Ulm University

Department of Internal Medicine I

Director: Professor Dr. Thomas Seufferlein

Novel molecular insights to discern and target migrating

cancer stem cells in pancreatic ductal adenocarcinomas

Dissertation submitted in partial fulfilment of the requirements for the degree of

„Doctor rerum naturalium” (Dr. rer. nat.) of the International Graduate School in
Molecular Medicine Ulm

Kanishka Tiwary
Sareshkunda, India
2023
Dean of the medical faculty:
Prof. Dr. Thomas Wirth

Chairman of the graduate school:

Prof. Dr. Bernd Knöll

Thesis Advisory Committee:

1st supervisor: Prof. Dr. Dr. Patrick Christian Hermann
2nd supervisor: Prof. Dr. Lisa Wiesmüller
3rd supervisor: Prof. Dr. Bruno Sainz Jr.

External reviewer:
Dr. Shiv K. Singh
Prof. Dr. Dr. Daniel Stange

Day doctorate awarded:

30th January 2023
Information about previous publications

Results gained in my thesis have previously been published in the following

publication:

1. Walter, K., Tiwary, K., Trajkovic-Arsic, M., Hidalgo-Sastre, A., Dierichs, L.,
T. Liffers, S., Gu, J., Gout, J., Schulte, L.-A., Münch, J., Seufferlein, T., Sainz,
B.-Jr., Hermann, P.-C. MEK inhibition targets cancer stem cells and impedes
migration of pancreatic cancer cells in vitro and in vivo. Stem Cells Int. 2019
Jun 2, 8475389. doi: 10.1155/2019/8475389.

Licensed under a Creative Commons Attribution License (CC BY 4.0,

https://creativecommons.org/licenses/by/4.0/) ©2019 (Walter et al. 2019).
Ref. No. 1

Contributions to further publications acquired during my PhD period:

1. Godbole, M., Togar, T., Patel K., Dharavath B., Yadav, N., Janjuha, S., Gardi,
N., Tiwary, K., Terwadkar, P., Desai, R.-P., Dhamne, H., Karve, K.,
Salunkhe, S., Kawle, D., Chandrani, P., Dutt, S., Gupta, S., Badwe, R.-A.,
Dutt, A. Up-regulation of the kinase gene SGK1 by progesterone activates
the AP-1–NDRG1 axis in both PR-positive and -negative breast cancer cells.
J. Biol. Chem. 2018 Dec 14, 293 (50): 19263-19276. doi:
10.1074/jbc.RA118.002894.

2. Valle, S., Alcalá, S., Martin-Hijano, L., Cabezas-Sáinz, P., Navarro, D.,
Ramos Muñoz, E., Yuste, L., Tiwary, K., Walter, K., Ruiz-Cañas, L., Alonso-
Nocelo, M., Rubiolo, J.-A., González-Arnay, E., Heeschen, C., Garcia-
Bermejo, L., Hermann, P.-C., Sánchez , L., Sancho, P., Fernández-Moreno
M.- Á., Sainz, B.-Jr. Exploiting oxidative phosphorylation to promote the stem
and immunoevasive properties of pancreatic cancer stem cells. Nat.
Commun. 2020 Oct 16, 11, 5265. https://doi.org/10.1038/s41467-020-
18954-z.
3. Walter, K., Rodriguez-Aznar, E., Ferreira, M.S.V., Frappart, P.-O., Dittrich,
T., Tiwary, K., Meessen, S., Lerma, L., Daiss, N., Schulte, L.-A., Najafova,
Z., Arnold, F., Usachov, V., Azoitei, N., Erkan, M., Lechel, A., Brümmendorf,
T.H., Seufferlein, T., Kleger, A., Tabarés, E., Günes, C., Johnsen, S.A., Beier,
F., Sainz, B.-Jr., Hermann, P.-C. Telomerase and Pluripotency Factors
Jointly Regulate Stemness in Pancreatic Cancer Stem Cells. Cancers. 2021
Jun 23, 13, 3145. https://doi.org/10.3390/cancers13133145.

4. Tsesmelis, M., Tiwary, K., Steiger, K., Sperb, N., Gerstenlauer, M., Manfras,
U., Maier, H.J., Hermann, P.C., Chan, L.K., Wirth, T. Deletion of NEMO
Inhibits EMT and Reduces Metastasis in KPC Mice. Cancers. 2021 Sep
10, 13, 4541. https://doi.org/10.3390/cancers13184541.
 TABLE OF CONTENTS

1 TABLE OF CONTENTS
2 LIST OF ABBREVIATIONS III
3 INTRODUCTION 1
3.1 Pancreatic ductal adenocarcinoma (PDAC) 1
3.1.1 KRAS: the driver mutation and an elusive target 1
3.1.2 Cellular plasticity and metastasis in PDAC 3
3.1.3 Current modalities in PDAC treatment 4
3.2 Cancer Stem Cells (CSCs) 5
3.2.1 CD133: A bona fide CSC marker in PDAC 5
3.2.2 Impact of CSCs on therapy resistance 7
3.2.3 Migrating CSCs: invasive front of metastasis 8
3.3 PDAC and its tumor microenvironment (TME) 10
3.3.1 TME and active players in metastasis 10
3.3.2 TME – CSC interplay 12
3.3.3 The tumor microenvironment mediates resistance to therapy 13
3.4 CXCR4 – CXCL12 signaling 15
3.4.1 CXCR4 and CXCL12 relevance in the metastasis paradigm 16
3.4.2 Importance of CXCR4 – CXCL12 signaling during PDAC development 17
3.4.3 Current therapies and recent advances in targeting CXCR4 18
3.5 Hypothesis and aims 21
4 MATERIAL AND METHODS 23
4.1 Chemicals and reagents 23
4.2 Consumables 24
4.3 Devices 26
4.4 Mice and Cell lines 27
4.5 Co-culture technique 29
4.5.1 Co-culture with PSCs 29
4.5.2 Condition media for PSCs maintenance 30
4.5.3 Primed and Non-primed PSCs – CM preparation 30
4.6 Organoid culture 30
4.7 Treatments 31
4.8 MTT assay 31
4.9 Clonogenic assay 31
4.10 Sphere formation assay 31
4.11 Scratch wound healing assay 32
4.12 Migration assay 32
4.12.1 Migration towards serum 32
4.12.2 Migration towards CXCL12 32
4.13 Molecular Cloning 33
4.13.1 Bacterial culture and plasmid isolation 33
4.13.2 Lentiviral particle production in Lenti-X cells 33
4.13.3 Lentiviral transduction and selection of target cells 34
4.14 RNA isolation and real-time PCR 34
4.14.1 RNA Isolation and cDNA synthesis 34
4.14.2 Real-time quantitative PCR 34
4.15 RNA Seq 36
4.16 Protein extraction and western blot analysis 36
4.16.1 Protein sample preparation 36
4.16.2 Western Blot 37
4.17 Immunofluorescence 38
4.17.1 E-cadherin 38
4.17.2 Nile Red and Phalloidin 38
4.17.3 Phalloidin 39
4.18 Flow Cytometry 39

I
 TABLE OF CONTENTS

4.18.1 Annexin V staining 39

4.18.2 Identifying CTCs in murine blood 39
4.18.3 Surface Protein Staining 39
4.19 Enzyme-linked immunosorbent assay (ELISA) 40
4.20 Antibody-competition assay 40
4.21 Microtiter based stability determination 41
4.22 Fatty acid conjugation with JM#21 41
4.23 Loading JM#21 to silica nanoparticles 41
4.24 Bioinformatics prediction 41
4.25 Software 42
4.26 Statistics 42
5 RESULTS 43
5.1 MEK signaling contributes to maintenance of CSCs phenotype and promotes migration of
pancreatic cancer cells 43
5.1.1 MEK inhibition reduces the growth of murine PDAC cells 43
5.1.2 MEK inhibition targets pancreatic cancer stem cells 44
5.1.3 MEK inhibition decreases migration in a dose-dependent manner 46
5.1.4 MEK Inhibition abrogates TGFβ-Induced EMT 47
5.1.5 MEK Inhibitors Prevent Organoid Formation and Decrease CTCs in vivo 49
5.2 CXCL12 – CXCR4 signaling is imperative for maintenance and metastatic propensity of miCSCs 50
5.2.1 CD133 and CXCR4 are highly expressed in PDAC 50
5.2.2 PSCs secreted factors upregulate CD133 and CXCR4 surface expression 52
5.2.3 Tumor cell – stellate cell crosstalk potentiates CSCs & miCSCs population via CXCL12 54
5.2.4 CXCL12 through CXCR4 maintains CSC and miCSC population 57
5.2.5 CXCL12 through CXCR4 sustains self-renewal capacity and metastatic propensity of miCSCs 60
5.2.6 BMI1 downstream CXCL12 – CXCR4 regulates EMT and stemness 63
5.2.7 CXCL12 – CXCR4 via BMI1 maintains CSCs and miCSCs population 66
5.3 Employing endogenous human peptide to target CXCR4 and eliminate miCSCs 68
5.3.1 JM#21, most potent EPI-X4 derivative to target CXCR4 68
5.3.2 JM#21 inhibits CXCL12 induced EMT and stemness 70
5.3.3 JM#21 sensitizes PDAC cells towards chemotherapy 73
5.3.4 Stabilization of JM#21 in serum conditions 75
5.3.5 Serum stable JM#21 reduces miCSCs in co-cultures with PSCs 77
6 DISCUSSION 81
6.1 MEK signaling in CSCs, CTCs and metastasis 81
6.2 Relevance of MEK inhibition in context of tumor microenvironment 83
6.3 Crucial nature of CXCR4 – CXCL12 pathway in miCSCs maintenance 85
6.4 Diverging roles of BMI1 downstream of CXCR4 – CXCL12 signaling 88
6.5 Endogenous human peptides as a novel agent to target CXCR4 90
6.6 Therapeutic efficacy of JM#21 92
7 SUMMARY 94
8 LIST OF REFERENCES 96
9 ACKNOWLEDGEMENTS 129
10 STATUTORY DECLARATION 131

II
 List of abbreviations

2 List of abbreviations

5-FU 5- Fluorouracil
ADM Acinar to ductal metaplasia
ALDH Aldehyde dehydrogenase
BMI1 B lymphoma Mo-MLV insertion region 1 homolog
CAF Cancer associated fibroblast
CSC Cancer stem cell
CXCL12 C-X-C motif chemokine 12
CXCR4 C-X-C chemokine receptor type 4
DAPI 4´,6-diamidin-2-phenylindol
DCs Dendritic cells
DFS Diesease free survival
DMSO Dimethyl sulfoxide
ECM Extra cellular matrix
EMT Epithelial to mesenchymal transition
EPI-X4 Endogenous Peptide Inhibitor of CXCR4
FBS Fetal bovine serum
FCS Fetal calf serum
FOLFIRINOX Folic acid, 5 Fluorouracil, irinotecan, oxaliplatin
GEMM Genetically engineered mouse model
GEPIA Gene expression profiling interactive analysis
GSH Glutathione
iCAF inflammatory cancer associated fibroblast
IF Immunofluorescence
IL Interleukins
IPMN Intraductal papillary mucinous neoplasm
KRAS Kirsten rat sarcoma virus
MAPK Mitogen-activated protein kinase
MCN Mucinous cystic neoplasm
MDSCs Myeloid-derived suppressor cells

III
 List of abbreviations

MEK Mitogen-activated protein kinase kinase

miCSC Migrating cancer stem cell
MMP Matrix metalloproteinase
MSN Mesoporous silica nanoparticles
myCAF Myofibroblastic cancer associated fibroblast
NF-κB Nuclear factor kappa-light-chain-enhancer of activated B cells
OS Overall survival
OSM Oncostatin M
PanIN Pancreatic intraepithelial neoplasia
PBS Phosphate-buffered Saline
PDAC Pancreatic ductal adenocarcinoma
PDO Patient-derived organoid
PDX Patient-derived xenograft
POU5F1 POU Class 5 Homeobox 1
PSC Pancreatic stellate cell
PTF1-α Pancreas transcription factor 1 subunit alpha
qRT-PCR Quantitative real-time polymerase chain reaction
RPM Revolution per minute
SHH Sonic Hedgehog
shRNA Small hairpin RNA
SOX9 SRY-box transcription factor 9
TAM Tumor associated macrophages
TCGA The Cancer Genome Atlas
TGFβ Transforming growth factor beta
TME Tumor microenvironment
Tregs Regulatory T cells
uPA Urokinase plasminogen activator
α-SMA Alpha smooth muscle actin

IV
 Introduction

3 Introduction
3.1 Pancreatic ductal adenocarcinoma (PDAC)
The pancreas is a spongy, tube-shaped organ that is about 15 cm long and located
in the retroperitoneum between the stomach and the spine, in the upper left
abdomen2. A normal healthy pancreas is comprised of acinar cells that secrete
digestive enzymes, ductal cells that secrete bicarbonate, centro-acinar cells as the
transitional region between acinar and ductal cells, endocrine islets which produce
hormones, and relatively inactive stellate cells3. Mutations in the pancreas can result
in pancreatic cancer, causing the pancreatic cells to uncontrollably grow and divide,
forming tumors4. Pancreatic cancer is characterized as a fatal disease and is one of
the most aggressive and lethal malignancies worldwide5. It accounts for more than
a fifth of all gastrointestinal cancer-related deaths and is projected to be the 2nd most
lethal cancer by 20306. Most cases (90%) are pancreatic ductal adenocarcinomas
(PDAC), which arise from the pancreatic epithelium7. Due to the lack of early
symptoms at the time of diagnosis, PDAC patients mostly present at an advanced
and/or metastasized stage8,9. Nearly 80% patients undergoing surgery with curative
intent ultimately relapse, with 2 out of 3 succumbing to distant recurrence9.
Moreover, 5-year overall survival is 3% for patients with metastasis, an outcome
that has largely remained unchanged in the past 15 years10,11.

3.1.1 KRAS: the driver mutation and an elusive target

In order to comprehend the causative agent(s) for such startling statistics, it is

imperative to first examine the so-called genetic rap sheet of pancreatic ductal
adenocarcinoma. PDAC is broadly assumed to originate due to induction of
environmental stress (e.g., tissue damage and inflammation) from acinar and ductal
cells12 within the exocrine pancreas leading to trans-differentiation of acinar to ductal
cells (acinar to ductal metaplasia, ADM)13-15. Since acinar cells acquire progenitor
cell-like features during ADM, they are more susceptible to pro-oncogenic mutations
or hits16,17. These changes lead ADM to the development of precursor lesions, the
most well-known being pancreatic intraepithelial neoplasias (PanINs)16. Mucinous
cystic, intraductal papillary mucinous, and intraductal tubular papillary neoplasms
have also been described as precursor lesions but are much less frequent18,19.

1
 Introduction

PanINs are part of a multistep tumor progression model in PDAC: Three stages of
PanINs show increasing cellular atypia and typically harbour step-specific
mutations20,21. Changes in the epithelium also result in increasing desmoplasia (Fig
3.1.1A). One of the earliest genetic alterations in this process is the activation of
Kirsten rat sarcoma viral oncogene homolog (KRAS), accompanied by the
inactivation of tumor suppressor genes cyclin-dependent kinase inhibitor p16
(p16)/cyclin-dependent kinase inhibitor 2A (CDKN2A)22,23. Activating mutations in
KRAS are the driver for additional inactivating mutations in tumor suppressor genes
in a rate-limiting step for tumor progression. Therefore, during later PanIN stages,
oncogenic hits in TP53 and inactivating mutations in SMAD4 pave the way for
invasiveness and metastatic burden24 (Fig 3.1.1). Moreover, epigenetic regulation
can repress key tumor suppressor genes and upregulate oncogenes25. In this
regard, tumor progression and metastasis formation have also been linked to
epigenetic (re-)programing26,27.

Given that mutations in one RAS protein isoform, KRAS, are found in nearly 90% of
pancreatic cancers, RAS signaling appears to play a critical role in both pancreatic
cancer initiation and maintenance28,29. Activated RAS can cause cell transformation,
proliferation, and metastasis by activating effector signaling pathways and
transcription factors30. RAS activation can also promote pro-inflammatory signaling
via activation of NF-κB, STAT3, and glycogen synthase kinase-3/nuclear factor of
activated T cell signaling31-33. Knockdown of KRAS in cancer cell lines diminishes
cell proliferation and activates apoptosis34, making KRAS a prime target for
therapeutic interventions in PDAC. Unfortunately, attempts to develop
adequate KRAS inhibitors have been hindered due to various unique features of
oncogenic KRAS, e.g., its high affinity for GTP and redundant pathways for
membrane localization. In this regard, Muzumdar et al. have shown that most PDAC
cell lines can survive in spite of CRISPR/Cas-mediated genetic ablation of KRAS,
pointing towards a potential for resistance to even the very best KRAS inhibitors35.
Hence, despite its central role in PDAC, KRAS still remains “undruggable”, shifting
the therapeutic focus to new approaches targeting crucial downstream effectors.

2
 Introduction

Figure 3.1.1 PanIN stages and tumor progression model for PDAC. (Reprinted with permission
from SPRINGER NATURE. Nature Reviews Cancer. Ref No.24. KRAS, Hedgehog, Wnt and the
twisted developmental biology of pancreatic ductal adenocarcinoma. Morris, J.-P. IV., Wang S.-C.,
Hebrok, M. ©2010).

3.1.2 Cellular plasticity and metastasis in PDAC

A feature associated with epithelial-to-mesenchymal transition (EMT) called cellular

plasticity contributes to tumor cell survival, migration, invasion, and therapy
resistance. In PDAC, phenotypic plasticity of the epithelium is a critical feature in an
oncogene- and tissue-specific context36-38. Many factors can drive cellular plasticity,
in fact, the four so-called hallmark mutations that ultimately give rise to PDAC
namely KRAS (>90%), p16INK4A (>90%), TP53 (~70%) and SMAD4 (55%)39, have
been previously reported to be involved in promoting EMT and thereby metastasis
40-43
. EMT is regulated at the genetic level by transcription factors, which uniformly
repress epithelial genes involved in the expression of adherens and tight junction
components (e.g. E-cadherin). Simultaneously, mesenchymal genes driving cell
plasticity and dissemination to local or distant sites are activated44 (Fig. 3.1.2).
Reichert et al. addressed the role of epithelial plasticity on metastatic organotropism
in PDAC. They observed that in PDAC organ-specific metastasis depends on the
functional form of P120CTN, with liver metastasis requiring intact or mono-allelic
P120CTN, whereas metastatic burden shifts towards pulmonary metastasis with bi-
allelic deletion of the gene45. In addition, Rhim et al. used a lineage-tracing system

3
 Introduction

to determine the kinetics of EMT and dissemination during the natural evolution of
PDAC in vivo and demonstrated that EMT occurs as early as during premalignant
lesions. They further discovered that the spread of pancreatic epithelial cells in mice
in fact precedes tumor formation46.

Figure 3.1.2 EMT activating signaling pathways. (Reprinted with permission from SPRINGER
NATURE. Nature Reviews Molecular Cell Biology. Ref No.47. New insights into the mechanisms of
epithelial–mesenchymal transition and implications for cancer. Dongre, A., Weinberg, R.-A. ©2018).

3.1.3 Current modalities in PDAC treatment

Even with improved surgical techniques, radiologically resectable PDAC patients

presumably already have micrometastases, which explain the frequent relapse in
patients that have been treated with curative intent. For patients with non-resectable
or metastatic disease, a systemic chemotherapy is commonly used as first-line
treatment. This includes nucleoside analogues, e.g., gemcitabine and capecitabine,
or the pyrimidine analogue 5-fluorouracil (5-FU) as monotherapy or in combination
with other chemotherapeutic agents or treatment modalities, such as radiotherapy48-
50
. Compared to gemcitabine alone, a poly-chemotherapeutic regimen called
FOLFIRINOX, comprised of folinic acid, 5-FU, irinotecan, and oxaliplatin, has been
described to almost double median survival in the metastasized stage51, and the
combination of gemcitabine and a nanoparticle albumin-bound paclitaxel (nab-
paclitaxel) has also been reported to significantly boost overall survival52. However,
while increasing the median survival time to approximately 11 months, these

4
 Introduction

regimen are associated with higher toxicity16,53, and can therefore be used only in
patients with high performance scores.

Although checkpoint inhibition has been approved for treatment of the small subset
of PDAC tumors with high microsatellite instability (1-2% of all cases54,55), the
therapeutic efficacy of immunotherapeutic approaches in PDAC has been
unfortunately limited56,57. Interestingly, approximately 7% of patients have mutations
involving DNA repair genes (e.g., BRCA2 and PALB2), which might allow the use
of PARP inhibitors in this subset of patients58. While new drugs are continuously
developed and improved, some of the major hurdles in PDAC therapy are the highly
heterogeneous tumor system consisting of cells with varying degrees of
differentiation (e.g., poorly differentiated cancer stem cells) and physical
inaccessibility of the tumors due to highly desmoplastic stroma.

3.2 Cancer Stem Cells (CSCs)

PDAC typically contains cells with a varying capacity of tumorigenicity59. Cancer
stem cells are a population of tumor cells that have specific defining abilities: they
can self-renew, differentiate into all types of cancer cells and generate tumors in a
secondary recipient. This definition stresses the importance of this population of
undifferentiated cells found within a highly heterogenous tumor, and it has been
widely described that although these cells are not true stem cells, they possess
stem-like characteristics60,61. CSCs are capable of not only mediating tumorigenicity
but also intrinsic cellular plasticity and the capacity to hierarchically organize the
tumor60,62, and have the ability to re-generate themselves regularly using
asymmetric and symmetric division. In both cases, the self-renewal capacity
remains unscathed, which secures the pool of CSCs over time63. Additionally, CSCs
show resistance to chemotherapeutics, multipotency and the ability to
metastasize63,64.

3.2.1 CD133: A bona fide CSC marker in PDAC

Li et al. and Hermann et al. independently identified CD44+CD24+EpCAM+ cells

and CD133+ cells, respectively as CSCs in PDAC60,62. CD133+ cells show some
overlap with CD44+CD24+EpCAM+ cells. Since then, pancreatic CSCs have been
identified using a variety of (bio)markers including but not limited to CD90, LGR5,

5
 Introduction

CXCR4, aldehyde dehydrogenase 1 (ALDH1), autofluorescence and hepatocyte

growth factor receptor C-MET60,62,65-68. Recently, Alcalá et al. have also described
anthrax toxin receptor 1 (ANTXR1) as a putative CSC marker in PDAC69. While
none of these markers can identify a pure CSC population, these markers are
helpful to further understand pancreatic CSC biology regarding unlimited self-
renewal capacity, exclusive tumorigenicity or inherent chemoresistance when
compared to non-CSCs (Fig. 3.2.1).

CD133 is a transmembrane protein. It consists of an extracellular domain with a

ganglioside-binding site and a cytoplasmic domain which is found in lipid rafts and
has the capability of tyrosine phosphorylation70-73. The activation of CD133
transcriptional regulation could be performed by a downstream signaling pathway
initiated by one of the extracellular signals–regulated kinases (ERKs), mainly by
ERK1 or ERK274. On the other hand, during carcinogenesis of PDAC, heat shock
proteins such as HIF-1α and HIF-2α show increased expression due to oxygen
deficiency in highly fibrotic and nutrient-deficient PDAC, and can also target the
CD133 promoter, increasing CD133 transcription74. Furthermore, Wnt/β-catenin
signaling and Janus kinases (JAK) or Signal Transducer and Activator of
Transcription protein 3 (STAT3) promote the expression and self-renewal of
CD133+ CSCs in pancreatic cancer75-77. CD133 and its substrates from the SRC
family tyrosine kinases (e.g., proto-oncogene tyrosine-protein kinases SRC and
FYN) trigger downstream regulatory signals for stemness (e.g., enhanced
telomerase reverse transcriptase expression, increased AKT phosphorylation and
ligand-independent EGFR activation) and EMT70,78-80. In normal pancreas CD133
expression is significantly lower (under 0.01% of cancer cells) than in pancreatic
cancer78,81, and CD133 levels are an independent prognostic marker for disease-
free survival in pancreatic cancer81. Thus, CD133 undeniably shows prognostic
value and is one of the most commonly used CSC markers in various solid
malignancies, e.g., PDAC, colorectal cancer and glioblastoma.

6
 Introduction

Figure 3.2.1 Pancreatic cancer stem cell markers for stemness, metastatic potential, EMT,
immune evasion and chemoresistance. (Reprinted from Valle, S., Martin-Hijano, L., Alcalá, S.,
Alonso-Nocelo, M., Sainz, B.-Jr. The ever-evolving concept of the cancer stem cell in pancreatic
cancer. doi: 10.3390/cancers10020033. Ref No.63. Licensed under a Creative Commons Attribution
License (CC BY 4.0) licence https://creativecommons.org/licenses/by/4.0/).

3.2.2 Impact of CSCs on therapy resistance

Standard chemotherapy has very limited or nearly no significant effect on CSCs,

only further enriching the CSC population due to the elimination of more
differentiated cells62. This observation holds true for in vitro cell cultures as well as
for fresh and in vivo expanded patient-derived pancreatic cancer cells82. Jagust et
al. have shown that CD133+ CSCs depend on glutathione (GSH) metabolism,
accumulating GSH in response to gemcitabine, aiding CSCs in their self-renewal
and chemoresistance83. Although gemcitabine treatment is effective in decreasing
tumor growth it shows virtually no effect on CSCs, resulting in rapid relapse and
increased tumor aggressiveness after gemcitabine withdrawal as a result of an
expanding CSC pool. CSCs are protected from damage caused by external agents
due to a number of favorable mechanism including anti-apoptosis, increased DNA-
repair and overexpression of efflux transporters (e.g., ABCG2) that eliminate the

7
 Introduction

drugs from these cells61,63,84,85. Recent studies also show that knockdown of NRF2
(a gene responsible for glutathione metabolism and antioxidant metabolism), re-
sensitizes PDAC cells towards 5-FU by downregulating CD133 and ABCG2
expression86.

ALDH is a member of a family of intracellular cytosolic enzymes and was

established as another CSC marker due to its high presence and activity in CSCs.
Interestingly, it mediates cyclophosphamide- as well as gemcitabine-resistance in
PDAC and thus represents a direct link between stemness and chemoresistance87.
Corroborating this link, reduction of the autophagy factor LC3 has been shown to
decrease ALDH1+ CSCs and thus sensitize PDAC cells towards gemcitabine. LC3
expression also positively correlates with other CSC markers like CD133 and CD44
in PDAC88. CSCs strongly rely on mitochondrial metabolism to maintain stemness
and chemoresistance89. A recent study describes targeting the mitochondrial fission
gene DNM1L (DRP1) induced apoptosis in CD133+ CSCs, resulting in a decrease
in self-renewal capacity and invasiveness, as well as in a sensitization of PDAC
cells to gemcitabine90. Given the functional characteristics of CSCs and their
eminent role in tumor biology, it is crucial to identify new therapeutic strategies to
selectively eliminate CSCs, or to develop combination therapies that target the vast
bulk of the tumor cells and CSCs as the root of the tumor to finally improve therapy
of pancreatic cancer.

3.2.3 Migrating CSCs: invasive front of metastasis

Mani et al. described the role of EMT induction in immortalized mammary epithelial
cells and subsequently increased stem cell marker expression, mammosphere
formation and tumorigenicity in vivo, showing a direct association between EMT and
the development of CSCs characteristics91. More recent studies have discovered
the underlying mechanisms by which EMT programs control stemness:
Pastushenko et al. identified transition states that occur in the tumor during the
process of EMT. In their model, hybrid EMT populations displayed a 5-fold increase
in tumor propagation as compared to epithelial cancer cells and demonstrated that
the intermediate EMT subpopulations were the most plastic92. Furthermore, the
EMT International Association (TEMTIA) uses the term ‘epithelial–mesenchymal
plasticity’ (EMP) to describe the ability of cells to adopt mixed

8
 Introduction

epithelial/mesenchymal features which interconvert between intermediate

phenotypic states along the epithelial–mesenchymal spectrum93. Interestingly,
these epithelial/mesenchymal hybrid states seem to be enriched in CTCs, i.e., cells
released from the primary tumor and circulating in the blood stream93-95. These
CTCs have EMT- and tumor-initiating properties (typical CSC characteristics), and
thus can successfully metastasize to distant secondary sites63. Therefore, studies
suggest that CTCs contain a highly metastatic CSC subpopulation, possibly in an
epithelial/mesenchymal hybrid state96-98 (Fig. 3.2.3).

Expanding this concept, Hermann et al. indeed identified not only a CD133+ CSC
population with tumor initiating properties in PDAC62 but their study also identified a
CD133+CXCR4+ subpopulation of CSCs that was highly and exclusively
metastatic. Due to this exclusive metastatic activity, they were defined as “migrating
CSC (miCSC)”. The expression of CXCR4 in miCSCs has significant therapeutic
potential: Hermann et al. used AMD3100 or CXCR4 neutralizing antibodies to inhibit
CXCR4, and thus were able to abrogate metastasis. A clinical correlation between
migrating CSCs and advanced disease stages was established by the significantly
higher presence of (CD133+ CXCR4+) population in patients with lymph node
metastases (pN1+). After this landmark study by Hermann et al., migrating CSCs
have been described in a number of other cancer types99-101. This provides
convincing evidence for the crucial role of CXCR4 signaling in metastasis.

Figure 3.2.3 The metastatic cascade. From Chaffer CL, Weinberg RA. A perspective on cancer
cell metastasis. Science. 2011 Mar 25;331(6024):1559-64. doi: 10.1126/science.1203543. Ref.
No.102. Reprinted with permission from AAAS.

9
 Introduction

3.3 PDAC and its tumor microenvironment (TME)

As mentioned previously, one of the major hurdles in PDAC therapeutics is the

inaccessibility of tumors to the administered drugs due to the highly desmoplastic
stroma typically observed in PDAC. The overexpression of ECM proteins and
extensive transformation of fibroblast-type cells to a myofibroblastic phenotype are
the clinical manifestations of desmoplasia103. The tumor microenvironment in PDAC
consists of various cells such as cancer-associated fibroblasts, tumor-associated
macrophages, endothelial cells, immune cells and extra-cellular matrix proteins. The
TME plays an important role in mediating and promoting tumor growth, therapy
resistance, and metastasis. Besides, the intra-tumoral architecture of PDAC is
modulated by the stromal microenvironment104. Therefore, there has been an
increased interest in the potential targeting of the tumor stroma by depleting
essential cellular elements, enhancing tumor perfusion, and inducing local
immunomodulation105. However, the translation of these strategies into treatment
remains controversial.

3.3.1 TME and active players in metastasis

Fundamentals for determining the capacity of cancer cells to escape from primary
tissues and subsequently seed and colonize distant tissues are (i) a permissive
tumor microenvironment and (ii) a receptive distant organ niche106. PDAC tumor
microenvironment components (Fig. 3.3.1) and factors secreted by them participate
in modulating metastasis. Tumor-associated macrophages (TAMs) are one of the
major immune cell types and cancer-associated fibroblasts (CAFs) are the primary
source of extracellular matrix deposition in the TME107. Studies have suggested a
crucial role of these components in cancer progression, metastasis, and
chemoresistance108. Furthermore, there is substantial evidence for tumor– stroma
crosstalk promoting metastasis109-111.

CAFs typically arise from tissue-resident fibroblasts (e.g., PSCs), but also from bone
marrow-derived cells and adipose tissue-derived mesenchymal stem cells (MSCs).
Interestingly, the heterogeneity in CAFs occurs based on their expression profile
and localization within the tumor microenvironment (e.g., myofibrotic CAFs,
[myCAFs]; inflammatory CAFs [iCAFs] and antigen presenting CAFs

10
 Introduction

[apCAFs])112,113). Pancreatic stellate cells (PSCs) are myofibroblast-like cells that

are usually located in the exocrine regions of the pancreas114. During PDAC
development, a variety of signaling factors (such as IL-1, IL-6, hypoxia inducible
factor 1α (HIF1α), and TGFβ) transforms quiescent PSCs into an activated
myofibroblast-like state or into CAFs115,116. Activated PSCs are characterized by a
loss of cytoplasmic lipid droplets, upregulation of ECM proteins, and increased
proliferative capacity114,117,118. Furthermore, activated PSCs play a vital role in
promoting metastasis (also tumor progression and desmoplasia) within the
microenvironment by secreting molecules such as TGFβ, IL-6, CXCL12, hepatocyte
growth factor (HGF) and galectin-1119-121. The typical extensive desmoplasia in
PDAC itself can generate a hypoxic microenvironment. Hypoxia, which is caused
by oxygen deprivation due to insufficient vasculature, is important for invasion and
metastasis in PDAC122. Other cellular components involved in inducing metastasis
within the TME are regulatory T cells (Tregs), myeloid-derived suppressor cells
(MDSCs), immature and tolerogenic dendritic cells (DCs) and M2 like TAMs123-126.

Figure 3.3.1 Components of PDAC microenvironment. (Reprinted from Truong, L.-H.; Pauklin, S.
Pancreatic Cancer Microenvironment and Cellular Composition: Current Understandings and

11
 Introduction

Therapeutic Approaches. doi: 10.3390/cancers13195028. Ref No.127. Licensed under a Creative

Commons Attribution License (CC BY 4.0) licence https://creativecommons.org/licenses/by/4.0/).

3.3.2 TME – CSC interplay

In order to maintain a supportive niche, CSCs proactively remodel their

microenvironment. CSCs communicate with other cellular components of the TME,
and understanding the dynamic crosstalk between CSCs and the TME is extremely
important128,129. In this regard, Begum et al., described the role of extracellular
matrix and focal adhesion kinase signaling in regulating CSC characteristics in
PDAC130. Besides, the hypoxic niche in the tumor microenvironment is one of the
major driving forces for the tumor-initiating population within: HIF1A regulates
stemness by upregulating self-renewal genes, especially in the CSC population131-
133
. Lonardo et al. have also shown that pancreatic stellate cells provide a niche for
the cancer stem cells134. In addition, Sainz et al. have recently demonstrated a
decisive (and therapeutically accessible) crosstalk between CSCs and the niche,
mediated by LL-37135.

Pancreatic CSCs utilize the dysregulation of Wnt/β-catenin, hedgehog, notch, NF-

κB, PI3K/Akt and PTEN signaling to maintain their stemness maintenance and
metastatic potential120,136-140. Xenografted tumors from pancreatic CSCs with low
Alk4/7 levels can enhance sensitivity to gemcitabine and result in prolonged survival
134
. Furthermore, stromal cells can maintain the pancreatic CSC population via
paracrine signaling pathways: HGF secreted by PSCs promotes self-renewal of c-
MetHigh pancreatic CSCs. In a similar context, pancreatic CSCs can secrete IFNβ to
stimulate TAMs to produce the IFN-stimulated factor ISG15 to enhance the CSC
phenotype in vitro and in vivo141. TAMs also produce leucine leucine (LL)-37 to
increase pluripotency-associated gene expression, self-renewal, invasion and
tumorigenicity of pancreatic CSCs via formyl peptide receptor 2 (FPR2)- and P2X
purinoceptor 7 receptor (P2X7R)-dependent mechanisms135. Szczerba et al.
recently demonstrated the relevance of supporting cells throughout metastasis
where surprisingly, clusters of CTCs and neutrophil leukocytes displayed a far
stronger metastatic potential than CTCs alone142.

12
 Introduction

3.3.3 The tumor microenvironment mediates resistance to therapy

Clinical studies have demonstrated that the expression of certain TME components
can correlate with a poor patient prognosis and can facilitate desmoplasia and
immunosuppression or promote metastasis in PDAC120. Non-specific targeting of
the ECM alone is not effective in pancreatic cancer as MMP inhibitors such as
marimastat and tanomastat or enzymatic degradation of hyaluronan by human
recombinant PH20 hyaluronidase (PEGPH20) failed to show significant clinical
benefit in patients with advanced stage pancreatic cancer143-146. Another approach
to target the TME involves cancer-associated stromal cells or cancer-associated
fibroblasts (CAFs), which are one of the major producers of ECM proteins. CAFs
are generally spindle-shaped cells and express one or more activated fibroblast
markers (e.g., fibroblast activation protein [FAP] and α-smooth muscle actin [α-
SMA])147. Specific targeting of fibroblasts was first assessed in colorectal cancer
using the FAP targeting inhibitor sibrotuzumab, which failed to meet the end point
in a phase II trial148. Other small molecule inhibitors of FAP have shown limited
efficacy against PDAC149.

In mouse models of PanIN or PDAC, Erkan et al. showed that genetic deletion of α-
SMA expressing fibroblasts resulted in more aggressive disease150 and
demonstrated that high CAF activity with low collagen deposition is associated with
poor prognosis. However, it also appears that the desmoplastic reaction might
restrain tumor cells from expanding outside the pancreas151. Therefore, instead of
eliminating the stromal fibroblasts from the TME, a more sophisticated approach
could be to alter CAF phenotype. Stromal reprogramming, as opposed to ablation,
is gaining acceptance in the field of stroma-targeting approaches for the treatment
of PDAC152-154. Chronopoulos et al. showed that ATRA is an active metabolite of
vitamin A that restores quiescence in PSCs. In 3D organotypic models, the authors
depicted ATRA treatment inhibits local cancer cell invasion153. Some of the
attractive targets for TME-related therapy in PDAC are depicted in (Fig. 3.3.3).

With regard to the immune compartment of the TME, studies show that CXCL12
secreted from PSCs reduces the migration of CD8+ T cells into the peritumoral
stroma, and galectin-1 induces T-cell apoptosis and Th2 cytokine secretion, thereby

13
 Introduction

promoting immunosuppression155,156. On the other hand, Tregs inhibit effector T cell

functions by secreting suppressive cytokines and molecules, such as IL-10 and
TGFβ, which is consistent with clinical findings157. Although these studies depict a
clear link between immunological processes and PDAC carcinogenesis, PDAC is
still known as an "immunologically cold” tumor, since only a very small subset of
pancreatic cancers is immunologically active158,159. However, there is a possibility to
exploit the altered immune TME that occurs as a result of stromal reprogramming
by tumour cells: A study in KPC mice showed the immune control of tumour growth
and an effective response to immune-checkpoint inhibitors (ICIs) when applied
together with depletion of FAP+ CAFs160. Such studies in addition to previously
described CXCL12-secreting PSCs promoting metastasis elucidate the importance
of CXCL12 – CXCR4 signaling as a means of stroma – tumor/stroma – immune cell
crosstalk, presenting yet another target for therapy.

Figure 3.3.3 Potential cancer stem cell and ECM targeting approaches in PDAC therapy.
(Reprinted from Truong, L.-H.; Pauklin, S. Pancreatic Cancer Microenvironment and Cellular
Composition: Current Understandings and Therapeutic Approaches. doi:
127
10.3390/cancers13195028. Ref No. . Licensed under a Creative Commons Attribution License (CC
BY 4.0) licence https://creativecommons.org/licenses/by/4.0/).

14
 Introduction

3.4 CXCR4 – CXCL12 signaling

CXCR4 (previously known as LESTR or fusin) was originally cloned as an orphan

chemokine receptor161-163. LESTR/fusin was identified as an essential co-factor for
T-tropic HIV-1 and HIV-2 env-mediated fusion and entry into CD4+ cells 164-166.
It was later re-classified as chemokine receptor CXCR4 when CXCL12 (or SDF-
1) was recognized as its biological ligand167,168. CXCL12 is one of the most
important CXC chemokines and is vital to normal B cell growth and cardiac tissue
initiation169 as well as leucocyte homing. Unlike other chemokines, CXCL12 is
expressed constitutively in a broad range of tissues and is a highly efficacious
lymphocyte and monocyte chemoattractant170,171.

CXCR4 is a G-protein coupled receptor and is expressed on the cell surface of

most leukocyte populations including the majority of T lymphocyte subsets, all
B cells and monocytes, and only weak expression on natural killer cells172. In the
absence of CXCL12, it is coupled to GDP-bound Giα, which forms an inactive
trimeric G protein with Gβγ. Upon stimulation with CXCL12, CXCR4 activates
signaling cascades downstream activated G-protein. CXCR4 undergoes a
conformational shift that favors the exchange of GDP for GTP, releasing GTP-bound
Giα from Gβγ. Free GTP-Giα and Gβγ then activate different downstream signaling
cascades (Fig. 3.4). GTP-Giα inhibits adenylyl cyclase and activates mitogen-
activated protein kinase (MAPK) signaling, amplifying cell proliferation and
migration. Gβγ induces phospholipase C (PLC)/protein kinase C (PKC)-Ca2+
signaling to augment chemotaxis and activates the phosphatidylinositol-3-kinase
(PI3K) pathway to promote cell survival173-175. CXCR4 can also signal in a G protein-
independent manner, although the majority of the stimulated CXCR4 signaling is G
protein-dependent. CXCL12 stimulation can activate the JAK/STAT pathway and
arrestin-2 and -3 can also enhance CXCR4-activated MAPK signaling, both in a G
protein-independent fashion. In addition, arrestin-3 has been described to activate
p38 MAP kinase directly to augment cell migration173,174,176.

15
 Introduction

Figure 3.4 CXCR4-related signaling. (Reprinted from Truong, Xu C, Zhao H, Chen H, Yao Q.
CXCR4 in breast cancer: oncogenic role and therapeutic targeting. doi: 10.2147/DDDT.S84932. Ref
No.177. Licensed under a Creative Commons Attribution 3.0 Unported License (CC BY-NC 3.0)
licence https://creativecommons.org/licenses/by-nc/3.0/).

3.4.1 CXCR4 and CXCL12 relevance in the metastasis paradigm

The CXCL12/CXCR4 axis promotes chemotaxis and invasion in normal cells.

Furthermore, this axis plays a pivotal role in promoting tumor cell invasion and the
development of metastasis in cellular and animal models, as well as patient-based
studies178,179. Another direct consequence of CXCL12/CXCR4 activation is
increased SHH signaling, which is associated with EMT, loss of cell adhesion, and
increased cancer progression and metastasis180-182. Wnt also promotes EMT
through increased expression of its targets such as vimentin and slug183. CXCL12
itself induces cytoskeletal changes in pancreatic cancer cells that are coherent with
those of metastatic cells and increases cell migration in vitro184,185. CXCL12 also
increases the expression of MMP-2, MMP-9, and uPA, which increase cell invasion
by matrix degradation, removing a physical barrier for cell mobility181,186,187.
CXCL12/CXCR4 signaling promotes EMT via MEK/ERK, PI3K/AKT or Wnt/β-
catenin signaling in chondrosarcoma, glioblastoma, colorectal cancer and
hepatocellular carcinoma188-191.

16
 Introduction

CXC12 is present in the basal lamina of endothelial venules and helps CXCR4+
cancer cells to migrate to blood vessels from the primary tumor192. Furthermore,
liver, bone marrow, lungs and lymph nodes exhibit peak expression levels of
CXCL12 and are most common target organs for metastasis193. Kim et al. showed
the liver to be a selective target of CXCR4+ colorectal cancer cells due to the
abundance of CXCL12194. Similarly, Gelmini et al. reported that treatment with a
CXCR4 neutralizing antibody dramatically reduced the number and size of
metastases in vivo in endometrial cancer195. This also indicates the relevance of
chemokine modulation for organ-specific migration of cancer cells, ultimately
leading to metastasis196. Moreover, tumors such as PDAC are often hypoxic, and
HIF-1α induces CXCL12 upregulation, assisting in the recruitment of CXCR4+
CSCs to peripheral vessels as a pool for metastasis197,198. All these findings indicate
that after being activated by its ligand CXCL12, CXCR4 initiates signaling in tumor
cells that promotes metastasis in a variety of cancer types.

3.4.2 Importance of CXCR4 – CXCL12 signaling during PDAC development

Deviations in the CXCL12 - CXCR4 signaling causes pathogenesis of different

diseases including neurological, cardiovascular, angiogenic, immunologic (WHIMS
syndrome), and myeloid precursor deficiencies199. In PDAC, CXCL12/CXCR4 axis
behaves in a similar fashion to that during embryologic development of normal
pancreatic ductal epithelium200. During the progression from PanIN to PDAC,
CXCR4 expression increases in both murine and human models201,202, and
increasing CXCL12 levels also correlate directly with increasing PanIN grade202.
Therefore, the CXCL12/CXCR4 axis might act as a critical component in the
malignant transformation of PanIN lesions203. Furthermore, numerous studies have
shown that high expression of CXCL12 and activation of CXCR4 in tumors utilizes
local autocrine and paracrine mechanisms to enhance growth and to restrict
immune surveillance within the tumor microenvironment204-207.

In terms of patient outcome, blood vessel density and perineural invasion are
prognostic markers for pancreatic cancer and are strongly associated with CXCR4
expression185,208. CXCR4 levels also correlate with lymph node and distant organ

17
 Introduction

metastasis, both leading to poor prognosis209-211. Moreover, increasing expression

of CXCR4 conform with an advanced stage and poor survival in pancreatic cancer
212-215
. Furthermore, Demir et al. demonstrated that CXCL12/CXCR4 signaling might
be involved in delaying symptomatic presentation of pancreatic cancer, leading to
delayed diagnosis216. In summary, the CXCL12/CXC4 axis plays an outstanding
role throughout PDAC development and malignant progression.

3.4.3 Current therapies and recent advances in targeting CXCR4

CXCR4 antagonists reduce macrophage infiltration, abrogate metastatic spread and

decrease tumor growth in PDAC animal models202,217-219. However, chemotherapies
such as gemcitabine increase CXCR4 expression and inadvertently might promote
aggressive and metastatic potential of PDAC tumors220-222. Therefore, combination
treatments with CXCR4 antagonists and chemo- or radiotherapy have the potential
to improve patient185. AMD3100 (Plerixafor) is the the first CXCR4 antagonist that
received FDA approval in 2008, initially for mobilization of hematopoietic stem cells
in bone marrow transplantation procedures223. In vitro AMD3100 can decrease
proliferation in pancreatic cancer cells and abrogate CXCL12-mediated
migration224,225. Singh et al. also demonstrated that AMD3100 sensitizes PDAC cells
towards gemcitabine treatment, outlining the benefits of co-administrating CXCR4
antagonists with conventional therapeutics220. However, AMD3100 is known to
cause cachexia, will influence leucocyte homing and is not CXCR4 specific and
therefore has off-target effects.

Apart from AMD3100, researchers have sought other methods to target CXCR4.
Some of them include utilizing neutralizing antibodies for immunotherapy for
colorectal cancer226, copolymer drug incorporating BKT140, for inhibiting prostate
cancer227, LFC131 peptide inhibitor conjugated PGLA nanoparticles for drug
delivery in lung cancer228 and HPMA block copolymer of hydroxychloroquine for
decreasing lung metastasis in breast cancer229. Compared to AMD3100, the
CXCR4 antagonist BL-8040 displayed a greater effect on the retention–mobilization
balance of bone marrow stem cells by providing a higher affinity and longer receptor
occupancy in mouse studies230,231. In preclinical models of pancreatic cancer,
inhibition of the CXCL12/CXCR4 axis enhanced T cell access to the TME and
increased tumor sensitivity to anti-PD-1 ligand-1 (PD-L1) therapy160,232. In this

18
 Introduction

regard, the COMBAT trial (NCT02826486) utilized BL-8040 in combination with

pembrolizumab and chemotherapy as second-line therapy for metastatic pancreatic
cancer, and patients receiving these drugs showed a median overall survival of 7.5
months. Moreover, BL-8040 reduced myeloid-derived suppressor cells (MDSCs),
promoted CD8+ effector T cell tumor infiltration, and depleted circulating regulatory
T cells. Taken together, these studies suggest that further randomized trials will be
necessary to demonstrate the benefit of CXCR4 and PD-L1 combination therapy for
advanced PDAC233. However, despite studies confirming the role of CXCR4 as a
valuable therapeutic target for the treatment of cancer, no new inhibitors have been
approved by the FDA since 2008. More importantly, no further studies have
investigated the therapeutic targeting of CXCR4 in pancreatic cancer besides the
few combination therapies mentioned above.

Only recently, Zirafi et al. screened an HF-derived (hemofiltrate) peptide library for
inhibitors of CXCR4-tropic (X4) HIV-1 and identified an endogenous antagonist of
CXCR4 generated by limited proteolysis of serum albumin (Fig. 3.4.3A), the most
abundant protein in human serum234. They identified EPI-X4 as an evolutionarily
conserved endogenous antagonist of CXCR4 that may play important roles in
physiological processes and diseases. Since its discovery and characterization in
2015 several derivatives of EPI-X4 have been generated by chemical modification,
most importantly WSCO2 and JM#21 (Fig. 3.4.3B)235. WSCO2 showed decreased
migratory capacity in vitro236 and while no immediate toxicity was observed, further
analysis for long-term toxicity of EPI-X4 and its derivatives is imperative. Recent
studies with Waldenström’s Macroglobulinemia (WM, B-cell lymphoma) confirmed
EPI-X4, and its optimized derivatives bind to CXCR4 on WM cells, and that they are
able to diminish growth of lymphoma cells in vivo237. Therefore, optimized
derivatives of EPI-X4 may represent a promising approach in targeting
CXCL12/CXCR4 signaling.

19
 Introduction

Figure 3.4.3 (A) EPI-X4 is generated from serum albumin under acidic conditions by cathepsin
D or E. (Reprinted with permission from John Wiley and Sons. JOURNAL OF LEUKOCYTE
BIOLOGY. Ref No.238. Proteolytic processing of human serum albumin generates EPI-X4, an
endogenous antagonist of CXCR4. Zirafi, O., Hermann, P.-C, Münch, J. ©2016). (B) Intramolecular
H-bonds in EPI-X4, WSCO2 and JM#21. (Color code: C = pink/blue, H = white, N = blue, and O =
red. H-bonds are shown as dotted lines in orange. (EPI-X4 from the previously published thesis
Harms M: Endogenous CXCR4 Antagonists: Role in HIV-1 Transmission and Therapy of CXCR4-
linked diseases239, PhD Dissertation, Institute of Molecular Virology, University of Ulm, 2021.
WSCO2 and JM#21 from Harms M, Habib MMW, Nemska S, Nicolò A, Gilg A, Preising N, Sokkar
P, Carmignani S, Raasholm M, Weidinger G, Kizilsavas G, Wagner M, Ständker L, Abadi AH, Jumaa
H, Kirchhoff F, Frossard N, Sanchez-Garcia E, Münch J. An optimized derivative of an endogenous
CXCR4 antagonist prevents atopic dermatitis and airway inflammation. doi:
235
10.1016/j.apsb.2020.12.005. Ref No. . Licensed under a Creative Commons Attribution License
(CC BY-NC-ND 4.0) licence https://creativecommons.org/licenses/by-nc-nd/4.0/).

20
 Introduction

3.5 Hypothesis and aims

The overall hypothesis of the present study is to discern molecular pathways that
are accountable for maintaining migrating cancer stem cells and their characteristic
features, and to exploit these targets in order to establish novel therapetic strategies.

Metastasis is still the major cause of mortality in patients with pancreatic ductal
adenocarcinoma (PDAC), yet this process is one of the most enigmatic aspects of
the diesease. KRAS, the key oncogenic driver mutation in PDAC, has been
implicated in promoting metastasis. During the different stages of metastasis,
circulating tumor cells (CTCs) act as a transition stage. We hypothesize that
RAS/MEK/ERK signaling may regulate CTC initiation. To address this issue, we
evaluated
(i) the role of MEK signaling downstream of RAS in aiding growth, self-renewal
and migratory propensity in PDAC cells,
(ii) the capacity of MEK signaling to inflect active EMT program, initiated by
TGFβ and
(iii) the effect of MEK inhibition on CTC numbers.

Interestingly, MAPK signaling relays, amplifies and integrates other signaling

pathways within the tumor microenvironment to promote stem-like cells and
metatstasis. Moreover, our own group showed previously that CD133+CXCR4+
migrating cancer stem cells (miCSCs), a subset of CD133+ cancer stem cells
(CSCs) are detectable within the invasive front of PDAC. Since CXCL12 is the ligand
for the chemokine receptor CXCR4, we hypothesized that CXCL12/CXCR4
signaling is imperative for the maintenance and metastatic capabilities of miCSCs.
In order to achieve this, we investigated
(i) the relevance of tumor – stroma crosstalk (w.r.t. pancreatic stellate cells) for
the maintenance of CSCs and miCSCs,
(ii) the underlying mechanism(s) by which CXCL12 sustains self-renewal and
metastatic capacity in miCSCs and
(iii) putative molecular links downstream of CXCL12/CXCR4 signaling between
EMT and stemness.

21
 Introduction

It is essential to develop and investigate novel approaches to target metastasis,

especially with the crucial role of CXCR4 in cell migration. Recently, a natural
CXCR4 antagonist, the endogenous human peptide EPI-X4 was characterized that
effectively blocks CXCL12-mediated receptor internalization. Thus, our final
objective was to employ EPI-X4 and derivatives thereof to target CXCR4 and
eliminate miCSCs. In order to address this, we evaluated
(i) the potency of different endogenous peptide to reduce CXCL12-dependent
CXCR4-mediated migration,
(ii) the capacity to restrain EMT, self-renewal and stem-cell like features by the
most potent peptide,
(iii) the therapeutic efficacy of EPI-X4 derivates in combination with established
chemotherapy in abrogating miCSCs.

22
 Material and Methods

4 Material and Methods

Material and methods are reprinted with permission of Stem Cell International
published by Hindawi Limited licensed under a Creative Commons Attribution
License (CC BY 4.0) ©2019 (Walter et al. 2019) with additions and adaptations.

4.1 Chemicals and reagents

Table1: Chemicals and reagents with respective catalogue numbers.

Chemical/Reagent Company Catalogue#
Acetic acid (10%) Merck 1.00063.2511
Acrylamide/Bis-acrylamide (30%) Sigma-Aldrich A3699
ALDEFLUORTM DEAB Reagent Stem Cell 01705
Technologies
Ampicillin (50 mg/ml stock solution) Roth K029
AnnexinV (APC) BD 550474
AnnexinV staining buffer BD 556454
APS Bio-Rad 161-0700
BSA (lyophilized powder) Sigma-Aldrich A9418
CASYton OLS OMNI Life OLS 5651808
Science
DAPI (4´,6-diamidin-2-phenylindol, Thermo Fisher D3571
10.9 mM or 0.838 mM stock solution) Scientific
DAPI (Prolong™ Gold Antifade) Invitrogen P36931
ddH2O (double-distilled, nuclease Qiagen 129117
free)
dH2O (demineralized, MiliQ®) Pharmacy, Hospital Ulm
Disodiumhydrogenphosphate Merck 1.06586.0500
(Na2HPO4)
EDTA disodium salt (Na2EDTA) Sigma-Aldrich 03710
Gamunex Grifols G130158
Giemsa solution Merck 1.09204.0100
Glycine Sigma-Aldrich 33226
Hydrochloric acid (HCl) Merck 30721
Immersion oil Roth H302-H410

23
 Material and Methods

Isopropanol (70%) Pharmacy, Hospital Ulm

Laemmli SDS-Sample buffer (6x) Boston BioProducts 111NR
LB broth (capsules) MP-Biomedicals 3002-011
LB broth with agar (capsules) Sigma-Aldrich L7025
Lenti-X Concentrator Clontech 631231
Methanol (99,8%) Sigma-Aldrich 32213
PageRulerTM Prestained Thermo Fisher 26617
Proteinladder Scientific
Paraformaldehyde (powder, 95%) Sigma-Aldrich 158127
Phalloidin-Atto 565 Sigma-Aldrich 94072
PhosStopTM Merck 4906845001
Polybrene (10 mg/ml stock) Sigma-Aldrich H9268
Polyethyleneimine (PEI, linear) Polysciences, Inc. 23966
Potassium chloride (KCl) J.T.Baker 7447-40-7
Potassium dihydrogen phosphate Merck 1.05108.0500
(KH2PO4)
PMSF (Phenylmethylsulphonyl Merck 11836170001
fluoride)
Puromycin (10 mg/ml stock) Sigma-Aldrich P9620
RIPA (Radioimmunoprecipitation Cell Signaling 9806S
assay buffer)
Sodium chloride (NaCl) Sigma-Aldrich 7647-14-5
Sodium dodecyl sulfate (SDS) Sigma-Aldrich L4509
TE (Tris-EDTA buffer) Invitrogen 8019005
TEMED Bio-Rad 110-18-9
Tris Roth 2449.2
Triton X-100 Fluka 93420
Tween20 Sigma-Aldrich 9005-64-5
ß-Mercaptoethanol Sigma-Aldrich M3148

4.2 Consumables
Table2: Consumables with type and the companies they were ordered from.

24
 Material and Methods

Consumable Type Company

Adhesive sealing sheet Micro Amp™ (#4306311) Applied Biosystems
Bacterial culture dish 100x15 mm (#663102) Greiner Bio-One
Blotting Pad 16x18 cm (#732-0606) VWR
Boyden chamber 8 μm pores, PET Corning
(#353182)
CASYcups #3926 9097 OLS OMNI Life Science
Cell culture dish 10 cm (#GB664160) Greiner Bio-One
Cover slips 24x60 mm (#1) Menzel-Gläser
Cover slips, round ∅18 mm (#ECN631-1580) VWR
Cryo conservation tube 1.5 ml (#5000-1020) Thermo Fisher Scientific
Cuvettes 12.5x12.5 mm (#2712120) ratiolab
Flow cytometry tubes 5 ml, 35 μm filter cap Corning
(#352235)
Object slide Superfrost (#630-0952) Gerhard Menzel GmbH
PCR Strips 8-er Strips (#710950) Biozym Scientific GmbH
PCR Strips, caps 8-er Caps (#710960) Biozym Scientific GmbH
Pipette stripettes 5 ml (#4487) Schubert & Weiss
10 ml (#4488) GmbH
25 ml (#4489)
50 ml (#4490)
Pipette filter tips 10 μl (#VT0200) Biozym Scientific GmbH
20 μl (#VT0220)
200 μl (#VT0240)
1250 μl (#VT0270)
Plates (cell culture) 6-well (#657160) Greiner Bio-One
12-well (#665180)

24-well (#353097) Falcon

Plate (cell culture, ultra- 24-well (#CLS3473-24EA) Corning
low attachment)
PVDF (Polyvinylidene 0.2 µm pores (#10600021) GE Healthcare
Fluoride) membrane
Q-tip #300200 Deltalab

25
 Material and Methods

Reaction tubes 15 ml (#188271) Greiner Bio-One

50 ml (#227261)
Safe-lock tubes 1.5 ml (#0030120.094) Eppendorf
Sterile syringe filter 0.45 μm pores (#28145- VWR International
481)
Syringe 20 ml (#300629) BD
Tubes (bacteria) Round bottom (#352059) Falcon
96-well plate, qPCR Micro Amp® (#4346906) Applied Biosystems
96-well plate flat bottom (#269620) Thermo Fisher Scientific

4.3 Devices
Table3: Devices with type and the companies they were ordered from.
Device Type Company
CASY TT Cell Counter (#5651697) OLS OMNI Life Science
Cell Counting chamber 0,100 mm depth, 0,0025 mm² Neubauer
Cell culture incubator Heracell Thermo Fisher Scientific
Centrifuges Centrifuge RC 6+ Thermo Fisher Scientific
Heraeus Multifuge X3R
Microcentrifuge 5417R Eppendorf
Chemiluminescence Fusion SL Vilber lourmat
imager
Flow cytometer LSR II BD
Fluorescence EVOS® FL AMF4300 Invitrogen
microscopes Axio Observer 7 Zeiss
Freezer/Refrigerator -20 °C/+4 °C (Premium Kombi) Liebherr
Freezer -80 °C (Hera Freeze KS 102) Thermo Fisher Scientific
Freezing container Mr. Frosty Thermo Fisher Scientific
Incubator (bacteria) Certomat® R Sartorius
Laminar airflow Herasafe HS9 (S/N 98105547) Heraeus
Magnetic stirrer #36560.00 IKA RCT classic
Microplate reader #185911 BioTek® Instruments
NanoDrop 2000 Thermo Fisher Scientific
PCR cycler DNAEngine® Bio-Rad

26
 Material and Methods

pH-meter SevenEasy Mettler Toledo

Pipettes 2.5 µl, 10 µl, 20 µl, 200 µl, Eppendorf
1000 µl (Research plus)
10 µl multi (8x, research plus)
Pipettus (#L7090280) Hirschmann
Power supply PowerPac™ HC High-Current Bio-Rad
qPCR QuantStudio™ 3 Applied Biosystems
Roller tube mixer KS501 digital IKA Labortechnik
Shaker (orbital) Polymax 1040 Heidolph
Vacuum pump KNF Laboport® Minipumpe Sigma-Aldrich
Vortexer Vortex–Genie 2™ Bender & Hobein AG
Water bath #3043 Kottermann

4.4 Mice and Cell lines

All cell lines were maintained at 37°C in a humidified atmosphere with 5% CO2 in
the indicated culture media (Table4) supplemented with 10% FBS (ThermoFischer
Scientific, 10270-106) (20% FBS + 10% PSCs - CM for pancreatic stellate cells),
1% Penicillin-Streptomycin (PAN Biosystems, P06-07100) and 1% Glutamine
(Gibco, 41965-039). For experiments, cells were used in early passages (Pass. 3 -
20) and were recovered from frozen stocks on a regular basis. Single cells were
counted with Neubauer chambers and seeded accordingly.

Primary murine pancreatic cancer cell lines were generated from PDAC tumors
resected from Kraswt/LSL-G12D; Trp53loxP/loxP; Ptf1awt/Cre;LSL-tdRFPKI/KI;Slug-YFP
(KPCRS)240. These mice express an oncogenic Kras mutation241, a conditional loss
of Trp53242, an R26-LSL-tdRFP243, a Cre recombinase under the control of a Ptf1a
promoter244, and a Slug-YFP reporter system245. Slug-YFP mice were generously
provided by Robert A. Weinberg, Whitehead Institute for Biomedical Research,
Cambridge, MA. Primary tumors were minced and digested with collagenase
(STEMCELL Technologies, 07902). After fibroblast removal, adherent pancreatic
cancer cells were expanded and cultured246.

27
 Material and Methods

The detailed experimental setup, tumor growth data, and imaging of the primary
tumor pertaining to circulating tumor cell counting in vivo (Fig. 5.1.5C) have already
been published in Ref. No.247.

Primary human pancreatic cell lines were generated from resected excess
pancreatic carcinoma tissue that was subcutaneously implanted into nude mice in
vivo. The PDO-derived cell line MetPO1 and MetPO2 originates from resected
excess liver metastasis grown into organoids in vitro. The organoids for generating
the MetPO1 and MetPO2 cell line were kindly provided by the Department of Internal
Medicine I at Ulm University (AG Kleger). Bo80 cell line was kindly provided by the
Department of Molecular Gastroenterological Oncology at Ruhr University Bochum
(AG Hahn). Gö5, Gö7 and Gö13 cell lines were kindly provided by the Department
of Gastroenterology (AG Heßmann) at Göttingen University. Human pancreatic
stellate cell line was kindly provided by Department of Internal Medicine I at Ulm
University (AG Seufferlein).

Table4: Overview of cell lines

Cell line Origin Culture Reference

Media
5493 Murine PDAC (primary tumor) DMEM Ref. No. 1
(Gibco,
41965-039)

8926 Murine PDAC (primary tumor) DMEM Ref. No.1

9228 Murine PDAC (primary tumor) DMEM Ref. No.1
Bo80 Human PDAC (primary tumor); RPMI 1640 unpublished
(Gibco,
PDX-derived
21875-034)

Gö5 Human PDAC (primary tumor); RPMI unpublished

PDX-derived

28
 Material and Methods

Gö7 Human PDAC (primary tumor); RPMI unpublished

PDX-derived
Gö13 Human PDAC (primary tumor); RPMI unpublished
PDX-derived
MetPO1 Human PDAC (liver metastasis); RPMI unpublished
PDO-derived
MetPO2 Human PDAC (liver metastasis); RPMI unpublished
PDO-derived
Panc185 Human PDAC (primary tumor); RPMI Ref. No. 248
PDX-derived
Panc215 Human PDAC (primary tumor); DMEM Ref. No. 248
PDX-derived
Panc354 Human PDAC (primary tumor); RPMI Ref. No. 134
PDX-derived
Pancreatic Human primary DMEM:F12 unpublished
stellate cell pancreatic stellate cells 1:1
line
Lenti-X Human embryonic kidney DMEM RRID:
HEK293T CVCL_6911

4.5 Co-culture technique

4.5.1 Co-culture with PSCs

PDX/O derived cell lines were cultured in 6-well plates at a density of 1 x 105 cells
per well. PSCs were cultured in inserts with 1μm pore size PET membranes at a
density of 1 x 105 cells per insert. Cells were kept separately at 37°C in a humidified
atmosphere with 5% CO2 in the indicated culture media respectively (section 4.1)
for 24 hours. Subsequently, media for both 6-well plate and inserts were replaced
to RPMI (with 10% FBS, 1% Pen-Strep and 1% Glutamine) and inserts with PSCs
were placed over PDX/O derived cell lines on 6-well plate and incubated additionally
for 72 hours to establish dual cultures. For control PDX/O derived cells were cultured
at the density of 1 x 105 cells per insert.

29
 Material and Methods

4.5.2 Condition media for PSCs maintenance

PSCs were thawed and cultured in a T75 flask in DMEM : F12 (1:1) together with
20% FBS, 1% Pen-Strep and 1% Glutamine for 7 days at 37°C in a humidified
atmosphere with 5% CO2. Media was collected in a 50ml tube and centrifuged at
1000 rpm for 5 minutes to remove debris. Media was next filter-sterilized (0.45μm
filter), labelled and stored as PSCs conditioned media (PSCs - CM) at 4°C. 10% of
PSCs – CM is used every time for PSCs media preparation.

4.5.3 Primed and Non-primed PSCs – CM preparation

PDX/O derived cell lines and PSCs were cultured separately in 10cm dishes with
their respective media (section 4.1) for 24 hours at 37°C in a humidified atmosphere
with 5% CO2. Subsequently, media from PDX/O derived cell lines was collected in
a 50ml tube and centrifuged at 1000 rpm for 5 minutes to remove debris. Media was
next filter-sterilized (0.45μm filter) and was labelled as PDX/O – CM. PSCs media
was replaced with this PDX/O – CM and was allowed to grow for additional 24 hours
to prime the PSCs. After 24 hours media from primed PSCs was collected in a 50ml
tube and centrifuged at 1000 rpm for 5 minutes to remove debris. Media was next
filter-sterilized (0.45μm filter) and was labelled as primed PSCs – CM. Non-primed
PSCs conditioned media was prepared by removing the PDX/O – CM based priming
step and adding fresh RPMI (with 10% FBS, 1% Pen-Strep and 1% Glutamine)
media to PSCs instead.

4.6 Organoid culture

5,000 single murine primary adherent cells in 25μl medium249 were mixed with equal
amounts of Matrigel GFR (growth factor reduced, Corning) per well and cultured at
37°C in a humidified atmosphere with 5% CO2. Treatment with MEK inhibitors was
performed on day 1 (for effect on organoid formation) or on day 4 (for effect on
established organoids) for 3 consecutive days. Medium was changed daily.
Metabolically active cells were measured with the Cell Titer-Glo Luminescent Cell
Viability Assay (Promega, G9681) according to the manufacturer’s instructions.

30
 Material and Methods

4.7 Treatments

PD0325901 was used at 0.5μM (5493 cells) or 5μM (8926 and 9228 cells), and
trametinib was used at 0.035μM (5493) or 0.175μM (8926 and 9228 cells) unless
stated otherwise. TGFβ, CXL12 and OSM was used at 10nM. Cell lines are pre-
treated for 30 minutes with EPI-X4, WSCO2, JM#21, inactive peptide before
experimental procedures unless otherwise stated. Concentrations used for EPI-X4,
WSCO2, JM#21, inactive peptide, paclitaxel and gemcitabine are mentioned in the
section 5.Results.

4.8 MTT assay

1,000 cells were seeded in a 96-well plate and incubated for 24 hours at 37°C. After
treatment (as described in section 5.Results), cells were incubated for 3 hours with
5 mg/ml MTT (Merck, M2128). DMSO (Roth, A994) was added, and the optical
density was measured using an Infinite 200 PRO plate reader (Tecan, Switzerland)
at 560 nm.

4.9 Clonogenic assay

10,000 cells per milliliter were seeded in a 24-well plate. After the treatment (as
described in section 5.Results), the media was removed, and the wells were washed
three times with PBS. After washing, 1 ml of diluted Giemsa solution (Merck,
1.09204.0100) (1: 20 in PBS) was added to each well and incubated for 20 min on
a shaker with 150 rpm. Later, Giemsa solution was removed, and the wells were
rinsed with water slowly. Dried plates were then used for imaging. For quantification,
10% acetic acid was added to Giemsa-stained cells and incubated for 20 min on a
shaker with 150 rpm for blue color to develop. 100μl of this solution was then
transferred into a 96-well plate, absorbance was measured using Infinite 200 PRO
plate reader (Tecan, Switzerland) at 590 nm.

4.10 Sphere formation assay

For first-generation spheres, following treatment or co-culture (as mentioned in

section 5.Results), 10,000 cells per milliliter were seeded in ultralow attachment
plates (Corning, 3473) and cultured at 37°C in a humidified atmosphere with 5%
CO2. Spheres were cultured in DMEM-F12 (Thermo Fisher Scientific, 10565018)

31
 Material and Methods

supplemented with B-27 (Thermo Fisher, 17504044) and basic fibroblast growth
factor (Novoprotein, CO46). After 7 days of incubation, > 40μm and < 120μm were
quantified using CASY TT (OMNI Life Science, 5651697). For second-generation
spheres, after counting the first-generation spheres, they were disrupted using
trypsin for 15 minutes at 37°C to make single cells. Equal number of cells per
milliliter were seeded back in ultralow attachment plates. After 7 days of
incubation, > 40μm and < 120μm were quantified using CASY TT.

4.11 Scratch wound assay

Cells were grown to confluency and then serum-starved for 24 hours before scratch
wounds were made using a sterile 10μl pipette tip. The cells were then cultured with
medium containing vehicle or PD0325901 for 24 hours. Images were captured after
24 hours and quantified using ImageJ (version 1.49, https://imagej.nih.gov/ij/).

4.12 Migration assay

Migration assays were performed using inserts with 8μm pore size PET membranes
(Corning, 353097). After 24 hours, invaded cells were fixed with 4% PFA and
stained with DAPI (Merck, 10236276001). Ten random 10x fields were chosen and
photographed (Keyence), and the pictures were quantified using ImageJ.

4.12.1 Migration towards serum

Cells in serum-free medium were added to the inserts. In the bottom well of the
companion plate, media containing 10% FBS was added.

4.12.2 Migration towards CXCL12

Cells in medium (with or without 10%FBS, as mentioned in section 5.Results) were

added to the inserts. In the bottom well of the companion plate, media (with or
without 10%FBS, as mentioned in section 5.Results) containing 10µM CXCL12
(Peprotech, 300-28A) was added.

32
 Material and Methods

4.13 Molecular Cloning

4.13.1 Bacterial culture and plasmid isolation

DH5α-T1R- E. coli transformed with BMI1, CXCR4 or scramble targeting shRNA

plasmids were streaked onto LB agar plates supplemented with 50mg/ml Ampicillin
and incubated overnight at 37°C. A single colony was picked from the plate and
transferred to a tube containing LB media. Bacterial growth was facilitated by
overnight (<18 hours) incubation at 37 °C and 220 rpm in Certomat® R. 1 mL of this
starting culture was used to inoculate 250mL LB in conical flask incubated overnight
(<18 hours) incubation at 37°C and 220 rpm. Bacterial work was conducted near an
open flame. For plasmid isolation, PureLinkTM HiPure Plasmid Mediprep Kit from
Invitrogen was utilized. LB media containing bacterial growth was centrifuged for
10min at 5000 rpm and 4°C, the plasmids isolated in accordance with the kits
protocol. Purified plasmids were resuspended in 100µL TE buffer and proper
isolation was confirmed by plasmid restriction.
Table5: Overview of plasmids used

Plasmid Number Manufacturer

pLKO.1-puro-shRNA-BMI1 TRCN0000020155 Sigma-Aldrich
pLKO.1-puro-shRNA-BMI1(2) TRCN0000020158 Sigma-Aldrich
pLKO.1-puro-shRNA-CXCR4 TRCN0000256864 Sigma-Aldrich
pLKO.1-puro-shRNA-CXCR4 (2) TRCN0000256865 Sigma-Aldrich
pLKO.1-puro-shRNA-scramble MFCD07785395 Sigma-Aldrich
pMD2.G 12259 Addgene
psPAX2 12260 Addgene

4.13.2 Lentiviral particle production in Lenti-X cells

Lenti-X (293T) cells were seeded at 5 x 106 cells per 10cm dish, cultured for 24
hours were transfected with 8µg plasmids of interest (shBMI1; shCXCR4; shSCR),
5.5µg psPAX2 (lentiviral packaging), 2µg pMD2 (lentiviral envelope) and 8.75% PEI
transfection reagent (50 mg PEI, 20 mM HEPES and 150mM NaCl; for DNA charge
opposition) in serum-free DMEM. After 4 hours at 37 °C, media was replaced with
DMEM supplemented with 10% FBS, 1% Pen-strep and 1% Glutamine. Virus was
harvested after 2 and 4 days after transfection, pooled and centrifuged at 1500 rpm,

33
 Material and Methods

room temperature for 2 min to pellet remaining cells. Supernatants were sterile
filtered (0.45µm), mixed with Lenti-X Concentrator (3:1) (Clontech, 631231), and
incubated at 4°C for 30 min prior to 45 min centrifugation at 1500 rpm and 4°C for
virus precipitation. Virus pellets were resuspended in DMEM.

4.13.3 Lentiviral transduction and selection of target cells

Human PDAC cell lines were seeded at a density of 1 x 106 cells per 2ml in 6 well
plates. They were then infected with 0 - 25µl virus polybrene (10µg/ml enhances
retroviral infection) containing RPMI after 24 hours, as well as 48 hours of culture.
Cell culture media was replaced 24 hours after the second infection and the cells
were allowed to recover for 48 hours. For drug selection, media was substituted with
puromycin (3mg/ml) supplemented RPMI every three days, until non-transduced
cells (0µl virus) in control wells died. Consequently, only successfully transduced
cells retaining the puromycin resistance gene survived the treatment. Upon
confluency, samples of each transduction were collected for qRT-PCR analysis to
assess the minimal virus concentration with the strongest downregulation. Cells with
the optimal virus concentration were expanded for conducting experiments.

4.14 RNA isolation and real-time PCR

4.14.1 RNA Isolation and cDNA synthesis

Total RNA was prepared using the RNeasy kit with on-column genomic DNA
digestion following the manufacturer’s instructions (Qiagen, 74106). First-strand
cDNA was prepared using the QuantiTect Reverse Transcription Kit (Qiagen,
205314).

4.14.2 Real-time quantitative PCR

Reactions were performed with the PerfeCTa SYBR Green FastMix PCR Reagent
(Qiagen, 204057) using a QuantStudioTM 3 machine (Applied Biosystems) using
primers (Table 6 and 7). Results were analyzed using the 2-ddCt method relative
to Ywhaz expression (for murine primary cell lines) and HPRT or GAPDH (for
human primary cell lines). Reactions were carried out from at least three
independent experiments.

Table6: Overview of mouse primers used with their sequences

34
 Material and Methods

Gene Sense Primer(5’-3’) Antisense Primer (5’-3’)

CD44 GCTTCCGAGAGGGGGCGACT TCCATTGTCGCCAGCAGCAGA

Cdh1 AGTACAGCTTGTTGTTAGTG TTGGAGCAAATGTTGATGAG
Sca1 ATGTCAACTCCTCCTTCTAC AATGATCTGAGCTATAGAGGC
Snai2 GATTGATGTCATGTATGAGGAG CTCTGTATTTCAATGGAAGTGG
Sox2 GAGAGAAAGTTAGAGAGGGG CGTGTAATTCTGAGAAACTGG
Sox9 TGCCCAGAAAATGAAAAAGG GGATGACACAGCGTGAGAGA
Vim CCCGCTTCTCTGAAAGGCTCTC CTCTGCTGCTGCTGCTGGTAG
Ywhaz TATCTTCAAGAGGTGCCTATC ACCTTTTCAACCAGACTTTC

Table7: Overview of human primers used with their sequences

Gene Sense Primer(5’-3’) Antisense Primer (5’-3’)

ALDH1 GCTTCCGAGAGGGGGCGACT TCCATTGTCGCCAGCAGCAGA

ABCC5 AGTACAGCTTGTTGTTAGTG TTGGAGCAAATGTTGATGAG
ABCG2 ATGTCAACTCCTCCTTCTAC AATGATCTGAGCTATAGAGGC
BMI1 GATTGATGTCATGTATGAGGAG CTCTGTATTTCAATGGAAGTGG
CDADC1 GAGAGAAAGTTAGAGAGGGG CGTGTAATTCTGAGAAACTGG
CDH1 TGCCCAGAAAATGAAAAAGG GGATGACACAGCGTGAGAGA
C-MYC CCCGCTTCTCTGAAAGGCTCTC CTCTGCTGCTGCTGCTGGTAG
CXCR4 GGTGGTCTATGTTGGCGTCT TGGAGTGTGACAGCTTGGAG
CXCL12 GCTTTTCAATGTTAGCCAC TTAAGCTCCATCACTAACAAC
DCK1 TATCTTCAAGAGGTGCCTATC ACCTTTTCAACCAGACTTTC
GAPDH CAGGAGCGAGATCCCT GGTGCTAAGCAGTTGGT
GLI1 CAGCCCAGATGAATCACCAA GCTCAGACTTCAGCTGGCAAGT
GLI2 CGCCAAGCACCAGAATCGCA TGCGGAGGTGCACGTCATTG
GPX1 CTACTTATCGAGAATGTGGC CAGAATCTCTTCGTTCTTGG
GPX2 AATTTGGACATCAGAACTGC GGCTGCTCTTCAAGATTTAG
HPRT TGCTCGAGATGTCATGAAGG AATCCAGCAGGTCAGCAAAG
KLF4 AGTCGCTTCATGTGGGAGAG TCCCATCTTTCTCCACGTTC

35
 Material and Methods

POU5F1 GAAGGATGTGGTCCGAGTGT GCCTCAAAATCCTCTCGTTG

PROM1 AAGCATTGGCATCTTCTATG TTTGCTCTGGAGTTTCATTC
NANOG AGATGCCTCACACGGAGACT AAGTGGGTTGTTTGCCTTTG
SHH CGGTGAAAGCAGAGAACTCGGTGG TTCACCAGCTTGGTGCCGCC
SNAI1 GCTCCTTCGTCCTTCTCCTC TGACATCTGAGTGGGTCTGG
SNAI2 ACAGCGAACTGGACACACAT CACAGTGATGGGGCTGTATG
SOD1 GTTTGGAGATAATACAGCAGG TGCCTCTCTTCATCCTTTG
SOX2 AGAACCCCAAGATGCACAAC CGGGGCCGGTATTTATAATC
TWIST1 CTCGGACAAGCTGAGCAAG CAGCTTGCCATCTTGGAGTC
VIM GACAATGCGTCTCTGGCACGTCT TCCTCCGCCTCCTGCAGGTTCTT
ZEB1 AAAGATGATGAATGCGAGTC TCCATTTTCATCATGACCAC

4.15 RNA Seq

Primary murine cell lines generated from PDAC of Ptf1awt/Cre; Kraswt/LSL-G12D;

Trp53loxP/loxP (CKP) animals250 were cultivated in standard cell culture dishes and
treated with respective IC50 concentrations of trametinib (4 cell lines, IC50 ranging
from 8 - 25nM). RNA was isolated using the Maxwell® RSC simply RNA Cells Kit
(Promega) according to the manufacturer’s instructions after 48 hours. RNA Seq
was performed by CeGaT (Tübingen, Germany). Library preparation was performed
with the TruSeq Stranded mRNA kit (Illumina), and 2x 100 bp was sequenced on
Hi-Seq 4000 (Illumina). Demultiplexing of the sequencing reads was performed with
Illumina CASAVA (2.17). Adapters were trimmed with Skewer (version 0.1.116)
(Jiang et al. 2014). RNA Seq data were quantified using the quasi-mapping
approach of Salmon251. TXImport252 and DESeq2253 were used to import transcript-
level counts and to perform differential expression analysis.

4.16 Protein extraction and western blot analysis

4.16.1 Protein sample preparation

Harvested cells were lysed in ice-cold RIPA buffer (Cell Signaling, 9806S)
supplemented with PhosSTOp™ (Merck, 4906845001) and a protease inhibitor
cocktail (Merck, 11836170001).

36
 Material and Methods

4.16.2 Western Blot

For each sample, equal amounts of protein were applied to a 10% SDS-
polyacrylamide gel and immunoblotted onto PVDF membranes (GE Healthcare,
10600021). Membranes were blocked for 2 hours in 5% BSA in 1x TBST, probed
with the primary antibodies (Table 8) overnight at 4°C, washed with 1x TBST, and
incubated with indicated secondary antibody (Table 8) for 2 hours. The
chemiluminescence detection was performed according to the manufacturer’s
instructions (Merck, WBKLS0500). Stripping of blots was performed when required
with RestoreTM Western Blot Stripping Buffer according to the manufacturer’s
instructions (Thermo Scientific, 21059). Blots were re-blocked after stripping
overnight at 4°C in 5% BSA in 1x TBST before probing with the primary antibody.

Table 8: Antibodies used in western blot analysis

Antibody (Clone) / Species Dilution Company Catalogue#
Primary Antibodies
BMI1 (F-9) / Mouse 1:1000 Santa Cruz sc390443
Biotechnology
Vimentin (D21H3) / Rabbit 1:1000 Cell Signaling 5741S
AKT / Rabbit 1:1000 Cell Signaling 9272
Phosphorylated (Ser473) AKT / 1:1000 Cell Signaling 9271
Rabbit
IκB-α / Rabbit 1:1000 Cell Signaling 9242
Phosphorylated (Ser32) (14D4) 1:1000 Cell Signaling 2859
IκB-α / Rabbit
Nanog (polyclonal) / Rabbit 1:1000 Cell Signaling 3580S
c-MYC (E5Q6W) / Rabbit 1:1000 Cell Signaling 18583S
Sox9 (polyclonal) / Rabbit 1:2000 Merck ab5535
N-Cadherin (D4R1H) / Rabbit 1:1000 Cell Signaling 13116S
E-Cadherin (24E10) / Rabbit 1:1000 Cell Signaling 3195S
Slug (H-140) / Rabbit 1:1000 Santa Cruz sc15391
Biotechnology
ERK1/2 (137F5) / Rabbit 1:1000 Cell Signaling 4695S

37
 Material and Methods

Phosphorylated ERK1/2 1:2000 Cell Signaling 4370S

(D13.14.4E) / Rabbit
GAPDH (polyclonal) / Rabbit 1:10000 Sigma-Aldrich G9545
Secondary Antibodies
Recombinant Anti-mouse IgGκ L 1:2000 Santa Cruz sc516102
(polyclonal), HRP-conjugated Biotechnology
Goat Anti-rabbit IgG H+L 1:2000 Thermo Fisher G21234
(polyclonal), HRP-conjugated Scientific

4.17 Immunofluorescence

4.17.1 E-cadherin

Murine primary pancreatic cells were cultured on coverslips (Hecht Assistant

41001115), then fixed with 4% PFA (Sigma), washed with 1x PBST (PBS with 0.3%
Triton X-100), and blocked for 1 h at room temperature with blocking solution (10%
goat serum and 0.3% Triton X-100 in PBS). An anti-E-cadherin primary antibody
(Cell Signaling) was diluted in blocking solution and incubated o/n at 4°C, whereas
the secondary antibody (Alexa Fluor 488 goat anti-rabbit, Invitrogen) was diluted in
blocking solution and incubated for 2 h at room temperature. All washes were done
with 1x PBST. Cells were visualized and photographed (Keyence).

4.17.2 Nile Red and Phalloidin

Pancreatic stellate cells (PSCs) were cultured at a density of 50,000 on coverslips

in a 6-well plate. PDX/O cell lines were grown in inserts with 1μm pore size PET
membranes. After 24 hours, PDX/O cell lines grown in inserts were placed over
PSCs on coverslips. After 72 hours, pancreatic stellate cells were washed three
times with PBS. Subsequently, PSCs were fixed with 2% PFA (Sigma, 158127) for
20 min at room temperature and then washed again three times with PBS. Next
PSCs were permeabilized with 0.7% TritonX-100 (Fluka, 93420) solution for 15 min
at room temperature followed by three PBS washing. Coverslips with PSCs were
then incubated with Nile red (1:500 diluted in PBS) (Sigma, N3013) and Phalloidin-
Atto (1:500 diluted in PBS) (Sigma, 94072) inside a dark humid chamber for 60 min.
PSCs were washed three times in PBS and mounted on slides using Prolong™ Gold
reagent with DAPI. Fluorescence at 565 nm was visualized and photographed

38
 Material and Methods

EVOS FL (Invitrogen, AMF4300).

4.17.3 Phalloidin

PDX/O cell lines were cultured at a density of 50,000 on coverslips in a 6-well plate.
After 24 hours or treatment (mentioned in section 5.Results), cells were washed
three times with PBS. Subsequently, they were fixed with 2% PFA for 20 min at
room temperature and then washed again three times with PBS. Next cells were
permeabilized with 0.7% TritonX-100 solution for 15 min at room temperature
followed by three PBS washing. Coverslips with cells were then incubated with
Phalloidin-Atto (1:500 diluted in PBS) inside a dark humid chamber for 60 min. Cells
were washed three times in PBS and mounted on slides using Prolong™ Gold
reagent with DAPI. Fluorescence at 565 nm was visualized and photographed Zeiss
Axio Vert.A1(Zen Blue).

4.18 Flow Cytometry

Flow cytometry analyses were performed using LSR II (BD). Dead cells were
excluded using DAPI.

4.18.1 Annexin V staining

Annexin V staining was performed using a BD Annexin V APC kit according to the
manufacturer’s instructions (BD, 550474). Data were analyzed using FlowJo v10
(Ashland, OR).

4.18.2 Identifying CTCs in murine blood

For the identification and quantification of circulating tumor cells in the blood of mice,
counting beads (Thermo Fisher Scientific, C36950) were added to the whole blood
aspirated from the right ventricle. After red blood cell lysis, samples were stained
with an EpCAM-APC antibody (Thermo Fisher) or an appropriate isotype control
(BD). The number of cells and beads in the final sample was recorded, and the total
quantity of cells in the original sample was calculated.

4.18.3 Surface Protein Staining

For surface protein staining (CD133; CXCR4), 1 x 106 cells were resuspended in
100µl ice-cold FACS buffer supplemented with 6 µl Gamunex for 15 min at 4°C.

39
 Material and Methods

Cells were then incubated with specific antibodies or an appropriate isotype control
(Table 9) for 30mins at 4°C. After PBS wash, cells were resuspended in 500µl of
ice-cold FACS buffer with DAPI in FACS tubes for analysis.

Table9: Antibodies used in flow cytometry

Antibody (Clone) Dye Company Catalogue#

Mouse CD133/1 IgG1κ (AC133) PE Miltenyi Biotec 130-113-108
CXCR4?
Recombinant BMI1 IgG1 (REA438) APC Miltenyi Biotec 130-124-301

Isotype control Mouse IgG1κ (MOPC-21) PE BD 559320

Isotype control Mouse IgG2ακ (G155-178) APC BD 555576

4.19 Enzyme-linked immunosorbent assay (ELISA)

Media levels of secreted CXCL12 were determined using CXCL12A Human ELISA
kit (Thermo Fischer Scientific, EHCXCL12A) according to the manufacturer
guidelines. Conditioned media from PSCs monoculture, PDX/O – PDX/O and
PDX/O – PSCs co-cultures was collected after 72 hours and stored at -80°C.
Thawed conditioned media samples were diluted 2-fold with assay diluent B.
Absorbance was measured using Infinite 200 PRO plate reader (Tecan,
Switzerland) at 450 nm.

4.20 Antibody-competition assay

50,000 SupT1 cells/well were seeded in 96-well V-bottom microtiter plates in FACS
buffer. Buffer was removed by centrifugation and cells precooled at 4°C for 15 min.
In the meantime, compounds were serially diluted in ice cold PBS and 12G5-APC
antibody was diluted in cold FACS buffer at a concentration of 0.49nM. Afterwards,
15µl compound directly followed by 15µl antibody was added to the cells. For
stability experiments, the mixture was carefully mixed after addition to the cells.
Cells were incubated at 4°C for 2 hours before unbound antibody and compounds
were removed by 2 washing steps followed by fixation in 2% PFA buffer. Cells were
analyzed for MFI values of bound antibody by flow cytometry using FACS
CytoFLEX. For calculation, the isotype control (Table 9) was subtracted, and values

40
 Material and Methods

normalized to 12G5 stained PBS control.

4.21 Microtiter based stability determination

JM#21 peptide was 150-fold diluted in media supplemented with PBS, 10% FCS or
to reach final concentrations of 20µM. The t = 0 sample was immediately taken and
stored at -80°C. Media/peptide mixture was then transferred to 37°C and shook at
350 rpm. At given time points samples were taken and stored at -80°C. For
measuring the functional activity of the media/peptide samples, the mixtures were
thawed and serially diluted in ice cold PBS (starting with 100 % sample). 12G5-APC
antibody competition was then performed as described before.

4.22 Fatty acid conjugation with JM#21

JM#21 conjugation to palmitic acid (JM#143 and JM#194) has been previously
described in the published thesis Harms M: Endogenous CXCR4 Antagonists: Role
in HIV-1 Transmission and Therapy of CXCR4-linked diseases239, PhD Dissertation,
Institute of Molecular Virology, University of Ulm, 2021.

4.23 Loading JM#21 to silica nanoparticles

The mesoporous silica nanoparticles with a radially (MSN) and dendritic (DMSN)
254
oriented pore system were synthesized similar to as mentioned in Ref. No. and
255
Ref. No. respectively. In brief, cetyltrimethylammonium bromide (CTAB) was
used as a structure directing agent and a mixture of tetramethyl orthosilicate
(TMOS) and (3-aminopropyl) trimethoxysilane (APTMS) acted as silica precursors.
The final synthesis mixture had a molar ratio of 0.68 TMOS: 0.1 APTMS: 1.00 CTAB:
0.21 NaOH: 1746 MeOH: 4142 H2O. Instead of extraction with acidic ethanol, the
surfactant was removed via calcination at 550°C for 5.5 hours. JM#21 was adsorbed
onto the radial mesoporous silica nanoparticles (MSN) and dendritic mesoporous
255
silica nanoparticles (DMSN) similar to as described previously in Ref. No. .

4.24 Bioinformatics prediction

GEPIA (http://gepia.cancer-pku.cn/intex.html) was applied to conduct tumor/normal
differential expression analysis and patient survival analysis (overall survival and
disease-free survival)256. The protein – protein interactions were carried out using
STRING database (https://string-db.org)257.

41
 Material and Methods

4.25 Software

Table10: Software used for data analysis.

Name Version Company
FlowJo 10.8.0 BD
GraphPad Prism 5, 8, 9.2 GraphPad Software Inc.
ImageJ 1.53 NIH
QuantStudio™ Design & Analysis 1.5.1 Applied Biosystems

4.26 Statistics

Results for continuous variables are presented as mean SEM unless stated
otherwise or as median with quartiles and min/max values for box plots.
Statistical analysis was performed using the GraphPad Prism software. In order
to detect normality Shapiro-wilk test was performed. Normally distributed data
was analyzed using two-way ANOVA (or one-way ANOVA for knockdown
experiments) otherwise Mann-Whitney U test was applied. Fold change data
was analyzed using student’s t test. *p values < 0.05 were considered
statistically significant. For each experiment, the sample size is indicated in the
figure legend.

42
 Results

5 Results
5.1 MEK signaling contributes to maintenance of CSCs phenotype
and promotes migration of pancreatic cancer cells

5.1.1 MEK inhibition reduces the growth of murine PDAC cells

To begin with, we generated KPCRS mice. These mice express genetic

modifications under the control of a pancreas-specific p48-Cre: an oncogenic
KrasG12D mutation, loss of p53, an RFP reporter construct and a YFP reporter
construct under the control of the Slug (Snail2) promoter. We analyzed the effects
of the small molecule MEK inhibitor PD0325901 on KPCRS mouse-derived primary
pancreatic cancer cell lines. Via MTT assays we detected a dose-dependent
response of the primary cells to MEK inhibition, confirming their dependency on a
functional RAS-RAF-MEK-ERK pathway. Interestingly, the different cell lines
displayed variable responsiveness to PD0325901 (Fig. 5.1.1A). PD0325901
concentrations slightly above the respective IC50 were used to perform the
subsequent experiments. Next, we demonstrated that no significant changes in
apoptosis or cell death were detected after 72 hours of treatment at the indicated
concentrations in two of the tested cell lines (Fig. 5.1.1B,C). Furthermore, to validate
the effectiveness of the compound, we confirmed the downregulation of ERK
phosphorylation as a downstream target of MEK, upon treatment with MEK
inhibitors (Fig. 5.1.1 D).

Figure 5.1.1. MEK inhibition reduces the growth of murine PDAC cells: (A) MTT assays to

43
 Results

determine the respective IC50 for three mouse tumor-derived primary cell lines (5493, 8926 and
9228) treated with PD0325901 (PD). The respective IC50 is depicted for each cell line. (B & C)
Apoptosis induction under MEK inhibitor treatment (3 days) as measured by Annexin V staining and
analyzed by flow cytometry in (B) 5493 and (C) 8926 cell lines respectively. (D) Western blot analysis
for phosphorylated and total ERK 1/2 was performed on the three cell lines treated with PD0325901
at the indicated concentrations. GAPDH was used as a loading control. n ≥ 3 for all experiments, ∗
p < 0.05 vs. control. ns = not significant. (Figures and figure legend are licensed under a Creative
Commons Attribution License (CC BY 4.0) https://creativecommons.org/licenses/by/4.0/ ©2019,
Walter et al. 2019 Ref. No. 1).

5.1.2 MEK inhibition targets pancreatic cancer stem cells

Sphere cultures enrich for cancer stem cells (CSCs)258. We therefore performed
sphere formation assays after 72 hours of pretreatment with PD0325901 to
elucidate the functional effects of MEK inhibition on CSCs. MEK inhibition
significantly reduced the number of spheres formed after treatment (Fig. 5.1.2A).
Additionally, the size of the spheres formed was notably smaller compared to the
vehicle-treated control (Fig. 5.1.2B). We next investigated the expression of genes
associated with pluripotency and stemness. Indeed, we observed a significant
downregulation in Sox9, Sox2, CD44, and (only 5493 cell line) Sca1 in adherent cell
cultures after MEK inhibition (Fig. 5.1.2C). In 3D sphere cultures, we observed
comparable, albeit slightly less pronounced effects of PD0325901 on stemness-
associated genes as in adherent cultures (Fig. 5.1.2D). Furthermore, treatment with
trametinib, a clinical-grade MEK inhibitor, also showed a significant down regulation
of stemness-associated genes (Figure 5.1.2E) in adherent cells. In order to
generalize our approach to a larger set of primary cell lines, we performed RNAseq
on 4 additional KPC-derived primary mouse cell lines treated with trametinib or
vehicle. The subsequent analysis revealed downregulation of Nanog, Sox9, and
Klf4 (Fig. 5.1.2 F), matching the qRT-PCR dataset.

44
 Results

Figure 5.1.2. Effects of MEK inhibition on stemness: (A) Sphere formation capacity of cells
pretreated with PD0325901 at the indicated concentrations. (B) Quantification of large spheres
(>120μm) and representative micrographs of sphere cultures after 7 days. Gene expression of the
indicated stemness-associated genes in (C) adherent cultures or (D) spheres after treatment with
vehicle or PD0325901 (PD). (E) Gene expression in adherent cultures and (F) RNA seq analysis of
murine primary cancer cell lines treated with DMSO vehicle (white) or trametinib (grey) for 48h. n ≥
3 for all experiments. ∗ p < 0.05 vs. control. (Figures and figure legend are licensed under a Creative
Commons Attribution License (CC BY 4.0) https://creativecommons.org/licenses/by/4.0/ ©2019,
Walter et al. 2019 Ref. No. 1).

45
 Results

5.1.3 MEK inhibition decreases migration in a dose-dependent manner

In order to delineate the role of MEK signaling in cell migration, we performed

scratch wound assays on three primary tumor cell lines. MEK inhibition resulted in
significantly reduced “wound closure” (i.e., migration capacity) in all primary cell
lines. A clear dose-dependent decrease in the wound closure activity was observed
(Fig. 5.1.3A, B). We also used transwell migration assays to further confirm the
effect of MEK inhibition on cell migration in a 3D set-up. After pretreatment with
PD0325901, a significant reduction in migration was observed (Fig. 5.1.3C, D),
confirming the results observed in the scratch wound assay. These results indicated
an essential role of MEK signaling on the migratory activity in PDAC cells.

46
 Results

Figure 5.1.3. Effect of MEK inhibition on Migration: (A) The percentage of wound closure 24h
after scratch wound induction in primary cells treated with vehicle or PD0325901 at the indicated
concentrations. (B) Representative micrographs of wound closure at 0 hours and 24 hours in the cell
line 5493. (C) Quantification and (D) representative micrographs (10x, DAPI nuclear staining) of
Transwell migration assays with vehicle or PD0325901 treated cells in the indicated concentrations.
n ≥ 3 for all experiments. ∗ p < 0.05 vs. control. (Figures and figure legend are licensed under a
Creative Commons Attribution License (CC BY 4.0) https://creativecommons.org/licenses/by/4.0/
©2019, Walter et al. 2019 Ref. No. 1).

5.1.4 MEK Inhibition abrogates TGFβ-Induced EMT

As the role of TGFβ in EMT induction and subsequent metastasis is well

documented, we wondered whether MEK inhibition could abrogate an active EMT
program initiated by TGFβ. We therefore treated the cells with TGFβ for 3 or 6 days,
respectively (Figure 5.1.4A). PD0325901 was added after 3 days of pretreatment
with TGFβ or vehicle control.

The treatment effects were then measured by western blotting of Slug and Vimentin
as key signaling molecules of EMT activation. Both proteins were strongly
upregulated following TGFβ treatment. As hypothesized, Slug protein levels were
reduced by subsequent MEK inhibition (Figure 5.1.4B). Interestingly, MEK inhibition
was unable to overcome the effects of continuous TGFβ stimulation (evident via
both Slug and E-cadherin protein levels). Furthermore, we utilized the endogenous
Slug-YFP reporter system contained in the mouse model in our cells: TGFβ
treatment resulted in robust Slug-YFP expression, as shown by the increase in

RFP+YFP+ cells after 3 days and 6 days (Figure 5.1.4C, D). Interestingly, MEK

inhibition with PD0325901 significantly diminished this RFP+YFP+ population by

almost 50% after 3 days of TGFβ treatment.

However, matching the observation in western blotting, MEK inhibition was not able
to abrogate the effects of continuous TGFβ treatment. In line with these
experiments, immunofluorescence staining revealed that while E-cadherin is
suppressed by TGFβ treatment, PD0325901 treatment induces re-expression of E-
cadherin, indicating the induction of a more epithelial and thus more differentiated
phenotype (Figure 5.1.4E). Altogether, the results indicate that pharmacological
MEK inhibition using PD0325901 inhibits TGFβ-induced EMT and migration in vitro.

47
 Results

Figure 5.1.4. Effects of MEK inhibition on TGFβ-induced EMT: (A) Experimental overview for
TGFβ and PD0325901 co–treatment. (B) Western blot analysis of key proteins involved in EMT with
treatment in vitro. Gapdh was used as a loading control. (C) Quantification of RFP+YFP+ cells under
treatment as indicated and (D) representative cytometry blots. (E) Immunofluorescence micrographs
of E-cadherin expression with treatment as indicated. n ≥ 3 for all experiments, n ≥ 2 for Western
blots. ∗ P < 0.05 vs. control, # P < 0.05 vs. TGFβ. (Figures and figure legend are licensed under a

48
 Results

Creative Commons Attribution License (CC BY 4.0) https://creativecommons.org/licenses/by/4.0/

©2019, Walter et al. 2019 Ref. No. 1).

5.1.5 MEK Inhibitors Prevent Organoid Formation and Decrease CTCs in vivo

3D organoid cultures cultures better reflect in vivo conditions by maintaining cell-

to-cell signaling and therefore are a more physiological cell culture model than 2D
monolayer. Thus, organoids better recapitulate the original tumor and are preferable
to predict treatment response as compared to monolayer cultures259. Therefore, we
performed organoid formation and treatment experiments by generating 3D
organoid cultures from our primary cell lines (Fig. 5.1.5A). With MEK inhibitior
treatment we observed a significant decrease in organoid formation and for the
treatment of already established organoids. This holds true for the treatment with
PD0325901 or trametinib (Fig. 5.1.5A, B).

As a functional in vivo readout for efficacy of MEK inhibitors on migration, we

quantified circulating tumor cells (CTCs) in KPC mice treated with refametinib,
another clinical-grade MEK inhibitor. In order to do so, we extracted blood from the
right ventricle of KPC mice treated either with refametinib or with vehicle control. In
agreement with the in vitro data, after refametinib treatment we observed
significantly fewer CTCs in the blood stream of these mice (Fig. 5.1.5C), suggesting
a significantly reduced metastatic potential.

49
 Results

Figure 5.1.5. MEK inhibitors diminish organoid formation and eliminate CTCs in vivo: (A)
Representative micrograph of RFP+ organoid cultures (left, 5493 cells). Organoid formation under
treatment with PD0325901 (PD) (middle) and trametinib (Tra) (right) in 3 primary cell lines treated as
indicated. (B) Viability of preformed organoids with PD0325901 treatment in vitro. (C) Absolute
numbers of circulating tumor cells (CTCs) in KPC mice treated with either vehicle control or
refametinib (as described in Ref. No. 247). n≥3 for all experiments. ∗ p < 0.05 vs. control. (Figures and
figure legend are licensed under a Creative Commons Attribution License (CC BY 4.0)
https://creativecommons.org/licenses/by/4.0/ ©2019, Walter et al. 2019 Ref. No. 1).

5.2 CXCL12 – CXCR4 signaling is imperative for maintenance and

metastatic propensity of miCSCs

So far we described that MAPK signaling pathway, acting through MEK, confers
invasive properties to pancreatic cancer cells via transcription factor Slug. However,
MEK inhibition could not overcome the effects of sustained TGFβ-indued EMT
activation. These results further shed light on the fact that single agent therapies,
even if targetted, may lead to resistance. Moreover, different cell types in the tumor
microenvironment (TME) release factors such as cytokines (e.g., TGFβ, SHH) and
chemokines which can interact with CSCs thereby promoting chemo-resistance and
metastasis. One such chemokine secreted in the TME is CXCL12 (or SDF1) which
is a ligand for the chemokine receptor CXCR4. As a next step we tried to delineate
the role of the CXCL12 – CXCR4 axis in the maintenance of migrating CSCs
(miCSCs).

5.2.1 CD133 and CXCR4 are highly expressed in PDAC

In previous studies, a distinct subpopulation of CD133 and CXCR4 expressing cells

were identified as migrating cancer stem cells in PDAC60,62. In order to understand
the mechanisms by which CD133 and CXCR4 may contribute to CSC features and
EMT, we first used in-silico analysis tools. Tumor/normal differential expression
analysis indicated that PROM1 (CD133) and/or CXCR4 show a significantly
elevated expression in PDAC tumor compared to normal tissue (Fig. 5.2.1A). We
next generated a protein – protein interaction (STRING) network of CD133 and
CXCR4 with other relevant factors involved in metastasis (green) and CSCs (red)
(Fig. 5.2.1B). These interactions demonstrate the indispensable nature of CD133
and CXCR4 in these contexts. Furthermore, patient survival analysis showed that
high content of PROM1 or CXCR4 or both together reduces disease-free survival

50
 Results

(Fig. 5.2.1C), and high PROM1 expression correlates with low overall survival of
patients (Fig. 5.2.1D).

Figure 5.2.1. CD133 and CXCR4 are highly expressed in PDAC: (A) Box plots represent relative
mRNA expression for the gene signatures PROM1, CXCR4 or PROM1 and CXCR4 in normal tissue
compared to PDAC tumor tissue (TCGA database, GEPIA). (B) Protein – protein interactions of
PROM1 (CD133) and CXCR4 with relevant factors involved in metastasis (green) and CSCs (red)
(STRING). (C) Kaplan – Meier curves showing disease-free survival for patients with high compared
to low relative mRNA expression of PROM1, CXCR4, or PROM1 and CXCR4 in PDAC (TCGA
database, GEPIA). (D) Kaplan – Meier curves showing overall survival for patients with high
compared to low relative mRNA expression of PROM1, CXCR4, or PROM1 and CXCR4 in PDAC
(TCGA database, GEPIA). The threshold for splitting high and low expression cohorts was set at the
median for box plots, *p < 0.05.

51
 Results

5.2.2 PSCs secreted factors upregulate CD133 and CXCR4 surface

expression

In this study, we used a total of 9 PDX- (patient-derived xenograft) or PDO- (patient-

derived organoids) derived cell lines, which genetically recapitulate the primary
tumor and maintain genetic features of PDAC260. In order to characterize these cell
lines, we first performed cell surface staining for CD133 and CXCR4. Flow cytometry
analysis of the 9 cell lines tested revealed differential quantities of CSC (CD133+)
(Fig. 5.2.2A) and miCSC (CD133+ CXCR4+) subpopulations (Fig. 5.2.2B). This
was in contrast to the previous in silico data that showed a significant elevation of
both CSCs and miCSCs subpopulation in the PDAC tumors. We therefore wanted
to investigate pathways responsible for CSC and miCSC upregulation.

Secreted factors from these activated PSCs play a crucial role in the development
of cellular heterogeneity in PDACs104. In order to evaluate the relative contribution
of factors secreted by stellate cells to CSC and miCSC maintenance, we
performed flow cytometry analysis for CD133 and CXCR4 in Panc354 (PDX cell
line) and MetPO1 (PDO cell line) that were exposed to different media conditioned
by PSCs (conditioned medium, CM) at three time points as depicted in (Fig.
5.2.2C). Interestingly, we observed a shift towards higher levels of CD133+,
CXCR4+ and CD133+CXCR4+ cells (Fig. 5.2.2D, E, F) in the presence of CM
from primed PSCs (pink) when compared to CM from non-primed PSCs (blue) or
no CM (gray) in all the time points tested. Since media from PDX/O cell lines was
used to prime PSCs, these results suggest that although secreted factors from
PSCs alone upregulate CD133 and CXCR4, tumor cell-primed PSCs either
secrete more or specific factors that can increase CSCs (CD133+) and miCSCs
(CD133+ CXCR4+) (Fig. 5.2.2G).

52
 Results

Figure 5.2.2. PSCs secreted factors upregulate CD133 and CXCR4 surface expression: (A)
Screening for percent CD133+ cells and (B) CD133+ CXCR4 + found in adherent cell culture of
indicated cell lines. (C) Experimental scheme to evaluate CD133 and CXCR4 surface expression to
identify patient derived xenografts (PDX) or patient derived organoids (PDO) crosstalk with
pancreatic stellate cells (PSCs) at the protein level. (D) FACS analysis performed on Panc354 and
MetPO1 when exposed to no conditioned media (grey), conditioned media from non - primed PSCs
(blue) or conditioned media from primed PSCs (pink) for CD133+ cells, (E) CXCR4+ cells and (F)
CD133+ CXCR4+ cells represented as fold change against no conditioned media. (G)
Representative cytometry blots. Error bars represent the standard deviation. n≥3 for all experiments.
*p < 0.05, ns = not significant.

53
 Results

5.2.3 Tumor cell – stellate cell crosstalk upregulates CSCs and miCSCs
population, and potentiates CXCL12 release

In order to delineate the extent of tumor cell – stellate cell crosstalk in maintaining
the CSC and miCSC subpopulations and their features, we used three PDX cell
lines (Panc354, Gö13 and Bo80) and one PDO cell line (MetPO1) for the
subsequent experiments. PDX/O cell lines were co-cultured with pancreatic stellate
cells (or PDX/O cell lines as control) for 72 hours. Then cells were harvested to
perform further analysis as depicted in (Fig. 5.2.3A). In line with our previous data,
flow cytometry analysis for CD133 showed a significantly increased CSC (CD133+)
population in Panc354 and MetPO1 cell lines (co-cultured with PSCs (pink)
compared to respective control (grey)) (Fig. 5.2.3B, D). Furthermore, flow cytometry
analysis for CD133 and CXCR4 also revealed a significant increase in miCSCs
(CD133+ CXCR4+) subpopulation in Panc354, MetPO1 and Gö13 cells (co-cultured
with PSCs compared to respective control) (Fig. 5.2.3C, D).

Gene expression analysis from PDX/O derived cell lines co-cultured with PSCs
(compared to control) (Fig. 5.2.3E) revealed significant upregulation of stemness-
associated genes (NANOG, POU5F1, ALDH1a1, BMI1). We also found DCK1, a
gene catalyzing the rate-limiting step in gemcitabine metabolism261 to be
downregulated, whereas genes involved in the metabolism of reactive oxygen
species (ROS), such as SOD1 and (only for Panc354 - PSCs) GPX1 were
upregulated. Moreover, ABCC5, which is involved in the transport of nucleotide
analogues, and which has been speculated to perform excessive efflux of nucleotide
analogue-based drugs such as 5-FU or gemcitabine262, was found to be
downregulated in all cell lines tested. Additionally, the epithelial marker CDH1 was
found to be significantly downregulated, while EMT markers (VIM, SNAI1, SNAI2)
were upregulated. This further confirms stark changes in the transcriptional profile
of PDX/O cell lines during co-culture with PSCs (compared to control). (Fig. 5.2.3E).

We then performed 3D transwell migration assays towards serum containing

medium (Fig. 5.2.3F, G) and sphere formation assays (Fig. 5.2.3H, I) with PDX/O
derived cell lines (co-cultured with PSCs or PDX/O cell lines) to validate the findings
of our flow cytometry and gene expression analysis on a functional level. PDX/O

54
 Results

derived cell lines co-cultured with PSCs showed tremendously increased migratory
capacity as well as sphere forming propensity.

We next wanted to confirm the changes in exogenous factors secreted by PSCs

when co-cultured with PDX/O cell lines. Since CXCL12 is one of the cytokines found
in abundance within the TME263 and is also ligand for CXCR4167,168, we performed
ELISAs for exogenous CXCL12 on conditioned medium from PSCs monoculture,
PDX/O – PDX/O and PDX/O – PSC co-cultures (Fig. 5.2.3J). Interestingly, PSC
monocultures secreted CXCL12 in significantly higher amounts than PDX/O –
PDX/O co-cultures. However, PDX/O – PSCs co-culture further potentiates CXCL12
secretion, which also corresponds to an activated (or primed) PSC state263.

To confirm activation of PSCs when co-cultured with PDX/O cell lines, we performed
immunofluorescence (IF) for lipid droplets (Fig. 5.2.3K) using Nile Red staining (red
dots) (Fig. 5.2.3L). IF revealed a loss of lipid droplets (marked by yellow
arrowheads) in PSCs co-cultured with PDX/O when compared to PSC monocultures
demonstrating an activation of PSCs. Altogether this confirms that tumor cell –
stellate cell crosstalk promotes self-renewal capacity and invasiveness of tumor
cells by promoting CSCs and miCSCs. Furthermore, the crosstalk promotes
CXCL12 release from stellate cells.

55
 Results

Figure 5.2.3. Tumor cell – stellate cell crosstalk upregulates CSC and miCSC population and
potentiates CXCL12 release: (A) Experimental scheme to evaluate CD133 and CXCR4 surface
expression, qRT-PCR, migration assay, sphere assay, ELISA and IF for co-culture PDX/O – PSCs
and PDX/O – PDX/O. (B) Flow cytometry analysis performed for percent CD133+ cells and (C)
CD133+ CXCR4+ cells for depicted cell lines. (D) Representative cytometry plots. (E) Gene
expression analysis for indicated genes using qRT-PCR for total RNA extracted from PDX/O cell
lines (indicated) co-cultured with PSCs or PDX/O cell lines represented as log2 fold change. (F)
Transwell migration assays for PDX/O cell lines (indicated) cultivated in dual culture with PSCs or
PDX/O cell lines and (G) representative micrographs (10x, DAPI nuclear staining). (H) Sphere
formation assays for PDX/O cell lines (indicated) cultivated in dual culture with PSCs or PDX/O cell
lines and (I) representative micrographs of sphere cultures after 7 days. (J) ELISA performed for
CXCL12 concentration (pg/ml) in media from cultures as indicated. (K) Quantifications and (L)
representative micrographs with yellow arrowheads marking lipid droplets stained with Nile Red (red
dots), nucleus with DAPI (blue) and actin filaments with Phalloidin (red). Error bars represent the
standard deviation. n≥3 for all experiments. *p < 0.05, ns = not significant.

56
 Results

5.2.4 CXCL12 maintains CSC and miCSC population through CXCR4

Our previous result suggests that CXCL12 is important in the activation of CSCs
and miCSCs. In order to confirm that CXCL12 indeed acts through CXCR4 to
maintain or promote CSCs and miCSCs and their functional characteristics, we
employed sh-RNA mediated knockdown of CXCR4 in one PDX cell line (Panc354)
and one PDO cell line (MetPO1). We confirmed the successful knockdown of
CXCR4 on gene (Fig. 5.2.4A), protein (Fig. 5.2.4B) and cell surface (Fig. 5.2.4C)
expression level for both cell lines.

Next, we selected the cell carrying the most efficient knockdown (sh1_CXCR4 and
sh2_CXCR4) and cultured Panc354 and MetPO1 (sh_SCR, sh1_CXCR4 and
sh2_CXCR4) in the presence or absence of CXCL12 for 24 hours or co-cultured the
cells with PSCs for 72 hours. Afterwards the harvested cells were subjected to
further experiments as depicted (Fig. 5.2.4D). Flow cytometry analysis for CD133
showed an increase in CSCs (CD133+) population for CXCL12-treated MetPO1
sh_SCR (but not in Panc354 sh_SCR) compared to sh_SCR control. A further
increase in CSCs was observed in MetPO1 sh_SCR (but not in Panc354 sh_SCR)
co-cultured with PSCs. Interestingly, after CXCR4 knockdown (sh1_CXCR4 and
sh2_CXCR4) in both cell lines, neither CXCL12 treatment nor PSCs co-culture were
able to upregulate CSCs (CD133+) population to sh_SCR levels (Fig. 5.2.4E,F,I).
supporting the critical role of the CXCR4-CXCL12 axis in CSC propagation. In line
with our previous data (Fig. 5.2.3C), an increase in miCSCs (CD133+ CXCR4+)
subpopulation was observed in both CXCL12-treated and PSCs co-cultured
Panc354 sh_SCR and MetPO1 sh_SCR cell lines. As expected, knockdown of
CXCR4 in both Panc354 and MetPO1 (sh1_CXCR4 and sh2_CXCR4) resulted in a
significant decrease in miCSCs (CD133+ CXCR4+), which neither CXCL12
treatment nor PSCs co-culture was able to “rescue”, i.e., upregulate to sh_SCR
levels (Fig. 5.2.4G,H,I).

In order to elucidate the underlying mechanisms of CXCL12 – CXCR4 signaling for

maintaining CSCs and miCSCs, we performed pathway-focused gene expression
analysis: We examined a number of genes associated with EMT, stemness,
chemoresistance and antioxidants in Panc354 and MetPO1 (sh_SCR, sh1_CXCR4
and sh2_CXCR4) treated with (or without) CXCL12 or co-cultured with PSCs (Fig.

57
 Results

5.2.4J,K). We found the epithelial marker CDH1 to be significantly downregulated

for sh_SCR (in both cell lines) when treated with CXCL12 or co-cultured with PSCs
(compared to sh_SCR control). We also found EMT markers (VIM, SNAI1, SNAI2)
to be upregulated for sh_SCR when treated with CXCL12 or co-cultured with PSCs
in both cell lines. CXCR4 knockdown in MetPO1 (sh1_CXCR4 and sh2_CXCR4),
on the other hand, showed noticeable decrease in expression of EMT markers,
which neither CXCL12 treatment nor PSCs co-culture was able to rescue. Panc354
CXCR4 knockdown cells (sh1_CXCR4 and sh2_CXCR4) showed a similar trend in
gene expression of CDH1 and EMT markers as MetPO1, albeit less pronounced.
NANOG expression, on the other hand, was significantly increased upon CXCL12
treatment, which was markedly reduced by CXC4 knockdown (sh1_CXCR4 and
sh2_CXCR4) in both cell lines. PSC co-culture had a similar effect on NANOG
expression for Panc354 sh_SCR and MetPO1 sh_SCR, although less pronounced
than CXCL12 treatment. However, CXCR4 knockdown (sh1_CXCR4 and
sh2_CXCR4) exhibited significantly decreased NANOG expression with PSCs co-
culture in both cell lines.

Genes involved in imparting chemoresistance in PDAC, such as CDA1 (catalyzes

the inactivation of gemcitabine to dFdU264 and ABCC5 was found to be significantly
overexpressed for sh_SCR when treated with CXCL12 or co-cultured with PSCs in
both cell lines. CXCR4 knockdown in MetPO1 (sh1_CXCR4 and sh2_CXCR4)
downregulated CDA1 and ABCC5 with (or without) CXCL12 treatment or PSCs co-
culture. Panc354 CXCR4 knockdown (sh1_CXCR4 and sh2_CXCR4) without
CXCL12 treatment were able to significantly reduced CDA1 and ABCC5 expression.
However, this effect seems to be diluted after CXCL12 treatment or PSCs co-
culture. DCK1, on the other hand, is significantly downregulated for sh_SCR when
treated with CXCL12 or co-cultured with PSCs in both cell lines. Although, CXCR4
knockdown (in both Panc354 and MetPO1) does result in a slight increase in DCK1
expression, the expression remains unchanged after treatment with CXCL12 or co-
culture with PSCs.

Furthermore, significant upregulation of antioxidant enzymes such as the GSH

peroxidase GPX1 and superoxide dismutase SOD1 was observed for sh_SCR
when treated with CXCL12 or co-cultured with PSCs in both cell lines. CXCR4

58
 Results

knockdown in both cell lines downregulated this increase in expression with (or
without) CXCL12 treatment or with PSC co-culture. Thus, we conclude that CXCL12
can recapitulate most of the downstream effects observed with PSC co-culture,
which were reversed or abrogated by CXCR4 knockdown. Altogether, these results
confirm that the regulatory axis between CXCL12 and CXCR4 maintains CSCs and
miCSCs.

59
 Results

Figure 5.2.4. CXCL12 maintains CSC and miCSC population through CXCR4: (A) CXCR4 gene
expression analysis for the depicted cell lines. (B) Western blot analysis of CXCR4. GAPDH was
used as a loading control. (C) Flow cytometry analysis of CXCR4 surface expression in depicted cell
lines. (D) Experimental setup to evaluate CD133 and CXCR4 surface expression and qRT-PCR for
Panc354 and MetPO1 (sh_SCR, sh1_CXCR4 and sh2_CXCR4). (E) Flow cytometry analysis of
CD133+ CSCs in Panc354 (sh_SCR, sh1_CXCR4 and sh2_CXCR4) treated with (or without)
CXCL12 or co-cultured with PSCs. (F) Flow cytometry analysis of CD133+ CSCs in MetPO1
(sh_SCR, sh1_CXCR4 and sh2_CXCR4) treated with (or without) CXCL12 or co-cultured with PSCs.
(G) Flow cytometry analysis of CD133+ CXCR4+ miCSCs in Panc354 (sh_SCR, sh1_CXCR4 and
sh2_CXCR4) treated with (or without) CXCL12 or co-cultured with PSCs. (H) Flow cytometry analysis
of CD133+ CXCR4+ miCSCs in MetPO1 (sh_SCR, sh1_CXCR4 and sh2_CXCR4) treated with (or
without) CXCL12 or co-cultured with PSCs and (I) representative cytometry plots for MetPO1 cell
line. (J) Targeted gene expression analysis for indicated genes in Panc354 (sh_SCR, sh1_CXCR4
and sh2_CXCR4) treated with (or without) CXCL12 or co-culture with PSCs. (K) Targeted gene
expression analysis for indicated genes in MetPO1 (sh_SCR, sh1_CXCR4 and sh2_CXCR4) treated
with (or without) CXCL12 or co-culture with PSCs. Error bars represent the standard deviation. n≥3
for all experiments. *p < 0.05, ns = not significant.

5.2.5 The CXCL12 – CXCR4 axis sustains self-renewal capacity and

metastatic propensity of miCSCs

We further wanted to validate our findings regarding the role of CXCL12 in the
maintenance of key functional characteristics of miCSCs, and to delineate
downstream pathways by which CXCL12 – CXCR4 signaling is able to impart these
abilities. In order to do so, we performed sphere formation assays with Panc354 and
MetPO1 (sh_SCR, sh1_CXCR4 and sh2_CXCR4) after 24 hours of CXCL12
treatment (Fig. 5.2.5A). We observed that CXCR4 knockdown indeed significantly
hampered sphere forming capacity of both Panc354 and MetPO1 (sh1_CXCR4 and
sh2_CXCR4) cells compared to sh_SCR (Fig. 5.2.5A). To confirm the effects on

60
 Results

stemness characteristics of treatment with CXCL12 in the context of a CXCR4

knockdown, we next measured NANOG and BMI1 protein expression by western
blotting. In line with the previous qRT-PCR data (Fig. 5.2.4 J,K), knockdown of
CXCR4 in Panc354 and MetPO1 (sh1_CXCR4 and sh2_CXCR4) resulted in a
decrease in NANOG levels with and without CXCL12 treatment (Fig. 5.2.5B). Even
though BMI1 protein levels did not change after CXCR4 knockdown in Panc354 and
MetPO1 (sh1_CXCR4 and sh2_CXCR4), the increase in BMI1 levels due to
CXCL12 treatment in sh_SCR (Panc354 and MetPO1) was reduced after CXCR4
knockdown (sh1_CXCR4 and sh2_CXCR4) in both cell lines (Fig. 5.2.5B).

We next performed 3D transwell migration assays towards medium containing

CXCL12 with Panc354 and MetPO1 cells (sh_SCR, sh1_CXCR4 and sh2_CXCR4)
(Fig. 5.2.5C). Knockdown of CXCR4 in Panc354 and MetPO1 (sh1_CXCR4 and
sh2_CXCR4) reduced the migrating capabilities of the cells by approximately 50%.
Further confirming this on a protein level (Fig. 3.2.5D), knockdown of CXCR4 in
Panc354 (sh1_CXCR4 and sh2_CXCR4) greatly increased E-CAHDERIN levels
with or without CXCL12 treatment. Additionally, VIMENTIN levels in MetPO1 cells
(sh1_CXCR4 and sh2_CXCR4) were repressed with (or without) CXCL12 treatment
(Fig. 5.2.5D).

We next looked into phosphorylation of AKT and IκB-α protein in Panc354 and
MetPO1 cells (sh_SCR, sh1_CXCR4 and sh2_CXCR4) treated with and without
CXCL12 (Fig. 5.2.5E). A decrease in phosphorylation of AKT was found in both
Panc354 and MetPO1 cells (sh1_CXCR4 and sh2_CXCR4) upon CXCL12
treatment. Similarly, phosphorylation of IκB-α was decreased in MetPO1
(sh1_CXCR4 and sh2_CXCR4) upon CXCL12 treatment (Fig. 5.2.5E). In order to
find the converging point through which CXCL12 – CXCR4 signaling maintains both
self-renewal and metastatic potential, we generated a protein – protein interaction
(STRING) network of CXCL12 – CXCR4 (green) with other known relevant factors
involved in metastasis (red), stemness (purple), sonic hedgehog signaling (blue),
AKT signaling (grey) and NFκB signaling (yellow) (Fig. 5.2.5F). These interactions
suggest that BMI1 (red arrow) could be one of the entities through which CXCL12 –
CXCR4 is acting, providing a mechanistic explanation of the relationship between
EMT and stemness in cancer cells. Taken together, these results demonstrate that

61
 Results

CXCL12 sustains self-renewal capacity and metastatic propensity of miCSCs

through CXCR4 and shed light on a possible role of BMI1 in both stemness and
EMT.

Figure 5.2.5. The CXCL12 - CXCR4 axis sustains self-renewal capacity and metastatic
propensity of miCSCs: (A) Sphere formation assay for depicted cell lines after CXCL12 treatment
and representative micrographs of sphere cultures after 7 days. (B) Western blot analysis of NANOG
and BMI1 for depicted cell lines and treatment conditions. GAPDH was used as a loading control.

62
 Results

(C) Migration assay towards CXCL12 for depicted cell lines and representative micrographs (10x,
DAPI nuclear staining). (D) Western blot analysis of CADHERIN-1 and VIMENTIN for depicted cell
lines and treatment conditions. GAPDH was used as a loading control. (E) Western blot analysis of
p-AKT, AKT, p- IκB-α and IκB-α for depicted cell lines and treatment conditions. GAPDH was used
as a loading control. (F) Protein – protein interactions of CXCL12 and CXCR4 with relevant factors
involved in metastasis (red), stemness (purple), sonic hedgehog signaling (blue), AKT signaling
(grey) and NFκB pathway (yellow) (STRING). Error bars represent the standard deviation. n≥3 for all
experiments. *p < 0.05, ns = not significant.

5.2.6 BMI1 downstream CXCL12 – CXCR4 regulates EMT and stemness

BMI1 is part of the PRC1 complex maintaining self-renewal and stemness together
with EZH2, which is a component of PRC2265. In order to delineate the underlying
role of BMI1 in EMT and stemness downstream of CXCL12 – CXCR4 signaling, we
employed sh-RNA mediated knockdown of BMI1 in one PDX cell line (Panc354)
and one PDO cell line (MetPO1). We confirmed knockdown of CXCR4 on gene (Fig.
5.2.6A) and protein (Fig. 5.2.6B) expression level for both cell lines.

We next performed gene expression analysis for stemness-associated genes in

Panc354 and MetPO1 (sh_SCR, sh1_BMI1 and sh2_BMI1) (Fig. 5.2.6C).
Interestingly, BMI1 knockdown in MetPO1 (sh1_BMI1 and sh2_BMI1) led to
significant downregulation of PROM1, ALDH1a1, NANOG, POU5F1, KLF4 and
SOX2. Panc354 (sh1_BMI1 and sh2_BMI1) showed similar trends regarding the
decreased expression of stemness-associated genes. We confirmed these findings
on a protein level performing western blots for NANOG, c-MYC and SOX9 (Fig.
5.2.6D). BMI1 knockdown in Panc354 (sh1_BMI1 and sh2_BMI1) showed an
evident decrease in NANOG, c-MYC and SOX9 compared to sh_SCR control.
Similarly, MetPO1 (sh1_BMI1 and sh2_BMI1) compared to sh_SCR control also
displayed reduced NANOG and SOX9 protein levels (Fig. 5.2.6D). However, sphere
formation assays with Panc354 and MetPO1 (sh_SCR, sh1_BMI1 and sh2_BMI1)
revealed no significant difference between sh_SCR and knockdown of BMI1
(sh1_BMI1 and sh2_BMI1) (Fig. 5.2.6E).

In order to understand the effect of BMI1 on EMT and migration capacity, gene
expression analysis on key EMT factors was performed (Fig. 5.2.6F). On the one
hand, BMI1 knockdown in MetPO1 (sh1_BMI1 and sh2_BMI1) showed a relevant
increase in expression of the epithelial marker CDH1, on the other hand it showed

63
 Results

no significant changes in Panc354 (sh1_BMI1 and sh2_BMI1). Similarly, gene

expression of EMT markers (SNAI1, SNAI2, TWIST1 and ZEB1) in MetPO1
(sh1_BMI1 and sh2_BMI1) decreased strongly, but in Panc354 cells (sh1_BMI1 and
sh2_BMI1), we observed varied results: Although SNAI1, TWIST1 and ZEB1
showed a marked decrease in expression in Panc354 BMI1 knockdown (sh1_BMI1
and sh2_BMI1) compared to Panc354 sh_SCR, VIM and SNAI2 expression were
significantly upregulated (Fig. 5.2.6F). We therefore performed western blots to
verify these findings on the protein level (Fig. 5.2.6G). In line with the gene
expression analysis, Panc354 BMI1 knockdown (sh1_BMI1 and sh2_BMI1) showed
no difference in E-CADHERIN expression compared to Panc354 sh_SCR. MetPO1
BMI1 knockdown (sh1_BMI1 and sh2_BMI1), however, showed a decrease in E-
CADHERIN expression when compared to MetPO1 sh_SCR. Interestingly, EMT
markers (N-CADHERIN and SLUG) for both Panc354 and MetPO1 (sh1_BMI1 and
sh2_BMI1) displayed a strong decrease in expression compared to sh_SCR control,
respectively (Fig. 5.2.6G). This effect of BMI1 knockdown on EMT was further
confirmed by migration assays (Fig. 5.2.6H) where we observed a significant
decrease in migratory capacity by BMI1 knockdown in both Panc354 and MetPO1
cells (sh1_BMI1 and sh2_BMI1) compared to sh_SCR control.

We also viewed BMI1 knockdown from a morphological standpoint. Therefore, we

stained actin filaments using phalloidin to observe the effect of BMI1 knockdown in
Panc354 and MetPO1 cells (sh_SCR, sh1_BMI1 and sh2_BMI1) (Fig. 5.2.6I). The
metastatic cell line MetPO1 sh_SCR displayed actin stress fibers and mesenchymal
structures (marked with white arrows) which were drastically diminished in the BMI1
knockdown in MetPO1 (sh1_BMI1 and sh2_BMI1), as they appeared more
epithelial. Statistically, BMI1 knockdown showed a strong decrease in mesenchymal
structures in MetPO1 (sh1_BMI1 and sh2_BMI1) compared to MetPO1 sh_SCR
control. In our previous objective, we had shown that MEK inhibition is associated
with reduced PDAC cell growth, migration capacity and stemness properties.
Therefore, we also looked at the effect of BMI1 knockdown on the phosphorylation
of ERK1/2 protein (Fig. 5.2.6J). Interestingly, BMI1 knockdown (sh1_BMI1 and
sh2_BMI1) showed a decrease in phosphorylation of ERK1/2 compared to
respective sh_SCR controls in both Panc354 and MetPO1 cells. Taken together,
our results demonstrate stemness- and migration-promoting capabilities of

64
 Results

BMI1.ther, our results depicted stemness and migration promoting capabilities of

BMI1.

Figure 5.2.6. BMI1 downstream CXCL12 – CXCR4 regulates EMT and stemness: (A) BMI1 gene

65
 Results

expression analysis for the depicted cell lines. (B) Western blot analysis of BMI1. GAPDH was used
as a loading control. (C) Gene expression analysis for indicated cell lines with genes involved in
stemness using qRT-PCR. (D) Western blot analysis of NANOG, c-MYC and SOX9 for indicated cell
lines. GAPDH was used as a loading control. (E) Sphere formation assays for depicted cell lines. (F)
Gene expression analysis for indicated cell lines with genes involved in EMT using qRT-PCR. (G)
Western blot analysis of CADHERIN-1, CADHERIN-2 and SLUG for indicated cell lines. GAPDH was
used as a loading control. (H) Migration assays towards serum containing media for depicted cell
lines. (I) Quantifications and representative micrographs for indicated cell lines with white
arrowheads marking mesenchymal structures of actin filaments stained with Phalloidin (pink) and
nucleus stained with DAPI (blue). (J) Western blot analysis of p-ERK for indicated cell lines. Total
ERK was used as a control. Error bars represent the standard deviation. n≥3 for all experiments. *p
< 0.05, ns = not significant.

5.2.7 CXCL12 – CXCR4 maintains CSCs and miCSCs population via BMI1

BMI1 knockdown exhibited its relevance in some of the most important pathways
involved in EMT and stemness. We further wanted to confirm that signaling via the
CXCL12 - CXCR4 axis critically involves BMI1 to maintain both CSCs and miCSCs.
Therefore, we treated Panc354 and MetPO1 (sh_SCR, sh1_BMI1 and sh2_BMI1)
cells with (or without) CXCL12 for 24 hours. The harvested cells were used to
perform flow cytometry for CD133 and CXCR4 (Fig. 5.2.6A).

In line with our hypothesis, we found a significant decrease in the CSC (CD133+)
population in BMI1 knockdown Panc354 and MetPO1 cells (sh1_BMI1 and
sh2_BMI1) compared to the respective sh_SCR after treatment with CXCL12 (Fig.
5.2.6B, D). Similarly, a significant decrease in the miCSC (CD133+ CXCR4+)
population was also observed in MetPO1 cells with BMI1 knockdown (sh1_BMI1
and sh2_BMI1) compared to the respective sh_SCR cells after CXCL12 treatment
(Fig. 5.2.6C, D). As expected, Panc354 sh_SCR showed a strong increase in
miCSCs (CD133+ CXCR4+) when treated with CXCL12 (compared to control-
treated sh_SCR). Surprisingly, we observed an increase in miCSC (CD133+
CXCR4+) after BMI1 knockdown in Panc354 (sh1_BMI1 and sh2_BMI1). But
CXCL12 treatment in Panc354 (sh1_BMI1 and sh2_BMI1) resulted in a strong
decrease in miCSCs (CD133+ CXCR4+) (compared to CXCL12 treated sh_SCR).

66
 Results

In a nutshell, these results demonstrate the significance of CXCL12 – CXCR4

signaling to sustain EMT and stemness features, and the indispensable nature of
BMI1 downstream to regulate and maintain both CSC and miCSC populations.

Figure 5.2.7. CXCL12 – CXCR4 maintains CSCs and miCSCs population via BMI1: (A)
Experimental scheme to evaluate CD133 and CXCR4 surface expression using flow cytometry for
Panc354 and MetPO1 (sh_SCR, sh1_BMI1 and sh2_BMI1) with (or without) CXCL12. (B) Flow
cytometry analysis of CD133+ cells in Panc354 and MetPO1 (sh_SCR, sh1_BMI1 and sh2_BMI1)
treated with (or without) CXCL12. (C) Flow cytometry analysis of CD133+ CXCR4+ cells in Panc354
and MetPO1 (sh_SCR, sh1_BMI1 and sh2_BMI1) treated with (or without) CXCL12 and (D)
representative cytometry plots for MetPO1 cell line. Error bars represent the standard deviation. n≥3
for all experiments. *p < 0.05, ns = not significant.

67
 Results

5.3 Employing endogenous human peptide to target CXCR4 and

eliminate miCSCs

In our previous objective we established that CXCL12/CXCR4 signaling engages in

forming a bidirectional tumor-stromal signaling loop which sustains and promotes
CSCs and miCSCs populations and their metastatic and chemoresistant phenotype.
It is therefore evident that targeting CXCR4 may provide a significant improvement
in the therapeutic efficacy of other targeted anticancer therapies. Endogenous
Peptide Inhibitor of CXCR4 (EPI-X4) is a natural antagonist of the CXC chemokine
receptor 4 (CXCR4) which was discovered from a human derived peptide library
screen234. EPI-X4 is a 16-mer peptide that is released from human serum albumin
(HSA) by acidic aspartic proteases such as Cathepsin D and E266. Our next objective
was to use these endogenous human peptides, EPI-X4 and other derivatives
thereof, to specifically target CXCR4 and explore whether such targeting could
eliminate miCSCs in PDAC and thus qualify as a potential improvement to therapy.

5.3.1 JM#21, most potent EPI-X4 derivative to target CXCR4

In order to uncover the potency of EPI-X4 and its derivatives to functionally target
CXCR4 in human pancreatic cancer cells, we utilized 3D transwell migration assays
to determine effects on the CXCR4-mediated migratory and metastatic activity. We
used the CXCR4 ligand CXCL12 as a chemoattractant to test the inhibitory capacity
of four peptides: EPI-X4, the modified EPI-X4 derivates WSCO2, JM#21 and an
inactive variant as control on Panc354 cells (Fig. 5.3.1A, B). Although 1µM and
10µM concentrations of EPI-X4 and WSCO2 significantly downregulated migration
capacity of Panc354 to approximately 50%. In case of JM#21 even lower
concentrations (0.1μM and 0.5μM) were sufficient to strongly abrogate migration.
As expected, the inactive peptide showed no effect on the migration of Panc354
cells towards CXCL12. We further confirmed the effect of JM#21 on migration in
other PDX-derived primary human pancreatic cancer cell lines, namely MetPO1,
Bo80 and Gö13 (Fig. 5.3.1C). Concentrations of 5μM and 10μM JM#21 significantly
downregulated the migratory capacity of all the cell lines. Taken together these
results confirmed that JM#21 is one of the most potent EPI-X4 derivatives regarding
the effect on migration of primary PDAC cells in vitro.

68
 Results

The IC50 was calculated for each time point (0, 2, 8 and 24 hours). JM#21 incubated
in 10% FCS was found to be inactivated (by the proteases contained in FCS) over
time, whereas there was no such inactivation observed in PBS or B27 serum
supplement (Fig. 5.3.1D), which is critical for the potential use of these peptides for
extended time periods in vitro. Next, dose-dependent inhibition of antibody binding
by JM#21 was measured. JM#21 was active up to 24 hours in culture media
supplemented with B27 instead of FCS (where more than 50% JM#21activity was
lost within the first 10 hours) (Fig. 5.3.1E). As a consequence of these peptide
stability assays and recognizing the low stability of these peptides in serum-
containing culture conditions234, we used B27 as serum supplement for the next
experiments. JM#21 was used at 10μM concentration for further experiments to
achieve effects with the same drug dose for all cell lines.

69
 Results

Figure 5.3.1. JM#21, most potent EPI-X4 derivative to target CXCR4: (A) Migration assays
towards CXCL12 using Panc354 for EPI-X4 and its derivatives at depicted concentrations. Pre-
treatment with EPI-X4, WSCO2, JM#21 and the inactive peptide was applied for 30 minutes. (B)
Representative micrographs (10x, DAPI staining) of transwell migration assays in Panc354 cells for
the indicated conditions and concentrations. (C) Migration assays towards CXCL12 for indicated cell
lines using JM#21 and the inactive peptide at depicted concentrations. (D) Peptide stability assays
for JM#21 in PBS, 10% FCS and B27 to detect activity up to 24 hours. n≥3 for all experiments. *p <
0.05, ns = not significant.

5.3.2 JM#21 inhibits CXCL12 induced EMT and stemness

Since we observed a strong inhibition in migration by JM#21, we next analyzed

morphological changes of the cells. During EMT, cells must acquire a mesenchymal
phenotype which requires changes in cytoskeletal dynamics and composition. We
stained actin structures using Phalloidin in JM#21- (or control-) treated Panc354 and
MetPO1 cells supplied with CXCL12 for 15 minutes and 6 hours, respectively (Fig.
5.3.2A, B). CXCL12 promoted actin stress fibers, mesenchymal structures and
spindle like cell shape within 15 minutes of exposure in both cell lines (white
arrowheads Fig. 5.3.2C). In line with the migration assays, JM#21 significantly
reduced these mesenchymal structures in presence of CXCL12 in Panc354 and
MetPO1 cells for both the time points tested.

70
 Results

We then performed pathway-focused gene expression analysis. As expected, the

key factors involved in the regulation of EMT were found to be upregulated by
CXCL12. JM#21 strongly inhibited expression of numerous EMT regulating factors
(SNAI1, SNAI2, ZEB2, TWIST1) in all cell lines tested, both in the presence and
absence of CXCL12 (Fig. 5.3.2D). Since the role of sonic hedgehog (SHH) signaling
in EMT has been widely discussed267,268, we analysed expression levels of SHH
pathway-related genes. We found that JM#21 can abrogate the increased
expression of SHH, GLI1 and GLI2 induced by CXCL12 (Fig. 5.3.2D). Gene
expression analysis of stemness-related factors revealed a similar decrease in
CXCL12-induced expression of POU5F1, NANOG and KLF4 albeit less
pronounced. The effects of JM#21 on ALDH1a1 expression were found to be
different, depending on the cell line used: Panc354 and Gö13 cells showed a slight
decrease in expression of ALDH1a1 after JM#21 treatment. However, MetPO1 and
Bo80 cells showed a marked increase (Fig. 5.3.2E). Further confirming these on a
protein level (Fig. 5.3.2F), CXCL12 was found to strongly decrease CADHERIN1
levels (in MetPO1 and Bo80) and increase VIMENTIN expression (in Panc354).
Oncostatin M (OSM) is a well-established inducer of EMT in pancreatic cancer
cells269 within the tumor microenvironment and was therefore used to corroborate
CXCL12 treatment effects on EMT. CXCL12 showed similar, although less
noticeable modulation of EMT factors as OSM. JM#21 treatment, across all cell lines
tested, showed prominent downregulation in CXCL12 induced expression of
VIMENTIN and CADHERIN-2 as well as upregulation of CADHERIN-1 (Fig. 5.3.2F).

Looking at the effects on stemness, a distinct downregulation of NANOG and BMI1

was observed after CXCL12 and JM#21 co-treatment (Fig. 5.3.2F). As anticipated,
CXCL12 increased sphere formation (Fig. 5.3.2F, G) in Panc354 cells. JM#21
treatment resulted in reduced sphere formation (for both 1st generation and 2nd
generation spheres) (independently of the presence of CXCL12). Altogether, these
results further established JM#21’s capacity to abrogate CXCL12-induced EMT and
stemness in primary PDAC cells.

71
 Results

Figure 5.3.2. JM#21 inhibits CXCL12 induced EMT and stemness: (A) Quantifications of percent
mesenchymal structures after 15 minutes and 6 hours of CXCL12 treatment in Panc354 and (B)
MetPO1 cells. JM#21 pre-treatment was applied for 30 minutes. (C) Representative micrographs for

72
 Results

indicated cell lines with white arrowheads marking mesenchymal structures of actin filaments stained
with Phalloidin (pink) and nuclear staining using DAPI (blue). (D) Gene expression analysis for
indicated cell lines with genes involved in EMT and SHH pathway. (E) Gene expression analysis for
indicated cell lines with genes involved in stemness. (F) Western blot analysis of CADHERIN-1,
VIMENTIN, CADHERIN-2, NANOG and BMI1 for indicated cell lines. GAPDH was used as a loading
control. (G) Sphere formation assays for 1st and 2nd generation of sphere formation and (H)
representative micrographs of sphere cultures after 7 days in Panc354 cells after the indicated
treatments. JM#21 and inactive peptide pre-treatment was applied for 30 minutes, respectively. Error
bars represent the standard deviation. n≥3 for all experiments. *p < 0.05, ns = not significant.

5.3.3 JM#21 sensitizes PDAC cells towards chemotherapy

We next wanted to ascertain the potential therapeutic value of JM#21. Clonogenic

assays on Panc354 cells pre-treated with CXCL12 revealed that JM#21 has no
effect on cell survival (Fig. 5.3.3A). Therefore, we next developed a combination
treatment plan (Fig. 5.3.3B), according to which PDX/O -derived cells cultured with
CXCL12 were given 2 days of JM#21 pre-treatment followed by a 3-day co-
treatment with either gemcitabine or paclitaxel (1μM or 5μM). Afterwards, cells were
given no further treatment for 7 days to enable and measure relapse growth.
Interestingly, the cells that survived treatment with gemcitabine or paclitaxel were
able to grow back during the relapse phase. However, the cells with combination
treatment could not, supporting the idea of a combined treatment regimen. Both
Panc354 and MetPO1 cells showed significantly reduced cell survival with
combination treatment using gemcitabine and paclitaxel (Fig. 5.3.3C, D). Gö13 cells
were sensitive to high concentrations (5μM) of gemcitabine and paclitaxel and
therefore combination treatment (JM#21 with gemcitabine or paclitaxel 5μM) did not
show a strong decrease in cell growth compared to single treatment (Fig. 5.3.3E).
However, low concentration combination treatments (JM#21 with gemcitabine or
paclitaxel 1μM) were indeed significantly lower than single treatment of gemcitabine
or paclitaxel (1μM). Bo80 cell line showed a similar effect with combination treatment
as Panc354 and MetPO1 (Fig. 5.3.3F).

In order to elucidate the underlying role of CSCs in the success of the above-
mentioned combination treatment, we used flow cytometry to quantify CD133+
CSCs in Panc354 and MetPO1 cells after day 12 for control, JM#21, gemcitabine
(1μM) and JM#21 + gem (1μM) treatments (Fig. 5.3.3G, H). We found that

73
 Results

gemcitabine treatment alone significantly increased the CD133+ population in

relapsed cells, which is well in line with previous data from our laboratory62.
Interestingly, JM#21 decreased CD133+ population, albeit less pronounced.
Combination treatment, however, led to a strongly reduced CD133+ population. In
summary, our data suggest that JM#21 can sensitize chemoresistant PDAC cells
and especially CSCs, and that a combination treatment of JM#21 and
chemotherapy could be beneficial for treatment.

74
 Results

Figure 5.3.3. JM#21 sensitizes PDAC cells towards chemotherapy: (A) Quantification of cell
viability and representative pictures for clonogenic assays after 48 hours of JM#21 treatment for
CXCL12 pretreated Panc354 cell line. (B) Experimental design for combination therapy analyzing
relapse using JM#21, gemcitabine (labelled as G) and paclitaxel (labelled as P). Quantification of
cell viability and representative pictures for clonogenic assays after treatment with JM#21 (10μM),
gemcitabine (Gem) or paclitaxel (Pac) for indicated concentrations as depicted in experimental
design in (C) Panc354, (D) MetPO1, (E) Gö13 and (F) Bo80 cells. (G) Flow cytometry for CD133 in
Panc354 and cells for the indicated treatments shown as fold change and (H) representative
cytometry plots for Panc354 cells. Error bars represent the standard deviation. n≥3 for all
experiments. *p < 0.05, ns = not significant.

5.3.4 Stabilization of JM#21 in serum conditions

As mentioned before, JM#21 was not stable in serum-containing culture conditions

(Fig. 5.3.1D, E). However, to exploit the therapeutic potential of this novel CXCR4
antagonistic peptide, two different techniques were utilized to increase its stability:
We either conjugated JM#21 to different types of palmitic acid (naming the resulting
compounds as JM#143 and JM#194239 (Fig. 5.3.4 A) or we encapsulated JM#21
within mesoporous silica nano particles with a radially (MSN) or dendritic (DMSN)
oriented pore system254,255 (Fig. 5.3.4 B). All these compounds (JM#143, JM#194,
MSN_JM#21 and DMSN_JM#21) were tested for their efficacy to hamper cell
migration towards CXCL12 in FBS-containing media. These migration assays
revealed that both JM#143 and JM#194 significantly decrease migration in a dose-
dependent manner across all the cell lines tested. Similarly, silica nanoparticle (Si-
NP) encapsulated JM#21 (MSN_JM#21 and DMSN_JM#21) showed abrogation of
migratory capacity of all cell lines towards CXCL12 in comparison to empty silica

75
 Results

nanoparticles (MSN_Empty and DMSN_Empty). In contrast, JM#21 alone failed to

have any influence on the migration capacity of the cell lines tested in the above-
mentioned FBS-containing media (Fig. 5.3.4C, D, E, F and G). The robust decrease
in migration potential confirmed the stability of JM#143, JM#194 and Si-NP
encapsulated JM#21 in serum conditions.

76
 Results

Figure 5.3.4. Stabilization of JM#21 in serum conditions: (A) Sequence and molecular weight of
fatty acid (palmitic) conjugated JM#21, namely JM#143 and JM#194 (B) TEM micrographs of silica
nanoparticles (MSN and DMSN). Migration assays towards CXCL12 in FBS-containing medium with
the indicated compounds, tested in (C) Panc354, (D) MetPO1, (E) Gö13 and (F) Bo80 cells. (G)
Representative micrographs (10x, DAPI staining) of transwell migration assays in MetPO1 cells for
the indicated conditions and concentrations. Error bars represent the standard deviation. n≥3 for all
experiments. *p < 0.05, ns = not significant.

5.3.5 Serum stable JM#21 reduces miCSCs in co-cultures with PSCs

We next wanted to delineate the capacity of serum stable JM#21 to target migrating
cancer stem cells. We have shown previously that co-cultures with PSCs increase
the miCSC subpopulation (Fig. 5.2.3C). We next investigated the effect of fatty acid
conjugated JM#21 (JM#143 and JM#194) and Si-NP encapsulated JM#21 pre-
treatment on PDX/O cell lines in co-culture with PSCs (Fig. 5.3.5A). In order to
characterize the effects, we performed pathway-focused gene expression analysis
for key genes involved in EMT and stemness. This indicated that JM#143 and
JM#194 significantly reduced expression of both stemness- (NANOG, ALDH1a1)
and EMT- (VIM, SNAI2) associated factors in Panc354 and MetPO1 cells.
Additionally, a substantial increase in CDH1 was also observed in both cell lines
after fatty acid conjugated JM#21 treatment (JM#143 and JM#194) (Fig. 5.3.5 B).
Genes involved in imparting chemoresistance exhibited a stark decrease in ABCC5
expression after JM#143 treatment. Furthermore, both JM#143 and JM#194
significantly upregulated the expression of DCK1 compared to vehicle-treated
control in both cell lines tested. Antioxidant enzymes (SOD1, GPX1, GPX2), on the

77
 Results

other hand, were found to be downregulated (Fig. 5.3.5 C). Flow cytometry analysis
revealed no changes in CD133+ expression after fatty acid conjugated JM#21
treatment in both Panc354 and MetPO1 cells, indicating no specific effect on the
general CSC population. In contrast, CD133+ CXCR4+ miCSCs and CXCR4+ cells
were practically abrogated (Fig. 5.3.5 D, H), indicating a strong effect on migratory
cells.

Correspondingly, treatment with Si-NP encapsulated JM#21 (MSN_JM#21 and

DMSN_JM#21) resulted in a substantial decrease in NANOG and BMI1 gene
expression. DMSN_JM#21 significantly increased CDH1 expression, whereas there
was a stark decrease in SNAI2 levels in both cell lines (Fig. 5.3.5 E), indicating an
induction of a more mesenchymal phenotype. A compelling decrease in CDA1 and
increase in DCK1 expression was observed for both cell lines treated with Si-NP
encapsulated JM#21 compared to empty Si-NP. A decrease in gene expression of
antioxidant enzymes (SOD1, GPX1, GPX2) was also observed (Fig. 5.3.5 F).
Similar to fatty acid conjugated JM#21, Si-NP encapsulated JM#21 showed no
significant change in CD133+ surface expression in flow cytometry analysis for both
cell lines. However, CD133+ CXCR4+ double expression and CXCR4+ expression
was found to be strongly downregulated (Fig. 5.3.5 G, I). In summary, we were able
to show that serum stable JM#21 (either by fatty acid conjugation or by
encapsulation in nanoparticles) abrogated the miCSC population in PDX/O – PSCs
co-culture.

78
 Results

Figure 5.3.5. Serum stable JM#21 reduces miCSCs in co-cultures with PSCs: (A) Experimental
design for treatment with fatty acid conjugated JM#21 and Si-NP encapsulated JM#21 in PDX/O –
PSCs co-culture. (B) Gene expression analysis for indicated cell lines with genes involved in EMT
and stemness after treatment with DMSO (vehicle control), JM#143 and JM#194. (C) Gene
expression analysis for indicated cell lines with genes involved in chemoresistance and antioxidant
signaling after treatment with DMSO, JM#143 and JM#194. (D) Flow cytometry analysis performed

79
 Results

for CD133+, CD133+CXCR4+ and CXCR4+ cells for depicted cell lines after treatment with DMSO,
JM#143 and JM#194. (E) Gene expression analysis for indicated cell lines with genes involved in
EMT and stemness after treatment with DMSN and MSN (Empty & JM#21). (F) Gene expression
analysis for indicated cell lines with genes involved in chemoresistance and antioxidant signaling
after treatment with DMSN and MSN (Empty & JM#21). (G) Flow cytometry analysis performed for
CD133+, CD133+CXCR4+ and CXCR4+ cells for depicted cell lines after treatment with DMSN and
MSN (Empty & JM#21). (H) Representative cytometry plots for MetPO1 cell line after treatment with
DMSO, JM#143 and JM#194 for CD133+, CD133+CXCR4+ and CXCR4+ cells. (I) Representative
cytometry plots for MetPO1 cells after treatment with DMSN and MSN (Empty & JM#21) for CD133+,
CD133+CXCR4+ and CXCR4+ cells. n≥3 for all experiments. *p < 0.05, ns = not significant.

80
 Discussion

6 Discussion
Pancreatic ductal adenocarcinoma (PDAC) is an extremely aggressive malignancy,
characterized by an unusually high metastatic burden already at the time of
diagnosis270. Furthermore, PDAC typically displays a high degree of resistance to
chemotherapy and radiation271. All this has been associated with the presence of
CSCs272. Considering that CSCs are essential for metastatic spread62, they might
be the population that gives rise to CTCs, and ultimately to metastasis. CTCs have
been demonstrated to have both tumor-initiating and EMT traits98. Therefore,
understanding the nuances of the regulatory mechanisms and the prerequisites for
the maintenance and expansion of migrating CSCs (and CTCs) will result in the
discovery and exploration of novel therapeutics, which in turn could result in
significantly improved therapy.

6.1 MEK signaling in CSCs, CTCs and metastasis

In spite of extensive efforts to improve therapy, the median survival in PDAC

patients is still far lower than desired, even with the (currently) most successful
therapies such as FOLFIRINOX (11.1 months)51 or gemcitabine+nab-paclitaxel (8.5
months), which have significantly improved survival in comparison with previous
standard therapies such as gemcitabine monotherapy52. In an effort to better
understand the regulation and significance of CSCs and miCSCs, we delineated the
effects of MEK inhibitors on stemness, migration, and circulating tumor cells using
primary cancer cells derived from genetically engineered mice that spontaneously
develop PDAC. The small molecule PD0325901 was used for MEK inhibition in
vitro, which was shown to compromise the growth and survival of the cells. This was
not surprising since MEK inhibition has been previously described to induce the
intrinsic apoptotic pathway in different types of cancer273-275. However, in our study
we observed the cells to be viable after treatment and we discovered no significant
differences with regard to apoptosis.

The activation of MAPK signaling components (RAS-RAF-MEK-ERK) can confer

stemness properties to cells240,276. We show that MEK inhibition functionally inhibits
CSC populations as confirmed by significant reduction in sphere formation capacity,
a surrogate marker for CSC activity. Additionally, we showed that CSCs seem to be

81
 Discussion

more MEK-dependent than the general cell population, since unspecific cytotoxicity
was not observed. This is further corroborated by the observation that stemness-
and pluripotency-associated markers are significantly downregulated upon MEK
inhibitor treatment. However, studies in colorectal cancer have shown that MEK
inhibition could lead to aberrant activation of Wnt/β-catenin signalling, which could
further modulate cancer stem cell homeostasis277. Additionally, signals from other
(non-cancer cells) cells within the tumor microenvironment can foster and potentiate
CSCs population278,279. It would therefore be interesting to further elucidate the
aspect that MEK inhibition has on other signaling pathways or in conjunction with
them in PDAC.

3D organoid cultures are a more physiological cell culture model than 2D monolayer
cultures that better reflects in vivo conditions by maintaining cell-to-cell signaling259.
Therefore, the investigation of MEK inhibition on primary organoid cultures is a very
promising way to contemplate treatment efficacy in a “physiological” in vitro
setting. We demonstrated that the capacity of primary murine PDAC tumor cells to
form organoids is hindered under MEK inhibition. Taken together with diminished
sphere forming capacity this suggests a decrease in their tumor-initiating potential.
We also observed similar results with the treatment performed in pre-formed
organoids, additionally validating our in vitro dataset.

The MAPK pathway has also been reported to promote the expression of EMT-
related transcription factors, in particular that of the Snail superfamily members
during development280,281, fibrosis282, and cancer progression and
metastasis283. The decrease in wound healing capacity of murine PDAC cells after
MEK inhibition was therefore in line with the current understanding of the role of the
MAPK pathway. Although we performed these experiments with matching cell
numbers after pretreatment, it still might be difficult to negate the effects of
proliferation on wound healing ability after scratch application. We therefore
performed 3D transwell migration assays towards serum-containing media and
found a strong dose-dependent (PD0325901-mediated) decrease in the migration
of murine PDAC cells. Additionally, MEK inhibitors can decrease tumor formation in
animal models, particularly in pancreatic cancer247. We demonstrated for the first
time that MEK inhibition additionally causes significant reduction in the number of

82
 Discussion

CSCs, organoids, and circulating tumor cells (CTCs) in vivo. Although, CTC
numbers do not naturally correlate directly with the metastatic load in patients, they
are indeed a strong indicator of prognosis284. Refametinib treatment significantly
247
reduced the size of primary tumors (previously published in Ref. No. ), the
observed effects on CTC numbers might (at least partially) also be as a result of a
general reduction in tumor size. However, the significant reduction of Slug-
expressing cells in vitro propose a strong inhibition of TGFβ-induced EMT by MEK
inhibition, which ultimately could result in the observed decrease in CTCs. This
offers a possible mechanism by which MEK can exert its function: promoting
survival, migration, stemness, and CTC initiation, and in turn supporting the
relevance of MEK signaling in PDAC.

6.2 Relevance of MEK inhibition in context of tumor

microenvironment

The primary tumor cell lines we used in this study are derived from a mouse model
which spontaneously develops metastatic PDAC (KPC mice), and was expanded
using an RFP reporter system to report cells of pancreatic origin, and a YFP reporter
system to indicate Slug activity. TGFβ promotes tumor progression in advanced
cancer stages by inducing tumor growth in itself. Most importantly it strongly induces
EMT, resulting in increased invasion and metastasis285 via upregulation of
transcription factors such as the zinc finger proteins Snail and Slug286. We therefore
utilized the RFP – YFP dual reporter system described above and TGFβ treatment
to delineate the role of MEK inhibition on an active EMT program. On the one hand
we found that increased mesenchymal differentiation (i.e., increased Slug activity)
after TGFβ pre-treatment was indeed diminished by MEK inhibition. A continuous
TGFβ treatment on the other hand displayed no effect on MEK inhibition. This
indicated that these two pathways, although capable of cooperating, act through
different downstream effectors or that compensatory feedback loop mechanisms
play a relevant role.

Taken together, these results also imply a possible interference of other pathways
(apart from TGFβ) involving cytokines and chemokines within the tumor
microenvironment which would hinder MEK inhibition efficacy. The TME is a
complex system containing several cell types such as immune cells, stellate cells,

83
 Discussion

and stromal fibroblasts as well as extracellular matrix, blood and lymphatic vessels,
all of which contribute to the growth and aggressiveness of a tumor287. The TME
provides necessary growth factors and signaling mediators like cytokines to the
tumor cells288. The extracellular matrix (ECM) acts as a scaffold and demonstrates
complex biochemical interactions inside the defined microenvironment. ECM
components constantly interact with integrin receptors from the cell surface which
forms a two-directional signaling regulated by GFs289. The concurrent response of
integrin–ECM–GF interactions inside the tumor microenvironment promotes the
activation as well as localization of RAS to the inner cell membrane290, promoting
resistance to MEK inhibitors. Studies have also shown that the TME is an important
root of resistance to MAPK pathway inhibitors by secretion of the macrophage-
derived mediator TNFα (Tumor necrosis factor alpha)291. Furthermore, inhibition of
MAPK interferes with the transactivation activity of p300/CREB (cAMP responsive
element binding protein 1)-binding protein (CBP) by HIF-1α and HIF-2α that
undergo oxygen-dependent degradation292. MAPK signaling hence promotes HIF-
1α activation which in turn is known to potentiate CSCs293 and metastasis in
PDAC294. These studies further strengthen the importance of TME components in
providing resistance towards MEK inhibitors.

Ligorio et al. recently described the role of stromal cancer-associated fibroblasts

(CAFs) in modulating the heterogeneity of PDAC model systems. Combining single-
cell RNA and protein analytics, they classified significant single-cell population shifts
toward invasive epithelial-to-mesenchymal transition (EMT) and proliferative (PRO)
phenotypes which were linked with mitogen-activated protein kinase (MAPK) and
signal transducer and activator of transcription 3 (STAT3) signaling104. Altogether,
these datasets depict MEK as an important therapeutic target due to its capacity to
interfere with complex molecular pathways289,295. Nevertheless, it is important to
emphasize the fact that other players involved in the progression of PDAC would
confer resistance using compensatory pathways. Therefore, using MEK inhibitors in
the current striving for precision medicine, a multi-targeted approach is required to
prevent resistance. Therefore, it will be crucial to precisely identify the factors and
mechanisms of resistance in order to further develop this line of therapy.

84
 Discussion

6.3 Crucial nature of CXCR4 – CXCL12 pathway in miCSCs

maintenance

PDAC tumors are heterogenous, comprising multiple cell populations. However,

within the majority of established cell lines used to model PDAC, there are no means
to determine how reflective these cultured cells are of the genetic features of the
tumor from which they were derived, and the extent to which they have diverged
after years in culture260. In recognition of these shortcomings of long-established
cell lines, we used PDX- (patient derived xenograft) or PDO- (patient derived
organoids) derived cell lines which are known to genetically recapitulate the primary
tumor and maintain genetic features of PDAC260).

The dissemination of tumor cells is the prerequisite of metastases and correlates

with a loss of epithelial differentiation and the acquisition of a migratory phenotype,
a hallmark of malignant tumor progression296. Migrating cancer stem cells (miCSCs)
play a pivotal role in malignant tumor formation and have been reported to form the
invasive front of the metastasis in different types of cancer, including PDAC101,209,297.
We define miCSC as a CXCR4 expressing subset of CSCs (i.e.,
CD133+CXCR4+)62 and using in silico analysis tools we show that both CSCs
(CD133+) and miCSCs (CD133+ CXCR4+) are overexpressed in PDAC. Longer
patient survival (disease free and overall survival analysis) correlated strongly with
low CD133 expression. However, no significant differences in patient survival were
observed when comparing high vs low CXCR4 expression and co-expression of
CD133 and CXCR4, although studies have previously shown that high CXCR4
expression may predict poor overall survival in resected PDAC patients209.
Additionally, in colon cancer and non-small cell lung carcinoma, Zhang et al. and Tu
et al. have respectively shown that CD133 and CXCR4 co-expression leads to poor
prognosis and CXCR4 drives CD133-induced EMT101,297. It is therefore imperative
to delineate the mechanisms for miCSC maintenance within the tumor
microenvironment in order to develop efficient therapeutic strategies.

CXCL12 (or SDF1) is critical for the homing of a variety of cell lineages during
development, tissue repair and metastasis298. All pancreatic tumor tissues express
CXCR4 and CXCL12 at higher expression levels compared to non-tumor tissue299.
CXCR4 is activated by CXCL12, which is abundantly present in the pancreatic tumor

85
 Discussion

microenvironment as well as at the typical sites of metastasis299. Within the tumor

microenvironment, activated pancreatic stellate cells (PSCs) are an important
source of CXCL12 secretion300. In our current study, PDX/PDO-derived cell lines
secrete only very low amounts of CXCL12, but quiescent PSCs secrete significantly
higher amounts of CXCL12. We observe a potentiated release of CXCL12 by PSCs
when co-cultured with PDX/PDO-derived cell lines (vs monoculture PSCs). Co-
cultured PSCs also display a dramatic decrease in lipid droplets, further confirming
the that co-culture conditions activates quiescent PSCs. Conditioned media from
activated PSCs can increase the expression of both CD133 and CXCR4, which
implies that secreted factors from PSCs can act to modulate CSCs and miCSCs.
This was further confirmed by increase in migratory activity and sphere forming
capabilities as well as upregulation of key EMT (VIM, SNAI1) and stemness
(NANOG, ALDH1a1) factors of PDX/PDO-derived cell lines when co-cultured with
PSCs.

PSCs are also known to play a role in PDAC chemoresistance through various
mechanisms, among which inducing hypoxic conditions is highly prevalent301-303.
Extensive fibrosis produced by the PSCs and the ECM results in significant
intratumoral hypoxia and a self-perpetuating hypoxia-fibrosis cycle303. PDX/O –
PSC (vs. PDX/O – PDX/O) co-culture elevated levels of ABCC5 expression, which
results in reduced drug uptake by the tumor cells. This is a somewhat controversial
result, since blocking ABCC5 has been shown to enhance effects of
chemotherapies like gemcitabine, but ABCC5 blocking has also been implicated in
a positive correlation to PDAC patient survival304. DCK1, encoding deoxycytidine
kinase, is downregulated, which further benefits the tumor cells against gemcitabine
as it catalyzes the reaction to convert gemcitabine to its active metabolite261.

Reactive oxygen species (ROS) are generally increased in pancreatic cancer cells
compared to normal cells. ROS play a vital role in the orchestration of the TME:
ROS can facilitate carcinogenesis and cancer progression with mild to moderate
elevation levels, while excessive ROS damage cancer cells dramatically and may
induce cell death305. The first line of defense against ROS-induced damage is the
conversion of superoxide to hydrogen peroxide by the antioxidant enzyme
superoxide dismutase (SOD)306. These hydrogen peroxide molecules are then

86
 Discussion

converted to water and oxygen by glutathione peroxidase (GPX)307. PDX/O – PSC

(vs PDX/O – PDX/O) co-culture show elevated SOD1 levels but decreased in GPX1
levels. Young et al. have shown in prostate cancer that SOD1 associates with
CXCR4 to regulate CXCR4 - CXCl12 signaling towards cell survival, apoptosis or
cell migration308. GPX1, however, plays more complex roles. It counteracts oxidative
stress and is pro-oncogenic in several malignancies. It also induces protective
autophagy in PDAC cells in response to glucose deprivation309. Additionally, low
GPX1 levels correlate with poor survival of PDAC patients treated with gemcitabine.
Due to these contradicting effects, the context-dependent role of GPX1 in PDAC
should be reviewed in more detail. Besides, studies show that glutathione
metabolism (and the genes involved e.g., GPXs, GSTs) is essential for self-renewal
and chemoresistance in CSCs83. Altogether, tumor cell – stellate cell crosstalk
promotes CSCs and miCSCs and their respective phenotypes.

We utilized sh-RNA mediated knockdown of CXCR4 to confirm that CXCL12 indeed

acts through CXCR4 in maintaining CSCs and miCSCs as well as their functional
characteristics. Flow cytometry revealed a significant decrease in CSCs (CD133+)
and miCSCs (CD133+ CXCR4+) after CXCR4 knockdown in control or CXCL12-
treated and PSC co-cultured PDX/PDO-derived cell lines. Additionally, cytidine
deaminase CDA1, which catalyzes the inactivation of gemcitabine to dFdU264, was
strongly downregulated after CXCR4 knockdown. Interestingly, low CDA mRNA
levels have been recently established as an independent prognostic
biomarker. When added to hENT1 (human equilibrative nucleoside transporter 1)
status, it may also provide treatment-specific predictive information and guide
towards either gemcitabine or 5FU therapy for the individual patient, supporting the
concept of a personalized treatment strategy310. CXCR4 is a GPCR protein receptor
which has multiple effector molecules spanning different pathways. Knockdown of
CXCR4 led to a decrease in both AKT and IκB-α phosphorylation. Previous studies
have shown that disruption of CXCR4 can lead to downregulation of NANOG by
interrupting the PI3K/AKT pathway, and CXCR4/CXCL12 can promote nuclear
accumulation of NF-κΒ by inducing the phosphorylation and destabilization of its
inhibitory protein, IκB-α189,311. Taken together, our results further strengthen the
importance of CXCR4 in modulating miCSCs. CXCR4 – CXCL12 signaling proved

87
 Discussion

to be an indispensable pathway to maintain miCSCs and its key physiological

features within PDAC and its microenvironment.

6.4 Diverging roles of BMI1 downstream of CXCR4 – CXCL12

signaling

BMI1 is a member of the polycomb-repressive complex 1 (PRC1) and it plays a

crucial role in the maintenance of chromatin silencing312,313. In neuronal,
haematopoietic and intestinal cells, BMI1 regulates in self-renewal through
repression of the INK4A–ARF locus in a polycomb repressor complex 2 (PRC2)
dependent manner314-317. Binding of PRC2 to its target genes allows EZH2 (a PRC2
subunit) to methylate Lysine 27 of histone H3 (H3K27)318,319. Trimethylated H3K27
(H3K27me3) is recognized by PRC1 and thus maintains the repression of target
genes320,321.

In head and neck squamous cell carcinoma (HNSCC), TWIST1 and BMI1 mutually
promote EMT by repressing both E-cadherin and p16INK4a322. Utilizing sh-RNA
mediated knockdown of BMI1, we showed that BMI1 and EMT factors are strongly
correlated. BMI1 knockdown causes significant downregulation of SNAI1 and
TWST1 mRNA, as well as N-CADHERIN protein. Functionally, migration assays
revealed a diminished migratory capacity of BMI1 knockdown cells compared to
scrambled controls. During EMT, tumor cells undergo a genetic and morphological
transformation, manifested by cell elongation, migration, and invasion, coordinated
by a reorganization of actin fibers in the cytoskeleton323. Immunofluorescent staining
in MetPO1 cells showed that BMI1 knockdown severely reduces these
morphological changes by reducing mesenchymal structures and actin stress fibers.
In intestinal tumors and HNSCC, BMI1 has been recognized as a cancer stem cell
marker324,325. Additionally, Dimri et al. have shown that BMI1 induces telomerase
activity and immortalizes human mammary epithelial cells326. Telomerase activity
has been shown in our laboratory to be a highly relevant factor in CSC
maintenance327. Although we observed no differences in sphere forming capacity
after BMI1 knockdown, there was a strong downregulation in key stemness-
associated genes (PROM1, ALDH1a1, POU5F1, NANOG, KLF4 and SOX2). BMI1
also had a pronounced effect on the CSC population, decreasing the expression of

88
 Discussion

CD133 as evidenced by flow cytometry. A possible explanation for the differences

in BMI1’s role on CSCs could be the association of BMI1 function to only a limited
number of stemness factors, which does not influence sphere forming capacity.
Furthermore, some pancreatic cancer stem cells have a slower cell cycle and
represent a quiescence-like stage89 and might not expand in sphere formation
assays328. This indicates a lower enrichment capability of sphere cultures than
expected, so that a potential effect might not have been observed experimentally.
Therefore, other methods for CSCs enrichment such as forced oxidative
phosphorylation89, which allows detection of quiescent CSCs could be applied to
confirm our observations on the role of BMI1 in CSC maintenance.

Paranjape et al. describe that BMI1 can modulate Nanog and affect both self-
renewal and EMT in breast cancer329. In recognition of the overlapping role of BMI1
in both CSC features and migration, deciphering the importance of BMI1 in miCSC
maintenance downstream of CXCR4-CXCL12 in such a highly metastatic disease
as PDAC is imperative. BMI1 knockdown led to a reduction of miCSCs in untreated
cell lines, and CXCL12 treatment was incapable of increasing the CSC or miCSC
population in BMI1 knockdown cells. However, the reverse effects of BMI1
knockdown on Panc354 miCSCs and EMT-associated genes could be due to an
aberrant CD133 upregulation after lentiviral transduction. CD133 overexpression
has been linked to SNAI2 upregulation330 and it would (at least partially) explain the
contradicting results in miCSC population, EMT-associated gene expression (VIM
and SNAI2), and the lack of changes in mesenchymal structures in Panc354 BMI1
knockdown cells. Nevertheless, BMI1 can potentiate PI3K/AKT and SHH/GLI1
signaling in a direct manner by induction of Nanog/NF-κB, and in an indirect manner
by hyperactivation of the PI3K/Akt/NF-κΒ axis331-333. In this context, we show
decreased ERK1/2 phosphorylation after BMI1 knockdown. Since we have
previously described the MEK/ERK axis to be highly relevant for both EMT and
stemness, this further underscores the role of BMI1 in maintaining (mi)CSCs.

From a therapeutic standpoint, Yin et al. have shown that BMI1 inhibition sensitizes
PDAC cells towards gemcitabine334. BMI1 is upregulated in PDX/PDO-derived cell
lines in PSC co-culture (with PSCs secreting high levels of CXCL12) or upon
treatment with CXCL12. On the other hand, CXCR4 knockdown cells show strongly

89
 Discussion

reduced BMI1 expression, which even CXCL12 or PSCs co-culture was not able to
upregulate. This establishes BMI1 as another possible regulator through which
CXCL12 exerts its effects on chemoresistance. In conclusion, we here demonstrate
that BMI1 is regulated by CXCL12 downstream CXCR4. CXCL12 conceivably
utilizes BMI1 to impart (at least partially) its distinct effects on EMT, stemness and
chemoresistance. BMI1 has diverse roles downstream of CXCR4 – CXCL12
signaling and represents a possible molecular link between EMT and stemness.

6.5 Endogenous human peptides as a novel agent to target CXCR4

CXCR4 and its primary chemokine ligand CXCL12 play a pivotal role not only in
pancreatic cancer development, progression and metastasis, but has been
implicated in almost every major malignancy185. Although a considerable number of
small molecule-peptide- or antibody-based CXCR4 antagonists are in preclinical
and clinical development at present335, the only clinically relevant CXCR4 antagonist
is Plerixafor (Mozobil, AMD3100), which unfortunately is not suitable for the
treatment of chronic disease because of extreme adverse effects336. Considering
the essential role of CXCR4 in cellular migration and metastasis, it is imperative to
discover further CXCR4 antagonists with less side effects in order to develop
therapies which can improve outcomes for (pancreatic) cancer patients.

EPI-X4 is a 16-mer peptide that is produced at low pH via proteolytic cleavage of

serum albumin by Cathepsin D and E234. In contrast to the inhibitor AMD3100, EPI-
X4 and its derivatives (e.g., JM#21) bind to CXCR4 and efficiently suppress CXCR4-
tropic HIV-1 infection235,237. We demonstrate for the first time the role of EPI-X4 and
its derivatives to target CXCR4 and to functionally eliminate migrating cancer (stem)
cells in PDAC. Using migration assays in FBS free media we establish JM#21 as
the most potent CXCR4 antagonist among four different EPI-X4 derivatives tested.
As little as 0.1μM of JM#21 caused significantly reduced migration towards the
CXCR4 ligand CXCL12. In Waldenström’s Macroglobulinemia, JM#21 inhibits ERK
and AKT phosphorylation as efficiently as AMD3100237. Since MAPK and AKT
signaling play a significant role in supporting both EMT and stemness, we utilized
immunofluorescence for actin filaments and sphere formation assays to show that
CXCR4 inhibition via JM#21. Even in the presence of CXCL12, it causes decreased
mesenchymal cytoskeletal changes and self-renewal capacity in PDAC cells. A

90
 Discussion

continued decrease in sphere forming capacity observed during second generation

spheres confirmed the inhibition of a CSC phenotype by JM#21 treatment. This was
further corroborated by gene and protein expression analysis for key EMT- and
cancer stemness-related factors.

A combination treatment using gemcitabine or paclitaxel together with JM#21

demonstrated the potential of JM#21 to sensitize PDAC cell towards chemotherapy,
which is especially beneficial for therapy resistant tumors such as PDAC.
Additionally, a combination treatment could allow the use of lower drug
concentrations, which in turn might lead to decreased drug toxicity and its side
effects337. An important limitation of JM#21, however, is its instability and short half-
life in serum234, with its rapid enzymatic degradation. Therefore, we tested serum
supplements and were able to show that EPI-X4 derivatives were stable under these
conditions. The further experiments such as clonogenic assays were therefore
performed using serum supplements rather than FBS. However, these conditions
might have an effect on PDAC cell capacity for relapse338 and therefore require
caution.

Gemcitabine treatment resulted in expansion of the CD133+ cancer stem cell

population, which is consistent with results from previous studies from our
group62,246. An increase in CD133+CXCR4+ miCSCs cells was also observed upon
gemcitabine treatment, which affects differentiated cells much more strongly than
CSCs82. We have previously shown the effectiveness of combination therapies
targeting specific stemness-associated pathways together with
chemotherapy82,135,246,339,340. In the present study, we aimed to establish a similar
combination therapy primarily targeting migration. Interestingly, we discovered the
ability of JM#21 to reduce CSC burden in addition to blocking CXCR4-mediated
migration and established that JM#21 inhibits CXCL12-induced chemoresistance in
order to make conventional therapies (e.g., gemcitabine) suitable. Altogether,
JM#21 appears to be a highly specific and highly active novel CXCR4 antagonist,
which could potentially improve therapy for pancreatic ductal adenocarcinomas.

91
 Discussion

6.6 Therapeutic efficacy of JM#21

While EPI-X4 has an extremely short half-life in serum, JM#21 is already

considerably more stable. However, in order to extend its half-life to fully establish
its therapeutic effects, we modified the compound itself or its delivery method. For
this purpose we utilized palmitic (fatty) acid conjugation to improve stability and silica
nanoparticle encapsulation to improve delivery. Both methods showed highly
increased stability of JM#21 even in serum-containing culture conditions as
evidenced by decreased migratory potential of all PDX/PDO cell lines tested in FBS
containing media.

In this study we describe the important role of PSCs in promoting migratory

potential, stemness and chemoresistance in PDAC cells. We also establish the role
of CXCR4 – CXCL12 signaling in maintaining CSCs and miCSCs. Serum-stable
JM#21 and its derivatives JM#143, JM#194, DMSN_JM#21 and MSN_JM#21
(either conjugated to fatty acids or encapsulated in nanoparticles) decreased EMT-
and stemness-related gene signatures in primary PDAC cells when co-cultured with
PSCs. Even after 72 hours of co-culture, serum-stable JM#21 blocked CXCR4 –
CXCL12 signaling, prompting stark upregulation of DCK1 and downregulation of
CDA1. Additionally, our experiments using shRNA mediated CXCR4 knockdown
showed a diminished CSC population with CXCL12 treatment and PSCs co-culture,
whereas serum stable JM#21 did not reduce CSCs in Panc354 or MetPO1 cells
after co-culture with PSCs. CXCR4 harbors AU-rich elements (ARE) in the 3’ UTR
region that bind and respond to RNA-binding proteins, tristetraprolin (TTP/ZFP36)
and HuR (ELAVL1)341. Latorre et. al., showed that long noncoding RNA MALAT1
and the RNA-binding protein HuR (ELAVL1) bind the CD133 promoter region to
regulate CD133 expression342. Possibly, shRNA-mediated knockdown of CXCR4
could also disrupt such posttranslational modifications, hindering CD133
expression, which might not be the case for serum stable JM#21 in PDX/O – PSC
co-culture. Furthermore, PSCs are known to secrete various cytokines in addition
to CXCL12 (e.g.,TGFβ), matrix metalloproteinase (e.g., MMP2) and interleukins
(e.g., IL1, IL8), which may regulate CSCs irrespective of CXCR4343-345).

92
 Discussion

Nevertheless, the miCSC population was completely abrogated by JM#21

treatment, further substantiating JM#21’s strong therapeutic effect in targeting
CXCR4. Moreover, developing a combination treatment for in vivo therapy with such
potent and stable CXCR4 modulators is of utmost importance and further research
in this regard is indispensable. In summary, we show for the first time that serum-
stable endogenous peptides such as JM#21 represent a potent novel therapeutic
entity, which robustly targets CXCR4 to reduce migration and eliminate miCSCs in
PDAC.

93
 Summary

7 Summary

Metastasis is the leading cause of cancer-related death in pancreatic ductal

adenocarcinomas (PDAC). All hallmark mutations in PDAC (KRAS, p16INK4A, TP53
and SMAD4) are known to modulate the metastatic process. The complex
metastatic cascade can be divided into three major phases, (i) physical translocation
of cancer cells from its primary tumor location to the microenvironment of a
secondary site, (ii) colonization and (iii) ultimately growth into a metastatic lesion. In
this regard, circulating tumor cells (CTCs) represent an intermediate stage of
metastasis, a subset of which require stem-like properties for tumor-initiation and
growth at the secondary site. It is therefore essential to delineate the pathways that
maintain these CTCs and miCSCs.

MEK is downstream of the oncogenic driver mutation KRAS and has been reported
to show outstanding relevance in PDAC biology. Thus, our first aim was to establish
the role of RAS/MEK/ERK signaling to promote EMT and support stem-like
phenotypes in disseminating cells. Using primary murine cancer cell lines with a
Slug – YFP reporter system and MEK inhibitors, we established the contribution of
MEK signaling on survival, migration and self-renewal capacity in PDAC. In brief,
TGFβ-induced EMT can be abrogated by MEK inhibition, evidenced by strong
decrease of Slug-expressing cells, which translates into a significant reduction in
CTCs in mice.

Furthermore, RAS/MEK signaling does not act in an isolated manner, as secreted

factors from within the tumor microenvironment, independently as well as in
combination with other secreted factors, promote metastasis. CXCL12 is one such
chemokine and exerts its effects primarily via CXCR4. Besides, a subset of cancer
stem cells (CSCs) called migrating cancer stem cells (miCSCs) with CXCR4 and
CD133 receptors can be detected in abundance within the invasive front of PDAC
and are exclusively responsible for metastasis. Therefore, our next aim was to
describe the relevance of CXCL12/CXCR4 signaling for the maintenance of CSCs
and miCSCs in context of the microenvironment. Using co-cultures of PDAC cells
and pancreatic stellate cells (PSCs), we demonstrate the importance of tumor –
stroma crosstalk on CSC and miCSC maintenance. PSCs in co-culture with PDAC

94
 Summary

cells secrete elevated levels of CXCL12, enabling tumor cells to self-renew and to
migrate, and ultimately to metastasize. Mechanistically, CXCL12 acts through
CXCR4 to activate multiple pathways that include sonic hedgehog, AKT and NF-κΒ,
which further promotes NANOG and BMI1 expression. In addition, BMI1 plays the
molecular link between inducing stemness and activating different EMT programs.
We established CXCR4 as a prominent target for therapeutic interventions to reduce
CSCs and abrogate miCSCs.

It is imperative to develop and investigate novel approaches to combat metastasis.

Having successfully established the crucial role of CXCR4, our final aim was to
exploit human endogenous peptides to eliminate miCSCs. We showed JM#21, an
EPI-X4 peptide derivative, to be a highly potent inhibitor of CXCR4, even in the
presence of CXCL12. JM#21 strongly reduced sphere formation capacity and
sensitized PDAC cells towards chemotherapy. Aiming at a potential therapy, we
further modified JM#21 by conjugation with fatty acids to improve stability, or
encapsulated JM#21 in silica nanoparticles for improved delivery. These modified
compounds drastically reduced chemoresistance and stemness features in PDAC
cells co-cultured with PSCs. In addition, they also decreased key factors responsible
for EMT and eliminated the miCSC population.

In conclusion, the data in the present study (i) demonstrate the importance of
RAS/MEK/ERK signaling on CTC initiation, (ii) provide novel molecular insights on
CSC and miCSC maintenance and (iii) explore human endogenous peptides as
innovative therapy for targeting miCSCs and migration / metastasis. In addition,
tumor-educated PSCs were shown to secrete elevated levels of CXCL12, which
promotes EMT, stemness and chemoresistance via CXCR4. CXCL12 was also
revealed to activate BMI1, which we describe as a downstream molecular link
between EMT and stemness.

95
 List of references

8 List of references

1 Walter, K. et al. MEK Inhibition Targets Cancer Stem Cells and Impedes
Migration of Pancreatic Cancer Cells In Vitro and In Vivo. Stem Cells Int
2019, 8475389, doi:10.1155/2019/8475389 (2019).
2 Board, P. D. Q. A. T. E. in PDQ Cancer Information Summaries (National
Cancer Institute (US), 2002).
3 Kleeff, J. et al. Pancreatic cancer. Nat Rev Dis Primers 2, 16022,
doi:10.1038/nrdp.2016.22 (2016).
4 Owens, D. K. et al. Screening for Pancreatic Cancer: US Preventive Services
Task Force Reaffirmation Recommendation Statement. Jama 322, 438-444,
doi:10.1001/jama.2019.10232 (2019).
5 Luo, W., Tao, J., Zheng, L. & Zhang, T. Current epidemiology of pancreatic
cancer: Challenges and opportunities. Chin J Cancer Res 32, 705-719,
doi:10.21147/j.issn.1000-9604.2020.06.04 (2020).
6 Rahib, L. et al. Projecting cancer incidence and deaths to 2030: the
unexpected burden of thyroid, liver, and pancreas cancers in the United
States. Cancer Res 74, 2913-2921, doi:10.1158/0008-5472.Can-14-0155
(2014).
7 Hu, J. X. et al. Pancreatic cancer: A review of epidemiology, trend, and risk
factors. World J Gastroenterol 27, 4298-4321, doi:10.3748/wjg.v27.i27.4298
(2021).
8 Aier, I., Semwal, R., Sharma, A. & Varadwaj, P. K. A systematic assessment
of statistics, risk factors, and underlying features involved in pancreatic
cancer. Cancer Epidemiol 58, 104-110, doi:10.1016/j.canep.2018.12.001
(2019).
9 Sperti, C., Pasquali, C., Piccoli, A. & Pedrazzoli, S. Recurrence after
resection for ductal adenocarcinoma of the pancreas. World J Surg 21, 195-
200, doi:10.1007/s002689900215 (1997).
10 Thomas, S. K., Lee, J. & Beatty, G. L. Paracrine and cell autonomous
signalling in pancreatic cancer progression and metastasis. EBioMedicine
53, 102662, doi:10.1016/j.ebiom.2020.102662 (2020).
11 Siegel, R. L., Miller, K. D. & Jemal, A. Cancer statistics, 2019. CA Cancer J
Clin 69, 7-34, doi:10.3322/caac.21551 (2019).

96
 List of references

12 Grimont, A., Leach, S. D. & Chandwani, R. Uncertain Beginnings: Acinar and

Ductal Cell Plasticity in the Development of Pancreatic Cancer. Cell Mol
Gastroenterol Hepatol 13, 369-382, doi:10.1016/j.jcmgh.2021.07.014 (2022).
13 Lee, A. Y. L. et al. Cell of origin affects tumour development and phenotype
in pancreatic ductal adenocarcinoma. Gut 68, 487-498, doi:10.1136/gutjnl-
2017-314426 (2019).
14 Puri, S., Folias, A. E. & Hebrok, M. Plasticity and dedifferentiation within the
pancreas: development, homeostasis, and disease. Cell Stem Cell 16, 18-
31, doi:10.1016/j.stem.2014.11.001 (2015).
15 Kopp, J. L. et al. Identification of Sox9-dependent acinar-to-ductal
reprogramming as the principal mechanism for initiation of pancreatic ductal
adenocarcinoma. Cancer Cell 22, 737-750, doi:10.1016/j.ccr.2012.10.025
(2012).
16 Orth, M. et al. Pancreatic ductal adenocarcinoma: biological hallmarks,
current status, and future perspectives of combined modality treatment
approaches. Radiat Oncol 14, 141, doi:10.1186/s13014-019-1345-6 (2019).
17 Del Poggetto, E. et al. Epithelial memory of inflammation limits tissue damage
while promoting pancreatic tumorigenesis. Science 373, eabj0486,
doi:10.1126/science.abj0486 (2021).
18 Ryan, D. P., Hong, T. S. & Bardeesy, N. Pancreatic adenocarcinoma. N Engl
J Med 371, 1039-1049, doi:10.1056/NEJMra1404198 (2014).
19 Cooper, C. L., O'Toole, S. A. & Kench, J. G. Classification, morphology and
molecular pathology of premalignant lesions of the pancreas. Pathology 45,
286-304, doi:10.1097/PAT.0b013e32835f2205 (2013).
20 Hruban, R. H. et al. Pancreatic intraepithelial neoplasia: a new nomenclature
and classification system for pancreatic duct lesions. Am J Surg Pathol 25,
579-586, doi:10.1097/00000478-200105000-00003 (2001).
21 Distler, M., Aust, D., Weitz, J., Pilarsky, C. & Grützmann, R. Precursor lesions
for sporadic pancreatic cancer: PanIN, IPMN, and MCN. Biomed Res Int
2014, 474905, doi:10.1155/2014/474905 (2014).
22 Hruban, R. H. et al. An illustrated consensus on the classification of
pancreatic intraepithelial neoplasia and intraductal papillary mucinous
neoplasms. Am J Surg Pathol 28, 977-987,
doi:10.1097/01.pas.0000126675.59108.80 (2004).

97
 List of references

23 Löhr, M., Klöppel, G., Maisonneuve, P., Lowenfels, A. B. & Lüttges, J.

Frequency of K-ras mutations in pancreatic intraductal neoplasias associated
with pancreatic ductal adenocarcinoma and chronic pancreatitis: a meta-
analysis. Neoplasia 7, 17-23, doi:10.1593/neo.04445 (2005).
24 Morris, J. P. t., Wang, S. C. & Hebrok, M. KRAS, Hedgehog, Wnt and the
twisted developmental biology of pancreatic ductal adenocarcinoma. Nat Rev
Cancer 10, 683-695, doi:10.1038/nrc2899 (2010).
25 Iguchi, E., Safgren, S. L., Marks, D. L., Olson, R. L. & Fernandez-Zapico, M.
E. Pancreatic Cancer, A Mis-interpreter of the Epigenetic Language. Yale J
Biol Med 89, 575-590 (2016).
26 Roe, J. S. et al. Enhancer Reprogramming Promotes Pancreatic Cancer
Metastasis. Cell 170, 875-888.e820, doi:10.1016/j.cell.2017.07.007 (2017).
27 McDonald, O. G. et al. Epigenomic reprogramming during pancreatic cancer
progression links anabolic glucose metabolism to distant metastasis. Nat
Genet 49, 367-376, doi:10.1038/ng.3753 (2017).
28 Sung, H. et al. Global Cancer Statistics 2020: GLOBOCAN Estimates of
Incidence and Mortality Worldwide for 36 Cancers in 185 Countries. CA
Cancer J Clin 71, 209-249, doi:10.3322/caac.21660 (2021).
29 Collins, M. A. et al. Oncogenic Kras is required for both the initiation and
maintenance of pancreatic cancer in mice. J Clin Invest 122, 639-653,
doi:10.1172/jci59227 (2012).
30 Wang, S. et al. The molecular biology of pancreatic adenocarcinoma:
translational challenges and clinical perspectives. Signal Transduct Target
Ther 6, 249, doi:10.1038/s41392-021-00659-4 (2021).
31 Tago, K. et al. Oncogenic Ras mutant causes the hyperactivation of NF-κB
via acceleration of its transcriptional activation. Mol Oncol 13, 2493-2510,
doi:10.1002/1878-0261.12580 (2019).
32 Loncle, C. et al. IL17 Functions through the Novel REG3β-JAK2-STAT3
Inflammatory Pathway to Promote the Transition from Chronic Pancreatitis
to Pancreatic Cancer. Cancer Res 75, 4852-4862, doi:10.1158/0008-
5472.Can-15-0896 (2015).
33 Baumgart, S. et al. GSK-3β Governs Inflammation-Induced NFATc2
Signaling Hubs to Promote Pancreatic Cancer Progression. Mol Cancer Ther
15, 491-502, doi:10.1158/1535-7163.Mct-15-0309 (2016).

98
 List of references

34 Fleming, J. B., Shen, G. L., Holloway, S. E., Davis, M. & Brekken, R. A.

Molecular consequences of silencing mutant K-ras in pancreatic cancer cells:
justification for K-ras-directed therapy. Mol Cancer Res 3, 413-423,
doi:10.1158/1541-7786.Mcr-04-0206 (2005).
35 Muzumdar, M. D. et al. Survival of pancreatic cancer cells lacking KRAS
function. Nat Commun 8, 1090, doi:10.1038/s41467-017-00942-5 (2017).
36 Arner, E. N., Du, W. & Brekken, R. A. Behind the Wheel of Epithelial Plasticity
in KRAS-Driven Cancers. Front Oncol 9, 1049, doi:10.3389/fonc.2019.01049
(2019).
37 Singh, A. et al. A gene expression signature associated with ''K-Ras
addiction'' reveals regulators of EMT and tumor cell survival. Cancer Cell 39,
441-442, doi:10.1016/j.ccell.2021.02.009 (2021).
38 Shao, D. D. et al. KRAS and YAP1 converge to regulate EMT and tumor
survival. Cell 158, 171-184, doi:10.1016/j.cell.2014.06.004 (2014).
39 Tuveson, D. A. & Neoptolemos, J. P. Understanding metastasis in pancreatic
cancer: a call for new clinical approaches. Cell 148, 21-23,
doi:10.1016/j.cell.2011.12.021 (2012).
40 Mazur, P. K. & Siveke, J. T. Genetically engineered mouse models of
pancreatic cancer: unravelling tumour biology and progressing translational
oncology. Gut 61, 1488-1500, doi:10.1136/gutjnl-2011-300756 (2012).
41 Muller, P. A. et al. Mutant p53 drives invasion by promoting integrin recycling.
Cell 139, 1327-1341, doi:10.1016/j.cell.2009.11.026 (2009).
42 Qiu, W. et al. Disruption of p16 and activation of Kras in pancreas increase
ductal adenocarcinoma formation and metastasis in vivo. Oncotarget 2, 862-
873, doi:10.18632/oncotarget.357 (2011).
43 Iacobuzio-Donahue, C. A. et al. DPC4 gene status of the primary carcinoma
correlates with patterns of failure in patients with pancreatic cancer. J Clin
Oncol 27, 1806-1813, doi:10.1200/JCO.2008.17.7188 (2009).
44 Ye, X. & Weinberg, R. A. Epithelial-Mesenchymal Plasticity: A Central
Regulator of Cancer Progression. Trends Cell Biol 25, 675-686,
doi:10.1016/j.tcb.2015.07.012 (2015).
45 Reichert, M. et al. Regulation of Epithelial Plasticity Determines Metastatic
Organotropism in Pancreatic Cancer. Dev Cell 45, 696-711.e698,
doi:10.1016/j.devcel.2018.05.025 (2018).

99
 List of references

46 Rhim, A. D. et al. EMT and dissemination precede pancreatic tumor

formation. Cell 148, 349-361, doi:10.1016/j.cell.2011.11.025 (2012).
47 Dongre, A. & Weinberg, R. A. New insights into the mechanisms of epithelial-
mesenchymal transition and implications for cancer. Nat Rev Mol Cell Biol
20, 69-84, doi:10.1038/s41580-018-0080-4 (2019).
48 Kretzschmar, C. S., Hamilton, J. W., Wissler, D. W., Yale, C. E. & Morrissey,
J. F. Balloon dilation for the treatment of stomal stenosis complicating gastric
surgery for morbid obesity. Surgery 102, 443-446 (1987).
49 Nigam, S. K. et al. Localization of the FK506-binding protein, FKBP 13, to the
lumen of the endoplasmic reticulum. Biochem J 294 ( Pt 2), 511-515,
doi:10.1042/bj2940511 (1993).
50 Teague, A., Lim, K. H. & Wang-Gillam, A. Advanced pancreatic
adenocarcinoma: a review of current treatment strategies and developing
therapies. Ther Adv Med Oncol 7, 68-84, doi:10.1177/1758834014564775
(2015).
51 Conroy, T. et al. FOLFIRINOX versus gemcitabine for metastatic pancreatic
cancer. N Engl J Med 364, 1817-1825, doi:10.1056/NEJMoa1011923 (2011).
52 Von Hoff, D. D. et al. Increased survival in pancreatic cancer with nab-
paclitaxel plus gemcitabine. N Engl J Med 369, 1691-1703,
doi:10.1056/NEJMoa1304369 (2013).
53 Gourgou-Bourgade, S. et al. Impact of FOLFIRINOX compared with
gemcitabine on quality of life in patients with metastatic pancreatic cancer:
results from the PRODIGE 4/ACCORD 11 randomized trial. J Clin Oncol 31,
23-29, doi:10.1200/JCO.2012.44.4869 (2013).
54 Le, D. T. et al. Mismatch repair deficiency predicts response of solid tumors
to PD-1 blockade. Science 357, 409-413, doi:10.1126/science.aan6733
(2017).
55 Lemery, S., Keegan, P. & Pazdur, R. First FDA Approval Agnostic of Cancer
Site - When a Biomarker Defines the Indication. N Engl J Med 377, 1409-
1412, doi:10.1056/NEJMp1709968 (2017).
56 Brahmer, J. R. et al. Safety and activity of anti-PD-L1 antibody in patients
with advanced cancer. N Engl J Med 366, 2455-2465,
doi:10.1056/NEJMoa1200694 (2012).

100
 List of references

57 Royal, R. E. et al. Phase 2 trial of single agent Ipilimumab (anti-CTLA-4) for

locally advanced or metastatic pancreatic adenocarcinoma. J Immunother
33, 828-833, doi:10.1097/CJI.0b013e3181eec14c (2010).
58 Waddell, N. et al. Whole genomes redefine the mutational landscape of
pancreatic cancer. Nature 518, 495-501, doi:10.1038/nature14169 (2015).
59 Quinonero, F. et al. The challenge of drug resistance in pancreatic ductal
adenocarcinoma: a current overview. Cancer Biol Med 16, 688-699,
doi:10.20892/j.issn.2095-3941.2019.0252 (2019).
60 Li, C. et al. Identification of pancreatic cancer stem cells. Cancer Res 67,
1030-1037, doi:10.1158/0008-5472.CAN-06-2030 (2007).
61 Hermann, P. C., Bhaskar, S., Cioffi, M. & Heeschen, C. Cancer stem cells in
solid tumors. Semin Cancer Biol 20, 77-84,
doi:10.1016/j.semcancer.2010.03.004 (2010).
62 Hermann, P. C. et al. Distinct populations of cancer stem cells determine
tumor growth and metastatic activity in human pancreatic cancer. Cell Stem
Cell 1, 313-323, doi:10.1016/j.stem.2007.06.002 (2007).
63 Valle, S., Martin-Hijano, L., Alcalá, S., Alonso-Nocelo, M. & Sainz, B., Jr. The
Ever-Evolving Concept of the Cancer Stem Cell in Pancreatic Cancer.
Cancers (Basel) 10, doi:10.3390/cancers10020033 (2018).
64 Vermeulen, L., Sprick, M. R., Kemper, K., Stassi, G. & Medema, J. P. Cancer
stem cells--old concepts, new insights. Cell Death Differ 15, 947-958,
doi:10.1038/cdd.2008.20 (2008).
65 Nguyen, P. H. et al. Characterization of Biomarkers of Tumorigenic and
Chemoresistant Cancer Stem Cells in Human Gastric Carcinoma. Clin
Cancer Res 23, 1586-1597, doi:10.1158/1078-0432.Ccr-15-2157 (2017).
66 Rasheed, Z. A. et al. Prognostic significance of tumorigenic cells with
mesenchymal features in pancreatic adenocarcinoma. J Natl Cancer Inst
102, 340-351, doi:10.1093/jnci/djp535 (2010).
67 Miranda-Lorenzo, I. et al. Intracellular autofluorescence: a biomarker for
epithelial cancer stem cells. Nat Methods 11, 1161-1169,
doi:10.1038/nmeth.3112 (2014).
68 Li, C. et al. c-Met is a marker of pancreatic cancer stem cells and therapeutic
target. Gastroenterology 141, 2218-2227.e2215,
doi:10.1053/j.gastro.2011.08.009 (2011).

101
 List of references

69 Alcalá, S., Martinelli, P., Hermann, P. C., Heeschen, C. & Sainz, B., Jr. The
Anthrax Toxin Receptor 1 (ANTXR1) Is Enriched in Pancreatic Cancer Stem
Cells Derived from Primary Tumor Cultures. Stem Cells Int 2019, 1378639,
doi:10.1155/2019/1378639 (2019).
70 Boivin, D. et al. The stem cell marker CD133 (prominin-1) is phosphorylated
on cytoplasmic tyrosine-828 and tyrosine-852 by Src and Fyn tyrosine
kinases. Biochemistry 48, 3998-4007, doi:10.1021/bi900159d (2009).
71 Taïeb, N. et al. The first extracellular domain of the tumour stem cell marker
CD133 contains an antigenic ganglioside-binding motif. Cancer Lett 278,
164-173, doi:10.1016/j.canlet.2009.01.013 (2009).
72 Roper, K., Corbeil, D. & Huttner, W. B. Retention of prominin in microvilli
reveals distinct cholesterol-based lipid micro-domains in the apical plasma
membrane. Nat Cell Biol 2, 582-592, doi:10.1038/35023524 (2000).
73 Giebel, B. et al. Segregation of lipid raft markers including CD133 in polarized
human hematopoietic stem and progenitor cells. Blood 104, 2332-2338,
doi:10.1182/blood-2004-02-0511 (2004).
74 Ohnishi, S. et al. hypoxia-inducible factors activate CD133 promoter through
ETS family transcription factors. PLoS One 8, e66255,
doi:10.1371/journal.pone.0066255 (2013).
75 Jiang, X. et al. Inactivating mutations of RNF43 confer Wnt dependency in
pancreatic ductal adenocarcinoma. Proc. Natl. Acad. Sci. U. S. A. 110,
12649-12654, doi:10.1073/pnas.1307218110 (2013).
76 Zhan, T., Rindtorff, N. & Boutros, M. Wnt signaling in cancer. Oncogene 36,
1461-1473, doi:10.1038/onc.2016.304 (2017).
77 Nagathihalli, N. S. et al. Inverse Correlation of STAT3 and MEK Signaling
Mediates Resistance to RAS Pathway Inhibition in Pancreatic Cancer.
Cancer Res 78, 6235-6246, doi:10.1158/0008-5472.Can-18-0634 (2018).
78 Xin, B. et al. Nerve growth factor regulates CD133 function to promote tumor
cell migration and invasion via activating ERK1/2 signaling in pancreatic
cancer. Pancreatology 16, 1005-1014, doi:10.1016/j.pan.2016.09.005
(2016).
79 Chen, Y. S. et al. CD133/Src axis mediates tumor initiating property and
epithelial-mesenchymal transition of head and neck cancer. PLoS One 6,
e28053, doi:10.1371/journal.pone.0028053 (2011).

102
 List of references

80 Weng, C. C. et al. Pancreatic Tumor Progression Associated With CD133

Overexpression: Involvement of Increased TERT Expression and Epidermal
Growth Factor Receptor-Dependent Akt Activation. Pancreas 45, 443-457,
doi:10.1097/mpa.0000000000000460 (2016).
81 Hou, Y. C., Chao, Y. J., Tung, H. L., Wang, H. C. & Shan, Y. S. Coexpression
of CD44-positive/CD133-positive cancer stem cells and CD204-positive
tumor-associated macrophages is a predictor of survival in pancreatic ductal
adenocarcinoma. Cancer 120, 2766-2777, doi:10.1002/cncr.28774 (2014).
82 Mueller, M. T. et al. Combined targeted treatment to eliminate tumorigenic
cancer stem cells in human pancreatic cancer. Gastroenterology 137, 1102-
1113, doi:10.1053/j.gastro.2009.05.053 (2009).
83 Jagust, P., Alcalá, S., Sainz Jr, B., Heeschen, C. & Sancho, P. Glutathione
metabolism is essential for self-renewal and chemoresistance of pancreatic
cancer stem cells. World J Stem Cells 12, 1410-1428,
doi:10.4252/wjsc.v12.i11.1410 (2020).
84 Visvader, J. E. & Lindeman, G. J. Cancer stem cells in solid tumours:
accumulating evidence and unresolved questions. Nat Rev Cancer 8, 755-
768, doi:10.1038/nrc2499 (2008).
85 Goodell, M. A., Brose, K., Paradis, G., Conner, A. S. & Mulligan, R. C.
Isolation and functional properties of murine hematopoietic stem cells that
are replicating in vivo. J Exp Med 183, 1797-1806,
doi:10.1084/jem.183.4.1797 (1996).
86 Kim, E. J. et al. NRF2 Knockdown Resensitizes 5-Fluorouracil-Resistant
Pancreatic Cancer Cells by Suppressing HO-1 and ABCG2 Expression. Int J
Mol Sci 21, doi:10.3390/ijms21134646 (2020).
87 Duong, H. Q. et al. Aldehyde dehydrogenase 1A1 confers intrinsic and
acquired resistance to gemcitabine in human pancreatic adenocarcinoma
MIA PaCa-2 cells. Int J Oncol 41, 855-861, doi:10.3892/ijo.2012.1516 (2012).
88 Yang, M. C. et al. Blockade of autophagy reduces pancreatic cancer stem
cell activity and potentiates the tumoricidal effect of gemcitabine. Mol Cancer
14, 179, doi:10.1186/s12943-015-0449-3 (2015).
89 Valle, S. et al. Exploiting oxidative phosphorylation to promote the stem and
immunoevasive properties of pancreatic cancer stem cells. Nat Commun 11,
5265, doi:10.1038/s41467-020-18954-z (2020).

103
 List of references

90 Courtois, S. et al. Inhibition of Mitochondrial Dynamics Preferentially Targets

Pancreatic Cancer Cells with Enhanced Tumorigenic and Invasive Potential.
Cancers (Basel) 13, doi:10.3390/cancers13040698 (2021).
91 Mani, S. A. et al. The epithelial-mesenchymal transition generates cells with
properties of stem cells. Cell 133, 704-715, doi:10.1016/j.cell.2008.03.027
(2008).
92 Pastushenko, I. et al. Identification of the tumour transition states occurring
during EMT. Nature 556, 463-468, doi:10.1038/s41586-018-0040-3 (2018).
93 Yang, J. et al. Guidelines and definitions for research on epithelial-
mesenchymal transition. Nat Rev Mol Cell Biol 21, 341-352,
doi:10.1038/s41580-020-0237-9 (2020).
94 Yu, M. et al. Circulating breast tumor cells exhibit dynamic changes in
epithelial and mesenchymal composition. Science 339, 580-584,
doi:10.1126/science.1228522 (2013).
95 Schliekelman, M. J. et al. Molecular portraits of epithelial, mesenchymal, and
hybrid States in lung adenocarcinoma and their relevance to survival. Cancer
Res 75, 1789-1800, doi:10.1158/0008-5472.Can-14-2535 (2015).
96 Grillet, F. et al. Circulating tumour cells from patients with colorectal cancer
have cancer stem cell hallmarks in ex vivo culture. Gut 66, 1802-1810,
doi:10.1136/gutjnl-2016-311447 (2017).
97 Aktas, B. et al. Stem cell and epithelial-mesenchymal transition markers are
frequently overexpressed in circulating tumor cells of metastatic breast
cancer patients. Breast Cancer Res 11, R46, doi:10.1186/bcr2333 (2009).
98 Poruk, K. E. et al. Circulating Tumor Cells Expressing Markers of Tumor-
Initiating Cells Predict Poor Survival and Cancer Recurrence in Patients with
Pancreatic Ductal Adenocarcinoma. Clin Cancer Res 23, 2681-2690,
doi:10.1158/1078-0432.Ccr-16-1467 (2017).
99 Zhang, S. S. et al. CD133(+)CXCR4(+) colon cancer cells exhibit metastatic
potential and predict poor prognosis of patients. BMC Med 10, 85,
doi:10.1186/1741-7015-10-85 (2012).
100 Cioffi, M. et al. Identification of a distinct population of CD133(+)CXCR4(+)
cancer stem cells in ovarian cancer. Sci Rep 5, 10357,
doi:10.1038/srep10357 (2015).

104
 List of references

101 Tu, Z. et al. CXCR4 is involved in CD133-induced EMT in non-small cell lung
cancer. Int J Oncol 50, 505-514, doi:10.3892/ijo.2016.3812 (2017).
102 Chaffer, C. L. & Weinberg, R. A. A perspective on cancer cell metastasis.
Science 331, 1559-1564, doi:10.1126/science.1203543 (2011).
103 Ayala, G. et al. Reactive stroma as a predictor of biochemical-free recurrence
in prostate cancer. Clin Cancer Res 9, 4792-4801 (2003).
104 Ligorio, M. et al. Stromal Microenvironment Shapes the Intratumoral
Architecture of Pancreatic Cancer. Cell 178, 160-175 e127,
doi:10.1016/j.cell.2019.05.012 (2019).
105 Carr, R. M. & Fernandez-Zapico, M. E. Pancreatic cancer microenvironment,
to target or not to target? EMBO Mol Med 8, 80-82,
doi:10.15252/emmm.201505948 (2016).
106 Fares, J., Fares, M. Y., Khachfe, H. H., Salhab, H. A. & Fares, Y. Molecular
principles of metastasis: a hallmark of cancer revisited. Signal Transduct
Target Ther 5, 28, doi:10.1038/s41392-020-0134-x (2020).
107 Haque, M. R. et al. Patient-derived pancreatic cancer-on-a-chip recapitulates
the tumor microenvironment. Microsyst Nanoeng 8, 36, doi:10.1038/s41378-
022-00370-6 (2022).
108 Rucki, A. A. et al. Heterogeneous Stromal Signaling within the Tumor
Microenvironment Controls the Metastasis of Pancreatic Cancer. Cancer Res
77, 41-52, doi:10.1158/0008-5472.Can-16-1383 (2017).
109 Chen, Q. et al. Tumour cell-derived debris and IgG synergistically promote
metastasis of pancreatic cancer by inducing inflammation via tumour-
associated macrophages. Br J Cancer 121, 786-795, doi:10.1038/s41416-
019-0595-2 (2019).
110 Sung, G. H., Chang, H., Lee, J. Y., Song, S. Y. & Kim, H. S. Pancreatic-
cancer-cell-derived trefoil factor 2 impairs maturation and migration of human
monocyte-derived dendritic cells in vitro. Anim Cells Syst (Seoul) 22, 368-
381, doi:10.1080/19768354.2018.1527721 (2018).
111 Sun, X. et al. Inflammatory cell-derived CXCL3 promotes pancreatic cancer
metastasis through a novel myofibroblast-hijacked cancer escape
mechanism. Gut 71, 129-147, doi:10.1136/gutjnl-2020-322744 (2022).

105
 List of references

112 Öhlund, D. et al. Distinct populations of inflammatory fibroblasts and

myofibroblasts in pancreatic cancer. J Exp Med 214, 579-596,
doi:10.1084/jem.20162024 (2017).
113 Elyada, E. et al. Cross-Species Single-Cell Analysis of Pancreatic Ductal
Adenocarcinoma Reveals Antigen-Presenting Cancer-Associated
Fibroblasts. Cancer Discov 9, 1102-1123, doi:10.1158/2159-8290.Cd-19-
0094 (2019).
114 Omary, M. B., Lugea, A., Lowe, A. W. & Pandol, S. J. The pancreatic stellate
cell: a star on the rise in pancreatic diseases. J Clin Invest 117, 50-59,
doi:10.1172/jci30082 (2007).
115 Zhan, H. X. et al. Crosstalk between stromal cells and cancer cells in
pancreatic cancer: New insights into stromal biology. Cancer Lett 392, 83-
93, doi:10.1016/j.canlet.2017.01.041 (2017).
116 Bachem, M. G. et al. Pancreatic carcinoma cells induce fibrosis by stimulating
proliferation and matrix synthesis of stellate cells. Gastroenterology 128, 907-
921, doi:10.1053/j.gastro.2004.12.036 (2005).
117 Apte, M. V. et al. Desmoplastic reaction in pancreatic cancer: role of
pancreatic stellate cells. Pancreas 29, 179-187, doi:10.1097/00006676-
200410000-00002 (2004).
118 Kim, N. et al. Formation of vitamin A lipid droplets in pancreatic stellate cells
requires albumin. Gut 58, 1382-1390, doi:10.1136/gut.2008.170233 (2009).
119 Wu, Q. et al. Functions of pancreatic stellate cell-derived soluble factors in
the microenvironment of pancreatic ductal carcinoma. Oncotarget 8, 102721-
102738, doi:10.18632/oncotarget.21970 (2017).
120 Ren, B. et al. Tumor microenvironment participates in metastasis of
pancreatic cancer. Mol Cancer 17, 108, doi:10.1186/s12943-018-0858-1
(2018).
121 Xu, Z. et al. Role of pancreatic stellate cells in pancreatic cancer metastasis.
Am J Pathol 177, 2585-2596, doi:10.2353/ajpath.2010.090899 (2010).
122 Erkan, M., Kurtoglu, M. & Kleeff, J. The role of hypoxia in pancreatic cancer:
a potential therapeutic target? Expert Rev Gastroenterol Hepatol 10, 301-
316, doi:10.1586/17474124.2016.1117386 (2016).

106
 List of references

123 Jiang, Y. et al. FOXP3+ lymphocyte density in pancreatic cancer correlates

with lymph node metastasis. PLoS One 9, e106741,
doi:10.1371/journal.pone.0106741 (2014).
124 Shields, J. D., Kourtis, I. C., Tomei, A. A., Roberts, J. M. & Swartz, M. A.
Induction of lymphoidlike stroma and immune escape by tumors that express
the chemokine CCL21. Science 328, 749-752, doi:10.1126/science.1185837
(2010).
125 Podgrabinska, S. et al. Inflamed lymphatic endothelium suppresses dendritic
cell maturation and function via Mac-1/ICAM-1-dependent mechanism. J
Immunol 183, 1767-1779, doi:10.4049/jimmunol.0802167 (2009).
126 Kurahara, H. et al. M2-polarized tumor-associated macrophage infiltration of
regional lymph nodes is associated with nodal lymphangiogenesis and occult
nodal involvement in pN0 pancreatic cancer. Pancreas 42, 155-159,
doi:10.1097/MPA.0b013e318254f2d1 (2013).
127 Truong, L. H. & Pauklin, S. Pancreatic Cancer Microenvironment and Cellular
Composition: Current Understandings and Therapeutic Approaches.
Cancers (Basel) 13, doi:10.3390/cancers13195028 (2021).
128 Prager, B. C., Xie, Q., Bao, S. & Rich, J. N. Cancer Stem Cells: The Architects
of the Tumor Ecosystem. Cell Stem Cell 24, 41-53,
doi:10.1016/j.stem.2018.12.009 (2019).
129 Albini, A. et al. Cancer stem cells and the tumor microenvironment: interplay
in tumor heterogeneity. Connect Tissue Res 56, 414-425,
doi:10.3109/03008207.2015.1066780 (2015).
130 Begum, A. et al. The extracellular matrix and focal adhesion kinase signaling
regulate cancer stem cell function in pancreatic ductal adenocarcinoma.
PLoS One 12, e0180181, doi:10.1371/journal.pone.0180181 (2017).
131 Nomura, A. et al. Microenvironment mediated alterations to metabolic
pathways confer increased chemo-resistance in CD133+ tumor initiating
cells. Oncotarget 7, 56324-56337, doi:10.18632/oncotarget.10838 (2016).
132 Meng, F. et al. Interaction between pancreatic cancer cells and tumor-
associated macrophages promotes the invasion of pancreatic cancer cells
and the differentiation and migration of macrophages. IUBMB Life 66, 835-
846, doi:10.1002/iub.1336 (2014).

107
 List of references

133 Zhan, H. X., Xu, J. W., Wu, D., Zhang, T. P. & Hu, S. Y. Pancreatic cancer
stem cells: new insight into a stubborn disease. Cancer Lett 357, 429-437,
doi:10.1016/j.canlet.2014.12.004 (2015).
134 Lonardo, E., Frias-Aldeguer, J., Hermann, P. C. & Heeschen, C. Pancreatic
stellate cells form a niche for cancer stem cells and promote their self-
renewal and invasiveness. Cell Cycle 11, 1282-1290, doi:10.4161/cc.19679
(2012).
135 Sainz, B., Jr. et al. Microenvironmental hCAP-18/LL-37 promotes pancreatic
ductal adenocarcinoma by activating its cancer stem cell compartment. Gut
64, 1921-1935, doi:10.1136/gutjnl-2014-308935 (2015).
136 Zhang, Y. et al. Canonical wnt signaling is required for pancreatic
carcinogenesis. Cancer Res 73, 4909-4922, doi:10.1158/0008-5472.Can-
12-4384 (2013).
137 Chen, K., Huang, Y. H. & Chen, J. L. Understanding and targeting cancer
stem cells: therapeutic implications and challenges. Acta Pharmacol Sin 34,
732-740, doi:10.1038/aps.2013.27 (2013).
138 Lee, J. Y., Song, S. Y. & Park, J. Y. Notch pathway activation is associated
with pancreatic cancer treatment failure. Pancreatology 14, 48-53,
doi:10.1016/j.pan.2013.11.011 (2014).
139 Abel, E. V. et al. The Notch pathway is important in maintaining the cancer
stem cell population in pancreatic cancer. PLoS One 9, e91983,
doi:10.1371/journal.pone.0091983 (2014).
140 Onishi, H. & Katano, M. Hedgehog signaling pathway as a new therapeutic
target in pancreatic cancer. World J Gastroenterol 20, 2335-2342,
doi:10.3748/wjg.v20.i9.2335 (2014).
141 Sainz, B., Jr., Martín, B., Tatari, M., Heeschen, C. & Guerra, S. ISG15 is a
critical microenvironmental factor for pancreatic cancer stem cells. Cancer
Res 74, 7309-7320, doi:10.1158/0008-5472.Can-14-1354 (2014).
142 Szczerba, B. M. et al. Neutrophils escort circulating tumour cells to enable
cell cycle progression. Nature 566, 553-557, doi:10.1038/s41586-019-0915-
y (2019).
143 Bramhall, S. R. et al. A double-blind placebo-controlled, randomised study
comparing gemcitabine and marimastat with gemcitabine and placebo as first

108
 List of references

line therapy in patients with advanced pancreatic cancer. Br J Cancer 87,

161-167, doi:10.1038/sj.bjc.6600446 (2002).
144 Bramhall, S. R., Rosemurgy, A., Brown, P. D., Bowry, C. & Buckels, J. A.
Marimastat as first-line therapy for patients with unresectable pancreatic
cancer: a randomized trial. J Clin Oncol 19, 3447-3455,
doi:10.1200/jco.2001.19.15.3447 (2001).
145 Moore, M. J. et al. Comparison of gemcitabine versus the matrix
metalloproteinase inhibitor BAY 12-9566 in patients with advanced or
metastatic adenocarcinoma of the pancreas: a phase III trial of the National
Cancer Institute of Canada Clinical Trials Group. J Clin Oncol 21, 3296-3302,
doi:10.1200/jco.2003.02.098 (2003).
146 Ramanathan, R. K. et al. Phase IB/II Randomized Study of FOLFIRINOX
Plus Pegylated Recombinant Human Hyaluronidase Versus FOLFIRINOX
Alone in Patients With Metastatic Pancreatic Adenocarcinoma: SWOG
S1313. J Clin Oncol 37, 1062-1069, doi:10.1200/jco.18.01295 (2019).
147 Chen, X. & Song, E. Turning foes to friends: targeting cancer-associated
fibroblasts. Nat Rev Drug Discov 18, 99-115, doi:10.1038/s41573-018-0004-
1 (2019).
148 Hofheinz, R. D. et al. Stromal antigen targeting by a humanised monoclonal
antibody: an early phase II trial of sibrotuzumab in patients with metastatic
colorectal cancer. Onkologie 26, 44-48, doi:10.1159/000069863 (2003).
149 Gunderson, A. J. et al. Blockade of fibroblast activation protein in
combination with radiation treatment in murine models of pancreatic
adenocarcinoma. PLoS One 14, e0211117,
doi:10.1371/journal.pone.0211117 (2019).
150 Özdemir, B. C. et al. Depletion of carcinoma-associated fibroblasts and
fibrosis induces immunosuppression and accelerates pancreas cancer with
reduced survival. Cancer Cell 25, 719-734, doi:10.1016/j.ccr.2014.04.005
(2014).
151 Herting, C. J., Karpovsky, I. & Lesinski, G. B. The tumor microenvironment in
pancreatic ductal adenocarcinoma: current perspectives and future
directions. Cancer Metastasis Rev 40, 675-689, doi:10.1007/s10555-021-
09988-w (2021).

109
 List of references

152 Sherman, M. H. et al. Vitamin D receptor-mediated stromal reprogramming

suppresses pancreatitis and enhances pancreatic cancer therapy. Cell 159,
80-93, doi:10.1016/j.cell.2014.08.007 (2014).
153 Chronopoulos, A. et al. ATRA mechanically reprograms pancreatic stellate
cells to suppress matrix remodelling and inhibit cancer cell invasion. Nat
Commun 7, 12630, doi:10.1038/ncomms12630 (2016).
154 Erkan, M. et al. The role of stroma in pancreatic cancer: diagnostic and
therapeutic implications. Nat Rev Gastroenterol Hepatol 9, 454-467,
doi:10.1038/nrgastro.2012.115 (2012).
155 Ene-Obong, A. et al. Activated pancreatic stellate cells sequester CD8+ T
cells to reduce their infiltration of the juxtatumoral compartment of pancreatic
ductal adenocarcinoma. Gastroenterology 145, 1121-1132,
doi:10.1053/j.gastro.2013.07.025 (2013).
156 Tang, D. et al. High expression of Galectin-1 in pancreatic stellate cells plays
a role in the development and maintenance of an immunosuppressive
microenvironment in pancreatic cancer. Int J Cancer 130, 2337-2348,
doi:10.1002/ijc.26290 (2012).
157 Strauss, L. et al. A unique subset of CD4+CD25highFoxp3+ T cells secreting
interleukin-10 and transforming growth factor-beta1 mediates suppression in
the tumor microenvironment. Clin Cancer Res 13, 4345-4354,
doi:10.1158/1078-0432.Ccr-07-0472 (2007).
158 Bailey, P. et al. Genomic analyses identify molecular subtypes of pancreatic
cancer. Nature 531, 47-52, doi:10.1038/nature16965 (2016).
159 Danilova, L. et al. Programmed Cell Death Ligand-1 (PD-L1) and CD8
Expression Profiling Identify an Immunologic Subtype of Pancreatic Ductal
Adenocarcinomas with Favorable Survival. Cancer Immunol Res 7, 886-895,
doi:10.1158/2326-6066.CIR-18-0822 (2019).
160 Feig, C. et al. Targeting CXCL12 from FAP-expressing carcinoma-
associated fibroblasts synergizes with anti-PD-L1 immunotherapy in
pancreatic cancer. Proc. Natl. Acad. Sci. U. S. A. 110, 20212-20217,
doi:10.1073/pnas.1320318110 (2013).
161 Federsppiel, B. et al. Molecular cloning of the cDNA and chromosomal
localization of the gene for a putative seven-transmembrane segment (7-

110
 List of references

TMS) receptor isolated from human spleen. Genomics 16, 707-712,

doi:10.1006/geno.1993.1251 (1993).
162 Nomura, H., Nielsen, B. W. & Matsushima, K. Molecular cloning of cDNAs
encoding a LD78 receptor and putative leukocyte chemotactic peptide
receptors. Int Immunol 5, 1239-1249, doi:10.1093/intimm/5.10.1239 (1993).
163 Loetscher, M. et al. Cloning of a human seven-transmembrane domain
receptor, LESTR, that is highly expressed in leukocytes. J Biol Chem 269,
232-237 (1994).
164 Feng, Y., Broder, C. C., Kennedy, P. E. & Berger, E. A. HIV-1 entry cofactor:
functional cDNA cloning of a seven-transmembrane, G protein-coupled
receptor. Science 272, 872-877, doi:10.1126/science.272.5263.872 (1996).
165 Deng, H. et al. Identification of a major co-receptor for primary isolates of
HIV-1. Nature 381, 661-666, doi:10.1038/381661a0 (1996).
166 Doranz, B. J. et al. A dual-tropic primary HIV-1 isolate that uses fusin and the
beta-chemokine receptors CKR-5, CKR-3, and CKR-2b as fusion cofactors.
Cell 85, 1149-1158, doi:10.1016/s0092-8674(00)81314-8 (1996).
167 Bleul, C. C. et al. The lymphocyte chemoattractant SDF-1 is a ligand for
LESTR/fusin and blocks HIV-1 entry. Nature 382, 829-833,
doi:10.1038/382829a0 (1996).
168 Oberlin, E. et al. The CXC chemokine SDF-1 is the ligand for LESTR/fusin
and prevents infection by T-cell-line-adapted HIV-1. Nature 382, 833-835,
doi:10.1038/382833a0 (1996).
169 Nagasawa, T., Kikutani, H. & Kishimoto, T. Molecular cloning and structure
of a pre-B-cell growth-stimulating factor. Proc. Natl. Acad. Sci. U. S. A. 91,
2305-2309, doi:10.1073/pnas.91.6.2305 (1994).
170 Bleul, C. C., Fuhlbrigge, R. C., Casasnovas, J. M., Aiuti, A. & Springer, T. A.
A highly efficacious lymphocyte chemoattractant, stromal cell-derived factor
1 (SDF-1). J Exp Med 184, 1101-1109, doi:10.1084/jem.184.3.1101 (1996).
171 Murdoch, C. CXCR4: chemokine receptor extraordinaire. Immunol Rev 177,
175-184, doi:10.1034/j.1600-065x.2000.17715.x (2000).
172 Hori, T. et al. Detection and delineation of CXCR-4 (fusin) as an entry and
fusion cofactor for T-tropic [correction of T cell-tropic] HIV-1 by three different
monoclonal antibodies. J Immunol 160, 180-188 (1998).

111
 List of references

173 Busillo, J. M. & Benovic, J. L. Regulation of CXCR4 signaling. Biochim

Biophys Acta 1768, 952-963, doi:10.1016/j.bbamem.2006.11.002 (2007).
174 Saini, V., Romero, J., Marchese, A. & Majetschak, M. Ubiquitin receptor
binding and signaling in primary human leukocytes. Commun Integr Biol 3,
608-610, doi:10.4161/cib.3.6.13375 (2010).
175 Blanchet, X., Langer, M., Weber, C., Koenen, R. R. & von Hundelshausen,
P. Touch of chemokines. Front Immunol 3, 175,
doi:10.3389/fimmu.2012.00175 (2012).
176 Cheng, Z. J. et al. beta-arrestin differentially regulates the chemokine
receptor CXCR4-mediated signaling and receptor internalization, and this
implicates multiple interaction sites between beta-arrestin and CXCR4. J Biol
Chem 275, 2479-2485, doi:10.1074/jbc.275.4.2479 (2000).
177 Xu, C., Zhao, H., Chen, H. & Yao, Q. CXCR4 in breast cancer: oncogenic
role and therapeutic targeting. Drug Des Devel Ther 9, 4953-4964,
doi:10.2147/dddt.S84932 (2015).
178 Krieg, A., Riemer, J. C., Telan, L. A., Gabbert, H. E. & Knoefel, W. T. CXCR4-
-A Prognostic and Clinicopathological Biomarker for Pancreatic Ductal
Adenocarcinoma: A Meta-Analysis. PLoS One 10, e0130192,
doi:10.1371/journal.pone.0130192 (2015).
179 Sun, X. et al. CXCL12 / CXCR4 / CXCR7 chemokine axis and cancer
progression. Cancer Metastasis Rev 29, 709-722, doi:10.1007/s10555-010-
9256-x (2010).
180 Katoh, M. & Katoh, M. Integrative genomic analyses of CXCR4:
transcriptional regulation of CXCR4 based on TGFbeta, Nodal, Activin
signaling and POU5F1, FOXA2, FOXC2, FOXH1, SOX17, and GFI1
transcription factors. Int J Oncol 36, 415-420, doi:10.3892/ijo_00000514
(2010).
181 Li, X. et al. SDF-1/CXCR4 signaling induces pancreatic cancer cell invasion
and epithelial-mesenchymal transition in vitro through non-canonical
activation of Hedgehog pathway. Cancer Lett 322, 169-176,
doi:10.1016/j.canlet.2012.02.035 (2012).
182 Singh, A. P. et al. CXCL12/CXCR4 protein signaling axis induces sonic
hedgehog expression in pancreatic cancer cells via extracellular regulated
kinase- and Akt kinase-mediated activation of nuclear factor κB: implications

112
 List of references

for bidirectional tumor-stromal interactions. J Biol Chem 287, 39115-39124,

doi:10.1074/jbc.M112.409581 (2012).
183 Wang, Z. et al. Blockade of SDF-1/CXCR4 signalling inhibits pancreatic
cancer progression in vitro via inactivation of canonical Wnt pathway. Br J
Cancer 99, 1695-1703, doi:10.1038/sj.bjc.6604745 (2008).
184 Billadeau, D. D. et al. Characterization of the CXCR4 signaling in pancreatic
cancer cells. Int J Gastrointest Cancer 37, 110-119, doi:10.1007/s12029-
007-0011-7 (2006).
185 Sleightholm, R. L. et al. Emerging roles of the CXCL12/CXCR4 axis in
pancreatic cancer progression and therapy. Pharmacol Ther 179, 158-170,
doi:10.1016/j.pharmthera.2017.05.012 (2017).
186 Pan, F. et al. SDF-1α upregulation of MMP-2 is mediated by p38 MAPK
signaling in pancreatic cancer cell lines. Mol Biol Rep 40, 4139-4146,
doi:10.1007/s11033-012-2225-4 (2013).
187 Shen, B. et al. CXCL12-CXCR4 promotes proliferation and invasion of
pancreatic cancer cells. Asian Pac J Cancer Prev 14, 5403-5408,
doi:10.7314/apjcp.2013.14.9.5403 (2013).
188 Yang, P. et al. SDF-1/CXCR4 signaling up-regulates survivin to regulate
human sacral chondrosarcoma cell cycle and epithelial-mesenchymal
transition via ERK and PI3K/AKT pathway. Med Oncol 32, 377,
doi:10.1007/s12032-014-0377-x (2015).
189 Lee, C. C. et al. Disrupting the CXCL12/CXCR4 axis disturbs the
characteristics of glioblastoma stem-like cells of rat RG2 glioblastoma.
Cancer Cell Int 13, 85, doi:10.1186/1475-2867-13-85 (2013).
190 Hu, T. H. et al. SDF-1/CXCR4 promotes epithelial-mesenchymal transition
and progression of colorectal cancer by activation of the Wnt/beta-catenin
signaling pathway. Cancer Lett 354, 417-426,
doi:10.1016/j.canlet.2014.08.012 (2014).
191 Li, X. et al. The SDF-1/CXCR4 axis induces epithelial–mesenchymal
transition in hepatocellular carcinoma. Mol Cell Biochem 392, 77-84,
doi:10.1007/s11010-014-2020-8 (2014).
192 Yang, B. G. et al. Binding of lymphoid chemokines to collagen IV that
accumulates in the basal lamina of high endothelial venules: its implications

113
 List of references

in lymphocyte trafficking. J Immunol 179, 4376-4382,

doi:10.4049/jimmunol.179.7.4376 (2007).
193 Müller, A. et al. Involvement of chemokine receptors in breast cancer
metastasis. Nature 410, 50-56, doi:10.1038/35065016 (2001).
194 Kim, J. et al. Chemokine receptor CXCR4 expression in patients with
melanoma and colorectal cancer liver metastases and the association with
disease outcome. Ann Surg 244, 113-120,
doi:10.1097/01.sla.0000217690.65909.9c (2006).
195 Gelmini, S. et al. The CXCR4/CXCL12 axis in endometrial cancer. Clin Exp
Metastasis 26, 261-268, doi:10.1007/s10585-009-9240-4 (2009).
196 Guo, F. et al. CXCL12/CXCR4: a symbiotic bridge linking cancer cells and
their stromal neighbors in oncogenic communication networks. Oncogene
35, 816-826, doi:10.1038/onc.2015.139 (2016).
197 Ceradini, D. J. et al. Progenitor cell trafficking is regulated by hypoxic
gradients through HIF-1 induction of SDF-1. Nat Med 10, 858-864,
doi:10.1038/nm1075 (2004).
198 Ratajczak, M. Z. et al. The pleiotropic effects of the SDF-1-CXCR4 axis in
organogenesis, regeneration and tumorigenesis. Leukemia 20, 1915-1924,
doi:10.1038/sj.leu.2404357 (2006).
199 Nagasawa, T. CXC chemokine ligand 12 (CXCL12) and its receptor CXCR4.
J Mol Med (Berl) 92, 433-439, doi:10.1007/s00109-014-1123-8 (2014).
200 Katsumoto, K. & Kume, S. The role of CXCL12-CXCR4 signaling pathway in
pancreatic development. Theranostics 3, 11-17, doi:10.7150/thno.4806
(2013).
201 Kure, S. et al. Expression of cancer stem cell markers in pancreatic
intraepithelial neoplasias and pancreatic ductal adenocarcinomas. Int J
Oncol 41, 1314-1324, doi:10.3892/ijo.2012.1565 (2012).
202 Thomas, R. M. et al. The chemokine receptor CXCR4 is expressed in
pancreatic intraepithelial neoplasia. Gut 57, 1555-1560,
doi:10.1136/gut.2007.143941 (2008).
203 Wang, Q. et al. BRCA2 dysfunction promotes malignant transformation of
pancreatic intraepithelial neoplasia. Anticancer Agents Med Chem 13, 261-
269, doi:10.2174/1871520611313020012 (2013).

114
 List of references

204 Balic, A., Dorado, J., Alonso-Gómez, M. & Heeschen, C. Stem cells as the
root of pancreatic ductal adenocarcinoma. Exp Cell Res 318, 691-704,
doi:10.1016/j.yexcr.2011.11.007 (2012).
205 Ma, Y., Hwang, R. F., Logsdon, C. D. & Ullrich, S. E. Dynamic mast cell-
stromal cell interactions promote growth of pancreatic cancer. Cancer Res
73, 3927-3937, doi:10.1158/0008-5472.Can-12-4479 (2013).
206 Park, B., Prasad, S., Yadav, V., Sung, B. & Aggarwal, B. B. Boswellic acid
suppresses growth and metastasis of human pancreatic tumors in an
orthotopic nude mouse model through modulation of multiple targets. PLoS
One 6, e26943, doi:10.1371/journal.pone.0026943 (2011).
207 Zhou, P. et al. The epithelial to mesenchymal transition (EMT) and cancer
stem cells: implication for treatment resistance in pancreatic cancer. Mol
Cancer 16, 52, doi:10.1186/s12943-017-0624-9 (2017).
208 Hirai, I. et al. Perineural invasion in pancreatic cancer. Pancreas 24, 15-25,
doi:10.1097/00006676-200201000-00003 (2002).
209 Maréchal, R. et al. High expression of CXCR4 may predict poor survival in
resected pancreatic adenocarcinoma. Br J Cancer 100, 1444-1451,
doi:10.1038/sj.bjc.6605020 (2009).
210 Wang, Z. et al. Aberrant expression of CXCR4 and β-catenin in pancreatic
cancer. Anticancer Res 33, 4103-4110 (2013).
211 Mizukami, T. et al. Immunohistochemical analysis of cancer stem cell
markers in pancreatic adenocarcinoma patients after neoadjuvant
chemoradiotherapy. BMC Cancer 14, 687, doi:10.1186/1471-2407-14-687
(2014).
212 Cui, K. et al. The CXCR4-CXCL12 pathway facilitates the progression of
pancreatic cancer via induction of angiogenesis and lymphangiogenesis. J
Surg Res 171, 143-150, doi:10.1016/j.jss.2010.03.001 (2011).
213 Bachet, J. B. et al. Contribution of CXCR4 and SMAD4 in predicting disease
progression pattern and benefit from adjuvant chemotherapy in resected
pancreatic adenocarcinoma. Ann Oncol 23, 2327-2335,
doi:10.1093/annonc/mdr617 (2012).
214 Xu, Q. et al. Stromal-derived factor-1α/CXCL12-CXCR4 chemotactic
pathway promotes perineural invasion in pancreatic cancer. Oncotarget 6,
4717-4732, doi:10.18632/oncotarget.3069 (2015).

115
 List of references

215 Wu, H. et al. Coexpression of EGFR and CXCR4 predicts poor prognosis in
resected pancreatic ductal adenocarcinoma. PLoS One 10, e0116803,
doi:10.1371/journal.pone.0116803 (2015).
216 Demir, I. E. et al. Early pancreatic cancer lesions suppress pain through
CXCL12-mediated chemoattraction of Schwann cells. Proc. Natl. Acad. Sci.
U. S. A. 114, E85-e94, doi:10.1073/pnas.1606909114 (2017).
217 Lee, Y. et al. PAUF functions in the metastasis of human pancreatic cancer
cells and upregulates CXCR4 expression. Oncogene 29, 56-67,
doi:10.1038/onc.2009.298 (2010).
218 Kim, J. et al. Identification of anti-malarial compounds as novel antagonists
to chemokine receptor CXCR4 in pancreatic cancer cells. PLoS One 7,
e31004, doi:10.1371/journal.pone.0031004 (2012).
219 Eberl, M. et al. Hedgehog-EGFR cooperation response genes determine the
oncogenic phenotype of basal cell carcinoma and tumour-initiating
pancreatic cancer cells. EMBO Mol Med 4, 218-233,
doi:10.1002/emmm.201100201 (2012).
220 Singh, S., Srivastava, S. K., Bhardwaj, A., Owen, L. B. & Singh, A. P.
CXCL12-CXCR4 signalling axis confers gemcitabine resistance to pancreatic
cancer cells: a novel target for therapy. Br J Cancer 103, 1671-1679,
doi:10.1038/sj.bjc.6605968 (2010).
221 Zlotnik, A., Burkhardt, A. M. & Homey, B. Homeostatic chemokine receptors
and organ-specific metastasis. Nat Rev Immunol 11, 597-606,
doi:10.1038/nri3049 (2011).
222 Arora, S. et al. An undesired effect of chemotherapy: gemcitabine promotes
pancreatic cancer cell invasiveness through reactive oxygen species-
dependent, nuclear factor κB- and hypoxia-inducible factor 1α-mediated up-
regulation of CXCR4. J Biol Chem 288, 21197-21207,
doi:10.1074/jbc.M113.484576 (2013).
223 Brave, M. et al. FDA review summary: Mozobil in combination with
granulocyte colony-stimulating factor to mobilize hematopoietic stem cells to
the peripheral blood for collection and subsequent autologous
transplantation. Oncology 78, 282-288, doi:10.1159/000315736 (2010).

116
 List of references

224 Marchesi, F. et al. Increased survival, proliferation, and migration in

metastatic human pancreatic tumor cells expressing functional CXCR4.
Cancer Res 64, 8420-8427, doi:10.1158/0008-5472.Can-04-1343 (2004).
225 Gao, Z., Wang, X., Wu, K., Zhao, Y. & Hu, G. Pancreatic stellate cells
increase the invasion of human pancreatic cancer cells through the stromal
cell-derived factor-1/CXCR4 axis. Pancreatology 10, 186-193,
doi:10.1159/000236012 (2010).
226 Ottaiano, A. et al. Inhibitory effects of anti-CXCR4 antibodies on human colon
cancer cells. Cancer Immunol Immunother 54, 781-791,
doi:10.1007/s00262-004-0636-3 (2005).
227 Peng, Z. H. & Kopeček, J. HPMA Copolymer CXCR4 Antagonist Conjugates
Substantially Inhibited the Migration of Prostate Cancer Cells. ACS Macro
Lett 3, 1240-1243, doi:10.1021/mz5006537 (2014).
228 Chittasupho, C., Lirdprapamongkol, K., Kewsuwan, P. & Sarisuta, N.
Targeted delivery of doxorubicin to A549 lung cancer cells by CXCR4
antagonist conjugated PLGA nanoparticles. Eur J Pharm Biopharm 88, 529-
538, doi:10.1016/j.ejpb.2014.06.020 (2014).
229 Yu, F., Li, J., Xie, Y., Sleightholm, R. L. & Oupický, D. Polymeric chloroquine
as an inhibitor of cancer cell migration and experimental lung metastasis. J
Control Release 244, 347-356, doi:10.1016/j.jconrel.2016.07.040 (2016).
230 Abraham, M. et al. Single Dose of the CXCR4 Antagonist BL-8040 Induces
Rapid Mobilization for the Collection of Human CD34(+) Cells in Healthy
Volunteers. Clin Cancer Res 23, 6790-6801, doi:10.1158/1078-0432.Ccr-16-
2919 (2017).
231 Peled, A. et al. The high-affinity CXCR4 antagonist BKT140 is safe and
induces a robust mobilization of human CD34+ cells in patients with multiple
myeloma. Clin Cancer Res 20, 469-479, doi:10.1158/1078-0432.Ccr-13-
1302 (2014).
232 Seo, Y. D. et al. Mobilization of CD8(+) T Cells via CXCR4 Blockade
Facilitates PD-1 Checkpoint Therapy in Human Pancreatic Cancer. Clin
Cancer Res 25, 3934-3945, doi:10.1158/1078-0432.Ccr-19-0081 (2019).
233 Bockorny, B. et al. BL-8040, a CXCR4 antagonist, in combination with
pembrolizumab and chemotherapy for pancreatic cancer: the COMBAT trial.
Nat Med 26, 878-885, doi:10.1038/s41591-020-0880-x (2020).

117
 List of references

234 Zirafi, O. et al. Discovery and characterization of an endogenous CXCR4

antagonist. Cell Rep 11, 737-747, doi:10.1016/j.celrep.2015.03.061 (2015).
235 Harms, M. et al. An optimized derivative of an endogenous CXCR4
antagonist prevents atopic dermatitis and airway inflammation. Acta Pharm
Sin B 11, 2694-2708, doi:10.1016/j.apsb.2020.12.005 (2021).
236 A., R. Synthesis of Peptide-Functionalized Poly(bis-sulfone) Copolymers
Regulating HIV-1 Entry and Cancer Stem Cell Migration. ACS Macro Lett 6
241-246, doi:DOI: 10.1021/acsmacrolett.7b00030 (2017 ).
237 Kaiser, L. M. et al. Targeting of CXCR4 by the Naturally Occurring CXCR4
Antagonist EPI-X4 in Waldenström's Macroglobulinemia. Cancers (Basel)
13, doi:10.3390/cancers13040826 (2021).
238 Zirafi, O., Hermann, P. C. & Münch, J. Proteolytic processing of human serum
albumin generates EPI-X4, an endogenous antagonist of CXCR4. J Leukoc
Biol 99, 863-868, doi:10.1189/jlb.2MR1115-521RR (2016).
239 Harms, M. Endogenous CXCR4 Antagonists: Role in HIV-1 Transmission
and Therapy of CXCR4-linked diseases PhD thesis, University of Ulm,
(2021).
240 Hermann, P. C. et al. Nicotine promotes initiation and progression of KRAS-
induced pancreatic cancer via Gata6-dependent dedifferentiation of acinar
cells in mice. Gastroenterology 147, 1119-1133.e1114,
doi:10.1053/j.gastro.2014.08.002 (2014).
241 Jackson, E. L. et al. Analysis of lung tumor initiation and progression using
conditional expression of oncogenic K-ras. Genes Dev 15, 3243-3248,
doi:10.1101/gad.943001 (2001).
242 Marino, S., Vooijs, M., van Der Gulden, H., Jonkers, J. & Berns, A. Induction
of medulloblastomas in p53-null mutant mice by somatic inactivation of Rb in
the external granular layer cells of the cerebellum. Genes Dev 14, 994-1004
(2000).
243 Luche, H., Weber, O., Nageswara Rao, T., Blum, C. & Fehling, H. J. Faithful
activation of an extra-bright red fluorescent protein in "knock-in" Cre-reporter
mice ideally suited for lineage tracing studies. Eur J Immunol 37, 43-53,
doi:10.1002/eji.200636745 (2007).

118
 List of references

244 Kawaguchi, Y. et al. The role of the transcriptional regulator Ptf1a in

converting intestinal to pancreatic progenitors. Nat Genet 32, 128-134,
doi:10.1038/ng959 (2002).
245 Guo, W. et al. Slug and Sox9 cooperatively determine the mammary stem
cell state. Cell 148, 1015-1028, doi:10.1016/j.cell.2012.02.008 (2012).
246 Lonardo, E. et al. Nodal/Activin signaling drives self-renewal and
tumorigenicity of pancreatic cancer stem cells and provides a target for
combined drug therapy. Cell Stem Cell 9, 433-446,
doi:10.1016/j.stem.2011.10.001 (2011).
247 Trajkovic-Arsic, M. et al. Apparent Diffusion Coefficient (ADC) predicts
therapy response in pancreatic ductal adenocarcinoma. Sci Rep 7, 17038,
doi:10.1038/s41598-017-16826-z (2017).
248 Rubio-Viqueira, B. et al. An in vivo platform for translational drug
development in pancreatic cancer. Clin Cancer Res 12, 4652-4661,
doi:10.1158/1078-0432.CCR-06-0113 (2006).
249 Reichert, M. et al. Isolation, culture and genetic manipulation of mouse
pancreatic ductal cells. Nat Protoc 8, 1354-1365,
doi:10.1038/nprot.2013.079 (2013).
250 Mazur, P. K. et al. Combined inhibition of BET family proteins and histone
deacetylases as a potential epigenetics-based therapy for pancreatic ductal
adenocarcinoma. Nat Med 21, 1163-1171, doi:10.1038/nm.3952 (2015).
251 Patro, R., Duggal, G., Love, M. I., Irizarry, R. A. & Kingsford, C. Salmon
provides fast and bias-aware quantification of transcript expression. Nat
Methods 14, 417-419, doi:10.1038/nmeth.4197 (2017).
252 Soneson, C., Love, M. I. & Robinson, M. D. Differential analyses for RNA-
seq: transcript-level estimates improve gene-level inferences. F1000Res 4,
1521, doi:10.12688/f1000research.7563.2 (2015).
253 Love, M. I., Huber, W. & Anders, S. Moderated estimation of fold change and
dispersion for RNA-seq data with DESeq2. Genome Biol 15, 550,
doi:10.1186/s13059-014-0550-8 (2014).
254 Rosenholm, J. M. et al. Targeting of porous hybrid silica nanoparticles to
cancer cells. ACS Nano 3, 197-206, doi:10.1021/nn800781r (2009).
255 Beitzinger, B. et al. Delivery by Dendritic Mesoporous Silica Nanoparticles
Enhances the Antimicrobial Activity of a Napsin-Derived Peptide Against

119
 List of references

Intracellular Mycobacterium tuberculosis. Adv Healthc Mater 10, e2100453,

doi:10.1002/adhm.202100453 (2021).
256 Tang, Z. et al. GEPIA: a web server for cancer and normal gene expression
profiling and interactive analyses. Nucleic Acids Res 45, W98-W102,
doi:10.1093/nar/gkx247 (2017).
257 Jensen, L. J. et al. STRING 8--a global view on proteins and their functional
interactions in 630 organisms. Nucleic Acids Res 37, D412-416,
doi:10.1093/nar/gkn760 (2009).
258 Dontu, G. et al. In vitro propagation and transcriptional profiling of human
mammary stem/progenitor cells. Genes Dev 17, 1253-1270,
doi:10.1101/gad.1061803 (2003).
259 Boj, S. F. et al. Organoid models of human and mouse ductal pancreatic
cancer. Cell 160, 324-338, doi:10.1016/j.cell.2014.12.021 (2015).
260 Knudsen, E. S. et al. Pancreatic cancer cell lines as patient-derived avatars:
genetic characterisation and functional utility. Gut 67, 508-520,
doi:10.1136/gutjnl-2016-313133 (2018).
261 Hu, Q. et al. dCK negatively regulates the NRF2/ARE axis and ROS
production in pancreatic cancer. Cell Prolif 51, e12456,
doi:10.1111/cpr.12456 (2018).
262 Hagmann, W., Faissner, R., Schnolzer, M., Lohr, M. & Jesnowski, R.
Membrane drug transporters and chemoresistance in human pancreatic
carcinoma. Cancers (Basel) 3, 106-125, doi:10.3390/cancers3010106
(2010).
263 Pan, B., Liao, Q., Niu, Z., Zhou, L. & Zhao, Y. Cancer-associated fibroblasts
in pancreatic adenocarcinoma. Future Oncol 11, 2603-2610,
doi:10.2217/FON.15.176 (2015).
264 Bjanes, T. K. et al. Intracellular Cytidine Deaminase Regulates Gemcitabine
Metabolism in Pancreatic Cancer Cell Lines. Drug Metab Dispos 48, 153-
158, doi:10.1124/dmd.119.089334 (2020).
265 Wu, K. J. & Yang, M. H. Epithelial-mesenchymal transition and cancer
stemness: the Twist1-Bmi1 connection. Biosci Rep 31, 449-455,
doi:10.1042/BSR20100114 (2011).

120
 List of references

266 Gilg, A. et al. Absence of the CXCR4 antagonist EPI-X4 from pharmaceutical
human serum albumin preparations. J Transl Med 19, 190,
doi:10.1186/s12967-021-02859-6 (2021).
267 Xu, X. et al. Genome-wide screening reveals an EMT molecular network
mediated by Sonic hedgehog-Gli1 signaling in pancreatic cancer cells. PLoS
One 7, e43119, doi:10.1371/journal.pone.0043119 (2012).
268 Xu, X. et al. Sonic hedgehog-Gli1 signaling pathway regulates the epithelial
mesenchymal transition (EMT) by mediating a new target gene, S100A4, in
pancreatic cancer cells. PLoS One 9, e96441,
doi:10.1371/journal.pone.0096441 (2014).
269 Smigiel, J. M., Parameswaran, N. & Jackson, M. W. Potent EMT and CSC
Phenotypes Are Induced By Oncostatin-M in Pancreatic Cancer. Mol Cancer
Res 15, 478-488, doi:10.1158/1541-7786.Mcr-16-0337 (2017).
270 Le Large, T. Y. S. et al. Key biological processes driving metastatic spread
of pancreatic cancer as identified by multi-omics studies. Semin Cancer Biol
44, 153-169, doi:10.1016/j.semcancer.2017.03.008 (2017).
271 Capurso, G. & Sette, C. Drug resistance in pancreatic cancer: New player
caught in act. EBioMedicine 40, 39-40, doi:10.1016/j.ebiom.2019.02.008
(2019).
272 Hermann, P. C. & Sainz, B., Jr. Pancreatic cancer stem cells: A state or an
entity? Semin Cancer Biol 53, 223-231,
doi:10.1016/j.semcancer.2018.08.007 (2018).
273 Chen, D., Wei, L., Yu, J. & Zhang, L. Regorafenib inhibits colorectal tumor
growth through PUMA-mediated apoptosis. Clin Cancer Res 20, 3472-3484,
doi:10.1158/1078-0432.Ccr-13-2944 (2014).
274 de la Puente, P. et al. MEK inhibitor, TAK-733 reduces proliferation, affects
cell cycle and apoptosis, and synergizes with other targeted therapies in
multiple myeloma. Blood Cancer J 6, e399, doi:10.1038/bcj.2016.7 (2016).
275 Hoshino, R., Tanimura, S., Watanabe, K., Kataoka, T. & Kohno, M. Blockade
of the extracellular signal-regulated kinase pathway induces marked G1 cell
cycle arrest and apoptosis in tumor cells in which the pathway is constitutively
activated: up-regulation of p27(Kip1). J Biol Chem 276, 2686-2692,
doi:10.1074/jbc.M006132200 (2001).

121
 List of references

276 Cheng, J. et al. MEK1 signaling promotes self-renewal and tumorigenicity of

liver cancer stem cells via maintaining SIRT1 protein stabilization.
Oncotarget 7, 20597-20611, doi:10.18632/oncotarget.7972 (2016).
277 Zhan, T. et al. MEK inhibitors activate Wnt signalling and induce stem cell
plasticity in colorectal cancer. Nat Commun 10, 2197, doi:10.1038/s41467-
019-09898-0 (2019).
278 Nallasamy, P. et al. Pancreatic Tumor Microenvironment Factor Promotes
Cancer Stemness via SPP1-CD44 Axis. Gastroenterology 161, 1998-
2013.e1997, doi:10.1053/j.gastro.2021.08.023 (2021).
279 Su, S. et al. CD10(+)GPR77(+) Cancer-Associated Fibroblasts Promote
Cancer Formation and Chemoresistance by Sustaining Cancer Stemness.
Cell 172, 841-856.e816, doi:10.1016/j.cell.2018.01.009 (2018).
280 Ciruna, B. & Rossant, J. FGF signaling regulates mesoderm cell fate
specification and morphogenetic movement at the primitive streak. Dev Cell
1, 37-49, doi:10.1016/s1534-5807(01)00017-x (2001).
281 Geary, L. & LaBonne, C. FGF mediated MAPK and PI3K/Akt Signals make
distinct contributions to pluripotency and the establishment of Neural Crest.
Elife 7, doi:10.7554/eLife.33845 (2018).
282 Strippoli, R. et al. Caveolin-1 deficiency induces a MEK-ERK1/2-Snail-1-
dependent epithelial-mesenchymal transition and fibrosis during peritoneal
dialysis. EMBO Mol Med 7, 102-123, doi:10.15252/emmm.201404127
(2015).
283 Lemieux, E. et al. Constitutively active MEK1 is sufficient to induce epithelial-
to-mesenchymal transition in intestinal epithelial cells and to promote tumor
invasion and metastasis. Int J Cancer 125, 1575-1586, doi:10.1002/ijc.24485
(2009).
284 Cristofanilli, M. et al. Circulating tumor cells, disease progression, and
survival in metastatic breast cancer. N Engl J Med 351, 781-791,
doi:10.1056/NEJMoa040766 (2004).
285 Massagué, J. TGFbeta in Cancer. Cell 134, 215-230,
doi:10.1016/j.cell.2008.07.001 (2008).
286 Brandl, M. et al. IKK(α) controls canonical TGF(ß)-SMAD signaling to
regulate genes expressing SNAIL and SLUG during EMT in panc1 cells. J
Cell Sci 123, 4231-4239, doi:10.1242/jcs.071100 (2010).

122
 List of references

287 Zhang, Y., Yang, J., Bronze, M. S., Houchen, C. W. & Li, M. Interplay of
Tumor Microenvironment Factors and Cancer Stem Cell Enrichment in
Pancreatic Ductal Adenocarcinoma. Gastroenterology 161, 1800-1802,
doi:10.1053/j.gastro.2021.09.024 (2021).
288 Barker, H. E., Paget, J. T., Khan, A. A. & Harrington, K. J. The tumour
microenvironment after radiotherapy: mechanisms of resistance and
recurrence. Nat Rev Cancer 15, 409-425, doi:10.1038/nrc3958 (2015).
289 Braicu, C. et al. A Comprehensive Review on MAPK: A Promising
Therapeutic Target in Cancer. Cancers (Basel) 11,
doi:10.3390/cancers11101618 (2019).
290 Ionescu, C. et al. TIMP-1 expression in human colorectal cancer is
associated with SMAD3 gene expression levels: a pilot study. J
Gastrointestin Liver Dis 23, 413-418, doi:10.15403/jgld.2014.1121.234.smad
(2014).
291 Smith, M. P. et al. The immune microenvironment confers resistance to
MAPK pathway inhibitors through macrophage-derived TNFα. Cancer
Discov 4, 1214-1229, doi:10.1158/2159-8290.Cd-13-1007 (2014).
292 Sang, N. et al. MAPK signaling up-regulates the activity of hypoxia-inducible
factors by its effects on p300. J Biol Chem 278, 14013-14019,
doi:10.1074/jbc.M209702200 (2003).
293 Zhang, Q., Han, Z., Zhu, Y., Chen, J. & Li, W. Role of hypoxia inducible factor-
1 in cancer stem cells (Review). Mol Med Rep 23,
doi:10.3892/mmr.2020.11655 (2021).
294 Tiwari, A. et al. Loss of HIF1A From Pancreatic Cancer Cells Increases
Expression of PPP1R1B and Degradation of p53 to Promote Invasion and
Metastasis. Gastroenterology 159, 1882-1897.e1885,
doi:10.1053/j.gastro.2020.07.046 (2020).
295 Pedersen, K. et al. Pancreatic cancer heterogeneity and response to Mek
inhibition. Oncogene 36, 5639-5647, doi:10.1038/onc.2017.174 (2017).
296 Brabletz, T., Jung, A., Spaderna, S., Hlubek, F. & Kirchner, T. Opinion:
migrating cancer stem cells - an integrated concept of malignant tumour
progression. Nat Rev Cancer 5, 744-749, doi:10.1038/nrc1694 (2005).

123
 List of references

297 Zhang, N. H. et al. Co-expression of CXCR4 and CD133 proteins is

associated with poor prognosis in stage II-III colon cancer patients. Exp Ther
Med 3, 973-982, doi:10.3892/etm.2012.527 (2012).
298 Penzo, M., Habiel, D. M., Ramadass, M., Kew, R. R. & Marcu, K. B. Cell
migration to CXCL12 requires simultaneous IKKα and IKKβ-dependent NF-
κB signaling. Biochim Biophys Acta 1843, 1796-1804,
doi:10.1016/j.bbamcr.2014.04.011 (2014).
299 Zhang, J., Liu, C., Mo, X., Shi, H. & Li, S. Mechanisms by which
CXCR4/CXCL12 cause metastatic behavior in pancreatic cancer. Oncol Lett
15, 1771-1776, doi:10.3892/ol.2017.7512 (2018).
300 Erkan, M. et al. StellaTUM: current consensus and discussion on pancreatic
stellate cell research. Gut 61, 172-178, doi:10.1136/gutjnl-2011-301220
(2012).
301 Koong, A. C. et al. Pancreatic tumors show high levels of hypoxia. Int J Radiat
Oncol Biol Phys 48, 919-922, doi:10.1016/s0360-3016(00)00803-8 (2000).
302 Evans, S. M. & Koch, C. J. Prognostic significance of tumor oxygenation in
humans. Cancer Lett 195, 1-16, doi:10.1016/s0304-3835(03)00012-0
(2003).
303 Erkan, M. et al. Cancer-stellate cell interactions perpetuate the hypoxia-
fibrosis cycle in pancreatic ductal adenocarcinoma. Neoplasia 11, 497-508,
doi:10.1593/neo.81618 (2009).
304 Adamska, A. & Falasca, M. ATP-binding cassette transporters in progression
and clinical outcome of pancreatic cancer: What is the way forward? World J
Gastroenterol 24, 3222-3238, doi:10.3748/wjg.v24.i29.3222 (2018).
305 Zhang, L. et al. Reactive Oxygen Species and Targeted Therapy for
Pancreatic Cancer. Oxid Med Cell Longev 2016, 1616781,
doi:10.1155/2016/1616781 (2016).
306 Taniguchi, N. Clinical significances of superoxide dismutases: changes in
aging, diabetes, ischemia, and cancer. Adv Clin Chem 29, 1-59,
doi:10.1016/s0065-2423(08)60221-8 (1992).
307 Mills, G. C. Hemoglobin catabolism. I. Glutathione peroxidase, an erythrocyte
enzyme which protects hemoglobin from oxidative breakdown. J Biol Chem
229, 189-197 (1957).

124
 List of references

308 Young, B., Purcell, C., Kuang, Y. Q., Charette, N. & Dupré, D. J. Superoxide
Dismutase 1 Regulation of CXCR4-Mediated Signaling in Prostate Cancer
Cells is Dependent on Cellular Oxidative State. Cell Physiol Biochem 37,
2071-2084, doi:10.1159/000438566 (2015).
309 Meng, Q. et al. GPx1 is involved in the induction of protective autophagy in
pancreatic cancer cells in response to glucose deprivation. Cell Death Dis 9,
1187, doi:10.1038/s41419-018-1244-z (2018).
310 Aughton, K. et al. hENT1 Predicts Benefit from Gemcitabine in Pancreatic
Cancer but Only with Low CDA mRNA. Cancers (Basel) 13,
doi:10.3390/cancers13225758 (2021).
311 Zhi, Y. et al. Involvement of the nuclear factor-kappaB signaling pathway in
the regulation of CXC chemokine receptor-4 expression in neuroblastoma
cells induced by tumor necrosis factor-alpha. Int J Mol Med 35, 349-357,
doi:10.3892/ijmm.2014.2032 (2015).
312 Valk-Lingbeek, M. E., Bruggeman, S. W. & van Lohuizen, M. Stem cells and
cancer; the polycomb connection. Cell 118, 409-418,
doi:10.1016/j.cell.2004.08.005 (2004).
313 Widschwendter, M. et al. Epigenetic stem cell signature in cancer. Nat Genet
39, 157-158, doi:10.1038/ng1941 (2007).
314 Molofsky, A. V. et al. Bmi-1 dependence distinguishes neural stem cell self-
renewal from progenitor proliferation. Nature 425, 962-967,
doi:10.1038/nature02060 (2003).
315 Iwama, A. et al. Enhanced self-renewal of hematopoietic stem cells mediated
by the polycomb gene product Bmi-1. Immunity 21, 843-851,
doi:10.1016/j.immuni.2004.11.004 (2004).
316 Sangiorgi, E. & Capecchi, M. R. Bmi1 is expressed in vivo in intestinal stem
cells. Nat Genet 40, 915-920, doi:10.1038/ng.165 (2008).
317 Jacobs, J. J., Kieboom, K., Marino, S., DePinho, R. A. & van Lohuizen, M.
The oncogene and Polycomb-group gene bmi-1 regulates cell proliferation
and senescence through the ink4a locus. Nature 397, 164-168,
doi:10.1038/16476 (1999).
318 Czermin, B. et al. Drosophila enhancer of Zeste/ESC complexes have a
histone H3 methyltransferase activity that marks chromosomal Polycomb
sites. Cell 111, 185-196, doi:10.1016/s0092-8674(02)00975-3 (2002).

125
 List of references

319 Müller, J. et al. Histone methyltransferase activity of a Drosophila Polycomb

group repressor complex. Cell 111, 197-208, doi:10.1016/s0092-
8674(02)00976-5 (2002).
320 Min, J., Zhang, Y. & Xu, R. M. Structural basis for specific binding of
Polycomb chromodomain to histone H3 methylated at Lys 27. Genes Dev
17, 1823-1828, doi:10.1101/gad.269603 (2003).
321 Bracken, A. P. et al. The Polycomb group proteins bind throughout the
INK4A-ARF locus and are disassociated in senescent cells. Genes Dev 21,
525-530, doi:10.1101/gad.415507 (2007).
322 Yang, M. H. et al. Bmi1 is essential in Twist1-induced epithelial-
mesenchymal transition. Nat Cell Biol 12, 982-992, doi:10.1038/ncb2099
(2010).
323 Yamazaki, D., Kurisu, S. & Takenawa, T. Regulation of cancer cell motility
through actin reorganization. Cancer Sci 96, 379-386, doi:10.1111/j.1349-
7006.2005.00062.x (2005).
324 Curtarelli, R. B. et al. Expression of Cancer Stem Cell Biomarkers in Human
Head and Neck Carcinomas: a Systematic Review. Stem Cell Rev Rep 14,
769-784, doi:10.1007/s12015-018-9839-4 (2018).
325 Yanai, H. et al. Intestinal cancer stem cells marked by Bmi1 or Lgr5
expression contribute to tumor propagation via clonal expansion. Sci Rep 7,
41838, doi:10.1038/srep41838 (2017).
326 Dimri, G. P. et al. The Bmi-1 oncogene induces telomerase activity and
immortalizes human mammary epithelial cells. Cancer Res 62, 4736-4745
(2002).
327 Walter, K. et al. Telomerase and Pluripotency Factors Jointly Regulate
Stemness in Pancreatic Cancer Stem Cells. Cancers (Basel) 13,
doi:10.3390/cancers13133145 (2021).
328 Pastrana, E., Silva-Vargas, V. & Doetsch, F. Eyes wide open: a critical review
of sphere-formation as an assay for stem cells. Cell Stem Cell 8, 486-498,
doi:10.1016/j.stem.2011.04.007 (2011).
329 Paranjape, A. N. et al. Bmi1 regulates self-renewal and epithelial to
mesenchymal transition in breast cancer cells through Nanog. BMC Cancer
14, 785, doi:10.1186/1471-2407-14-785 (2014).

126
 List of references

330 Ding, Q. et al. CD133 facilitates epithelial-mesenchymal transition through

interaction with the ERK pathway in pancreatic cancer metastasis. Mol
Cancer 13, 15, doi:10.1186/1476-4598-13-15 (2014).
331 Song, L. B. et al. The polycomb group protein Bmi-1 represses the tumor
suppressor PTEN and induces epithelial-mesenchymal transition in human
nasopharyngeal epithelial cells. J Clin Invest 119, 3626-3636,
doi:10.1172/jci39374 (2009).
332 Liu, S. et al. Hedgehog signaling and Bmi-1 regulate self-renewal of normal
and malignant human mammary stem cells. Cancer Res 66, 6063-6071,
doi:10.1158/0008-5472.Can-06-0054 (2006).
333 Jiang, L. et al. Bmi-1 promotes the aggressiveness of glioma via activating
the NF-kappaB/MMP-9 signaling pathway. BMC Cancer 12, 406,
doi:10.1186/1471-2407-12-406 (2012).
334 Yin, T. et al. Bmi1 inhibition enhances the sensitivity of pancreatic cancer
cells to gemcitabine. Oncotarget 7, 37192-37204,
doi:10.18632/oncotarget.9293 (2016).
335 Tahirovic, Y. A. et al. Small molecule and peptide-based CXCR4 modulators
as therapeutic agents. A patent review for the period from 2010 to 2018.
Expert Opin Ther Pat 30, 87-101, doi:10.1080/13543776.2020.1707186
(2020).
336 De Clercq, E. Mozobil® (Plerixafor, AMD3100), 10 years after its approval by
the US Food and Drug Administration. Antivir Chem Chemother 27,
2040206619829382, doi:10.1177/2040206619829382 (2019).
337 Lei, F., Xi, X., Batra, S. K. & Bronich, T. K. Combination Therapies and Drug
Delivery Platforms in Combating Pancreatic Cancer. J Pharmacol Exp Ther
370, 682-694, doi:10.1124/jpet.118.255786 (2019).
338 Fang, C. Y., Wu, C. C., Fang, C. L., Chen, W. Y. & Chen, C. L. Long-term
growth comparison studies of FBS and FBS alternatives in six head and neck
cell lines. PLoS One 12, e0178960, doi:10.1371/journal.pone.0178960
(2017).
339 Gallmeier, E. et al. Inhibition of ataxia telangiectasia- and Rad3-related
function abrogates the in vitro and in vivo tumorigenicity of human colon
cancer cells through depletion of the CD133(+) tumor-initiating cell fraction.
Stem Cells 29, 418-429, doi:10.1002/stem.595 (2011).

127
 List of references

340 Hermann, P. C. et al. Multimodal Treatment Eliminates Cancer Stem Cells

and Leads to Long-Term Survival in Primary Human Pancreatic Cancer
Tissue Xenografts. PLoS One 8, e66371, doi:10.1371/journal.pone.0066371
(2013).
341 Al-Souhibani, N., Al-Ghamdi, M., Al-Ahmadi, W. & Khabar, K. S.
Posttranscriptional control of the chemokine receptor CXCR4 expression in
cancer cells. Carcinogenesis 35, 1983-1992, doi:10.1093/carcin/bgu080
(2014).
342 Latorre, E. et al. The Ribonucleic Complex HuR-MALAT1 Represses CD133
Expression and Suppresses Epithelial-Mesenchymal Transition in Breast
Cancer. Cancer Res 76, 2626-2636, doi:10.1158/0008-5472.Can-15-2018
(2016).
343 Aoki, H. et al. Autocrine loop between TGF-beta1 and IL-1beta through
Smad3- and ERK-dependent pathways in rat pancreatic stellate cells. Am J
Physiol Cell Physiol 290, C1100-1108, doi:10.1152/ajpcell.00465.2005
(2006).
344 Shek, F. W. et al. Expression of transforming growth factor-beta 1 by
pancreatic stellate cells and its implications for matrix secretion and turnover
in chronic pancreatitis. Am J Pathol 160, 1787-1798, doi:10.1016/s0002-
9440(10)61125-x (2002).
345 Apte, M. V. et al. Pancreatic stellate cells are activated by proinflammatory
cytokines: implications for pancreatic fibrogenesis. Gut 44, 534-541,
doi:10.1136/gut.44.4.534 (1999).

128
 Acknowledgements

9 Acknowledgements

I would like to first thank my supervisor PD Dr. Dr. Patrick Christian Hermann (Dept.
of Internal Medicine I, University of Ulm) for providing me with an excellent
opportunity to perform my doctoral studies under his guidance. I express my sincere
gratitude for his continuous support and also thank him for allowing me to contribute
to different intriguing research projects. His supervision and motivation modelled my
learning capabilities to autonomously design projects, collaborate, guide other
students in the lab as well as for writing this thesis. I am convinced that his input
and encouragement will help me further in my scientific career.

Beside my supervisor, I would also like to thank the whole Hermann Lab. In
particular, I would express my earnest gratitude to Dr. Karolin Walter for her
exceptional support, cooperation, remarkable patience and kind words for the last
four years. Additionally, I would like to thank Dr. Eva Rodriguez-Aznar for critiquing
together numerous publications and inculcating in me the scientific urge to read.
Moreover, I would also like to thank all the students I got the pleasure to supervise
and/or work together, Arijan, Tabea, Inaas, Anton, Selina and Steffi. Even though it
was my responsibility to help and guide them, I did learn a lot from them. They made
some of the most stressful moments within the lab very joyous. In addition, I like to
acknowledge Andrea for her support and technical assistance. Without all of your
help, I wouldn’t have been able to achieve such great aims during my PhD period.

I would also like to thank people from the Dept. of Internal Medicine I, specially,
Akshaya, Sabine, Jessica and Florian for lengthy discussions of both scientific and
non-scientific manner. It made my time working in the department cheerful and
memorable.

Special acknowledgements go to my other two Thesis Advisory Committee (TAC)

members, Prof. Dr. Lisa Wiesmüller (University of Ulm) and Prof. Dr. Bruno Sainz
Jr. (Universidad Autónoma de Madrid) for their constructive support at either the
intermediate evaluations or other scientific meetings. Moreover, I would also like to
thank them both for kindly agreeing to be part of my TAC and evaluating my PhD
thesis.

129
 Acknowledgements

Further, I would like to acknowledge Prof. Jan Münch and Mirja Harms (Institute of
Molecular Virology, University of Ulm) as well as Prof. Mika Lindèn, Bastian
Beitzinger and Roman Schmid (Institute of Inorganic Chemistry II, University of Ulm)
for their invaluable collaboration and continuous support.

Furthermore, I would like to thank CRC 1279 for an excellent collaborative research
program and funding of my position and research work. I further express my
gratitude to the International Graduate School in Molecular Medicine (IGradU) from
the University of Ulm for providing funding for attendances on conferences and
workshops. I thank the IGradU for the activities that allowed the opportunity to
develop my skills for my future career.

Moreover, I would like to fondly thank my family and all of my friends who
continuously motivated me and more importantly listened to me throughout my
whole study time.

And last, I would like to share my utmost love and heartfelt gratitude with Frank
Arnold. He, for the past 4.5 years, not only comforted, encouraged and consoled
me but also challenged me to become the best version of myself. I cannot thank you
enough.

130
 Statutory declaration

10 Statutory declaration

I hereby declare that I wrote the present dissertation with the topic:

Novel molecular insights to discern and target migrating cancer stem cells
in pancreatic ductal adenocarcinomas

independently and used no other aids that those cited. In each individual case, I
have clearly identified the source of the passages that are taken word for word or
paraphrased from other works.

I also hereby declare that I have carried out my scientific work according to the
principles of good scientific practice in accordance with the current „Satzung der
Universität Ulm zur Sicherung guter wissenschaftlicher Praxis“ [Rules of the
University of Ulm for Assuring Good Scientific Practice].

Ulm, 30.05.2022

Kanishka Tiwary

131