatonal competton. The appointments are for a three-year duraton.
Postdoctoral Fellowships carry a compettve annual salary, and ofer an annual
research expense fund. PhDFellowships will partcipate in the Vienna graduate
program CoQuS.
TheVienna QuantumFellowshipprogramhas beenestablishedwiththesupport
from the Austrian Ministry of Science and Research to ofer young scientsts
the best possible opportunity to develop their talents in the environment of
the Vienna Center for Quantum Science and Technology (VCQ).
The VCQ faculty provides a broad variety of research opportunites in the
areas of Experimental and Theoretcal Quantum Science (see htp://vcq.
quantum.at):
Mater-wave interferometry and quantum atom optcs
Micro- and nanoscale quantum optcs and quantum optomechanics
Microoptcs and novel quantum states of light
Cold atoms and degenerate quantum gases
Many-body quantum physics and quantum simulatons
Entanglement-based quantum communicaton on Earth and via satellites
Quantum informaton and foundatons of physics
Applicatonmaterial shouldbe sent tovcq@quantum.at Detailedandadditonal
informaton regarding applicaton can be obtained from htp://vcq.quantum.
at/fellowships
Deadline for the applicaton is December 1, 2011. Fellowship candidates will
automatcally be considered for other available postdoctoral positons in their
felds of interest.
The Vienna Center for Quantum Science and Technology (VCQ) invites
applicatons for the
Vienna Quantum Fellowships
More information about ATTRACT and PEARL as well as the other funding opportunities
offered by the National Research Fund Luxembourg can be found on the FNRs website.
Go and see whats behind on www.fnr.lu/pearl and www.fnr.lu/attract
For an overview on research in Luxembourg, have a look at www.publicresearch.lu
ATTRACT
LUXEMBOURGS RESEARCH PROGRAMME FOR OUTSTANDING
YOUNG RESEARCHERS FROM ALL OVER THE WORLD
If you are an outstanding young researcher, our
research programme ATTRACT allows you to set up an
independent research team within a research institution
in Luxembourg. Through your innovation, dynamism and
creativity as well as the high scientic quality of your
research project, you will help to enhance Luxembourgs
position in the international world of research. Projects
selected under ATTRACT have a lifespan of ve years
and the nancial contribution will be up to 1.5M EUR.
The 6
th
ATTRACT Call will be launched in December 2011.
PEARL
LUXEMBOURGS RESEARCH PROGRAMME
FOR INTERNATIONALLY RECOGNISED SENIOR RESEARCHERS
If you are an internationally recognised senior
researcher, our research programme PEARL gives you
the opportunity to establish a high-prole research
programme in a research institution in Luxembourg
(University of Luxembourg and other Research Centres)
and thus to accelerate the development of and
to strengthen the countrys research priorities.
The nancial contribution will be up to 5M EUR over
a duration of ve years.
Research Opportunities in Luxembourg.
See whats behind.
INVESTIGATING FUTURE CHALLENGES
challenges love solutions
We think theres a solution for every problem. Thats why BASF researchers from all disciplines
always work on innovations with passion. Help us, in a modern environment, to nd not just
products but comprehensive solutions for tomorrows challenges. Thats how we create
chemistry. At BASF. Find out more now and send your application to: www.basf.com/career
o
n
l
i
n
e

@
s
c
i
e
n
c
e
c
a
r
e
e
r
s
.
o
r
g
f
o
c
u
s

o
n

e
u
r
o
p
e
The Leipzig School of Human Origins
- An International Max Planck Research School -
by
The University of Leipzig
and
The Max Planck Institute for Evolutionary Anthropology
The Leipzig School of Human Origins offers a unique interdisciplinary
graduate program to study the evolutionary history of humans and great
apes. Graduate students are accepted into one of the following areas,
but are encouraged to take part in courses and seminars from all three
disciplines:
Comparative and Molecular Primatology
Evolutionary and Functional Genomics, Ancient DNA,
Molecular Anthropology and Genome Bioinformatics
Human Paleontology, Prehistoric Archaeology
and Archaeological Science
The language of the school is English. Visit www.leipzig.de for informati-
on on living in Leipzig, Germany, in the center of Europe.
For project and application details go to www.leipzig-school.eva.mpg.de
or contact us at:
e-mail: leipzig-school@eva.mpg.de
phone: ++49 (0) 341 3550-0
fax ++49 (0) 341 3550-119
Application deadline: January 31, 2012
The Norwegian University of Science and Technology (NTNU) in Trondheim
represents academic eminence in technology and the natural sciences as well as
in other academic disciplines ranging from the social sciences, the arts, medicine,
architecture to ne art. Cross-disciplinary cooperation results in innovative break-
throughs and creative solutions with far-reaching social and economic impact.
J
o
b
b
n
n
o
r
g
e
.
n
o
Faculty of Medicine
Professorship/
Qualication Fellowship
in Systems Neuroscience
The Kavli Institute for Systems Neuroscience of the Faculty of
Medicine at the Norwegian University of Science and Technology
(NTNU) invites applications for a faculty position in systems
neuroscience. The new position is part of NTNUs strategic effort in
the eld of neuroscience.
We seek applicants with experience and interest in using state-of-
the-art molecular and cellular technologies to understand neural
networks and behaviour. The successful candidate holds a Ph.D. and
has experience leading a research team and attracting international
funding. The candidate has a track record in both molecular-cellular
and systems neuroscience, with outstanding publications, and
she/he demonstrates ability to develop an internationally competitive
research programme. Participation in teaching activities at masters
and PhD level is required.
The applicant will benet from the strong infrastructure at the Kavli
Institute for Systems Neuroscience at NTNU (www.ntnu.no/cbm).
Start-up funding, including scientic equipment and PhD/post
doctoral fellows, is negotiable. The position is advertised at the
rank of Professor but may alternatively be dened as a qualication
fellowship for a period of no longer than 3 years in case of sufcient
future potential.Young applicants not yet qualied for full
professorship are thus encouraged to apply.
Applicants should submit a cover letter, a CV with a complete
publication list, copies of 5 selected papers, a summary of research
achievements, a research plan, and 3 letters of reference (referees
should send letters directly to Edvard Moser, Director of the Kavli
Institute, edvard.moser@ntnu.no).
The complete advertisement is available at
www.jobbnorge.no.
Submit applications through www.jobbnorge.no
within 1 December, 2011.
For further information about the position contact Edvard Moser,
edvard.moser@ntnu.no, tel. +47 73598278; information about the
application process contact Brit Lbeck Fladvad, HR- section,
Faculty of Medicine, brit.advad@ntnu.no.
See also http://www.medisin.ntnu.no/eng/
The Molecular Biology Program of the Sloan-Kettering Institute, Memorial
Sloan-Kettering Cancer Center (www.ski.edu), has initiated a faculty search
at the Assistant Member level (equivalent to Assistant Professor). We are
interested in outstanding individuals who have demonstrated records of
significant accomplishment and the potential to make substantial contributions
to the biological sciences as independent investigators. Successful applicants
will have research interests that move the Program into exciting new areas that
complement and expand our existing strengths in the areas of maintenance
of genomic integrity, regulation of the cell cycle, and regulation of gene
expression. Faculty will be eligible to hold appointments in the Gerstner
Sloan-Kettering Graduate School of Biomedical Sciences, the Weill Cornell
Graduate School of Medical Sciences, as well as the Tri-Institutional MD/PhD
Training Program.
The deadline for applications is November 1, 2011. Interested candidates
should visit http://facultysearch.ski.edu to apply via the on-line faculty
application. Please visit the site as soon as possible, as it contains important
information on the required application materials, including deadlines for
submission of letters of reference.
Informal inquiries may be sent to Julie Kwan at kwanj@mskcc.org
or to Dr. Kenneth Marians, Chair, Molecular Biology Program at
kmarians@sloankettering.edu. MSKCC is an equal opportunity and affirmative
action employer committed to diversity and inclusion in all aspects of recruiting
and employment. All qualified individuals are encouraged to apply.
www.mskcc.org
Faculty Position
Molecular Biology
Sloan-Kettering Institute
o
n
l
i
n
e

@
s
c
i
e
n
c
e
c
a
r
e
e
r
s
.
o
r
g
FOCUS ON EUROPE GRADUATE PROGRAM
POSITIONS OPEN
Senior Level Research Position
WEILL CORNELL MEDICAL COLLEGE and
NEWYORK-PRESBYTERIAN HOSPITAL
Thrombosis research with particular reference to diagnosis, treatment
and prevention of occlusive vascular diseases.
Since 1958, the laboratories have been located at the New York VA
Medical Center and Weill Cornell Medical College.
The major focus is on the etiology and pathogenesis of increased,
therapeutically resistant blood cell and vessel wall reactivity.
The main emphasis of the research is on the vascular and molecular
biology of all blood cell and vessel wall components. Metabolic effects
of cell-cell interactions as initiated with multiple agonists are under
study for identification of transcellular metabolism, both in vitro and ex
vivo. The guiding premise is that thrombosis is strictly a multicellular
process and must be treated by newer methodologies.
Requirements: The applicant should have a Ph.D. or an M.D.-Ph.D.
with a minimum of five years of post-doctoral experience. It would be
important to have a publication record-especially in high impact
journals. The incumbent should also have a proven ability to obtain
independent funding previously. The ability to develop a research
project and work independently on the main theme of the laboratory is
desirable. The ability to interact with colleagues and synergize with them
to advance the research of the laboratory is highly important. The ability
to write and co-author research and clinical publications and also to
develop specific aims for a research grant would be expected.
Qualified candidates should submit their complete curriculum vitae and
bibliography in addition to three reference letters to the Department of
Human Resources, WEILL CORNELL MEDICAL COLLEGE,
1300 York Avenue, Room C610 (Box 113), New York, NY 10065;
ajmarcus@med.cornell.edu.
EOE/M/F/D/V
Climate Change & Sea Level Rise Initiative
Old Dominion University has created an exciting new Climate Change and Sea
Level Rise Initiative and seeks a senior level faculty member in any discipline
related to climate change and sea level rise to help lead this initiative. The goal of
the Initiative is to foster research, education and outreach on the impact that climate
change and sea level rise may have on metropolitan communities that are situated
at or close to sea level. Old Dominion University, a state-assisted Carnegie
doctoral/research-extensive institution that serves almost 25,000 students including
more than 6,000 graduate students, is itself located in the city of Norfolk in the
metropolitan Hampton Roads region of coastal Virginia. Abroad range of faculty
fromacross the entire university, ranging fromscience and engineering to the social
sciences, education, business and health sciences, are currently involved in this
initiative (see http://www.odu.edu/ao/research/ccslri/). The successful candidate
will be part of the leadership team, will be an effective leader and advocate for the
Initiative, and will be expected to contribute to the Initiative through research,
teaching and service activities.
Applicants should possess an appropriate terminal degree and an academic record
that merits a tenured appointment at the rank of associate or full professor in one of
the academic departments within the University. Asuccessful record in research and
grant writing is required, as is evidence of leadership and the ability to interact and
communicate clearly with internal and external communities.
Applications should include a letter of interest that addresses the Initiatives goals, a
curriculum vitae, and contact information for three professional references including
email addresses and phone numbers. Review of applications will begin November
11, 2011 and the position will remain open until an appointment is made.
Applications and nominations should be submitted electronically to: CCSLRSearch
Committee, attention Judy Bowman (jbowman@odu.edu), 222 Koch Hall,
Norfolk, VA23529.
Old Dominion University is an affirmative action, equal opportunity institution and
requires compliance with the Immigration Reform and Control Act of 1986.
o
n
l
i
n
e

@
s
c
i
e
n
c
e
c
a
r
e
e
r
s
.
o
r
g
FOCUS ON EUROPE
Tenure/Tenure Track Faculty Position
in Biological Sciences
The Department of Biological Sciences at Wayne State University
(http://www.clas.wayne.edu/biology/) anticipates hiring a tenure-track
professor with research expertise in systems or computational biology.
Preference will be given to candidates working in areas complementing
the departments existing strengths in transcription and gene regula-
tion, organismal and evolutionary development, intra- and intercellular
signaling, genomics, and community and landscape ecology. Rank will
be dependent upon qualications.
Wayne State University is a large, comprehensive, nationally ranked
research institution that offers state-of-the-art research facilities and
highly competitive start-up packages. The metropolitan Detroit area
offers a rich cultural and educational environment, an excellent standard
of living, and easy proximity to Michigans lakes, forests and recreational
sites. Applicants must have a Ph.D. degree, postdoctoral experience and
an outstanding record of research achievement. Successful applicants
are expected to establish and maintain vigorous, externally funded
research programs and to participate in graduate and undergraduate
education. All positions are posted on-line at jobs.wayne.edu. In addition
to an online application that includes cover letter and curriculum vitae,
applicants must submit a 2-page statement of their research plans and
have three letters of reference sent to the Faculty Search Committee:
ad5348@wayne.edu. Please apply by November 30, 2011 for full
consideration. Applications will be considered only when all materials
have been received.
Wayne State University is an afrmative action/equal opportunity
employer. Women and members of minority groups are especially
encouraged to apply.
Tenure-Track Position in Cell Biology
McGill University
The Department of Biology at McGill University invites applications
for a tenure-track position in cell biology. We are seeking an energetic,
interactive individual who will complement the Departments recognized
strengths, across a wide range of model organisms, in developmental
genetics, neuroscience, biophysics, and cell biology. The ideal applicant
will employ advanced modern techniques to address questions of broad
biological signicance. The candidate will have convenient access to state-
of-the-art imaging facilities and other major core services as a member of
the McGill Life Sciences Complex. Applicants should possess a Ph.D. or
equivalent degree in Biology or a related discipline, postdoctoral experi-
ence, and a signicant track record of research excellence. The successful
applicant will be expected to conduct a vigorous programof independent,
externally funded research and to contribute to teaching at both the under-
graduate and graduate levels. We anticipate that this position will be lled
at the ASSISTANT PROFESSOR (tenure track) level, but applications
from more established candidates will be considered for recruitment at
the ASSOCIATE or FULL PROFESSOR rank. Competitive startup
and equipment funding packages are available. Persons wishing to be
considered for this position should forward via e-mail: a curriculumvitae,
a statement of research interests, a statement of teaching interests, PDF
les of major publications, and arrange to have three letters of reference
submitted directly by e-mail to: recruit.biology@mcgill.ca In the subject
line, please enter Biology Faculty Search and your name. Acceptable le
formats are Microsoft Word and PDF.
The application deadline is December 2
nd
, 2011.
All qualied applicants are encouraged to apply; however, Canadian and
permanent residents will be given priority. McGill University is committed
to diversity and equity in employment. It welcomes applications from
indigenous peoples, visible minorities, ethnic minorities, persons with
disabilities, women, persons of minority sexual orientations and gender
identities and others who may contribute to further diversication.
Assistant Professor of Cell Biology and Neuroscience
The Department of Cell Biology and Neuroscience at Rutgers, The
State University of New Jersey, Piscataway, seeks to ll two tenure-
track positions at the Assistant Professor level. Individuals working
on all aspects of cell biology or neurobiology are invited to apply;
those with research programs that integrate with and complement cur-
rent faculty research will be given highest priority. The new faculty
may be part of the Brain Health Institute initiative. The Department
is located on the Rutgers Busch Campus and is part of the Division
of Life Sciences, a group of Departments and Institutes that serves to
provide opportunities for interdisciplinary research. Current collabora-
tions within the DLS range from biomaterials and nanotechnology
to human genetics and stem cells. The Campus is located near the
UMDNJ-Robert Wood Johnson Medical School and is less than one
hour away from New York City and Philadelphia.
Applicants must have a Ph.D. and/or M.D. with a minimum of
three years postdoctoral experience. The successful candidate will
be expected to establish an independent research program sup-
ported by external funding and to contribute to undergraduate and
graduate education. The Department offers excellent facilities and
competitive start-up packages. Interested individuals are encouraged
to apply online through the CBN website (http://cbn.rutgers.edu)
with a curriculum vitae, a brief statement of research plans, and the
names, addresses, and contact information of three individuals who
will provide a letter of reference.
Applications received after December 1, 2011 will be considered
if positions remain available.
Rutgers University is an Equal Opportunity/Afrmative Action
Employer.
The School of Life Sciences and The Biodesign Institute at Arizona State
University invite applications for a tenure-track faculty position at the level
of Assistant Professor whose research is at the interface between Genomics
and Life Sciences. Research methods may include theoretical, computational,
and empirical approaches in population genetics, functional and comparative
genomics, and bioinformatics. The successful candidate will be expected to
develop an innovative, extramurally-funded, research program, teach at the
undergraduate and graduate levels, and have a commitment to outreach and
service. The successful candidate will be expected to mentor undergraduate
and graduate students as well as postdoctoral fellows. A competitive start-up
package and teaching load compatible with high research productivity will be
provided.
Arizona State University has made a commitment to accelerating the
translation of basic discoveries into practical benets for society through
the construction of state-of-the-art research facilities and the recruitment
of world-class faculty members. The successful candidate will participate
in university-wide health and/or sustainability initiatives supported by core
facilities for functional genomics and next generation sequencing, functional
proteomics, high throughput cellular screen, bioinformatics, high performance
computing, and imaging. More information on genomic research opportunities
at the Biodesign Institute and the School of Life Sciences at ASU can be found
at http://genomics.asu.edu.
Candidates must have a Ph.D. (or equivalent) in an appropriate eld, and
a minimum of 2 years of postdoctoral training is preferred. Demonstrated
teaching and research excellence is preferred.
To apply, send cover letter, your curriculum vitae, three representative
publications, separate statements of future research plans and teaching
philosophy and interests, and contact information for three references to be sent
to Alan Rawls, Chair, Genomics Faculty Search Committee, School of Life
Sciences, PO Box 874501, Tempe, AZ 85287-4501. Electronic applications
sent as PDF les to solsfacultysearch@asu.edu are preferred. The initial
closing date for receipt of applications is November 11, 2011; applications
will be reviewed weekly thereafter until the search is closed. For additional
information on this position and the School of Life Sciences, please visit http:
//sols.asu.edu/jobs. A background check is required for employment. Arizona
State University is an equal opportunity/afrmative action employer committed
to excellence through diversity. Women and minorities are encouraged to apply.
Faculty Position
in Genomics
o
n
l
i
n
e

@
s
c
i
e
n
c
e
c
a
r
e
e
r
s
.
o
r
g
The University of Michigan seeks outstanding applicants for tenure-track and tenured clinical and
biomedical informatics faculty positions. UM hosts a Clinical and Translational Sciences Award (CTSA)
Biomedical Informatics Program, 48 current Bioinformatics PhD students and NIGMS bioinformatics and
NCI proteome informatics training grants. We have active national outreach for minority candidates.
We are currently recruiting up to 3 senior and 5 junior faculty members to establish independent
individual and team-based research programs. Informatics research teams are welcome to apply as a
unit. After multiple recent recruitments in the area of basic bioinformatics, our emphasis this cycle is in
biomedical/clinical informatics. Specifc joint appointments may be considered, for appropriate candidates,
with the Dept. of Human Genetics, Electrical Engineering and Computer Science, School of Information,
School of Public Health, School of Nursing or other appropriate units. In addition, affliation with Centers,
including the Comprehensive Cancer Center, Depression Center, Cardiovascular Center and Metabolomics
and Obesity Center will be encouraged. There are extensive computational and information infrastructural
resources are available to individual recruits and teams. For appropriate individuals, opportunities exist for
faculty leadership roles in Research Information Technology management and operations and to infuence
institutional priorities in clinical and biomedical informatics.
Successful candidates will have a PhD and/or MD degree, or equivalent, with post-doctoral training,
on topics such as biomedical data mining and machine learning; multi-scale integrative analysis; natural
language processing (NLP) and ontologies applied to biomedicine; informatics related to healthcare delivery
and personalized medicine (e.g. clinical decision support or pharmacogenomics). Publications and funding,
as evidence of research productivity, a detailed research plan as well as evidence for an interest in in graduate
and post-doctoral education will be essential components of the application. The rank of selected candidates
will depend upon experience and qualifcations.
Applicants should send a letter of interest with Curriculum Vitae, Research Plan, and a list of three or
more references with current contact information to: Search Committee, Department of Computational
Medicine and Bioinformatics, Job Code 200. The University of Michigan, 2017 Palmer Commons,
100 Washtenaw Ave, Ann Arbor, MI. 48109-2218, email: ccmbrecruit@umich.edu. For appropriately
qualifed candidates, simultaneous application to the UM Biological Sciences Scholars Program, BSSP, is
strongly encouraged (for more information see: http://www.med.umich.edu/medschool/research/bssp/).
Applications will be reviewed through March 2012 beginning October 2011.
Ann Arbor is a remarkable cultural and living environment. The University of Michigan is responsive to
the needs of dual career families and is an Equal Opportunity Affrmative Action Employer committed to
diverse faculty, staff and student body.
http://ccmb.med.umich.edu
University of Michigan Medical School
Department of Computational Medicine
and Bioinformatics
Assistant, Associate and Full Professors
FACULTY POSITION
Department of Chemistry and Biochemistry
Texas State University -San Marcos
The Department of Chemistry and Biochemistry at Texas State University-San Marcos
seeks to ll a new faculty position in any area of Chemistry or Biochemistry, beginning fall
2012. Applicants at all ranks will be considered. The successful candidate at theAssistant Profes-
sor rank must have a Ph.D. in Chemistry, Biochemistry, Chemical Education or a closely-related
eld; postdoctoral research experience; a track record substantiating the potential to establish an
externally-funded research programthat involves undergraduate and/or graduate students; and the
capability to teach courses in chemistry, biochemistry, and/or chemical education. The successful
candidate at the Associate Professor or Professor rank must meet all qualications for the Assis-
tant Professor rank, in addition to the following: ve or more years of experience as a university
faculty member; and a record consistent with (1) success in obtaining externally-funded research
grants, (2) quality peer-reviewed publications, and (3) signicant professional service activities.
At all ranks, preference will be given to candidates whose areas of teaching and research expertise
complement the departments goals; and to candidates who can effectively mentor students.
Texas State University- San Marcos is located in the burgeoning Austin-San Antonio corridor
at the edge of the hill country, and is the 5
th
largest campus in Texas with more than 34,100
students. The Department of Chemistry and Biochemistry currently has 20 faculty, 30 M.S.
graduate students, and 350 undergraduate majors, most of whom participate in research. The
Department also is a participant in a proposed interdisciplinary Ph.D. program in Materials
Science, Engineering, and Commercialization. For additional information, please visit http:
//www.txstate.edu/chemistry.
All application materials should be emailed to Chem_Search@TxState.edu. A complete
application consists of the following: (1) a cover letter identifying the rank for which you are
applying, the area(s) of your teaching and research interests, and a list of the three individuals
(with contact information) who will be submitting letters of reference; (2) a CV; (3) a one-page
summary of teaching philosophy and interests; (4) an outline of research plans (3 pages or less);
(5) undergraduate and graduate transcripts; and (6) three letters of reference, emailed directly
from each referee. To receive full consideration, application submissions should be complete
by December 4, 2011. Reviewof complete applications will begin December 5 and will continue
until the position has been lled.
Texas State University - San Marcos is an equal opportunity employer; women and members of
underrepresented minorities and individuals with disabilities are encouraged to apply.
Executive Director
Division of Earth and Ecosystem Sciences
Position #30-001
o
n
l
i
n
e

@
s
c
i
e
n
c
e
c
a
r
e
e
r
s
.
o
r
g
ScienceCareers.org
An exceptional career requires insightful planning and management.
Thats where Science Careers comes in. From job search to career
enhancement, Science Careers has the tools and resources to help
you achieve your goals. Get yourself on the right track today and get a
real career plan that works. Visit ScienceCareers.org.
Get a Career Plan that Works.
The Department of Civil and Environmental Engineering at Northwest-
ern University invites applications for a tenure-track or tenured faculty
position in Environmental Engineering and Science. We seek outstand-
ing applicants in any area of environmental research who have a strong
fundamental knowledge base, perform cutting-edge cross-disciplinary
research, and have the ability to contribute substantially to our core teach-
ing efforts environmental science and engineering. Areas of particular
interest include: environmental microbiology, biotechnology, and micro-
bial ecology, innovative approaches to sustainable water use, design and
development of advanced treatment technologies, and linkages between
water, ecosystems, biogeochemical cycles, and global change.
Reviewof applications will begin in October 2011, and the search will
proceed until the position is lled. Preference will be given to applications
submitted by November 30, 2011.
Applications should be submitted electronically as a PDF document
containing a cover letter, curriculum vitae, a two-to-three-page descrip-
tion of research accomplishments and plans for future work, a one-to-two
page description of teaching interests, and a list of at least three persons
who can provide letters of recommendation. Application materials should
be submitted to the Environmental Search Committee Chair via the web
interface at http://facultysearch.mccormick.northwestern.edu/apply/
index/NDI
It is anticipated that this position will be lled at the junior level but
outstanding senior candidates will also be considered.
Northwestern University is an Afrmative Action, Equal Opportunity
Employer. Women and individuals in underrepresented groups in science
and engineering are encouraged to apply. Hiring is contingent upon
eligibility to work in the United States.
Faculty Tenure Track Positions
Faculty Position
in Environmental Engineering
and Science
at Northwestern University
Weinviteapplicants withaPhD, MDor equivalent, andarecordof outstanding
promise and achievements, for our faculty positons in cardiovascular and
metabolic disorders. The Duke-NUS Cardiovascular and Metabolic Disorders
(CVMD) Signature ResearchProgramwill bring together top-fight researchers
investgatng theinteractons betweenmetabolic disorders andcardiovascular
disease, focusing on translatonal discoveries that can impact clinical care.
Investgators will address diverse aspects of the problem from diabetes and
dyslipidemia to hypertension and vascular disease.
Positons include full salary as well as generous start-up and 5 years of annual
research funding to assure a stable base of support to be supplemented by
compettve awards. Candidates should submit a cover leter, curriculum
vitae, summary of research accomplishment and an outline of future plans
by February 1, 2012 to:
Thomas Cofman, MD, Director
Cardiovascular and Metabolic Disorders Program
Duke-NUS Graduate Medical School Singapore
8 College Road, Singapore, 169857
E-mail cvmd.recruit@duke-nus.edu.sg
ABOUT DUKE-NATIONAL UNIVERSITY OF SINGAPORE GRADUATE MEDICAL
SCHOOL
DukeNUS brings post-baccalaureate, research-intensive medical educaton
and research to Asia, and represents a truly global partnership between Duke
University in the U.S. and Natonal University of Singapore (NUS). Duke-NUS
shares a moderncampus withSingapores largest hospital, Singapore General
Hospital, and several natonal centers, including the Natonal Heart Centre.
The facilites innearby Biopolis alsoprovide a unique array of complementary
resources. In additon, the CVMD program has close associatons with the
Duke Cardiovascular Research Center and Duke Global Health Insttute.
www.duke-nus.edu.sg
Faculty Positons in Cardiovascular & Metabolic Disorders
Chair,
Biomolecular Chemistry
University of Wisconsin School of Medicine
and Public Health, Madison, Wisconsin
The UWSMPH invites applications and nomina-
tions for the position of Chair of the Department
of Biomolecular Chemistry (BMC), to succeed
Dr. Robert Fillingame, who will be stepping down
after ten years as Chair. The BMC Department has
thirteen faculty members, each with an active
research program in a variety of areas ranging from
biophysical chemistry to cellular and developmental
biology. The Departments graduate program, which
is collaboratively directed by faculty in BMC and the
Biochemistry Department in the College of Agricul-
ture and Life Sciences attracts, on average, 25 very
high quality students annually. The department also
attracts graduate students from other multidisci-
plinary programs on campus and currently houses
40 - 45 students in the various research laboratories in
addition to 30 post-doctoral trainees and other
research staff. The department currently participates
in several multidisciplinary training grants and
has an impressive portfolio of extramural grants
to the faculty.
We seek an accomplished scientist with a strong
record of academic and administrative leadership.
Special emphasis will be given to experience in both
graduate student and medical student education, the
mentoring of junior faculty, and a sustained record
of extramural research funding. The successful
candidate will have an exciting vision for the future
of biochemical research, education, and training in a
leading academic medical center. Candidates must
have a PhD degree or equivalent and must have
academic credentials for a tenured faculty appoint-
ment at the University of Wisconsin-Madison.
Send a letter of application or nomination, with
curriculum vitae, to: K. Craig Kent, M.D.,
and Rod Welch, Ph.D., BMC Chair Search
Committee, c/o Jamie Edge, 4150 HSLC,
750 Highland Avenue, Madison, WI,
53705-2111, jledge@wisc.edu.
To receive full consideration, applications
should arrive by February 1, 2012.
Unless confidentiality is requested in writing, information regarding
applicants must be released upon request. Finalists cannot be
guaranteed confidentiality. Wisconsin Caregiver Law applies.
The University of Wisconsin is an equal opportunity, affirmative
action employer. For more information: www.med.wisc.edu,
http://www.bmolchem.wisc.edu/index.html
o
n
l
i
n
e

@
s
c
i
e
n
c
e
c
a
r
e
e
r
s
.
o
r
g
Heal the sick, advance the science, share the knowledge.
The Division of Pulmonary and Critical Care Medicine, in
conjunction with the Mayo Clinic Cancer Center in Rochester,
MN, is seeking a highly productive Ph.D. and/or M.D. scientist
to lead a developing Program in Lung Cancer. The ideal
candidate will be an established investigator at the Associate or
Full Professor level with a strong record of extramural funding,
exceptional productivity in cancer research and leadership
experience. Academic appointment is available in any of the
Mayo Graduate School of Medicine basic science departments.
The Mayo Clinic has over 200 biomedical research laboratories,
institutionally supported state-of-the-art animal, molecular and
microscopic core facilities, a number of unique human disease
tissue banks and opportunities to readily translate discovery to
the bedside. Very competitive start-up and sustained intramural
funding will be provided.
Interested applicants should submit curriculum vitae,
description of research plans and names and addresses of three
references by December 1, 2011 online at: www.mayoclinic.
org/physician-jobs (job #6442BR), or email to:
Ms. Trish Iverson
Search Committee Secretary
Email: iverson.patricia@mayo.edu
Mayo Foundation is an affrmative action and equal opportunity employer and
educator. Post-offer/pre-employment drug screening is required.
Faculty Position in Cell/Molecular
Biology of Lung Cancer
Your
career is
our cause.
Job Postings
Job Alerts
Resume/CV
Database
Career Advice
Career Forum
www.sciencecareers.org
Get help
fromthe
experts.
o
n
l
i
n
e

@
s
c
i
e
n
c
e
c
a
r
e
e
r
s
.
o
r
g
Why not
Postdoctoral Position in
Astronomy/Astrochemistry
NY Center for Astrobiology
Rensselaer Polytechnic Institute
The successful applicant will join a team seeking
to identify major chemical pathways that lead
from simple molecules in interstellar clouds to
complex organic molecules in protoplanetary
disks around new stars. The appointee will partic-
ipate in acquisition and analysis of new and exist-
ing astronomical observations, modeling, and
publication of results. APh.D. (or foreign equiva-
lent) in astrophysics or a related field is required;
prior relevant research experience is advanta-
geous. The position will be for one year with
expectation of renewal.
Apply by December 15, 2011 to Postdoctoral
Research Associate, NY Center for
Astrobiology, search number 20110001 at:
https://rpijobs.rpi.edu/applicants/jsp/shared/
frameset/Frameset.jsp?time=1313683304474
Vitae, publications, research interests and names
of 3 professional references may also be mailed
to: AM Strack, Astrobiology, Cogswell, RPI,
110 8th Street, Troy, NY 12180 or by email to
straca@rpi.edu
Rensselaer is an AA/EOE
change the world?
The University invites applications for two tenure track Assistant Professor positions in RNAscience
and technology.
CELL/DEVELOPMENTALBIOLOGIST, Department of Biological Sciences: conducting research
on the role of RNA, including but not exclusive to non-coding or microRNAmolecules, in post-transcrip-
tional gene regulation or other cellular and/or developmental processes.
CHEMIST/BIOCHEMIST, Department of Chemistry: conducting research in RNA science and its
applications in areas such as, but not limited to, modified nucleosides, synthesis, imaging, and analytical
chemistries as it pertains to RNAstructure/function relationships, including interactions with proteins and
other RNAs.
Both positions will be affiliated with the RNAInstitute (http://www.albany.edu/rna) with state-of-the-
art laboratories housed in the Life Sciences Research Building (http://www.albany.edu/lifesciences).
The Institute brings together more than 35 investigators from the College of Arts & Sciences, the College
of Nanoscale Science and Engineering, the School of Public Health, and regional institutions including
the Wadsworth Center, Rensselaer Polytechnic Institute, andAlbany Medical College. This creates an out-
standing environment for research collaborations.
Instructional responsibilities will be consistent with the position and those of the faculty in the home
department, and the interests of the candidate.
Submit applications for CELL/DEVELOPMENTALBIOLOGY at:
http://albany.interviewexchange.com/jobofferdetails.jsp?JOBID=27938
Submit applications for CHEMIST/BIOCHEMIST at
http://albany.interviewexchange.com/jobofferdetails.jsp?JOBID=27907
Applications must include a CV with publications cited in detail and any present or past grant funding,
statement of research interests, statement of teaching interests, and a minimum of three references with
contact information.
The successful candidates for both positions will be offered a competitive salary, start-up package, and
research space in the Life Sciences Research Building.
Qualifications for both Candidates: Ph.D. froma college or university accredited by the U.S. Department
of Education or an internationally recognized accrediting organization and a strong publication record
reflecting significant scientific accomplishments. Applicants must address in their applications their abili-
ty to work with and instruct a culturally diverse population. Preferred qualifications include productive
post-doctoral training and the potential or demonstrated ability, to obtain independent extramural funding.
Review of applications will begin on November 15, 2011 and continue until the positions are filled.
The University at Albany is an EEO/AA/IRCA/ADA employer.
Two Faculty Positions in RNA Research
The Biology Department at City College of the City University of New York invites applications for a tenure-track or tenured position
in Molecular Neurobiology at the level of either Assistant or Associate Professor to begin Fall 2012. We are searching for an
outstanding molecular neurobiologist performing cutting-edge research in fundamental cellular or developmental processes in the
nervous system. Candidates should have demonstrated research excellence and collaborative skills to interact with a vibrant,
expanding neurobiology group. The candidates research program should complement the current departmental research in
molecular neurobiology, systems neurobiology, and behavior. The successful candidate will be expected to teach in both
undergraduate and doctoral programs and work collaboratively within the City University of New York. For areas of departmental
strengths, see www.sci.ccny.cuny.edu/biology.
QUALIFICATIONS
Junior candidates should have a Ph.D., postdoctoral experience, and a strong record of publications; senior candidates should have
a strong history of federal funding, research productivity, and teaching at the undergraduate and graduate level.
COMPENSATION
Commensurate with qualifications and experience. Competitive start-up package available.
HOW TO APPLY
If you are viewing this job posting in CUNYFirst, please click on "Apply Now" on the bottom of this page and follow the instructions.
If you are viewing this job posting externally, please apply as follows:
- Go to www.cuny.edu and click on "Employment"
- Click "Search job listings"
- Click on "More options to search for CUNY jobs- Search by Job Opening ID number
- Click on the "Apply Now" button and follow the instructions.
To be considered for this position, you must include a curriculum vitae (CV), summary of past research accomplishments and future
research plans, and a statement of teaching and mentoring experience in one document in any of the following formats: doc, .docx,
.pdf, .rtf, or text format.
Letters of recommendation from at least three referees should be sent directly to the search committee at:
Molecular Neurobiology Search Committee - Department of Biology, J526
The City College of New York, 160 Convent Avenue, New York, NY 10031
biosearches@sci.ccny.cuny.edu
fax: 212 650-8585
EQUAL EMPLOYMENT OPPORTUNITY
We are committed to enhancing our diverse academic community by actively encouraging people with disabilities, minorities,
veterans, and women to apply. We take pride in our pluralistic community and continue to seek excellence through diversity and
inclusion. EO/AA Employer.
Assistant or Associate Professor
Molecular Neurobiologist
Job Opening IDNumber: 4678
Closing Date: Until Filled; Applications will be reviewed starting December 1, 2011
The City College, host site of the New York Structural
Biology Center, home campus of The City University of
New York Macromolecular Assemblies Institute, and a
CUNY flagship campus, invites applications for a tenure-
track faculty position in Structural Biology with emphasis
on Electron Microscopy and/or X-ray crystallography. We
are seeking a candidate who will develop a world-class,
externally funded research program that aims to
elucidate structure/function correlations in biomolecules.
This Chemistry Department position may be at the rank
of Assistant or Associate Professor. The new faculty
member will be expected to teach both undergraduate
and graduate level courses and mentor undergraduate
and graduate students. The successful candidate should
have a Ph.D. in biophysics, chemistry, or biochemistry
and substantial postdoctoral experience in structural
biology or related fields and ability to build a productive
externally funded research program.
HOW TO APPLY
Please apply as follows:
- Go to www.cuny.edu and click on "Employment"
- Click "Search job listings"
- Click on "More search options"
- Search by Job Opening ID number
- Click on the "Apply Now" button and follow the
instructions.
To be considered for this position, a curriculum vitae
(CV), list of publications, a description of current and
future research plans, a description of teaching
philosophy and 2-3 representative journal articles should
be combined and uploaded as a single PDF. The 3 letters
of reference should be sent directly by email to Ms.
Denise Addison (dma@sci.ccny.cuny.edu).
Assistant or Associate Professor
Structural Biology
Tenure Track
Job Opening ID Number: 4543
Closing Date: November 15, 2011
o
n
l
i
n
e

@
s
c
i
e
n
c
e
c
a
r
e
e
r
s
.
o
r
g
POSITIONS OPEN
Outstanding POSTDOCTORAL FELLOWS are
sought to conduct research into the pathogenesis of
AIDS. Using the nonhuman primate model of AIDS
we are examining host and viral determinants and
mechanisms that underlie host control of and protec-
tion from infection.
The successful candidate will have a DVM, M.D., or
Ph.D. in the life sciences along with evidence of first
author papers published or accepted in peer-reviewed
journals. The candidate will also have excellent written
and verbal communication skills and analytical capa-
bilities and a solid understanding of immunology and
virology and techniques such as polychromatic flow
cytometry, immunohistochemistry, and in situ hybrid-
ization. Molecular biology skills will also be useful. To
apply, send a cover letter and curriculum vitae and the
names of three individuals who may be contacted for
references to Andrew A. Lackner, DVM, Ph.D.; Pro-
fessor of Microbiology, Immunology and Pathology;
Director, Tulane National Primate Research Center at
e-mail: rita@tulane.edu.
Tulane University and Louisiana State University are Affir-
mative Action/Equal Opportunity Educators and Employers. In-
dividuals from under-represented minorities are strongly encouraged
to apply.
FACULTY POSITION in
Microbiology
San Jose State University
San Jose State University (SJSU) invites applications
for a tenure-track ASSISTANT PROFESSOR position
in microbiology to begin August 20, 2012 (website:
http://www.sjsu.edu/facultyaffairs/Unit_3/
Tenure_Track/Employment/index.htm). Preference
will be given to applicants in such areas as eukaryotic
microbiology, virology, microbial genomics, microbial
proteomics, and host-microbial areas will also be con-
sidered. For full consideration send a letter of appli-
cation, curriculum vitae, university undergraduate and
graduate transcripts, statement of teaching interests/
philosophy and research interests, and at least three of-
ficial letters of reference with contact information to:
Cleber Ouverney, Department of Biological Sciences,
San Jose State University, San Jose, CA95192-0100.
E-mail: cleber.ouverney@sjsu.edu. For full considera-
tion, apply by November 4, 2011. Please include Job
Opening Identification (JOID) 14215 on all corre-
spondence. SJSU is an Equal Opportunity/Affirmative Action
Employer committed to the core values of inclusion, civility, and
respect for each individual.
FACULTY POSITION in
Neurophysiology
San Jose State University
San Jose State University (SJSU) invites applications
for a tenure-track ASSISTANT PROFESSOR posi-
tion in neurophysiology to begin August 20, 2012
(website: http://www.sjsu.edu/facultyaffairs/
Unit_3/Tenure_Track/Employment/index.htm).
Preference will be given to applicants whose back-
grounds complement and augment existing faculty and
programs. Experience in vertebrate neurophysiology is
essential. For full consideration send a letter of appli-
cation, curriculum vitae, university undergraduate and
graduate transcripts, statement of teaching interests/
philosophy and research interests, and at least three
official letters of reference with contact information
to: Michael Sneary, Chair, Department of Biolog-
ical Sciences, San Jose State University, San Jose, CA
95192-0100. E-mail: michael.sneary@sjsu.edu. For
full consideration, apply by November 4, 2011. Please
include Job Opening Identification (JOID) 14216 on
all correspondence. SJSU is an Equal Opportunity/Affirmative
Action Employer committed to the core values of inclusion, civility,
and respect for each individual.
POSITIONS OPEN
ASSISTANT PROFESSOR
of Genome Biology
The School of Biological Sciences at Illinois
State University (website: http://www.bio.
illinoisstate.edu) invites applications for a
tenure-track faculty position at the level of As-
sistant Professor in the area of Genome Biol-
ogy. The successful applicant should be engaged
in research that blends bioinformatics with mo-
lecular experimentation to address fundamental
questions in broad areas of genome structure,
function, evolution, or related area. We seek ap-
plicants with a Ph.D., or equivalent, and post-
doctoral experience, a demonstrated potential to
secure external funding and an interest in work-
ing within a diverse intellectual community. The
successful applicant will be expected to devel-
op an independent externally funded research
program, and be involved in both graduate and
undergraduate teaching and mentoring. The
School of Biological Sciences at Illinois State
University is currently home to 85 M.S. and
Ph.D. candidates and over 500 B.S. majors. To
apply, send descriptive cover letter, curriculum
vitae, a one- to two-page statement of future re-
search goals, and contact information for three
references as a single PDF file to Dr. Wade
Nichols, c/o Sally Little via e-mail: salitt2@
ilstu.edu. Review of applications will begin on
November 1, 2011 and continue until the po-
sition is filled. Intended start date August 16,
2012. Illinois State University is an Equal Opportunity
University encouraging diversity.
ASSISTANT/ASSOCIATE PROFESSOR
Department of Neuroscience & Physiology
State University of New York
Upstate Medical University
We seek applications to fill two tenure-track posi-
tions at either the Assistant or Associate Professor
level from individuals studying any area of Neurosci-
ence. The successful applicants will be expected to de-
velop well-funded research programs and to contribute
to graduate and medical teaching. We offer a highly
competitive startup package and salary. Appointment
at the Associate Professor level will require demonstra-
tion of outstanding achievement and current extramural
funding. Further information about the Department
can be found at website: http://www.upstate.edu/
neuroscience/.
Candidates should have a Ph.D. or equivalent, post-
doctoral experience, and a strong publication record.
Applicants should e-mail a PDF file containing cur-
riculum vitae, summary of research accomplishments,
and future research plans to e-mail: neurosci@upstate.
edu. In addition, three letters of reference should be
addressed to: Dr. Barry E. Knox, Chair, Department
of Neuroscience & Physiology, WH 3223, 750 East
Adams Street, Syracuse, New York 13210.
Review of applications will begin November 1, 2011,
and continue until the positions are filled.
SUNY Upstate Medical University is an Affirmative Action/
Equal Opportunity Employer engaging excellence through diver-
sity. Women and minorities are strongly encouraged to apply.
MEDICAL DIRECTOR
Physician, biomedical researcher, or other medical/
bioscience professional sought by Manhattan family
to research and coordinate family medical and health
care issues. This person will manage a small team of
professionals and interface with physicians, medical re-
searchers, and consultants (in academia and otherwise)
to ensure delivery of highest-quality medical care to
family members. Considerable weight will be given
to unusual academic distinction and other intellectu-
al achievements. Excellent communication skills are a
must, a Ph.D. or M.D. is strongly preferred, and clini-
cal experience is a plus. This is a full-time position
with a highly attractive compensation package and sig-
nificant upside potential. Resume to e-mail: mdrecr@
gmail.com.
POSITIONS OPEN
The Department of Biological Sciences, Uni-
versity of Denver (DU), invites applications for
an Ecologist in a tenure-track position at the
ASSISTANT PROFESSOR level to begin Sep-
tember 1, 2012. We are particularly interested
in candidates in the fields of population ecology,
community ecology, plant-animal interactions,
evolutionary ecology, and physiological ecolo-
gy. Field-oriented ecologist studying whole or-
ganisms; potential for research in the Rocky
Mountain region is especially desirable. The suc-
cessful candidate will have a Ph.D. and post-
doctoral experience in appropriate fields, will
develop an extramurally funded research program,
will supervise undergraduate research projects
and M.S. and Ph.D. students, and will teach
undergraduate and graduate courses in ecology
and specific areas of specialty. All candidates
must submit their application through website:
https://www.dujobs.org. The online applica-
tion should include: curriculum vitae, statements
of teaching philosophy and research interests
and two recent publications. Under a separate
cover, send three letters of recommendation
to: Dr. Shannon Murphy, Chair, Ecologist
Search Committee, Department of Biolog-
ical Sciences, University of Denver, Denver,
CO 80208. Review of applications will begin
December 1, 2011. The University of Denver is com-
mitted to enhancing the diversity of its faculty and staff and
encourages applications from women, minorities, people
with disabilities and veterans. DU is an Equal Employ-
ment Opportunity/Affirmative Action Employer. Please
see our extensive benefit package at website: http://
www.du.edu/hr/benefits.
ASSISTANT PROFESSOR
Institute of Bioinformatics and the
Department of Computer Science
University of Georgia
The Institute of Bioinformatics and the Depart-
ment of Computer Science at the University of Georgia
invite applications for a joint tenure-track faculty po-
sition at the Assistant Professor level, to start August
2012. The tenure home is in Computer Science. We are
interested in strengthening and complementing our
existing broad program in Bioinformatics at the Uni-
versity of Georgia. The candidate should have a Ph.D.
or equivalent degree in Computer Science, Bioinfor-
matics or a closely related field and a strong research
track record at the interface of computer science and
biology. The candidate will be expected to maintain a
rigorous, externally funded research program and to
contribute to undergraduate and graduate teaching
in both Computer Science and Bioinformatics degree
programs. For information about the breadth of the
units, see websites: http://iob.uga.edu and http://
cs.uga.edu.
Applications should be uploaded to website:
http://recruitment.franklin.uga.edu and should in-
clude a cover letter, curriculum vitae, and brief state-
ments of research and teaching interests. Three letters
of recommendation should be uploaded separately
to the same website. The committee will begin re-
viewing applications on November 15, 2011 until the
position is filled.
The Institute of Bioinformatics, the Department of Computer
Science, the Franklin College of Arts and Sciences, its many units,
and the University of Georgia are committed to increasing the
diversity of its faculty and students, and to sustaining a work and
learning environment that is inclusive. Women, minorities and
people with disabilities are encouraged to apply. The University is
an Equal Employment Opportunity/Affirmative Action Institution.
Translational Neuroscience: The University of Ari-
zona seeks to hire a tenure-eligible physician-scientist
at the ASSISTANT or ASSOCIATE level in Neu-
roscience. Visit website: https://www.uacareertrack.
com (Job Number 48123) or contact Search Com-
mittee Chair, Carol A. Barnes, Ph.D. at e-mail: carol@
nsma.arizona.edu. Affirmative Action/Equal Opportunity
Employer.
o
n
l
i
n
e

@
s
c
i
e
n
c
e
c
a
r
e
e
r
s
.
o
r
g
21 OCTOBER 2011 VOL 334 SCIENCE www.sciencecareers.org 414
Read inspiring stories of women working
in Green Science who are blending
a unique combination of enthusiasm for
science and concern for others
to make the world a better place.
This booklet is brought to you by the
AAAS/Science Business Ofce in partnership
with the LOreal Foundation
Download this free booklet
ScienceCareers.org/LOrealWiS
WOMEN IN SCIENCE
forging new pathways
in green science
Download your free copy today at
ScienceCareers.org/booklets
Brought to you by the
AAAS/Science Business Ofce
o
n
l
i
n
e

@
s
c
i
e
n
c
e
c
a
r
e
e
r
s
.
o
r
g
POSITIONS OPEN
The Department of Biological Sciences (website:
http://biology.fiu.edu) at Florida International Uni-
versity (FIU) is seeking applicants for a full-time,
OPEN RANK FACULTY POSITION in molecu-
lar biology to begin Fall 2012. The position is part of
the strategic initiative of the School of Integrated Science
and Humanity (SISH) to support the planned Insti-
tute of Biomolecular and Biomedical Sciences (IBBS).
The area of molecular biology is broadly defined and in-
cludes research that will build the department_s strength
in genomics, proteomics, and biomedical applications.
The successful candidate will have a strong record of
scholarly activity and external funding. Rank is open,
but preference will be given to senior applicants who
are attracted to the challenges of playing a leadership
role in building the IBBS and fostering its multidis-
ciplinary environment. Advanced junior applicants with
strong research credentials and demonstrated poten-
tial for obtaining external funding will also be consid-
ered. A Ph.D. in the appropriate discipline is required.
Interested candidates should apply online at website:
http://www.fiujobs.org and reference Position Num-
ber 35539. Please include curriculum vitae, list of
publications, contact information for three potential
references, and a letter describing research interests. In-
quiries may be addressed to Dr. Walter Van Hamme,
Chair, Biomolecular Science Search Committee (e-mail:
vanhamme@fiu.edu). Review of applications will be-
gin on December 1, 2011, and continue until the po-
sition is filled.
FIU is a multi-campus public research university lo-
cated in Miami, a vibrant, international city.It has a
newly established medical school that emphasizes med-
ical research and community based medicine. FIU of-
fers more than 180 baccalaureate, masters, professional,
and doctoral degree programs to over 42,000 students.
FIU is a member of the State University System of Florida
and is an Equal Opportunity/Equal Access/Affirmative Action
Employer.
TWO NEUROBIOLOGY POSITIONS: As
part of a University-wide expansion in neuroscience
research, West Virginia University (WVU) invites ap-
plications for two tenure-track positions at the AS-
SISTANT PROFESSOR level in the Department
of Biology, to begin August 2012. These individuals
will join a rapidly-growing neuroscience community
at WVU, which numbers almost 50 full-time faculty
(website: http://www.hsc.wvu.edu/wvucn/). We
seek a systems-level neurophysiologist and a compu-
tational neuroscientist. The neurophysiologist will use
a multidisciplinary approach in non-mammalian or
invertebrate systems to ask integrative questions in either
sensory or motor neurobiology. For the computa-
tional position, we are especially interested in someone
working at the level of cells, circuits, and/or systems.
For both positions, we seek individuals with a broad
biology background and strong interdisciplinary skills.
They will develop externally funded, independent re-
search programs, and contribute to the undergraduate
and graduate teaching missions of the department.
Two years of postdoctoral experience, excellent writ-
ten and oral communication skills, and the potential
to secure external funding are required. Qualified ap-
plicants should submit research and teaching state-
ments, curriculum vitae, and arrange for three letters
of reference to be sent to e-mail: wvubiology@mail.
wvu.edu. Review of applications will commence on
November 15 and continue until the position is filled.
For more information about the position, contact
Jim Belanger, e-mail: jim.belanger@mail.wvu.edu.
For more information about WVU and the city of
Morgantown, West Virginia website: http://www.
as.wvu.edu/biology/faculty/positions.htm.
Women, minorities, and persons with disabilities are strongly
encouraged to apply, and the university is supportive of the needs
of dual career couples. West Virginia University is an Equal Op-
portunity/Affirmative Action Employer and the recipient of an
NSF ADVANCE award for gender equity.
POSITIONS OPEN
MARINE GENETICIST
The Florida International University (FIU) is seek-
ing applicants for a Marine Geneticist (rank open, senior
applicants are encouraged) for a tenure-track position
in the Department of Biological Sciences (website:
http://casgroup.fiu.edu/Biology/). He or she will
participate in the Marine Science Program (website:
http://casgroup.fiu.edu/marine/), a newand grow-
ing interdisciplinary initiative emphasizing research
and teaching in coastal marine science. The Marine
Science Program is housed in a recently completed
building in FIU Biscayne Bay Campus in the city of
North Miami, and currently is home to 12 research
laboratories, two teaching laboratories, wet labs, a
mesocosm facility, and running sea water systems.
The successful candidate will be expected to maintain
an externally funded research program, supervise grad-
uate students in our Ph.D. program, as well as teach
undergraduate courses including General Genetics
and other courses in their areas of expertise. Areas of
prospective research foci may include but are not lim-
ited to: genomics/phylogenetics, metagenomics, popu-
lation genetics, ecological genetics, and environmental
microbiology. To ensure full consideration, applications
should be received by November 21, 2011. Screen-
ing of applications will begin on that date and continue
until a suitable candidate is selected. Applications will
only be accepted electronically as PDF files. Applica-
tions should include a cover letter, curriculum vitae, a
summary of research interests and teaching goals.
Please submit applications to e-mail: trexlerj@fiu.edu.
In addition, applicants should arrange for three let-
ters of reference to be sent directly to the same e-mail
address. Interested applicants are also required to
submit their applications online via website: http://
www.fiujobs.org reference Position Number 35538.
FIU is a member of the State University System of Florida
and is an Equal Opportunity/Equal Access/Affirmative Action
Employer.
RANK-OPEN FACULTY POSITION
in Bioinformatics
Florida International University
The Department of Biological Sciences at Florida
International University (FIU) (website: http://
casgroup.fiu.edu/Biology/) is seeking applicants
for an open-rank tenure-track position in Bioinfor-
matics. The successful candidate will be expected to
maintain an externally funded research program, su-
pervise graduate students in our Ph.D. program, as
well as teach undergraduate courses including genet-
ics and other courses in their areas of expertise. Areas
of prospective research foci may include but are not
limited to: genomics, metagenomics, proteomics, com-
putational biology, systems biology, genome-wide as-
sociation mapping, and phylogenetics. Candidates
utilizing theoretical approaches are encouraged to ap-
ply. To ensure full consideration, applications should
be received by November 21, 2011. Screening of ap-
plications will begin on that date and continue until a
suitable candidate is selected. Applications will only
be accepted electronically as PDF files. Interested ap-
plicants should submit a (1) cover letter, (2) curricu-
lum vitae, (3) statement of research interests, teaching
goals, and service interests, and (4) arrange to have
three or more references sent independently to Eric
von Wettberg at e-mail: eric.vonwettberg@fiu.edu.
Interested applicants are also required to submit their
applications online via website: http://www.fiujobs.
org, reference Position Number 35537. FIU is a mem-
ber of the State University System of Florida and is an Equal
Opportunity/Equal Access/Affirmative Action Employer.
We deliver
customized job alerts.
www.ScienceCareers.org
POSITIONS OPEN
FACULTY POSITION BIOCHEMISTRY
The Department of Chemistry and Biochemistry at
Auburn University invites applications for a tenure-
track, nine-month position in biochemistry at the level
of ASSISTANT PROFESSOR. The candidate is
expected to develop a vigorous, externally funded re-
search program. Although we are particularly inter-
ested in persons with expertise in biomacromolecular
X-ray crystallography, we welcome applications from
individuals with research interests in any area of bio-
chemistry. Duties also include teaching at the under-
graduate and graduate levels. A Ph.D. in chemistry or
biochemistry and at least one year of postdoctoral ex-
perience are required. The candidate selected for this
position must meet eligibility requirements to work in the
United States on the date the appointment is scheduled to
begin (August 2012) and continue working legally for
the proposed term of employment. Excellent communica-
tion skills required. Minorities and women are encour-
aged to apply. Applicants should submit curriculum
vitae, statements of research plans and teaching phi-
losophy, and have three letters of reference sent elec-
tronically to e-mail: bchesearch@auburn.edu or by
mail to: Biochemistry Search Committee, Depart-
ment of Chemistry and Biochemistry, Auburn
University, AL 36849-5312. Review of applications
will begin November 28, 2011, and continue until the
position is filled. Auburn University is an Affirmative Action/
Equal Opportunity Employer.
RANK-OPEN FACULTY POSITION
in Comparative Immunology
Florida International University
The Department of Biological Sciences at Florida
International University (FIU) is seeking a Compara-
tive Immunologist (open rank) to fill a tenure-track
position. All candidates must have postdoctoral expe-
rience and strong publication and funding records.
Senior candidates should have a demonstrated his-
tory of leadership and consistent extramural funding
in the field of comparative immunology. The depart-
ment has a broad range of faculty for potential col-
laborative interactions (website: http://casgroup.
fiu.edu/Biology/). Teaching expectations include a
general undergraduate course in immunobiology and
graduate specialty courses. All application materials
should be submitted electronically as PDF files to:
Dr. Laurie Richardson, Chair, Comparative Immu-
nology Search Committee, e-mail: laurie.richardson@
fiu.edu and received by November 21, 2011. Each
application should include a cover letter, curriculum
vitae, and a summary of professional and teaching in-
terests. In addition, applicants should arrange for
three to four letters of reference to be sent directly
to the same e-mail address. Interested applicants are
also required to submit their applications online via
website: http://www.fiujobs.org reference Posi-
tion Number 34952. FIU is a member of the State Uni-
versity System of Florida and is an Equal Opportunity/Equal
Access/Affirmative Action Employer.
POSTDOCTORAL FELLOWSHIPS
The Geophysical Laboratory, Carnegie Institution
of Washington, invites applications for postdoctoral
fellowships. The Geophysical Laboratory emphasizes
interdisciplinary experimental and theoretical research
in fields spanning geoscience, microbiology, chemistry,
and physics. The Laboratory supports world-class fa-
cilities in high-pressure research; organic, stable isotope
and biogeochemistry; mineral physics and petrology;
and astrobiology.
Please see website: http://www.gl.ciw.edu/
employment/Postdoctoral_Positions for informa-
tion on the application process. Also, see website:
http://www.gl.ciw.edu/ for a listing of personnel,
current research interests, major facilities, and appli-
cation information.
Completed applications for Carnegie fellowships
should be submitted by December 31, 2011, to:
Russell J. Hemley, Director, Attn: Danielle Appleby,
Geophysical Laboratory, 5251 Broad Branch Road,
NW, Washington, DC 20015-1305, USA (e-mail:
dappleby@ciw.edu). E-mail is preferred.
The Geophysical Laboratory is an Equal Opportunity
Employer.
o
n
l
i
n
e

@
s
c
i
e
n
c
e
c
a
r
e
e
r
s
.
o
r
g
21 OCTOBER 2011 VOL 334 SCIENCE www.sciencecareers.org 416
V
i
s
i
t
o
u
r
E
N
H
A
N
C
E
D
W
E
B
S
I
T
E
!
ScienceCareers.org
Science Careers
is the forum that
answers questions.
Science Careers is dedicated to
opening new doors and answering
questions on career topics that
matter to you. Were the go-to
career site for connecting with top
employers, industry experts, and
your peers. Were the source for
the latest and most relevant career
information across the globe.
With community feedback and
a professional atmosphere, our
careers forum allows you to
connect with colleagues and
associates to get the advice
and guidance you seek.
Science Careers Forum:
Relevant Career Topics
Advice and Answers
Community, Connections,
and More!
Visit the forum and get your
questions answered today!
Your Future
Awaits.
Come work with us!
http://web.mit.edu
Faculty Searches
Department of Earth,
Atmospheric and Planetary
Sciences
Applications for all of the following positions
are being accepted at Academic Jobs
Online (https://academicjobsonline.org/ajo).
Please do not ask your referees to upload
letters at the time of application; letters will
be requested directly by MIT. To receive
consideration, a completed application
must be received.
Faculty Positions Climate-related Fields
The MIT Department of Earth, Atmospheric
and Planetary Sciences announces a major
expansion of its activities in climate science
and seeks applicants for up to three faculty
positions in climate-related fields. Preference
will be given to junior appointments at the
assistant professor level, but a senior appointment
can be considered for an individual with
exceptional qualifications. Areas of specific
interest include observations, models and
theory of the atmosphere, ocean and cryosphere,
and climates, biogeochemical cycles, and
ecology. Successful candidates must have
a strong record of accomplishment in their
discipline, a strong commitment to teaching
and student advising, a keen interest in relating
their work to complementary research in the
Department and in the MIT/Woods Hole Joint
Program in Oceanography. Joint appointments
with other MIT departments are also potentially
negotiable where appropriate.
A completed application will include a
curriculum vitae, a statement of research and
teaching objectives, and the names of five
potential referees. More information about this
position can be obtained by writing Professor
Kerry A. Emanuel at emanuel@mit.edu.
Tenure-Track Junior Faculty Position
Exoplanets
We seek an outstanding scientist with interest
in and potential for innovation and leadership
in teaching and research. The search is in the
broad area of exoplanets, including theory,
observation, and instrumentation. However,
we are especially interested in individuals
whose research complements existing MIT
expertise. The Department encourages strong
interaction with research programs within the
Earth, Atmospheric, and Planetary Sciences.
A completed application will include a
curriculum vitae, a one-page description
of research plans, and the names of three
potential referees. The deadline for this
opening will be December 31, 2011; all
applications received by this date will receive
full consideration. More information about this
position can be obtained by writing Professor
Sara Seager at seager@mit.edu.
Junior Faculty Position
Sedimentary Geology
We seek an individual with a field-based
observational program, broad research interests,
and a commitment to interdisciplinary studies.
Applicants should submit a curriculum vitae,
one-page descriptions of research and teaching
plans, and the names, email addresses, and
phone numbers of three professional referees.
Review of applications will take place beginning
October 1, 2011, but will not continue past
March 1, 2012. Questions may be addressed
to Prof. Leigh Royden, Search Committee
Chair, at lhroyden@mit.edu.
Search contact for all of the above-listed
positions: Mr. Michael Richard, HR
Administrator, EAPS, Massachusetts Institute
of Technology, 54-926, 77 Massachusetts
Avenue, Cambridge, MA 02139-7307;
mjr@mit.edu; 617-253-5184; 617-253-8298 (fax).
MIT is an Equal Opportunity/Affirmative Action
employer. Applications from women and
underrepresented minority candidates are
encouraged.
o
n
l
i
n
e

@
s
c
i
e
n
c
e
c
a
r
e
e
r
s
.
o
r
g
POSITIONS OPEN
FACULTY POSITION IN GENETICS
Department of Biological Sciences
The George Washington University
The Department of Biological Sciences at the George
Washington University is accepting applications for a
tenure-track faculty position at the rank of ASSIST-
ANT PROFESSOR with expertise in the field of Ge-
netics. We are searching for candidates who use genetic
or genomic experimental methods to address funda-
mental questions about gene expression or developmen-
tal biology or who are able to integrate gene function
and regulation of complex traits from the genome to
the phenotype level. Research activities should be
conducted within the context of comparative and/or
evolutionary biology and complement the research fo-
cus of the department. The successful candidate will
be expected to establish and maintain an externally
funded research program that involves undergraduate
and graduate students. Teaching responsibilities will
include an undergraduate introductory course in ge-
netics that includes a lab. Basic Qualifications: a Ph.D.
in an appropriate discipline, postdoctoral experience,
ability to teach basic genetics, and accomplishment
in biological research in genetics as demonstrated by
publications in peer-reviewed journals. Application Pro-
cedure: to be considered please send electronically a
letter of application, a complete curriculum vitae, brief
descriptions of teaching and research plans, three pub-
lications, and the names and addresses of three people
who will be willing to submit letters of recommenda-
tion. Applications should be sent to the search chair
at e-mail: gwgeneticssearch@gmail.com. Only com-
plete applications will be considered. Review of appli-
cations will begin on November 18 and will continue
until the position is filled.
The George Washington University is an Equal Opportunity/
Affirmative Action Employer. The University Search Committee
seeks to attract an active, culturally, and academically diverse fac-
ulty of the highest caliber.
FACULTY POSITION IN CELL BIOLOGY
Department of Biological Sciences
The George Washington University
The Department of Biological Sciences at the
George Washington University is accepting appli-
cations for a tenure-track position at the rank of
ASSISTANT PROFESSOR in Cell Biology. We
seek broadly trained candidates working on problems
of cell-cell interactions including, but not limited to,
host-pathogen interactions, symbiosis, developmental
processes, or neurobiology. Research activities should
be conducted within the context of comparative and/
or evolutionary biology and complement the research
focus of the department. Preference will be given to
candidates who integrate cellular analyses with modern
imaging approaches. The successful candidate will
be expected to establish and maintain an externally
funded research program that involves undergraduate
and graduate students. Teaching responsibilities will
include an undergraduate, introductory course in cell
biology and a second course within the candidate_s
area of expertise. Basic Qualifications: a Ph.D. in an
appropriate discipline, postdoctoral experience, ability
to teach cell biology, and accomplishments in cell bi-
ology as demonstrated by publications in peer-reviewed
journals. Application Procedure: to be considered please
send electronically a letter of application, a complete
curriculum vitae, brief descriptions of teaching and re-
search plans, three publications, and the names and
addresses of three people who will be willing to submit
letters of recommendation. Applications should be
sent to the search chair at e-mail: gwcellbiolsearch@
gmail.com. Only complete applications can be eval-
uated. Assessment of applications will begin on
November 18, 2011, and will be ongoing until the
position is filled.
The George Washington University is an Equal Opportunity/
Affirmative Action Employer. The University Search Committee
seeks to attract an active, culturally, and academically diverse fa-
culty of the highest caliber.
POSITIONS OPEN
FACULTY POSITION
in Animal Behavior (Invertebrate)
Department of Biological Sciences
The George Washington University
The Department of Biological Sciences of the George
Washington University is accepting applications for a
tenure-track faculty position at the rank of ASSIST-
ANT PROFESSOR with expertise in the field of
Animal Behavior, focusing on invertebrates. We are
searching for candidates who use comparative and/or
experimental approaches to investigate any aspect of
invertebrate behavior, including but not limited to mat-
ing, foraging, feeding, communication, learning, ag-
gression, and cooperation. Applicants who examine
the genetic and/or neurological basis of behavior are
also encouraged to apply. Teaching responsibilities
include an undergraduate course with laboratory in
Animal Behavior and a second course in their area of
expertise. The successful candidate is expected to es-
tablish and maintain a vigorous research program
capable of attracting external funding that involves
graduate and undergraduate students and will add to
our growing departmental strength in environmental
biology and ecology. Basic Qualifications: a Ph.D. in
an appropriate discipline, postdoctoral experience, abil-
ity to teach animal behavior, and accomplishments in
biological research in animal behavior demonstrated
by publications in peer-reviewed journals. Applica-
tion Procedure: to be considered please send electron-
ically a complete curriculum vitae, brief descriptions
of teaching and research plans, three publications,
and the names and contact information for three ref-
erences to the search chair, Dr. John Lill, at e-mail:
gwbehaviorsearch@gmail.com.
Only complete applications will be considered. Re-
view of applications will begin on November 18, 2011,
and will continue until the position is filled.
The George Washington University is an Equal Opportunity/
Affirmative Action Employer. The University Search Committee
seeks to attract an active, culturally, and academically diverse fac-
ulty of the highest caliber.
UNIVERSITY OF CALIFORNIA, IRVINE-
FACULTY POSITION IN ATMOSPHERIC
CHEMISTRY. The Department of Chemistry invites
applications for a tenure-track position as ASSIST-
ANT PROFESSOR in the area of Atmospheric Chem-
istry. We are seeking an outstanding Ph.D. who will
establish a vigorous research program in areas, which
will complement existing strong programs in atmo-
spheric chemistry within the Department of Chem-
istry, and preferably also take advantage of a breadth
of related programs and collaborations on campus
(see website: http://airuci.uci.edu/). Typical (but not
exclusive) areas include the development of new an-
alytical methods, studies of the fundamental chemistry
and photochemistry of organic or inorganic atmo-
spheric systems, and molecular level studies of inter-
actions of outdoor or indoor atmospheric species with
biological or energy-related systems. The ability to
effectively teach chemistry at the undergraduate and
graduate levels is required. Applications must be sub-
mitted electronically via the Internet at website: https://
recruit.ap.uci.edu. Applicants should upload a cover
letter, curriculum vitae (including publication list), a
concise statement of research plans, and the names
and contact information for at least three references.
Applications and supporting materials should be
received by November 30, 2011 for full consideration.
The University of California, Irvine is an Equal Opportunity/
Affirmative Action Employer committed to excellence through di-
versity and strongly encourages applications from all qualified ap-
plicants, including women and minorities. UC Irvine has an active
ADVANCE Gender Equity Program.

Find your future here.

www.ScienceCareers.org
POSITIONS OPEN
COMMUNITY ECOLOGIST
The Department of Biological Sciences at Bowling
Green State University (BGSU) seeks applicants for an
ASSISTANT PROFESSOR level, tenure-track
Community Ecology position. Applicants with post-
doctoral experience and research expertise that com-
plements existing strengths in evolution, population,
and conservation ecology are preferred. Successful can-
didates are expected to develop a highly productive,
externally funded research program, as well as contrib-
ute to the teaching missions of our Ph.D./M.S. pro-
gram (80 students) and undergraduate program,
which includes a Specialization in Ecology and Con-
servation Biology. To apply, electronically submit cov-
er letter, curriculum vitae, statements of research plans
and teaching philosophy/experience, representative
publications, and three reference letters to e-mail:
dmclean@bgsu.edu or by mail to: Community Ecol-
ogist Search Committee, Department of Biological
Sciences, BGSU, Bowling Green, OH 43403-0208
by November 30, 2011. Contact Helen Michaels at
e-mail: hmichae@bgsu.edu for additional informa-
tion. Information about our department can be found
at website: http://www.bgsu.edu/departments/
biology. BGSU is an Affirmative Action/Equal Opportunity
Employer/Educator and encourages applications from women,
minorities, veterans, and persons with disabilities.
TENURE-TRACK POSITION in
Theoretical or Computational Biophysics
Department of Biophysics
Johns Hopkins University
The Thomas C. Jenkins Department of Biophysics
seeks candidates for a tenure-track, ASSISTANT PRO-
FESSOR position in theory and simulation of bio-
logical systems. Areas of special interest include, but
are not restricted to the application of theory and
computation to biological macromolecules and as-
semblies, cell dynamics, cellular physiology, or applica-
tions of statistical thermodynamics, polymer physics,
and physical chemistry in biology, systems biology,
and biological networks. Applicants should electron-
ically send a single PDF file with a cover letter, cur-
riculum vitae, and a brief description of research plans
to e-mail: biophysicssearch@jhu.edu. Three letters
of recommendation should be sent to this e-mail ad-
dress or to: Faculty Search Committee, T. C. Jenkins
Department of Biophysics, Johns Hopkins Univer-
sity, 3400 N. Charles Street, Baltimore, M.D. 21218-
2685; telephone: 410-516-7245. All materials should
be received by November 30, 2011.
Johns Hopkins University is an Affirmative Action/Equal
Opportunity Employer.
More
scientists agree
we are the most
useful website.

www.ScienceCareers.org
MARKET PLACE
Promab Biotechnologies Inc.
Custom Monoclonal
Antibody $4,200
>3,000 CLONES WILL BE SCREENED
1-866-339-0871
www.promab.com info@promab.com
o
n
l
i
n
e

@
s
c
i
e
n
c
e
c
a
r
e
e
r
s
.
o
r
g
21 OCTOBER 2011 VOL 334 SCIENCE www.sciencecareers.org 418
1200 NewYork Ave., NW Washington, DC 20005 +1 202-326-6417 ScienceTranslationalMedicine.org
AAASmember price:
US$52 online only
US$225 print and online (add US$100 for non-US delivery)
AAAS membership number required ______________________________
Nonmember price:
US$156 online only
US$465 print and online (add US$100 for non-US delivery)
Name ___________________________________________
Address _________________________________________
_________________________________________________
City _____________________________________________
State/Province ____________________________________
Zip/Postal Code ___________________________________
Country __________________________________________
E-mail ________________________________________
Phone ___________________________________________
Payment
Check (payable to AAAS Science Translational Medicine)
Mail check and this formto:
AAAS
Attn: Membership Department
1200 NewYork Avenue, NW
Washington, DC 20005 USA
Charge my:
VISA MasterCard American Express
Card Number________________________________________
Expiration Date______________________________________
Signature___________________________________________
Date _______________________________________________
If paying by credit card, you may FAST FAX your order to
+1 202-842-1065
Subscribe now for 1 years access to Science Translational Medicine
(required for subscription activation)
Science Translational Medicine, the newest weekly journal from AAAS,
is focused on applications of basic research knowledge to improve
human health.
The goal of Science Translational Medicine is simple: help the scientic
community harness decades of progress in research at the basic level
and translate these biological discoveries into medical advances.
Science Translational Medicine publishes:
Peer-reviewed primary research papers
Perspectives and reviews on research from basic science
and clinical viewpoints
Survey of recent literature and ndings in other journals
Commentary on policy, funding, regulatory issues, and more.
As a AAAS member, add Science Translational Medicine
access for over 60% off of the regular price. Subscribe
in any of these ways:
Go to ScienceTranslationalMedicine.org
Call +1 202-326-6417
Mail or fax this form with your
payment to +1 202-842-1065
ScienceTranslational Medicine
ScienceTranslationalMedicine.org
Weekly journal integrating science and clinical medicine
Prices valid until December 31, 2011.
Sitewide access available for
your institution. Recommend a
subscription to your library today.
ScienceOnline.org/recommend
Our stock is growing.
With 670 sandwich ELISAkits nowavailable covering
10 species, you have more choices under one roof.
More kits, more choices! Raybiotech now carries ELISA kits for human, rhesus macaque, mouse,
rat, cotton rat, pig, cow, horse, cat, and dog target proteins.
Pre-coated strip plates
Reproducible and reliable
Ideal for validation of antibody array data
Extensively validated in publications
Money back guarantee
ELISA testing service available!
Discover more. Publish faster.
90-day money-back guarantee. Applicable
to U.S. customers only. For details, visit:
www.raybiotech.com/risk_free_sci.asp
Try any RayBio

ELISA kit
RISK-FREE!