1. health and fitness
2. disease
3. cancer

NOVEL MECHANISMS OF STAT3 ACTIVATION
by
Rozanne Arulanandam
A thesis submitted to the Department of Pathology and Molecular Medicine
In conformity with the requirements for the degree of
Doctor of Philosophy
Queen’s University
Kingston, Ontario, Canada
(February/2010)
Copyright © Rozanne Arulanandam, 2010
Abstract
Stat3 (signal transducer and activator of transcription-3) is activated by a number of
receptor and non-receptor tyrosine kinases, while a constitutively active form of Stat3 alone is
sufficient to induce neoplastic transformation. Results presented in this thesis reveal that Stat3
can also be activated through homophilic interactions by the epithelial (E)-cadherin and cadherin11, two members of the classical type I and II cadherin family of surface receptors, responsible
for the formation of cell to cell junctions. Indeed, by plating cells onto surfaces coated with
fragments encompassing the two outermost domains of these cadherins, we definitively
demonstrate that cadherin engagement can activate Stat3, even in the absence of direct cell to cell
contact. At the same time, levels of the extracellular signal regulated kinase (Erk)1/2, which is
often coordinately activated by growth factor receptors and oncogenes, remain unchanged upon
cadherin ligation. Most importantly, we report, for the first time, an unexpected surge in total
Rac1 and Cdc42 protein levels, triggered by cadherin engagement, and an increase in Rac1 and
Cdc42 activity, which is responsible for the Stat3 stimulation observed. Inhibition of cadherin
interactions reduced Rac/Cdc42 and Stat3 levels and induced apoptosis, pointing to a significant
role of this pathway in cell survival signalling, a finding which could also have important
therapeutic implications.
To better understand the role of Rac/Cdc42 in the cadherin-mediated Stat3 activation, we
compared Stat3 activity in mouse HC11 cells before and after expression of the mutationally
activated, RacV12. We demonstrate a dramatic increase in protein levels and activity of both the
endogenous Rac and RacV12 with cell density, which was due to inhibition of proteasomal
degradation. Moreover, we clearly show that RacV12 expression can activate Stat3 through an
increase in expression of members of the IL6 family of cytokines, known potent Stat3 activators.
In fact, knockdown experiments indicate that gp130 receptor function, and Stat3 activation, are
ii
essential for the migration and proliferation of RacV12-expressing cells, thereby demonstrating
that the gp130/Stat3 axis represents an essential target of activated Rac in the regulation of both
of these fundamental cellular functions.
iii
Co-authorship
The data presented in this thesis result from collaborations with a number of researchers
from Queen’s University, including Dr. Adina Vultur, Dr. Jun Cao, Dr. Mulu Geletu, Dr. Bruce
Elliott, Dr. Peter F. Truesdell, Esther Carefoot, as well as Dr. Hélène Feracci and Dr. Sébastien
Chevalier, from the Université Bordeaux 1, Centre de Recherche Paul Pascal, Pessac, France, and
Dr. Lionel Larue, from the Institut Curie, Paris, France. The preface to each chapter describes the
contribution of each author to the work presented. Most of the results shown have been published
or are being submitted for publication. The full citation for each published manuscript is found at
the beginning of each chapter.
iv
Acknowledgements
I am forever grateful to my supervisor, Dr. Leda Raptis, for her mentoring, advice and
support. Her depth of knowledge, as well as her enthusiasm and exceptional dedication to
research, teaching and her students, represent to me a profound source of inspiration. I would
like to thank Dr. Adina Vultur for showing me the ropes, Dr. Jun Cao, Marilyn Garrett, Dr. Mulu
Geletu, Reva Mohan, Sam Greer, Chrystele Chaize and the past and present members of the
Raptis lab, who have been a true joy to work with. I am most appreciative of the constructive
advice, encouragement and reagents received from my committee members, Dr. Bruce Elliott and
Dr. Graham Côté, as well as Dr. Peter Greer. Over the course of my degree, I have thoroughly
enjoyed the opportunity to collaborate with Dr. Hélène Feracci and Dr. Olivier Courjean, who
provided us with several nice batches of cadherin fragments, and Dr. Lionel Larue who sent us
the mouse embryonic stem cells as well as precious lots of serum that caused much grief with
customs. I owe many thanks to Jerry Dering, Chris Boer, Emilia Furmaniak-Kazmierczak and
Karoline Fisher for their technical assistance and expertise. I would also like to acknowledge
members of the Microbiology and Pathology departments, especially John Wu, Dr. Simon Weli,
Dr. Sébastien Fraud, Dr. Ray Amith, Dr. Preethi Jayanth, Victoria Au, Jalna Meens, Sara
Sharifpoor, Blerta Starova and Maike Bossert, for their continuing friendship, support and advice.
I am immensely grateful to God for granting me strength, perseverance, and a thirst for
knowledge, and above all for blessing me with a truly exceptional family. This thesis is dedicated
to Mark, whose patience, understanding, sacrifice and devotion have guided me through this
degree, to my mother, Cynthia, for her continuing faith, support and more reasons than I could
ever fit on this page, and to Anjali, my greatest accomplishment and the source of my inspiration
v
and joy. I would also like to thank my grandparents, our dear aunt Maria Ralli, my aunt and
uncle, friends and cousins, especially Geetha, for all their help, encouragement and prayers.
This work is supported by a Doctoral and Master’s award from the Canadian Institutes of
Health Research (CIHR) and the Women’s Health Scholars Award from the Ontario Council of
Graduate Studies (OCGS).
vi
Table of Contents
Abstract............................................................................................................................................ii
Co-authorship..................................................................................................................................iv
Acknowledgements.......................................................................................................................... v
Table of Contents...........................................................................................................................vii
List of Figures .................................................................................................................................xi
List of Tables ................................................................................................................................xiv
Abbreviations................................................................................................................................. xv
Chapter 1 – Introduction............................................................................................................... 1
Chapter 2 – Literature Review ..................................................................................................... 4
1. Signal Transducers and Activators of Transcription (STATs) ................................................ 5
STAT activation pathways .................................................................................................. 5
Structure of STATs .............................................................................................................. 7
Nuclear trafficking of STATs ............................................................................................ 10
The multifunctional role of Stat3 ...................................................................................... 11
Regulation of Stat3 signalling........................................................................................... 13
Stat3 and cancer ............................................................................................................... 15
2. The Cadherin Family of Cell Adhesion Receptors ................................................................ 18
Structure of classical cadherins and mechanism of adherens junction formation ........... 19
Cytoplasmic interactions of cadherins.............................................................................. 25
Regulation and role of E-cadherin in cancer ................................................................... 26
Cadherin switching in cancer ........................................................................................... 28
3. Rho Family GTPases ............................................................................................................. 29
Regulation of Rho GTPases .............................................................................................. 30
Structural features of Rho GTPases ................................................................................. 32
Rho GTPases and cell-cell adhesion ................................................................................ 35
Rho GTPases and Stat3 .................................................................................................... 38
4. The gp130/JAK/Stat3 axis ..................................................................................................... 41
Cytokines and the JAK-STAT pathway ............................................................................. 41
vii
Janus kinases (JAKs) ........................................................................................................ 42
IL6-type cytokines and the gp130 receptor....................................................................... 43
IL6 and cancer .................................................................................................................. 48
5. Previous results from our lab: cell to cell adhesion activates Stat3 ....................................... 51
Hypothesis and Objectives............................................................................................................. 53
Significance ................................................................................................................................... 55
References...................................................................................................................................... 56
Chapter 3 – Transfection techniques affecting Stat3 activity levels........................................ 72
Abstract.......................................................................................................................................... 73
Introduction.................................................................................................................................... 74
Materials and Methods................................................................................................................... 76
Cell lines and culture techniques ...................................................................................... 76
Western blotting................................................................................................................ 77
Luciferase assays .............................................................................................................. 77
In situ electroporation ....................................................................................................... 78
Results............................................................................................................................................ 79
Calcium phosphate transfection activates Stat3................................................................ 79
Lipofectamine transfection activates Stat3 ....................................................................... 82
Retroviral infection does not affect Stat3 phosphorylation .............................................. 82
In situ electroporation does not affect Stat3 phosphorylation or transcriptional activity . 84
Discussion ...................................................................................................................................... 87
References...................................................................................................................................... 88
Chapter 4 – Cadherin engagement promotes cell survival via Rac/Cdc42 and Stat3 ........... 90
Abstract.......................................................................................................................................... 91
Introduction.................................................................................................................................... 92
Materials and Methods................................................................................................................... 95
Cell lines and culture techniques ...................................................................................... 95
Cadherin recombinant fragments and peptides................................................................. 95
Gene expression................................................................................................................ 96
Lentivirus vector production............................................................................................. 97
Immunofluorescence and apoptosis assays....................................................................... 97
viii
Immunohistochemical staining for E-cadherin and Stat3-ptyr705 ................................... 98
Western blotting and immunoprecipitation ...................................................................... 99
Reverse transcriptase-polymerase chain reaction (RT-PCR) assays .............................. 100
Results.......................................................................................................................................... 102
E-cadherin engagement can activate Stat3 ..................................................................... 102
E-cadherin engagement is sufficient to activate Stat3 .................................................... 107
Cadherin-11 engagement in Balb/c 3T3 can also activate Stat3..................................... 112
Cell to cell adhesion increases protein levels and activity of Rac1 and Cdc42 .............. 116
E-cadherin and cadherin-11 engagement can increase protein levels and activity of Rac1
and Cdc42 ....................................................................................................................... 121
Rac1 and Cdc42 are required for the cell to cell adhesion-mediated, Stat3 activation... 122
JAKs are required for the cell to cell adhesion-mediated, Stat3 activation .................... 126
Cell to cell adhesion triggers cytokine gene expression ................................................. 128
Inhibition of E-cadherin engagement at confluence promotes apoptosis ....................... 129
Discussion .................................................................................................................................... 135
References.................................................................................................................................... 140
Chapter 5 – Activated Rac requires gp130 for Stat3 activation, cell proliferation and
migration..................................................................................................................................... 145
Abstract........................................................................................................................................ 146
Introduction.................................................................................................................................. 147
Materials and Methods................................................................................................................. 149
Cell lines, culture techniques and gene expression......................................................... 149
Western blotting and immunoprecipitation .................................................................... 150
qRT-PCR assays ............................................................................................................. 151
Results.......................................................................................................................................... 152
Cell density inhibits cRac1 proteasomal degradation ..................................................... 152
Cell density upregulates activated RacV12 levels through inhibition of proteasomal
degradation...................................................................................................................... 158
Mutationally activated RacV12 and Cdc42V12 can activate Stat3 ..................................... 160
NFκB and JAKs are required for the RacV12-mediated, Stat3 activation ........................ 161
RacV12 triggers cytokine gene expression and requires gp130 for Stat3 activation ........ 163
RacV12-induced IL6 upregulation is unable to activate Erk1/2 in confluent cultures..... 164
ix
RacV12-induced cell proliferation and migration require IL6 family cytokines .............. 167
Discussion .................................................................................................................................... 170
References.................................................................................................................................... 174
Chapter 6 – Discussion .............................................................................................................. 178
Direct cadherin engagement activates Stat3 ................................................................................ 179
Cadherin engagement upregulates Rac1 and Cdc42 protein levels through inhibition of
proteasomal degradation .............................................................................................................. 181
Mutationally activated Rac and Cdc42 are also targeted to the proteasome, in the
absence of cell-cell contact............................................................................................. 181
IL6 family cytokines transmit the signal from activated Rac1 to Stat3 ....................................... 184
Rac1 and Cdc42 transduce the cadherin signal to Stat3 .............................................................. 185
Cell-cell adhesion induces the autocrine production of IL6 family cytokines ................ 185
Calcium phosphate and lipofectamine transfection increase Stat3 activity ................................. 188
Role of the cadherin/Rac1/gp130/JAK/Stat3 axis in cell survival and cancer............................. 189
Conclusions.................................................................................................................................. 192
References.................................................................................................................................... 199
Appendix..................................................................................................................................... 206
References.................................................................................................................................... 237
x
List of Figures
Figure 2.1: Classical model of Stat3 activation .............................................................................. 6
Figure 2.2: Domain structure of Stat3............................................................................................. 8
Figure 2.3: General structure of classical cadherins ..................................................................... 21
Figure 2.4: Proposed mechanisms for cadherin adhesion............................................................. 23
Figure 2.5: Ribbon drawing of the trans dimer structure of human E-cadherin EC1-EC2 .......... 24
Figure 2.6: Regulation of Rho GTPases ....................................................................................... 31
Figure 2.7: Structure of Rho GTPases .......................................................................................... 34
Figure 2.8: Activation of Rac1 by E-cadherin-mediated cell-cell adhesion involves PI3-kinase. 37
Figure 2.9: Role of the Rac1 and Cdc42 effector, IQGAP, in E-cadherin-mediated cellcell adhesion............................................................................................................... 39
Figure 2.10: Domain structure of JAKs........................................................................................ 44
Figure 2.11: Receptor complexes of the IL6-type family of cytokines ........................................ 46
Figure 2.12: Extracellular structure of the gp130 receptor, which binds IL6-type cytokines....... 47
Figure 2.13: Cytoplasmic domain of gp130 in signal transduction .............................................. 49
Figure 3.1A-D: Calcium-phosphate transfection activates Stat3.................................................. 80
Figure 3.2A-B: Effect of lipofectamine and retroviral infection upon Stat3 activity ................... 83
Figure 3.3A-C: In situ electroporation does not affect Stat3 phosphorylation or
transcriptional activity.......................................................................................... 85
Figure 4.1A-B: Cell density triggers a dramatic activation of Stat3 in HC11 cells.................... 103
Figure 4.2A-C: Disruption of cadherin engagement by soluble cadherin fragments or synthetic
peptides reduces Stat3-tyr705 phosphorylation and activity............................. 105
Figure 4.3A-B: E-cadherin ablation inhibits the density-mediated, Stat3 activation ................ 108
Figure 4.4A-B: Expression of Stat3-ptyr705 and E-cadherin in normal mouse breast ductal
epithelium ......................................................................................................... 109
Figure 4.5A-B: E-cadherin engagement is sufficient to activate Stat3....................................... 111
Figure 4.6A-D: Cadherin-11 engagement can trigger Stat3 phosphorylation in Balb/c 3T3 cells
.......................................................................................................................... 114
Figure 4.7A-B: Cadherin-11 engagement is sufficient to activate Stat3 in Balb/c 3T3 fibroblasts
.......................................................................................................................... 115
Figure 4.8A-B: Cell density increases the activity as well as total protein levels of Rac1 and
Cdc42 in HC11 cells ......................................................................................... 117
xi
Figure 4.9A-B: E-cadherin engagement increases Rac1 protein levels and activity in the absence
of direct cell to cell contact ............................................................................... 118
Figure 4.10A-C: Cadherin-11 increases Rac1/Cdc42 activity and protein levels in Balb/c 3T3
cells .................................................................................................................. 119
Figure 4.11A-C: Inhibition of Rac1 or Cdc42 prevents the density-induced, Stat3 activation .. 123
Figure 4.12A-B: JAKs are required for the cadherin-dependent Stat3-tyr705 phosphorylation 127
Figure 4:13A-C: Cell to cell adhesion triggers cytokine gene expression ................................. 130
Figure 4.14A-D: E-cadherin engagement promotes cell survival............................................... 132
Figure 5.1A-D: RacV12 and Cdc42 V12 activate Stat3 in HC11 cells at all densities.................... 153
Figure 5.2A-E: Cell density inhibits the proteasomal degradation of Rac in HC11 cells .......... 155
Figure 5.3A-C: Cell density inhibits the proteasomal degradation of Rac in Balb/c 3T3
fibroblasts.......................................................................................................... 157
Figure 5.4A-C: Rac V12-dependent Stat3-ptyr705 phosphorylation requires JAK and NFκB .... 162
Figure 5.5A-B: gp130 is required for the Rac V12-mediated, Stat3 activation............................. 165
Figure 5.6A-C: Rac-induced cell migration and proliferation require gp130 ............................ 168
Figure 5.7: Proposed model for Rac/Cdc42-mediated, Stat3 activation..................................... 173
Figure 6.1: Model of Stat3 activation by cadherin engagement ................................................. 195
Appendix Fig. A.1: The effect of cell density on Stat3 activity ................................................. 207
Appendix Fig. A.2: Weak Stat3 activation in ES cells by the LIF present in the growth medium
................................................................................................................... 208
Appendix Fig. A.3: ES cells do not differentiate when cultured to high densities, in the presence
of LIF ......................................................................................................... 209
Appendix Fig. A.4: E-cadherin ablation inhibits the density-mediated, Stat3 activation........... 210
Appendix Fig. A.5: Plating HC11 cells on cadherin-coated surfaces for more than 10 hours or at
high cell densities does not activate Stat3.................................................. 212
Appendix Fig. A.6: E-cadherin engagement is sufficient to activate Stat3 in ES cells.............. 213
Appendix Fig. A.7: Stat3 activation after plating different numbers of cells for 24 hours ........ 214
Appendix Fig. A.8: Toxin B can inhibit the induction of the density-mediated, Stat3 tyr705
phosphorylation.......................................................................................... 215
Appendix Fig. A.9: Reduction of Rac1 or Cdc42 levels or activity prevents Stat3-tyr705
phosphorylation in mouse 10T1/2 fibroblasts............................................ 216
xii
Appendix Fig. A.10: Reduction of Cdc42 levels or activity prevents Stat3-tyr705
phosphorylation......................................................................................... 218
Appendix Fig. A.11: Rac V12 and Cdc42 V12 activate Stat3 in 10T½ cells at all densities........... 220
Appendix Fig. A.12: Rac-dependent Stat3-ptyr705 phosphorylation requires JAK and NFκB
in Balb/c 3T3 cells ................................................................................... 221
Appendix Fig. A.13: Stat3 can be phosphorylated at critical tyrosine 705 in E-cadherin deficient
cell lines ................................................................................................... 222
Appendix Fig. A.14: N-cadherin expression increases Stat3 tyr705 phosphorylation in ES cells
.................................................................................................................. 223
Appendix Fig. A.15: E-cadherin levels increase with cell density............................................. 224
xiii
List of Tables
Table 1.1: Cadherin switching during tumorigenesis ................................................................... 29
Table 4.1: Stat3 downregulation promotes apoptosis in confluent HC11 cells .......................... 125
Appendix Table A.1, A-B: Quantitation of Rac1, Cdc42 and Stat3 levels in HC11 cells and
their derivatives with reduced Rac1 or Cdc42 activity or protein
levels ............................................................................................... 225
Appendix Table A.2A-B: qRT-PCR array for cytokines secreted by densely-growing HC11
or Balb/c 3T3 cells ........................................................................... 227
Appendix Table A.3: Stat3 tyr-705 phosphorylation and apoptosis following cadherin
ablation or inhibition with peptides......................................................... 232
Appendix Table A.4: shRNA sequences.................................................................................... 234
Appendix Table A.5: qRT-PCR array for cytokines secreted by H-RacV12 cells ....................... 235
xiv
Abbreviations
AJ
adherens junctions
APC
adenomatous polyposis coli
β-gal
β-galactosidase
BES
N,N’-bis[2Hydroxyethyl]-2-aminoethanesulfonic acid
C/EBPδ
CCAAT/enhancer binding protein δ
CAPE
caffeic acid phenethyl ester
CBP
cAMP-response element-binding protein (CREB) binding protein
CHR
cytokine binding homology region
CLC
cardiotrophin-like cytokine
CMV
cytomegalovirus
CNF
cytotoxic necrotizing factor
CNTF
ciliary neurotrophic factor
CR
cadherin-specific repeats
CRIB
Cdc42/Rac-interactive binding
CRM1
chromosome region maintenance 1
CT-1
cardiotrophin-1
DAB
diaminobenzidine
DH
dbl-homology
E2F
elongation factor-2
EC
extracellular region of classical cadherins
ECL
enhanced chemiluminescence
E-cadherin
epithelial cadherin
EGFP
enhanced green fluorescent protein
EGFR
epidermal growth factor receptor
EMT
epithelial mesenchymal transition
ERK
extracellular signal regulated kinase
ES
embryonic stem cell
FACS
fluorescence activated cell sorting
FAK
focal adhesion kinase
FERM
a band four-point-one, ezrin, radixin, moesin)
FGF
fibroblast growth factor
xv
FNIII
fibronectin type III
GAP
GTPase activating protein
GAPDH
glyceraldehyde-3-phosphate dehydrogenase
GAS
gamma activation sequence
GDP
guanosine 5’-diphosphate
GEFs
guanine nucleotide exchange factors
GRIM-19
genes Associated with Retinoid IFN Induced Immortality
GSK-3β
glycogen synthase-3β
GST
glutathione-S-transferase
GTP
guanosine 5’-triphosphate
HCl
hydrochloric acid
HGF
hepatocyte growth factor
HHV-8
human herpes virus-8
HNSCC
head and neck squamous cell carcinoma
HRP
Horseradish Peroxidase
Hrs
HGF-regulated substrate
hSIE
high affinity mutant, sis-inducible-element
Hsp90
Heat shock protein 90
IFN
interferon
IKK
IκB kinase
Ig
immunoglobulin
IL
interleukin
IRF
interferon regulatory factor
ITO
indium-tin oxide
JH
janus homology
JNK
c-Jun N-terminal kinase
LEF/TCF
lymphoid enhancer factor/T-cell factor
LIF
leukaemia inhibitory factor
LNCaP
lymph node carcinoma of the prostate
MAPK
mitogen activated protein kinases
M-cadherin
myotubule-cadherin
MgcRacGAP
male germ cell Rac1 GAP
MDCK
Madin-Darby canine kidney
xvi
Mlk
mixed lineage kinase
N-cadherin
neural-cadherin
NES
nuclear export sequences
NFκB
nuclear factor-κB
NLS
nuclear localization sequence
NPC
nuclear pore complex
NSCLC
non-small cell lung carcinoma
OB-cadherin
osteoblast cadherin
OSM
oncostatin M
PAK
p21-activated kinase
PARP
poly (ADP-ribose) polymerase
PBD
PAK-binding domain
P-cadherin
placental cadherin
PDGF(R)
platelet-derived growth factor receptor
pfu
plaque forming units
PH
plekstrin homology
PI3-kinase
phosphatidyl-inositol-3 kinase
PIAS
protein inhibitors of activated STATs
POSH
plenty of SH3 domains
pser
phosphorylated serine
PTP
protein tyrosine phosphatases
ptyr
phosphorylated tyrosine
RA
retinoid acid
RANK-L
receptor activator of NFκB (RANK) ligand
R-cadherin
retinal-cadherin
Rho-GDIs
Rho guanine nucleotide dissociation inhibitors
Rm
Rhesus macaque rhadinovirus
RT-PCR
reverse transcriptase-polymerase chain reaction
SDS-PAGE
sodium dodecyl sulphate-polyacrylamide gel electrophoresis
SH2
Src-homology-2
SLIM
STAT-interacting LIM protein
Smurf
Smad-ubiquitin regulatory factor
SOCS
suppressors of cytokine signaling
xvii
SRE
serum response element
SRF
serum response factor
STAT
signal transducer and activator of transcription
SUMO
small ubiquitin-related modifier
TBS
tris-buffered saline
TCCR
T cell cytokine signaling receptor
TC-PTP
T-cell-PTP
TGF
Transforming growth factor
TNF
tumor necrosis factor
TUNEL
terminal deoxynucleotidyl transferase dUTP nick end labeling
Tyk2
tyrosine kinase-2
VE-cadherin
vascular endothelial-cadherin
VEGF
vascular endothelial growth factor
xviii
Chapter 1
Introduction
1
Human cancer invariably occurs through the accumulation of stepwise mutations that
allow normal cells to effectively evade homeostasis and continue proliferating. The mitotic
decision is a highly regulated process, controlled by the coordinated integration of molecular
signalling networks. Cell surface receptor activation through ligand binding or the interaction
between adhesion molecules can communicate the status of the extracellular environment to the
nucleus, where target gene regulatory proteins become altered, thereby generating an appropriate
response from the cell concerning proliferation, differentiation or apoptosis (Weinberg, 2007;
Juliano, 2002). Cytoplasmic modulators of signal transduction pathways include small GTPases
and tyrosine or serine/threonine kinases, or phosphatases, which are often aberrantly expressed in
cancer. A number of different mitogenic pathways can often converge onto a limited set of
transcription factors, the final switches that control the gene expression patterns that ultimately
lead to malignancy (Yu and Jove, 2004). Hence, targeting a single transcription factor may block
the effects of several upstream mutations that cause its persistent activation.
It has emerged over the last decade that the Signal Transducer and Activator of
Transcription-3 (Stat3) can be a critical point of convergence for many intercellular pathways that
are often exploited in malignancy (Yu and Jove, 2004; see Figure 2.1). Hyperactive Stat3 is
present in a large number of cancers, and is required for tumor cell growth and survival, as well
as angiogenesis, metastasis and immune evasion (Bromberg et al, 1999; Bowman et al., 2001, Yu
and Jove, 2004; Yu et al., 2009). Stat3 inhibition in tumors was shown to induce apoptosis while
having little effect upon normal tissues, possibly because tumor cells may have become
irreversibly dependent on Stat3 signalling to sustain their growth and survival, while normal cells
may be able to use alternate pathways to compensate for Stat3 loss (Yu and Jove, 2004). As a
result, drugs inhibiting Stat3 may be specific for the tumor, with little effect upon normal tissues.
It follows that a comprehensive understanding of Stat3 signalling mechanisms is indispensable
for the development of effective anticancer therapies.
2
Cells in normal tissues or tumors have extensive opportunities for adhesion to their
neighbors in a three-dimensional organization, and it recently became apparent that in the study
of fundamental cellular processes it is important to take into account the effect of surrounding
cells. The relevance of interactions observed in densely growing, cultured cells to human cancer
is clearly evident from findings demonstrating a close correspondence of genes expressed in
prostate cancer LNCaP cells cultured to high, but not low, densities, with genes associated with
prostate cancer in vivo (Chen et al., 2006).
We recently discovered the existence of a novel pathway of activation of Stat3, brought
about by cell density. This Stat3 activation was triggered by cell to cell adhesion and resistant to
inhibition of a number of tyrosine kinases often activated in a variety of cancers. We now further
examine the molecular details of this cascade, to enable the identification of prognostic markers
and promote the validation of this pathway as a target for drug design, especially for the treatment
of tumors which may be resistant to inhibition of many tyrosine kinase oncogenes known to be
hyperactive in many cancers.
3
Chapter 2
Literature Review
4
1. Signal Transducers and Activators of Transcription (STATs)
STAT activation pathways
James E. Darnell and colleagues cloned the first members of the STAT family by
purifying factors bound to promoters of interferon (IFN)-inducible genes (Fu et al., 1992). The
IFN-α ligand-induced complex consisted of a 91 kDa protein identified as Stat1 and a 113 kDa
protein termed Stat2, whereas IFN-γ stimulation seemed to trigger Stat1 homodimerization
(Schindler and Darnell, Jr., 1995; Xu et al., 1996). Since then seven STAT family members have
been identified in mammalian cells: Stat1, Stat2, Stat3, Stat4, Stat5a, Stat5b and Stat6 (Kisseleva
et al., 2002).
STATs are a family of latent transcription factors that become activated upon tyrosine
phosphorylation. STAT activators include growth factor receptors with intrinsic kinase activity
such as epidermal growth factor receptor (EGFR) or platelet-derived growth factor receptor
(PDGFR), as well as cytokine receptors which are non-covalently associated with the JAK family
of tyrosine kinases (Darnell, Jr., 1997; Murray, 2007; Yu and Jove, 2004). Cytoplasmic Src
kinases and the Bcr-Abl oncoprotein were also shown to phosphorylate certain STATs
independent of receptor engagement (Yu and Jove, 2004). Phosphorylated STATs dimerize via
reciprocal phosphorylated tyrosine (ptyr)-Src-homology-2 (SH2) interactions and migrate to the
nucleus where they bind target regions of DNA to regulate the transcriptional activation of genes
involved in immune function, development, differentiation, proliferation and survival (Darnell,
Jr., 1997; Murray, 2007; Yu and Jove, 2004). The classical model of Stat3 activation is presented
in Figure 2.1.
5
Figure 2.1: Classical model of Stat3 activation:
Stat3 (red) is an important point of convergence for signalling pathways often activated in cancer.
Binding of growth factors (a) or cytokines (b) to their receptors results in the stimulation of the
intrinsic receptor tyrosine-kinase activity or that of receptor-associated kinases, such as JAKs
(blue) or cSrc (pink), which may phosphorylate the cytoplasmic tail of the receptor. Cytoplasmic
Stat3 then binds onto the activated receptor with its SH2 domain, is phosphorylated by the JAK
kinases, dimerized through reciprocal SH2-ptyr interactions and translocates to the nucleus where
it activates transcription of a number of genes involved in cell division or survival. Stat3 also
downregulates p53 by direct binding to the p53 promotor (Niu et al., 2005). (c) Mutationally
activated Src, or other non-receptor tyrosine kinases such as Abl or Btk autophosphorylate and
activate Stat3 in a similar manner (Zhang et al., 2000).
Figure adapted from Yu and Jove, 2004.
6
Structure of STATs (Figure 2.2)
STAT proteins contain five structurally and functionally conserved domains and a
variable carboxy (C)-terminal end. The amino (N)-terminal region is believed to be involved in
the oligomerization of STAT dimers, which can promote cooperative binding to tandem DNA
binding elements resulting in more efficient transcriptional control (Kisseleva et al., 2002;
Bowman et al., 2001). The coiled-coiled region of STATs consists of four alpha helices which
form a large hydrophilic surface that can interact with proteins involved in transcriptional
regulation, for example the Interferon Regulatory Factor (IRF) family of DNA binding proteins
which are important in IFN signalling responses (Kisseleva et al., 2002; Horvath et al., 1996).
The DNA binding domain allows for direct association of STATs at various promoter regions.
Most STATs recognize an 8-10 base pair consensus sequence of 5’ –TT (N4-6)AA – 3’ referred to
as the GAS (gamma activation sequence) element, as it was initially characterized as the IFN-γ
activation sequence recognized by Stat1 homodimers (Xu et al., 1996). The SH2 domain can
recruit STATs to specific ptyr motifs on ligand-activated receptors and also mediates STAT
homo- or heterodimerization by recognizing, once phosphorylated, the preserved tyrosine, located
approximately 700 residues from the N-terminus (Kisseleva et al., 2002).
The carboxy-terminal domain of STATs is variable among members and encodes a
transactivation domain (Kisseleva et al., 2002). Stat1, Stat3, Stat4, Stat5a, Stat5b and Stat6 can
become phosphorylated on a serine (ser) residue within this region (pser727 in the case of Stat1
and Stat3), and this is believed to be required for maximal transcriptional activation (Darnell, Jr.,
1997; Wick and Berton, 2000). In fact, ser727 phosphorylation was shown to be required for full
neoplastic transformation by the constitutively active form of Stat3, Stat3C (Bromberg et al.,
1999). Stat1β and Stat3β are naturally occurring splice variants which lack the transactivation
domain as well as this serine residue; they can often block the function of the full-length proteins
in a dominant negative fashion (Caldenhoven et al., 1996; Bowman et al., 2001). Darnell’s group
7
Figure 2.2: Domain structure of Stat3
STAT proteins contain five structurally and functionally conserved domains: the Nterminal domain is involved in oligomerization of STAT dimers, the coiled-coiled
domain mediates interaction with other regulatory proteins, the DNA binding domain
enables direct association with STAT binding sites at promoter regions, the linker
region connects the DNA binding and SH2 domains, and the SH2 domain binds
phosphorylated tyrosine motifs of activated receptors. Approximately 700 residues
from the N-terminal domain of STATs is a preserved tyrosine residue, tyr705 in the
case of Stat3, which becomes phosphorylated by kinases and allows for STAT
dimerization through reciprocal SH2-ptyr interactions. The carboxy-terminal
transcriptional activation domain is variable among STAT family members and
contributes to STAT specificity. This C-terminal domain also encompasses a site of
serine phosphorylation, ser727 for Stat3, which is believed to be required for maximal
transcriptional activity of all STAT members, with the exception of Stat2 (Kisseleva et
al., 2002). Figure adapted from Yu and Jove, 2004.
8
claims that although serine phosphorylation of Stat1 and Stat3 may enhance gene expression by
these STATs, is not required for their DNA binding (Wen and Darnell, Jr., 1997). Further work
has demonstrated that ser727 phosphorylation of Stat1 may instead optimize transcriptional
activity by increasing its binding affinity for nuclear coregulatory proteins such as MCM5, a
member of the mini-chomosome maintenance family involved in initiating DNA replication
(Zhang et al., 1998).
While tyrosine phosphorylation of STATs takes place at the plasma membrane, their
phosphorylation on serine has been shown to occur in the cytoplasm (Darnell et al., 1997).
Reports have claimed that the Mitogen Activated Protein Kinases (MAPKs) such as p38, c-Jun
N-terminal kinase (JNK) or ERK can trigger serine phosphorylation of certain STATs, thus
enabling cross-talk between signalling cascades (Decker and Kovarik, 2000). More specifically,
Turkson et al. have demonstrated that Ras- and Rac1- mediated p38 and JNK signalling can lead
to Stat3, ser727 phosphorylation, and this is required for Stat3 activation by the Src kinase
(1999).
STAT proteins typically exhibit long half-lives, however their stability may be affected
by ubiquitin-modification, which can lead to the degradation of polyubiquitinated STATs by the
26S proteasome (Ulane et al., 2003; Tanaka et al., 2005). E3 ubiquitin ligases are the largest
group of enzymes involved in the ubiquitination process and act as the key determinants of
reaction specificity by catalyzing the transfer of ubiquitin to lysine residues of the substrate
protein (Finley, 2009). The V protein of Mumps virus plays an important role in host immune
evasion, and this may be attributed to its ability to act as an E3 ligase for Stat1, Stat2 or Stat3,
resulting in the suppression of cytokine signaling (Ulane et al., 2003; Horvath, 2004).
Interestingly, by removing Stat3 from cells, Mumps virus has also demonstrated clear oncolytic
properties and high potential for use in cancer therapy (Ulane et al., 2003). In the nucleus, Stat4
has been shown to associate with the STAT-interacting LIM protein (SLIM) E3 ligase which can
result in STAT ubiquitination and degradation (Tanaka et al., 2005).
9
Nuclear trafficking of STATs
In order to exert their function as transcriptional regulators, STAT members must transit
nuclear pore complexes (NPC) within the nuclear membrane. Proteins larger than 50 kDa,
including most extracellular signal-activated transcription factors, generally achieve this by
binding to soluble carriers of the karyopherin-β family (Reviewed in: Reich and Liu, 2006; Terry
et al., 2007; Herrmann et al., 2007). In general, transcription factors, or binding partners of the
transcription factor acting as chaperones, may associate with any of six characterized importin-α
adaptor proteins (1-6), through a specific motif called the nuclear localization sequence (NLS).
The importin-α complex then interacts with the karyopherin-β member, importin-β1, which
mediates docking of the protein cargo at the NPC where translocation occurs through an energydependant active transport mechanism dependant on the NPC-associated Ran GTPase (Terry et
al., 2007). While conventional NLSs, consisting of a single stretch or bipartite sequence of basic
amino acids, have not yet been identified for all STATs, several of the IRFs and other interacting
factors which bind to STATs have shown to contain known or putative NLS and may assist in
directing them to the nucleus (Reich and Liu, 2006; Subramanian et al., 2001). Interestingly, in
the case of Stat1, a conditional NLS has been uncovered, which can be recognized by importinα5 member only following its tyrosine phosphorylation and dimerization (Sekimoto et al., 1996;
McBride et al., 2002; Reich and Liu, 2006). In contrast, tyrosine phosphorylation is not a
prerequisite for nuclear import of Stat3 (Liu et al., 2005). Unexpectedly, studies have revealed
that Stat3 can enter the nucleus even in an unphosphorylated state and dynamically shuttles
between cytoplasmic and nuclear compartments (Yang et al., 2005). Indeed, Stat3 was found to
contain a constitutive NLS within its coiled-coiled domain which can be recognized by importinsα3 and -α6 regardless of its phosphorylation status (Liu et al., 2005). The role of
unphosphorylated Stat3 in the nucleus is presently unclear. Controversial data have revealed that
Stat3 with a mutation in tyr705 can retain transcriptional activity to affect a different subset of
10
genes than wild type Stat3 (Yang et al., 2005). Others have identified breast tumor kinase (btk)
as a nuclear tyrosine kinase able to phosphorylate and activate Stat3 (Liu et al., 2006). Nuclear
export of STATs is believed to be regulated by Nuclear Export Sequences (NES), which may be
recognized by the exportin carrier, chromosome region maintenance 1 (CRM1; Reich and Liu,
2006).
The multifunctional role of Stat3
Stat3 was originally discovered as a mediator of the acute phase response, activated by
Interleukin-6 (IL6) in the liver (Wegenka et al., 1993). Stat3 is ubiquitously expressed in most
tissues and its phosphorylation can be brought about by activated receptors of cytokines, growth
factors and hormones as well as through intracellular kinases (reviewed in: Groner et al., 2008;
Yu and Jove, 2004; see Figure 2.1). The most common ligands known to activate Stat3 include
IL6, IL10, EGF and PDGF (Groner et al., 2008).
Of all the Stat genes, only Stat3 knockouts display embryonic lethality, at the gastrulation
phase, which argues for a crucial role in development (Takeda et al., 1997). Tissue-specific
inhibition of Stat3 using the Cre-loxP recombination system adapted from the P1 bacteriophage
has been used to further characterize Stat3 function (reviewed in: Akira et al., 2000; Levy and
Lee, 2002). For this, paired loxP sites can be inserted into introns surrounding critical regions of
Stat3, such as the site for tyrosine phosphorylation (Levy and Lee, 2002). Introduction of tissue
specific promoters driving the expression of Cre recombinase enables targeted deletion of the
DNA segment flanked by the loxP sites (Levy and Lee, 2002). This approach can be used to
generate mice that are viable despite the absence of Stat3 in selected tissues. Expression of the
Cre recombinase gene under the keratin-5 promoter has uncovered a role for Stat3 in epithelial
tissues, more specifically, in hair and skin remodeling, as well as wound healing, likely in
response to growth factors such as EGF and PDGF (Sano et al., 1999; Kisseleva et al., 2002).
Further studies from DiGiovani’s lab demonstrating that skin-specific Stat3-deficient mice were
11
resistant to chemically-induced skin carcinogenesis offered support to their model indicating that
Stat3 plays a critical role in the initiation, promotion and progression of epithelial carcinogenesis
(Chan et al., 2004; Kim et al., 2007). Paradoxically however, studies where Cre was expressed
under the β-lactoglobulin promoter have revealed that loss of Stat3 in the mammary gland can
delay involution and epithelial apoptosis after lactation (Clevenger, 2004). In this case Stat3 may
be inducing expression of CCAAT/enhancer binding protein δ (C/EBPδ), shown to activate proapoptotic genes such as p53 (Thangaraju et al., 2005).
Stat3 is also a key player in modulating immune responses. Stat3 deletion in T cells was
shown to impair IL6-mediated T-cell survival (Takeda et al., 1998). Stat3 deficiency in
macrophages and neutrophils can cause deregulated cytokine expression where IL10 production
is reduced allowing for an increase in TNFα-mediated inflammation (Riley et al., 1999).
Myeloid cell-specific Stat3 deficiency also enhances the susceptibility of these mice to bacterial
infection resulting in endotoxic shock (Takeda et al., 1999). Furthermore, certain
paramyxoviruses, such as mumps or measles, have evolved, as a means of escaping the host
response to infection, the ability to target Stat3 for degradation, revealing a vital role for Stat3 in
antiviral immunity (Horvath, 2004).
In cell culture systems, Stat3 is required to maintain the pluripotency of mouse
embryonic stem cells by preventing their differentiation (Raz et al., 1999). Stat3 can however
promote the differentiation of a variety of lines including lymphocytes, granulocytes and
astrocytes (reviewed in: Hirano et al., 2000). Stat3 has also been linked to B lymphocyte
proliferation by inhibiting the anti-apoptotic gene bcl-2, and a similar survival function has been
attributed to Stat3 in motor and sensory neurons in vitro, as few of these cells, following Stat3
knockdown, survive in vivo (Heinrich et al., 1998; Alonzi et al., 2001; Levy and Lee, 2002). In
sharp contrast, Stat3 was shown stimulate the growth arrest and terminal differentiation of
12
monocytes by downregulating c-myc and c-myb and upregulating junB and IRF-1 (Yamanaka et
al., 1996; Hirano et al., 2000; Levy and Lee, 2002).
In sum, Stat3 can induce a wide variety of responses including proliferation and growth
arrest, cell division and differentiation, as well as survival and apoptosis. This may be due to the
upregulation of a different set of Stat3 target genes in different cell types (Levy and Lee, 2002).
Nevertheless, Stat3 seems to be a central mediator of key cellular functions and efficient control
of its signalling is vital to maintaining homeostasis.
Regulation of Stat3 signalling
Positive regulation of Stat3 is thought to occur through the MAPKs which mediate its
phosphorylation on serine, thus enabling cross-talk between signalling cascades (Turkson et al.,
1999; Kisseleva et al., 2002). Studies have also revealed an association between Stat3 and
various proteins which facilitate transcription through chromatin modification, such as the
transcriptional coactivators CREB binding protein (CBP) and p300 (Yuan et al., 2005). Further
research has demonstrated that CBP/p300 can also acetylate Stat3 at lys685, and this can
contribute to Stat3 dimerization and activation (Yuan et al., 2005).
To prevent overstimulation, three main classes of negative regulators of STAT signalling
have been identified. These include: (1) protein tyrosine phosphatases (PTPs), (2) suppressors of
cytokine signalling (SOCS), and (3) protein inhibitors of activated STATs (PIAS) (reviewed in:
Chen et al., 2004).
Classical PTPs consist of the transmembrane receptor-like PTPs, and the intracellular
PTPs, which include SH2 domain-containing PTP1b, and T-cell (TC)-PTP (Chen et al., 2004).
The defining feature of all PTPs is the presence of a signature motif within the active site of their
catalytic domain which contains a preserved cysteine (Burke and Zhang, 1998). This critical
cysteine can mediate nucleophilic attack on substrate phosphotyrosine residues to promote their
hydrolysis (Burke and Zhang, 1998). Studies have identified the tandem ptyr motif of Jak2
13
(pYpY1007/1008) as a substrate of PTP1b (Myers et al., 2001). Further studies have shown that
PTP1b activation in glioma cells can decrease pro-survival Stat3 signalling by dephosphorylating
Jak2, its upstream activator, thereby priming these cells for apoptosis (Gu et al., 2003; Akasaki et
al., 2006). Mouse embryonic fibroblasts isolated from TC-PTP knockout mice also show
increased Jak2 and Stat3 activity, however in this case Stat3 may be a direct target of TC-PTP
(Yamamoto et al., 2002).
SOCS are a family of eight proteins (SOCSs 1 through 7 and cytokine inducible SH2containing protein (CIS)) that can be stimulated by the very cytokines that enhance STAT
activation, to act as inhibitors of JAKs (reviewed in: Chen et al., 2000). Hence SOCS proteins
can function in a classical negative feedback loop. Structurally, SOCS proteins consist of an Nterminal domain of variable length followed by a central SH2 domain and a C-terminal SOCS
box (Chen et al., 2000). Both SOCS-1 and SOCS-3 have shown to concurrently inhibit Stat3
activation by binding to upstream JAK kinases through their SH2 domain (Chen et al., 2000). A
kinase inhibitory region (KIR) at the N-terminus of these SOCS may act to suppress the
enzymatic activity of JAKs by competing with the access of substrates and/or ATP into their
catalytic pocket (Yasukawa et al., 1999; Chen et al., 2000). More recent studies have revealed
that the SOCS box may function in ubiquitin-mediated proteasomal degradation by associating
with factors to form an E3 ubiquitin ligase complex (reviewed in: Chen et al., 2000). Consistent
with this notion, SOCS-1 was shown to target Jak2 for proteasomal degradation (Ungureanu et
al., 2002). Interestingly, SOCS-1 and SOCS-3 are found to be methylated and silenced in certain
types of cancer (Baus and Pfitzner, 2006; Tischoff et al., 2007; Fernández-Mercado et al., 2008).
PIAS family members, which are encoded by four genes in humans, PIAS-1, PIAS-2
(PIASx), PIAS-3 and PIAS-4 (PIASy), function to regulate the activity of several transcription
factors, notably STATs (reviewed in: Rytinki et al., 2009). PIAS-3 can bind directly to
phosphorylated Stat3 homodimers or Stat1:Stat3 heterodimers to prevent DNA recognition,
possibly by recruiting co-repressor proteins to the complex (Chung et al., 1997). PIAS proteins
14
are also known to exhibit SUMO (small ubiquitin-related modifier)-ligase activity, similar that of
an E3 ubiquitin ligase, however this may not play a role in the repression of Stat3 by PIAS-3
(Chung et al., 1997; Rytinki et al., 2009).
More recent evidence has identified Gene Associated with Retinoid IFN-Induced
Immortality (GRIM)-19, a novel factor involved in Retinoid acid (RA)/IFN-β triggered apoptosis,
as an inhibitor of Stat3 nuclear translocation and transcriptional activity (Zhang et al., 2003).
Overexpression of GRIM-19 inhibits Stat3 through a mechanism involving direct binding.
Interestingly, GRIM-19 expression is found to be lost in a number of different cancers and its reexpression in a cervical cancer line was shown to restore growth suppression in vivo through
downregulation of Stat3 and its targets (Zhou et al., 2009).
Stat3 and cancer
Given that Stat3 has the intrinsic ability to promote cell proliferation and survival, it is
not surprising that overexpression of Stat3 has been demonstrated in a large number of solid and
blood tumors (reviewed in: Bromberg et al., 1999; Bowman et al., 2001; Yu and Jove, 2004;
Germain and Frank, 2007). Indeed, immunohistochemical analysis has revealed the presence of
nuclear, tyrosine phosphorylated Stat3 in 75% of primary breast and prostate cancers (Alvarez et
al., 2005). While oncogenic mutations within the Stat3 gene itself have not been reported,
hyperactive Stat3 can result from persistant signalling by upstream tyrosine kinases, which
themselves have become overactive due to genetic or epigenetic alterations or through silencing
or deregulation of its repressors (Yu and Jove, 2004).
While Stat3 tyr-705 phosphorylation is transient and tightly controlled in normal cells,
generally peaking within 30 minutes after stimulation by cytokines or growth factors and
declining in the following 30 minutes, Stat3 has been shown to be persistently phosphorylated in
cells transformed by certain oncoproteins such as v-Src (viral sarcoma; Turkson et al., 1998).
Data have shown that Stat3 is in fact required for full neoplastic transformation by v-Src, so that
15
dominant negative Stat3 can block v-Src-induced transformation by inducing cell cycle arrest and
apoptosis (Turkson et al., 1998). Furthermore, Stat3 activation by v-Src is not affected by
overexpression of SOCS-1 indicating that in transformed cells, hyperactive Stat3 may have
become desensitized to normal negative regulatory mechanisms (Iwamoto et al., 2000).
Stat3 is also known to be activated by growth factor receptors such as EGFR, the
dimerization of which may promote direct phosphorylation of Stat3 on tyr705 (Grandis et al.,
1998). Several studies have shown that constitutive activation of this receptor through
amplification, mutation or autocrine signalling, can lead to cellular transformation (reviewed in:
Quesnelle et al., 2007; Normanno et al., 2006). Stat3 has been shown to play a major role in
oncogenic signalling by the EGFR, by promoting the transcriptional activation of genes involved
in cell cycle progression, cell survival and metastasis (Grandis et al., 1998, Vigneron et al., 2008).
Very recent evidence has emerged indicating that Stat3 may also promote cellular
transformation by oncogenes which lack kinase activity. Gough et al. have shown that Stat3
could cooperate with activated (G12V) Ras to form colonies in soft agar and further demonstrated
a requirement for Stat3 by RasV12 to generate subcutaneous tumors in nude mice (2009). This
was demonstrated by the fact that RasV12-transformed, Stat3-deficient cells showed restricted
colony formation and tumor growth when injected into mice compared with RasV12 cells alone.
Interestingly, expression of Y705F Stat3 in shStat3 cells restored transformation by V12Ras and
tumor formation, indicating a lack of requirement for tyrosine phosphorylation. Expression of
various Stat3 deletion mutants revealed that pser-727 Stat3 plays a role in Ras-mediated
transformation. Furthermore, cell fractionation experiments demonstrated that Stat3 localizes to
the mitochondria in RasV12 cells. The data further indicate that mitochondrial Stat3 could support
transformation by Ras specifically, but not vSrc, through the promotion of glycolytic and
oxidative phosphorylation, exposing a novel, uncharacterized role of Stat3 (Gough et al., 2009).
The introduction of two cysteine residues in the SH2 domain of Stat3 has led to the
generation of a constitutively active form of Stat3 (Stat3C), which dimerizes spontaneously and
16
shows increased DNA binding affinity than its wild type counterpart (Bromberg et al., 1999; Li
and Shaw, 2006). Stat3C expression is sufficient to transform cells in culture and induce tumors
in nude mice (Bromberg et al., 1999). Stat3C has also enabled the identification of several key
Stat3-regulated genes. Targets of Stat3 include c-myc, cyclin D1/D2, and hepatocyte growth
factor (hgf), the expression of which can increase cell cycle progression and cell proliferation as
well as motility (Bowman et al., 2001; Hung and Elliott, 2001). Stat3 activation can also promote
cell survival through expression of members of the bcl-2 family of anti-apoptotic genes, such as
mcl-1 (Epling-Burnette et al., 2001) and bcl-xL (Bromberg et al., 1999), as well as survivin,
shown to block the mitochondrial-induced activation of caspases (Gritsko et al., 2006; Aoki et al.,
2003). Stat3 also plays a role in preventing programmed cell death by inhibiting p53 expression
through direct binding to its promoter (Niu et al., 2005). Furthermore, Stat3 expression can
promote angiogenesis by activating transcription of the vascular endothelial growth factor (vegf)
gene; VEGF is the most potent angiogenic signal and can induce the formation of new blood
vessels to facilitate the metastasis of cancer cells (Niu et al., 2002). Hyperactive Stat3 can
restrain anti-tumor immune responses by inhibiting the expression of key mediators of T-cell
mediated anti-tumor immunity and promoting the production of immunosuppressive factors such
as IL10 (Yu et al., 2009). In addition, Stat3 promotes tumor-associated inflammation and the
progression of inflammatory cancers, such as colorectal and gastric cancers, through the
production of pro-inflammatory cytokines like IL6 (Wang et al., 2004, Yu et al., 2009).
Inhibition of Stat3 using various methods [dominant-negative Stat3β (Catlett-Falcone et
al., 1999; Garcia et al., 2001), peptides (Turkson et al., 2001), antisense oligonucleotides (Niu et
al., 2002), RNA interference (Konnikova et al., 2003), DNA decoy (Gu et al., 2008; Leong et al.,
2003) or small molecule inhibitors (Turkson et al., 2004; Sun et al., 2005; Schust et al., 2006)]
could induce growth arrest or apoptosis of a number of tumor cell lines derived from multiple
myeloma, breast carcinoma, head and neck cancer, glioma, astrocytoma and hepatocellular
carcinoma, amongst others. Data from a number of labs have shown that inhibition of
17
constitutive Stat3 signalling using in vivo model systems, can induce apoptosis of tumor cells
with minimal effect on normal host tissues (reviewed in: Germain and Frank, 2007; Yu et al.,
2009). These findings suggest that cancer cells depend on higher Stat3 levels for growth and
survival. The effectiveness of Stat3 inhibitors in animal models has offered precedence for their
further validation in the clinic, and trials are currently underway. Blocking Stat3 can also
suppress angiogenesis and stimulate host immune responses. Hence targeting Stat3 in cancer can
have a double benefit, by preventing the growth of tumor cells and by enhancing adaptive and
innate immunity. While targeting a molecule which lacks enzymatic activity and does not bind
low molecular weight ligands, such as growth factors, may be unconventional, the current
problem facing anticancer therapy remains that oncogenic events in tumors can be highly diverse
and unpredictable. Stat3 lies at the convergence of a number of common deregulated pathways in
cancer and blocking this “central regulatory node” can overcome the upstream effects (Yu et al.,
2009). Collectively, these findings point towards Stat3 as a leading target in cancer therapy. It
follows that a comprehensive understanding of Stat3 signalling mechanisms is indispensable for
promoting its anti-cancer drug development.
2. The Cadherin Family of Cell Adhesion Receptors
Multicellular organisms rely on the complex recognition properties of adhesion
molecules for the selective aggregation of cells into fully functional, distinct anatomical
structures. In addition to providing an essential role in the genesis and architecture of tissues, cell
adhesion molecules are also dynamic units that can capture and integrate signals from the
extracellular environment thereby affecting key processes such as cytoskeletal organization and
cell proliferation (Jeanes et al., 2008; Takeichi 1990). Loss of adhesive properties has been
shown to be ssociated with the aberrant morphogenetic effects of cancer such as the metastasis of
tumour cells (Lodish et al., 2000).
18
Structure of classical cadherins and mechanism of adherens junction formation
The formation of cell-cell adhesion junctions is primarily modulated by cadherins, a
calcium-dependent family of membrane-spanning glycoproteins. Over 100 different cadherins
have been identified and can be subdivided into five groups: the classical cadherins, the
desmosomal cadherins, the seven-pass transmembrane cadherins, the larger cadherins of the fat
and daschous family and the protocadherins (reviewed in: Stemmler, 2008). The classical
cadherins are the best characterized and are subdivided into type I and type II members. The term
“classical” refers to the ability of the cytoplasmic domain of these cadherins to interact with βcatenin (Stemmler, 2008). Type I cadherins, which include epithelial (E)-cadherin, placental (P)cadherin, neural (N)-cadherin, as well as the retinal (R)-cadherin and myotubule (M)-cadherin,
are found in most tissues and are crucial to the developmental process (Yap et al., 1997;
Stemmler, 2008; Kaufmann et al., 1999). E-cadherin is the major receptor involved in
establishing and maintaining cell-cell or adherens junctions (AJ). Knockouts of the gene for Ecadherin die in utero due to defective formation of the first epithelial layer, the trophectoderm
(Larue et al., 1994). N-cadherin is expressed in the early neuroepithelium and murine knockouts
of this gene are viable only until day 10 of gestation, corresponding to the early stages of neural
tube formation (Radice et al., 1997; Takechi 1990). P-cadherin was initially found to be
expressed in mouse placenta during early pregnancy and also plays a crucial role in the terminal
differentiation of the epidermis in mouse and human tissues (Takechi 1990). P-cadherin
knockout mice are indeed viable and fertile, suggesting the possibility that other cadherins may
substitute for its function. Together with N-cadherin, M- and R-cadherins are known to be
involved in skeletal muscle differentiation (Kauffman et al., 1999).
Type I cadherins share low amino acid homology with type II cadherins, which include
human cadherins-5 (also known as vascular endothelial (VE)-cadherin) to -12 (Yap et al., 1997).
The functions of type II cadherins are often overlapping and mostly associated with the
developing nervous system (Patel et al., 2006). Aside from neural development, cadherin-11
19
plays a role in the formation of bone and the synovial lining of joints, and is also known as
osteoblast (OB)-cadherin (Okazaki et al., 1994). Cadherin-11 knockout mice display defects in
bone growth and are also resistant to inflammatory arthritis (Okazaki et al., 1994).
Classical cadherins contain an extracellular region, a single pass membrane spanning
region and a highly conserved cytoplasmic domain (Leckband and Prakasam, 2006; Figure 2.3).
The extracellular region of classical cadherins is divided into five distinct domains (EC1-EC5)
that each exhibit characteristic β-strand architecture (Boggon et al., 2002; Leckband and
Prakasam, 2006). Aggregation assays have confirmed that calcium is indeed required at
millimolar concentrations for AJ formation, which can be reversibly inactivated by EDTA
(Leckband and Prakasam, 2006; Courjean et al., 2008). The binding of a total of twelve calcium
ions to the interface between successive EC domains contributes to a more rigidified cadherin
structure that is resistant to proteolytic degradation (Parsini et al., 2007; Boggon et al., 2002;
Leckband and Prakasam, 2006). Mutations in the calcium binding sites of E-cadherin, such as the
D103A mutation in the outermost pocket, have shown to decrease homophilic adhesion
(Handschuh et al., 2001; Courjean et al., 2008).
Two models of adherens junction assembly in classical cadherins exist, the first involves
an initial cis interaction of receptors, followed by a trans binding between cadherin dimers. The
current model suggests a zippering between trans homodimers on adjacent cells (Parsini et al.,
2007; Perret et al., 2004; Leckband and Prakasam, 2006; Figure 2.4). Both models indicate that
the EC1 domain of classical cadherins plays a critical role in homotypic recognition of their
binding partner (Stemmler, 2008). In fact, adhesion is mediated by an exchange of the outermost
20
Figure 2.3: General structure of classical cadherins.
Left: Classical cadherins consist of an extracellular domain, a single-pass transmembrane region
and a cytoplasmic domain, which interacts with catenins and the actin cytoskeleton. Right: The
ectodomain of classical cadherins (represented here by C-cadherin from Xenopus) includes five
tandem extracellular repeats (EC1-5). Three calcium ions (green) bind at each of the four EC
domain junctions imposing rigidity to the structure. Figure adapted from Leckband and
Prakasam, 2006.
21
β-strands in the EC1 domain, and the docking of a conserved tryptophan (W) side chain of one
molecule into the hydrophobic pocket of its partner (Boggon et al., 2002; Chen et al., 2005;
Figure 2.5). Type I cadherins mediate strong cell-cell adhesion through the binding of W2
within a site surrounding a His-Ala-Val (HAV) triple amino acid sequence in EC1 (Stemmler,
2008). Type II cadherins, on the other hand, have two W residues (W2 and W6) but lack the
HAV motif within their much larger acceptor pocket, and the adhesive force is seven fold less
with type II cadherins compared to type I (Patel et al., 2006; Chu et al., 2006). Mutation of W2 to
Alanine (W2A) abolishes the adhesive capacity of several cadherins (Perret et al., 2002).
It has been suggested that both the EC1 and EC2 domains are sufficient for classical
cadherin engagement. Indeed, results have shown that a fragment consisting of the two outermost
domains of E-cadherin (E/EC1/2) can retain biological activity and be selectively recognized by
E-cadherin expressing cells (Perret et al., 2002). Later studies however showed that at least EC13 are involved in maintaining strong adhesive bonds and constructs containing additional EC
domains may be necessary to restore full activity (Leckband and Prakasam, 2006; Boggon, 2002).
Accordingly, it is believed that the further interdigitation of cadherins may result in more tightly
bound states (see Figure 2.4). Analysis of the binding properties of type I cadherins at the single
molecule level has revealed two attachment subpopulations, one population maintaining
attachments from 0.1 to 1 second and another from 102 to 105 seconds suggesting a role for
cadherins in transient recognition as well as stable tissue formation (Perret et al., 2002; Perret et
al., 2004; Pierres et al., 2007).
22
Figure 2.4: Proposed mechanisms for cadherin adhesion.
Lateral (cis) dimers may either (1) dissociate to form adhesive trans dimers or (2) remain
dimerized, while binding to opposing cadherins. The trans adhesive complexes may further
interdigitate to form additional bound states (right). Adapted from Leckband and Prakasam,
2006.
23
Figure 2.5: Ribbon drawing of the trans dimer structure of human E-cadherin EC1EC2 (Parsini et al., 2007).
Adhesion between cadherins is mediated by the docking of the W2 side chain of one
molecule into the hydrophobic pocket of the other molecule. This is the final step for the
correct folding of EC1 allowing trans-interactions. Mutation of Trp to Ala (W2A)
abolishes the binding capacity of several cadherins (Perret et al., 2002).
24
Cytoplasmic interactions of cadherins
Intercellular adhesion is strengthened by the clustering of cadherin receptors at adherens
junctions and the association between cadherin complexes and the actin cytoskeleton (Yap et al.,
1997). The cytoplasmic domain of cadherins is linked to underlying actin filaments through a
group of proteins termed catenins. Classical cadherins encompass a carboxy-terminal,
intracellular region for β- or γ-catenin (plakoglobin)-binding and a juxtamembrane region for
p120 catenin (p120ctn) binding (Yap et al., 1997). β-catenin and γ-catenin bind to α-catenin,
which is able to interact with actin binding proteins such as α-actinin, vinculin or ZO-1, as well
as actin itself (Stemmler, 2008). Cadherin, β-catenin and α-catenin form the core cadherincatenin complex and mutations affecting the expression or association of any of these proteins
can lead to reduced cell-cell adhesion (Oyama et al., 1994; Stemmler, 2008). p120 catenin is a
substrate of the Src kinase and may be recruited to the core cadherin complex to generally
strengthen adhesions by clustering cadherins to specific sites on the cell surface (Yap et al., 1998;
Reynolds et al., 1994). The small GTPases Rac, Rho, Cdc42, as well as Rap1, can also associate
with cadherin adhesion complexes to regulate their assembly; the role of Rho family GTPases in
AJ formation is discussed in section 3 below.
The adhesive interactions between cells are dynamic, so that normal morphological
remodeling events linked to embryogenesis, motility and wound healing can involve controlled
changes in cadherin function. In cultured cells, the depletion of extracellular calcium or sparse
growth conditions can lead to the internalization of cadherin-catenin complexes through
endocytic trafficking (Le et al., 1999). In response to extracellular factors, tyrosine
phosphorylation of cadherins or catenins by receptor or non-receptor kinases also correlates with
their disassembly at the membrane leading to decreased adhesion (Stemmler, 2008). Dissociation
of β-catenin from cadherins can lead to its accumulation in the cytoplasm, where its levels are
kept low through targeted destruction by the proteasome (Cavallaro and Christofori, 2004; Yap
25
and Kovacs, 2003). More specifically, cytoplasmic β-catenin associated with adenomatous
polyposis coli (APC) can be sequentially phosphorylated by casein kinase (CK)II and glycogen
synthase-3β (GSK-3β) leading to its ubiquitination and subsequent degradation (Gao et al., 2002;
Cavallaro and Christofori, 2004). In tumor cells, the normal regulatory processes which control
AJ assembly or disassembly can be perturbed.
Regulation and role of E-cadherin in cancer
Germ line mutations in the E-cadherin gene (CDH1) can lead to a predisposition to
gastric, lobular breast and colorectal cancers (Conacci-Sorrell et al., 2002). Furthermore, certain
tumors or tumor cell lines which harbor mutations in one CDH1 allele often acquire deletions in
the other one thereby characterizing E-cadherin as a classic tumor suppressor gene (reviewed in:
Conacci-Sorrell et al., 2002; Hirohashi, 1998).
Adherens junctions are essential to the structural integrity of a tissue and can thereby
restrict cell proliferation and motility (Conacci-Sorrell, et al., 2002; Oka et al., 1993). E-cadherin
is expressed in normal epithelial tissues and loss of E-cadherin expression has been observed in a
variety of epithelium-derived cancer cells and tumors in situ (reviewed in: Wheelock et al., 2001;
Wheelock et al., 2008). This may occur through: (1) downregulation of E-cadherin expression
following hypermethylation of its promoter or the binding of transcriptional repressors, (2)
mutations in the coding region, as well as (3) the activation of pathways that prevent AJ assembly
(Conacci-Sorrell et al., 2002). E-cadherin expression is negatively regulated by members of the
zinc-finger-family of transcription factors, including Snail, Slug, Twist, SIP1 and EF1, which are
able to bind to and repress transcription from the CDH1 promoter (Batlle et al., 2000; Wheelock
et al., 2007). Indeed, examination of the levels of these transcription factors in various cancer cell
lines reveals an upregulation correlating with E-cadherin repression (Batlle et al., 2000; Bolos et
al., 2003). The EGF receptor, as well as Src, Fer and Abl non-receptor tyrosine kinases, which
are often activated in cancer, can phosphorylate β-catenin, thereby disrupting its binding to
26
cadherin, or, in the case of Fer, α-catenin, leading to AJ disassembly (reviewed in: Roura et al.,
1999; Nelson, 2008). EGFR, HGFR/Met and Src can also tyrosine-phosphorylate E-cadherin on
its cytoplasmic tail (reviewed in Pece and Gutkind, 2002). More recent studies have identified
Hakai as an E3 ubiquitin ligase able to interact with tyrosine-phosphorylated E-cadherin to
promote its endocytosis and subsequent degradation in lysosomes (Fujita et al., 2002; Palacios et
al., 2005; Shen et al., 2007). While the region in E-cadherin known to bind Hakai is conserved
amongst all classical cadherins, the tyrosine residues which permit Hakai binding are only found
in E-cadherin (Pece and Gutkind, 2002). Interestingly, expression of Hakai was recently shown
to be upregulated in colon and gastric cancer (Figueroa et al., 2009).
The tyrosine phosphatase PTP1b, on the other hand, was shown to positively regulate Ncadherin-based adhesion by dephosphorylating β-catenin (Balsamo et al., 1996). Phosphorylation
of serine residues in the cytoplasmic tail of cadherins by GSK-3β or CKII also results in
increased adhesion by enhancing the binding of cadherins to β-catenin (Lickert et al., 2000).
However, as discussed above, in the absence of Wnt signalling, serine phosphorylation of
cytoplasmic β-catenin by GSK-3β or CKII kinases can lead to its ubiquitin-mediated degradation.
Activation of the Wnt signalling pathway can repress GSK-3β activity, leading to the nuclear
translocation of β-catenin, which may bind to Lymphoid Enhancer Factor/T-cell Factor
(LEF/TCF) transcription factors leading to the expression of genes which play a role in tumor
progression, such as c-myc and cyclin D1 (reviewed in: Cavallaro and Christofori, 2004; Yap and
Kovacs, 2003). Similarly, dissociation of p120ctn can lead to its accumulation in the nucleus
resulting in the activation of Kaiso, a recently characterized transcriptional repressor involved in
cancer development (Daniel and Reynolds, 1999; Kim et al., 2004).
Certain receptor tyrosine kinases were shown to coimmunoprecipitate with cadherins,
and cadherins may also in turn influence tyrosine kinase signalling. Qian et al. have reported that
the growth inhibition observed in confluent cell cultures may occur as a result of the E-cadherin-
27
mediated suppression of RTKs, such as EGF or IGFR, or the activation of PTPs (Qian et al.,
2004). They claim that the extracellular domain of E-cadherin may associate with RTKs to
inhibit their ligand-mediated activation, more specifically by causing a decrease in their affinity
for the ligand and reduced receptor mobility (Qian et al., 2004). This may explain why
overexpression of E-cadherin in carcinoma cells can lead to a reversion to a normal epithelial
phenotype (Vleminckx et al., 1991; Conacci-Sorrell et al., 2002). Paradoxically however, in
other cell systems, interactions between EGFR or HGFR/Met and E-cadherin or FGFR and Ncadherin have shown to sustain receptor signalling through the ERK, MAPK pathway (Li et al.,
2001; Pece and Gutkind, 2000; Hoschuetzky et al., 1994; Stemmler, 2008). Interestingly, FGFR
also contains a HAV domain which mediates binding to N-cadherin, somewhat analogous to
homophilic cadherin engagement (Williams et al., 2001).
Cadherin switching in cancer
Loss of E-cadherin is often concomitantly associated with a de novo gain of expression in
the mesenchymal cadherins, N-cadherin and cadherin-11 (Stemmler, 2008; Conacci-Sorrell, et al.,
2002; Oka et al., 1993; Hazan et al., 1997; Feltes et al., 2002 Pishivaian et al., 1999). While
cadherin switching is common during normal embryonic development, it is the rate limiting step
in the malignant progression of cancer cells, also known as the epithelial to mesenchymal
transition, or EMT (Conacci-Sorrell, et al., 2002; Oka et al., 1993; Table 1.1). N-cadherin
activation in cancer cell lines is associated with a highly invasive and motile phenotype (Hazan et
al., 1997). Cadherin switching from E- to cadherin-11 has been observed in breast and prostate
cancers and is linked with bone metastasis (Feltes et al., 2002 Pishivaian et al., 1999; Tomita et
al., 2000; Chu et al., 2008). In sharp contrast, E-cadherin was also shown to promote
tumorigenesis in tissues which do not normally express this adhesion molecule. Ovarian surface
epithelial cells commonly express N-cadherin, and a switch from N- to E-cadherin has been
reported to promote aberrant differentiation, characteristic of ovarian carcinogenesis (Wheelock
28
et al., 2008; Patel et al., 2003; Table 1.1). There is also strong evidence that once tumor cells
reach distant metastatic sites, they may re-express E-cadherin to levels equal to or higher than that
of the primary tumor (Kowalski et al., 2003).
Table 1.1
Cadherin switching during tumorigenesis
(Wheelock et al., 2008)
Switch
Example
From E-cadherin to N-cadherin
Melanoma
TGFβ-induced EMT in mammary epithelial cells
Prostate cancer
Breast cancer
Pancreatic cancer
From E-cadherin to T-cadherin
Hepatocellular carcinoma
From E-cadherin to P-cadherin
Pancreatic cancer
Gastric cancer
From E-cadherin to cadherin-11
Prostate cancer
Breast cancer
From E- and P- to N-cadherin
Oral squamous cell carcinoma
From N-cadherin to E-cadherin
Ovarian cancer
3. Rho Family GTPases
Rho family GTPases are members of the RAS superfamily of small GTPases which cycle
between an active GTP (guanosine 5’-triphosphate)-bound state and an inactive GDP (guanosine
5’-diphosphate)-bound form (reviewed in: Ellenbroek and Collard, 2007; Vega and Ridley,
2008). The first small GTPase to be discovered was Ras. While the ras gene is found to be the
most frequently mutated in cancer, oncogenic transformation by aberrantly activated Ras has
29
been shown to depend on Rho family members (Yamamoto et al., 1999; Sahai et al., 2001; BarSagi and Hall, 2000).
Regulation of Rho GTPases (Figure 2.6)
Ras and Rho GTPases have very low intrinsic GTP hydrolytic activity hence the cellular
control of GTP/GDP cycling must be modulated by regulatory proteins (Ellenbroek and Collard,
2007). Rho GTPases remain inactive in the cytosol. The association of these GTPases to the
cytoplasmic membrane following post-translational modifications by isoprenoid lipids
(prenylation) at the C-terminus is a process critical for their activation (Vega and Ridley, 2008).
Following membrane targeting, guanine nucleotide exchange factors (GEFs) can support the
dissociation of GDP from Rho GTPases thereby favoring the uptake of GTP from the cytosol
(Vega and Ridley, 2008). The Rho family of GEFs is very large compared to the Ras family of
GEFs and includes at least 60 members, which may be either specific or non-specific towards the
GTPase they act upon (Mackay & Hall, 1998). Rho GEFs are also referred to as Dbl family
proteins, named after their founding member identified as a transforming protein from a human
diffuse B-cell lymphoma (Dbl) cell line (Schmidt and Hall, 2002). All Rho GEFs contain a Dblhomology (DH) domain, which is required for their catalytic activity, and a Pleckstrin-homology
(PH) domain, which mediates membrane association but has also shown to modulate DH domain
activation (Schmidt and Hall, 2002; Karnoub et al., 2004). Inactivation of small GTPases is
promoted by GTPase activating proteins (GAPs), which increase the rate of bound GTP
hydrolysis (Vega and Ridley, 2008). More than 80 known Rho GAPs exist, and amongst these is
BCR, first identified as the translocation partner of the Abl tyrosine kinase in the Philadelphia
chromosome, found in the majority of leukemias (Heisterkamp et al., 1993). Rho family
GTPases are further negatively regulated by the effects of Rho guanine nucleotide dissociation
inhibitors (Rho-GDIs; Mackay and Hall, 1998). Rho-GDIs may inhibit Rho GTPase signalling
by: 1) preventing nucleotide dissociation and GDP/GTP exchange to antagonize GEF activity,
30
Figure 2.6: Regulation of Rho GTPases (Fukata and Kaibuchi, 2001).
Rho GTPases are inactive when complexed to GDP and one of three known RhoGDIs. Upon stimulation by extracellular factors, Rho is released from Rho-GDI and
associates with the membrane through its C-terminal prenyl group. Rho-GEFs promote
Rho-GTP exchange leading to the activation of various effector proteins. A Rho-GAP
will then catalyze GTP hydrolysis and Rho-GDI will extract the GTPase from the
membrane locking it once again in an inactive state.
31
2) by interfering with GTP hydrolysis and restricting GAP activity and 3) stimulating the release
of Rho GTPases from the membrane, where they are active, to the cytosol (Karnoub et al., 2004).
Hence Rho GTPases exist in a stable, cytosolic, inactive, GDP-bound state while complexed with
Rho-GDIs. Cytokine and growth factor receptors, cell-cell (cadherin) and cell-matrix (integrin)
adhesion receptors as well as G-protein coupled receptors may all activate Rho GTPases mainly
via stimulation of Rho GEFs, and most GEFs contain additional domains such as SH2, SH3,
serine/threonine or tyrosine kinase, Ras-GEF, Rho-GAP, PDZ or extra PH domains which may be
involved in the interaction of GEFs with upstream modulators (Schmidt and Hall, 2002). By
assisting in the membrane translocation of Rho GTPases and specifically promoting guanine
nucleotide exchange and activation, Rho GEFs may disrupt the cytosolic complex of Rho-GDI
(Bokoch et al., 1994; Schmidt and Hall., 2002). In their GTP-bound form, Rho GTPases are able
to bind to various effectors to regulate a number of cellular responses in the cytoplasm and the
nucleus including actin cytoskeletal rearrangement, cell cycle progression, proliferation and
apoptosis (Ellenbroek and Collard, 2007; Karnoub et al., 2004). RhoA, Rac1 and Cdc42 are the
prototype Rho GTPase members, most known for their ability to regulate the assembly of focal
complexes associated with actin stress fibers, lamellipodia and filopodia, respectively (Ellenbroek
and Collard, 2007).
Structural features of Rho GTPases (Figure 2.7)
For Rho GTPases, as for Ras and other small GTPases, the exchange of GDP for GTP is
accompanied by conformational changes in two N-terminal regions termed switch regions I and II
(Ihara et al., 1998; Bishop and Hall, 2000). In Rac1, switch I and switch II span amino acids 2549 and 59-76, respectively (Karnoub et al., 2004). It is interesting to note that these two regions
are essentially identical in RhoA, Rac1 and Cdc42 with the exception of position 38 in switch I
which is aspartic acid (D) in Rac1/Cdc42 and glutamic acid (E) in RhoA (Bishop and Hall, 2000).
The Cdc42/Rac-interactive binding (CRIB) motif, present in many Rac and Cdc42 effectors, may
32
preferentially recognize D38, preventing its association with Rho (Bishop and Hall, 2000).
Indeed, while the switch I region is important for the interaction of Rho GTPases with effector
molecules, and has been termed the effector domain, point mutations to this site inhibit the
binding of some but not all downstream targets (Karnoub et al., 2004). For example, while
binding of the Rac/Cdc42 effector p21-activated kinase (PAK) is prevented by a tyrosine to
cysteine substitution at residue 40, the same mutation does not affect the association of Rac with
mixed lineage kinase (Mlk)-2 and -3 (Bishop and Hall, 2000). Mutagenesis and structural studies
have also identified a role for the switch I region in proper GTPase function, indicating that it
constitutes a docking site for GEFs and GDIs (Karnoub et al., 2004). However, while GDIs may
contact the C-terminal end of switch I, they primarily associate with the switch II region
(Hakoshima et al., 2003). In contrast to the Ras family, Rho GTPases possess a 13 amino acid
insert region close to the C-terminus (Karnoub et al., 2004). This insert region has also been
implicated in some but not all effector interactions. For instance, the insert region of Rac1 and
Cdc42 is necessary for IQGAP binding, however it is not required for their association with any
of the known CRIB-containing effectors (Bishop and Hall, 2000).
As mentioned above, Rho proteins, like Ras GTPases, undergo posttranslational
modification at the C-terminus which contains a CAAX tetrapeptide motif (cysteine (C) followed
by two aliphatic amino acids (AA) and a terminal amino acid (X)) (Roberts et al., 2008). For
RhoA, Rac1 and Cdc42, the first step involves covalent addition of isoprenoid moieties to the
cysteine residue by geranylgeranyltransferase I (Roberts et al., 2008). Next, the AAX residues
are removed by proteolysis and a methyl group is added to the prenylated cysteine by a
carboxymethyltransferase (Roberts et al., 2008).
Characterization of the biological roles of Rho GTPases has been facilitated by the
generation of activated and dominant-negative mutants, as well as bacterial toxins. Amino acid
substitutions of either valine for glycine at codons 12 in Rac1/Cdc42, or 14 in RhoA, as well as
leucine for glutamine at codons 61 in Rac1/Cdc42, or 63 in RhoA, can prevent intrinsic and GAP-
33
Figure 2.7: Structure of Rac1/Cdc42.
Rho GTPases share significant sequence identity with Ras proteins. The switch I and switch II
domains adopt a conformational change upon binding to GTP and are involved in the association
of these GTPases with various effectors. In contrast to Ras, Rho GTPases contain a 13 amino
acid insert domain which may also regulate effector binding. The CAAX motif at the C-terminal
end, along with another upstream membrane targeting motif, enables membrane association of
these GTPases, a process critical for their activation. Constitutively activating (G12V, Q61L) and
dominant negative (T17N) mutations have been identified in Rho GTPases that have enabled the
characterization of their role in various biological processes. Figure adapted from Karnoub et al.,
2004.
34
induced GTP hydrolysis resulting in GTP-bound proteins which may be activated independently
of ligand binding (Bishop and Hall, 2002). On the other hand, substitution of asparagine for
threonine at codons 17 in Rac1/Cdc42, or 19 in RhoA, can allow these mutants to bind GEFs with
higher affinity than their wild type counterparts (Feig, 1999). Hence, these mutants display a
dominant negative phenotype by competing with the endogenous GTPases for GEF binding,
thereby blocking downstream signalling from effectors (Feig, 1999; Figure 2.7).
Certain bacterial toxins are also able to exert negative effects on Rho GTPases. The C3
transferase from Clostridium botulinum can inactivate RhoA, and other subtypes RhoB and
RhoC, through ADP-ribosylation at asparagine 41 (Wilde and Aktories, 2001). It has been
suggested that this modification may impair the ability of Rho proteins to interact with GEFs and
may also sequester it in the cytoplasm by preventing its release from GDIs (Wilde and Aktories,
2001). On the other hand, the toxin B from Clostridium difficile is known to inhibit the activity
of most, if not all, Rho GTPase members (Just et al., 1994; Bishop and Hall, 2000). In this case
toxin B may modify Rho GTPases by glucosylation at threonine 37, thereby inhibiting effector
binding and also efficient cycling between the cytosol and membrane (Genth et al., 1999).
Rho GTPases and cell-cell adhesion
Rho GTPases are key regulators of cell morphology and motility through their intimate
association with the actin cytoskeleton. Incidentally, Rac1 and Cdc42 proteins can co-localize
with E-cadherin at cell-cell AJs and are believed to contribute to the stabilization of cadherin
receptors (reviewed in: Fukata and Kaibuchi, 2001; Nelson, 2008). Calcium chelation, which
leads to the rapid disassembly of E-cadherin complexes, has also shown to promote the
translocation of these GTPases to the cytosol in Madin-Darby Canine Kidney (MDCK) cells
(Kovacs et al., 2002a). Further data has revealed an increase in active Rac-GTP levels in
epithelial cells plated on surfaces coated with an E-cadherin fragment indicating that Rac
activation is a specific consequence of cadherin ligation (Kovacs et al., 2002a). Indeed, a number
35
of labs have provided evidence of a positive feedback loop enabling cadherin assembly to
increase the activity of Rac1 and Cdc42, which may in turn stabilize adherens junctions
(reviewed in: Fukata and Kaibuchi, 2001; Braga et al., 1999).
Cadherins may communicate with Rho GTPases through phosphatidyl-inositol-3 (PI3)
kinase signalling pathways (Sander et al., 1998). Cell to cell contact can bring about the
association of PI3-kinase with E-cadherin containing complexes, possibly through catenins which
may serve as docking proteins, and its subsequent activation (Pece et al., 1999; Rivard, 2009;
Larue and Bellacosa, 2005). This may lead to the recruitment of GEFs such as Tiam1, which is
Rac-specific, to the membrane, through binding of its PH domain to the 3-phosphorylated
phosphoinositides (PIP3) products of PI3-kinase (Sander et al., 1998; Rivard, 2009; Figure 2.8).
While inhibition of PI3-kinase prevented activation of Rac1 by E-cadherin, it did not abolish it
completely indicating that other mechanisms may exist for E-cadherin to activate Rac1 and other
GTPases (Kovacs et al., 2002a). Apart from E-cadherin, other cadherin family members may
also activate Rho GTPases, as Tiam1 has been implicated in the activation of Rac1 by the type II
cadherin, VE-cadherin, in endothelial cells (Lampugnani et al., 2002).
Rho GTPases may also function in stabilizing cadherin junctions. In fact, Tiam1 is also
known to inhibit cell scattering brought about by HGF in MDCK cells, likely by increasing Ecadherin based adhesion (Hordijk et al., 1997). In addition, Rac1 and Cdc42 have shown to
reinforce cadherin mediated cell-cell contacts by sequestering IQGAP, a downstream effector of
these GTPases, which is known to negatively regulate E-cadherin adhesion by binding and
displacing β- and α-catenin from the cell adhesion complex (Sander and Collard, 1999; Briggs
and Sacks, 2003; see Figure 2.9). Furthermore, cadherins may act to strengthen AJs by
expanding sites of newly formed cell-cell contacts (reviewed in Yap and Kovacs, 2003). Indeed,
activated Rac1 and Cdc42 may stimulate the catalytic activity of the Arp2/3 actin nucleator
complex, which may be recruited to the membrane upon cadherin ligation, thereby enabling
cadherins to mark sites for actin assembly (Kovacs et al., 2002b).
36
Figure 2.8: Activation of Rac1 by E-cadherin involves PI3-kinase
In the absence of cell-cell adhesion, Rac1-GDP exists in an inactive complex bound to GDI.
Rac1 has been shown localize to sites of nascent cell contact, however the specific manner in
which Rho GTPases can be targeted to the membrane is yet unknown. Calcium chelation or
treatment of epithelial cells with an E-cadherin blocking antibody leads to cytosolic Rac1 (and
Cdc42). E-cadherin engagement can result in the activation of PI3-kinase, which may lead to the
membrane translocation of the Rac1-GEF Tiam1, or other potential GEFs with PH domains that
recognize the PIP3 products of PI3-kinase. Inhibition of PI3-kinase activity can inhibit, at least in
part, the activation of Rac1.
(Yap and Kovacs, 2003; Braga, 2000; Fukata and Kaibuchi; 2001)
Figure adapted from Fukata and Kaibuchi, 2001.
37
Unlike Rac1 and Cdc42, RhoA is found to localize in the cytosol of confluent cells and
might instead influence adhesive activity through its effects on the actin cytoskeleton (Takaishi et
al., 1997). While some studies have demonstrated its activation upon N-cadherin engagement,
others have revealed RhoA-GTP levels to generally decrease following E-cadherin engagement,
and in confluent epithelial cells (Charrasse et al., 2002; Yap and Kovacs, 2003; Fukata and
Kaibuchi, 2001). In line with the latter findings, Noren et al. have identified a RhoA-specific
GAP, p190RhoGAP, which is activated upon cadherin engagement, through a mechanism
dependant on its tyrosine phosphorylation by Src kinases (2002). Incidentally, crosstalk between
Rac1 and RhoA GTPases has also been demonstrated and studies have revealed that Rac1, which
is generally activated at sites of cell-cell contact, may antagonize RhoA through activation of
Rho-GAPs, such as p190 (Bustos et al., 2008), or the inhibition of Rho-GEFs.
Rho GTPases and Stat3
By targeting a variety of different effectors, Rho GTPases have shown to play a role in
cell migration as well as cell cycle progression and proliferation (Benitah et al., 2004). The
overexpression of Rho proteins in human tumors suggests that Rho GTPases are involved in
carcinogenesis. High levels of Rac1, Cdc42 and RhoA have been found in breast and testicular
cancers amongst others (Ellenbroek and Collard, 2007). In constrast to Ras proteins, where direct
mutational activation is prelavent in a number of cancers, the evidence associating Rho GTPases
with cancer is more indirect and may occur through abberant signalling from upstream signalling
components such as receptor tyrosine kinases in which genetic or epigenetic alterations are more
common. Studies over the last decade have indicated that the role of Rho GTPases in tumor
initiation, tumor progression and metastasis may be linked to their ability to affect gene
transcription (Benitah et al., 2004).
38
Figure 2.9: Role of the Rac1 and Cdc42 effector, IQGAP, in E-cadherin-mediated cell-cell
adhesion.
When Rac1 and Cdc42 are inactive in the cytosol, IQGAP can negatively regulate cadherin
activity by binding to β-catenin, thereby displacing α-catenin from the complex. Active, GTPbound Rac1 or Cdc42 can sequester IQGAP, to allow the association of the E-cadherin complex
with α-catenin and the actin cytoskeleton, leading to strong cell-cell adhesion. E: E-cadherin; β:
β-catenin; α: α-catenin.
Figure adapted from Fukata and Kaibuchi, 2001.
39
RhoA, Rac1 and Cdc42 can modulate the expression of several transcription factors,
including nuclear Factor (NF)κB, elongation factor-2 (E2F), Serum Response Factor (SRF), as
well as Stat3 (Benitah et al., 2004). In addition to tyr705, Stat3 requires phosphorylation at
ser727 for maximal transcriptional activity (Wen et al., 1995). In vitro kinase assays using
purified full length Stat3 as the substrate showed that Rac1 mediates ser727 phosphorylation of
Stat3 in vSrc transformed fibroblasts through activation of the stress-activated serine/threonine
MAP kinases, p38 and JNK (Turkson et al., 1999). The Rac1/Cdc42 effector, p21-activated
kinase (PAK) shows high homology to an upstream regulator of the pheromone MAP kinase
pathway in yeast and therefore might link Rac1 and Cdc42 activation to JNK activity (Mackay &
Hall, 1998). Reports have claimed that although constitutively active PAK can stimulate p38 and
JNK activity, Rac1 mutants which do not bind PAK can still stimulate JNK activation suggesting
the possibility of other effectors (Aznar and Lacal, 2001). One such candidate is the Rac effector
POSH (plenty of SH3 domains), which has been shown to lead to JNK activation as well as
nuclear translocation of NFκB in Cos-1 cells (Aznar and Lacal, 2001; Tapon et al., 1998). POSH
was also revealed to have a scaffolding role by forming a multiprotein complex linking together
Rac1 and modulators of the JNK cascade, MLK3 and MEKK4 which may act as MAPK kinases
(Mackay & Hall, 1998; Aznar and Lacal, 2001; Tapon et al., 1998).
Independent studies have demonstrated that Rho GTPases could also effectively
phosphorylate Stat3 at tyr-705, leading to its activation (Debidda et al., 2005; Aznar and Lacal,
2001; Pelletier et al., 2003; Simon et al., 2000; Faruqi et al., 2001). Controversial studies by
Simon et al. have suggested that a direct association between Rac1 and Stat3 could lead to
tyrosine and serine phosphorylation of the latter (Simon et al., 2000). On the other hand, Faruqi
et al. have reported that Stat3 activation by Rho GTPases may occur indirectly through
transcriptional activation of IL6 and IL6 receptor genes by NFκB, thus forming an autocrine
activation loop (Faruqi et al., 2001). Once secreted, the IL6 cytokine is the most potent activator
of the Stat3 response. Rac1 and Cdc42-mediated tyrosine 705 phosphorylation of Stat3 was also
40
shown to occur via G-protein coupled receptor (GPCR)-mediated Jak2 signalling in two peaks of
activation, the second more sustained response requiring autocrine IL6 production (Pelletier et al.,
2003). A more recent paper has disproved any stable association between active RhoA, Rac1 or
Cdc42 and Stat3 and further demonstrates that Rho GTPases can also effectively induce Stat3
activation independently of IL6, suggesting the possibility of yet unappreciated mechanisms of
Stat3 activation by the Rho proteins (Debidda et al., 2005).
4. The gp130/JAK/Stat3 axis
Cytokines and the JAK-STAT pathway
The JAK-STAT pathway was first discovered through the study of cytokines, more
specifically interferon-induced signal transduction (Imada and Leonard, 2000; Schindler and
Plumlee, 2008). The large hematopoetin subfamily of cytokines which trigger signalling to
STATs now include IFN-α, -β and -γ; Interleukins 2-7, 10-13, 15, 19-24, 27 and 35; growth
hormone, erythropoetin, prolactin, thrombopoetin and other polypeptides (Darnell, Jr., 1997;
Schindler and Plumlee, 2008).
Like growth factors and hormones, cytokines are important mediators of intercellular
communication. Unlike hormones which are preformed and stored in glands, cytokines can be
rapidly synthesized and are secreted by cells following an immune stimulus, particularly during
infection or inflammation (Heinrich et al., 1998). They are pleiotropic molecules, affecting many
different target cells by binding to their matching cell surface receptor in an autocrine or
paracrine manner. Cytokines are also redundant in that similar responses can be achieved by
several different members of this large family. These molecules have been classified on the basis
of (1) their actions as pro- or anti-inflammatory cytokines, (2) according to the receptors used or
(3) their three-dimensional structure (Heinrich et al., 1998).
While lacking intrinsic enzymatic activity themselves, cytokine receptors are found to be
non-covalently associated with the JAK family of kinases (Jak1, Jak2 and Jak3 and tyrosine
41
kinase-2 (Tyk2); Imada and Leonard, 2000). Cytokine binding to its receptor can lead to the
activation of JAKs, which phosphorylate themselves and the receptor at specific tyrosines that
may be recognized by the SH2-domain of STATs or other SH2-containing adaptor proteins
(Imada and Leonard, 2000). Cytokine signalling can ultimately modulate gene transcription
through the activation of two major transduction pathways: (1) the JAK-STAT pathway and (2)
the MAP kinase cascade (Heinrich et al., 1998). Signalling through these pathways can result in
the production of more cytokines, their specific receptors or a suppression of their own effect
through feedback inhibition (Heinrich et al., 1998).
Janus kinases (JAKs)
The connection between JAKs and cytokine signalling was first observed following the
analysis of Tyk2 mutant cell lines, where it was shown that Tyk2 was required for interferon
signalling (Velazquez et al., 1992). Members of this family now include Jak1, Jak2, Jak3, as well
as Tyk2 (O’Shea et al., 2004). JAKs are fairly large kinases of about 120-140kDa, that associate
with a membrane-proximal, proline-rich domain of cytokine receptors (Kisseleva et al., 2002;
Imada and Leonard, 2000). Jak1, Jak2 and Tyk2 are expressed in many tissue types and bind
several cytokine receptors whereas Jak3 is found predominantly in myelocytic or lymphocytic
lineages and selectively binds the common gamma cytokine (γc)-receptor subunit which
associates with the interleukins 2, 4, 7, 9, 15 and 21 (Schindler and Darnell, Jr., 1995; O’Shea et
al., 2004). That is to say, many different cytokines can activate the same JAKs. JAKs contain
seven conserved regions, termed Janus Homology domains 1-7 (Imada and Leonard, 2000;
Ortmann et al., 2000; Figure 2.10). The presence of a pseudokinase domain in JH2 in addition to
a kinase domain in JH1 is what makes this family unique. In fact, the name “Janus” kinase refers
back to the Roman mythological god of gates and doorways and implies that JAKs are two-faced
with reference to these domains (Imada and Leonard, 2000). While the JH1 and JH2 regions are
structurally similar, in terms of their function the pseudokinase domain lacks catalytic activity,
42
although it appears to be involved in the regulation of the kinase domain (reviewed in: Imada and
Leonard, 2000; O’Shea et al., 2004). Other studies have suggested that the JH2 region serves as a
docking site for STATs (Fujitani et al., 1997). The amino terminal JAK region (JH3-JH7) is less
well conserved and contains a FERM (a band four-point-one, ezrin, radixin, moesin) region that
is crucial for receptor association, as well as an SH2-like domain in JH4 of unknown function
(Kisseleva et al., 2002; O’Shea et al., 2004; Schindler and Plumlee, 2008).
IL6-type cytokines and the gp130 receptor
Binding of cytokines of the IL6 family, consisting of IL6, IL11, IL27, Leukaemia
Inhibitory Factor (LIF), Oncostatin M (OSM), ciliary neurotropic factor (CNTF), cardiotropin-1
(CT-1) and cardiotropin-like cytokine (CLC), to the common gp130 receptor subunit can trigger
Stat3 phosphorylation via Jak1, Jak2 or Tyk2 (Heinrich et al., 2003).
Members of the IL6 family all share a common structure consisting of four long chain alpha-helix
bundles (reviewed in: Hirano et al., 2000; Heinrich et al., 2003). Aside from their role in
inflammation and the acute phase response to infection, IL6 cytokines are also involved in
haematopoesis, liver and neuronal regeneration, embryonal development and fertility (Hirano et
al., 2000). Aberrant signalling from these cytokines has been shown to contribute to the initiation
and progression of several diseases such as rheumatoid arthritis, inflammatory bowel disease,
osteoporosis, multiple sclerosis as well as a number of different cancers (Heinrich et al., 2003).
IL6 in particular is found to be overexpressed in many tumor derived cell lines, breast and lung
cancers as well as metastases (Scheller and Rose-John, 2006; Selander et al., 2004; Garcia et al.,
2001; Gao et al., 2007).
Receptors involved in the recognition of IL6 type cytokines include the non-signalling αreceptors (IL6R, IL11R and CTNF-R) and the signal transducing receptors (gp130, LIFR and
OSMR). IL6, IL11, and CTNF must first bind to their respective α-receptors, via their site I and
43
Figure 2.10: Domain structure of JAKs.
All JAKs contain seven JH domains (JH1-JH7). JH1 represents the catalytic domain and JH2 the
pseudo-kinase domain, which may serve to regulate kinase activity. JH4 includes an SH2-like
domain of unknown function. JH4-JH7 comprise a FERM domain which mediates binding to
cytokine receptors. JAKs autophosphorylate at multiple sites, including two key tyrosine residues
(pY) in the activation loop of the kinase domain.
Figure adapted from Schindler and Plumlee, 2008.
44
II regions, in order to recruit signalling receptor subunits (Wang et al., 2009; Heinrich et al.,
2003). The CTNFα receptor is unique in that it can recognize CTNF, CT-1 and CLC (Wang et
al., 2009; Figure 2.11).
All IL6 family cytokines possess a site III receptor binding site which recognizes gp130
(Wang et al., 2009). IL6 and IL11 can signal through gp130 receptor homodimers, while the rest
of the IL6 family cytokines can signal via gp130-LIF heterodimers (LIF, CT-1, CTNF, OSM and
CLC) or through gp130-OSMR heterodimers (OSM) (Heinrich et al., 2003; Wang et al., 2009).
Gp130 has also shown to engage the more recently identified IL-27 in conjunction with the T cell
cytokine signalling receptor (TCCR) (Pflanz et al., 2004). Interestingly, two viral homologs of
IL6 have been found in Human Herpes Virus-8 (HHV-8) and Rhesus macaque rhadinovirus
(Rm), which bind to gp130 alone (Wang et al., 2009; Figure 2.12). Alternatively spliced
isoforms of interleukin receptors have also been identified, soluble IL6 and IL11 receptors may
lack the cytoplasmic domains and function as antagonists by competing with native IL6 or IL11
but failing to transmit signals.
The extracellular region of gp130 consists of six domains (D1-D6) and is characterized
by an Immunoglobulin (Ig)-like region at the N-terminus (D1), a cytokine binding homology
region (CHR; D2 and D3) and three Fibronectin type III (FNIII) domains (D4-D6; Wang et al.,
2009; see Figure 2.12). Both the Ig and CHR sites are necessary for the binding and full
activation of gp130 (Wang et al., 2009). Binding of an IL6 family cytokine to its receptor results
in phosphorylation and activation of the gp130-associated JAKs (Jak1, Jak2 and Tyk2) (Hirano et
al., 2000). The JAKs then phosphorylate gp130 at specific tyrosines which act as binding sites
for the SH2 domains of (1) Stat3, (2) the protein tyrosine phosphatase SHP-2 and (3) SOCS-3
(Figure 2.13). Tyr759 of the gp130 receptor can recruit SHP-2, which itself can be
phosphorylated by JAKs (Jak1) to attract the SH2 domain of Grb2 resulting in activation of the
MAP kinase pathway, or the scaffolding protein Gab1 which can associate with p85, the
regulatory subunit of PI3-kinase (Hirano et al., 2000). Tyr759 is also known to bind SOCS-3, a
45
Figure 2.11: Receptor complexes of the IL6-type family of cytokines.
The seven members of the IL6-type family of cytokines are represented by dark grey circles.
Signalling receptors include gp130 (dark pink), OSMR and LIFR (light pink). The non-signalling
α-receptor subunits for IL6, IL11, CT-1 and CNTF are shown in light grey. Soluble isoforms of
these receptors exist which lack the cytoplasmic domain; these may compete for cytokine binding
in an inhibitory manner.
Figure adapted from Heinrich et al., 2003
46
Figure 2.12: Extracellular structure of the gp130 receptor, which binds IL6-type cytokines
(Wang et al., 2009).
The extracellular region of the gp130 receptor is characterized by an Ig-like region (IgD) at the
N-terminus (D1), a CHR region for cytokine binding (D2 and D3) and three FNIII repeats (D4D6). Both the IgD and CHR sites are necessary for the binding and full activation of gp130. IL6type cytokines with known three-dimensional structure are shown with cylinder representations
of the four helix bundles. Viral homologs of IL6 have been identified in Human Herpes Virus-8
(HHV8) and Rhesus macaque rhadinovirus (Rm) (Wang et al., 2009).
47
known inhibitor of Stat3 signalling. Mutation of the tyr759 binding site has been shown to result
in enhanced Stat3 signalling suggesting a negative regulatory role for SOCS-3 and SHP-2 in
gp130/JAK/Stat3 signalling (Hirano et al., 2000; Fairlie et al., 2003).
Stat3 can bind to any one of four YXXQ motifs in the carboxy-terminus of the gp130
receptor, these contain tyrosines 767, 814, 905 or 915. Subsequently, Stat3 phosphorylation at
tyr705 by JAKs results in its homodimerization or heterodimerization with Stat1 (Hirano et al.,
2000). IL6, which binds to gp130 homodimers can specifically trigger Stat3:Stat3 binding
(Darnell, Jr., 1997).
IL6 and cancer
The link between the IL6/JAK/Stat3 axis and survival signalling was first demonstrated
in 1996 through data indicating that Stat3 is needed for gp130-induced proliferation in a pro-B
cell line (Hirano et al., 2000). While gp130 can activate MAPK and PI3K pathways as well as
Stat3, it was revealed that tyr759 of this receptor was essential for cell cycle progression but not
for preventing apoptosis. In fact, the tyrosines involved in Stat3 activation were required for bcl2 induction and cell survival (Hirano et al., 2000).
Further studies have indicated that the specific gp130 ligands, IL6, OSM and LIF, can
play a role in the differentiation and growth inhibition of a variety of cultured cells (Selander et
al., 2004). However, IL6/gp130 signalling in particular can trigger constitutive activation of
Stat3 in a number of tumors and tumor-derived cell lines including myeloma, head and neck
squamous cell carcinoma (HNSCC), breast and prostate cancers, cholangiocarcinoma, and nonsmall cell lung carcinoma (NSCLC) (Catlett-Falcone et al., 1999; Sriuranpong et al., 2003; Gao et
al., 2007). Recently activating mutations in the gp130 receptor have been identified, enabling
Stat3 binding in the absence of ligand; this was shown to occur in 60% of inflammatory
hepatocellular adenomas (Rebouissou et al., 2009). Persistent IL6/JAK/Stat3 activation can also
result from feedforward signalling as Stat3 can transcriptionally increase IL6 expression (Yu et
48
Figure 2.13: Cytoplasmic domain of gp130 in signal transduction (Hirano et al., 2000).
Upon gp130 activation, receptor associated JAKs may phosphorylate tyrosines (Y) 759, 767, 814,
905 or 915 of human gp130. Y759 is a binding site for the SH2 domain of SHP-2, which can
result in MAPK signalling or PI3-kinase activation. Y759 also recognizes the SOCS-3 inhibitor
of cytokine signalling. Y767, Y814, Y905 and Y915 all have the conserved YXXG motif which
contains a glutamine at position +3 of tyrosine; any of these tyrosines are able to bind the SH2
domain of Stat3. (Hirano et al., 2000)
49
al., 2009). Indeed high levels of serum IL6 have been detected in colorectal, liver, lung and
breast cancer patients and in the latter this correlates with a more advanced tumor stage,
metastases and an overall poor prognosis (Sullivan et al., 2009). In myeloma, as well as other
cancer types, IL6 production has also been observed in non-transformed stromal cells suggesting
both autocrine and paracrine mechanisms of Stat3 activation (Yu et al., 2009).
Inhibition of gp130 through dominant negative mutants was reported to downregulate
tumor invasion and angiogenesis by blocking constitutive Stat3 signalling in the breast cancer
line MDA-MB-231 (Selander et al., 2004). Similarly, treatment of NSCLC-derived lines with
anti-gp130 or anti-IL6 antibodies reduced Stat3 activity and decreased in vitro and in vivo growth
of H2355 cells (Gao et al., 2007). Gao et al. further demonstrated that these tumor-derived lines
produced high levels of IL6, which was secreted into the growth medium and could trigger Stat3ptyr705 phosphorylation in normal MCF-10A cells (Gao et al., 2007). In addition, tissue
microarray samples of lung adenocarcinomas revealed a positive correlation between pStat3positive samples and high IL6 levels (Gao et al., 2007). These data taken together identify the
gp130/JAK/Stat3 axis as a critical marker in carcinogenesis and drugs inhibiting this pathway
may be valuable therapeutic targets for the malignancies which exploit it.
50
5. Previous results from our lab: cell to cell adhesion activates Stat3
Interactions with neighboring cells can profoundly influence signal transduction
pathways, many of which are activated or inactivated by phosphorylation at specific tyrosine
residues (Juliano, 2002; Stemmler, 2008). Dense cell cultures, albeit only two-dimensional, may
in part mimic some of the physiological stress signals that occur in fast-growing tumors. Indeed,
many of the cells in tumor nests lack direct interaction with surrounding connective tissue and
they may depend on survival signalling generated by cell-cell contacts.
We recently discovered that cell density causes a dramatic increase in Stat3 activity in
breast carcinoma as well as normal epithelial cells or fibroblasts (Vultur et al., 2004; Vultur et al.,
2005). Stat3 activation levels peaked at approximately 1-2 days post-confluence, when cell-cell
contacts were highest and cell-matrix contacts were minimal. The density-dependent Stat3
activation was substantially greater than that brought about by serum or EGF stimulation (Vultur
et al., 2004). Treatment of postconfluent cells with a calcium chelating solution (EGTA/EDTA)
abrogated Stat3 tyr705 phosphorylation, DNA binding activity and transcription, yet when the
dispersed cells were allowed to aggregate on agar-coated petris, Stat3 activity was significantly
increased, indicating that cell-cell adhesion is likely the cause behind the Stat3 activation in dense
cultures.
Exploration of the nature of the kinase(s) involved revealed that this Stat3 activation was
independent of Src/Fyn/Yes, Fer, Abl, EGFR and the Insulin Growth Factor-1 Receptor, often
activated in a variety of cancers. In addition, cell density did not lead to activation of the
Ras/Raf/Erk pathway, which is downstream of a number of receptor and non-receptor tyrosine
kinases. These results indicate that, individually, these kinases may not play a role in activating
Stat3 in our system. Stat3 activation was partially suppressed following inhibition of JAKs with
the JAK-selective inhibitor, AG490. Most significantly, inhibition of Stat3 in postconfluent cells
resulted in a dramatic increase in apoptotic cell death indicating that this pathway constitutes a
strong survival signal (Vultur et al., 2004). By elucidating the underlying mechanism of this
51
novel cascade we may be able to identify the tumors that rely on it for growth and survival; this
may lead to the further validation of this pathway as an important target for anti-cancer drug
design.
52
Hypothesis and Objectives
Given our previous findings, our aim was to determine the upstream signals involved in
triggering the density-mediated Stat3 activation. Cadherins are the primary mediators of calciumdependent intercellular adhesion, and cadherin engagement has been shown to trigger signalling
to the Rho family of GTPases, notably Rac1 and Cdc42. Independent studies have shown that
Rho GTPases can lead to Stat3 activation in a JAK-dependent manner, possibly by promoting
transcription of the IL6 cytokine, a key transducer of Stat3 signalling.
Our hypothesis was that cell-cell adhesion by cadherins could activate Rho family
GTPases, which would lead to a potent activation of Stat3, through the transcriptional
upregulation of IL6 family cytokines.
More specifically our objectives were:
(1)
To identify the type of cadherin(s) involved in the density-mediated, Stat3 activation.
(2)
To determine the nature of the Rho family GTPase(s), which are activated by the
cadherin(s).
(3)
To examine the role of the IL6 family of cytokines in the density-dependent Stat3
signalling.
While investigating the mechanism of the density-mediated Stat3 activation, we
discovered that certain common transfection techniques such as lipofectamine or calcium
phosphate precipitation could upregulate Stat3 tyr-705 phosphorylation and activity. Our
observations prompted us to generate cell lines stably expressing specific genes or their mutants
for all subsequent studies where Stat3 function was examined.
Most significantly, our results uncovered a novel pathway of Stat3 activation triggered by
homophilic interactions between E-cadherin or cadherin-11, prototype members of type I and II
classical cadherins, in different cell systems. Stat3 activation was preceded by an increase in
53
Rac1 and Cdc42 activity which could, in turn, promote the expression of members of the IL6
family of cytokines, responsible for the Stat3 stimulation observed. Unexpectedly, in addition to
this cadherin-induced activation of Rac1/Cdc42, we demonstrated, for the first time, a dramatic
increase in the endogenous levels of total Rac1/Cdc42 proteins at high cell densities. Our results
revealed that cadherin engagement could inhibit the ubiquitination and degradation of these
GTPases, leading to a further increase in their activity. Inhibition of cadherin engagement
induced apoptosis, indicating that this pathway plays a crucial role in cell survival signalling.
Upon further examination of the mechanism of Stat3 activation by Rac1/Cdc42, we also
observed increased stability of mutationally activated forms of Rac1 (RacV12 or RacL61) and Cdc42
(Cdc42 V12) with cell density which was due to the inhibition of their modification by ubiquitin.
Moreover, we clearly showed that RacV12 expression can activate Stat3 through an NFκBmediated increase in mRNA of members of the IL6 family of cytokines. In fact, knockdown
experiments indicated that gp130 receptor function, and Stat3 activation, are essential for the
migration and proliferation of RacV12-expressing cells, thereby demonstrating that the
gp130/Stat3 axis represents an essential target of activated Rac in the regulation of both of these
fundamental cellular functions.
54
Significance
Cells in normal tissues or tumors have extensive opportunities for adhesion to their
neighbors in a three-dimensional organization, and it has recently become apparent that in the
study of fundamental cellular processes it is important to take into account the effect of
surrounding cells. Indeed, the relevance of interactions observed in densely growing, cultured
cells to human cancer is clearly evident from findings demonstrating a close correspondence of
genes expressed in lymph node carcinoma of the prostate (LNCaP) cells cultured to high, but not
low, densities, with genes associated with prostate cancer in vivo.
The aim of my project is based on previous data from our lab demonstrating that cell
density can activate Stat3 independently of a number of known tumor-promoting tyrosine kinases,
including the Src family, IGF1-R, Abl, Fer and EGFR, often activated in a variety of cancers.
Most importantly, this density-dependant Stat3 activation constitutes a potent survival signal,
which could be exploited by cancer cells. Hence, given that Stat3 is associated with malignant
cell proliferation, cell survival, angiogenesis, and metastasis, our identification of a novel
mechanism of Stat3 activation triggered by cadherin engagement and the upregulation of
Rac1/Cdc42 GTPases has obvious implications in the development of targeted therapies. This
pathway could be a prognostic marker, as well as a promising predictive target for the treatment
of tumors which may be resistant to inhibition of many tyrosine kinase oncogenes known to be
hyperactive in many cancers.
55
References
Aaronson, D. S. and Horvath, C. M. (2002). A road map for those who don't know JAK-STAT.
Science 296, 1653-1655.
Akasaki, Y., Liu, G., Matundan, H. H., Ng, H., Yuan, X., Zeng, Z., Black, K.L., and Yu, J.S.
(2006). A peroxisome proliferator-activated receptor-gamma agonist, troglitazone, facilitates
caspase-8 and -9 activities by increasing the enzymatic activity of protein-tyrosine phosphatase1B on human glioma cells. J. Biol. Chem. 281, 6165–6174.
Akira, S. (2000). Roles of STAT3 defined by tissue-specific gene targeting. Oncogene 19, 26072611.
Alonzi, T., Middleton, G., Wyatt, S., Buchman, V., Betz, U.A., Muller, W., Musiani, P., Poli, V.,
and Davies, A.M. (2001). Role of Stat3 and PI3-kinase/Akt in mediating the survival actions of
cytokines on sensory neurons. Mol Cell Neurosci. 18, 270-282.
Alvarez, J.V., Febbo, P.G., Ramaswamy, S., Loda, M., Richardson, A., Frank, D.A. (2003).
Identification of a genetic signature of activated signal transducer and activator
of transcription 3 in human tumors. Cancer Res. 65, 5054-5062.
Aoki, Y., Feldman, G. M., and Tosato, G. (2003). Inhibition of STAT3 signalling induces
apoptosis and decreases survivin expression in primary effusion lymphoma. Blood 101, 15351542.
Aplin, A. E., Howe, A., Alahari, S. K., and Juliano, R. L. (1998). Signal transduction and signal
modulation by cell adhesion receptors: the role of integrins, cadherins, immunoglobulin-cell
adhesion molecules, and selectins. Pharmacol. Rev. 50, 197-263.
Aznar, S. and Lacal, J.C. (2001). Rho signals to cell growth and apoptosis. Cancer Letters 165, 110.
Aznar, S., Valeron, P.F., del Rincon, S.V., Perez, L.F., Perona, R. and J.C. Lacal. (2001).
Simultaneous tyrosine and serine phosphorylation of Stat3 transcription factor is involved in
RhoA GTPase oncogenic transformation. Mol. Biol. Cell 12, 3282-3294.
Balsamo, J., Leung, T., Ernst, H., Zanin, M.K., Hoffman, S. and Lilien, J. (1996). Regulated
binding of PTP1B-like phosphatase to N-cadherin: control of cadherin-mediated adhesion by
dephosphorylation of β-catenin. J. Cell Biol. 134, 801–813.
Bar-Sagi, D. and Hall, A. (2000). Ras and Rho GTPases: a family reunion. Cell 103, 227-238.
Batlle, E., Sancho, E., Franci, C., Dominguez, D., Monfar, M., Baulida, J., and Garcia, D. H.
(2000). The transcription factor snail is a repressor of E-cadherin gene expression in epithelial
tumour cells. Nat. Cell Biol. 2, 84-89.
Baus, D. and Pfitzner, E. (2006). Specific function of STAT3, SOCS1, and SOCS3 in the
regulation of proliferation and survival of classical Hodgkin lymphoma cells. Int. J. Cancer 118,
1404-1413.
56
Benitah, S. A., Valeron, P. F., van Aelst, L., Marshall, C. J., and Lacal, J. C. (2004). Rho
GTPases in human cancer: an unresolved link to upstream and downstream transcriptional
regulation. Biochim. Biophys. Acta 1705, 121-132.
Boggon, T.,J., Murray, J., Chappuis-Flament, S., Wong, E., Gumbiner, B.M., and Shapiro, L.,
(2002). C-cadherin ectodomain structure and implications for cell adhesion mechanisms. Science.
296, 1308-1318.
Bokoch, G.M., Bohl, B.P., and Chuang, T.H. (1994). Guanine nucleotide exchange regulates
membrane translocation of Rac/Rho GTP-binding proteins. J. Biol. Chem. 269, 31674-31679.
Bolos, V., Peinado, H., Perez-Moreno, M. A., Fraga, M. F., Esteller, M., and Cano, A. (2003).
The transcription factor Slug represses E-cadherin expression and induces epithelial to
mesenchymal transitions: a comparison with Snail and E47 repressors. J. Cell Sci. 116, 499-511.
Bowman, T., Garcia, R., Turkson, J., and Jove, R. (2001). STATs in oncogenesis. Oncogene. 19:
2474-2488.
Braga, V.M. (2000). Epithelial cell shape: cadherins and small GTPases. Exp. Cell Res. 25, 8390.
Braga, V. M., Del Maschio, A., Machesky, L., and DeJana, E. (1999). Regulation of cadherin
function by Rho and Rac: modulation by junction maturation and cellular context. Mol. Biol. Cell
10, 9-22.
Briggs, M. W. and Sacks, D. B. (2003). IQGAP proteins are integral components of cytoskeletal
regulation. EMBO Rep. 4, 571-574.
Bromberg, J. F., Wrzeszczynska, M. H., Devgan, G., Zhao, Y., Pestell, R. G., Albanese, C., and
Darnell, J. E., Jr. (1999). Stat3 as an oncogene. Cell 98, 295-303.
Bustos, R.I., Forget, M.-A., Settleman, J.E., and Hansen, S.H. (2008). Coordination of Rho and
Rac GTPase function via p190 RhoGAP. Curr. Biol. 18, 1606-1611.
Caldenhoven, E., van Dijk, T.B., Solari, R., Armstrong, J., Raaijmakers, J.A., Lammers, J.W.,
Koenderman, L., de Groot, R.P. (1996). Stat3beta, a splice variant of transcription factor STAT3,
is a dominant negative regulator of transcription. J. Biol. Chem. 31, 13221-13227.
Catlett-Falcone, R., Landowski, T.H., Oshiro, M.M., Turkson, J., Levitzki, A., Savino, R.,
Ciliberto, G., Moscinski, L., Fernandez-Luna, J.L., Nunez, G., Dalton, W.S. and Jove, R. (1999).
Constitutive activation of Stat3 signalling confers resistance to apoptosis in human U266
myeloma cells. Immunity. 10: 105-115.
Cavallaro, U. and Christofori, G. (2004). Cell adhesion and signalling by cadherins and Ig-CAMs
in cancer. Nat Rev Cancer. 4: 118-132.
Chan, K.S., Carbajal, S., Kiguchi, K., Clifford, J., Sano, S. and DiGiovanni, J. (2004). Epidermal
growth factor receptor-mediated activation of Stat3 during multistage skin carcinogenesis. Cancer
Res. 64, 2382-2389.
57
Charrasse, S., M. Meriane, F. Comunale, A. Blangy, and Gauthier-Rouviere, C.
(2002). N-cadherin-dependent cell–cell contact regulates Rho GTPases and β-catenin localization
in mouse C2C12 myoblasts. J. Cell Biol. 158, 953-965.
Chen, Q., Watson, J.T., Marengo, S.R., Decker, K.S., Coleman, I., Nelson, P.S., and Sikes, R.A.
(2006). Gene expression in the LNCaP human prostate cancer progression model: progression
associated expression in vitro corresponds to expression changes associated with prostate cancer
progression in vivo. Cancer Lett. 244, 274-288.
Chen, C. P., Posy, S., Ben Shaul, A., Shapiro, L., and Honig, B. H. (2005). Specificity of cell-cell
adhesion by classical cadherins: Critical role for low-affinity dimerization through beta-strand
swapping. Proc. Natl. Acad. Sci. U. S. A 102, 8531-8536.
Chen, W., Daines, M. O., and Khurana Hershey, G. K. (2004). Turning off signal transducer and
activator of transcription (STAT): the negative regulation of STAT signalling. J. Allergy Clin.
Immunol. 114, 476-489.
Chen, X.P., Losman, J.A. and Rothman, P. (2000). SOCS proteins, regulators of intracellular
signalling.. Immunity 13, 287-290.
Chu, K., Cheng, C., Ye, X., Lee, Y., Zurita, A.J., Chen, D., Yu-Lee, L., Zhang, S., Yeh, E.T., Hu,
M.C., Logothetis, C.J., and Lin, S. (2008). Cadherin-11 promotes the metastasis of prostate
cancer cells to bone. Mol. Cancer Res 6, 1259-67.
Chu, Y.-S., Eder, O., Thomas, W.A., Simcha, I., Pincet, F., Ben-Ze'ev, A., Perez, E., Thiery, J.P.,
and Dufour, S. (2006). Prototypical Type I E-cadherin and Type II Cadherin-7 Mediate Very
Distinct Adhesiveness through Their Extracellular Domains. J. Biol. Chem. 281, 2901-2910.
Chung, C.D., Liao, J., Liu, B., Rao, X., Jay, P., Berta, P., and Shuai, K. (1997). Specific
inhibition of Stat3 signal transduction by PIAS3. Science 278, 1803-1805.
Clevenger, C. V. (2004). Roles and regulation of stat family transcription factors in human breast
cancer. Am. J. Pathol. 165, 1449-1460.
Conacci-Sorrell, M., Zhurinsky, J., and Ben Ze'ev, A. (2002). The cadherin-catenin adhesion
system in signalling and cancer. J. Clin. Invest 109, 987-991.
Courjean, O., Chevreux, G., Perret, E., Morel, A., Sanglier, S., Potier, N., Engel, J., van
Dorsselaer, A., and Feracci, H. (2008). Modulation of E-cadherin monomer folding by
cooperative binding of calcium ions. Biochemistry 47, 2339-2349.
Daniel, J. M. and Reynolds, A. B. (1999). The catenin p120(ctn) interacts with Kaiso, a novel
BTB/POZ domain zinc finger transcription factor. Mol. Cell. Biol. 19, 3614-3623.
Darnell, J. E., Jr. (1997). STATs and gene regulation. Science 277, 1630-1635.
Debidda, M., Wang, L., Zang, H., Poli, V., and Zheng, Y. (2005). A role of STAT3 in Rho
GTPase-regulated cell migration and proliferation. J. Biol. Chem. 280, 17275-17285.
Decker, T. and Kovarik, P. (2000). Serine phosphorylation of STATs. Oncogene 19, 2628-2637.
58
Ellenbroek, S. I. and Collard, J. G. (2007). Rho GTPases: functions and association with cancer.
Clin. Exp. Metastasis 24, 657-672.
Epling-Burnette, P.K., Liu, J.H., Catlett-Falcone, R., Turkson, J., Oshiro, M., Kothapalli, R., Li,
Y.X., Wang, J.M., Yang-Yen, H.F., Karras, J., Jove, R., and Loughran, T.P. (2001). Inhibition of
STAT3 signalling leads to apoptosis of leukemic large granular lymphocytes and decreased Mcl1 expression. J. Clin. Invest. 107, 351-361.
Fairlie, W. D., De Souza, D., Nicola, N. A., and Baca, M. (2003). Negative regulation of gp130
signalling mediated through tyrosine-757 is not dependent on the recruitment of SHP2. Biochem.
J. 372, 495-502.
Faruqi, T. R., Gomez, D., Bustelo, X. R., Bar-Sagi, D., and Reich, N. C. (2001). Rac1 mediates
STAT3 activation by autocrine IL-6. Proc. Nat. Acad. Sci. USA 98, 9014-9019.
Feig, L.A. (1999). Tools of the trade: Use of dominant-inhibitory mutants of Ras-family GTPase.
Nature Cell Biol. 1, E25–E27.
Feltes, C. M., Kudo, A., Blaschuk, O., and Byers, S. W. (2002). An alternatively spliced
cadherin-11 enhances human breast cancer cell invasion. Cancer Res. 62, 6688-6697.
Fernandez-Mercado, M., Cebrian, V., Euba, B., Garcia-Granero, M., Calasanz, M. J., Novo, F. J.,
Vizmanos, J. L., and Garcia-Delgado, M. (2008). Methylation status of SOCS1 and SOCS3 in
BCR-ABL negative and JAK2V617F negative chronic myeloproliferative neoplasms. Leuk. Res.
32, 1638-1640.
Figueroa, A., Kotani, H., Toda, Y., Mazan-Mamczarz, K., Mueller, E. C., Otto, A., Disch, L.,
Norman, M., Ramdasi, R. M., Keshtgar, M., Gorospe, M., and Fujita, Y. (2009). Novel roles of
hakai in cell proliferation and oncogenesis. Mol. Biol. Cell 20, 3533-3542.
Finley, D. (2009). Recognition and processing of ubiquitin-protein conjugates by the proteasome.
Annu. Rev. Biochem. 78, 477-513.
Fu, X. Y., Schindler, C., Improta, T., Aebersold, R., and Darnell, J. E., Jr. (1992). The proteins of
ISGF-3, the interferon alpha-induced transcriptional activator, define a gene family involved in
signal transduction. Proc. Natl. Acad. Sci. U. S. A 89, 7840-7843.
Fujita, Y., Krause, G., Scheffner, M., Zechner, D., Leddy, H. E., Behrens, J., Sommer, T., and
Birchmeier, W. (2002). Hakai, a c-Cbl-like protein, ubiquitinates and induces endocytosis of the
E-cadherin complex. Nat. Cell Biol. 4, 222-231.
Fujitani, Y., Hibi, M., Fukada, T., Takahashitezuka, M., Yoshida, H., Yamaguchi, T., Sugiyama,
K., Yamanaka, Y., Nakajima, K. and Hirano, T. (1997). An alternative pathway for STAT
activation that is mediated by the direct interaction between JAK and
STAT. Oncogene 14, 751-761.
Fukata, M. and Kaibuchi, K. (2001). Rho-family GTPases in cadherin-mediated cell-cell
adhesion. Nat. Rev. Mol. Cell Biol. 2, 887-897.
Gao, S. P., Mark, K. G., Leslie, K., Pao, W., Motoi, N., Gerald, W. L., Travis, W. D., Bornmann,
W., Veach, D., Clarkson, B., and Bromberg, J. F. (2007). Mutations in the EGFR kinase domain
59
mediate STAT3 activation via IL-6 production in human lung adenocarcinomas. J. Clin. Invest
117, 3846-3856.
Gao, Z.-H., Seeling, J.M., Hill, V., Yochum, A., and Virshup, D.M. (2002). Casein kinase I
phosphorylates and destabilizes the β-catenin degradation complex. Proc. Nat. Acad. Sci. USA
99, 1182-1187.
Garcia, R., Bowman, T. L., Niu, G., Yu, H., Minton, S., Muro-Cacho, C. A., Cox, C. E., Falcone,
R., Fairclough, R., Parsons, S., Laudano, A., Gazit, A., Levitzki, A., Kraker, A., and Jove, R.
(2001). Constitutive activation of Stat3 by the Src and JAK tyrosine kinases participates in
growth regulation of human breast carcinoma cells. Oncogene 20, 2499-2513.
Genth, H., Aktories, K. and Just, I. (1999). Monoglucosylation of RhoA at threonine 37 blocks
cytosol-membrane cycling. J. Biol. Chem. 274, 29050-29056.
Germain, D. and Frank, D. A. (2007). Targeting the cytoplasmic and nuclear functions of signal
transducers and activators of transcription 3 for cancer therapy. Clin. Cancer Res. 13, 5665-5669.
Gough, D. J., Corlett, A., Schlessinger, K., Wegrzyn, J., Larner, A. C., and Levy, D. E. (2009).
Mitochondrial STAT3 supports Ras-dependent oncogenic transformation. Science 324, 17131716.
Grandis, J. R., Drenning, S. D., Chakraborty, A., Zhou, M. Y., Zeng, Q., Pitt, A. S., and Tweardy,
D. J. (1998). Requirement of Stat3 but not Stat1 activation for epidermal growth factor receptormediated cell growth in vitro. J. Clin. Invest. 102, 1385-1392.
Gritsko, T., Williams, A., Turkson, J., Kaneko, S., Bowman, T., Huang, M., Nam, S., Eweis, I.,
Diaz, N., Sullivan, D., Yoder, S., Enkemann, S., Eschrich, S., Lee, J.H., Beam, C.A., Cheng, J.,
Minton, S., Muro-Cacho, C.A., and Jove, R. (2006). Persistant activation of Stat3 signalling
induces survivin gene expression and confers resistance to apoptosis in human breast cancer cells.
Clin. Cancer Res. 12, 11-19.
Groner, B., Lucks, P., and Borghouts, C. (2008). The function of Stat3 in tumor cells and their
microenvironment. Semin. Cell Dev. Biol. 19, 341-350.
Gu, J., Li, G., Sun, T., Zhang, X., Shen, J., Tian, Z., and Zhagn, J. (2008). Blockage of the
STAT3 signalling pathway with a decoy oligonucleotide suppresses growth of human malignant
glioma cells. J. Neurooncol. 89, 9-17.
Gu, F., Dube, N., Kim, J. W., Cheng, A., Ibarra-Sanchez, M. J., Tremblay, M. L., and Boisclair,
Y. R. (2003). Protein tyrosine phosphatase 1B attenuates growth hormone-mediated JAK2-STAT
signalling. Mol. Cell Biol. 23, 3753-3762.
Handschuh, G., Luber, B., Hutzler, P., Hofler, H., and Becker, K.F. (2001). Single amino acid
substitutions in conserved extracellular domains of E-cadherin differ in their functional
consequences. J. Mol. Biol. 314, 445-454.
Hakoshima, T., Shimizu, T., and Maesaki, R. (2003). Structural basis of the Rho GTPase
signalling. J. Biochem. 134, 327-331.
60
Hazan, R. B., Kang, L., Whooley, B. P., and Borgen, P. I. (1997). N-cadherin promotes adhesion
between invasive breast cancer cells and the stroma. Cell Adhes. Commun. 4, 399-411.
Heinrich, P. C., Behrmann, I., Haan, S., Hermanns, H. M., Muller-Newen, G., and Schaper, F.
(2003). Principles of interleukin (IL)-6-type cytokine signalling and its regulation. Biochem. J.
374, 1-20.
Heinrich, P. C., Behrmann, I., Muller-Newen, G., Schaper, F., and Graeve, L. (1998). Interleukin6-type cytokine signalling through the gp130/Jak/STAT pathway. Biochem. J 334, 297-314.
Heisterkamp, N., Kaartinen, V., van Soest, S., Bokoch, G.M., and Groffen, J. (1993). Human
ABR encodes a protein with GAPrac activity and homology to the DBL nucleotide exchange
factor. J. Biol. Chem. 268, 16903-16906.
Herrmann, A., Vogt, M., Monnigmann, M., Clahsen, T., Sommer, U., Haan, S., Poli, V.,
Heinrich, P.C., and Muller-Newen, G. (2007). Nucleocytoplasmic shuttling of persistantly
activated STAT3. J. Cell Science. 120, 3249-3261.
Hirano, T., Ishihara, K., and Hibi, M. (2000). Roles of STAT3 in mediating the cell growth,
differentiation and survival signals relayed through the IL-6 family of cytokine receptors.
Oncogene 19, 2548-2556.
Hirohashi, S. (1998). Inactivation of the E-cadherin-mediated cell adhesion system in human
cancers. Am. J. Pathol. 153, 333-339.
Hordijk, P. L., ten Klooster, J.P., van der Kammen, R.A., Michiels, F., Oomen, L.C.J.M., and
Collard, J.G. (1997). Inhibition of invasion of epithelial cells by Tiam1–Rac signalling. Science
278, 1464–1466.
Horvath, C. M., Stark, G. R., Kerr, I. M., and Darnell, J. E., Jr. (1996). Interactions between
STAT and non-STAT proteins in the interferon-stimulated gene factor 3 transcription complex.
Mol. Cell Biol. 16, 6957-6964.
Horvath, C.M. (2004). Silencing STATs: lessons from paramyxovirus interferon evasion.
Cytokine Growth Factor Rev. 15, 117-127.
Hoschuetzky , H., Aberle, H., Kemler, R. (1994). Beta-catenin mediates the interaction of the
cadherin-catenin complex with epidermal growth factor receptor. J. Cell Biol. 127, 1375-1380.
Hung, W. and Elliott, B. (2001). Co-operative effect of c-Src tyrosine kinase and Stat3 in
activation of hepatocyte growth factor expression in mammary carcinoma cells. J. Biol. Chem.
276, 12395-12403.
Ihara, K., Muraguchi, S., Kato, M., Shimizu, T., Shirakawa, M., Kuroda, S., Kaibuchi,
K. and Hakoshima, T. (1998) Crystal structure of human RhoA in a dominantly active
form complexed with a GTP analogue. J. Biol. Chem. 273, 9656-9666.
Imada, K. and Leonard, W. J. (2000). The Jak-STAT pathway. Mol. Immunol. 37, 1-11.
Iwamoto, T., Senga, T., Naito, Y., Matsuda, S., Miyake, Y., Yoshimura, A., and
61
Hamaguchi, M. (2000). The JAK-inhibitor, JAB/SOCS-1 selectively inhibits cytokine-induced,
but not v-Src induced JAK-STAT activation. Oncogene. 19, 4795-4801.
Jeanes, A., Gottardi, C. J., and Yap, A. S. (2008). Cadherins and cancer: how does cadherin
dysfunction promote tumor progression? Oncogene 27, 6920-6929.
Juliano, R. L. (2002). Signal transduction by cell adhesion receptors and the cytoskeleton:
functions of integrins, cadherins, selectins, and immunoglobulin- superfamily members. Annu.
Rev. Pharmacol. Toxicol. 42, 283-323.
Just, I., Fritz, G., Aktories, K., Giry, M., Popoff, M.R., Boquet, P., Hegenbarth, S., and von
Eichel-Streiber, C. (1994). Clostridium difficile toxin B acts on the GTP-binding protein Rho. J.
Biol. Chem. 269, 10706-10712.
Karnoub, A.E., Symons, M., Campbell, S.L., and Der, C.J. (2004). Molecular basis for Rho
GTPase signalling specificity. Breast Cancer Res. Treat. 84, 61-71.
Kaufmann, U., Martin, B., Link, D., Witt, K., Zeitler, R., Reinhard, S., Starzinski-Powitz, A.
(1999). M-cadherin and its sisters in development of striated muscle. Cell Tissue Res. 296, 191198.
Kim, D.J., Chan, K.S., Sano, S., and DiGiovanni, J. (2007). Signal transducer and activator of
transcription 3 (Stat3) in epithelial carcinogenesis. Mol. Carcinog. 46, 725-731.
Kim, S.W., Park, J.I., Spring, C.M., Sater, A.K., Otchere, A.A., Daniel, J.M., and McCrea, P.D.
(2004). Non-canonical Wnt signals are modulated by the Kaiso transcriptional repressor and
p120-catenin. Nat. Cell Biol. 6, 1212-1220.
Kisseleva, T., Bhattacharya, S., Braunstein, J., and Schindler, C. W. (2002). Signalling through
the JAK/STAT pathway, recent advances and future challenges. Gene 285, 1-24.
Konnikova, L., Kotecki, M., Kruger, M.M. and Cochran, B.H. (2003). Knockdown of STAT3
expression by RNAi induces apoptosis in astrocytoma cells. BMC Cancer 3, 23.
Kovacs, E.M., Ali, R.G., McCormack, A.J., and Yap, A.S. (2002a). E-cadherin homophilic
ligation directly signals through Rac and PI3-kinase to regulate adhesive
contacts. J. Biol. Chem. 277, 6708–6718.
Kovacs, E.M., M. Goodwin, R.G. Ali, A.D. Paterson, and A.S. Yap. (2002b). Cadherindirected actin assembly: E-cadherin physically associates with the Arp2/3 complex to direct actin
assembly in nascent adhesive contacts. Curr.Biol. 12, 379–382.
Kowalski, P. J., Rubin, M. A., and Kleer, C. G. (2003). E-cadherin expression in primary
carcinomas of the breast and its distant metastases. Breast Cancer Res. 5, R217-R222.
Lampugnani, M.G., Zanetti, A., Breviario, F., Balconi, G., Orsenigo, F., Corada, M., Spagnuolo,
R., Betson, M., Braga, V., and Dejana, E. (2002). VE-cadherin regulated endothelial actin
activating Rac and increasing membrane association of Tiam. Mol. Biol. Cell 13, 1175-1189.
Larue, L. and Bellacosa, A. (2005). Epithelial-mesenchymal transition in development and
cancer: role of phosphatidylinositol 3' kinase/AKT pathways. Oncogene 24, 7443-7454.
62
Larue, L., Ohsugi, M., Hirchenhain, J., and Kemler, R. (1994). E-cadherin null mutant embryos
fail to form a trophectoderm epithelium. Proc. Nat. Acad. Sci. USA 91, 8263-8267.
Le, T. L., Yap, A. S., and Stow, J. L. (1999). Recycling of E-cadherin: a potential mechanism for
regulating cadherin dynamics. J. Cell Biol. 146, 219-232.
Levy, D.E. and Lee, C. (2002). What does Stat3 do? J. Clin. Invest. 109, 1143-1148.
Leckband, D. and Prakasam, A. (2006). Mechanism and dynamics of cadherin adhesion. Annu.
Rev. Biomed. Eng 8, 259-287.
Leong, P. L., Andrews, G. A., Johnson, D. E., Dyer, K. F., Xi, S., Mai, J. C., Robbins, P. D.,
Gadiparthi, S., Burke, N. A., Watkins, S. F., and Grandis, J. R. (2003). Targeted inhibition of
Stat3 with a decoy oligonucleotide abrogates head and neck cancer cell growth. Proc. Natl. Acad.
Sci. U. S. A 100, 4138-4143.
Li, G., Schaider, H., Satyamoorthy, K., Hanakawa, Y., Hashimoto, K. and Herlyn, M. (2001).
Downregulation of E-cadherin and desmoglein 1 by autocrine hepatocyte growth factor during
melanoma development. Oncogene 20, 8125-8135.
Li, L., and Shaw, P.E. (2006). Elevated activity of STAT3C due to higher DNA binding affinity
of phosphotyrosine dimer rather than covalent dimer formation. J. Biol. Chem. 281, 3317233181.
Lickert, H., Bauer, A., Kemler, R., and Stappert, J. (2000). Casein kinase II phosphorylation of Ecadherin increases E-cadherin/β-catenin interaction and strengthens cell-cell adhesion. J. Biol.
Chem. 275, 5090-5095.
Liu, L., McBride, K.M., and Reich, N.C. (2005). Stat3 nuclear import is independent of tyrosine
phosphorylation and mediated by importin-alpha3. Proc. Nat. Acad. Sci. USA 102, 8150-8155.
Liu, L., Gao, Y., Qiu, H., Miller, W.T., Poli, V., and Reich, N.C. (2006). Identification of STAT3
as a specific substrate of breast tumor kinase. Oncogene 25, 4904-4912.
Lodish, H., Berk, A., Zipursky, S.L., Matsudaria, P., Baltimore, D., and Darnell, J. 2000.
Molecular Cell Biology. 4th ed. W.H. Freeman and Company, N.Y., New York. 22.1-22.2
Mackay, D. J. and Hall, A. (1998). Rho GTPases. J. Biol. Chem. 273, 20685-20688.
McBride, K.M., Banninger, G., McDonald, C., and Reich, N.C. (2002). Regulated nuclear import
of the STAT1 transcription factor by direct binding of importin-α. EMBO J. 21, 1754-1763.
Murray, P. J. (2007). The JAK-STAT signalling pathway: input and output integration. J.
Immunol. 178, 2623-2629.
Myers, M. P., Andersen, J. N., Cheng, A., Tremblay, M. L., Horvath, C. M., Parisien, J. P.,
Salmeen, A., Barford, D., and Tonks, N.K. (2001). TYK2 and JAK2 are substrates of proteintyrosine phosphatase 1B. J. Biol. Chem. 276, 47771–47774.
Nelson, W.J. (2008). Regulation of cell-cell adhesion by the cadherin-catenin complex. Biochem.
Soc. Trans. 36, 149-155.
63
Niu, G., Wright, K. L., Ma, Y., Wright, G. M., Huang, M., Irby, R., Briggs, J., Karras, J., Cress,
W. D., Pardoll, D., Jove, R., Chen, J., and Yu, H. (2005). Role of Stat3 in regulating p53
expression and function. Mol. Cell Biol. 25, 7432-7440.
Niu, G.L., Wright, K.L., Huang, M., Song, L.X., Haura, E., Turkson, J., Zhang, S.M., Wang,
T.H., Sinibaldi, D., Coppola, D., Heller, R., Ellis, L.M., Karras, J., Bromberg, J., Pardoll, D.,
Jove, R., and Yu, H. (2002). Constitutive Stat3 activity up-regulates VEGF expression and
tumour angiogenesis. Oncogene 21, 2000-2008.
Noren, N.K., Arthur, W.T. and Burridge, K. (2003). Cadherin engagement inhibits RhoA via
p190RhoGAP. J. Biol. Chem. 278, 13615-13618.
Normanno, N., De Luca, A., Bianco, C., Strizzi, L., Mancino, M., Maiello, M. R., Carotenuto, A.,
De Feo, G., Caponigro, F., and Salomon, D. S. (2006). Epidermal growth factor receptor (EGFR)
signalling in cancer. Gene 366, 2-16.
O'Shea, J. J., Pesu, M., Borie, D. C., and Changelian, P. S. (2004). A new modality for
immunosuppression: targeting the JAK/STAT pathway. Nat. Rev. Drug Discov. 3, 555-564.
Oka, H., Shiozaki, H., Kobayashi, K., Inoue, M., Tahara, H., Kobayashi, T., Takatsuka, Y.,
Matsuyoshi, N., Hirano, S., Takeichi, M., and . (1993). Expression of E-cadherin cell adhesion
molecules in human breast cancer tissues and its relationship to metastasis. Cancer Res. 53, 16961701.
Okazaki, M., Takeshita, S., Kawai, S., Kikuno, R., Tsujimura, A., Kudo, A., and Amann, E.
(1994). Molecular cloning and characterization of OB-cadherin, a new member of cadherin
family expressed in osteoblasts. J. Biol. Chem. 269, 12092-12098.
Ortmann, R. A., Cheng, T., Visconti, R., Frucht, D. M., and O'Shea, J. J. (2000). Janus kinases
and signal transducers and activators of transcription: their roles in cytokine signalling,
development and immunoregulation. Arthritis Res. 2, 16-32.
Oyama, T., Kanai, Y., Ochiai, A., Akimoto, S., Oda, T., Yanagihara, K., Nagafuchi, A., Tsukita,
S., Shibamoto, S., Ito, F., Takeichi, M., Matsuda, H., and Hirohashi, S. (1994). A truncated betacatenin disrupts the interaction between E-cadherin and alpha-catenin: a cause of loss of
intercellular adhesiveness in human cancer cell lines. Cancer Res. 54, 6282-6287.
Palacios, F., Tushir, J. S., Fujita, Y., and D'Souza-Schorey, C. (2005). Lysosomal
targeting of E-cadherin: a unique mechanism for the down-regulation of cell-cell
adhesion during epithelial to mesenchymal transitions. Mol. Cell. Biol 25, 389–402.
Parisini, E., Higgins, J.M., Liu, J.H., Brenner, M.B. and Wang, J.H. (2007). The crystal
structure of human E-cadherin domains 1 and 2, and comparison with other cadherins in
the context of adhesion mechanism. J. Mol. Biol. 373, 401-411.
Patel, I. S., Madan, P., Getsios, S., Bertrand, M. A., and MacCalman, C. D. (2003). Cadherin
switching in ovarian cancer progression. Int. J. Cancer 106, 172-177.
64
Patel, S.D., Ciatto, C., Chen, C.P., Bahna, F., Rajebhosale, M., Arkus, N., Schieren, I., Jessell,
T.M., Honig, B., Price, S.R., and Shapiro, L. (2006). Type II cadherin ectodomain structures:
implications for classical cadherin specificity. Cell 124, 1255-1268.
Pece, S., and Gutkind, J.S. (2002). E-cadherin and Hakai: signalling, remodeling or destruction?
Nat. Cell Biol. 4, E72-74.
Pece, S., Chiariello, M., Murga, C., and Gutkind, J. S. (1999). Activation of the protein kinase
Akt/PKB by the formation of E-cadherin-mediated cell-cell junctions. Evidence for the
association of phosphatidylinositol 3-kinase with the E-cadherin adhesion complex. J. Biol.
Chem. 274, 19347-19351.
Pelletier, S., Duhamel, F., Coulombe, P., Popoff, M. R., and Meloche, S. (2003). Rho family
GTPases are required for activation of Jak/STAT signalling by G protein-coupled receptors. Mol.
Cell. Biol. 23, 1316-1333.
Perrais, M., Chen, X., Perez-Moreno, M., and Gumbiner, B. M. (2007). E-Cadherin Homophilic
Ligation Inhibits Cell Growth and Epidermal Growth Factor Receptor Signalling Independently
of Other Cell Interactions. Mol. Biol. Cell 18, 2013-2025.
Perret, E., Benoliel, A. M., Nassoy, P., Pierres, A., Delmas, V., Thiery, J. P., Bongrand, P., and
Feracci, H. (2002). Fast dissociation kinetics between individual E-cadherin fragments revealed
by flow chamber analysis. EMBO J. 21, 2537-2546.
Perret, E., Leung, A., Feracci, H., and Evans, E. (2004). Trans-bonded pairs of E-cadherin exhibit
a remarkable hierarchy of mechanical strengths. Proc. Nat. Acad. Sci. USA 101, 16472-16477.
Pierres, A., Prakasam, A., Touchard, D., Benoliel, A. M., Bongrand, P., and Leckband, D. (2007).
Dissecting subsecond cadherin bound states reveals an efficient way for cells to achieve ultrafast
probing of their environment. FEBS Lett. 581, 1841-1846.
Pishvaian, M. J., Feltes, C. M., Thompson, P., Bussemakers, M. J., Schalken, J. A., and Byers, S.
W. (1999). Cadherin-11 is expressed in invasive breast cancer cell lines. Cancer Res. 59, 947952.
Qian, X., Karpova, T., Sheppard, A. M., McNally, J., and Lowy, D. R. (2004). E-cadherinmediated adhesion inhibits ligand-dependent activation of diverse receptor tyrosine kinases.
EMBO J. 23, 1739-1748.
Quesnelle, K. M., Boehm, A. L., and Grandis, J. R. (2007). STAT-mediated EGFR signalling in
cancer. J. Cell Biochem. 102, 311-319.
Radice, G. L., Rayburn, H., Matsunami, H., Knudsen, K. A., Takeichi, M., and Hynes, R. O.
(1997). Developmental defects in mouse embryos lacking N-cadherin. Dev. Biol. 181, 64-78.
Raz, R., Lee, C. K., Cannizzaro, L. A., D'Eustachio, P., and Levy, D. E. (1999). Essential role of
STAT3 for embryonic stem cell pluripotency. Proc. Natl. Acad. Sci. U. S. A 96, 2846-2851.
Rebouissou, S., Amessou, M., Couchy, G., Poussin, K., Imbeaud, S., Pilati, C., Izard, T.,
Balabaud, C., Bioulac-Sage, P., and Zucman-Rossi, J. (2009). Frequent in-frame somatic
deletions activate gp130 in inflammatory hepatocellular tumours. Nature 457, 200-204.
65
Reich, N.C. and Liu, L. (2006). Tracking STAT nuclear traffic. Nat. Rev. Immunol. 6, 612-612.
Reynolds, A.B., Daniel, J., McCrea, P.D., Wheelock, M.J., Wu, J., and Zhang, Z. (1994).
Identification of a new catenin: the tyrosine kinase substrate p120cas associates with E-cadherin
complexes. 14, 8333-8342.
Riley, J.K., Takeda, K., Akira, S., and Schreiber, R.D. (1999). Interleukin-10 receptor signalling
through the JAK-STAT pathway. Requirement for two distinct receptor-derived signals for antiinflammatory action. J. Biol. Chem. 274, 16513-16521.
Rivard, N. (2009). Phosphatidylinositol 3-kinase: a key regulator in adherens junction formation
and function. Front Biosci. 14, 510-522.
Roberts, P.J., Mitin, N., Keller, P.J., Chenette, E.J., Madigan, J.P., Currin, R.O., Cox, A.D.,
Wilson, O., Kirschmeier, P. and Der, C.J. (2008). Rho family GTPase modification and
dependence on CAAX motif-signaled posttranslational modification. J. Biol. Chem. 12, 2515025163.
Roura, S., Miravet, S., Piedra, J., Garcı́a de Herreros, A., and Duñach, M. (1999). Regulation of
E-cadherin/Catenin Association by Tyrosine Phosphorylation. J. Biol. Chem. 274, 36734-36740.
Rytinki, M.M., Kaikkonen, S., Pehkonen, P., Jaaskelainen, T., and Palvimo J.J. (2009). PIAS
proteins: pleitrophic interactors associated with SUMO. 66, 3029-3041.
Sahai, E., Olson, M. F., and Marshall, C. J. (2001). Cross-talk between Ras and Rho signalling
pathways in transformation favours proliferation and increased motility. EMBO J. 20, 755-766.
Sander, E. E. and Collard, J. G. (1999). Rho-like GTPases: their role in epithelial cell-cell
adhesion and invasion. Eur. J. Cancer 35, 1905-1911.
Sander, E. E., van Delft, S., ten Klooster, J. P., Reid, T., van der Kammen, R. A., Michiels, F.,
and Collard, J. G. (1998). Matrix-dependent Tiam1/Rac signalling in epithelial cells promotes
either cell-cell adhesion or cell migration and is regulated by phosphatidylinositol 3-kinase. J.
Cell Biol. 143, 1385-1398.
Sano, S., Itami, S., Takeda, K., Tarutani, M., Yamaguchi, Y., Miura, H., Yoshikawa, K., Akira, S.
and Takeda, J. (1999). Keratinocyte-specific ablation of Stat3 exhibits impaired skin remodeling
but does not affect skin morphogenesis. EMBO J. 18, 4657-4668.
Scheller, J. and Rose-John, S. (2006). Interleukin-6 and its receptor: from bench to bedside. Med.
Microbiol. Immunol. 195, 173-183.
Schindler, C. and Darnell, J. E., Jr. (1995). Transcriptional responses to polypeptide ligands: the
JAK-STAT pathway. Annu. Rev. Biochem. 64, 621-651.
Schindler, C. and Plumlee, C. (2008). Inteferons pen the JAK-STAT pathway. Semin. Cell Dev.
Biol. 19, 311-318.
Schust, J., Sperl, B., Hollis, A., Mayer, T.U., and Berg, T. (2006). Stattic: a small-molecule
inhibitor of STAT3 activation and dimerization. Chemistry & Biology 13, 1123-1124.
66
Schmidt, A., and Hall, A. (2002). Guanine nucleotide exchange factors for Rho GTPases: turning
on the switch. Genes Dev. 16, 1587-1609.
Sekimoto, T., Makajima, K., Tachibana, T., Hirano, T., Yoneda, Y. (1996). Interferon-βdependent nuclear import of STAT1 is mediated by the GTPase activity of Ran/TC4. J. Biol.
Chem. 271, 31017–31020.
Selander, K. S., Li, L., Watson, L., Merrell, M., Dahmen, H., Heinrich, P. C., Muller-Newen, G.,
and Harris, K. W. (2004). Inhibition of gp130 signalling in breast cancer blocks constitutive
activation of Stat3 and inhibits in vivo malignancy. Cancer Res. 64, 6924-6933.
Semb, H. and Christofori, G. (1998). The tumor-suppressor function of E-cadherin. Am. J. Hum.
Genet. 63, 1588-1593.
Shen, Y., Hirsch, D. S., Sasiela, C. A., and Wu, W. J. (2007). CDC42 regulates E-cadherin
ubiquitination and degradation through an EGF receptor to SRC-mediated pathway. J. Biol.
Chem 283, 5127–5137.
Simon, A. R., Vikis, H. G., Stewart, S., Fanburg, B. L., Cochran, B. H., and Guan, K. L. (2000).
Regulation of STAT3 by direct binding to the Rac1 GTPase. Science 290, 144-147.
Sriuranpong, V., Park, J. I., Amornphimoltham, P., Patel, V., Nelkin, B. D., and Gutkind, J. S.
(2003). Epidermal growth factor receptor-independent constitutive activation of STAT3 in head
and neck squamous cell carcinoma is mediated by the autocrine/paracrine stimulation of the
interleukin 6/gp130 cytokine system. Cancer Res. 63, 2948-2956.
Stemmler, M. P. (2008). Cadherins in development and cancer. Mol. Biosyst. 4, 835-850.
Stuible, M., Doody, K. M., and Tremblay, M. L. (2008). PTP1B and TC-PTP: regulators of
transformation and tumorigenesis. Cancer Metastasis Rev. 27, 215-230.
Subramanian, P., Torres, B.A., and Johnson, H.M. (2001). So many ligands and so few
transcription factors: a new paradigm for signalling through the STAT transcription factors.
Cytokine 15, 175-187.
Sullivan, N. J., Sasser, A. K., Axel, A. E., Vesuna, F., Raman, V., Ramirez, N., Oberyszyn, T. M.,
and Hall, B. M. (2009). Interleukin-6 induces an epithelial-mesenchymal transition phenotype in
human breast cancer cells. Oncogene 28, 2940-2947.
Sun, J., Blaskovich, M. A., Jove, R., Livingston, S. K., Coppola, D., and Sebti, S. M. (2005).
Cucurbitacin Q: a selective STAT3 activation inhibitor with potent antitumor activity. Oncogene
24, 3236-3245.
Takaishi, K., Sasaki, T., Kotani, H., Nishioka, H. and Takai, Y. (1997). Regulation of cell–cell
adhesion by Rac and Rho small G proteins in MDCK cells. J. Cell Biol. 139, 1047–1059.
Takeda, K., Noguchi, K., Shi, W., Tanaka, T., Matsumoto, M., Yoshida, N., Kishimoto, T., and
Akira, S. (1997). Targeted disruption of the mouse Stat3 gene leads to early embryonic lethality.
Proc. Natl. Acad. Sci. U. S. A. 94, 3801-3804.
67
Takeda, K., Kaisho, T., Yoshida, N., Takeda, J., Kishimoto, T., and Akira, S. (1998). Stat3
activation is responsible for IL-6 dependent T-cell proliferation through preventing apoptosis:
Generation and characterization of T cell-specific Stat3-deficient mice. J. Immunol. 161, 46524660.
Takeda, K., Clausen, B.E., Kaisho, T., Tsujimura, T., Terada, N., Forster, I., and Akira, S. (1999).
Enhanced Th1 activity and development of chronic enterocolitis in mice devoid of Stat3 in
macrophages and neutrophils. Immunity. 10, 39-49.
Takeichi, M. (1990). Cadherins: a molecular family important in selective cell-cell adhesion.
Annu. Rev. Biochem. 59, 237-252.
Tanaka, T., Soriano, M.A., and Grusby, M.J. (2005). SLIM is a nuclear ubiquitin E3 ligase that
negatively regulates STAT signaling. Immunity. 22, 729-736.
Tapon, N., Nagata, K., Lamarche, N., and Hall, A. (1998). A new Rac target POSH is an SH3containing scaffold protein involved in the JNK and NF-κB signalling pathways. EMBO J. 17,
1395-1404.
Terry, L.J., Shows, E.B., and Wente, S.R. (2007). Crossing the nuclear envelope: hierarchical
regulation of nucleocytoplasmic transport. Science. 318, 1412-1416.
Thangaraju, M., Rudelius, M., Bierie, B., Raffeld, M., Sharan, S., Hennighausen, L., Huang, AM., and Sterneck, E. (2005). C/EBPδ is a crucial regulator of pro-apoptotic gene expression
during mammary gland involution. Development. 132, 4675-4685.
Tischoff, I., Hengge, U. R., Vieth, M., Ell, C., Stolte, M., Weber, A., Schmidt, W. E., and
Tannapfel, A. (2007). Methylation of SOCS-3 and SOCS-1 in the carcinogenesis of Barrett's
adenocarcinoma. Gut 56, 1047-1053.
Tomita, K., van Bokhoven, A., van Leenders, G. J., Ruijter, E. T., Jansen, C. F., Bussemakers, M.
J., and Schalken, J. A. (2000). Cadherin switching in human prostate cancer progression. Cancer
Res. 60, 3650-3654.
Turkson, J., Kim, J.S., Zhang, S., Yuan, J., Huang, M., Glenn, M., Haura, E., Sebti, S., Hamilton,
A.D. and Jove, R. (2004). Novel peptidomimetic inhibitors of signal transducer and activator of
transcription 3 dimerization and biological activity. Mol. Cancer Ther. 3, 261-269.
Turkson, J., Ryan, D., Kim, J.S., Zhang, Y., Chen, Z., Haura, E., Laudano, A., Sebti, S.,
Hamilton, A.D., and Jove, R. (2001). Phosphotyrosyl peptides block Stat3-mediated DNA
binding activity, gene regulation, and cell transformation. J. Biol. Chem. 276, 45443-45455.
Turkson, J., Bowman, T., Adnane, J., Zhang, Y., Djeu, J. Y., Sekharam, M., Frank, D. A.,
Holzman, L. B., Wu, J., Sebti, S., and Jove, R. (1999). Requirement for Ras/Rac1-mediated p38
and c-Jun N-terminal kinase signalling in Stat3 transcriptional activity induced by the Src
oncoprotein. Mol. Cell. Biol. 19, 7519-7528.
Turkson, J., Bowman, T., Garcia, R., Caldenhoven, E., de Groot, R. P., and Jove, R. (1998). Stat3
activation by Src induces specific gene regulation and is required for cell transformation. Mol.
Cell. Biol. 18, 2545-2552.
68
Ulane, C.M., Rodriguez, J.J., Parisien, J.-P., and Horvath, C.M. (2003). Stat3 ubiquitylation and
degradation by Mumps virus suppress cytokine and oncogene signaling. J. Virology. 77, 63856393.
Ungureanu, D., Saharinen, P., Junttila, I., Hilton, D.J., Silvennoinen, O. (2002). Regulation of
Jak2 through the ubiquitin-proteasome pathway involves phosphorylation of Jak2 on Y1007 and
interaction with SOCS-1. Mol Cell Biol. 22,3316-3326.
Vega, F. M. and Ridley, A. J. (2008). Rho GTPases in cancer cell biology. FEBS Lett. 582, 20932101.
Velazquez, L., Fellous, M., Stark, G. R., and Pellegrini, S. (1992). A protein tyrosine kinase in
the interferon alpha/beta signalling pathway. Cell 70, 313-322.
Vigneron, A., Gamelin, E., and Coqueret, O. (2008). The EGFR-STAT3 oncogenic pathway upregulates the Eme1 endonuclease to reduce DNA damage after topoisomerase I inhibition. Cancer
Res. 68, 815-825.
Vleminckx, K., Vakaet, L., Jr., Mareel, M., Fiers, W., and van Roy, F. (1991). Genetic
manipulation of E-cadherin expression by epithelial tumor cells reveals an invasion suppressor
role. Cell 66, 107-119.
Vultur, A., Arulanandam, R., Turkson, J., Niu, G., Jove, R., and Raptis, L. (2005). Stat3 is
required for full neoplastic transformation by the Simian Virus 40 Large Tumor antigen.
Molecular Biology of the Cell 16, 3832-3846.
Vultur, A., Cao, J., Arulanandam, R., Turkson, J., Jove, R., Greer, P., Craig, A., Elliott, B. E., and
Raptis, L. (2004). Cell to cell adhesion modulates Stat3 activity in normal and breast carcinoma
cells. Oncogene 23, 2600-2616.
Wang, T., Niu, G., Kortylewski, M., Burdelya, L., Shain, K., Zhang, S., Bhattacharya, R.,
Gabrilovich, D., Heller, R., Coppola, D., Dalton, W., Jove, R., Pardoll, D., and Yu, H. (2004).
Regulation of the innate and adaptive immune responses by Stat-3 signalling in tumor cells. Nat.
Med. 10, 48-54.
Wang, X., Lupardus, P., Laporte, S. L., and Garcia, K. C. (2009). Structural biology of shared
cytokine receptors. Annu. Rev. Immunol. 27, 29-60.
Weinberg, R.A. (2007). The Biology of Cancer. London, England: Garland Science Taylor and
Francis Group.
Wegenka, U.M., Buschmann, J., Lutticken, C., Heinrich, P.C., and Horn, F. (1993). Acute-phase
response factor, a nuclear factor binding to acute-phase response elements, is rapidly activated by
interleukin-6 at the posttranslational level. Mol. Cell Biol. 13, 276-288.
Wen, Z., Zhong, Z., and Darnell, J. E., Jr. (1995). Maximal activation of transcription by
Stat1 and Stat3 requires both tyrosine and serine phosphorylation. Cell 82, 241-250.
Wen, Z. and Darnell, J. E., Jr. (1997). Mapping of Stat3 serine phosphorylation to a single residue
(727) and evidence that serine phosphorylation has no influence on DNA binding of Stat1 and
Stat3. Nucleic Acids Res. 25, 2062-2067.
69
Wheelock, M. J., Soler, A. P., and Knudsen, K. A. (2001). Cadherin junctions in mammary
tumors. J. Mammary. Gland. Biol. Neoplasia. 6, 275-285.
Wheelock, M. J., Shintani, Y., Maeda, M., Fukumoto, Y., and Johnson, K. R. (2008). Cadherin
switching. J. Cell Sci. 121, 727-735.
Wick, K. R., and Berton, M. T. (2000). IL-4 induces serine phosphorylation of the STAT6
transactivation domain in B lymphocytes. Mol. Immunol. 37, 641-652.
Wilde, C., and Aktories, K. (2001). The Rho-ADP-ribosylating C3 exoenzyme from Clostridium
botulinum and related C3-like transferases. Toxicon. 39, 1647-1660.
Williams, E.-J., Williams, G., Howell, F.V., Skaper, S.D., Walsh, F.S., and Doherty, P. (2001).
Identification of an N-cadherin motif that can interact with the fibroblast growth factor receptor
and is required for axonal growth. J. Biol. Chem. 23, 43879-43866.
Wormald, S., and Hilton, D.J. (2003). Inhibitors of cytokine signal transduction. J. Biol. Chem.
279: 821-824.
Xu, X., Sun, Y. L., and Hoey, T. (1996). Cooperative DNA binding and sequence-selective
recognition conferred by the STAT amino-terminal domain. Science 273, 794-797.
Yamamoto, T., Taya, S., and Kaibuchi, K. (1999). Ras-induced transformation and signalling
pathway. J. Biochem. 126, 799-803.
Yamamoto, T., Sekine, Y., Kashima, K., Kubota, A., Sato, N., Aoki, N., and Matsuda, T. (2002).
The nuclear isoform of protein-tyrosine phosphatase TC-PTP regulates interleukin-6-mediated
signalling pathway through STAT3 dephosphorylation. Biochem.
Biophys. Res. Commun. 297, 811–817.
Yamanaka, Y., Nakajima, K., Fukada, T., Hibi, M. and Hirano, T. (1996). Differentiation and
growth arrest signals are generated through the cytoplasmic region of gp130 that is essential for
Stat3 activation. EMBO J. 15, 1557-1565.
Yang, J., Chatterjee-Kishore, M., Staugaitis, S. M., Nguyen, H., Schlessinger, K., Levy, D. E.,
and Stark, G. R. (2005). Novel roles of unphosphorylated STAT3 in oncogenesis and
transcriptional regulation. Cancer Res. 65, 939-947.
Yap, A. S., Brieher, W. M., and Gumbiner, B. M. (1997). Molecular and functional analysis of
cadherin-based adherens junctions. Annu. Rev. Cell Dev. Biol. 13, 119-146.
Yap, A. S., Niessen, C. M., and Gumbiner, B. M. (1998). The Juxtamembrane Region of the
Cadherin Cytoplasmic Tail Supports Lateral Clustering, Adhesive Strengthening, and Interaction
with p120ctn. J. Cell Biol. 141, 779-789.
Yap, A. S. and Kovacs, E. M. (2003). Direct cadherin-activated cell signalling: a view from the
plasma membrane. J. Cell Biol. 160, 11-16.
Yasukawa, H., Misawa, H., Sakamoto, H., Masuhara, M., Sasaki, A., Wakioka, T., Ohtsuka, S.,
Imaizumi, T. Matsuda, T., Ihle, J.N., and Yoshimura, A. (1999). The JAK-binding protein JAB
70
inhibits Janus tyrosine kinase activity through binding in the activation loop. EMBO J. 18, 1309–
1320.
Yu, H. and Jove, R. (2004). The STATs of cancer--new molecular targets come of age. Nat. Rev.
Cancer 4, 97-105.
Yu, H., Pardoll, D., and Jove, R. (2009). STATs in cancer inflammation and immunity: a leading
role for STAT3. Nat. Rev. Cancer 9, 798-809.
Yuan, Z. L., Guan, Y. J., Chatterjee, D., and Chin, Y. E. (2005). Stat3 dimerization regulated by
reversible acetylation of a single lysine residue. Science 307, 269-273.
Zhang, J.J., Zhao, Y., Chait, B.T., Lathem, W.W., Ritzi, M., Knippers, R., and Darnell, J.E., Jr.
(1998). Ser727-dependent recruitment of MCM5 by Stat1α in IFN-γ-induced transcriptional
activation. EMBO J. 17, 6963-6971.
Zhang, J., Yang, J., Roy, S. K., Tininini, S., Hu, J., Bromberg, J. F., Poli, V., Stark, G. R., and
Kalvakolanu, D. V. (2003). The cell death regulator GRIM-19 is an inhibitor of signal transducer
and activator of transcription 3. Proc. Natl. Acad. Sci. U. S. A. 100, 9342-9347.
Zhang, Y., Turkson, J., Carter-Su, C., Smithgall, T., Levitzki, A., Kraker, A., Krolewski, J. J.,
Medveczky, P., and Jove, R. (2000). Activation of Stat3 in v-Src transformed fibroblasts requires
cooperation of Jak1 kinase activity. J. Biol. Chem. 275, 24935
24944.
Zhon, I.M., Campbell, S.L., Khosravi-Far, R., Rossman, K.L. and Der, C.J. (1998). Rho family
proteins and Ras transformation: the RHOad less traveled gets congested. Oncogene 17, 14151438.
Zhou, Y., Li, M., Wei, Y., Feng, D., Peng, C., Weng, H., Ma, Y., Bao, L., Nallar, S., Kalakonda,
S., Xiao, W., Kalvakolanu, D. V., and Ling, B. (2009). Down-regulation of GRIM-19 expression
is associated with hyperactivation of STAT3-induced gene expression and tumor growth in
human cervical cancers. J. Interferon Cytokine Res. 29, 695-703.
71
Chapter 3
Transfection techniques affecting Stat3 activity levels
Rozanne Arulanandam, Adina Vultur, and Leda Raptis.
Departments of Microbiology and Immunology and Pathology and Molecular Medicine, and
Cancer Research Center, Queen's University, Kingston, Ont. Canada.
(2005). Analytical Biochemistry. 338, 83-89.
Author contributions:
RA performed most of the experiments, AV and LR conducted the electroporation work.
72
Abstract
Transfection techniques, such as calcium-phosphate or liposome-mediated gene transfer,
are commonly used for the examination of the effect of a gene upon cellular phenotype and
biochemical properties. We previously demonstrated that cell to cell adhesion causes a dramatic
increase in Stat3 activity. Given that the opportunities for cell to cell adhesion could be altered
due to the presence of the DNA-containing complexes, we examined the effect of the calciumphosphate transfection procedure upon Stat3 activity levels. The results revealed a dramatic
increase in Stat3 phosphorylation at the critical tyr705 site and Stat3 activity following calciumphosphate transfection. This increase was noted even in the absence of DNA and was not due to
the mere presence of calcium ions. In contrast, DNA introduction through electroporation or
infection with a retroviral vector did not affect Stat3 activity, while cationic lipids such as
lipofectamine or Fugene6 had a less pronounced effect than calcium-phosphate transfection.
These results indicate that caution is required in the interpretation of results with regard to
activity of Stat3 following certain commonly used transient transfection regimens.
73
Introduction
The signal transducer and activator of transcription-3 (Stat3) protein is activated by
cytokines as well as growth factors such as EGF or PDGF through their receptors, or the nonreceptor tyrosine kinase Src (reviewed in Yu and Jove, 2004). Stat3 is invariably latent in the
cytoplasm and, subsequent to binding to an activated receptor through its SH2 domain, becomes
activated through phosphorylation by the receptor itself or by the associated JAKs, or cSrc
tyrosine kinases (Wang et al., 2000; Bromberg et al., 1998; Turkson et al., 1998).
Phosphorylation of a single critical tyrosine residue, tyr705, activates Stat3 by stabilizing the
association of two Stat3 monomers through reciprocal SH2-ptyr interactions to form a dimer
which migrates to the nucleus and binds specific DNA sites to initiate transcription from a
number of genes. Intense interest in Stat3 signalling was sparked by the discovery that
inappropriate activation of Stat3, along with Stat5, occurs with high frequency in a number of
human cancers, while constitutively high Stat3 activity is required for the growth and survival of
many tumor-derived, cultured cells (Song and Grandis, 2000; Coffer et al., 2000; Niu et al., 2002;
Yu and Jove, 2004).
The investigation of Stat3 function has often been accomplished through the introduction
of a variety of genes, such as activated forms or dominant negative mutants of different signal
transducers, through DNA transfection into cultured cells. In many cases, the experiments
involved transient expression of these genes, followed by examination of their effect upon Stat3
activity at 48-72 hours after transfection (Turkson et al., 1999; Turkson et al., 1998). A very
commonly used transfection method is the calcium-phosphate precipitation technique (Graham
and van der Eb, 1973); a precipitate is formed between calcium and phosphate ions and the added
DNA, the coarseness of which can be adjusted through precise control of the pH. The mixture is
added to the culture through the medium and the grains are taken up by the cells.
74
We previously demonstrated that cell to cell adhesion causes a dramatic increase in Stat3
activity levels in carcinoma as well as normal epithelial cells or fibroblasts (Vultur et al., 2004).
This is unique to Stat3 since no activation of other pathways, often coordinately activated with
Stat3 by growth factors or oncogenes, such as the Ras/Raf/Erk1/2 or PI3-kinase/Akt, was
observed under these conditions (Vultur et al., 2004). Since the opportunities for cell to cell
adhesion could be altered due to the presence of the calcium phosphate/DNA precipitate, we
examined the effect of this transfection procedure upon Stat3 activity levels. The results revealed
a dramatic increase in Stat3-tyr705 phosphorylation and consequently Stat3 activity following
calcium phosphate transfection, as determined by measurement of phosphorylation of the critical
tyrosine 705 residue, or transcriptional activity using a Stat3-specific reporter. This increase was
noted even in the absence of DNA and was not due to the mere presence of calcium ions. In
contrast, DNA introduction through electroporation or infection with a retroviral vector did not
affect Stat3 activity, while cationic lipids such as Lipofectamine or Fugene6 had a less
pronounced effect than calcium phosphate transfection. These results indicate that caution is
required in the interpretation of results regarding activity of Stat3, following certain commonly
used, transient transfection regimens.
75
Materials and Methods
Cell lines and culture techniques
Mouse NIH 3T3, 10T½ or rat F111 fibroblasts (Raptis et al., 2000) were grown in plastic
3 cm dishes in Dulbecco’s Modified Eagle Medium (DMEM) supplemented with 10% calf
serum, in a 7% CO2 incubator. For calcium phosphate transfection, cells were transfected with
the β-gal plasmid (Turkson et al., 1998), using the technique of Chen and Okayama (Chen and
Okayama, 1987) with some modifications. When they reached confluences of 15%, 35% or 60%,
the DNA was prepared by dissolving a dried pellet of 4 μg DNA with 135 μl sterile H2O and
adding 15 μl 2.5 M CaCl2.6 H2O. The precipitate was formed through the addition of 150 μl BES
buffer [50 mM BES (N,N’-bis[2Hydroxyethyl]-2-aminoethanesulfonic acid), Sigma, 280 mM
NaCl, 1.5 mM Na2HPO4.2H2O] and incubation at room temperature for 2 or 10 min, when
opalescence appeared. The suspension was added to the cells and following a 4 hr incubation at
37oC and every day thereafter, the medium was changed to DMEM with 10% calf serum.
For Lipofectamine transfection, 8 μg β-gal DNA in the form of ethanol precipitate were
resuspended with 290 μl of serum-free DMEM and 10 μl of Plus reagent added (Life
Technologies Inc). In a separate tube, 15 μl of Lipofectamine reagent were added to 285 μl
DMEM. The two tubes were mixed, incubated for 15 min at room temperature and the mixture
added to NIH 3T3 cells at 35% or 60% confluence, growing in 6 cm petris and covered with 3
mls of serum-free DMEM, according to the manufacturer’s protocol. Cells were placed in the
incubator for 3-4 hrs, followed by a medium change to DMEM containing 10% calf serum.
For retroviral infection, NIH 3T3 cells growing at confluences of 15%, 35% or 60% in 6
cm dishes were infected with the vector MSCV (murine stem cell virus, 106 pfu/ml) which codes
for the enhanced green fluorescence protein (EGFP), in the presence of 8 μg/ml polybrene
76
(McLemore et al., 2001). Following incubation at 37oC for 4 hours, the medium was changed to
DMEM with 10% calf serum.
Western blotting
Total cell lysates were obtained three days following transfection, when cells were at
30%, 70% or 100% confluence, respectively. Proteins were extracted using 50 mM Hepes, pH
7.4, 150 mM NaCl, 10 mM EDTA, 10mM Na4P2O7, 100 mM NaF, 2 mM Na3VO4, 0.5 mM
PMSF, 10 μg/ml aprotinin, 10 μg/ml leupeptin, and 1% Triton X-100 (Raptis et al., 2000). 50 μg
of clarified cell extract were resolved on a 10% sodium dodecyl sulfate-polyacrylamide gel
electrophoresis (SDS-PAGE) gel and transferred to a nitrocellulose membrane (Bio-Rad). The
membranes were blocked with 5% nonfat milk for at least one hour followed by an overnight
incubation in primary antibody.
Immunodetection was performed using antibodies against the tyrosine-705
phosphorylated, i.e. activated form of Stat3 (Biosource International); or phosphorylated Erk1/2
(Biosource), followed by alkaline phosphatase-conjugated goat secondary antibodies (Biosource)
as described (Vultur et al., 2004). The bands were visualized using enhanced chemiluminescence
(ECL), according to the manufacturer’s instructions (PerkinElmer Life Sciences, Cat.# NEL602).
As a control for protein loading, parallel blots were probed with the mouse monoclonal anti-heat
shock protein 90 (Hsp90) antibody (Stressgen).
Luciferase Assays
vSrc-transformed NIH 3T3 cells were transfected with the pLucTKS3 reporter plasmid
which harbors seven copies of a sequence corresponding to the Stat3-specific binding site in the
C-reactive gene promoter (termed APRE, TTCCCGAA) upstream from a firefly luciferase coding
sequence (Turkson et al., 1998) with G418-resistance co-selection. Individual clones were grown
77
up and screened for firefly Luciferase activity, measured in total cell extracts according to the
manufacturer’s protocol (Promega, Cat. # E4030). As a control, pLucTKS3-expressing cells were
stably transfected with a different reporter, pRLSRE, which contains two copies of the serum
response element (SRE) of the c-fos promoter, subcloned into the Renilla luciferase reporter,
pRL-null (Promega), and Zeocin-resistance co-selection (Turkson et al., 2001). The firefly and
Renilla luciferases utilize different substrates and thus can be assayed independently in the same
lysates using this kit. Cells were subject to calcium phosphate transfection as above and Stat3
activity assessed.
In situ electroporation
The electroporation equipment was provided by Ask Science Products Inc., Kingston,
Ont. Canada. The conditions described before (Raptis et al., 2004; Raptis et al., 2003) were used
with some modifications. Briefly, cells were grown on conductive and transparent indium-tin
oxide (ITO)-coated glass slides with a cell growth area of 15x7 mm. When 35 or 60% confluent,
the oligonucleotide solution (10-30 μg/ml, in calcium-free DMEM) was added to the cells and the
electrode set placed on the slide. The optimal voltage and capacitance settings determined before
were used (Raptis et al., 2004; Raptis et al., 2000) to ensure efficient incorporation of the DNA
without damage to the cells (six pulses of 30 volts from a 2 μF capacitor).
78
Results
Calcium phosphate transfection activates Stat3
As shown in Fig. 3.1A (lanes 1, 5 and 9), cell density caused an increase in Stat3-ptyr705
levels in untreated cells, as expected based on our previous results (Vultur et al., 2004). The
addition of the calcium phosphate/DNA precipitate caused a further increase in Stat3 tyr705
phosphorylation. This increase was evident in cells that were at a confluence of 30% at the time
of harvest, as well as in 100% confluent cells, at which time Stat3 tyr705 phosphorylation
reached approximately half the levels in NIH 3T3 cells transformed by the potent Stat3 activator,
vSrc (Fig. 3.1A, lane 12 vs 13). Stat3 activation was caused by the precipitate itself, since it was
strong even in the absence of DNA (lanes 3, 7 and 11 vs 4, 8 and 12), while Stat3 levels gradually
waned to background by 6 days following transfection. When the DNA preparation was allowed
to stand for only 2 minutes at room temperature following addition of the BES buffer, the
precipitate was finer and Stat3 activation less pronounced (lanes 2, 6 and 10 vs 3, 7 and 11), while
the efficiency of gene expression was reduced (~10% vs 18%, respectively), as revealed by
staining for β-galactosidase (not shown). The addition of lower amounts of DNA resulted in a
finer precipitate and slightly reduced Stat3 levels. Similar results were obtained with mouse
10T½ or rat F111 fibroblasts, or using pUC19 plasmid DNA (not shown), indicating that the
calcium phosphate transfection procedure can have a profound effect upon Stat3 tyrosine
phosphorylation levels in cultured cells. At the same time, examination of Erk1/2 activity levels
showed no differences between transfected and control cells (Fig. 3.1B), indicating that the
calcium phosphate effect is specific for Stat3.
The vSrc oncoprotein is known to activate Stat3, as part of the process of vSrc-mediated
malignant transformation (Turkson et al., 1998). As a result, vSrc expression increases
transcription from a Stat3-specific binding element (termed APRE, TTCCCGAA, derived from
79
80
Figure 3.1
Calcium-phosphate transfection activates Stat3.
A-C: Stat3 tyr-705 phosphorylation.
NIH 3T3 cells were subject to the Calcium phosphate transfection procedure in the
presence or absence of DNA (β-gal plasmid), as indicated. The transfection mix was allowed to
stand for 2 or 10 minutes before addition to the cells, which were harvested three days later, when
they were at a confluence of 30%, 70% or 100%, as indicated. Lanes 1, 5, 9: Untransfected cells.
Lane 13: NIH 3T3 cells transformed by vSrc, harvested at 100% confluence. Detergent lysates
were resolved by gel electrophoresis and Western immunoblots were probed for, A, the tyr705
phosphorylated form of Stat3; B, the phosphorylated form of Erk1/2; and C, Hsp90, as indicated.
40 μg of lysate were loaded in all cases. Numbers at the left refer to molecular weight markers.
Arrows point to the position of Stat3 ptyr705, Erk1/2 or Hsp90, respectively. Cell confluence
was estimated by microscopic observation and quantitated by imaging analysis of live cells under
phase contrast using a Leitz Diaplan microscope and the MCID-elite software (Imaging Research,
St. Catharine’s Ont.).
D: Stat3 Transcriptional activity.
vSrc-transformed, NIH 3T3 cells, stably expressing a plasmid encoding the Stat3dependent luciferase reporter, pLucTKS3, were subjected to the Calcium phosphate transfection
procedure. Untransfected cells served as controls, as indicated. Three days later, when cells
reached a confluence of 30% (□) or 100% (▨), Luciferase activity was examined in cytosolic
extracts, as light emission. Values shown in each panel represent arbitrary luciferase units, means
plus standard deviations of at least 3 experiments, each performed in triplicate.
the C-reactive protein gene promoter), driving a luciferase reporter construct (pLucTKS3, see
Materials and Methods), when co-expressed through transfection into mouse NIH 3T3
fibroblasts. Previous results indicated that, although vSrc-transformed cells have higher Stat3
activity than the parental NIH 3T3 line when sparse, cell density causes a further increase, in an
additive manner (Vultur et al., 2004). To examine whether, in addition to tyr705 phosphorylation,
the calcium phosphate treatment can also increase Stat3 transcriptional activity, we measured
Stat3-specific transcription in vSrc-transformed NIH 3T3 cells and Firefly Luciferase activity was
measured in total cell extracts (see Materials and Methods).
As shown in Fig. 3.1D, cells treated with the calcium phosphate precipitate had ~2.5
times higher Stat3 transcriptional activity levels compared to untreated controls. As a further
control, given that cell to cell adhesion is often mediated through cadherin engagement, which is
known to require calcium, we examined whether calcium ions alone might be able to increase
81
Stat3 activity. Cells were grown in DMEM containing 1.8 mM CaCl2 (the normal concentration)
or 3, 5, 7.5, 10, 20 or 40 mM CaCl2. While up to 5 mM CaCl2 had no significant effect upon
Stat3 phosphorylation or activity levels, the highest concentration was visibly inhibitory (not
shown), indicating that the increase in Stat3 activity seen following calcium phosphate
transfection is not due to increases in free calcium concentration.
Lipofectamine transfection activates Stat3
Another commonly used transfection technique is the use of cationic lipids, such as
Lipofectamine. As shown in Fig. 3.2A, Lipofectamine treatment (see Materials and Methods)
brought about an increase in Stat3 tyrosine phosphorylation. Contrary to calcium-phosphate
transfection, this increase was slightly more pronounced in the absence of added DNA (lanes 2 vs
3 and 5 vs 6). Addition of 8 μg β-gal DNA reduced the levels almost to the background of
untreated cells (lanes 1 vs 3 and 4 vs 6). Similar results were obtained with Fugene6 (Roche)
while the efficiency of gene expression was ~20% with both treatments. The above findings
taken together indicate that treatment with cationic lipids can affect Stat3 activity.
Retroviral infection does not affect Stat3 phosphorylation
Retroviral infection is an effective way of gene transfer. To examine its potential effect
upon Stat3 activity, NIH 3T3 cells were infected with the retroviral vector MSCV which codes
for the enhanced green fluorescence protein (EGFP, see Materials and Methods). Examination of
EGFP expression by fluorescence microscopy three days later indicated that for all three sets
more than 80% of the cells had been transduced with this vector. As shown in Fig. 3.2B,
examination of Stat3 activity levels three days following infection, at which time cells were 30%,
70% or 100% confluent, revealed an increase with cell density, as observed before (Vultur et al.,
2004). However, no differences between infected and uninfected cells were noted, indicating
that, unlike transfection, retroviral infection itself does not increase Stat3 phosphorylation.
82
Figure 3.2
Effect of Lipofectamine and retroviral infection upon Stat3 activity
A: Lipofectamine transfection results in increased Stat3 phosphorylation
NIH 3T3 cells were transfected with Lipofectamine and harvested three days later, when
they reached a confluence of 70% or 100%, as indicated. Lane 7: NIH 3T3 cells transformed by
vSrc, harvested at 100% confluence. Detergent lysates were resolved by gel electrophoresis and
Western immunoblots were probed for, top panel, the tyr705 phosphorylated form of Stat3;
bottom panel, Hsp90, as indicated. 40 μg of lysate were loaded in all cases. Numbers at the left
refer to molecular weight markers. Arrow points to the position of Stat3 ptyr705 or Hsp90,
respectively.
B: Retroviral infection does not affect Stat3 tyrosine phosphorylation
NIH 3T3 cells were infected with the MSCV retroviral vector and harvested three days
later, when they reached a confluence of 30%, 70% or 100%, as indicated. Lane 7: vSrctransformed NIH 3T3 cells, harvested at 100% confluence. Detergent lysates were resolved by
gel electrophoresis and Western immunoblots were probed for, top panel, the tyr705
phosphorylated form of Stat3; bottom panel, Hsp90, as indicated. 40 μg of lysate were loaded in
all cases. Numbers at the left refer to molecular weight markers. Arrow points to the position of
Stat3 ptyr705 or Hsp90, respectively.
83
In situ electroporation does not affect Stat3 phosphorylation or transcriptional activity
Electroporation in situ has been extensively employed for the introduction of peptides
(Raptis et al., 2000; Boccaccio et al., 1998), proteins (Nakashima et al., 1999), radioactive
nucleotides (Boussiotis et al., 1997; Raptis et al., 2003; Tomai et al., 2003), oligonucleotides
(Bardelli et al., 1998; Gambarotta et al., 1996), as well as a variety of prodrugs and other
nonpermeant molecules (Marais et al., 1997; reviewed in Raptis et al., 2004). We previously
demonstrated that electroporation of a decoy double-stranded oligonucleotide, corresponding to
the hSIE, Stat3 DNA binding site (high affinity mutant, sis-inducible-element, derived from the cfos promoter 5’-GTGCATTTCCCGTAAATCTTGTCTA-3’(Turkson et al., 1998) into vSrctransformed cells which stably express the pLucTKS3 plasmid caused a potent reduction in the
levels of luciferase activity obtained 24 hrs after electroporation of the hSIE decoy, indicating
that this oligonucleotide can dramatically reduce transcription, presumably by dislodging the
activated Stat3 from its endogenous binding site on the APRE element (Raptis et al., 2003). At
the same time, an unrelated control oligonucleotide (FIRE sequence:
AGCGCCTCCCCGGCCGGGG (Turkson et al., 1998), which does not bind Stat3 did not lead to
a reduction (Raptis et al., 2003). To compare the effect of electroporation with transfection
concerning Stat3 activation, we repeated the experiment with the control oligonucleotide, or with
the tracking dye, Lucifer yellow alone in DMEM lacking calcium and compared the results with
nonelectroporated cells serving as controls.
As previously demonstrated (Raptis et al., 2004), incorporation of the tracking dye,
Lucifer yellow indicated that essentially 100% of the cells had been permeated under these
conditions, while immunostaining with antibodies against activated forms of the stress-activated
kinases (SAPK/JNK or p38hog) indicated the absence of stress to the cells (Brownell et al., 1998).
Following pulse application and a 24 hr incubation (see Materials and Methods), Stat3 tyr705
phosphorylation levels were examined. As shown in Fig. 3.3B, electroporation with or without
the control oligonucleotide did not affect Stat3 activity levels (lane 1 vs 2 and 3, or lane 4 vs 5
84
Figure 3.3
In situ electroporation does not affect Stat3 phosphorylation or transcriptional
activity
A: In situ electroporation does not affect Stat3 phosphorylation
vSrc-transformed, NIH 3T3 cells were plated on ITO-coated slides at a confluence of
35% or 60%. 30 μg/ml of the control oligonucleotide were introduced by electroporation (2 μF,
30V, 6 pulses) and 24 hours later, when 75% or 100% confluent, detergent lysates were resolved
by gel electrophoresis and Western immunoblots probed for, top panel, the tyr705 phosphorylated
form of Stat3; bottom panel, Hsp90, as indicated. Untreated cells served as controls, as indicated.
40 μg of lysate were loaded in all cases. Numbers at the left refer to molecular weight markers.
Arrow points to the position of Stat3 ptyr705 or Hsp90, respectively.
B: In situ electroporation does not affect Stat3 transcriptional activity
vSrc-transformed, NIH 3T3 cells were plated on ITO-coated slides at a confluence of
50%. Increasing amounts of the hSIE decoy oligonucleotide (■), or the same amounts of a control
85
oligonucleotide (□) as indicated were introduced by electroporation (2 μF, 30V, 6 pulses) and 24
hours later Luciferase activity was examined in cytosolic extracts, as light emission. Untreated
cells (▩), or cells electroporated with the tracking dye, Lucifer yellow (▨) served as controls, as
indicated. Values shown in each panel represent arbitrary luciferase units, means plus standard
deviations of at least 3 electroporation experiments, each performed in triplicate.
and 6). Similarly, examination of luciferase activity in cytosolic extracts showed that
electroporation did not engender a change to Stat3 activity; luciferase values obtained were
essentially the same as for parallel, nonelectroporated cultures of identical confluence, or cultures
electroporated with the tracking dye alone (Fig. 3.3C). At the same time, cells electroporated
with the decoy oligonucleotide had progressively lower luciferase activity levels, indicating that
this decoy can effectively compete with the APRE site which was already occupied by Stat3,
activated by vSrc in the transformed cells. This indicates that efficient interruption of
transcription can be achieved with this technique, under conditions that Stat3 activity is not
affected by the procedure. Similarly, electroporation of the β-gal plasmid had no effect upon
Stat3 activity (not shown).
86
Discussion
Transfection techniques utilizing calcium phosphate or cationic lipids are widely
employed for the study of Stat3 function in signal transduction. Given the importance of precise
Stat3 activity measurement, the observation that these procedures can have a profound effect
upon Stat3 activity levels, indicates that caution is required in the interpretation of results
regarding activity of Stat3 following transient transfection. Our results demonstrate that
electroporation or retroviral infection have no detectable effect, indicating that these could be
alternative approaches for Stat3 studies.
87
References
Bardelli, A., Longati, P., Gramaglia, D., Basilico, C., Tamagnone, L., Giordano, S., Ballinari, D.,
Michieli, P., and Comoglio, P. M. (1998). Uncoupling signal transducers from oncogenic MET
mutants abrogates cell transformation and inhibits invasive growth. Proc. Nat. Acad. Sci. USA
95, 14379-14383.
Boccaccio, C., Ando, M., Tamagnone, L., Bardelli, A., Michielli, P., Battistini, C., and Comoglio,
P. M. (1998). Induction of epithelial tubules by growth factor HGF depends on the STAT
pathway. Nature 391, 285-288.
Boussiotis, V. A., Freeman, G. J., Berezovskaya, A., Barber, D. L., and Nadler, L. M. (1997).
Maintenance of human T cell anergy: blocking of IL-2 gene transcription by activated Rap1.
Science 278, 124-128.
Bromberg, J. F., Horvath, C. M., Besser, D., Lathem, W. W., and Darnell, J. E., Jr. (1998). Stat3
activation is required for cellular transformation by v-src. Mol. Cell. Biol. 18, 2553-2558.
Brownell, H. L., Lydon, N., Schaefer, E., Roberts, T. M., and Raptis, L. (1998). Inhibition of
Epidermal Growth Factor-mediated ERK1/2 activation by in situ electroporation of nonpermeant
[(alkylamino)methyl]acrylophenone derivatives. DNA & Cell Biol. 17, 265-274.
Chen, C. and Okayama, H. (1987). High-Efficiency Transformation of Mammalian Cells by
Plasmid DNA. Mol. Cell. Biol. 7, 2745-2752.
Coffer, P. J., Koenderman, L., and de Groot, R. P. (2000). The role of STATs in myeloid
differentiation and leukemia. Oncogene 19, 2511-2522.
Gambarotta, G., Boccaccio, C., Giordano, S., Ando, M., Stella, M. C., and Comoglio, P. M.
(1996). Ets up-regulates met transcription. Oncogene 13, 1911-1917.
Graham, F. L. and van der Eb, A. J. (1973). A new technique for the assay of infectivity of human
Adenovirus 5 DNA. Virology 52, 456-467.
Marais, R., Spooner, R. A., Stribbling, S. M., Light, Y., Martin, J., and Springer, C. J. (1997). A
cell surface tethered enzyme improves efficiency in gene-directed enzyme prodrug therapy. Nat.
Biotechnol. 15, 1373-1377.
McLemore, M. L., Grewal, S., Liu, F., Archambault, A., Poursine-Laurent, J., Haug, J., and Link,
D. C. (2001). STAT-3 activation is required for normal G-CSF-dependent proliferation and
granulocytic differentiation. Immunity. 14, 193-204.
Nakashima, N., Rose, D., Xiao, S., Egawa, K., Martin, S., Haruta, T., Saltiel, A. R., and Olefsky,
J. M. (1999). The functional role of crk II in actin cytoskeleton organization and mitogenesis. J.
Biol. Chem. 274, 3001-3008.
Niu, G., Bowman, T., Huang, M., Shivers, S., Reintgen, D., Daud, A., Chang, A., Kraker, A.,
Jove, R., and Yu, H. (2002). Roles of activated Src and Stat3 signaling in melanoma tumor cell
growth. Oncogene 21, 7001-7010.
88
Raptis, L., Vultur, A., Brownell, H. L., and Firth, K. L. (2004). Dissecting pathways; in situ
electroporation for the study of signal transduction and gap junctional communication. Cell
Biology A laboratory Handbook.
Raptis, L., Brownell, H. L., Vultur, A. M., Ross, G., Tremblay, E., and Elliott, B. E. (2000).
Specific inhibition of Growth Factor-stimulated ERK1/2 activation in intact cells by
electroporation of a Grb2-SH2 binding peptide. Cell Growth Differ. 11, 293-303.
Raptis, L., Vultur, A., Balboa, V., Hsu, T., Turkson, J., Jove, R., and Firth, K. L. (2003). In situ
electroporation of large numbers of cells using minimal volumes of material. Analytical
Biochemistry 317, 124-128.
Song, J. I. and Grandis, J. R. (2000). STAT signalling in head and neck cancer. Oncogene 19,
2489-2495.
Tomai, E., Vultur, A., Balboa, V., Hsu, T., Brownell, H. L., Firth, K. L., and Raptis, L. (2003). In
situ electroporation of radioactive compounds into adherent cells. DNA and Cell Biology 22,
339-346.
Turkson, J., Bowman, T., Adnane, J., Zhang, Y., Djeu, J. Y., Sekharam, M., Frank, D. A.,
Holzman, L. B., Wu, J., Sebti, S., and Jove, R. (1999). Requirement for Ras/Rac1-mediated p38
and c-Jun N-terminal kinase signaling in Stat3 transcriptional activity induced by the Src
oncoprotein. Mol. Cell. Biol. 19, 7519-7528.
Turkson, J., Bowman, T., Garcia, R., Caldenhoven, E., de Groot, R. P., and Jove, R. (1998). Stat3
activation by Src induces specific gene regulation and is required for cell transformation. Mol.
Cell. Biol. 18, 2545-2552.
Turkson, J., Ryan, D., Kim, J. S., Zhang, Y., Chen, Z., Haura, E., Laudano, A., Sebti, S.,
Hamilton, A. D., and Jove, R. (2001). Phosphotyrosyl peptides block Stat3-mediated DNA
binding activity, gene regulation, and cell transformation. J. Biol. Chem. 276, 45443-45455.
Vultur, A., Cao, J., Arulanandam, R., Turkson, J., Jove, R., Greer, P., Craig, A., Elliott, B. E., and
Raptis, L. (2004). Cell to cell adhesion modulates Stat3 activity in normal and breast carcinoma
cells. Oncogene 23, 2600-2616.
Wang, Y. Z., Wharton, W., Garcia, R., Kraker, A., Jove, R., and Pledger, W. J. (2000). Activation
of Stat3 preassembled with platelet-derived growth factor beta receptors requires Src kinase
activity. Oncogene 19, 2075-2085.
Yu, H. and Jove, R. (2004). The STATs of cancer--new molecular targets come of age. Nat. Rev.
Cancer 4, 97-105.
89
Chapter 4
Cadherin engagement promotes cell survival
via Rac/Cdc42 and Stat3
These data are included in the following manuscripts:
Arulanandam, R., Vultur, A., Cao, J., Carefoot, E., Truesdell, P.F., Elliott, B.E., Larue, L.,
Feracci, H. and Raptis, L. (2009). Cadherin engagement promotes survival via Rac/Cdc42 and
Stat3. Mol. Cancer Res. 7, 1310-27.
Arulanandam, R., Geletu, M., Chevalier, S., Feracci, H. and Raptis, L. Cadherin-11 engagement
activates Stat3 and promotes survival of mouse fibroblasts. In preparation.
Author contributions:
RA performed the bulk of the work in this chapter. JC helped with the RT-PCR experiments. AV
produced preliminary data and discussed results. MG performed the siCad11 experiments. As
collaborators, LL from the Institut Curie, Paris, France, provided the ES cell lines and HF and SC
from the Universite Bordeaux 1, Centre de Recherche Paul Pascal, Pessac, France, generously
provided the recombinant cadherin fragment constructs. As collaborators from the Cancer
Research Institute at Queen’s, EC and BEE performed the tissue immunostaining and PFT
generously provided primary epithelial cell lysates. All authors contributed insightful advice and
helped revise the manuscripts.
90
Abstract
Stat3 (signal transducer and activator of transcription-3) is activated by a number of
receptor and non-receptor tyrosine kinases, while a constitutively active form of Stat3 alone is
sufficient to induce neoplastic transformation. In the present report we demonstrate that Stat3 can
also be activated through homophilic interactions by the epithelial (E)-cadherin and cadherin-11,
two members of the classical type I and II cadherin family of surface receptors, responsible for
the formation of cell to cell junctions. Indeed, by plating cells onto surfaces coated with
fragments encompassing the two outermost domains of these cadherins, we clearly demonstrate
that cadherin engagement can activate Stat3, even in the absence of direct cell to cell contact.
Most importantly, our results also reveal for the first time a dramatic surge in total Rac1 and
Cdc42 protein levels triggered by cadherin engagement, and an increase in Rac1 and Cdc42
activity, which is responsible for the Stat3 stimulation observed. Inhibition of cadherin
interactions using a peptide, a soluble cadherin fragment or genetic ablation induced apoptosis,
pointing to a significant role of this pathway in cell survival signalling, a finding which could also
have important therapeutic implications.
91
Introduction
Stat3 is found to be overexpressed in a number of carcinomas and tumor cell lines
[reviewed in (Yu and Jove, 2004; Germain and Frank, 2007)]. The fact that a constitutively
active form of Stat3 alone is sufficient to induce transformation points to an etiological role for
Stat3 in neoplasia (Bromberg et al., 1999). Stat3 can be stimulated by cytokines and receptor
tyrosine kinases, as well as the non-receptor tyrosine kinase Src (Yu et al., 1995). Stat3 is
invariably latent in the cytoplasm and, subsequent to binding to an activated receptor through its
SH2 domain, becomes activated through phosphorylation by the receptor itself or by the
associated JAKs or Src family kinases. Phosphorylation at the critical tyrosine-705 activates
Stat3 by stabilizing the association of two monomers through reciprocal SH2-phosphotyrosine
interactions. The Stat3 dimer then migrates to the nucleus where it binds to target sequences,
leading to the transcriptional activation of specific genes, such as myc, bcl-xL, cyclin D, survivin,
hgf and others (Yu and Jove, 2004).
In addition to providing structure and integrity to the cell, adhesion molecules can play a
key role in the initiation of signalling through multiple intracellular pathways which involve
protein phosphorylation. The formation of cell to cell adhesion junctions is primarily modulated
by the calcium-dependent family of cadherin receptors, and leads to a reorganization of the
underlying actin cytoskeleton. Classical cadherins contain an extracellular domain organized into
five distinct cadherin-specific repeats (EC) which display calcium ion binding, a membrane
spanning region and a highly conserved cytoplasmic domain, which enables association with the
cytoskeletal network through catenin interactions. Classical cadherins are subdivided into type I
and type II (Halbleib and Nelson, 2006); type I cadherins have a conserved HAV motif in the
most distal EC and include E-cadherin, P-cadherin and N-cadherin, which are found in most
tissues and are crucial to the developmental process, while type II cadherins, such as cadherin-11,
92
lack this motif in their hydrophobic acceptor pocket. E-cadherin represents the best characterized
classical cadherin; it is responsible for the formation and maintenance of epithelial structures and
is abundant in cultured cells of epithelial origin (Jeanes et al., 2008). Although it is generally
thought that cadherin expression results in a tight cell association, mesenchymal cells, which are
loosely associated, express mesenchyme-specific cadherins such as cadherin-11. Cadherin-11
was originally identified in mouse osteoblasts (Okazaki et al., 1994), but it was later found to be
expressed in a variety of normal tissues of mesodermal origin, such as areas of the kidney and
brain (Hoffmann and Balling, 1995), as well as in cultured fibroblasts (Orlandini and Oliviero,
2001), and recent results showed that cadherin-11 promotes the metastasis of prostate cancer cells
to the bone (Chu et al., 2008).
One of the signalling targets of cadherin engagement is the Rho family of proteins,
which consists of a number of small GTPases acting as intracellular molecular switches cycling
between the active, GTP-bound state and the inactive, GDP-bound form (Etienne-Manneville and
Hall, 2002). The Rho proteins, Rac1, Cdc42 and RhoA are best known as master regulators of
the actin cytoskeleton and promote the formation of lamellipodia, filopodia and stress fibers,
respectively. Previous work has indicated that following ligation, the juxtamembrane domain of
cadherins interacts with p120 catenin, which can activate Rac1 and Cdc42, possibly by binding to
Vav2, an exchange factor for these GTPases, although the precise mechanism is not clearly
established (Jaffer and Chernoff, 2004).
We and others recently demonstrated a dramatic increase in the activity of Stat3 triggered
by cell confluence in a variety of cell lines (Vultur et al., 2004; Onishi et al., 2008; Steinman et
al., 2003; Su et al., 2007; Kreis et al., 2007; Vultur et al., 2005). Given the generally accepted,
positive role of Stat3 in proliferation, the Stat3 activity increase observed in post-confluent cells,
at a time when cells do not divide, was a surprising observation. In this communication we
definitively demonstrate that E-cadherin or cadherin-11 engagement alone activates Stat3, even in
the absence of direct cell to cell contact. This activation occurs through an increase in the activity
93
of Rac1 and Cdc42. Unexpectedly, we also found that cadherin engagement induces a dramatic
increase in total Rac1 and Cdc42 protein levels, which points towards a novel mechanism of Rac1
and Cdc42 upregulation. We demonstrate that this event causes a further increase in Rac1, Cdc42
and Stat3 activity, and constitutes a potent survival signal.
94
Materials and Methods
Cell lines and culture techniques
The normal mouse mammary epithelial line HC11 is a prolactin-responsive cell clone
originally isolated from the COMMA-1D mouse mammary epithelial cell line derived from a
female Balb/c mouse in mid-gestation (Ball et al., 1988). HC11 cells transformed by an activated
Y527F-Src have been described previously (Wojcik et al., 2006). Normal mouse Balb/c 3T3,
Balb/c 3T3 transformed by vSrc and mouse 10T1/2 fibroblasts have been previously described
(Vultur et al., 2004). ES cell cultures were grown as before (Larue et al., 1996). Cell confluence
was estimated visually and quantitated by imaging analysis of live cells under phase contrast
using a Leitz Diaplan microscope and the MCID-elite software (Imaging Research, St.
Catharine’s Ont.). For the isolation of mouse breast epithelial cells, mammary glands at day 1214 of pregnancy were aseptically removed and transferred to a culture dish. The digestion of
tissue and epithelial cell growth was adapted from (Medina and Kittrell, 2000). These cells grow
in islands, so that it is difficult to obtain sparse cultures.
JAK-inhibitor-1 (EMD Biosciences, Gibbstown, NJ) and Toxin B (Sigma, St. Louis,
MO) were added at the indicated concentrations. Cell viability was assessed by trypan blue
exclusion and by replating cells in medium lacking the inhibitors. IL6 was purchased from R&D
Systems (Minneapolis, MN).
Conditioned medium was prepared by adding 3 mls of serum-free RPMI to a 10 cm plate
of HC11 cells grown to 3 days post-confluence.
Cadherin recombinant fragments and peptides
The plasmid constructs coding for the first two extracellular domains of E-cadherin
(E/EC12) or cadherin-11 (11/EC12), fused with a C-terminal hexahistidine tag were previously
95
described (Perret et al., 2002). Cadherin fragment expression was induced by the addition of
IPTG for 2h. Cell pellets were resuspended in lysis buffer: 4M urea, 50mM Na2HPO4 pH 7.8,
20mM Imidazole and 20mM β-mercaptoethanol. Purification was carried out by Ni-affinity
chromatography as previously described (Perret et al., 2002). Purified preparations of each
cadherin fragment yielded only one band upon analysis by 15% SDS-PAGE and Coomassie blue
staining. Purified fragments were attached to plastic petris as follows: petris were treated with
poly-DL-lysine (Sigma, 0.1 mg/ml in H2O, for 1 hour at room temperature), followed by
glutaraldehyde treatment (Sigma, 2.5% in PBS, for 1 hour at room temperature). The cadherin
fragments were subsequently added at the indicated concentrations for 2 hours, followed by
glycine treatment (0.2 M in PBS, overnight). Petris were incubated with serum-free medium for
one hour, prior to the addition of cells in complete medium. Growth of cells on the E/EC12 or
cadherin-11 fragments did not cause any detectable change in morphology. Peptides were
synthesized by CPC (San Jose, CA) and used at the indicated concentrations.
Gene expression
Rac1N17 was expressed with a myc-tag at the carboxy-end by plasmid transfection.
Expression was verified by blotting against the myc-tag using the 9E10 antibody (Sigma). Rac1
shRNA was purchased from Open Biosystems (Huntsville, AL, Cat# RMN 1766-97047533), in
the retroviral vector pSM2c. For gp130 knockdown, a mouse pSM2 retroviral target gene shRNA
set (Cat#. RMM4530-NM_010560) was purchased from Open Biosystems. A cadherin-11
shRNA set was also purchased from Open Biosystems (Cat#. RMM4530-NM_009866). Cells
infected with the pSM2 constructs were selected for puromycin resistance. For the complete
shRNA sequences see Appendix, Table A.4.
96
Lentivirus vector production
293T cells were plated at a density of 4×106 cells per 10-cm culture dish. The cells were
co-transfected by calcium phosphate co-precipitation with 15 μg of pLKO1-Stat3 shRNA
(Mission shRNA TRCN0000020842 and TRCN0000020840 ) or pLKO1-control shRNA (Sigma)
and 10 μg of pPACK packaging plasmid mix (SBI, Mountain View, CA). The culture medium
was replaced with fresh medium after 6 h. The supernatant was collected 16 h after the
transfection and stored at −80°C. To determine the vector titers, 105 HT1080 cells were seeded in
a six-well plate and infected with various dilutions of the vector in the presence of 4 μg/ml
polybrene. The culture medium was replaced 48 h later with fresh medium containing puromycin
at a concentration of 1.5 μg/ml. Puromycin-resistant colonies were counted 10 days after
transduction. To examine the efficiency of Stat3 downregulation, cells were infected at 5 plaque
forming units (pfu)/cell and extracts probed for total Stat3 (Cell Signalling, Danvers, MA).
Immunofluorescence and apoptosis assays
Immunofluorescence staining for E-cadherin was performed using a mAb (BD
Transduction Labs, Mississauga, ON), biotinylated anti-mouse IgG secondary and Rhodamineavidin D (Vector Labs, Burlington, Ontario). Fixed cells were assayed for apoptosis by terminal
deoxynucleotidyl transferase dUTP nick end labeling (TUNEL, Roche, Mississauga, ON) and
fluorescence activated cell sorting (FACS) analysis as described (Vultur et al., 2004). Images
were captured using a 20x objective (NA 0.45) on a Nikon TE200 epifluorescence microscope
equipped with a cooled CCD camera (Roper Scientific, Cool Snap HQ) and Metamorph
Software. Fluorescent images were also analyzed on a laser scanning confocal microscope (Leica
SP2) with Tsunami 2 photon IR laser using a 100x oil immersion lens (NA 1.4) and LCS
software for initial acquisition, as well as Image Pro Plus 6.0 software for further analysis.
97
Immunohistochemical staining for E-cadherin and Stat3-ptyr705
For immunohistochemical staining, consecutive tissue sections (6 µm) were cut from a
formalin-fixed, paraffin-embedded tissue block of the #4 mammary fatpad from a nulliparous (812 wk old) strain 129 mouse bred in the Queen’s University Animal Care Facility. Sections were
first deparaffinized in toluene for 4 min, three times and then rehydrated in a graded series of
100%, 85% and 70% ethanol, prior to being submerged in a water bath (10 oC) for 4 min.
Antigen retrieval used a 1:10 dilution of 10 mM sodium citrate buffer, pH 6.0 at 95°C for 30 min.
Tissues were then cooled to room temperature before washing twice with TBS (tris-buffered
saline). Excess buffer was removed and 3% hydrogen peroxide was applied for 5 min. The
tissue sections were then permeabilized in 25 μg/ml Digitonin in TBS, and blocked with 5%
horse serum in TBS for 30 min. Excess blocking medium was removed and a PAP pen was used
to concentrate antibodies over the tissue. Monoclonal anti-E-cadherin antibody (BD
Transduction Labs), mouse IgG2α control (Invitrogen Inc., Burlington, ON), or rabbit antiStat3ptyr705 antibody (Cell Signalling NEB, Pickering ON) were applied at a predetermined
dilution (200 μl) to each tissue section. For Stat3-ptyr705 staining, sections underwent an
additional trypsinization step after antigen-retrieval in which sections were incubated in 0.01N
hydrochloric acid (HCl) and 0.4% Trypsin for 20 min at 37°C, prior to application of rabbit antiStat3ptyr705 antibody. Sections were incubated with antibodies overnight at 4°C in a humidified
chamber. The next day, the biotinylated link (Dako North America Inc.) was applied and
incubated for 15 min. Streptavidin-peroxidase (Dako North America Inc.) was then applied to the
specimens and incubated for another 15 min, according to the manufacturer’s instructions.
Finally, 200 µL of DAB (diaminobenzidine) with DAB chromogen substrate (1 drop/mL) (Dako
North America Inc.) was added to each section and incubated for 10 min. Tissues were
counterstained lightly with hematoxylin, dipped three times in 1% HCl in 70% ethanol and placed
in a water bath. The sections were then dehydrated in a graded ethanol series (70%, 85%, 100%)
98
followed by three toluene baths, and coverslipped using permount. Representative images were
acquired under brightfield illumination using a Leitz microscope.
Western blotting and immunoprecipitation
Detergent cell extracts were prepared as described (Vultur et al., 2004). Following a
careful protein determination (BCA-1 Protein assay kit, Sigma), 30 or 100 μg of clarified cell
extract were loaded, as indicated. Blots were cut into strips and probed with antibodies specific
for the tyr-705 phosphorylated Stat3 or total Stat3 protein (Cell Signalling NEB), Rac1 or Cdc42
(BD Transduction Labs), the dually phosphorylated form of Erk1/2 (Biosource, Burlington, ON),
survivin (Cell Signalling NEB), p21 (Biosource), poly (ADP-ribose) polymerase (PARP, Roche),
E-cadherin (BD Transduction Labs), gp130 (Sigma) or Hsp90 (Stressgen, Ann Arbor, Michigan)
followed by alkaline phosphatase, or HRP-conjugated secondary antibodies (Jackson Labs, Bar
Harbor, Maine). Bands were visualized using enhanced chemiluminescence (ECL, PerkinElmer
Life Sciences, Woodbridge, Ontario), or SuperSignal West Femto Maximum Sensitivity
Substrate (Pierce, Rockford, IL). Quantitation was achieved by fluorimager analysis using the
FluorChem program (AlphaInnotech Corp). In all cases Stat3-ptyr705 band intensities were
normalized to Hsp90 levels of the same samples. Jak1 phosphorylation was examined by
immunoprecipitation against total Jak1, followed by blotting with a Jak-ptyr1022/1023 antibody
(Cell Signalling NEB). Jak kinase assays were performed as described (Vultur et al., 2004).
Rac1/Cdc42 activation assays were performed using GST-PAK-PBD (glutathione-Stransferase – p21-activated kinase binding domain) pull-down assays with the Rac1/Cdc42
activation assay kit (Cytoskeleton, Denver, Colorado, #BK035). Photoshop (Adobe) or Corel
Draw software were used for the organization of non-adjusted, original images and blots.
99
Reverse transcriptase-polymerase chain reaction (RT-PCR) assays
RT-PCR was performed using Superscript III First-Strand Synthesis kit (Invitrogen) in an
Eppendorf Personal Mastercycler. cDNAs were synthesized using gene-specific primers and the
Superscript III First-Strand Synthesis System kit for RT-PCR (Invitrogen, Cat. #180080-051).
PCR was performed in 20μl on an Eppendorf Mastercycler Personal thermo cycler (Mississauga,
ON). For the detection of cadherin levels, cDNA was denatured for 2 minutes at 94oC, then the
reaction conducted for 30 cycles (94oC for 5 seconds, 55oC for 5 seconds, 72oC for 5 seconds)
and the reaction completed at 72oC for 2 minutes. Primer pairs for amplification were, for Ecadherin: forward primer 5’CAT CGC CAC AGA TGA TGG TT and the reverse primer 5’ACC
TGC ATG TTT CGA GGT TCT to generate a 111bp fragment; for cadherin-11: forward primer
5’ACC CCC TGA AAT CAT GCA CAA and the reverse primer 5’ACC GGC ATT CAG GAT
GTA G to generate a 251bp fragment.
For the detection of Rac1 and Cdc42 levels, total cDNA from cells grown to different
densities was denatured for 2 minutes at 94oC, then the reaction conducted for 30 cycles (94oC for
15 seconds, 55oC for 30 seconds, 72oC for 1 minute) and the reaction completed at 72oC for 6
minutes. The primer pairs for Rac1 amplification were: forward primer 5’GGA CAC AGC TGG
ACA AGA AGA and the reverse primer 5’GGA CAG AGA ACC GCT CGG ATA to generate a
368 bp fragment. For Cdc42 the primer pairs were: forward primer 5’CGA CCG CTA AGT TAT
CCA CAG and the reverse primer 5’GCA GCT AGG ATA GCC TCA TCA to generate a 325 bp
fragment. 18S RNA (153 bp; forward primer 5’AAA CGG CTA CCA CAT CCA AG and
reverse primer 5’CCT CCA ATG GAT CCT CGT TA) was used as a control (Sigma).
All quantitative RT-PCR reactions were performed with 1x SYBR Green Master Mix
(BioRad, Philadelphia, PA) using the Corbett Rotor-Gene 6000. Serial ten-fold dilutions of 18S
RNA were used as a reference for the standard curve calculation. The delta ct (Δct) value was
calculated from the given ct value by the formula: Δct = (ctsample – ctcontrol). The fold change was
calculated as the value of 1.94-Δct, 1.94 being the average PCR efficiency. For the qRT-PCR
100
cytokine array, we used the PAMM-021A kit (SA Biosciences, Frederick, MD) with an RT-PCR
for IL6 run in parallel, according to the manufacturer’s protocol.
101
Results
E-cadherin engagement can activate Stat3
To explore the nature of the molecule(s) that might be involved in the cell densitymediated Stat3 activation, we first used the normal mouse breast epithelial line HC11, which
expresses high amounts of E-cadherin, an extensively studied, calcium-dependent, cell to cell
adhesion molecule and member of the classical type I cadherin family [(Merlo et al., 1995);
Appendix, Fig. A.1A]. HC11 cells were plated in plastic Petri dishes, and when ~30-80%
confluent, and over several days thereafter, detergent cell extracts were probed by western
blotting with an antibody specific for the tyrosine-705 phosphorylated form of Stat3 (Stat3ptyr705, see Materials and Methods). To reduce the variability that might be caused by nutrient
depletion in post-confluent cultures, the medium was changed every 24 hours. As a loading
control, the same extracts were probed for the abundant heat shock protein, Hsp90. As shown in
Fig. 4.1, Stat3-ptyr705 levels were almost undetectable in sparsely growing HC11 cells (A, panel
a, and B, lane 1), while density caused a dramatic increase (>50-fold), which peaked at 2-3 days
after confluence (A, panel d, and B, lanes 1 vs 6) and eventually decreased slightly at later times.
The Stat3-ptyr705 observed in densely growing, normal HC11 cells was nuclear (Appendix, Fig.
A.1C) and reached approximately half the levels present in cells transformed by the potent Stat3
activator, Src-527F (Fig. 4.1B, lane 9). Examination of the levels of total Stat3 protein revealed a
modest increase with cell density (~2 fold, Fig. 4.1B), possibly due to the fact that the Stat3
promotor itself is one of the Stat3 targets (Narimatsu et al., 2001), although the differences were
not as pronounced as the Stat3-tyr705 phosphorylation observed (Appendix Table A.1). This
activation was found to be specific to Stat3, since the levels of the extracellular signal regulated
kinase (Erk1/2), a signal transducer often coordinately activated with Stat3 by a number of
growth factors and oncogenes, remained unaffected by cell confluence (Fig. 4.1B). In addition,
102
Figure 4.1
Cell density triggers a dramatic activation of Stat3 in HC11 cells.
A: HC11 cells grown to the indicated confluences were photographed under phase-contrast
illumination. Bar represents 100 μm.
B: Cell density upregulates Stat3-ptyr705 levels.
Lysates from normal mouse breast epithelial HC11 cells grown to increasing densities
were resolved by gel electrophoresis. The blot was cut into strips which were probed for Stat3ptyr705, phospho-Erk1/2, total Stat3 or Hsp90 as a loading control, as indicated. Numbers under
the lanes of the top panels refer to Stat3-ptyr705 and total Stat3 band intensities obtained through
quantitation by fluorimager analysis and normalized to Hsp90 levels, with the peak value of the
control, HC11 cells transformed by src-527F (lane 9) taken as 100% (see Materials and Methods
and Appendix, table A.1). Numbers at the left refer to molecular weight markers.
103
cell density increased Stat3 transcription in HC11 cells (Appendix, Fig. A.1D). The above results
taken together indicate that cell density causes a specific increase in Stat3-tyr705 phosphorylation
and activity in HC11 cells.
To examine whether E-cadherin engagement is directly required for the cell to cell
adhesion-mediated Stat3 activation, cadherin adhesive interactions were interrupted through the
introduction of a recombinant E-cadherin fragment, encompassing the two distal, extracellular
domains of E-cadherin (E/EC12), which was previously shown to retain biological activity
(Perret et al., 2002), into the medium of densely growing HC11 cells. In solution, this fragment is
expected to compete with the cadherin molecules interacting from opposite cells and dissociate
the junctions. As shown in Fig. 4.2A, addition of 40 μM, soluble E/EC12 to the growth medium
caused a ~6-fold decrease in Stat3-ptyr705 levels (lane 2 vs 1), while the E/W2A mutant
fragment, which is largely defective in cadherin engagement (Perret et al., 2002), had a weaker
effect (<2-fold, lanes 3 and 4), strongly suggesting that cadherin engagement is required for the
Stat3-ptyr705 increase. To further substantiate this conclusion, we used the peptide Ac-S-HAVSA-NH2 (SHAVSA), which has been shown to inhibit E-cadherin homophilic interactions in
epithelial cells (Makagiansar et al., 2001). As shown in Fig. 4.2A (lanes 6-9), addition of this
peptide to the growth medium resulted in a substantial decrease in Stat3-tyr705 phosphorylation
(up to 20-fold, lane 9 vs 6). In contrast, the peptide Ac-S-HAV-SS-NH2 (SHAVSS), previously
shown to be less effective at inhibiting cadherin engagement had a less pronounced effect (lane
12), while the control, inactive peptide Ac-S-HGV-SA-NH2 (SHGVSA) had no detectable effect
(lane 10). On the other hand, addition of the SHAVSA peptide even at 10 mM to HC11 cells
expressing the constitutively active form of Stat3, Stat3C (McLemore et al., 2001), was unable to
reduce Stat3-ptyr705 levels substantially (Fig. 4.2C). The reduction in Stat3-tyr705
phosphorylation following treatment with the SHAVSA peptide or soluble E/EC12 fragment was
accompanied by a specific decrease in Stat3 transcriptional activity as well (Fig. 4.2B). Similar
results were obtained with the MDCK, canine kidney epithelial line (not shown). These data
104
Figure 4.2
Disruption of cadherin engagement by soluble cadherin fragments or synthetic
peptides reduces Stat3-tyr705 phosphorylation and activity
A: Reduction of Stat3, ptyr705 phosphorylation.
Left: HC11 cells were grown to 1 day post-confluence (lanes 1-4) or 40% confluence (lane 5)
and the E/EC12, E-cadherin fragment encompassing the two outermost extracellular domains
(lane 2), or the corresponding fragment from the E/W2A mutant (lanes 3 and 4) were added to the
medium at concentrations of 40 μM for 48 hours, redosing at 24 hours. Detergent cell extracts
were resolved by gel electrophoresis and probed for Stat3-ptyr705 or Hsp90 as a loading control,
as indicated. Lanes 1 and 5: Control, untreated cells.
Right: HC11 cells grown to 1 day post-confluence were treated with increasing concentrations of
the SHAVSA peptide (lanes 7-9), the inactive SHGVSA (10 mM, lane 10) or the weakly active
SHAVSS (10 mM, lane 12) peptides for 48 hours, as indicated, and Stat3-ptyr705 and Hsp90
levels examined as above. Lanes 6 and 11: Control, untreated cells grown to 1 day postconfluence. Lane 13: Untreated HC11 cells grown to 40% confluence.
Numbers under the lanes of the top panels refer to Stat3-ptyr705 band intensities obtained
through quantitation by fluorimager analysis and normalized to Hsp90 levels, with the peak value
of the control, untreated cells for each blot (lanes 1, 6 or 11) taken as 100% (see Materials and
Methods).
105
B: Reduction of Stat3 transcriptional activity.
HC11 cells, transfected with the Stat3-dependent pLucTKS3 and the Stat3-independent
pRLSRE reporters, were grown to the indicated densities and postconfluent cells treated with the
E/EC12 or E/W2A cadherin fragments, the E-cadherin-disrupting SHAVSA, or the control
SHGVSA peptides, as indicated. Firefly (■) or Renilla ( ) luciferase activities were determined
in detergent cell extracts (Vultur et al., 2004). Values shown represent luciferase units expressed
as a percentage of the highest value obtained, means ± s.e.m. of at least 3 experiments, each
performed in triplicate.
C: Stat3C expression rescues Stat3-ptyr705 levels following inhibition of cadherin
engagement.
Stat3C was stably expressed in HC11 cells, which were grown to 1 day post-confluence.
Following addition of 5 mM or 10 mM of the cadherin-disrupting, SHAVSA peptide, detergent
cell extracts were probed as above for Stat3-ptyr705 or Hsp90 as a loading control, as indicated.
Numbers under the lanes refer to Stat3-ptyr705 band intensities obtained through
quantitation by fluorimager analysis and normalized to Hsp90, with levels in the control,
untreated HC11-Stat3C cells taken as 100%.
further reinforce the conclusion that E-cadherin engagement is required for the cell to cell
adhesion-mediated, Stat3-tyr705 phosphorylation, in both mammalian species.
To examine the generality of the E-cadherin requirement for the density-mediated, Stat3
activation, we employed an E-cadherin-defective, mouse embryonic stem (ES) cell line (null),
and the parental ES line (Ecad+/+) as a control (Larue et al., 1996). As expected, density caused
a 10-fold increase in Stat3-ptyr705 phosphorylation in the wild-type (Ecad+/+) cells, which
peaked at 2-3 days after confluence (Fig. 4.3A, top, lanes 1 vs 6). In stark contrast however,
Stat3-ptyr705 levels were approximately 6-fold lower in null cells, at all densities examined (Fig.
4.3B vs A, top). The low, background Stat3-ptyr705 levels present in null cells (Fig. 4.3B) could
conceivably be due to the LIF, a known Stat3 activator, present in the growth medium of ES cells
(Appendix Figs. A.2, and A.3). However, the possibility of other cadherins, or alternative
pathways of Stat3 activation being present, cannot be excluded. In any event, the dramatic Stat3ptyr705 increase with cell confluence (Fig. 4.3A, lane 1 vs 6), and decrease with cadherin
ablation (Fig. 4.3B vs A and Appendix Fig. A.4A), indicates that the E-cadherin-induced, Stat3
phosphorylation is much stronger than the activation due to LIF. The decrease in Stat3-705
106
phosphorylation in null cells was accompanied by a similar decrease in Stat3 DNA binding and
transcriptional activity (Appendix Fig. A.4C and D). The above data taken together indicate that
E-cadherin ablation in ES cells can have a striking and specific effect upon the density-induced
increase in Stat3-tyr705 phosphorylation and activity, and further suggest that E-cadherin is
required for this activation.
The HC11 line was originally derived from normal mouse breast epithelium (Danielson
et al., 1984; Kimball et al., 1984). Therefore, to assess the relevance of the cadherin-Stat3
interactions observed at high cell densities to the in vivo microenvironment, the expression
patterns of E-cadherin and Stat3-ptyr705 were examined in normal mouse breast tissues by
immunohistochemistry staining. As shown in Fig. 4.4A, there is intense E-cadherin staining at
the plasma membrane, while Stat3-ptyr705 staining is concentrated mostly in the nuclei of
luminal breast epithelial cells. In contrast, there is little expression of either protein in the
surrounding adipose tissue. These findings demonstrate the presence of constitutively activated
Stat3 in the corresponding normal breast tissue where E-cadherin is engaged, thus revealing a
distinct correlation in the state of these same molecular markers observed in HC11 cells growing
at high, but not low densities in culture. This conclusion was reinforced by examination of Stat3ptyr705 levels in cultured primary mouse breast epithelial cells, which display similar levels of
Stat3-ptyr705 as confluent HC11 cells (Fig. 4.4B).
E-cadherin engagement is sufficient to activate Stat3
The above results demonstrate that E-cadherin is required for the cell to cell adhesionmediated Stat3 activation in normal cells. The question still remaining is whether Stat3 activation
is a direct consequence of E-cadherin engagement or whether E-cadherin interactions are simply
required to bring cell surfaces into proximity, to initiate signals which are not direct effects of
107
Figure 4.3
E-cadherin ablation inhibits the density-mediated, Stat3 activation.
Ecad+/+ (A) or null (B) mouse ES cells were grown to increasing densities, up to 8 days
post-confluence. Detergent cell lysates were resolved by gel electrophoresis and probed for
Stat3-ptyr705 (top), total Stat3 (middle) or Hsp90 as a loading control (bottom), as indicated.
108
Figure 4.4
Expression of Stat3-ptyr705 and E-cadherin in normal mouse breast ductal
epithelium
A. Serial tissue sections from a formalin-fixed, paraffin embedded mouse mammary (#4) fatpad
from a nulliparous 8 wk old mouse were subjected to immunohistochemical staining for Stat3ptyr705 (top) or E-cadherin (bottom), as described in Materials and Methods. Representative
images taken with a 40x or 100x objective are shown. Scale bar = 100 µm.
B. Lysates from primary mouse breast epithelial cells (lane 3) and normal mouse breast epithelial
HC11 cells grown to 40% (lane 1) or 100% (lane 2) confluence were resolved by gel
electrophoresis. The blot was cut into strips which were probed for Stat3-ptyr705, E-cadherin or
Hsp90 as a loading control, as indicated. Numbers at the left refer to molecular weight markers.
109
cadherin ligation. To definitively demonstrate whether the Stat3 activation observed at high cell
densities is a direct effect of cadherin engagement per se, the E/EC12 fragment was used to
functionalize petri dishes by covalent immobilization (Fig. 4.5A). This fragment has been shown
to retain biological activity when immobilized on surfaces and is recognized specifically by Ecadherin-expressing cells (Perret et al., 2002).
Plastic petri dishes were functionalized with increasing amounts of purified E/EC12 or
E/W2A fragments and HC11 cells plated on these surfaces (see Materials and Methods). No
change in the morphology of HC11 cells upon growth on these fragments was noted (Fig. 4.5A).
Detergent cell extracts were prepared 48 hours later, when cells were 40% confluent, and probed
for Stat3-ptyr705 as above. As shown in Fig. 4.5B, there was a dramatic and graded increase in
Stat3-ptyr705 levels, in proportion to the amounts of E/EC12 used to decorate these surfaces
(lanes 4-12 and 16). As expected, coating with the E/W2A fragment had only a small effect (~4x
lower than the E/EC12 fragment, Fig. 4.5B, bar graph), consistent with its impaired ability to
engage wild- type E-cadherin (Perret et al., 2002). The increase in Stat3-ptyr705 was most
pronounced when cells were grown for more than 20 hrs on E/EC12-coated dishes (Appendix,
Fig. A.5A), and when the confluence at harvest time was 40% or less (Appendix, Fig. A.5B).
The increase in Stat3-ptyr705 levels was specific to Stat3, since no differences were noted in the
levels of activated Erk1/2 in the same cell extracts (Fig. 4.5B). Similar results were obtained with
MDCK cells (not shown). At the same time, Balb/c 3T3 cells, which are naturally devoid of Ecadherin (Fig. 4.6A and Appendix, Fig. A.1A) showed no increase in Stat3-ptyr705 upon growth
on surfaces coated with the E/EC12 fragment (Fig. 4.5B, lanes 1-2). The above findings
demonstrate a dramatic increase in Stat3-ptyr705 upon direct E-cadherin engagement in mouse
epithelial cells, which is specific to Stat3 and proportional to the density of E/EC12 present on the
culture surface, indicating that E-cadherin engagement alone is sufficient to activate Stat3, in the
absence of direct cell to cell contact.
110
111
Figure 4.5
E-cadherin engagement is sufficient to activate Stat3.
A. Schematic drawing of epithelial cells plated at low density on plastic petris coated with an Ecadherin fragment encompassing the two outermost EC domains (E/EC12) (Boggon et al., 2002).
Right: Phase-contrast photographs of HC11 cells grown for 48 hours on an E/EC12-coated
surface (0, 100 or 1,000 µg/ml). Bar corresponds to 100 μm.
B: 0.3x105 HC11 cells were grown on plastic 3 cm dishes coated with increasing amounts of
E/EC12, from 0.5 to 1,000 µg/ml. 48 hours later, extracts were probed for Stat3-ptyr705 or total
Stat3, as indicated. As a further control (lanes 1 and 2), Balb/c 3T3 cells, which are devoid of Ecadherin (Appendix, Fig. A.1A), were grown on the same surfaces. Lanes 13 and 14: Surfaces
were coated with the largely inactive, mutant fragment of E-cadherin, E/W2A.
Bar graph: The average relative Stat3-ptyr705 levels across 3 independent experiments were
quantitated by fluorimager analysis, and graphed as a function of the loading control, Hsp90 ( ),
or as a function of total Stat3 (■). The peak value of HC11 cells grown to 2 days postconfluence
on an uncoated surface was taken as 100%.
The same extracts were probed for phospho-Erk1/2 or Hsp90 as a loading control, as
indicated (lower panels).
Cadherin-11 engagement in Balb/c 3T3 can also activate Stat3
Cadherins were originally identified as cell surface glycoproteins responsible for
calcium-dependent, cell to cell adhesion in development. More than 100 family members have
been discovered to date with diverse structures, but with the characteristic extracellular, EC
repeats [reviewed in (Halbleib and Nelson, 2006; Gumbiner, 2005)]. To examine whether other
cadherins might also be able to activate Stat3, we used the mouse fibroblast line Balb/c 3T3
which expresses cadherin-11 (Orlandini and Oliviero, 2001), a prototype type II classical
cadherin, but not E-cadherin (Fig. 4.6A). Examination of Stat3-ptyr705 levels at different
densities revealed a distinct increase (Fig. 4.6B), similar to HC11 (Fig. 4.1B) or 10T½ fibroblasts
[(Vultur et al., 2004; Vultur et al., 2005) and Appendix, Fig. A.11], suggesting that cadherin-11
homophilic interactions may be indispensable for the cell to cell adhesion-mediated, Stat3
activation. The levels of Erk1/2 also remained unaffected by density in this cell line (Fig 4.6B)
indicating that cell confluence causes a specific increase in Stat3-ptyr705 levels in mouse Balb/c
3T3 fibroblasts. Stat3 activity levels mirrored the increase in ptyr705 (Fig 4.6C). In addition,
112
cell density increased the levels of the Stat3 transcriptional target, survivin (Fig. 4.6C). These
observations are in keeping with previous data indicating that Stat3 transcriptional activity
increases with the density of cultured fibroblasts (Vultur et al., 2004; Vultur et al., 2005), and
further point to the possibility that cadherin-11 may be responsible for the Stat3 activation
observed at high cell densities in Balb/c 3T3 fibroblasts.
To examine whether cadherin-11 was responsible for the density-mediated increase in
Stat3-ptyr705 levels, cadherin-11 was knocked down through shRNA expression (see Materials
and Methods). As shown in Fig. 4.6D, downregulation of cadherin-11 resulted in a reduction in
Stat3-ptyr705 levels, indicating that cadherin-11 is indeed responsible for the Stat3, tyr705
phosphorylation observed at high densities.
To further demonstrate whether, like E-cadherin, direct cadherin-11 engagement might
also be able to lead to Stat3 activation, Balb/c 3T3 cells were plated on petri dishes coated with a
cadherin-11 fragment encompassing the two outermost EC domains (11/EC12). The results
revealed that this treatment caused a dramatic increase in Stat3-ptyr705 (Fig. 4.7C, lanes 2-5),
while there was no increase in Stat3-ptyr705 when Balb/c 3T3 cells were plated on petri dishes
coated with the E-cadherin-derived fragment, E/EC12 (Fig. 4.7C, lane 1). As a further control,
HC11 cells were plated on surfaces coated with 11/EC12 or on E/EC12, and Balb/c3T3 cells were
plated on surfaces coated with the E/EC12 fragment. As expected, there was no increase in Stat3ptyr705 when HC11 cells, which are naturally devoid of cadherin-11 (Fig. 4.6A), were grown on
surfaces coated with 11/EC12 (Fig. 4.7B, lane 8), while there was an increase upon growth of
HC11 cells on E/EC12-coated surfaces (lane 9). The increase was specific to Stat3, since no
increase in Erk1/2 upon plating of Balb/c 3T3 cells on 11/EC12-coated surfaces was noted (Fig.
4.7C). These results indicate that, apart from E-cadherin, cadherin-11 can also increase Stat3ptyr705 phosphorylation levels in the appropriate cellular context.
113
Figure 4.6
Cadherin-11 engagement can trigger Stat3 phosphorylation in Balb/c 3T3 cells
A: Examination of E-cadherin and cadherin-11 mRNA levels in HC11 and Balb/c 3T3 cells.
Extracts from Balb/c3T3 or HC11 cells were probed by RT-PCR for E-cadherin (lanes 1
and 2) or cadherin-11 (lanes 3 and 4), with 18S RNA as a control (lanes 6 and 7) (see Materials
and Methods). Numbers on the left refer to the molecular weight marker lane (M).
B: Cell density upregulates Stat3-ptyr705 levels in Balb/c3T3 cells.
Lysates from Balb/c3T3 fibroblasts grown to increasing densities were resolved by gel
electrophoresis and probed for Stat3-ptyr705, total Stat3, phospho-Erk1/2 or Hsp90 as a loading
control, as indicated.
C: Cell density upregulates Stat3 transcriptional activity in Balb/c 3T3 cells.
Upper panel: Balb/c 3T3 cells were transfected with the Stat3-dependent pLucTKS3 reporter
driving a firefly luciferase gene under control of the C-reactive gene promoter element, and the
Stat3-independent pRLSRE reporter driving a Renilla luciferase gene under control of the c-fos
serum response element (SRE) promotor, respectively. Cells were grown to the indicated
densities with daily media changes and firefly (■) or Renilla ( ) luciferase activities determined
114
in cytosolic extracts with the peak value of the control, v-src transformed cells taken as 100%
(see Materials and Methods). Values shown represent luciferase units expressed as a percentage
of the highest value obtained, means±SEM of at least 3 experiments, each performed in triplicate.
Lower panel: Detergent cell extracts of Balb/c 3T3 cells were probed for survivin or Hsp90 as a
loading control, as indicated.
D: Cadherin-11 knockdown dramatically reduces Stat3 phosphorylation
Lysates from Balb/c 3T3 fibroblasts or Balb/c 3T3 stably expressing an shcad11
construct (see Materials and Methods) were grown to increasing densities. Lysates were resolved
by gel electrophoresis and probed for Stat3-ptyr705, and Hsp90 as a loading control.
Figure 4.7
Cadherin-11 engagement is sufficient to activate Stat3 in Balb/c 3T3 fibroblasts
A: Schematic drawing of fibroblasts plated at low density in plastic petris coated with a
cadherin-11 fragment encompassing the two outermost EC domains (11/EC12) (Boggon et al.,
2002).
115
B: Balb/c 3T3 cells were grown in plastic 3cm dishes, coated with increasing amounts of the
cadherin-11 fragment, 11/EC12, as indicated. 48 hours later, cell lysates were probed for Stat3ptyr705, phospho-Erk1/2 or Hsp90 as a loading control, as indicated. As controls Balb/c 3T3
cells, which are devoid of E-cadherin (see Fig. 4.6A above), were grown on a surface coated with
the corresponding E-cadherin fragment (E/EC12, lane 1) and HC11 cells which are devoid of
cadherin-11 were grown on surfaces coated by cadherin-11 (lanes 7 and 8), or by E-cadherin
(lane 9).
Cell to cell adhesion increases protein levels and activity of Rac1 and Cdc42
Results from a number of laboratories have indicated that E-cadherin-mediated cell to
cell contacts can activate the Rac1 and Cdc42, Rho family GTPases (Kovacs et al., 2002; Noren
et al., 2001; Nakagawa et al., 2001). Therefore, to assess the potential role of Rac1 in the
cadherin-dependent, Stat3 activation, we investigated whether Rac1 activity does, in fact,
increase with cell density; cells were plated in petri dishes and when ~30% confluent, and over
several days thereafter, Rac1 activity was determined by measuring the binding between Rac1GTP and its effector, p21-activated kinase (PAK), in cell extracts using pull-down assays (see
Materials and Methods). As shown in Fig. 4.8A, left, there was a gradual increase in Rac1
activity with density in HC11 cells, which plateaued at approximately 80-100% confluence, that
is ~24 hours before the rise in Stat3-ptyr705. Cdc42 activity displayed a similar increase with
cell confluence (Fig. 4.8A, right), in a manner parallel to Rac1. In addition, cell density led to an
upregulation of Rac1-GTP and Cdc42-GTP levels in Balb/c 3T3 cells (Fig. 4.10A), which express
cadherin-11. The above results demonstrate that cell density causes an increase in Rac1-GTP and
Cdc42-GTP (~6-fold), which is detectable before the rise in Stat3 activity, in both HC11
epithelial cells and Balb/c 3T3 fibroblasts.
It was previously demonstrated that epithelial cell scattering brought about by HGF can
induce the proteasome-mediated degradation of Rac1, pointing towards a new mechanism of
Rac1 regulation involving protein stability (Lynch et al., 2006). Therefore, to explore the
potential effect of cell density upon the levels of total Rac1 protein, detergent extracts from cells
116
Figure 4.8
Cell density increases the activity as well as total protein levels of Rac1 and Cdc42 in
HC11 cells.
A: Rac1-GTP, Cdc42-GTP as well as total Rac1 and Cdc42 protein levels are dramatically
increased by cell density.
HC11 cells were grown to increasing densities, up to 11 days post-confluence, and total
protein extracted. Rac1/Cdc42 activation assays were performed using beads coated with
glutathione-S-transferase (GST) fused to the binding domain of PAK (PAK-PBD) in pulldown
assays (see Materials and Methods). Detergent lysates were resolved by gel electrophoresis and
blots were cut into strips which were probed for Stat3-ptyr705, active Rac1-GTP and total Rac1
(left) or active Cdc42-GTP and total Cdc42 (right), or Hsp90 as a loading control, as indicated.
See Appendix, table A.1 for quantitation.
B: Rac1 and Cdc42 mRNA levels are not affected by density.
Total mRNA from HC11 cells grown to different densities was subjected to RT-PCR
analysis, with the 18S RNA as a control (see Materials and Methods).
Bottom: Quantitative RT-PCR was performed as described in Materials and Methods. The
relative expression of each sample was determined according to 18S RNA expression levels as an
internal control.
117
Figure 4.9
E-cadherin engagement increases Rac1 protein levels and activity in the absence of
direct cell to cell contact
A: Cadherin engagement is sufficient to increase Rac1 activity and protein levels.
Lanes 1-4: HC11 cells were grown in plastic petris coated with 1 mg/ml of the E/EC12 (lane 3)
or the largely inactive E/W2A (lane 2) fragment. 48 hours later, detergent cell extracts were
resolved by gel electrophoresis and blots were cut into strips which were probed for Stat3ptyr705, total Rac1 or Hsp90 as a loading control, as indicated. Lanes 5-6: Cadherin
engagement increases Rac1 activity: Sparsely growing HC11 cells were plated on uncoated
petris (lane 5) or petris coated with 1 mg/ml of the E/EC12 fragment (lane 6) and Rac1 activity
determined (see Materials and Methods).
Numbers under the lanes refer to Stat3-ptyr705, Rac1-GTP and total Rac1 band intensities
obtained through quantitation by fluorimager analysis and normalized to Hsp90 levels, with the
peak value of the control untreated cells taken as 100%.
B: Inhibition of E-cadherin engagement prevents Stat3 phosphorylation and the increase in
Rac1 activity and protein levels.
118
Lanes 1-5: HC11 cells were grown to 1 day post-confluence (lanes 1-4) or 40% confluence (lane
5) and treated with 40 µM soluble E/EC12 fragment (lane 2), or the E/W2A mutant (lanes 3 and
4).
Lanes 6-13: HC11 cells were grown to 1 day post-confluence (lanes 6 or 11, untreated) or 40%
confluence (lane 13, untreated) and subject to increasing concentrations of the SHAVSA peptide
(lanes 7-9), the inactive SHGVSA (lane 10) or the weakly active SHAVSS (lane 12). Rac1 levels
were determined in the same extracts and the same blot as in Fig. 4.2A.
Lanes 14-17: Inhibition of cadherin engagement reduces Rac1 activity. HC11 cells were
grown to 1 day post-confluence, treated with the indicated amounts of peptide for 48 hours and
Rac1 activity determined (see Materials and Methods).
Figure 4.10
Cadherin-11 increases Rac1/Cdc42 activity and protein levels in Balb/c 3T3 cells.
A: Rac1 and Cdc42 activity, as well as expression, and Stat3-ptyr705 levels are dramatically
increased by cell density in Balb/c 3T3 cells.
119
Balb/c 3T3 cells were grown to different densities, up to 5 days post-confluence.
Detergent cell lysates were probed for Stat3-ptyr705, active Rac1-GTP, total Rac1, active Cdc42,
total Cdc42 or Hsp90 as a loading control, as indicated.
B: Cadherin-11 engagement is sufficient to increase Rac1 protein levels and activity.
Balb/c 3T3 cells were grown in plastic petris coated with 1mg/ml of the 11/EC12
fragment or the the E-cadherin-derived, E/EC12 fragment. 48 hours later, detergent cell extracts
were probed for Stat3-ptyr705, Rac1 or Hsp90 as a loading control, as indicated.
C: Rac1 and Cdc42 mRNA’s are not affected by cell density.
Balb/c 3T3 cells were grown to different densities as indicated and Rac1 or Cdc42
mRNA levels examined by RT-PCR, using 18S RNA as a control (see Materials and Methods).
grown to different degrees of confluence were probed for Rac1. As shown in Fig. 4.8A, left, and
Fig. 4.10A, there was a ~50-fold increase in total Rac1 protein levels with cell density. The
increase in total Rac1 protein was sharp, in contrast to the gradual increase in Rac1 activity, and
could explain the further increase in active Rac1-GTP at higher densities. Total Cdc42 levels
mirrored Rac1 and increased at approximately the same time (Fig. 4.8A, right, and Fig. 4.10A).
At the same time, Rac1 and Cdc42 mRNA levels did not change significantly with cell density, as
shown by quantitative RT-PCR (Figs. 4.8B and 4.10B), pointing to a post-mRNA mechanism.
The above results indicate that, in addition to Rac1 and Cdc42 activity, cell to cell adhesion also
causes a dramatic increase in total Rac1 and Cdc42 protein levels in HC11 and Balb/c 3T3.
To better control the degree of cell to cell contact, irrespective of differences in cell
growth patterns, we repeated the experiment by plating different numbers of HC11 cells (ranging
from 0.05-3x106 per 3cm petri), so that they would reach the same densities as above within 24
hrs. At that time, cells were lysed and Rac1, Cdc42 and Stat3 phosphorylation levels analysed as
above. In all cases, very similar results were obtained, indicating that it is the extent of cell to cell
contact, regardless of time in culture beyond 24h, that is responsible for the increase in Rac1 or
Cdc42 protein levels and Stat3-ptyr705 (Appendix, Fig. A.7). Similar results were obtained with
Balb/c 3T3 (not shown).
120
E-cadherin and cadherin-11 engagement can increase protein levels and activity of Rac1
and Cdc42
To examine whether E-cadherin or cadherin-11 engagement alone is able to increase
Rac1 protein levels in the absence of cell to cell contact, HC11 cells or Balb/c 3T3 cells were
plated in E/EC12- or 11/EC12-coated dishes, as in Fig. 4.5B or Fig. 4.7B, respectively, and Rac1
levels examined. As shown in Fig. 4.9A, plating HC11 cells on E/EC12-coated surfaces (lanes 14), besides leading to an increase in Stat3-ptyr705 (top), also caused a ~50-fold increase in Rac1
protein levels (lanes 1 vs 3) and activity (lanes 5 and 6), while the mutant E/W2A fragment had a
less pronounced effect (lane 2). Similarly, plating Balb/c 3T3 cells on surfaces coated with the
11/EC12 fragment led to a striking increase in Rac1 protein levels (Fig. 4.10C, lanes 1 vs 3) and
activity (lanes 8 vs 9), while plating them on E/EC12 had no effect on Stat3-ptyr705 or Rac1
levels (Fig. 4.10C, lanes 1 vs 2).
To further examine whether cadherin engagement is required for the increase in Rac1
protein levels, E-cadherin interactions were disrupted through the introduction of the E/EC12
fragment into the medium, as in Fig. 4.2A. As shown in Fig. 4.9B, besides a reduction in Stat3ptyr705, addition of 40 μM, soluble E/EC12 fragment to the medium of densely growing HC11
cells caused a ~6-fold decrease in Rac1 levels (lane 2), while the mutant E/W2A had a weaker
effect (~3-fold, lanes 3 and 4). Inhibition of cadherin engagement with the SHAVSA peptide had
a similar effect upon Rac1 activity (Fig. 4.9B, lanes 14-17) as well as total Rac1 and Stat3ptyr705 levels (Fig. 4.9B, lanes 6-9). Preliminary data also show that Balb/c 3T3 cells stably
expressing shcad11 have reduced Rac1 protein levels and activity (not shown). These results
taken together further demonstrate that E-cadherin or cadherin-11 engagement is required for the
cell to cell adhesion-mediated increase in total Rac1 protein levels and activity in HC11 or Balb/c
cells, respectively.
121
Rac1 and Cdc42 are required for the cell to cell adhesion-mediated, Stat3 activation
To examine whether the increase in cellular Rac1 and Cdc42 we observe in confluent
cultures is actually required for the Stat3 increase, their activity was blocked in HC11, Balb/c 3T3
or 10T½ cells using the potent inhibitor of the Rho family GTPases, toxin B of Clostridium
difficile (Aktories, 1997). As shown in Fig. 4.11A, besides a reduction in Rac1 and Cdc42
activity (right), toxin B treatment caused a dramatic reduction in Stat3-ptyr705 levels in HC11 at
all densities (~5-fold, left). Similar results were obtained with 10T½ fibroblasts, which express
cadherin-11 (Appendix, Fig. A.9, A-B) and Balb/c 3T3 fibroblasts (not shown), indicating that
these GTPases may be required for the density-induced, Stat3 stimulation in both cell types.
To further examine the relative contribution of each Rho family GTPase individually, we
assessed the ability of their dominant-negative mutants (Rac1N17 and Cdc42N17, respectively), or
Rac1- and Cdc42-specific shRNAs, to restrain the density-mediated increase in Stat3-ptyr705
levels. These constructs were expressed in HC11, 10T1/2 or Balb/c 3T3 cells and Stat3-ptyr705
examined at different cell densities as above. As shown in Fig. 4.11B and C, Appendix, Figs.
A.9D, A.10 and Tables A.1A and B, there was a significant reduction in Stat3-ptyr705 levels (~2fold), upon expression of Rac1N17 or shRac1 and Cdc42N17, or shCdc42 at all densities examined.
The residual Stat3-ptyr705 following Rac1 downregulation (Fig. 4.11B, lanes 5-8 and C, lanes 712 and Appendix, Fig. A.9D, top panel, lanes 8-14) could be due to Stat3 phosphorylation
mediated by Cdc42, and vice versa, following Cdc42 dowregulation. Rac1 downregulation with
the Rac1N17 or shRNA caused a similar decrease in Stat3 transcriptional activity (Fig. 4.11B and
C). These results were confirmed using the Rac1 specific inhibitor NSC23766 (Appendix, Fig.
A.9C). Similar results were obtained in Balb/c 3T3 upon shRac1 expression (not shown).
Conversely, to ensure that Rac1 activation is not caused by Stat3, Stat3 activity was
reduced by [1], expressing a Stat3-specific, shRNA with a lentiviral vector, using a scrambled
sequence as a control (see Materials and Methods), or [2], treating the cells with the compound
122
123
Figure 4.11
Inhibition of Rac1 or Cdc42 prevents the density-induced, Stat3 activation.
A: Toxin B, a Rho GTPase protein family inhibitor prevents the density-mediated, Stat3tyr705 phosphorylation.
Left: HC11 cells were grown from 70% to 7 days post-confluence with daily media changes and
treated with 50 pg/ml toxin B for 5 hours (lanes 6-10), or the DMSO carrier alone (lanes 1-5), and
lysates resolved by gel electrophoresis. Extracts were probed for Stat3-ptyr705 (top) or Hsp90 as
a loading control (bottom), as indicated.
Right: HC11 cells were grown to 1 day post-confluence and treated with 50 pg/ml toxin B for 5
hours (lane 2) or the DMSO carrier alone (lane 1) and the activity of Rac1 or Cdc42 determined
(see Materials and Methods).
B: The dominant-negative Rac1N17 mutant prevents the density-mediated, Stat3-tyr705
phosphorylation and transcriptional activity
Left: Increasing numbers of HC11 cells as indicated, before (lanes 1-4) or after (lanes 5-8)
transfection with the dominant-negative mutant Rac1N17 were plated in 3 cm dishes. Detergent
cell extracts were resolved by gel electrophoresis and blots were probed for Stat3-ptyr705, active
Rac1-GTP, total Rac1, the myc-tag or Hsp90, as above, as indicated.
Right: Increasing numbers of HC11 cells as indicated, before or after transfection with Rac1N17
and the Stat3-dependent pLucTKS3 and the Stat3-independent pRLSRE reporters, were lysed and
firefly (■) or Renilla ( ) luciferase activities determined (Vultur et al., 2004). Values shown
represent luciferase units expressed as a percentage of the highest value obtained, means ± s.e.m.
of at least 3 experiments, each performed in triplicate.
C: Rac1 downregulation through expression of an shRNA prevents the density-mediated,
Stat3-tyr705 phosphorylation and transcriptional activity
Top: HC11 cells before (lanes 1-6) or after (lanes 7-12) stable expression of a Rac1-shRNA were
grown to different densities and lysates probed, as above, for Stat3-ptyr705, active Rac1-GTP
(right), total Rac1 or Hsp90, as indicated.
Bottom: HC11 cells before or after expression of a Rac1-shRNA as indicated, were transfected
with the Stat3-dependent pLucTKS3 and the Stat3-independent pRLSRE reporters, grown to
different densities, then lysed and firefly (■) or Renilla ( ) luciferase activities determined
(Vultur et al., 2004). Values shown represent luciferase units expressed as a percentage of the
highest value obtained, means ± s.e.m. of at least 3 experiments, each performed in triplicate.
S3I-201 (Siddiquee et al., 2007), shown to inhibit Stat3 by binding to its SH2 domain (see
Materials and Methods). No significant change in Rac1 levels or activity was noted in HC11
cells under either condition (Fig. 4.12B and Table 4.1). Similar results were obtained in 10T½
and Balb/c 3T3.
124
Table 4.1
Stat3 downregulation promotes apoptosis in confluent HC11 cells
Cell line
Treatment
c-RNA
shRNA
S3I-201
HC11
HC11-Stat3C
Confluence
Stat3-ptyr705
Apoptosis
Rac1-GTP
Confluence
Stat3C-ptyr705
Apoptosis
(%)*
(%)†
(%)‡
(%)§
(%)*
(%)†
(%)‡
80
32±11
2±1
78±10
80
85±11
1±0.2
100
80±15
2±1
85±11
100
190±21
2±1
+3d
100±12
5±1
100±18
+3d
300±34
2±1
80
10±3
8±1
75±11
80
32±5
5±1
100
30±9
22±2
82±12
100
89±7
7±2
+3d
39±8
69±7
100±19
+3d
122±19
10±3
80
18±6
5±1
72±10
80
24±4
6±2
100
29±5
18±4
88±13
100
62±7
8±3
+3d
51±11
55±8
98±17
+3d
101±12
12±3
HC11 or HC11-Stat3C cells were grown to different densities and treated with the Stat3shRNA, a control, scrambled c-RNA or 200μM of the inhibitor S3I-201, as indicated, and Stat3ptyr705, active Rac1-GTP and apoptosis examined.
Confluence was estimated visually and quantitated by imaging analysis of live cells under phase
contrast (see Materials and Methods).
*
Stat3-ptyr705 levels were quantitated by fluorimager analysis and normalized to Hsp90 levels,
with the values for HC11 cells at 2 days post-confluence, treated with the control RNA taken as
100%. The results are averages of at least three independent experiments ± s.e.m. In all cases, the
EMSA and transcriptional activity values obtained paralleled the Stat3-705 phosphorylation
levels indicated.
†
To quantitate apoptosis, cells were stained with propidium iodide and their sub-G1 profile
analysed by FACS sorting. Numbers refer to cells in subG1 as a % of total cells. The results are
averages of at least three independent experiments ± s.e.m.
‡
§
Rac1 activity was measured as described in Materials and Methods.
125
The above data taken together indicate that the Rac1 and Cdc42 Rho family GTPases are
essential components of the pathway whereby cadherin engagement triggers the Stat3
phosphorylation and activity increase observed at post-confluence in epithelial cells and
fibroblasts.
JAKs are required for the cadherin-mediated, Stat3 activation
Previous results indicated that cell density induces an increase in Jak1 kinase activity
(Vultur et al., 2004). To further investigate the role of Jak1 in the cadherin-mediated, Stat3
activation, we at first examined whether Jak1 is activated by E-cadherin engagement, by
examining Jak1 phosphorylation at tyr1022/1023, shown to be important in the regulation of Jak1
activity (Gauzzi et al., 1996; Leonard and O'Shea, 1998). HC11 cells were grown in plastic petris
coated with the E/EC12 fragment and Jak1 phosphorylation at 1022/1023 measured by blotting
with a phospho-specific antibody (see Materials and Methods). As shown in Fig. 4.12A (lanes 8
and 9), there was a 5-fold increase in Jak1 1022/1023 phosphorylation upon growth on E/EC12coated surfaces, which paralleled the phosphorylation of Stat3 (lanes 1 and 2). To examine
whether Jak1 is actually required for Stat3 phosphorylation, HC11 cells were plated on the
E/EC12 fragment and treated with the pan-JAK inhibitor, JAK inhibitor 1. The results showed a
reduction in Stat3-ptyr705 levels, proportional to the concentration of the inhibitor added (Fig.
4.12A, lanes 3-7). Treatment of Balb/c 3T3 with JAK inhibitor I also reduced the cell-cell
adhesion dependant Stat3-tyr705 phosphorylation (see Appendix, Fig. A.12B). Similar results
were obtained with the AG490 JAK inhibitor ((Zhang et al., 2000); not shown). These data
suggest that the JAK kinases may be involved in the cadherin-mediated increase in Stat3-ptyr705
levels.
We next examined the effect of downregulation of the Rho GTPases upon Jak1,
1022/1023 phosphorylation (Fig. 4.12A). HC11 cells were treated with toxin B as in Fig. 4.11A,
left and Jak1 ptyr1022/1023 phosphorylation measured as above. As shown in Fig. 4.12A,
126
Figure 4.12
JAKs are required for the cadherin-dependent Stat3-tyr705 phosphorylation.
A: JAK inhibitor-1 reduces the cadherin-dependent, Stat3-tyr705 phosphorylation.
Left: HC11 cells were grown to a confluence of 40% on 3 cm dishes coated with the E-cadherin
fragment, E/EC12, as indicated, and treated with the pan-JAK, JAK inhibitor-1 at the indicated
concentrations (lanes 4-7) or not (lanes 1-3) for 48 hours, at which time cell lysates were probed
as above, for Stat3-ptyr705, Rac1 or Hsp90.
Right: Lysates were immunoprecipitated against total Jak1 and blotted against p-Jak1022/1023,
total Jak1, or Hsp90 as a loading control, as indicated.
127
Bottom: HC11 cells were grown to 1 day post-confluence and treated with 50 pg/ml toxin B for
5 hours (lane 2) or the DMSO carrier alone (lane 1) (see Materials and Methods). HC11 (lane 3)
and HshRac1 (lane 4) cells were grown to 1 day post confluence. Lysates were
immunoprecipitated against total Jak1 and blotted against p-Jak1022/1023, total Jak1, or Hsp90
as a loading control, as indicated.
B: Stat3 inhibition does not affect Rac1 protein levels. HC11 cells were grown to different
densities as indicated and treated with 100 μM of the S3I-201, Stat3 inhibitor (lanes 2,4,6,8), or
the DMSO carrier alone (lanes 1,3,5,7) for 24 hrs, at which time cell lysates were resolved by gel
electrophoresis and blots cut into strips which were probed for Stat3-ptyr705, Rac1 or Hsp90.
bottom, lane 2, toxin B treatment caused a 5-fold reduction in Jak1 phosphorylation. To further
demonstrate the effect of Rac1 upon Jak phosphorylation, we examined Jak1, ptyr1022/1023
levels in HC11 cells expressing the Rac1 shRNA. As shown in Fig. 4.12A, bottom, lane 4, Rac1
knockdown reduced Jak1 1022/1023 phosphorylation by ~2-fold. Jak1 in vitro kinase assays
gave similar results (Vultur et al., 2004). These findings taken together indicate that Rac1
activity is required for the activation of Jak1 observed at high densities.
Cell to cell adhesion triggers cytokine gene expression
As shown in Figs. 4.8A and 4.10A, the increase in Stat3-tyr705 phosphorylation occurs
~24 hrs after the surge in Rac1 levels, which points to the possibility of an indirect mechanism.
To explore the likelihood that the Stat3 activation by cadherin may be mediated by secreted
factors, medium conditioned by HC11 cells grown to high densities was added to sparsely
growing HC11 cells for 15 minutes and Stat3-ptyr705 examined. The results revealed a ~20-fold
increase in Stat3-ptyr705, indicating the presence of autocrine factors, able to activate Stat3 (Fig.
4.10A, lane 2 vs 1). To examine the nature of the cytokines that may be expressed, we conducted
a quantitative RT-PCR array for 86 cytokines, and compared sparsely growing HC11 or Balb/c
3T3 cells to cells grown as dense cultures (see Materials and Methods). The results revealed an
increase in mRNA levels of a number of cytokines, including the IL6 family, known to act
through the common gp130 subunit, shared by a number of Stat3 activating cytokines, such as
128
IL6, LIF, CT-1 and IL27 in HC11 [76-fold increase for IL6 mRNA, see Appendix, Table A.2A;
(Fischer and Hilfiker-Kleiner, 2007)]. In Balb/c 3T3 cells IL6 and LIF mRNA gave a similar
increase in confluent cells (Appendix, Table A.2B). To examine whether these cytokines are
required for the Stat3 activation observed in confluent cultures, gp130 levels were reduced
through expression of shRNA (see Materials and Methods). As shown in Fig. 4.13B, a 70%
gp130 knockdown reduced Stat3-ptyr705 levels in HC11 by ~60%, indicating that gp130
activation is at least partly responsible for the Stat3-ptyr705 increase. Stat3-ptyr705 levels were
also reduced in Balb/c 3T3 cells upon gp130 knockdown (not shown).
Results from a number of labs indicated that, besides Stat3, IL6 stimulation results in
activation of Erk1/2 (Erk) (Fischer and Hilfiker-Kleiner, 2008). However, as shown in Figs. 4.1B
and 4.7B, E-cadherin or cadherin-11 engagement does not activate Erk. To solve this apparent
paradox, we examined the ability of IL6, or conditioned medium, to activate Erk in confluent
cultures. HC11 cells were grown to 60% or 100% confluence, serum-starved and following IL6
stimulation, cell extracts were probed for Erk or Stat3-ptyr705. As shown in Fig. 4.13C, at a
confluence of 60%, IL6 addition caused an increase (~10-fold) in both Erk and Stat3, in
agreement with previous results (Fischer and Hilfiker-Kleiner, 2008). As expected, cell density
per se caused an increase in Stat3-ptyr705 levels (Fig. 4.13C, lanes 1 vs 3), and IL6 caused a
further activation (~30%, lane 4). However, IL6 did not bring about an increase in Erk levels in
densely growing cultures (Fig. 4.13C, middle, lane 4 vs 3). The above findings indicate that IL6
is not able to activate Erk in cells grown to high densities, thus revealing a profound effect of
confluence on the response of HC11 cells to IL6 addition.
Inhibition of E-cadherin engagement at confluence promotes apoptosis
Previous results have shown that Stat3 signalling contributes to the induction of antiapoptotic genes, such as bcl-xL and mcl-1 (Grandis et al., 2000; Epling-Burnette et al., 2001)
while it downregulates the p53 promotor (Niu et al., 2005), thus protecting tumor cells from
129
Figure 4.13
Cell to cell adhesion triggers cytokine gene expression.
A: Confluence induces autocrine secretion of Stat3-activity factors. Conditioned medium
from HC11 cells grown to 100% confluence was added for 15 min to HC11 cells grown to 40%
(lane 2) or 100% (lane 4) confluence. At this time detergent cell extracts were resolved by gel
electrophoresis and blots were cut into strips which were probed for Stat3-ptyr705, p-Erk1/2 or
Hsp90 as indicated. Numbers under the lanes refer to Stat3-ptyr705 or p-Erk1/2 band intensities
obtained through quantitation by fluorimager analysis and normalized to Hsp90 levels, with the
highest value taken as 100%.
130
B: gp130 downregulation reduces the density-mediated Stat3 activation. HC11 cells before
(lanes 1-8) or after (lanes 9-16) expression of shRNA for gp130 were grown to different densities
and lysates probed for Stat3-ptyr705, gp130 (bottom), Rac1 or Hsp90, as indicated. Numbers
under the lanes refer to Stat3-ptyr705 or gp130 band intensities obtained through quantitation by
fluorimager analysis and normalized to Hsp90 levels, with the highest value taken as 100%.
C: IL6 activates Stat3 but not Erk at high densities. IL6 was added at 10 ng/ml for 15 min to
HC11 cells grown to 60% (lane 2) or 100% (lane 4) confluence and cell extracts probed for Stat3ptyr705, p-Erk1/2 or Hsp90 as indicated. Numbers under the lanes refer to Stat3-ptyr705 or pErk1/2 band intensities obtained through quantitation by fluorimager analysis and normalized to
Hsp90, with the highest value taken as 100%.
apoptosis. The importance of Stat3 in survival of confluent, HC11 cells was examined by
reducing Stat3 activity by shRNA expression or treatment with S3I-201 as above. As a control,
the constitutively active mutant, Stat3C, was expressed with a retroviral vector in HC11 cells
prior to treatment (McLemore et al., 2001). As shown in Table 4.1, Stat3 downregulation in
confluent HC11 epithelial cells using either approach induced apoptosis, in agreement with
previous data from mouse NIH3T3 or 10T1/2 fibroblasts (Anagnostopoulou et al., 2006; Vultur et
al., 2004), indicating that the increased Stat3 activity may reflect a survival mechanism that is
engaged as cells reach confluence. Similar results were obtained with the ES, Ecad+/+ cells, as
well as following reduction of Stat3 levels through Rac1 (Fig. 4.11) or Cdc42 (Appendix, Fig.
A.10) or gp130 (Fig. 4.13B) downregulation (not shown). At the same time, Stat3C expression
rescued HC11 cells from apoptosis induced by Stat3 inhibition, thus reinforcing this conclusion
(Table 4.1).
To evaluate the functional consequences of the E-cadherin-mediated, Stat3 activation, we
examined the effect of E-cadherin ablation in ES cells. Ecad+/+ and null cells were plated in
plastic petri dishes and apoptosis examined by TUNEL and PARP cleavage experiments as well
as by FACS analysis, at different densities. As shown in Fig. 4.14, the E-cadherin null cells
succumbed to apoptosis (~70%) when confluent (A, panel f, B, bottom, and D, lanes 4-6), while
no significant apoptotic death was noted with the Ecad+/+ cells (A, panel b, B, top, and D, lanes
131
132
Figure 4.14
E-cadherin engagement promotes survival.
A: Ecad+/+ (a-d) or null (e-h) cells were grown to densities of two days post-confluence and
stained for E-cadherin (a, e), DAPI (c, g) or TUNEL to examine apoptosis (b, f) (see Materials
and Methods). Scale bar in overlayed images (d, h) represents 30 μm.
B: Ecad+/+ (top) or null (bottom) cells were grown to densities of two days post-confluence,
stained with propidium iodide and subject to FACS analysis (see Materials and Methods). Bar
indicates cells with a sub-G1 DNA content.
C: Inhibition of cadherin engagement with the SHAVSA peptide induces apoptosis, which
can be abrogated by Stat3C expression.
b: HC11 cells were grown to a density of two days post confluence, treated with 10 mM
of the SHAVSA peptide for 48 hours and apoptosis examined by TUNEL staining. d: Stat3Cexpressing, HC11 cells, similarly treated (note the absence of apoptosis). a, c: Same panels as in
b and d, phase contrast illumination. Scale bar represents 100 μm.
D: Cadherin engagement is required for cellular survival.
Left: Ecad+/+ (lanes 1-3) or null (lanes 4-6) cells were grown to different densities and apoptosis
examined by PARP cleavage analysis.
Right: HC11 (lanes 2 and 3) or HC11-Stat3C cells were treated with the cadherin-inhibitory
peptide, SHAVSA (lanes 1 and 3) or not (lane 2) and apoptosis examined by PARP cleavage
analysis.
Arrows point to PARP (113 kDa) and cleaved PARP (89 kDa), respectively.
1-3). This observation is in agreement with previous data pointing to a positive role for Ecadherin engagement in cell survival signalling (Liu et al., 2006). To further reinforce this
conclusion, we examined the consequences of inhibition of cadherin engagement in HC11 cells
using the cadherin-antagonistic, SHAVSA, or the inactive SHGVSA peptides. Following
treatment for 48 hours, confluent cells were fixed and apoptotic death assessed. As shown in Fig.
4.14C, b (and Appendix, Table A.3), concomitant with a Stat3 activity reduction, addition of the
cadherin-antagonizing peptide induced apoptosis (~65%), which was more pronounced in
confluent cultures. Expression of the constitutively active Stat3C mutant rendered HC11 cells
(line HC11-Stat3C, Fig. 4.2C) resistant to apoptosis triggered by inhibition of cadherin
133
engagement (Fig. 4.14C, c and d, and D, right, and Appendix, Table A.3) indicating that Stat3
activation can in fact restore survival.
In the context of cancer, a more aggressive role has been assigned to cadherin-11, which
was shown to be aberrantly expressed in tumor cells of epithelial lineage with a more invasive
phenotype. Our preliminary results indicate that genetic knockdown of cadherin-11 through
shRNA in Balb/c 3T3 leads to an increase in apoptotic death as observed by TUNEL staining.
These data further emphasize the central role of cadherin/Stat3 signalling in promoting cell
survival, a mechanism which could be exploited even in the later stages of cancer.
134
Discussion
Cellular interactions with neighboring cells profoundly influence a variety of signalling
events including those involved in mitogenesis, survival and differentiation. Unlike cells cultured
in two dimensions, cells in a tumor have extensive opportunities for adhesion to their neighbors in
a three-dimensional structure. For this reason, in the study of these cellular processes it is
important to take into account the effect of surrounding cells. In the present communication we
demonstrate that homophilic interactions by E-cadherin or cadherin-11 can dramatically activate
Stat3. Furthermore, Stat3 activation is preceded by a striking increase in the total levels of Rac1
and Cdc42 Rho family GTPases, as well as their activity, and is followed by potent survival
signalling. Our finding that downregulation of Rac1 and Cdc42 through expression of dominantnegative mutants, shRNA or pharmacological inhibitors causes a dramatic reduction in Stat3ptyr705 levels, demonstrates that their activation is part of the pathway(s) whereby cadherin
engagement leads to Stat3 activation and survival.
These findings raise important questions: [1] How does cadherin engagement and
Rac1/Cdc42 activate Stat3? [2] What is the functional significance of the cadherin-mediated
Stat3 activation? These are crucial questions in the context of understanding the molecular
mechanisms involving cadherin-mediated adhesion in cellular survival.
Cadherin engagement increases Rac1 and Cdc42 total protein levels and activity
Earlier reports indicated that cell to cell adhesion results in the rapid activation of the
Rac1 (Noren et al., 2001; Nakagawa et al., 2001) and Cdc42 (Kim et al., 2000) Rho family
GTPases (Fukuyama et al., 2006). Our findings, revealing an increase in Rac1 and Cdc42
activity as cells approach confluence, are in agreement with the above data. Still, the activity of
these enzymes at confluences greater than 100%, when the opportunity for cell to cell adhesion is
135
maximized is unknown. Our results reveal a continuing increase in Rac1-GTP and Cdc42-GTP at
higher densities. Unexpectedly, we also found that in addition to this cadherin-induced activation
of Rac1, there is a dramatic increase in total Rac1 protein levels with confluence, leading to a
further increase in activity. This could be due, at least in part, to inhibition of proteasomemediated degradation of Rac1 following cadherin engagement (Lynch et al., 2006). A similar
mechanism could hold true for Cdc42, which mirrored Rac1 activity and protein levels at all
densities examined. Moreover, the absence of a significant increase in Rac1 or Cdc42 mRNA
levels with cell density further points to a post-mRNA mechanism.
In transmitting a proliferative signal, cadherins can act directly as receptors propagating
an intracellular signal, or they may function primarily to bring cells into contact with each other,
to signal via other, juxtacrine receptors. Plating epithelial cells on surfaces coated with the
E/EC12 fragment, or fibroblasts on 11/EC12, demonstrated that cadherin engagement alone is
sufficient to increase Rac1 and Cdc42 activity and protein levels, pointing to a direct mechanism
in the HC11, MDCK and Balb/c 3T3 cell systems. The above findings add to the growing body
of evidence indicating that cadherins can provide direct, functionally relevant signalling beyond
their structural role.
Rac1 and Cdc42 transduce the cadherin signal to Stat3
Data from a number of labs have demonstrated a functional link between the Stat3 and
Rho family pathways (Simon et al., 2000; Faruqi et al., 2001; Debidda et al., 2005). Our results
showing that both pharmacological and genetic Rac1/Cdc42 inhibitors blocked the cadherinmediated Stat3 activation, clearly demonstrate that Rac1/Cdc42 are the mediators of the cadherin
signal to Stat3. Given the complexity of the immediate effector networks controlled by cadherins
(Jaffer and Chernoff, 2004), it is likely that these Rho GTPases may engage multiple effectors,
which could indirectly lead to Stat3 activation. Our findings further indicate that both tyr-705
136
and ser-727 are phosphorylated and JAKs are required for Stat3 activation following cadherin
engagement [(Figs. 4.12 and (Vultur et al., 2004)], expanding on the above studies.
Extensive previous studies using both genetic ablation and pharmacological inhibition
approaches indicated that the cell confluence-mediated, Stat3 activation is resistant to inhibition
of the Src, Fyn, Yes, EGFR, Insulin-like Growth Factor-1 Receptor and Fer kinases, individually
(Vultur et al., 2004). Our present data demonstrate that cell density causes the expression of a
number of cytokines, most importantly the IL6 family. The role of Rac1/Cdc42 activation in
increasing mRNAs for IL6 family cytokines will be discussed in Chapter 5. Although it is
certainly possible that other receptors may also be involved, the fact that downregulation of the
common subunit, gp130, did reduce Stat3 activity indicates that this family is at least in part
responsible for the Stat3 activity increase observed at high cell densities. It is especially
noteworthy that IL6 cannot activate Erk in densely growing cultures, which explains the
observation that cadherin engagement activates Stat3 exclusively, while Erk remains unaffected
in our system. The reasons for the inability of IL6 to activate Erk in confluent cultures is
presently under investigation. In any event, these results further reveal that, despite the fact that
these two pathways are often both activated by growth factors or oncogenes, they are not
coordinately regulated by cadherin engagement.
Cadherin-mediated Stat3 activation leads to survival signalling
The role of E-cadherin in cell proliferation and survival is complex. It is thought to be a
tumor suppressor molecule largely because it is frequently down-regulated in carcinomas
(Gottardi et al., 2001), while in cultured, human colon carcinoma and mammary carcinoma cell
lines, E-cadherin plays a negative role in cell proliferation (Perrais et al., 2007). However, there
is also evidence that E-cadherin is associated with increased cell proliferation. Ovarian cancers
up-regulate E-cadherin, the suppression of which inhibits their proliferation (Sundfeldt, 2003;
Reddy et al., 2005). More recently, E-cadherin engagement in cultured normal rat kidney or
137
mammary epithelial cells was shown to stimulate cell proliferation (Liu et al., 2006). The results
presented here indicate that in our cellular systems, E-cadherin engagement promotes cell
survival. Indeed, we observed a dramatic induction of apoptosis in ES cells lacking E-cadherin
when grown to high densities, or in epithelial cells such as HC11 or MDCK following inhibition
of cadherin engagement.
Preliminary data indicate that cadherin-11 inhibition also results in apoptosis of Balb/c
3T3 fibroblasts. Cadherin-11 is aberrantly expressed in cancer cells of epithelial lineage with a
more invasive phenotype (Pishvaian et al., 1999; Feltes et al., 2002). It was also recently shown
that although cadherin-11 is not expressed in normal human prostate epithelial cells, it is detected
in prostate cancer, with its expression increasing from primary to metastatic disease to the bone, a
tissue where cadherin-11 is abundantly expressed. In the same report it was shown that cadherin11 knockdown with shRNA in PC3, prostate carcinoma cells, which are derived from a bone
metastasis, decreased their metastatic ability to the bone following intracardiac injection in mice.
These findings suggest that cadherin-11 promotes metastasis of prostate cancer cells specifically
to the bone, and it has been proposed that cadherin-11 inhibition may inhibit bone metastasis
(Chu et al., 2008). Our observations, combined with the fact that cadherin-11 has been shown to
promote metastasis, may point to Stat3 as a central survival, rather than metastasis, factor. Most
importantly, cadherin-11 inhibition could induce apoptosis (through Stat3 downregulation) in
metastatic cells specifically, while normal cells expressing E-cadherin would be spared.
Our findings further demonstrate that this cell survival signal may be mediated by an
increase in Rac1 and Cdc42 activity, leading to Stat3 stimulation. Such a schema could offer an
explanation for our previous data showing that Stat3 inhibition under conditions of extensive cell
to cell adhesion can induce apoptosis, even in normal cells (Anagnostopoulou et al., 2006).
Apoptosis could be mediated by p53 downregulation, through direct binding of Stat3 to the p53
promotor (Niu et al., 2005). The above findings illustrate the fact that cadherin function is
modulated by the cellular context.
138
The relevance of cell interactions observed in densely growing cultured epithelial cells to
that of normal breast is clearly demonstrated by our findings. We observed high levels of Stat3
activity in the luminal epithelial cells of normal mouse breast duct lobule, where E-cadherin is
engaged, thus revealing a distinct association of the signalling of these molecular markers with
that of HC11 cells growing at high but not low densities. Furthermore, the relevance of cell
interactions observed in densely growing, cultured cells to human cancer is emphasized by recent
findings demonstrating a close correspondence of genes expressed specifically in lymph node
carcinoma of the prostate (LNCaP) cells cultured to high densities, with genes associated with
prostate cancer in vivo (Chen et al., 2006). The fact that these genes were different from genes
identified in LNCaP cells grown under log-phase conditions further underscores the importance
of the examination of signalling pathways in densely growing cells. The dramatic activation of
Stat3 by cell density may also explain previous discrepancies regarding activation of Stat3 by the
Rho GTPases in cultured cells (Faruqi et al., 2001; Simon et al., 2000; Debidda et al., 2005). In
any event, disruption of cadherin engagement inhibits Stat3 activity and induces apoptosis, which
may indicate that the role of Stat3 in densely growing cells may be to counteract apoptotic death.
At least two different cadherins can perform this function by activating two distinct Rho family
GTPases in an apparently redundant fashion in at least two mammalian species, a finding which
could have important therapeutic implications.
139
References
Aktories, K. (1997). Bacterial toxins that target Rho proteins. J. Clin. Invest. 99, 827-829.
Anagnostopoulou, A., Vultur, A., Arulanandam, R., Cao, J., Turkson, J., Jove, R., Kim, J. S.,
Glenn, M., Hamilton, A. D., and Raptis, L. (2006). Differential effects of Stat3 inhibition in
sparse vs confluent normal and breast cancer cells. Cancer Lett. 242, 120-132.
Ball, R. K., Friis, R. R., Schoenenberger, C. A., Doppler, W., and Groner, B. (1988). Prolactin
regulation of beta-casein gene expression and of a cytosolic 120-kd protein in a cloned mouse
mammary epithelial cell line. EMBO J. 7, 2089-2095.
Boggon, T. J., Murray, J., Chappuis-Flament, S., Wong, E., Gumbiner, B. M., and Shapiro, L.
(2002). C-cadherin ectodomain structure and implications for cell adhesion mechanisms. Science
296, 1308-1313.
Bromberg, J. F., Wrzeszczynska, M. H., Devgan, G., Zhao, Y., Pestell, R. G., Albanese, C., and
Darnell, J. E., Jr. (1999). Stat3 as an oncogene. Cell 98, 295-303.
Chen, Q., Watson, J. T., Marengo, S. R., Decker, K. S., Coleman, I., Nelson, P. S., and Sikes, R.
A. (2006). Gene expression in the LNCaP human prostate cancer progression model: progression
associated expression in vitro corresponds to expression changes associated with prostate cancer
progression in vivo. Cancer Lett. 244, 274-288.
Chu, K., Cheng, C. J., Ye, X., Lee, Y. C., Zurita, A. J., Chen, D. T., Yu-Lee, L. Y., Zhang, S.,
Yeh, E. T., Hu, M. C., Logothetis, C. J., and Lin, S. H. (2008). Cadherin-11 promotes the
metastasis of prostate cancer cells to bone. Mol. Cancer Res. 6, 1259-1267.
Danielson, K. G., Oborn, C. J., Durban, E. M., Butel, J. S., and Medina, D. (1984). Epithelial
mouse mammary cell line exhibiting normal morphogenesis in vivo and functional differentiation
in vitro. Proc. Natl. Acad. Sci. U. S. A 81, 3756-3760.
Debidda, M., Wang, L., Zang, H., Poli, V., and Zheng, Y. (2005). A role of STAT3 in Rho
GTPase-regulated cell migration and proliferation. J. Biol. Chem. 280, 17275-17285.
Epling-Burnette, P. K., Liu, J. H., Catlett-Falcone, R., Turkson, J., Oshiro, M., Kothapalli, R., Li,
Y., Wang, J. M., Yang-Yen, H. F., Karras, J., Jove, R., and Loughran, T. P., Jr. (2001). Inhibition
of STAT3 signaling leads to apoptosis of leukemic large granular lymphocytes and decreased
Mcl-1 expression. J. Clin. Invest. 107, 351-362.
Etienne-Manneville, S. and Hall, A. (2002). Rho GTPases in cell biology. Nature 420, 629-635.
Faruqi, T. R., Gomez, D., Bustelo, X. R., Bar-Sagi, D., and Reich, N. C. (2001). Rac1 mediates
STAT3 activation by autocrine IL-6. Proc. Nat. Acad. Sci. USA 98, 9014-9019.
Feltes, C. M., Kudo, A., Blaschuk, O., and Byers, S. W. (2002). An alternatively spliced
cadherin-11 enhances human breast cancer cell invasion. Cancer Res. 62, 6688-6697.
140
Fischer, P. and Hilfiker-Kleiner, D. (2007). Survival pathways in hypertrophy and heart failure:
the gp130-STAT axis. Basic Res. Cardiol. 102, 393-411.
Fischer, P. and Hilfiker-Kleiner, D. (2008). Role of gp130-mediated signalling pathways in the
heart and its impact on potential therapeutic aspects. Br. J. Pharmacol. 153 Suppl 1, S414-S427.
Fukuyama, T., Ogita, H., Kawakatsu, T., Inagaki, M., and Takai, Y. (2006). Activation of Rac by
cadherin through the c-Src-Rap1-phosphatidylinositol 3-kinase-Vav2 pathway. Oncogene 25, 819.
Gauzzi, M. C., Velazquez, L., McKendry, R., Mogensen, K. E., Fellous, M., and Pellegrini, S.
(1996). Interferon-alpha-dependent activation of Tyk2 requires phosphorylation of positive
regulatory tyrosines by another kinase. J. Biol. Chem. 271, 20494-20500.
Germain, D. and Frank, D. A. (2007). Targeting the cytoplasmic and nuclear functions of signal
transducers and activators of transcription 3 for cancer therapy. Clin. Cancer Res. 13, 5665-5669.
Gottardi, C. J., Wong, E., and Gumbiner, B. M. (2001). E-cadherin suppresses cellular
transformation by inhibiting beta-catenin signaling in an adhesion-independent manner. J. Cell
Biol. 153, 1049-1060.
Grandis, J. R., Drenning, S. D., Zeng, Q., Watkins, S. C., Melhem, M. F., Endo, S., Johnson, D.
E., Huang, L., He, Y., and Kim, J. D. (2000). Constitutive activation of Stat3 signaling abrogates
apoptosis in squamous cell carcinogenesis in vivo. Proc. Nat. Acad. Sci. USA 97, 4227-4232.
Gumbiner, B. M. (2005). Regulation of cadherin-mediated adhesion in morphogenesis. Nat. Rev.
Mol. Cell Biol. 6, 622-634.
Halbleib, J. M. and Nelson, W. J. (2006). Cadherins in development: cell adhesion, sorting, and
tissue morphogenesis. Genes Dev. 20, 3199-3214.
Hoffmann, I. and Balling, R. (1995). Cloning and expression analysis of a novel mesodermally
expressed cadherin. Dev. Biol. 169, 337-346.
Jaffer, Z. M. and Chernoff, J. (2004). The cross Rho'ds of cell-cell adhesion. J. Biol. Chem. 279,
35123-35126.
Jeanes, A., Gottardi, C. J., and Yap, A. S. (2008). Cadherins and cancer: how does cadherin
dysfunction promote tumor progression? Oncogene 27, 6920-6929.
Kim, S. H., Li, Z., and Sacks, D. B. (2000). E-cadherin-mediated cell-cell attachment activates
Cdc42. J. Biol. Chem. 275, 36999-37005.
Kimball, E. S., Bohn, W. H., Cockley, K. D., Warren, T. C., and Sherwin, S. A. (1984). Distinct
high-performance liquid chromatography pattern of transforming growth factor activity in urine
of cancer patients as compared with that of normal individuals. Cancer Res. 44, 3613-3619.
Kovacs, E. M., Ali, R. G., McCormack, A. J., and Yap, A. S. (2002). E-cadherin homophilic
ligation directly signals through Rac and phosphatidylinositol 3-kinase to regulate adhesive
contacts. J. Biol. Chem. 277, 6708-6718.
141
Kreis, S., Munz, G. A., Haan, S., Heinrich, P. C., and Behrmann, I. (2007). Cell density
dependent increase of constitutive signal transducers and activators of transcription 3 activity in
melanoma cells is mediated by Janus kinases. Mol. Cancer Res. 5, 1331-1341.
Larue, L., Antos, C., Butz, S., Huber, O., Delmas, V., Dominis, M., and Kemler, R. (1996). A
role for cadherins in tissue formation. Development 122, 3185-3194.
Leonard, W. J. and O'Shea, J. J. (1998). Jaks and STATs: biological implications. Annu. Rev.
Immunol. 16, 293-322.
Liu, W. F., Nelson, C. M., Pirone, D. M., and Chen, C. S. (2006). E-cadherin engagement
stimulates proliferation via Rac1. J. Cell Biol. 173, 431-441.
Lynch, E. A., Stall, J., Schmidt, G., Chavrier, P., and D'Souza-Schorey, C. (2006). Proteasomemediated degradation of Rac1-GTP during epithelial cell scattering. Mol. Biol. Cell 17, 22362242.
Makagiansar, I. T., Avery, M., Hu, Y., Audus, K. L., and Siahaan, T. J. (2001). Improving the
selectivity of HAV-peptides in modulating E-cadherin-E-cadherin interactions in the intercellular
junction of MDCK cell monolayers. Pharm. Res. 18, 446-453.
McLemore, M. L., Grewal, S., Liu, F., Archambault, A., Poursine-Laurent, J., Haug, J., and Link,
D. C. (2001). STAT-3 activation is required for normal G-CSF-dependent proliferation and
granulocytic differentiation. Immunity. 14, 193-204.
Medina,D. and Kittrell,F.S. (2000). Establishment of mouse mammary cell lines. In Methods in
mammary gland Biology, M.M.Ip and Asch B.B., eds. Kluwer Academic/Plenum Publishers,
New York), pp. 137-145.
Merlo, G. R., Basolo, F., Fiore, L., Duboc, L., and Hynes, N. E. (1995). p53-dependent and p53independent activation of apoptosis in mammary epithelial cells reveals a survival function of
EGF and insulin. J. Cell Biol. 128, 1185-1196.
Nakagawa, M., Fukata, M., Yamaga, M., Itoh, N., and Kaibuchi, K. (2001). Recruitment and
activation of Rac1 by the formation of E-cadherin-mediated cell-cell adhesion sites. J. Cell Sci.
114, 1829-1838.
Narimatsu, M., Maeda, H., Itoh, S., Atsumi, T., Ohtani, T., Nishida, K., Itoh, M., Kamimura, D.,
Park, S. J., Mizuno, K., Miyazaki, J., Hibi, M., Ishihara, K., Nakajima, K., and Hirano, T. (2001).
Tissue-specific autoregulation of the stat3 gene and its role in interleukin-6-induced survival
signals in T cells. Mol. Cell. Biol. 21, 6615-6625.
Niu, G., Wright, K. L., Ma, Y., Wright, G. M., Huang, M., Irby, R., Briggs, J., Karras, J., Cress,
W. D., Pardoll, D., Jove, R., Chen, J., and Yu, H. (2005). Role of Stat3 in regulating p53
expression and function. Mol. Cell Biol. 25, 7432-7440.
Noren, N. K., Niessen, C. M., Gumbiner, B. M., and Burridge, K. (2001). Cadherin engagement
regulates Rho family GTPases. J. Biol. Chem. 276, 33305-33308.
142
Okazaki, M., Takeshita, S., Kawai, S., Kikuno, R., Tsujimura, A., Kudo, A., and Amann, E.
(1994). Molecular cloning and characterization of OB-cadherin, a new member of cadherin
family expressed in osteoblasts. J. Biol. Chem. 269, 12092-12098.
Onishi, A., Chen, Q., Humtsoe, J. O., and Kramer, R. H. (2008). STAT3 signaling is induced by
intercellular adhesion in squamous cell carcinoma cells. Exp. Cell Res. 314, 377-386.
Orlandini, M. and Oliviero, S. (2001). In fibroblasts Vegf-D expression is induced by cell-cell
contact mediated by cadherin-11. J. Biol. Chem. 276, 6576-6581.
Perrais, M., Chen, X., Perez-Moreno, M., and Gumbiner, B. M. (2007). E-Cadherin Homophilic
Ligation Inhibits Cell Growth and Epidermal Growth Factor Receptor Signaling Independently of
Other Cell Interactions. Mol. Biol. Cell 18, 2013-2025.
Perret, E., Benoliel, A. M., Nassoy, P., Pierres, A., Delmas, V., Thiery, J. P., Bongrand, P., and
Feracci, H. (2002). Fast dissociation kinetics between individual E-cadherin fragments revealed
by flow chamber analysis. EMBO J. 21, 2537-2546.
Pishvaian, M. J., Feltes, C. M., Thompson, P., Bussemakers, M. J., Schalken, J. A., and Byers, S.
W. (1999). Cadherin-11 is expressed in invasive breast cancer cell lines. Cancer Res. 59, 947952.
Reddy, P., Liu, L., Ren, C., Lindgren, P., Boman, K., Shen, Y., Lundin, E., Ottander, U., Rytinki,
M., and Liu, K. (2005). Formation of E-cadherin-mediated cell-cell adhesion activates AKT and
mitogen activated protein kinase via phosphatidylinositol 3 kinase and ligand-independent
activation of epidermal growth factor receptor in ovarian cancer cells. Mol. Endocrinol. 19, 25642578.
Siddiquee, K., Zhang, S., Guida, W. C., Blaskovich, M. A., Greedy, B., Lawrence, H. R., Yip, M.
L., Jove, R., McLaughlin, M. M., Lawrence, N. J., Sebti, S. M., and Turkson, J. (2007). Selective
chemical probe inhibitor of Stat3, identified through structure-based virtual screening, induces
antitumor activity. Proc. Natl. Acad. Sci. U. S. A 104, 7391-7396.
Simon, A. R., Vikis, H. G., Stewart, S., Fanburg, B. L., Cochran, B. H., and Guan, K. L. (2000).
Regulation of STAT3 by direct binding to the Rac1 GTPase. Science 290, 144-147.
Steinman, R. A., Wentzel, A., Lu, Y., Stehle, C., and Grandis, J. R. (2003). Activation of Stat3 by
cell confluence reveals negative regulation of Stat3 by cdk2. Oncogene 22, 3608-3615.
Su, H. W., Yeh, H. H., Wang, S. W., Shen, M. R., Chen, T. L., Kiela, P. R., Ghishan, F. K., and
Tang, M. J. (2007). Cell confluence-induced activation of signal transducer and activator of
transcription-3 (Stat3) triggers epithelial dome formation via augmentation of sodium hydrogen
exchanger-3 (NHE3) expression. J. Biol. Chem. 282, 9883-9894.
Sundfeldt, K. (2003). Cell-cell adhesion in the normal ovary and ovarian tumors of epithelial
origin; an exception to the rule. Mol. Cell Endocrinol. 202, 89-96.
Vultur, A., Arulanandam, R., Turkson, J., Niu, G., Jove, R., and Raptis, L. (2005). Stat3 is
required for full neoplastic transformation by the Simian Virus 40 Large Tumor antigen.
Molecular Biology of the Cell 16, 3832-3846.
143
Vultur, A., Cao, J., Arulanandam, R., Turkson, J., Jove, R., Greer, P., Craig, A., Elliott, B. E., and
Raptis, L. (2004). Cell to cell adhesion modulates Stat3 activity in normal and breast carcinoma
cells. Oncogene 23, 2600-2616.
Wojcik, E. J., Sharifpoor, S., Miller, N. A., Wright, T. G., Watering, R., Tremblay, E. A., Swan,
K., Mueller, C. R., and Elliott, B. E. (2006). A novel activating function of c-Src and Stat3 on
HGF transcription in mammary carcinoma cells. Oncogene 25, 2773-2784.
Yu, C. L., Meyer, D. J., Campbell, G. S., Larner, A. C., Carter-Su, C., Schwartz, J., and Jove, R.
(1995). Enhanced DNA-binding activity of a Stat3-related protein in cells transformed by the Src
oncoprotein. Science 269, 81-83.
Yu, H. and Jove, R. (2004). The STATs of cancer--new molecular targets come of age. Nat. Rev.
Cancer 4, 97-105.
Zhang, Y., Turkson, J., Carter-Su, C., Smithgall, T., Levitzki, A., Kraker, A., Krolewski, J. J.,
Medveczky, P., and Jove, R. (2000). Activation of Stat3 in v-Src transformed fibroblasts requires
cooperation of Jak1 kinase activity. J. Biol. Chem. 275, 24935
24944.
144
Chapter 5
Activated Rac1 requires gp130 for Stat3 activation,
cell proliferation and migration
These data are included in the following manuscripts:
Arulanandam, R., Geletu, M., Feracci, H., and Raptis, L. Activated Rac1 required gp130 for Stat3
activation, cell proliferation and migration. (2009). Experimental Cell Research. Oct 21. In press.
R. Arulanandam, M. Geletu, S. Chevalier, H. Feracci, and L. Raptis, Cadherin-11 engagement
activates Stat3 and promotes survival of mouse fibroblasts. In preparation.
Author contributions:
RA generated all the cell lines and performed the bulk of the experiments, MG conducted some
of the migration assays and growth curve analysis, HF provided insightful advice and was
involved in discussing results and revising the manuscript. Special thanks to Dr. Graham Côté
and Emilia Furmaniak-Kazmierczak for the Rac1 and Cdc42 vectors and Dr. Harvey Ozer and
Douglas Gray for the ts20 cell line.
145
Abstract
Rac1 (Rac) is a member of the Rho family of small GTPases which controls cell
migration by regulating the organisation of actin filaments. Previous results suggested that
mutationally activated forms of the Rho GTPases can activate Stat3, but the exact mechanism is a
matter of controversy. We recently demonstrated that Stat3 activity of cultured cells increases
dramatically following E-cadherin or cadherin-11 engagement in epithelial cells or fibroblasts,
respectively. To better understand this pathway, we now compared Stat3 activity levels in mouse
HC11, 10T1/2 or Balb/c 3T3 cells before and after expression of the mutationally activated Rac1
(RacV12), at different cell densities. The results revealed for the first time a dramatic increase in
protein levels and activity of both the endogenous Rac and RacV12 with cell density, which was
due to inhibition of proteasomal degradation. In addition, RacV12-expressing cells had higher
Stat3, tyrosine-705 phosphorylation and activity levels at all densities, indicating that RacV12 is
able to activate Stat3. Further examination of the mechanism of Stat3 activation showed that
RacV12 expression caused a surge in mRNA of IL6 family cytokines, known potent Stat3
activators. Knockdown of gp130, the common subunit of this family reduced Stat3 activity,
indicating that these cytokines may be responsible for the Stat3 activation by RacV12. The
upregulation of IL6 family cytokines was required for cell migration and proliferation induced by
RacV12, as shown by gp130 knockdown experiments, thus demonstrating that the gp130/Stat3
axis represents an essential effector of activated Rac for the regulation of key cellular functions.
146
Introduction
Rho family GTPases are intracellular molecular switches cycling between the active,
GTP-bound state and the inactive, GDP-bound form (Etienne-Manneville and Hall, 2002). Upon
activation, the Rho GTPases can activate a distinct panel of effectors to regulate cellular
functions. In fact, the Rho family members RhoA, Rac1 and Cdc42 are best known as master
regulators of the actin cytoskeleton, promoting the formation of stress fibers, lamellipodia or
filopodia, respectively. In addition, Rho GTPases are known growth stimulators that act by
modulating key cell cycle regulators, such as cyclin D1 and the transcription factor NFκB, at the
transcriptional level (Joyce et al., 1999; Perona et al., 1997).
STATs were discovered as latent cytoplasmic transcription factors that are activated by a
number of cytokines and growth factors. Among seven mammalian STAT genes identified, Stat3
is found to be overexpressed in a number of tumor cell lines and carcinomas (Frank, 2007). The
fact that a constitutively active form of Stat3 alone is sufficient to transform cultured fibroblasts
to anchorage-independence and tumorigenicity points to an etiological role for Stat3 in neoplasia
(Bromberg et al., 1999). Like other STAT proteins, Stat3 is latent in the cytoplasm in an
unstimulated cell. Following ligand engagement and receptor phosphorylation, Stat3 binds the
activated receptor through its Src homology 2 (SH2) domain and is activated through
phosphorylation by the receptor itself or by the associated JAKs or Src family kinases.
Phosphorylation at a critical tyr-705 activates Stat3 by stabilizing the association of two
monomers through reciprocal SH2-phosphotyrosine interactions. The Stat3 dimer then migrates
to the nucleus where it binds to target sequences, leading to the transcriptional activation of
specific genes, such as myc, bcl-xL, cyclin D, survivin, hgf (Hung and Elliott, 2001) and to the
downregulation of the p53 antioncogene (Niu et al., 2005; Yu and Jove, 2004).
147
Previous results suggested a functional link between the Rho GTPases and Stat3 but the
mechanism is still unclear. Indeed, while in one study it was reported that mutationally activated
Rac1 can directly interact with Stat3 in co-immunoprecipitation assays (Simon et al., 2000), other
data showed that Rac1 indirectly activates Stat3 through autocrine induction of IL6 (Faruqi et al.,
2001), while another group reported that the Rho GTPases can activate Stat3 independent from
IL6 action (Debidda et al., 2005). We and others recently demonstrated a dramatic increase in the
activity of Stat3 triggered by cell to cell contact in a variety of cell lines (Kreis et al., 2007;
Onishi et al., 2008; Steinman et al., 2003; Su et al., 2007; Vultur et al., 2004). For this reason, the
modulation of Stat3-ptyr705 levels by cell density must be taken into account in experiments
assessing the effect of proto-oncogenes such as the Rho GTPases upon Stat3 function.
In this communication, we revisited the question of the mechanism of Stat3 activation by
Rac1 and Cdc42, in light of the above findings. The results indicate that cell density alone causes
a dramatic increase in protein levels and activity of both the endogenous cRac1 (Rac) and Cdc42,
and the mutationally activated Rac1 (RacV12) and Cdc42V12 in mouse epithelial cells and
fibroblasts, through inhibition of proteasomal degradation. Furthermore, lines expressing
activated RacV12 had higher Stat3 activity levels at all cell densities examined, indicating that
RacV12 is, in fact, able to activate Stat3. The Stat3 increase was mediated through the
expression of IL6 family cytokines, as shown through knockdown of gp130, common subunit of
the family. Gp130 function and Stat3 activation were required for cell migration and increase in
proliferation induced by mutationally activated Rac and Cdc42, as shown by genetic knockdown
experiments, thus demonstrating that the gp130/Stat3 axis represents an essential effector of
activated Rac in the regulation of both of these essential cellular functions.
148
Materials and Methods
Cell lines, culture techniques and gene expression
The normal mouse mammary epithelial line HC11 is a prolactin-responsive cell clone
originally isolated from the COMMA-1D mouse mammary epithelial cell line derived from a
female Balb/c mouse in mid-gestation (Ball et al., 1988). Normal mouse Balb/c 3T3 and mouse
10T1/2 fibroblasts have been previously described (Vultur et al., 2004). The ts20, Balb/c3T3
A31-derived line was grown as described (Chowdary et al., 1994). Cell confluence was
estimated visually and quantitated by imaging analysis of live cells under phase contrast using a
Leitz Diaplan microscope and the MCID-elite software (Imaging Research, St. Catharine’s Ont.).
For NFκB inhibition, cells were treated with 20 µM IκB kinase (IKK)-inhibitorIII (BMS345541) or 20 µg/ml caffeic acid phenethyl ester (CAPE, EMD Biosciences). JAK-inhibitor-1
(EMD Biosciences) or MG132 (Sigma) were added at the indicated concentrations. Treatment
with the JAK-inhibitor-1 was for 24 hrs and with MG132 for 8 hrs. The CPA7, platinum Stat3
inhibitor was prepared as described (Littlefield et al., 2008) and used at a 50μM concentration.
Cell viability was assessed by trypan blue exclusion and by replating cells in medium lacking the
inhibitors. IL6 was purchased from Invitrogen.
For gp130 knockdown, a mouse pSM2 retroviral target gene shRNA set (Cat#.
RMM4530-NM_010560, Appendix, Table A.4) was purchased from Open Biosystems. V2MM70734 was the most efficient. Infected cells were selected for puromycin resistance. RacV12 was
expressed with a retroviral vector (a gift of Dr John Collard (Sander et al., 1998)). RacL61,
Cdc42V12 and wtRac1 were expressed through plasmid transfection under control of the
cytomegalovirus (CMV) promotor (plasmids were a gift of Dr. Graham Côté, Queen’s
University). Transfections for wtRac1 and RacL61 expression were performed with Lipofectamine
Plus (Invitrogen). To effectively compare the consequences of wtRac1 vs RacL61 expression upon
149
Stat3 activity, HC11 cells were transfected with each of their respective plasmids and the next
day plated at 3x106 cells/3 cm petri, equivalent to 2 days post confluence, as before (Arulanandam
et al., 2009). 24 hrs later, cell extracts were prepared and Western blots probed with the indicated
antibodies.
For neutralisation of IL6 the #ab6672 antibody (Abcam) was used, while for
neutralisation of LIF we used the antibody #L9152 (Sigma), at concentrations of 0.425 μg/ml and
1 ng/ml, respectively, according to the manufacturers’ protocols.
Western blotting and immunoprecipitation
Detergent cell extracts were prepared as described (Vultur et al., 2004). Following a
careful protein determination (BCA-1 Protein assay kit, Sigma), 30 or 100 μg of clarified cell
extract were submitted to SDS-PAGE, as indicated. Blots were cut into strips and probed with
antibodies specific for the tyr-705 phosphorylated Stat3 (Cell Signalling), total Stat3 (Cell
Signalling), the dually phosphorylated form of Erk1/2 (Biosource), survivin (Cell Signalling), p21
(Biosource), gp130 (Sigma) or Hsp90 (Stressgen) followed by alkaline phosphatase, or HRPconjugated secondary antibodies (Jackson Labs). Rac1 and Cdc42 antibodies (BD Transduction
Labs) recognised both the endogenous and mutant Rac or Cdc42, respectively. To examine the
degree of Rac ubiquitination, extracts were immunoprecipitated with anti-ubiquitin antibodies
(Biomol) and blotted against Rac1 or myc-tag (for RacL61, antibody 9E10, Sigma). In all cases,
bands were visualized using enhanced chemiluminescence (PerkinElmer Life Sciences), or
SuperSignal West Femto Maximum Sensitivity Substrate (Pierce). Quantitation was achieved by
fluorimager analysis using the FluorChem program (AlphaInnotech Corp). In all cases Stat3ptyr705 band intensities were normalized to Hsp90 levels of the same samples. Jak1
phosphorylation was examined by immunoprecipitation against total Jak1, followed by blotting
with a Jak-ptyr1022/1023 antibody (Cell Signalling).
150
Rac/Cdc42 activation assays were performed using GST-PAK-PBD pull-down assays
with the Rac/Cdc42 activation assay kit (Cytoskeleton, #BK035). Briefly, the beads were coated
with GST fused to the binding domain of PAK (PAK-PBD) in pulldown assays. Adding twice
the amount of PBD-coated beads did not increase the signal, indicating that the amount of binding
partner used in the detection was not limiting.
Photoshop (Adobe) or Corel Draw software were used for the organization of non
adjusted, original images and blots.
qRT-PCR assays
For quantitative RT-PCR, the delta ct (Δct) value was calculated from the given ct value
by the formula: Δct = (ctsample – ctcontrol). For the qRT-PCR cytokine array, we used the PAMM021A kit (SA Biosciences) with an RT-PCR for IL6 run in parallel, according to the
manufacturer’s protocol.
151
Results
Cell density inhibits cRac1 proteasomal degradation
We have demonstrated that cell density causes a dramatic increase in the cellular Rac1
(Rac) and Cdc42 protein levels in HC11 normal mouse breast epithelial cells, or Balb/c 3T3 and
10T1/2 normal mouse fibroblasts, through homophilic engagement of E-cadherin or cadherin-11,
respectively (Arulanandam et al., 2009; Arulanandam et al., in preparation). To explore the
possibility that the increase in Rac/Cdc42 with cell density observed in HC11 cells (Figs. 5.1A
and B, lanes 1-7) might be due to inhibition of proteasome-mediated degradation, we at first
made use of the MG132 proteasome inhibitor (Tsubuki et al., 1996). As shown in Fig. 5.2A,
MG132 treatment of sparsely growing HC11 cells caused a substantial increase in Rac protein
levels, as well as the p21CIP/WAF, p53 substrate serving as a positive control (Zhu et al., 2007).
This observation points to a role for the proteasome in Rac degradation. At the same time, the
levels of phosphorylated, i.e. activated Erk1/2 remained unchanged following MG132 treatment,
suggesting that Rac may be selectively degraded by the proteasome. As shown in Fig. 4.10 and
Appendix, Fig. A.10, cell density caused a similar increase in Rac/Cdc42 protein levels in Balb/c
3T3 and 10T1/2 cells, while MG132 treatment of sparsely growing Balb/c 3T3 resulted in high
Rac levels compared to untreated cells (Fig. 5.3A).
To further demonstrate the involvement of ubiquitination, we took advantage of the ts20,
Balb/c 3T3 A31-derived line. Due to a mutation in the gene for the ubiquitin activating enzyme
E1, which makes it susceptible to accelerated destruction, this line is defective in protein
ubiquitination at high temperatures (39oC), while ubiquitination is normal at 34oC (Chowdary et
al., 1994). To examine the effect of E1 inactivation upon Rac levels, ts20 cells were grown to
different densities at 34o or 39oC, along with the parental Balb/c 3T3 A31, and Rac levels
examined. The results revealed that in cells grown to low densities (0.5x105 cells/3 cm petri)
152
Figure 5.1
RacV12 and Cdc42V12 activate Stat3 in HC11 cells at all densities
A: Both endogenous and mutant RacV12 protein levels and activity are dramatically
increased by cell density, while RacV12 expression increases Stat3-ptyr705 levels at all cell
densities
HC11 cells (lanes 1-7) or their counterparts stably expressing activated RacV12 (lanes 814) were grown to increasing densities, up to 11 days post-confluence. Detergent cell lysates
were probed for total Rac, active Rac-GTP, Stat3-ptyr705, total Stat3 or Hsp90 as a loading
control, as indicated (see Materials and Methods). Numbers at the left refer to molecular weight
markers.
B: Both endogenous and mutant Cdc42V12 protein levels and activity are dramatically
increased by cell density, while Cdc42V12 expression increases Stat3-ptyr705 levels at all cell
densities
HC11 cells (lanes 1-7) or their counterparts stably expressing activated Cdc42V12 (lanes
8-14) were grown to increasing densities, up to 11 days post-confluence. Detergent cell lysates
153
were probed for total Cdc42, active Cdc42-GTP, Stat3-ptyr705, total Stat3 or Hsp90 as a loading
control, as indicated.
C: RacV12 and Cdc42V12-expression increases Stat3 transcriptional activity
HC11 and RacV12 (top panel)- and Cdc42 V12 (bottom panel)-expressing cells were
transfected with the Stat3-dependent pLucTKS3 and the Stat3-independent pRLSRE reporters
and grown to the indicated densities. Firefly (■) or Renilla ( ) luciferase activities were
determined in detergent cell extracts (Vultur et al., 2004). Values shown represent luciferase
units expressed as a percentage of the highest value obtained, means ± s.e.m. of at least 3
experiments, each performed in triplicate.
D: RacL61 is more effective than wtRac1 at increasing Stat3-705 phosphorylation.
HC11 cells (lane 1), or their counterparts transfected with wtRac1 (lane 2) or RacL61 (lane
3) were plated at 3 x 106 cells/3cm plate (see Materials and Methods) and detergent cell lysates
probed for total Rac, active Rac-GTP, Stat3-ptyr705, total Stat3 or Hsp90 as a loading control, as
indicated. Numbers at the left refer to molecular weight markers.
under permissive conditions, that is in ts20 cells grown at 34oC (Fig. 5.3B, lane 1) or in Balb/c
3T3 cells at either temperature (lanes 7 and 9), levels of Rac were low. In contrast, ts20 cells
grown to the same low densities at 39oC had substantially higher Rac levels (Fig. 5.3B, lanes 1 vs
4), pointing to the possibility of a ubiquitination-inhibition effect. The requirement for a
continuous inhibition of ubiquitination for the maintenance of high Rac levels was examined next
(Fig. 5.3B, lanes 11-14). ts20 cells were grown at the permissive temperature, to 100%
confluence at first, at which time Rac levels were high, as expected (lane 13). Trypsinizing and
replating them at a low density caused a dramatic decrease in Rac (lane 14). However, replating
ts20 cells at 39oC did not cause a decrease in Rac levels, possibly due to low ubiquitination
activity because of the destruction of the temperature-sensitive, E1 ligase at this temperature (lane
12). This treatment did not affect Rac levels in the parental Balb/c 3T3 cells (not shown). The
above data taken together demonstrate that inhibition of ubiquitin ligase E1 may play an
important role in the dramatic increase in Rac levels observed at high cell densities.
We next examined the effect of ubiquitin overexpression upon Rac protein levels, by
transfecting HC11 cells with a plasmid consisting of a CMV promotor driving the transcription of
an mRNA encoding a multimeric precursor molecule composed of eight, N-terminal hexa
154
Figure 5.2
Cell density inhibits the proteasomal degradation of Rac in HC11 cells
A: The proteasome inhibitor, MG132 increases total Rac and Stat3-ptyr705 levels.
Sparsely growing HC11 cells were treated with 10μM of the proteasome inhibitor,
MG132 (lane 2), or buffer alone (lane 1) for 8 hrs. Detergent cell extracts were probed for total
Rac, Stat3-ptyr705, total Stat3, p21CIP/WAF, pErk1/2 or Hsp90 as a loading control, as indicated.
B: Ubiquitin overexpression leads to Rac degradation.
HC11 cells were transfected with a plasmid consisting of a CMV promotor driving the
transcription of an mRNA encoding eight ubiquitin units with a his6 tag (His-Ubq, lanes 1 and 3)
and grown to 30% (lanes 1 and 2) or 100% (lanes 3 and 4) confluence. Extracts were probed for
Rac, Stat3-ptyr705, total Stat3, pErk1/2 or Hsp90, as indicated. Numbers under the lanes refer to
total Rac or Stat3-ptyr705 band intensities obtained through quantitation by fluorimager analysis
155
and normalized to Hsp90 levels, with the peak value of the control, HC11 cells grown to
confluence (lane 4) taken as 100% (see Materials and Methods).
C-E: Rac is ubiquitinated in vivo, at low cell densities. HC11 or H-RacV12 cells were grown to
30% or 100% confluence and anti-ubiquitin immunoprecipitates of detergent cell extracts (lanes
5-9) blotted against Rac1. As a control, extracts from cells grown to 30% confluence were
immunoprecipitated with normal rabbit serum (lanes 3-4) or buffer alone (lanes 1-2).
Bracket points to the ubiquitinated Rac.
D: Extracts from HC11 or H-RacV12 cells grown to 30% or 100% confluence were blotted
against Rac1.
E: HC11 cells before (lane 1) or after (lanes 2 and 3) expression of a myc-tagged, L61 mutant
were grown to 30% or 100% confluence and anti-ubiquitin immunoprecipitates of detergent cell
extracts blotted against the myc-tag. H.C., L.C.: IgG heavy and light chains, respectively.
Bracket points to the ubiquitinated Rac.
-histidine tagged ubiquitin units (Treier et al., 1994). The results revealed that ubiquitin
overexpression caused a reduction in Rac total protein levels which was more pronounced at
lower cell densities (Fig. 5.2B).
To further investigate whether Rac itself might be a substrate of the proteasome in
sparsely growing cells, we examined whether Rac is modified by ubiquitin tagging. To this
effect, we searched for the presence of Rac in the pool of ubiquitinated proteins, by probing antiubiquitin immunoprecipitates from HC11 cells for Rac by Western blotting. As shown in Fig.
5.2C (left panel), a diffuse band of ubiquitin-tagged Rac, consistent with short chain
polyubiquitination (Lynch et al., 2006), was detected in immunoprecipitates from HC11 cells
(Fig. 5.2, lane 5) grown to 30% confluence (lane 5). This band was very weak at 100%
confluence (Fig, 5.2C, lane 6), possibly due to inhibition of ubiquitination at high cell densities.
When the anti-ubiquitin antibody was omitted (Fig. 5.2C, lane 1) or replaced with normal mouse
IgG (Fig, 5.2C, lane 3) or an unrelated, control mouse monoclonal antibody (not shown), no Rac
was found in the immunoprecipitates from 30% confluent cultures. Similar results were obtained
in Balb/c 3T3 (Fig. 5.3C) and 10T1/2 cells (not shown). The above data taken together indicate
that Rac itself is, in fact, a substrate of the proteasome in sparsely growing cultures, and that cell
confluence inhibits Rac ubiquitination and proteasomal degradation.
156
Figure 5.3
Cell density inhibits the proteasomal degradation of Rac in Balb/c 3T3 fibroblasts
A: The proteasome inhibitor, MG132 increases total Rac and Stat3-ptyr705 levels.
A: Sparsely growing Balb/c 3T3 cells were treated with 10μM of the proteasome inhibitor,
MG132 (lane 2), or not (lane 1) for 8 hrs. Detergent cell extracts were probed for Stat3-ptyr705,
total Rac, or α-tubulin as a loading control, as indicated.
B: Inhibition of the E1 ubiquitin activating enzyme increases Rac protein levels.
Left panel: ts20 (lanes 1-6) or Balb/c 3T3 (lanes 7-10) cells were grown to different densities, at
34oC (lanes 1-3 and 7-8) or 39oC (lanes 4-6 and 9-10), as indicated. Detergent cell extracts were
157
probed for Stat3-ptyr705, Rac or Hsp90 as a loading control. Note the high levels of Stat3ptyr705 at a low confluence, at 39oC (lane 4 vs lane 1).
Right panel: ts20 cells were grown to a confluence of 100% at 39oC (lane 11) or 34oC (lane 13).
Cells were subsequently trypsinized and replated at a low confluence at 39 oC (lane 12) or 34 oC
(lane 14). Detergent cell extracts were probed for Stat3-ptyr705, Rac or Hsp90 as a loading
control.
Note the dramatic reduction of both Stat3-ptyr705 and Rac levels upon replating sparsely
at 34oC (lane 14), which was abrogated when the E1 enzyme was inhibited by replating at 39oC
(lane 12).
C: Rac1 is ubiquitinated in vivo. Balb/c 3T3 cells were grown to 30% or 100% confluence and
anti-ubiquitin immunoprecipitates of detergent cell extracts (lanes 3 and 4) blotted against Rac
(bracket). Lanes 5 and 6: As a control, extracts were immunoprecipitated with normal rabbit
serum and probed for Rac in a similar manner. H.C.: Heavy chain.
Cell density upregulates activated RacV12 levels through inhibition of proteasomal
degradation
It was previously shown that the active, GTP-bound form of Rac is preferentially
degraded by the proteasome (Pop et al., 2004). Therefore, to examine the impact of cell density
upon the levels of mutationally activated RacV12 or RacL61, the latter labeled with a myc-tag, these
genes were stably expressed in HC11 cells (termed H-RacV12 or H-RacL61, see Materials and
Methods). A number of colonies were picked and, since cell density affects Rac protein levels,
screening for total Rac or myc-Rac by Western blotting was performed at 40% confluence. Cells
were plated in 3 cm dishes and when 30% confluent and at different times thereafter, Rac protein
levels were evaluated by Western blotting. As shown in Fig. 5.1A (lanes 8-14), the levels of
RacV12 increased dramatically with cell density. Interestingly, H-RacV12 cells had ~3x higher total
Rac levels than the parental HC11, at all densities examined. Two other RacV12 and three RacL61expressing clones gave the same results (not shown). Similar results were found in 10T1/2 lines
stably expressing RacV12 (see Appendix, Fig. A.11, lanes 7-12). The above data, taken together,
indicate that, similar to cRac1, activated RacV12 and RacL61 are also subject to density-dependent
upregulation.
158
The effect of cell density upon RacV12 activity was examined next. Cells were plated in
Petri dishes and when ~30% confluent, and over several days thereafter, Rac activity was
measured by assessing the binding between Rac-GTP and its effector p21-activated kinase (PAK)
in cell extracts using pull-down assays (see Materials and Methods). As shown in Fig. 5.1, there
was a dramatic increase in Rac activity with cell density, in both the parental HC11 (Fig. 5.1,
lanes 1-7), and in RacV12-expressing cells (Fig. 5.1, lanes 8-14), at all densities examined, in
parallel with Rac protein levels. 10T-RacV12 (Appendix, Fig. A.11) and clones expressing the
mutationally activated, Cdc42V12 gave similar results (Fig. 5.1B and Appendix, Fig. A.11). The
above findings taken together indicate that cell density, besides an increase in total Rac/Cdc42
protein levels, it also causes a dramatic increase in both Rac/Cdc42 and RacV12/Cdc42V12 activity.
It follows that, cell density has to be taken into account when the effect of RacV12 upon Stat3 is
examined.
To better control the degree of cell-cell contact, irrespective of differences in cell growth
patterns, we repeated the experiments by plating different numbers of HC11 and or H-RacV12
cells (0.4x106, 0.7x106, 1.0x106, 1.3x106, 1.8x106 or 2.5x106 per 3cm petri, respectively), so that
they would reach the same densities as above within 24 hours (Vultur et al., 2004). At that time,
cells were lysed and Rac protein levels and activity analyzed as above. In all cases, very similar
results were obtained, indicating that it is the extent of cell-cell contact, regardless of time in
culture beyond 24 hrs that is responsible for the increase in Rac protein and activity (Appendix,
Fig. A.7 and not shown).
The potential role of ubiquitination upon the levels of the mutationally activated, RacV12
was examined next. As for Rac in HC11 cells (Fig 5.2C, lanes 5 and 6), anti-ubiquitin
immunoprecipitates of H-RacV12 cells grown to different densities were probed with an anti-Rac1
antibody. As shown in Fig. 5.2C (lane 8), a diffuse band of ubiquitin-tagged Rac (Lynch et al.,
2006), was detected in immunoprecipitates from cells grown to 30% confluence. Cells grown to
100% confluence on the other hand, had very low levels of ubiquitin-tagged Rac (lane 9),
159
consistent with inhibition of ubiquitination at high cell densities. When the anti-ubiquitin
antibody was omitted (lane 2) or replaced with normal mouse IgG (lane 4), no RacV12 was
detected in the immunoprecipitates. At the same time, total Rac protein levels were much higher
at high densities, both in the HC11 and H-RacV12 cell lines (Fig. 5.2D). Taken together, these
data strongly suggest that RacV12 itself is also a substrate of the proteasome, in sparsely growing
cells exclusively. To ensure that it is the mutationally activated Rac which is ubiquitinated, we
made use of the myc-tagged, RacL61. As shown in Fig. 5.2E, blotting anti-ubiquitin
immunoprecipitates from H-RacL61 cells, expressing a myc-tagged RacL61, grown to 30%
confluence, with an anti-myc-tag antibody, yielded a strong band (lane 2), while at 100%
confluence no myc-tagged band was detected (lane 3). Similar results were obtained with 10T½
and 10T-Rac V12 (not shown). The above data taken together indicate that cell density can cause a
dramatic increase in both Rac and RacV12 protein levels, through inhibition of proteasomal
degradation.
Mutationally activated RacV12 and Cdc42V12 can activate Stat3
Previous results demonstrated that RacV12 expression leads to the activation of Stat3.
However, the effect of cell density was not taken into account in any previous report (Debidda et
al., 2005; Faruqi et al., 2001; Simon et al., 2000), and this could be a source of apparently
conflicting results. To definitively demonstrate the effect of RacV12 upon Stat3 activity, H-RacV12
cells were plated in 3 cm dishes and at different times thereafter Stat3-ptyr705 levels measured by
western blotting and compared to the parental HC11 cells, respectively. As shown in Fig. 5.1A
(lanes 8-14), H-RacV12 cells had higher Stat3-ptyr705 levels at all cell densities examined.
Similar results were obtained with H-RacL61 and 10T- RacV12 (Appendix, Fig. A.11), while
overexpression of the cellular Rac1 in HC11 cells (wtRac1, Fig. 5.1D) had a substantially
reduced effect upon Stat3-ptyr705 levels, compared to RacV12 or RacL61. Cdc42V12 mirrored the
effect of RacV12 regarding Stat3 activation when expressed in both HC11 and 10T1/2 cell lines
160
(Fig. 5.1B, lanes 8-14 and Appendix, Fig. A.11, lanes 8-14). Examination of the levels of total
Stat3 protein revealed a modest increase with cell density, as previously reported (Vultur et al.,
2004), and with RacV12, RacL61 or Cdc42V12 expression compared to parental HC11 (Fig. 5.1, A,
B, D) or 10T½ lines (not shown), possibly due to the fact that the Stat3 promotor itself is one of
the Stat3 targets (Narimatsu et al., 2001), although the differences were not as pronounced as the
Stat3-tyr705 phosphorylation observed. Stat3 transcriptional activity also increased upon RacV12
or Cdc42V12 expression (Fig. 5.1C), following the pattern of Stat3-tyr705 phosphorylation.
The effect of Rac ubiquitination upon Stat3-ptyr705 levels was examined next. As
shown in Fig. 5.2A and Fig. 5.3A, treatment with MG132, which increased Rac levels, also
increased the levels of Stat3-ptyr705, while expression of his-ubiquitin, which reduced Rac
levels, had a proportional effect upon Stat3-ptyr705 (Fig. 5.2B). Stat3-ptyr705 levels were also
increased in subconfluent ts20 cells at the higher, nonpermissive temperature (Fig. 5.3B, lanes 1
vs 4 or lane 14 vs 12).
The above results taken together indicate that mutationally activated forms of both Rac1
and Cdc42 Rho family GTPases are indeed able to activate Stat3 in epithelial cells and
fibroblasts, above and beyond the activation due to cell density.
NFκB and JAKs are required for the RacV12-mediated, Stat3 activation
Early data showed that Rac1 can activate NFκB (Sulciner et al., 1996). To examine
whether NFκB may be actually required for the RacV12-mediated Stat3 activation, sparsely
growing HC11 cells were trypsinized and plated to a high density (2.5x106 cells/3 cm petri).
Following attachment, cells were treated with the IKK-inhibitor-III (BMS-345541) or the DMSO
carrier alone, for 48 hours. As shown in Fig. 5.4A, treatment with the inhibitor caused a dramatic
reduction in Stat3-ptyr705 levels, in both the parental HC11 (lanes 3 vs 2), or H-RacV12 (lanes 5
vs 4). These data indicate that both the cell density-mediated, and RacV12-mediated increase in
Stat3-tyr705 phosphorylation requires NFκB. On the other hand, Rac levels were unaffected by
161
Figure 5.4
RacV12-dependent Stat3-ptyr705 phosphorylation requires JAK and NFκB
A: The IKK inhibitor III inhibits the density-mediated, Stat3, tyr705 phosphorylation in
both HC11 and H-RacV12 cells. HC11 (lanes 1-3) or RacV12 (lanes 4 and 5) cells were
trypsinised and plated at a high density (2.5x106 cells/3 cm plate). Following attachment 30 min
later, cells were treated with the IKK inhibitor III (lanes 3 and 5), or the DMSO carrier (lanes 2
and 4) for 48 hrs. Cell extracts were probed for Stat3-ptyr705, total Stat3, Rac or Hsp90, as a
loading control. Lane 1: HC11 cells immediately after attachment to the plastic.
B: RacV12-expression increases Jak1-1022/1023 phosphorylation.
Lysates from HC11 (lanes 1 and 3) or H-RacV12 (lanes 2 and 4) cells were grown to 30%
confluence (lanes 1 and 2), or one day after confluence (lanes 3 and 4) and probed for pJak1022/1023, total Jak1 or Hsp90 as a loading control, as indicated.
C: JAK inhibitor 1 reduces the RacV12-dependent, Stat3-tyr705 phosphorylation.
HC11 (lanes 1-3) or RacV12 (lanes 4-6) cells were grown to a high density and treated
with 0 (lanes 1 and 4), 10 (lanes 2 and 5) or 2 (lanes 3 and 6) μg/ml JAK inhibitor 1. Cell lysates
were probed for Stat3-ptyr705, total Stat3, Rac or Hsp90 as a loading control, as indicated.
162
NFκB inhibition (Fig. 5.4A), which further reinforces the conclusion that NFκB is downstream of
Rac, ie it is required for the Rac-mediated, Stat3 activation exclusively. Similar results were
obtained with caffeic acid phenethyl ester (CAPE), another known inhibitor of NFκB activity (not
shown) and in parental Balb/c 3T3 (Appendix, Fig. A.12A).
To explore the role of Jak1 in the RacV12-mediated, Stat3 activation, we at first
investigated whether Jak1 is activated by RacV12, by examining Jak1 phosphorylation at
tyr1022/1023, shown to be important in the regulation of Jak1 activity (Gauzzi et al., 1996;
Leonard and O'Shea, 1998). Jak1 phosphorylation at 1022/1023 was measured by blotting lysates
from H-RacV12 and HC11 cells with a phospho-specific antibody (see Materials and Methods).
As shown in Fig. 5.4B, there was a ~2 fold increase in Jak1, 1022/1023 phosphorylation upon
RacV12 expression at one day after 100% confluence (lanes 3 vs 4), which paralleled the
phosphorylation of Stat3 at this density (Fig. 5.4A, lanes 2 vs 4), while at 30% confluence the
difference was more pronounced (Fig. 5.4B, lane 1 vs 2). To examine whether Jak1 is actually
required for the RacV12-mediated increase in Stat3 activity, HC11 cells grown to densities of 1
day post confluence were treated with the pan-JAK inhibitor, JAK inhibitor 1. The results
showed a dramatic reduction in Stat3-ptyr705 levels, which essentially plateaued at 2μM of
inhibitor, consistent with previous findings (Arulanandam et al., 2009), while Rac levels
remained unaffected (Fig. 5.4C, lanes 1-6). Similar results were obtained with the AG490 JAK
inhibitor and in parental Balb/c 3T3 cells [(Zhang et al., 2000), and Appendix, Fig. A.12B].
These findings suggest that the JAK kinases are required for the RacV12-mediated increase in
Stat3-ptyr705 levels.
RacV12 triggers cytokine gene expression and requires gp130 for Stat3 activation
To explore the possibility that the RacV12-induced, Stat3 activation may be mediated by
autocrine factors, medium conditioned by sparsely growing, H-RacV12 cells was added to the
parental HC11 cells growing to a low density and Stat3-ptyr705 examined. The results revealed a
163
3-fold increase in Stat3-ptyr705, indicating the presence of autocrine factors (not shown). To
explore the nature of the cytokines responsible, we conducted a quantitative RT-PCR array
experiment for 86 cytokines, comparing mRNA levels in H-RacV12 cells with levels in the
parental HC11 line (see Materials and Methods). Given the effect of density upon Stat3 levels, to
examine the effect of RacV12 per se, we compared cells expressing RacV12 with the parental HC11,
while both were grown to densities of 40%. The results revealed an increase in mRNA levels of
two cytokines of the IL6 family, IL6 (18-fold) and LIF (~10-fold), known to act through the
common gp130 subunit, shared by a number of Stat3 activating cytokines [(Hilfiker-Kleiner et
al., 2004), Appendix, Table A.5]. In fact, addition of neutralizing antibodies against two
members of the family, IL6 and LIF, separately reduced Stat3-ptyr705 levels by approximately
30%, while a combination of the two caused a reduction of ~50% (Fig. 5.5A, top panel). To
further demonstrate the requirement for IL6 family cytokines for the RacV12-mediated, Stat3
activation, the levels of gp130 were reduced through stable expression of a specific shRNA, using
a retroviral vector (see Materials and Methods). As shown in Fig. 5.5A, gp130 knockdown
reduced Stat3-ptyr705 levels in H-RacV12 cells (line H-RacV12-shgp130) at all densities examined
(lanes 1-3 vs 4-6), pointing to the possibility that the two IL6 family cytokines play an important
role in the activation of Stat3 by RacV12. Stat3-tyr705 phosphorylation was also reduced in 10TRacV12 cells following shgp130 expression (not shown).
RacV12-induced IL6 upregulation is unable to activate Erk1/2 in confluent cultures
Results from a number of labs indicated that, in addition to Stat3, IL6 is able to activate
the Erk1/2 kinase [Erk (Fischer and Hilfiker-Kleiner, 2008)]. Since RacV12 expression leads to
IL6 upregulation, we investigated whether RacV12 can activate Erk. H-RacV12 and HC11 cells
were grown to 30% confluence, serum-starved for 24 hrs and levels of the dually phosphorylated,
i.e. activated Erk1/2 (pErk) examined by Western immunoblotting. As shown in Fig. 5.5B, pErk
levels were higher in H-RacV12 cells when grown to 30% confluence than the parental HC11 (lane
164
Figure 5.5
gp130 is required for the RacV12-mediated, Stat3 activation
A: Top: Reduction in Stat3-ptyr705 levels by neutralising antibodies to IL6 and LIF. HRacV12 cells were grown to 30% confluence (lane 1) or one day post confluence and an antibody
to IL6 (lane 3) or to LIF (lane 4), or a combination of the two (lane 6) added to the growth
medium for 6 hrs. Detergent cell extracts were blotted for Stat3-ptyr705, total Stat3 or Hsp90 as a
loading control. Lanes 2 and 5: Control lanes, cells treated with buffer alone.
Bottom panel: Lysates from H-RacV12 cells before (lanes 1-3) or after (lanes 4-6) expression of a
gp130-specific, shRNA were probed for Stat3-ptyr705, total Stat3 or Hsp90 as a loading control,
as indicated.
165
Right panel: Lysates from H-RacV12 cells, before (lane 1) or after (lane 2) knockdown of gp130
were probed for gp130 or Hsp90 as a loading control, as indicated.
B: RacV12 does not activate Erk at high cell densities.
Left panel: Extracts from the indicated cell lines, before (lanes 1-4) or after (lanes 5-8)
expression of the gp130-specific, shRNA were probed for pErk1/2 or Hsp90 as a loading control,
as indicated.
Right panel: IL6 was added at 10 ng/ml for 15 min to HC11 cells grown to 60% (lane 2) or
100% (lane 4) confluence and cell extracts probed for Erk1/2 or Hsp90 as a loading control, as
indicated.
2 vs 1). At high density however, there was no Erk activity increase upon RacV12 expression
(lanes 4 vs 3), although Stat3 phosphorylation was increased (Fig. 5.1A). To investigate whether
the Erk activation by RacV12 might be due to the IL6 family cytokines produced, pErk levels were
examined in the gp130 knockdown cells. As shown in Fig. 5.5B (lanes 5-8), shgp130-expressing
cells had low pErk levels, even after RacV12 expression (lane 6 vs 2), indicating that IL6 activity is
required for pErk upregulation by RacV12. Similar results were obtained in 10T½ and 10T-Rac
V12
, with or without shgp130. To further verify the effect of IL6 upon Erk, we directly examined
the role of density upon the activation of Erk by purified IL6. HC11 cells were grown to
densities of 60% or 2 days after confluence, stimulated with IL6 for 15 min and Erk activity
examined by probing for pErk (Fig. 5.5B, right panel). The results demonstrated that although
IL6 could activate Erk in subconfluent cultures (lane 2 vs 1), as previously reported (Fischer and
Hilfiker-Kleiner, 2008), there was no Erk activity increase upon IL6 addition in cells grown to
high confluence (lane 4). The above findings clearly indicate that, although RacV12 induces the
expression of IL6, a known Erk activator, IL6 is unable to activate Erk in cells grown to high
densities. This observation explains earlier data indicating that pErk levels are unaffected by
confluence in a number of cellular systems (Vultur et al., 2004), and reveals a profound effect of
cell to cell adhesion upon the response of HC11 cells to IL6.
166
RacV12-induced cell proliferation and migration require IL6 family cytokines
The Rho family GTPases are known to play a role in the control of cell proliferation (Guo
and Zheng, 2004). To examine the effect of IL6 family cytokines upon the RacV12-induced cell
growth, HC11, H-shgp130, H-RacV12 and H-RacV12- shgp130 cells were plated in 3 cm dishes and
their growth rate determined. As shown in Fig. 5.6A, shgp130 expression caused a ~2.2 fold
reduction in the growth rate of both H-RacV12 cells and the parental HC11, indicating that gp130
is an essential component of a pathway leading to Rac-induced cell proliferation. Similar results
were observed when comparing 10T1/2, 10T-shgp130, 10T-RacV12 and 10T-RacV12-shgp130 cell
lines (not shown).
It is well established that increased Rac activity can increase cell motility through the
formation of lamellipodia at the leading edge of cells in a wound healing assay (Nobes and Hall,
1999). To examine the role of the gp130 subunit in the Rac-induced cell migration, HC11, HRacV12, H-shgp130 and H-RacV12-shgp130 cells were plated in plastic petri dishes. At two days
post-confluence, a scratch-wound was introduced to the monolayer using a plastic pipette tip, and
the cells allowed to migrate into the gap. As shown in Figs. 5.6B and C, RacV12-expressing cells
were able to move faster to the open wound than the parental HC11 (panel f vs b). However,
gp130 downregulation abolished this effect (panel h), indicating that gp130 is an integral
component of the RacV12 pathway leading to the increase in cell migration. gp130 shRNA
expression in the parental HC11 line also decreased cell migration (panel d). Very similar
observations were noted when comparing 10T1/2, 10T-shgp130, 10T-RacV12 and 10T-RacV12shgp130 cell lines (not shown).
Although RacV12 cells grow faster than cells expressing both RacV12 and shgp130, the
accelerated growth rate cannot account for the increase in rate of migration and gap closure.
Downregulation of Stat3 through shRNA knockdown (Geletu et al., 2009), or treatment with the
CPA7 inhibitor (Littlefield et al., 2008) caused a similar decrease in cell migration of H-RacV12
cells (not shown), as previously reported in other systems (Debidda et al., 2005). These findings
167
Figure 5.6
Rac-induced cell migration and proliferation require gp130
A: Cell proliferation: HC11, H-shgp130, H-RacV12, or H-RacV12-shgp130 cells were grown in
Petri dishes in 10% serum and cell numbers obtained over several days, as indicated. Values
represent averages of 3 independent experiments.
B: Cell migration: HC11, HC11-shgp130, H-RacV12 or H-RacV12-shgp130 cells were cultured to
confluence before a scratch was made through the monolayer using a plastic pipette tip. Cells
were photographed at 0 (panels a, c, e, g) or 24 hrs (panels b, d, f, h) after 12 hrs of culturing in
0.5% fetal calf serum.
C: Quantitation of the time (hrs) required by the different cell lines for gap junction
closure.
HC11, HC11-shgp130, H-RacV12 or H-RacV12-shgp130 cells were cultured to confluence
before a scratch was made through the monolayer using a plastic pipette tip. The time taken for
gap junction closure was determined by microscopic observation. Numbers represent averages of
3 independent experiments.
168
indicate that gp130 is required for cell migration triggered by the cellular Rac, and that expression
of the mutationally activated RacV12 cannot overcome the gp130 knockdown. Taken together,
these results demonstrate that gp130 represents an essential effector of activated Rac in the
regulation of both cell proliferation and migration.
169
Discussion
We previously demonstrated that cell to cell adhesion induces a dramatic increase in the
levels of Stat3 activity in a number of cell lines [(Arulanandam et al., 2009; Vultur et al., 2004;
Vultur et al., 2005); reviewed in (Raptis et al., 2009)]. An important piece of information that
came out of these studies is that cell density must be taken into account in experiments to assess
the effect of an oncogene upon Stat3 activity. Therefore, to investigate the effect of activated Rac
upon Stat3, Stat3 activity levels were examined at different cell densities, following expression of
two mutationally activated Rac constructs, RacV12 and RacL61. The results revealed that Stat3
activity was increased with activated Rac expression at all cell densities examined, indicating that
activated Rac can, in fact, activate Stat3, above and beyond the activation due to cell density.
These observations raise important questions: 1. What is the mechanism of increase in
Rac and RacV12 levels with cell density? 2. What is the mechanism of Stat3 activation by
RacV12? 3. What is the functional significance of the Stat3 activation in RacV12-expressing cells?
Cell density inhibits the proteasomal degradation of Rac.
We have demonstrated that cell density also increases Rac levels in both HC11 epithelial
cells and Balb/c 3T3 or 10T1/2 fibroblasts (Arulanandam et al., 2009; Arulanandam et al., in
preparation). Our data indicate that this increase could be due to inhibition of Rac proteasomal
degradation with confluence. In fact, previous results indicated that Rac can be degraded through
the proteasome pathway (Pop et al., 2004). A mutational analysis further indicated that
constitutive activation of Rac, as well as binding of effectors, which might be acting as ubiquitin
E3 ligases, are necessary for Rac degradation (Pop et al., 2004). Our results, demonstrating that,
although activated, RacV12 and RacL61 protein levels are increased dramatically with cell density,
indicate that cell density can overcome the degradative effect of activation. These findings
170
extend and reinforce previous data indicating that epithelial cell scattering brought about by HGF
can induce the proteasome-mediated degradation of Rac1 (Lynch et al., 2006). A similar
mechanism could hold true for Cdc42, which mirrored Rac levels and stability increases with cell
density.
IL6 family cytokines transmit the signal from activated RacV12 to Stat3.
Although earlier reports (Simon et al., 2000) indicated that Rac can directly interact with
Stat3 in co-expression/co-immunoprecipitation assays, later findings unequivocally demonstrated
that Stat3 activation by the Rho GTPases can occur without the formation of a stable complex
[(Debidda et al., 2005); and data not shown]. In addition, Faruqi et al. have demonstrated that
RacV12 could indirectly activate Stat3 through autocrine induction of IL6 (Faruqi et al., 2001).
However, through the use of specific antibodies added to the medium it was later demonstrated
that Stat3 activation can also occur independently of IL6 stimulation (Debidda et al., 2005). To
resolve this apparent controversy, we examined mRNA levels for 86 cytokines in an array
analysis. The results revealed an increase in two cytokines of the IL6 family (IL6 and LIF),
indicating that this family may be involved in Stat3 activation by mutationally activated Rac.
Furthermore, downregulation of gp130, the common subunit of the family, abolished the RacV12mediated, Stat3 activation indicating that this subunit is actually required for Stat3 activation.
The fact that the LIF also, rather than IL6 alone, is upregulated by RacV12 explains earlier data
indicating that IL6-specific antibodies did not reduce Stat3 activity (Debidda et al., 2005), since
these antibodies, directed against IL6 would inhibit IL6 exclusively, while LIF would still be free
to activate Stat3. In any event, these results demonstrate that the total Stat3 activity in the cell is
the sum of effects of both cell to cell adhesion, plus the Stat3 activating, RacV12 or Cdc42V12
oncogene present.
Previous results indicated that IL6 activates Erk, as well as Stat3. Since RacV12 causes the
upregulation of IL6 family cytokines, it is expected that RacV12 expression would activate Erk.
171
However, although RacV12-expressing cells had higher pErk levels than the parental HC11, when
sparsely grown, there was no pErk increase in H-RacV12 cells grown to high densities. To solve
this apparent paradox, HC11 cells were grown to different densities and stimulated with IL6.
Indeed IL6 itself, although able to activate Erk in subconfluent cultures, was unable to do so once
cells reached confluence (Fig. 5.5B), although Stat3 was still activated (Fig. 5.1). It is possible
that Erk-specific phosphatases such as Cdc25A are activated at high cell densities (Lazo et al.,
2002), or that other adaptors required for Erk activation by IL6, or phosphorylation of IL6-R
sites, are downregulated following cadherin engagement and establishment of cell to cell
contacts. In any event, these results further demonstrate that, despite the fact that these two
pathways are often both activated by oncogenes, cytokines or growth factors, they are not
coordinately regulated by IL6 in densely growing cells.
Gp130 is required for Rac-mediated cell proliferation and migration.
It is well established that Rac activation can increase cell motility, in part due to
regulation of the actin cytoskeleton. However, the role of the IL6/Stat3 axis in this effect has not
been examined. Our results showed that, as observed before (Nobes and Hall, 1999), RacV12expressing cells were able to move faster into an open wound introduced with a plastic pipette tip
into a cell monolayer than the parental line (Etienne-Manneville and Hall, 2002). Furthermore,
the Rac-mediated motility required the activity of IL6 family cytokines, since downregulation of
the gp130, common subunit of the family, reduced the rate of migration. The fact that activated
RacV12 upregulates IL6 and Stat3, and that this may occur, at least, in both epithelial cells and
fibroblasts, gives further credence to the model of Stat3 activation by E-cadherin or cadherin-11
engagement, respectively, via Rac and IL6 (Arulanandam et al., 2009).
Conclusions: Our results definitively demonstrate that activated Rac/Cdc42 expression leads to
Stat3 activation, by a mechanism requiring NFκB, gp130 and JAK. This pathway is required for
172
cell proliferation and migration, that is, the gp130/Stat3 axis represents an essential effector of
activated Rac for the regulation of key cellular functions (Fig. 5.7).
Figure 5.7
Proposed model for Rac/Cdc42-mediated, Stat3 activation
Constitutively active Rac/Cdc42 activates Stat3 through induction of IL6 family
cytokines, such as IL6 and LIF which possess the common subunit, gp130. This leads to
activation of their respective receptors, Jak activation and phosphorylation of Stat3, resulting in
cell proliferation and migration.
173
References
Arulanandam, R., Geletu, M., Chevalier, S., Feracci, H., and Raptis, L. (2010). Cadherin-11
engagement activates Stat3 and promotes survival of mouse fibroblasts. Manuscript in
preparation.
Arulanandam, R., Vultur, A., Cao, J., Carefoot, E., Truesdell, P., Elliott, B., Larue, L., Feracci,
H., and Raptis, L. (2009). Cadherin-cadherin engagement promotes survival via Rac/Cdc42 and
Stat3. Molecular Cancer Research 17, 1310-1327.
Ball, R. K., Friis, R. R., Schoenenberger, C. A., Doppler, W., and Groner, B. (1988). Prolactin
regulation of beta-casein gene expression and of a cytosolic 120-kd protein in a cloned mouse
mammary epithelial cell line. EMBO J. 7, 2089-2095.
Bromberg, J. F., Wrzeszczynska, M. H., Devgan, G., Zhao, Y., Pestell, R. G., Albanese, C., and
Darnell, J. E., Jr. (1999). Stat3 as an oncogene. Cell 98, 295-303.
Chowdary, D. R., Dermody, J. J., Jha, K. K., and Ozer, H. L. (1994). Accumulation of p53 in a
mutant cell line defective in the ubiquitin pathway. Mol. Cell Biol. 14, 1997-2003.
Debidda, M., Wang, L., Zang, H., Poli, V., and Zheng, Y. (2005). A role of STAT3 in Rho
GTPase-regulated cell migration and proliferation. J. Biol. Chem. 280, 17275-17285.
Etienne-Manneville, S. and Hall, A. (2002). Rho GTPases in cell biology. Nature 420, 629-635.
Faruqi, T. R., Gomez, D., Bustelo, X. R., Bar-Sagi, D., and Reich, N. C. (2001). Rac1 mediates
STAT3 activation by autocrine IL-6. Proc. Nat. Acad. Sci. USA 98, 9014-9019.
Fischer, P. and Hilfiker-Kleiner, D. (2008). Role of gp130-mediated signalling pathways in the
heart and its impact on potential therapeutic aspects. Br. J. Pharmacol. 153 Suppl 1, S414-S427.
Frank, D. A. (2007). STAT3 as a central mediator of neoplastic cellular transformation. Cancer
Lett. 251, 199-210.
Gauzzi, M. C., Velazquez, L., McKendry, R., Mogensen, K. E., Fellous, M., and Pellegrini, S.
(1996). Interferon-alpha-dependent activation of Tyk2 requires phosphorylation of positive
regulatory tyrosines by another kinase. J. Biol. Chem. 271, 20494-20500.
Geletu, M., Chaize, C., Arulanandam, R., Vultur, A., Kowolik, C., Anagnostopoulou, A., Jove,
R., and Raptis, L. (2009). Stat3 activity is required for gap junctional permeability in normal
epithelial cells and fibroblasts. DNA & Cell Biol. 28, 319-327.
Guo, F. and Zheng, Y. (2004). Involvement of Rho family GTPases in p19Arf- and p53-mediated
proliferation of primary mouse embryonic fibroblasts. Mol. Cell Biol. 24, 1426-1438.
Hilfiker-Kleiner, D., Hilfiker, A., Fuchs, M., Kaminski, K., Schaefer, A., Schieffer, B., Hillmer,
A., Schmiedl, A., Ding, Z., Podewski, E., Podewski, E., Poli, V., Schneider, M. D., Schulz, R.,
Park, J. K., Wollert, K. C., and Drexler, H. (2004). Signal transducer and activator of
174
transcription 3 is required for myocardial capillary growth, control of interstitial matrix
deposition, and heart protection from ischemic injury. Circ. Res. 95, 187-195.
Hung, W. and Elliott, B. (2001). Co-operative effect of c-Src tyrosine kinase and Stat3 in
activation of hepatocyte growth factor expression in mammary carcinoma cells. J. Biol. Chem.
276, 12395-12403.
Joyce, D., Bouzahzah, B., Fu, M., Albanese, C., D'Amico, M., Steer, J., Klein, J. U., Lee, R. J.,
Segall, J. E., Westwick, J. K., Der, C. J., and Pestell, R. G. (1999). Integration of Rac-dependent
regulation of cyclin D1 transcription through a nuclear factor-kappaB-dependent pathway. J. Biol.
Chem. 274, 25245-25249.
Kreis, S., Munz, G. A., Haan, S., Heinrich, P. C., and Behrmann, I. (2007). Cell density
dependent increase of constitutive signal transducers and activators of transcription 3 activity in
melanoma cells is mediated by Janus kinases. Mol. Cancer Res. 5, 1331-1341.
Lazo, J. S., Nemoto, K., Pestell, K. E., Cooley, K., Southwick, E. C., Mitchell, D. A., Furey, W.,
Gussio, R., Zaharevitz, D. W., Joo, B., and Wipf, P. (2002). Identification of a potent and
selective pharmacophore for Cdc25 dual specificity phosphatase inhibitors. Mol. Pharmacol. 61,
720-728.
Leonard, W. J. and O'Shea, J. J. (1998). Jaks and STATs: biological implications. Annu. Rev.
Immunol. 16, 293-322.
Littlefield, S. L., Baird, M. C., Anagnostopoulou, A., and Raptis, L. (2008). Synthesis,
characterization and Stat3 inhibitory properties of the prototypical platinum(IV) anticancer drug,
[PtCl3(NO2)(NH3)2] (CPA-7). Inorg. Chem. 47, 2798-2804.
Lynch, E. A., Stall, J., Schmidt, G., Chavrier, P., and D'Souza-Schorey, C. (2006). Proteasomemediated degradation of Rac1-GTP during epithelial cell scattering. Mol. Biol. Cell 17, 22362242.
Narimatsu, M., Maeda, H., Itoh, S., Atsumi, T., Ohtani, T., Nishida, K., Itoh, M., Kamimura, D.,
Park, S. J., Mizuno, K., Miyazaki, J., Hibi, M., Ishihara, K., Nakajima, K., and Hirano, T. (2001).
Tissue-specific autoregulation of the stat3 gene and its role in interleukin-6-induced survival
signals in T cells. Mol. Cell. Biol. 21, 6615-6625.
Niu, G., Wright, K. L., Ma, Y., Wright, G. M., Huang, M., Irby, R., Briggs, J., Karras, J., Cress,
W. D., Pardoll, D., Jove, R., Chen, J., and Yu, H. (2005). Role of Stat3 in regulating p53
expression and function. Mol. Cell Biol. 25, 7432-7440.
Nobes, C. D. and Hall, A. (1999). Rho GTPases control polarity, protrusion, and adhesion during
cell movement. J. Cell Biol. 144, 1235-1244.
Onishi, A., Chen, Q., Humtsoe, J. O., and Kramer, R. H. (2008). STAT3 signaling is induced by
intercellular adhesion in squamous cell carcinoma cells. Exp. Cell Res. 314, 377-386.
Perona, R., Montaner, S., Saniger, L., Sanchez-Perez, I., Bravo, R., and Lacal, J. C. (1997).
Activation of the nuclear factor-kappaB by Rho, CDC42, and Rac-1 proteins. Genes Dev. 11,
463-475.
175
Pop, M., Aktories, K., and Schmidt, G. (2004). Isotype-specific degradation of Rac activated by
the cytotoxic necrotizing factor 1. J. Biol. Chem. 279, 35840-35848.
Raptis, L., Arulanandam, R., Vultur, A., Geletu, M., Chevalier, S., and Feracci, H. (2009).
Beyond structure, to survival: Stat3 activation by cadherin engagement. Biochemistry and Cell
Biology in press.
Sander, E. E., van Delft, S., ten Klooster, J. P., Reid, T., van der Kammen, R. A., Michiels, F.,
and Collard, J. G. (1998). Matrix-dependent Tiam1/Rac signaling in epithelial cells promotes
either cell-cell adhesion or cell migration and is regulated by phosphatidylinositol 3-kinase. J.
Cell Biol. 143, 1385-1398.
Simon, A. R., Vikis, H. G., Stewart, S., Fanburg, B. L., Cochran, B. H., and Guan, K. L. (2000).
Regulation of STAT3 by direct binding to the Rac1 GTPase. Science 290, 144-147.
Steinman, R. A., Wentzel, A., Lu, Y., Stehle, C., and Grandis, J. R. (2003). Activation of Stat3 by
cell confluence reveals negative regulation of Stat3 by cdk2. Oncogene 22, 3608-3615.
Su, H. W., Yeh, H. H., Wang, S. W., Shen, M. R., Chen, T. L., Kiela, P. R., Ghishan, F. K., and
Tang, M. J. (2007). Cell confluence-induced activation of signal transducer and activator of
transcription-3 (Stat3) triggers epithelial dome formation via augmentation of sodium hydrogen
exchanger-3 (NHE3) expression. J. Biol. Chem. 282, 9883-9894.
Sulciner, D. J., Irani, K., Yu, Z. X., Ferrans, V. J., Goldschmidt-Clermont, P., and Finkel, T.
(1996). rac1 regulates a cytokine-stimulated, redox-dependent pathway necessary for NF-kappaB
activation. Mol. Cell Biol. 16, 7115-7121.
Treier, M., Staszewski, L. M., and Bohmann, D. (1994). Ubiquitin-dependent c-Jun degradation
in vivo is mediated by the delta domain. Cell 78, 787-798.
Tsubuki, S., Saito, Y., Tomioka, M., Ito, H., and Kawashima, S. (1996). Differential inhibition of
calpain and proteasome activities by peptidyl aldehydes of di-leucine and tri-leucine. J. Biochem.
119, 572-576.
Vultur, A., Arulanandam, R., Turkson, J., Niu, G., Jove, R., and Raptis, L. (2005). Stat3 is
required for full neoplastic transformation by the Simian Virus 40 Large Tumor antigen.
Molecular Biology of the Cell 16, 3832-3846.
Vultur, A., Cao, J., Arulanandam, R., Turkson, J., Jove, R., Greer, P., Craig, A., Elliott, B. E., and
Raptis, L. (2004). Cell to cell adhesion modulates Stat3 activity in normal and breast carcinoma
cells. Oncogene 23, 2600-2616.
Yu, H. and Jove, R. (2004). The STATs of cancer--new molecular targets come of age. Nat. Rev.
Cancer 4, 97-105.
Zhang, Y., Turkson, J., Carter-Su, C., Smithgall, T., Levitzki, A., Kraker, A., Krolewski, J. J.,
Medveczky, P., and Jove, R. (2000). Activation of Stat3 in v-Src transformed fibroblasts requires
cooperation of Jak1 kinase activity. J. Biol. Chem. 275, 24935-24944.
176
Zhu, Q., Wani, G., Yao, J., Patnaik, S., Wang, Q. E., El Mahdy, M. A., Praetorius-Ibba, M., and
Wani, A. A. (2007). The ubiquitin-proteasome system regulates p53-mediated transcription at
p21waf1 promoter. Oncogene 26, 4199-4208.
177
Chapter 6
Discussion
178
Malignant cancer progression is associated with mutational events that may lead to persistent
mitogenic signalling or loss of negative control, resulting in activation of cell division even in the
absence of stimuli. While Stat3 has recently emerged as an ideal target in anticancer therapy due
to its ability to coordinate proliferative, anti-apoptotic, immunosuppressive and angiogenic
responses triggered by a number of different aberrant molecular signalling events, its mechanism
of action is incompletely understood. The current literature suggests that constitutive signalling
to Stat3 can occur from its persistent phosphorylation by growth factor receptors, cytokine
receptors and cytoplasmic tyrosine kinases. The loss of key negative regulators of Stat3 has also
been observed in a number of tumors and tumor-derived cell lines. Our lab has done pioneering
research in demonstrating a novel role for cell-cell adhesion in the activation of Stat3. Indeed,
these findings indicate that the Stat3 activation observed in confluent cultures may constitute a
potent survival signal which could be exploited by cancer cells. The work presented here,
revealing that homophilic cadherin ligation and increased protein stability of Rac1 and Cdc42
GTPases can increase ptyr705-Stat3 phosphorylation and activity levels, broadens our
understanding of Stat3 signalling while promoting the validation of this pathway for anticancer
drug design.
Direct cadherin engagement activates Stat3
Cells in normal tissues or in tumors have extensive opportunities for adhesion to their
neighbors in a three-dimensional organization, and it recently became apparent that in the study
of fundamental cellular processes, such as mitogenesis, survival or differentiation, it is important
to take into account the effect of surrounding cells. Dense cell cultures, albeit only two
dimensional, may in part mimic some of the physiological stress signals that occur in fastgrowing tumors. Indeed, many of the cells in tumor nests lack direct interaction with surrounding
connective tissue and they may depend on survival signalling generated by cell-cell contacts. Not
179
surprisingly, the world of cells growing at high densities in 2-dimensions can be dramatically
different from the one of cells dividing peacefully at subconfluence.
While both structural and signalling functions have been extensively characterized for the
integrin family of cell-matrix adhesion receptors, which are known to stimulate the activity of the
focal adhesion tyrosine kinase (FAK), our understanding of the signals generated at adherens
junctions is more recent and less well understood. Early data have demonstrated an increase in
total cellular tyrosine phosphorylation in confluent MCF-10A human mammary epithelial
cultures, and more specifically upon E-cadherin engagement, which was reduced following
calcium chelation or treatment with E-cadherin blocking antibodies (Kinch et al., 1997; Batt and
Roberts, 1998). In line with these findings, my work has identified Stat3 as a tyrosinephosphorylated substrate downstream of signals brought about by homophilic ligation of Ecadherin in both murine and canine epithelial cells (Arulanandam et al., 2009b). In fact, our
approach of employing E/EC12 fragments to functionalize plastic petri dishes offered the
possibility for a definitive demonstration that the Stat3 activity increase observed at high densities
is a direct effect of cadherin engagement, rather than a secondary effect of cell to cell proximity.
Interestingly, examination of cell lines which are known to lack E-cadherin, such as the
chinese hamster ovary (CHO) and the cervical carcinoma A431D, also revealed the same increase
in Stat3 activity with density (Appendix, Fig. A.13). Hence, these data led us to investigate
whether other cadherins may also be activating Stat3. In the context of metastasis, loss of Ecadherin is associated with a concomitant gain in expression of the mesenchymal N-cadherin and
cadherin-11. This prompted the examination of the role of cadherins associated with the
metastatic phenotype in activating Stat3. Accordingly, our data reveal that cadherin-11, a well
characterized type II classical cadherin, is also a potent stimulator of Stat3 activity in Balb/c 3T3
and 10T½ fibroblasts (Arulanandam et al., in preparation). My data so far also attribute a role for
N-cadherin in Stat3 activation (Appendix, Fig. A.14). Indeed, comparison of ptyr-705 levels in
ES, E-cad null cells, which express very low levels of known cadherins, including N-cadherin, to
180
null cells where the N-cadherin gene was added back [Ncad-addback cells (Larue et al., 1996)],
reveals a dramatic increase in Stat3 upon N-cadherin expression. Interestingly, independent
studies have identified the short type PB-cadherin, a novel adhesion receptor in the rat testes, as a
direct activator of Jak2 and Stat3 (Wu et al., 2005). Hence, the fact that a number of different
cadherins activate Stat3 in tissues originating from at least three mammalian species emphasizes
the potentially universal nature and significance of this pathway. In addition, our findings add to
the growing body of evidence indicating that cadherins can provide direct, functionally relevant
signalling beyond their structural role.
Cadherin engagement upregulates Rac1 and Cdc42 protein levels through inhibition of
proteasomal degradation
Further examination of the mechanism of Stat3 activation by cadherin engagement in
HC11 cells revealed a role for the Rac1 and Cdc42, Rho family GTPases. While our results
confirm an increase in GTP-bound Rac1 and Cdc42 as cells approach confluence (Noren et al.,
2001; Nakagawa et al., 2001), we further expose an unanticipated rise in Rac1 and Cdc42 total
protein levels, brought about by cadherin ligation. Our results reveal for the first time that the
dramatic rise in the activity of these GTPases is due to an increase in the stability of Rac1/Cdc42,
brought about by an increase in their protein levels, at high confluences. Indeed, Rac1 was
shown to be modified by ubiquitin tagging in the absence of cell to cell contact, thereby leading
to its degradation by the proteasome. Interestingly, the degradation of GTP-bound Rac1 and its
subsequent ubiquitination was also shown in the context of epithelial cell scattering brought about
by HGF (Lynch et al., 2006).
Mutationally activated Rac and Cdc42 are also targeted to the proteasome, in the absence of cellcell contact
In addition, our results demonstrate that the levels of mutationally activated forms of
Rac1 (RacV12 or RacL61) and Cdc42 (Cdc42V12) also increase with cell density, and can be subject
to ubiquitination, which would occur in the absence of cell to cell contact. The ubiquitination of
181
Rac1 has been demonstrated following its deamidation at Gln61, which causes its constitutive
activation by the bacterial toxin, cytotoxic necrotizing factor (CNF) found in Escherichia coli
(Lerm et al., 2002; Pop et al., 2004). Others have reported the proteasomal degradation of
mutationally activated RacV12 following treatment of endothelial cells with reactive oxygen
species (ROS), and have identified a destruction box domain, originally found in mitotic cyclins,
in the primary structure of Rac, which may be a signal for its modification by ubiquitin (Kovacic
et al., 2001; Jacobs et al., 2001). Hence, the present literature implies that ubiquitin tagging of
Rac requires its modification and activation, which leads to effector binding. Studies have shown
that point mutations in the switch I domain, involved in mediating effector interactions, may
prevent the degradation of Rac following CNF treatment (Pop et al., 2004). One promising
candidate is the Rac1 effector, POSH, which, in addition to serving as a scaffolding protein
linking together activated Rac1 with downstream modulators of the JNK cascade, was shown to
also target proteins for degradation via the ubiquitin proteasome pathway (Xu et al., 2003).
Examination of the POSH sequence has revealed a zinc ring finger domain near the N-terminus
with characteristic E3 ligase activity (Xu et al., 2003). POSH was first shown to direct its own
proteasomal degradation, resulting in its autoregulation (Xu et al., 2003). Further studies have
revealed POSH to be involved in the proteasomal targeting of a number of other substrates,
including the HGF-regulated tyrosine kinase substrate, Hrs, an early endosomal resident sorting
factor (Tuvia et al., 2007; Kim et al., 2006; Alroy et al., 2005; Tsuda et al., 2005). Most
interestingly, overexpression of the Rac1-binding domain of POSH in HEK 293 cells following
CNF treatment, or in MDCK epithelial cells after HGF stimulation could prevent Rac1
degradation (Pop et al., 2004; Lynch et al., 2006). However, these results are at variance with
findings by Visvikis et al., who have observed the continuous degradation of RacV12, but not
endogenous Rac1, in HEK 293 cells, but still claim that POSH expression did not promote RacV12
ubiquitination (Visvikis et al., 2008). However, cell density was not taken into account in these
experiments and might explain this apparent discrepancy. Given the above findings, the
182
examination of the role of POSH in the ubiquitin modification of Rac1, or activated Rac1
mutants, in the absence of cell-cell adhesion, would be essential to the further characterization of
this GTPase as a binding partner and substrate of this unique effector with E3 ligase activity.
The regulation of other Rho GTPase family members by ubiquitin-mediated proteasomal
degradation is less well understood. However, an E3 ligase responsible for the ubiquitination of
dominant negative, or GDP-bound, RhoA in HEK 293 cells has been identified as the Smadubiquitin regulatory factor (Smurf) (Wang et al., 2006; Zhang et al., 2004). On the other hand,
while the ubiquitin targeting of Cdc42 has not yet been validated, the degradation of Cdc42 was
demonstrated following its cleavage by caspases during Fas ligand-induced apoptosis (Tu and
Cerione, 2001). However, it is interesting to note that activated Cdc42 can form complexes with
the E3 ligases, Cbl and Hakai (Wu et al., 2003; Shen et al., 2008). While Cdc42 was shown to
function in sequestering Cbl, thereby preventing its binding to its substrate, the EGF receptor,
activated Cdc42 could promote the ubiquitination of E-cadherin by Hakai, following calcium
chelation (Wu et al., 2003; Shen et al., 2008). Indeed, E-cadherin harbours a PEST sequence, a
motif associated with ubiquitin targeting, which overlaps the site for β-catenin binding (Huber et
al., 2001). Hence, cadherins that are uncoupled from β-catenin may be endocytosed and
degraded in order to reduce the population of cadherins at the cell surface following calcium
chelation and even in the context of sparsely growing cells (Le et al., 1999; Shen et al., 2008;
Fujita et al., 2002). These findings taken together, further point towards the possibility that
Cdc42, or Rac1, could also be targeted for destruction in a complex with adherens junction
components in the absence of cell-cell contact. In line with these findings, my unpublished data
indicate that E-cadherin, along with Rac1 and Cdc42 total protein levels, are reduced in the
absence of cell-cell contacts (Appendix, Fig. A.15). However, whether Rac1/Cdc42 associates
with inactive E-cadherin to form a complex which may be destroyed by the proteasome remains
to be seen. Nevertheless, our results, demonstrating that, although Cdc42 and Rac1, as well as
183
Cdc42V12, RacV12, RacL61, protein levels are increased dramatically with cell density, strongly
suggest that cell density can overcome the degradative effect of activation.
IL6 family cytokines transmit the signal from activated Rac to Stat3.
Data from a number of labs have demonstrated a functional link between the Stat3 and
Rho family pathways in a variety of cell systems (Faruqi et al., 2001; Debidda et al., 2005;
Turkson et al., 1999; Simon et al., 2000). However, the exact mechanism whereby these
GTPases may affect Stat3 is a topic of some controversy. While a direct association between
Rac1V12 and Stat3 has been proposed (Simon et al., 2000), later findings unequivocally
demonstrated that Stat3 activation by the Rho GTPases can occur without the formation of a
stable complex (Debidda et al., 2005). My unpublished data, demonstrating that Rac1 and Stat3
do not co-immunoprecipitate in confluent cultured cells are in agreement with the latter
observations.
In addition to their role in the regulation of the actin cytoskeleton, Rho GTPases have
been implicated in the modulation of gene expression, mainly due to their ability to activate the
NFκB transcription factor (Perona et al., 1997; Aznar and Lacal, 2001). Further data by Faruqi et
al. demonstrated that RacV12, and Cdc42 V12, could indirectly activate Stat3 through NF-κBmediated autocrine induction of IL6 (Faruqi et al., 2001). However, through the use of blocking
antibodies added to the medium, Debidda et al. demonstrated that Stat3 activation by RacV12 can
also occur independently of IL6 stimulation (Debidda et al., 2005). To resolve this apparent
controversy, we examined H-RacV12 mRNA levels for 86 cytokines in an array analysis. The
results revealed an increase in two cytokines of the IL6 family (IL6 and LIF), indicating that this
family may be involved in Stat3 activation by mutationally activated Rac. In addition, we
demonstrate that inhibition of NF-κB in H-RacV12 cells may prevent Stat3 activation, suggesting
that Rac-NFκB activation may be inducing the expression of IL6 family cytokines. Furthermore,
downregulation of gp130, the common subunit of the family, abolished the RacV12-mediated,
184
Stat3 activation indicating that this subunit is actually required for Stat3 activation. The fact that
the LIF also, rather than IL6 alone, is upregulated by RacV12 explains earlier data indicating that
IL6-specific antibodies did not reduce Stat3 activity (Debidda et al., 2005), since these antibodies,
directed against IL6 would inhibit IL6 exclusively, while LIF would still be free to activate Stat3.
In any event, these results demonstrate that the total Stat3 activity in the cell is the sum of the
effects of both cell to cell adhesion, plus the Stat3 activating, RacV12 or Cdc42V12 oncogene
present.
Rac1 and Cdc42 transduce the cadherin signal to Stat3
Our results reveal that in addition to being activated by Rac1V12 and Cdc42V12, Stat3
activation can also result from the dramatic upregulation of endogenous levels and activity of
these GTPases following direct cadherin ligation. In contrast to Rac and Cdc42, expression of a
dominant-negative RhoA construct did not seem to affect Stat3 activity levels (not shown).
Interestingly, Kinch et al. have observed increased tyrosine phosphorylation of Ras-GAP in
confluent cells (Kinch et al., 1997), while Ras-GAP was shown to associate with Rho-GAP,
which leads to a corresponding decline in Rho activity (Settleman et al., 1992). This could, in
part, explain the reduced Rho-GTP levels previously observed upon E-cadherin engagement
(Noren et al., 2001; Fukata and Kaibuchi, 2001) and the fact that RhoA was found to not play a
role in the density-mediated Stat3 activation in our system.
Cell-cell adhesion induces the autocrine production of IL6 family cytokines
The mechanism whereby Rac1 and Cdc42 may activate Stat3 with cell density is likely
autocrine, since conditioned medium from densely growing cells did increase Stat3 activity in
cells grown to low density. Our data further demonstrate that cell confluence causes the
expression of a number of cytokines in both E-cadherin and cadherin-11 expressing cells, most
importantly members of the IL6 family. Indeed, knockdown of the common gp130 subunit by
185
stable transfection could significantly inhibit the density-mediated Stat3 activation indicating that
this family of cytokines is at least in part responsible for the Stat3 activity increase observed at
high cell densities.
In addition to the IL6 family, our qRT-PCR array analysis in HC11 cells pointed to other
cytokines which employ different pathways from gp130/Stat3. Prominent among them is a tumor
necrosis factor (TNF) family member, the receptor activator of NFκB (RANK) ligand, which was
found to be upregulated ~50-fold in densely growing, compared to sparse HC11 cells. RANK-L
was shown to play a critical role in mammary gland development in pregnancy, by promoting the
proliferation and survival of mammary gland epithelial cells (Fata et al., 2000). The role of
RANK-L in the activation of Stat3 by E-cadherin is the subject of further study.
In both HC11 and Balb/c 3T3, the most significant increase was noted in IL6 mRNA in
confluent vs. sparse cells (76 and 32 fold, respectively). IL6 is the most potent activator of Stat3
and relies heavily on receptor-associated JAK kinases to fulfill this role. Interestingly, recent
findings have revealed that phosphorylation of Stat3 on tyr-705 by JAKs may also depend on its
association with the male germ cell Rac1-GAP (MgcRacGAP). Previous data identified a crucial
role for MgcRacGAP in cytokinesis through its binding to tubulin and Rho GTPases (Hirose et
al., 2001). Conditional knockout of MgcRacGAP can also result in apoptosis even before
cytokinesis in B220+ T cells, indicating that it may also function in promoting cell survival
(Yamada et al., 2008). More recently, MgcRacGAP was revealed to directly associate with a
region in the DNA binding domain of STATs (Kawashima et al., 2006). A classical NLS was
identified in MgcRacGAP, which enables it to act, together with GTP-bound Rac1, as a nuclear
chaperone for phosphorylated STATs by enabling the association of this complex with importinα (Kawashima et al., 2006). In addition, downregulation of MgcRacGAP was also shown to
prevent the tyrosine phosphorylation and activation of Stat3 following its stimulation by IL6 in a
chicken B cell line (Kawashima et al., 2009). The authors indicate that MgcRacGAP can bind
Jak2 and hence, along with Rac1-GTP, may mediate Stat3 phosphorylation by serving as a
186
scaffold for the interaction between Jak2 and Stat3 (Kawashima et al., 2009). Hence, it would be
interesting to study the role of MgcRacGAP in the activation of Rac1 and Stat3 observed in our
cell systems, in the context of cell-cell adhesion.
The autocrine model of Stat3 activation leaves a question unanswered: The IL6 receptor
family invariably activates other pathways, such as Erk1/2 (Fischer and Hilfiker-Kleiner, 2008).
To solve this apparent paradox, HC11 cells were grown to different densities and stimulated with
IL6. Indeed IL6, although able to activate Erk in subconfluent cultures, it was found to be unable
to do so once cells reach confluence, although Stat3 was still activated. This could not have been
due to a general inability of the cells to activate Erk in this setting, since my data indicate that
confluent cells did respond to EGF with Erk activation (not shown). It is possible that IL6associated phosphatases that are specific for Erk are activated at high cell densities, or that other
adaptors required for Erk activation by IL6, or phosphorylation of certain IL6 receptor sites, are
downregulated following cadherin engagement. Our preliminary results indicate a role for the
phosphatase Cdc25A, known for its role as an essential regulator of cell cycle transition in
mammalian cells, by dephosphorylating the Raf kinase (Xia et al., 1999) as well as the cyclindependent kinases, Cdk2 and Cdk4 (Draetta and Eckstein, 1997). Cdc25A was also shown to
bind to, dephosphorylate and inactivate Erk as well, both in purified preparations, and in cultured
Hep3B hepatoma cells (Wang et al., 2005). Indeed, treatment of densely growing, HC11 cells
with the potent, Cdc25A-selective inhibitor, NSC-67212 (Nemoto et al., 2004) caused a dramatic
increase in Erk levels in densely growing HC11 cells (Geletu et al., unpublished data). The fact
that in confluent cells, Stat3 could be immune to inactivation by Cdc25A, and possibly other
phosphatases (not shown), at a time when Erk activity is drastically curtailed, further emphasizes
the signalling specificity of this pathway triggered by cadherin ligation and points towards a
primordial significance. These data reveal that, despite the fact that the Erk and Stat3 pathways
are often both activated by growth factors or oncogenes, they are not coordinately regulated by
cadherin engagement.
187
Calcium phosphate and lipofectamine transfection increase Stat3 activity
The fact that transient transfection regiments alone could activate Stat3 prompted us to
generate stable cell lines for all studies on the effect of Stat3 modulators. While the specific
mechanism of tyr-705 phosphorylation upon plasmid DNA transfection using calcium phosphate
precipitation or certain cationic lipid-based methods has not been fully elucidated, it has been
demonstrated that the gene transfer of nucleic acids, including unmethylated plasmid DNA, or
siRNA, can trigger proinflammatory responses in vitro or in vivo through IL6 production (Yew et
al., 1999; Yoo et al., 2006). Indeed, one of the problems facing non-viral gene therapy for the
correction of genetic deficiencies is the activation of a strong innate immune response. While
inflammatory cytokine production was generally found to be triggered by unmethylated CpG
residues found in plasmid DNA (Yonenaga et al., 2007; Scheule, 2000), a study by Yoo et al. has
demonstrated that certain cationic lipid based transfection reagents alone may increase the
production of IL6, IL8, IL18 and IL1A (Yoo et al., 2006). Inhibition of endosomal maturation by
chloroquine treatment was found to significantly lower IL8, and potentially IL6 expression,
brought about by transfection of siRNA or lipid reagent alone (Yoo et al., 2006). Cytokine
expression following transient transfection was suggested to occur as a result of nucleic acid
binding Toll-like receptor signaling, which could lead to NFκB activation. Interestingly, studies
have revealed that Stat3 can also be targeted to signalling or sequestering endosomes, thereby
adding a new twist to the original paradigm of Stat3 activation by cytokine receptors (Xu et al.,
2007; Sehgal, 2008). Hence, it is possible that the increased Stat3 activity observed following
lipofectamine transfection may occur through endosomal interactions.
We have shown that in addition to cationic lipid-based methods, calcium phosphate
transfection can also trigger Stat3 activation. An additional explanation for our findings could be
that the physical presence of a precipitate in the Petri dish or the increase in calcium
concentration may increase the opportunities for cell-cell adhesion. While my unpublished
results suggest that adding higher amounts of calcium in the presence or absence of DNA did not
188
increase Stat3 tyr-705 phosphorylation, it remains to be seen whether these transfection
techniques may encourage cadherin engagement.
Role of the cadherin/Rac1/gp130/JAK/Stat3 axis in cell survival and cancer
The biological consequences of the cell-cell adhesion mediated Stat3 activation were first
examined in our lab by downregulating Stat3 activity in normal mouse epithelial or fibroblast
cultures using a variety of methods, such as electroporation of peptides or peptidomimetics to
block the Stat3, SH2 domain, or through treatment with two platinum compound inhibitors
(Turkson et al., 2004a; Turkson et al., 2004b). The results demonstrated a profound difference in
the effect of Stat3 inhibition in sparsely growing vs. confluent cells (Vultur et al., 2004;
Anagnostopoulou et al., 2006; Vultur et al., 2005). Although Stat3 levels are very low in normal
mouse NIH 3T3 fibroblasts or human breast epithelial MCF-10A cells grown to a low
confluence, Stat3 inhibition does cause a growth retardation, indicating a role for Stat3 in cell
proliferation in this setting. In sharp contrast, in densely growing, normal cells, Stat3 inhibition
induced apoptosis. In fact, both growth retardation and apoptosis following Stat3 inhibition
appear to be mediated by a rise in p53 levels. Treatment with the CPA-7 inhibitor (Turkson et al.,
2004a; Turkson et al., 2004b) resulted in a dramatic increase in p53 at all cell densities, consistent
with previous findings that Stat3 dowregulates p53 by binding directly to its promotor (Niu et al.,
2005). Interestingly, p21, the substrate responsible for p53-mediated growth retardation, was
elevated at lower densities exclusively (Anagnostopoulou, 2009, thesis). Whether Stat3
inhibition at high densities induces alternate p53 targets associated with apoptosis, such as the
PUMA and NOXA members of the Bcl-2 family (Pietsch et al., 2008), is currently under
investigation.
The results presented demonstrate that direct cadherin engagement can promote cell
survival through Stat3 activation. Indeed, we observed a dramatic induction of apoptosis in cells
lacking E-cadherin when grown to high densities, or in HC11 cells following inhibition of
189
cadherin engagement. Contrary to E-cadherin, the mesenchymal cadherins, cadherin-11 and Ncadherin, both have key functions in promoting the metastatic spread of cancer cells, therefore,
the demonstration that they also activate Stat3 may point to Stat3 as a central survival, rather than
metastasis, factor. Most importantly, cadherin-11 or N-cadherin inhibition could induce
apoptosis (through Stat3 downregulation) in metastatic cells specifically, while normal cells
expressing E-cadherin would be spared.
Our findings demonstrate the presence of constitutively activated Stat3 in normal breast
tissue where cadherin is engaged, thus revealing a distinct correlation of the state of these
molecular markers between cells growing to high, but not low densities with the same type of cell
in vivo. The expression of Stat3 in normal tissues surrounding human breast tumors has been
examined by Diaz et al. (Diaz et al., 2006). Using tissues fixed within less than 15 min after
surgery, to preserve the integrity of Stat3-ptyr705, which is known to be rapidly
dephosphorylated after surgery (Diaz et al., 2006; Mora et al., 2002), they definitively
demonstrated that Stat3-ptyr705 was present in luminal cells, so that overall levels in the nonneoplastic tissues were only two-fold lower than the levels in the tumor itself, as determined by
immunohistochemistry and computer-assisted quantitation (Diaz et al., 2006). Remarkably,
Her2/neu levels were very high in the tumors but were undetectable in the normal tissues
surrounding the breast tumor. Furthermore survivin levels were undetectable in normal tissues
but were increased in the tumors pointing towards a marked difference in Stat3 target gene
expression profiles in normal vs tumor tissues. The presence of phosphorylated Stat3 staining in
normal luminal cells by this group is consistent with our data from the established, mouse HC11
cells, cultured primary breast epithelial cells and breast tissues. These findings by Diaz et al. also
correlate with our observations that Stat3-ptyr705 levels in normal cells are generally two-fold
lower than cells transformed by oncogenes such as vSrc or activated Rac.
Our demonstration of the central role of gp130 in the density-dependent Stat3 activation
explains earlier findings indicating that inhibition of certain tyrosine kinases commonly activated
190
in cancer such as EGFR, IGFR, Fer or the Src family in mouse epithelial cells or fibroblasts do
not play a role (Vultur et al., 2004; Steinman et al., 2003). It is interesting to note that the EGFR
independence could perhaps explain why inhibition of members of this family of receptors (e.g.
ErbB2) has been ineffective in a large number of ErbB2-positive cancers (Hynes and Lane, 2005;
Vogel et al., 2002; Harries and Smith, 2002; Cobleigh et al., 1999). Indeed, emerging evidence
has identified a critical role for IL6 signalling in activating Stat3 and promoting malignancy in a
number of cancers (Catlett-Falcone et al., 1999a; Lou et al., 2000; Berishaj et al., 2007). A study
by Gao et al. has shown that while the presence of nuclear, tyrosine phosphorylated Stat3 was
found in 50% of lung adenocarcinomas, its activation was not directly affected by the EGFR, but
likely triggered by transcriptional induction of IL6 and the activation of its receptor (Gao et al.,
2007). Similar findings were noted in breast cancer cell lines where inhibition of EGFR or Src
had minimal effect on Stat3 compared to pJAK or gp130 downregulation (Berishaj et al., 2007).
These results indicate that targeting the gp130/JAK/Stat3 axis may present a more effective
strategy compared to EGFR or Src inhibition in these cancers.
In addition to autocrine stimulatory loops, paracrine feedforward mechanisms of cytokine
signalling to Stat3 have been demonstrated in certain human cancers and can play a role in
inducing and maintaining a cancer-promoting, inflammatory environment. Hyperactive Stat3 in
tumor cells can lead to the production of cytokines and growth factors that in turn may stimulate
Stat3 signalling in stromal cells (Yu et al., 2009). Hence a feedforward stimulation of Stat3 can
result from the tumor microenvironment. This is exemplified in multiple myeloma, where bone
marrow stromal cells contribute to the production of IL6, thereby fueling cancer progression
through the persistent activation of the IL6/JAK/Stat3 axis in tumor cells and the induction of
antiapoptotic genes (Shain et al., 2009; Catlett-Falcone et al., 1999b). Our demonstration of a
novel pathway of Stat3 signalling through cadherin ligation further suggests that in addition to the
production of cytokines such as IL6, stromal cells may also contribute to Stat3 signalling through
direct cell-cell adhesion, thereby emphasizing the role of cell communication with the
191
extracellular milieu in cancer progression. Indeed, this promotion of Stat3 activation by the
microenvironment could also be crucial to tumor cell survival following the administration of
chemotherapeutic agents, thereby indicating that the cadherin-mediated Stat3 activation could be
a key contributing factor in multidrug resistance (Shain and Dalton, 2001).
Conclusion
Over the past decade, Stat3 has received much attention because its constitutive
activation has shown to promote the growth and survival of tumor cells. The work presented here
furthers our understanding of this transcription factor by identifying three novel pathways leading
to its activation (Fig. 6.1A). The fact that Stat3 tyr-705 phosphorylation, DNA-binding and
transcriptional activation, as well as Rac1 and Cdc42 activity and levels, may increase through
simple cell-cell contact by cadherins, or that ptyr705-Stat3 can result from certain common
transfection techniques further indicates that caution is required upon the interpretation of data
involving these extensively studied signalling proteins.
Pioneering work from our lab revealed that the world of densely growing cells is vastly
different from that of cells growing peacefully at subconfluence. Our data has greatly contributed
to the growing list of proteins which are affected by cell to cell contact; these now include pJak1
(Vultur et al., 2004; Arulanandam et al., 2009b), Rac1/Cdc42 (Arulanandam et al., 2009b),
activated Rac1V12/Rac1L61/Cdc42V12 (Arulanandam et al., 2009a), caveolin-1 (Mohan et al.,
unpublished), cdk2 (Steinman et al., 2003), connexin-43 (Geletu et al., 2009), key housekeeping
proteins such as αtubulin and glyceraldehyde-3-phosphate dehydrogenase (GAPDH) (Greer et
al., 2009), as well as Stat3 targets such as p21 and p27 (Steinman et al., 2003), VEGF-D
(Orlandini and Oliviero, 2001), survivin (Arulanandam et al., 2009b; Arulanandam et al., 2010),
and p53 (Anagnostopoulou, Thesis). The significance of our work is brought to light by findings
indicating that confluent cell systems may be more representative to the in vivo situation, as
revealed by a close correspondence of genes expressed specifically in the prostate carcinoma line,
192
LNCaP, cultured to high densities, with genes associated with prostate cancer (Chen et al., 2006).
The fact that these genes were different from genes identified in LNCaP cells grown under logphase conditions further stresses the importance of the examination of signalling pathways in
densely growing cells. In line with these data, our observations that cultured breast epithelial
cells, known to express RANK-L in vivo, have high RANK-L levels when grown to high
densities, and the absence of RANK-L in fibroblasts at all densities, that is in a cell type that does
not express it in vivo, further confirm the relevance of data from cells cultured to high densities to
the in vivo situation. Hence, dense cultures offer a more relevant view of the internal mechanisms
of cellular communication which is crucial to the identification and in depth understanding of the
signalling pathways exploited in cancer. It is interesting to note that this novel mechanism of
Stat3 activation by cadherin-mediated cell-cell interactions not only promotes cell survival, but
may also enhance resistance to chemotherapeutic agents (Nakamura et al., 2003; Shain and
Dalton, 2001), a finding which could have significant therapeutic implications.
While significant advances have been made in cancer diagnosis, conventional approaches
in cancer therapy, including surgery, radiation and chemotherapy, appear to be reaching a plateau
in terms of their efficacy. The complexity of cancer as a disease has now become apparent and
rather than generally interfering with cell division, which was the goal of traditional
chemotherapy, the use of targeted therapies will assist in the identification and specific inhibition
of signaling pathways unique to each tumor type. This application of personalized medicine to
the field of oncology is presently seeing some promising results. However, while many current
FDA-approved targeted therapies have focused on tyrosine kinases, our identification of a novel
pathway of Stat3 activation that is independent of a number of kinases would be a promising
target in the treatment of cancers where such inhibitors, e.g. Herceptin, are ineffective. Indeed,
the data included in this thesis clearly reveal that the gp130/Stat3 axis is highly activated upon the
engagement of a number of cadherins, including ones activated during the metastatic switch, and
also lies downstream of Rac1 and Cdc42, which are often overexpressed in cancer. Our findings
193
further expose Stat3 at a central determinant of the balance between cell proliferation and
apoptosis (Fig. 6.1B). Inhibition of Stat3 in tumors can promote apoptosis, through p53dependant or p53-independent mechanisms. My unpublished work has shown that CPA-7
treatment of SV40 Large T expressing cells or tumors in vivo could lead to cell death and tumor
regression possibly through E2F-mediated apoptosis, which is normally inhibited by Stat3. From
our findings presented here, we anticipate that Stat3 inhibitors will be effective in patients with
tumors expressing strong, nuclear ptyrStat3 levels and high serum IL6, indicative of autocrine
gp130 signaling. As Stat3 inhibitors are currently undergoing clinical trials across the United
States, this work will further our understanding of how these inhibitors may exert their anti-tumor
action and also contributes to the growing body of evidence in support of the development of
targeted cancer therapies against Stat3.
194
Figure 6.1
A: Stat3 activation by cadherin engagement
1. Ligation of E-cadherin or cadherin-11 results in the upregulation of Rac1 and Cdc42
protein levels and activity. This may lead to the autocrine induction of IL6 family
cytokines via NFκB. Stat3 activation by cadherin engagement generates a potent survival
signal, possibly via upregulation of survivin, and others such as bcl-xL or mcl-1.
195
2. Mutationally activated Rac and Cdc42 may activate Stat3 by increasing the expression of
IL6 family cytokines leading to Stat3 signalling via the gp130 receptor. Activation of
Stat3 is required for RacV12-mediated migration and cell proliferation.
3. Lipofectamine or calcium phosphate transfection can activate Stat3, even in the absence
of DNA. This may occur through endosomal trafficking or Toll-like receptor activation
of NFκB, which may trigger the cytokine mediated-Stat3 activation. Alternatively, the
physical presence of a precipitate may encourage cell-cell contact and cadherin
engagement.
B: Stat3 inhibition can promote apoptosis
The E2F transcription factor lies downstream of a number of oncogenes, such as Ras,
Src, and Myc as well as the SV40 Large T antigen (TAg), which may inhibit the binding
between the retinoblastoma-susceptibility gene product (Rb) and E2F, leading to
transcriptional activation of genes involved in cell division and apoptosis. Stat3 controls
cell fate by inhibiting E2F-mediated apoptosis and promoting cell survival. Stat3
inhibition may lead to apoptosis through E2F-dependant mechanisms.
The following publications were the result of this work:
1. Arulanandam, R., Geletu, M., Feracci, H. and Raptis, L. (2009). Activated Rac1 requires
gp130 for Stat3 activation, cell proliferation and migration. Experimental Cell Research. In
press.
2. Raptis, L., Arulanandam, R., Vultur, A., Geletu, M., Chevalier, S., and Feracci, H. (2009).
Beyond structure, to survival: activation of Stat3 by cadherin engagement. Biochemistry and
Cell Biology. 87, 835-43.
3. Arulanandam, R., Vultur, A.*, Cao, J.*, Carefoot, E., Elliott, B.E, Truesdell, P., Larue, L.,
Feracci, H., and Raptis, L. (2009). Cadherin-cadherin engagement promotes survival via
Rac/Cdc42 and Stat3. Molecular Cancer Research. 7, 1310-27. *equal contribution.
4. Anagnostopoulou, A.*, Vultur, A.*, Arulanandam, R., Cao, J., Turkson, J., Jove, R., Kim,
J.S., Glenn, M., Hamilton, A.D., and Raptis, L. (2006). Differential effects of Stat3 inhibition in
sparse vs. confluent normal and breast cancer cells. Cancer Letters. 242, 120-32. *equal
contribution.
5. Vultur, A., Arulanandam, R., Turkson, J., Niu, G., Jove, R., and Raptis, L. (2005). Stat3 is
required for full neoplastic transformation by the Simian Virus 40 Large Tumor antigen.
Molecular Biology of the Cell. 16, 3832-46.
6. Arulanandam, R., Vultur, A., and Raptis, L. (2005). Transfection techniques affecting Stat3
activity levels. Analytical Biochemistry. 338, 83-89.
7. Vultur, A., Cao, J., Arulanandam, R., Turkson, J., Jove, R., Greer, P., Craig, P., Elliott, B.,
and Raptis, L. (2004). Cell to cell adhesion modulates Stat3 activity in normal and breast
carcinoma cells. Oncogene. 23, 2600-2616.
196
Submitted, or manuscripts in preparation
8. Greer, S. *, Honeywell, R. *, Geletu, M., Arulanandam, R., and Raptis, L. Housekeeping
genes; expression levels change with density of cultured cells. Journal of Immunological
Methods. Submitted, September 2009. *equal contribution.
9. Arulanandam, R., Geletu, M, Feracci, H., Chevalier, S. and Raptis, L. Cadherin-11
engagement activates Stat3 and promotes survival of mouse fibroblasts. Manuscript in
preparation.
10. Arulanandam, R, D’Abreo C, Geletu M and Raptis L. Activated Src increases total Rac
protein levels and requires Rac and gp130 for Stat3-tyr705 phosphorylation and full neoplastic
transformation. Manuscript in preparation.
11. Vultur, A., Geletu, M., Arulanandam, R., Turkson, J. and Raptis, L. Differential
phosphorylation of Stat3a and Stat3b following cell to cell adhesion or expression of the Simian
Virus 40 Large Tumor antigen. Manuscript in preparation.
I also contributed to the following publications which do not constitute part of this thesis:
1. Arulanandam, R., Geletu, M., and Raptis, L. (2009). The Simian Virus 40 Large Tumor
antigen activates cSrc and requires cSrc for full neoplastic transformation. Anticancer Research.
In press.
2. Geletu M, Chaize C, Arulanandam R, Vultur A, Kowolik C, Anagnostopoulou A, Jove R and
Raptis, L. (2009). Stat3 activity is required for gap junctional permeability in normal rat liver
epithelial cells. DNA and Cell Biology. 28, 319-27.
3. Mohan, R.*, Arulanandam, R.*, Vultur, A., Lo, O., Cao, J., Grammatikakis, N., and Raptis,
L. (2008). Differential effects of Adenovirus E1A and Simian Virus 40 Large Tumor antigen
upon the activity of the extracellular-signal-regulated kinase and the Signal Transducer and
Activator of Transcription-3 in rodent cells. Trends in Cell and Molecular Biology. 3, 59-68.
*equal contribution.
4. Raptis, L., Vultur, A., Brownell, H., Tomai, E., Anagnostopoulou, A., Arulanandam, R.,
Cao, J., and Firth, K. (2008). Electroporation of adherent cells in situ for the study of signal
transduction and gap junctional communication. Methods in Molecular Biology: Electroporation
Protocols. 423, 173-189.
5. Cao, J., Arulanandam, R., Vultur, A., Anagnostopoulou, A., and Raptis, L. (2007).
Adenovirus E1A requires c-Ras for full neoplastic transformation or suppression of
differentiation of murine preadipocytes. Molecular Carcinogenesis. 46, 284-302.
6. Cao, J., Arulanandam, R., Vultur, A., Anagnostopoulou, A., and Raptis, L. (2007).
Differential effects of c-Ras upon transformation, adipocytic differentiation and apoptosis
mediated by the simian virus 40 large tumor antigen. Biochemistry and Cell Biology. 85, 32-48.
7. Cao, J.*, Arulanandam, R.*, Vultur, A., Preston, T., Jaronczyk, Tomai, E., K., Zandi,
K., and Raptis, L. (2005). Adenovirus-5 E1A suppresses differentiation of 3T3 L1 preadipocytes
197
at lower levels than required for induction of apoptosis. Molecular Carcinogenesis. 43, 38-50.
*equal contribution.
Manuscripts in preparation
8. Geletu, M., Chaize, C., Arulanandam, R., Anagnostopoulou, A., and Raptis, L. Src activity
and gap junctional communication in cultured, human lung carcinoma cells. Manuscript in
preparation.
198
References
Alroy, I., Tuvia, S., Greener, T., Gordon, D., Barr, H. M., Taglicht, D., Mandil-Levin, R., BenAvraham, D., Konforty, D., Nir, A., Levius, O., Bicoviski, V., Dori, M., Cohen, S., Yaar, L.,
Erez, O., Propheta-Meiran, O., Koskas, M., Caspi-Bachar, E., Alchanati, I., Sela-Brown, A.,
Moskowitz, H., Tessmer, U., Schubert, U., and Reiss, Y. (2005). The trans-Golgi networkassociated human ubiquitin-protein ligase POSH is essential for HIV type 1 production. Proc.
Natl. Acad. Sci. U. S. A 102, 1478-1483.
Anagnostopoulou, A. The role of Stat3 in cell division and apoptosis. 2009.
Ref Type: Thesis/Dissertation
Anagnostopoulou, A., Vultur, A., Arulanandam, R., Cao, J., Turkson, J., Jove, R., Kim, J. S.,
Glenn, M., Hamilton, A. D., and Raptis, L. (2006). Differential effects of Stat3 inhibition in
sparse vs confluent normal and breast cancer cells. Cancer Lett. 242, 120-132.
Arulanandam, R., Geletu, M., Chevalier, S., Feracci, H., and Raptis, L. (2010). Cadherin-11
engagement activates Stat3 and promotes survival of mouse fibroblasts. Manuscript in
preparation.
Arulanandam, R., Geletu, M., Feracci, H., and Raptis, L. (2009a). Activated Rac1 requires gp130
for Stat3 activation, cell proliferation and migration. Exp. Cell Res.
Arulanandam, R., Vultur, A., Cao, J., Carefoot, E., Elliott, B. E., Truesdell, P. F., Larue, L.,
Feracci, H., and Raptis, L. (2009b). Cadherin-cadherin engagement promotes cell survival via
Rac1/Cdc42 and signal transducer and activator of transcription-3. Mol. Cancer Res. 7, 13101327.
Aznar, S. and Lacal, J. C. (2001). Rho signals to cell growth and apoptosis. Cancer Lett. 165, 110.
Batt, D. B. and Roberts, T. M. (1998). Cell density modulates protein-tyrosine phosphorylation. J.
Biol. Chem. 273, 3408-3414.
Berishaj, M., Gao, S. P., Ahmed, S., Leslie, K., Al Ahmadie, H., Gerald, W. L., Bornmann, W.,
and Bromberg, J. F. (2007). Stat3 is tyrosine-phosphorylated through the interleukin6/glycoprotein 130/Janus kinase pathway in breast cancer. Breast Cancer Res. 9, R32.
Catlett-Falcone, R., Landowski, T. H., Oshiro, M. M., Turkson, J., Levitzki, A., Savino, R.,
Ciliberto, G., Moscinski, L., Fernandez-Luna, J. L., Nunez, G., Dalton, W. S., and Jove, R.
(1999a). Constitutive activation of Stat3 signaling confers resistance to apoptosis in human U266
myeloma cells. Immunity 10, 105-115.
Catlett-Falcone, R., Landowski, T. H., Oshiro, M. M., Turkson, J., Levitzki, A., Savino, R.,
Ciliberto, G., Moscinski, L., Fernandez-Luna, J. L., Nunez, G., Dalton, W. S., and Jove, R.
(1999b). Constitutive activation of Stat3 signaling confers resistance to apoptosis in human U266
myeloma cells. Immunity. 10, 105-115.
199
Chen, Q., Watson, J. T., Marengo, S. R., Decker, K. S., Coleman, I., Nelson, P. S., and Sikes, R.
A. (2006). Gene expression in the LNCaP human prostate cancer progression model: progression
associated expression in vitro corresponds to expression changes associated with prostate cancer
progression in vivo. Cancer Lett. 244, 274-288.
Cobleigh, M. A., Vogel, C. L., Tripathy, D., Robert, N. J., Scholl, S., Fehrenbacher, L., Wolter, J.
M., Paton, V., Shak, S., Lieberman, G., and Slamon, D. J. (1999). Multinational study of the
efficacy and safety of humanized anti-HER2 monoclonal antibody in women who have HER2overexpressing metastatic breast cancer that has progressed after chemotherapy for metastatic
disease. J. Clin. Oncol. 17, 2639-2648.
Debidda, M., Wang, L., Zang, H., Poli, V., and Zheng, Y. (2005). A role of STAT3 in Rho
GTPase-regulated cell migration and proliferation. J. Biol. Chem. 280, 17275-17285.
Diaz, N., Minton, S., Cox, C., Bowman, T., Gritsko, T., Garcia, R., Eweis, I., Wloch, M.,
Livingston, S., Seijo, E., Cantor, A., Lee, J. H., Beam, C. A., Sullivan, D., Jove, R., and MuroCacho, C. A. (2006). Activation of stat3 in primary tumors from high-risk breast cancer patients
is associated with elevated levels of activated SRC and survivin expression. Clin. Cancer Res. 12,
20-28.
Draetta, G. and Eckstein, J. (1997). Cdc25 protein phosphatases in cell proliferation. Biochim.
Biophys. Acta 1332, M53-M63.
Faruqi, T. R., Gomez, D., Bustelo, X. R., Bar-Sagi, D., and Reich, N. C. (2001). Rac1 mediates
STAT3 activation by autocrine IL-6. Proc. Nat. Acad. Sci. USA 98, 9014-9019.
Fata, J. E., Kong, Y. Y., Li, J., Sasaki, T., Irie-Sasaki, J., Moorehead, R. A., Elliott, R., Scully, S.,
Voura, E. B., Lacey, D. L., Boyle, W. J., Khokha, R., and Penninger, J. M. (2000). The osteoclast
differentiation factor osteoprotegerin-ligand is essential for mammary gland development. Cell
103, 41-50.
Fischer, P. and Hilfiker-Kleiner, D. (2008). Role of gp130-mediated signalling pathways in the
heart and its impact on potential therapeutic aspects. Br. J. Pharmacol. 153 Suppl 1, S414-S427.
Fujita, Y., Krause, G., Scheffner, M., Zechner, D., Leddy, H. E., Behrens, J., Sommer, T., and
Birchmeier, W. (2002). Hakai, a c-Cbl-like protein, ubiquitinates and induces endocytosis of the
E-cadherin complex. Nat. Cell Biol. 4, 222-231.
Fukata, M. and Kaibuchi, K. (2001). Rho-family GTPases in cadherin-mediated cell-cell
adhesion. Nat. Rev. Mol. Cell Biol. 2, 887-897.
Gao, S. P., Mark, K. G., Leslie, K., Pao, W., Motoi, N., Gerald, W. L., Travis, W. D., Bornmann,
W., Veach, D., Clarkson, B., and Bromberg, J. F. (2007). Mutations in the EGFR kinase domain
mediate STAT3 activation via IL-6 production in human lung adenocarcinomas. J. Clin. Invest
117, 3846-3856.
Geletu, M., Chaize, C., Arulanandam, R., Vultur, A., Kowolik, C., Anagnostopoulou, A., Jove,
R., and Raptis, L. (2009). Stat3 activity is required for gap junctional permeability in normal rat
liver epithelial cells. DNA Cell Biol. 28, 319-327.
200
Greer, S., Honeywell, R., Geletu, M., Arulanandam, R., and Raptis, L. (2009). Housekeeping
gene products; levels may change with confluence of cultured cells. Journal of Immunological
Methods JIM-D-09-00266.
Harries, M. and Smith, I. (2002). The development and clinical use of trastuzumab (Herceptin).
Endocr. Relat Cancer 9, 75-85.
Hirose, K., Kawashima, T., Iwamoto, I., Nosaka, T., and Kitamura, T. (2001). MgcRacGAP is
involved in cytokinesis through associating with mitotic spindle and midbody. J. Biol. Chem.
276, 5821-5828.
Huber, A. H., Stewart, D. B., Laurents, D. V., Nelson, W. J., and Weis, W. I. (2001). The
cadherin cytoplasmic domain is unstructured in the absence of beta-catenin. A possible
mechanism for regulating cadherin turnover. J. Biol. Chem. 276, 12301-12309.
Hynes, N. E. and Lane, H. A. (2005). ERBB receptors and cancer: the complexity of targeted
inhibitors. Nat. Rev. Cancer 5, 341-354.
Jacobs, H. W., Keidel, E., and Lehner, C. F. (2001). A complex degradation signal in Cyclin A
required for G1 arrest, and a C-terminal region for mitosis. EMBO J. 20, 2376-2386.
Kawashima, T., Bao, Y. C., Minoshima, Y., Nomura, Y., Hatori, T., Hori, T., Fukagawa, T.,
Fukada, T., Takahashi, N., Nosaka, T., Inoue, M., Sato, T., Kukimoto-Niino, M., Shirouzu, M.,
Yokoyama, S., and Kitamura, T. (2009). A Rac GTPase-activating protein, MgcRacGAP, is a
nuclear localizing signal-containing nuclear chaperone in the activation of STAT transcription
factors. Mol. Cell Biol. 29, 1796-1813.
Kawashima, T., Bao, Y. C., Nomura, Y., Moon, Y., Tonozuka, Y., Minoshima, Y., Hatori, T.,
Tsuchiya, A., Kiyono, M., Nosaka, T., Nakajima, H., Williams, D. A., and Kitamura, T. (2006).
Rac1 and a GTPase-activating protein, MgcRacGAP, are required for nuclear translocation of
STAT transcription factors. J. Cell Biol. 175, 937-946.
Kim, G. H., Park, E., Kong, Y. Y., and Han, J. K. (2006). Novel function of POSH, a JNK
scaffold, as an E3 ubiquitin ligase for the Hrs stability on early endosomes. Cell Signal. 18, 553563.
Kinch, M. S., Petch, L., Zhong, C., and Burridge, K. (1997). E-cadherin engagement stimulates
tyrosine phosphorylation. Cell Adhes. Commun. 4, 425-437.
Kovacic, H. N., Irani, K., and Goldschmidt-Clermont, P. J. (2001). Redox regulation of human
Rac1 stability by the proteasome in human aortic endothelial cells. J. Biol. Chem. 276, 4585645861.
Larue, L., Antos, C., Butz, S., Huber, O., Delmas, V., Dominis, M., and Kemler, R. (1996). A
role for cadherins in tissue formation. Development 122, 3185-3194.
Le, T. L., Yap, A. S., and Stow, J. L. (1999). Recycling of E-cadherin: a potential mechanism for
regulating cadherin dynamics. J. Cell Biol. 146, 219-232.
Lerm, M., Pop, M., Fritz, G., Aktories, K., and Schmidt, G. (2002). Proteasomal degradation of
cytotoxic necrotizing factor 1-activated rac. Infect. Immun. 70, 4053-4058.
201
Lou, W., Ni, Z., Dyer, K., Tweardy, D. J., and Gao, A. C. (2000). Interleukin-6 induces prostate
cancer cell growth accompanied by activation of stat3 signaling pathway. Prostate 42, 239-242.
Lynch, E. A., Stall, J., Schmidt, G., Chavrier, P., and D'Souza-Schorey, C. (2006). Proteasomemediated degradation of Rac1-GTP during epithelial cell scattering. Mol. Biol. Cell 17, 22362242.
Mora, L. B., Buettner, R., Seigne, J., Diaz, J., Ahmad, N., Garcia, R., Bowman, T., Falcone, R.,
Fairclough, R., Cantor, A., Muro-Cacho, C., Livingston, S., Karras, J., Pow-Sang, J., and Jove, R.
(2002). Constitutive activation of Stat3 in human prostate tumors and cell lines: direct inhibition
of Stat3 signaling induces apoptosis of prostate cancer cells. Cancer Res. 62, 6659-6666.
Nakagawa, M., Fukata, M., Yamaga, M., Itoh, N., and Kaibuchi, K. (2001). Recruitment and
activation of Rac1 by the formation of E-cadherin-mediated cell-cell adhesion sites. J. Cell Sci.
114, 1829-1838.
Nakamura, T., Kato, Y., Fuji, H., Horiuchi, T., Chiba, Y., and Tanaka, K. (2003). E-cadherindependent intercellular adhesion enhances chemoresistance. Int. J. Mol. Med. 12, 693-700.
Nemoto, K., Vogt, A., Oguri, T., and Lazo, J. S. (2004). Activation of the Raf-1/MEK/Erk kinase
pathway by a novel Cdc25 inhibitor in human prostate cancer cells. Prostate 58, 95-102.
Niu, G., Wright, K. L., Ma, Y., Wright, G. M., Huang, M., Irby, R., Briggs, J., Karras, J., Cress,
W. D., Pardoll, D., Jove, R., Chen, J., and Yu, H. (2005). Role of Stat3 in regulating p53
expression and function. Mol. Cell Biol. 25, 7432-7440.
Noren, N. K., Niessen, C. M., Gumbiner, B. M., and Burridge, K. (2001). Cadherin engagement
regulates Rho family GTPases. J. Biol. Chem. 276, 33305-33308.
Orlandini, M. and Oliviero, S. (2001). In fibroblasts Vegf-D expression is induced by cell-cell
contact mediated by cadherin-11. J. Biol. Chem. 276, 6576-6581.
Perona, R., Montaner, S., Saniger, L., Sanchez-Perez, I., Bravo, R., and Lacal, J. C. (1997).
Activation of the nuclear factor-kappaB by Rho, CDC42, and Rac-1 proteins. Genes Dev. 11,
463-475.
Pietsch, E. C., Sykes, S. M., McMahon, S. B., and Murphy, M. E. (2008). The p53 family and
programmed cell death. Oncogene 27, 6507-6521.
Pop, M., Aktories, K., and Schmidt, G. (2004). Isotype-specific degradation of Rac activated by
the cytotoxic necrotizing factor 1. J. Biol. Chem. 279, 35840-35848.
Scheule, R. K. (2000). The role of CpG motifs in immunostimulation and gene therapy. Adv.
Drug Deliv. Rev. 44, 119-134.
Sehgal, P. B. (2008). Paradigm shifts in the cell biology of STAT signaling. Semin. Cell Dev.
Biol. 19, 329-340.
Settleman, J., Albright, C. F., Foster, L. C., and Weinberg, R. A. (1992). Association between
GTPase activators for Rho and Ras families. Nature 359, 153-154.
202
Shain, K. H. and Dalton, W. S. (2001). Cell adhesion is a key determinant in de novo multidrug
resistance (MDR): new targets for the prevention of acquired MDR. Mol. Cancer Ther. 1, 69-78.
Shain, K. H., Yarde, D. N., Meads, M. B., Huang, M., Jove, R., Hazlehurst, L. A., and Dalton, W.
S. (2009). Beta1 integrin adhesion enhances IL-6-mediated STAT3 signaling in myeloma cells:
implications for microenvironment influence on tumor survival and proliferation. Cancer Res. 69,
1009-1015.
Shen, Y., Hirsch, D. S., Sasiela, C. A., and Wu, W. J. (2008). Cdc42 regulates E-cadherin
ubiquitination and degradation through an epidermal growth factor receptor to Src-mediated
pathway. J. Biol. Chem. 283, 5127-5137.
Simon, A. R., Vikis, H. G., Stewart, S., Fanburg, B. L., Cochran, B. H., and Guan, K. L. (2000).
Regulation of STAT3 by direct binding to the Rac1 GTPase. Science 290, 144-147.
Steinman, R. A., Wentzel, A., Lu, Y., Stehle, C., and Grandis, J. R. (2003). Activation of Stat3 by
cell confluence reveals negative regulation of Stat3 by cdk2. Oncogene 22, 3608-3615.
Tsuda, M., Langmann, C., Harden, N., and Aigaki, T. (2005). The RING-finger scaffold protein
Plenty of SH3s targets TAK1 to control immunity signalling in Drosophila. EMBO Rep. 6, 10821087.
Tu, S. and Cerione, R. A. (2001). Cdc42 is a substrate for caspases and influences Fas-induced
apoptosis. J. Biol. Chem. 276, 19656-19663.
Turkson, J., Bowman, T., Adnane, J., Zhang, Y., Djeu, J. Y., Sekharam, M., Frank, D. A.,
Holzman, L. B., Wu, J., Sebti, S., and Jove, R. (1999). Requirement for Ras/Rac1-mediated p38
and c-Jun N-terminal kinase signaling in Stat3 transcriptional activity induced by the Src
oncoprotein. Mol. Cell. Biol. 19, 7519-7528.
Turkson, J., Kim, J. S., Zhang, S., Yuan, J., Huang, M., Glenn, M., Haura, E., Sebti, S., Hamilton,
A. D., and Jove, R. (2004a). Novel peptidomimetic inhibitors of signal transducer and activator of
transcription 3 dimerization and biological activity. Mol. Cancer Ther. 3, 261-269.
Turkson, J., Zhang, S., Palmer, J., Kay, H., Stanko, J., Mora, L., Sebti, S., Yu, H., and Jove, R.
(2004b). Inhibition of constitutive Stat3 activation by novel platinum complexes with potent antimalignant activity. Mol. Cancer Ther. 3, 1533-1542.
Tuvia, S., Taglicht, D., Erez, O., Alroy, I., Alchanati, I., Bicoviski, V., Dori-Bachash, M., BenAvraham, D., and Reiss, Y. (2007). The ubiquitin E3 ligase POSH regulates calcium homeostasis
through spatial control of Herp. J. Cell Biol. 177, 51-61.
Visvikis, O., Lores, P., Boyer, L., Chardin, P., Lemichez, E., and Gacon, G. (2008). Activated
Rac1, but not the tumorigenic variant Rac1b, is ubiquitinated on Lys 147 through a JNKregulated process. FEBS J. 275, 386-396.
Vogel, C. L., Cobleigh, M. A., Tripathy, D., Gutheil, J. C., Harris, L. N., Fehrenbacher, L.,
Slamon, D. J., Murphy, M., Novotny, W. F., Burchmore, M., Shak, S., Stewart, S. J., and Press,
M. (2002). Efficacy and safety of trastuzumab as a single agent in first-line treatment of HER2overexpressing metastatic breast cancer. J. Clin. Oncol. 20, 719-726.
203
Vultur, A., Arulanandam, R., Turkson, J., Niu, G., Jove, R., and Raptis, L. (2005). Stat3 is
required for full neoplastic transformation by the Simian Virus 40 Large Tumor antigen.
Molecular Biology of the Cell 16, 3832-3846.
Vultur, A., Cao, J., Arulanandam, R., Turkson, J., Jove, R., Greer, P., Craig, A., Elliott, B., and
Raptis, L. (2004). Cell-to-cell adhesion modulates Stat3 activity in normal and breast carcinoma
cells. Oncogene 23, 2600-2616.
Wang, H. R., Ogunjimi, A. A., Zhang, Y., Ozdamar, B., Bose, R., and Wrana, J. L. (2006).
Degradation of RhoA by Smurf1 ubiquitin ligase. Methods Enzymol. 406, 437-447.
Wang, Z., Zhang, B., Wang, M., and Carr, B. I. (2005). Cdc25A and ERK interaction: EGFRindependent ERK activation by a protein phosphatase Cdc25A inhibitor, compound 5. J. Cell
Physiol 204, 437-444.
Wu, J., Jester, W. F., Jr., and Orth, J. M. (2005). Short-type PB-cadherin promotes survival of
gonocytes and activates JAK-STAT signalling. Dev. Biol. 284, 437-450.
Wu, W. J., Tu, S., and Cerione, R. A. (2003). Activated Cdc42 sequesters c-Cbl and prevents
EGF receptor degradation. Cell 114, 715-725.
Xia, K., Lee, R. S., Narsimhan, R. P., Mukhopadhyay, N. K., Neel, B. G., and Roberts, T. M.
(1999). Tyrosine phosphorylation of the proto-oncoprotein Raf-1 is regulated by Raf-1 itself and
the phosphatase Cdc25A. Mol. Cell. Biol. 19, 4819-4824.
Xu, F., Mukhopadhyay, S., and Sehgal, P. B. (2007). Live cell imaging of interleukin-6-induced
targeting of "transcription factor" STAT3 to sequestering endosomes in the cytoplasm. Am. J.
Physiol Cell Physiol 293, C1374-C1382.
Xu, Z., Kukekov, N. V., and Greene, L. A. (2003). POSH acts as a scaffold for a multiprotein
complex that mediates JNK activation in apoptosis. EMBO J. 22, 252-261.
Yamada, T., Kurosaki, T., and Hikida, M. (2008). Essential roles of mgcRacGAP in multilineage
differentiation and survival of murine hematopoietic cells. Biochem. Biophys. Res. Commun.
372, 941-946.
Yew, N. S., Wang, K. X., Przybylska, M., Bagley, R. G., Stedman, M., Marshall, J., Scheule, R.
K., and Cheng, S. H. (1999). Contribution of plasmid DNA to inflammation in the lung after
administration of cationic lipid:pDNA complexes. Hum. Gene Ther. 10, 223-234.
Yonenaga, Y., Mori, A., Fujimoto, A., Nagayama, S., Tachibana, T., Onodera, H., and Uemoto,
S. (2007). The administration of naked plasmid DNA into the liver induces antitumor innate
immunity in a murine liver metastasis model. J. Gene Med. 9, 299-307.
Yoo, J. W., Hong, S. W., Kim, S., and Lee, D. K. (2006). Inflammatory cytokine induction by
siRNAs is cell type- and transfection reagent-specific. Biochem. Biophys. Res. Commun. 347,
1053-1058.
Yu, H., Pardoll, D., and Jove, R. (2009). STATs in cancer inflammation and immunity: a leading
role for STAT3. Nat. Rev. Cancer 9, 798-809.
204
Zhang, Y., Wang, H. R., and Wrana, J. L. (2004). Smurf1: a link between cell polarity and
ubiquitination. Cell Cycle 3, 391-392.
205
Appendix
206
Figure A.1
The effect of cell density on Stat3 activity levels.
A: Examination of E-cadherin levels in HC11 cells.
Detergent cell lysates from HC11 (lane 1) or Balb/c 3T3 (lane 2) cells were resolved by
gel electrophoresis. The blot was cut into strips which were probed for E-cadherin (top) or Hsp90
(bottom). Arrows at the right point to the positions of E-cadherin or Hsp90, as indicated.
Numbers on the left refer to molecular weight markers.
B: HC11 cells are growth arrested at high densities.
0.3x105 HC11 cells were plated in 3 cm petris and cell numbers obtained at different
times, as indicated. Cell numbers plateaued at 2 days post-confluence, indicating the absence of
overt cell division during this time.
C: Cell density upregulates nuclear Stat3-ptyr705.
HC11 cells were grown to a confluence of 30% (lanes 1-3) or 2 days post confluence
(lanes 4-6). Cell fractionation experiments were conducted by lysing cells in Hypotonic buffer
(40 mM Hepes pH 6.0, 2 mM EGTA, 2 mM EDTA) and protease and phosphatase inhibitors,
followed by Dounce homogenisation. Nuclei were collected by low speed centrifugation and
proteins extracted with Triton X-100 from the nuclear pellet (Vultur et al., 2004). Nuclear and
cytoplasmic cell fractions were subsequently resolved by gel electrophoresis and blots were cut
into strips which were probed for Stat3-ptyr705 or Hsp90 as a loading control.
D: Cell density upregulates Stat3 transcriptional activity
HC11 cells were transfected with the Stat3-dependent pLucTKS3 reporter driving a
firefly luciferase gene under control of the C-reactive gene promoter element, and the Stat3independent pRLSRE reporter driving a Renilla luciferase gene under control of the c-fos SRE
promotor, respectively. Cells were grown to the indicated densities with daily media changes and
firefly (■) or Renilla ( ) luciferase activities determined in detergent cell extracts (see Chapter 3,
Materials and Methods). Values shown represent luciferase units expressed as a percentage of the
highest value obtained, means ± s.e.m. of at least 3 experiments, each performed in triplicate.
207
Figure A.2
Weak Stat3 activation in ES cells by the LIF present in the growth medium.
To examine the contribution of the LIF present in the ES cell medium, to Stat3 activation,
the E-cadherin defective null cells were grown to 30% or 100% confluence in complete ES
medium, containing 15% fetal calf serum and 1x103 units LIF/ml, then grown in medium lacking
LIF for 24 hours (lanes 1, 4), followed by complete ES medium containing LIF (lanes 3, 6).
Starved null cells were also grown in medium conditioned by growing 100% confluent, Ecad+/+
cells in 10 cm petris with 10 mls serum-free DMEM (lanes 2, 5). Detergent cell extracts were
probed for Stat3-ptyr705 or Hsp90, as indicated. Note the small but detectable increase in Stat3ptyr705 levels upon addition of ES medium (lanes 3, 6 vs 1, 4), but not by conditioned medium
(lanes 2, 5). Note that 100 μg protein were loaded to better visualize the differences.
208
Figure A.3
ES cells do not differentiate when cultured to high densities, in the presence of LIF
Ecad+/+ cells were grown to 30% confluence in the absence of LIF for 5 days (a, b), or
to 5 days post-confluence, in the presence of LIF (c, d), both using a particular lot of fetal calf
serum which did not allow differentiation in the presence of LIF. Cells were subsequently fixed
and stained for the epithelial cell marker, cytokeratin-8 with the TROMA-1 rat monoclonal
antibody (Larue et al., 1996). Note the absence of stained, differentiated cells in d and their
presence in a and b (arrow).
209
Figure A.4
E-cadherin ablation inhibits the density-mediated, Stat3 activation.
A: Proteins in detergent cell extracts (30 μg) from cell lines established from wild-type
(Ecad+/+), E-cadherin-/- (null) and null cells where E-cadherin was reexpressed through
transfection (addback) were grown to 2 days postconfluence, resolved by gel electrophoresis, and
Western immunoblots cut into strips which were probed for E-cadherin, ptyr-705 Stat3, phosphoErk1/2, or Hsp90 as a loading control, as indicated.
B: Stat3 tyr-705 phosphorylation: The addback cells were grown to increasing densities, up to
8 days post-confluence. Detergent cell lysates were resolved by gel electrophoresis and the blot
was probed for Stat3-ptyr705 (top), total Stat3 (middle) or Hsp90 as a loading control (bottom),
as indicated.
210
C: Stat3 DNA binding activity:
Autoradiogram of Stat3:sis inducible element (hSIE) complexes, in nuclear extracts from
Ecad+/+ (lanes 1-4) or null (lanes 5-8) cells prepared at the indicated densities, in EMSA assays
(Vultur et al., 2005). Arrow points to the position of the Stat3:hSIE complex. In lanes 4, 8, 12
and 15, the same samples as in lanes 3, 7, 11 and 13, respectively, were incubated with an antiStat3 antibody. Lane 14: The same sample as in lane 13 was incubated with an anti-Stat1
antibody. The position of the supershifted Stat3-hSIE-antibody complex (lanes 4, 8, 12, 15) is
indicated with an asterisk. Note the absence of supershifting with the Stat1 antibody.
D: Stat3 transcriptional activity:
Ecad+/+ and null cells were transfected with the Stat3-dependent pLucTKS3 reporter
driving a firefly luciferase gene under control of the C-reactive gene promoter element, and the
Stat3-independent pRLSRE reporter driving a Renilla luciferase gene under control of the c-fos
SRE promotor, respectively. Cells were grown to the indicated densities and firefly (■) or
Renilla ( ) luciferase activities determined in detergent cell extracts with the peak value of the
control, src-527F transformed cells taken as 100% (Vultur et al., 2004). Values shown represent
luciferase units expressed as a percentage of the highest value obtained, means ± s.e.m. of at least
3 experiments, each performed in triplicate.
211
Figure A.5
Plating HC11 cells on cadherin-coated surfaces for more than 10 hours or at high
cell densities does not activate Stat3.
A: Plating HC11 cells on E/EC12-coated surfaces for 10 hrs is not sufficient to activate
Stat3.
0.3x105 HC11 cells were plated in 3 cm plastic petris coated with the indicated amounts
of the E/EC12, E-cadherin fragment. 10 hours after plating, when confluence was ~40%, 30 μg
of detergent cell lysates were resolved by gel electrophoresis. The blot was cut and probed for
Stat3-ptyr705 (top) or Hsp90 (bottom). Lane 5: HC11 cells grown on control, uncoated plastic
petris and harvested at 2 days post-confluence, at which time Stat3-ptyr705 levels were examined
as above (see Materials and Methods).
B: Stat3 activation upon plating HC11 cells on E/EC12-coated surfaces at 70% confluence
is less pronounced than at lower cell densities.
1x105 Balb/c 3T3 (lanes 1 and 2) or HC11 (lanes 3-11) cells were plated in 3 cm plastic
petris coated with the indicated amounts of the E/EC12, E-cadherin fragment (lanes 2 and 4-8) or
the E/W2A mutant (lanes 9-10), as indicated. 48 hours after plating, when confluence was ~70%,
30μg of detergent cell lysates were probed, as above, for Stat3-ptyr705 (top) or Hsp90 (bottom).
Lane 11: HC11 cells grown on control, uncoated plastic petris and harvested at 1 day postconfluence, at which time Stat3-ptyr705 levels were examined as above (see Materials and
Methods). Arrows point to the positions of Stat3-ptyr705 or Hsp90, respectively. Numbers at the
left refer to molecular weight markers.
212
Figure A.6
E-cadherin engagement is sufficient to activate Stat3 in ES cells.
Ecad+/+ (lanes 5-8), null (lanes 1 and 2) and addback (lanes 3 and 4) cells were grown in plastic
petris coated with 100 µg/ml E/EC12 E-cadherin fragment or the E/W2A mutant (lane 7), as
indicated. To adjust for differences in growth rates among the different cell lines, 0.3x105,
0.4x105 or 0.5x105 cells per 3 cm plate were seeded for the Ecad+/+, addback (i.e. null cells
where E-cadherin was re-expressed through transfection) and null cells, respectively. Detergent
cell extracts were prepared 48 hours thereafter, when confluences were ~40% for all three lines,
and resolved by gel electrophoresis. The blot was cut and probed for Stat3-ptyr705 (top panel) or
Hsp90 (bottom panel). Lane 8: Ecad+/+ cells were grown on control, uncoated plastic petris and
harvested at 2 days post-confluence, at which time Stat3-ptyr705 levels were examined as above.
Numbers under the lanes of the top panel refer to Stat3-ptyr705 band intensities obtained through
quantitation by fluorimager analysis and normalized to Hsp90 levels, with the peak value of the
control, Ecad+/+ cells at 2 days after confluence taken as 100% (see Materials and Methods).
Arrows point to the positions of Stat3-ptyr705 or Hsp90, respectively. Numbers at the left refer
to molecular weight markers.
213
Figure A.7
Stat3 activation after plating different numbers of cells for 24 hours.
Different numbers of HC11 cells (0.05x106, 0.1x106, 0.5x106, 1x106, 1.8x106 , 2x106 or
3x10 per 3cm petri, respectively) were plated in plastic petri dishes. 24 hours later, detergent
cell lysates were resolved by gel electrophoresis and Western blots cut into strips which were
probed for Stat3-ptyr705, Rac1, Cdc42 or Hsp90 as a loading control, as indicated.
6
214
Figure A.8
Toxin B can inhibit the induction of the density-mediated, Stat3 tyr705
phosphorylation
Sparsely growing, HC11 cells (lane 1) were re-plated to a high density (3 x 106 cells / 3
cm plate, lanes 2-4). Following attachment for one hour (lane 2), cells were treated with the
DMSO carrier (lane 3) or 50 pg/ml toxin B (lane 4) for 5 hours and cell extracts resolved by gel
electrophoresis. The blot was cut into strips and probed for Stat3-ptyr705 or Hsp90 as a loading
control, as indicated.
215
216
Figure A.9
Reduction of Rac1 or Cdc42 levels or activity prevents Stat3-tyr705 phosphorylation
in mouse 10T1/2 fibroblasts
A: Examination of cadherin-11 mRNA levels in 10T1/2, HC11 and Balb/c 3T3 cells.
Extracts from 10T1/2, HC11 or Balb/c 3T3 cells were probed by RT-PCR for cadherin-11 (lanes
4-6), with 18S RNA as a control (lanes 1-3, see Materials and Methods). Numbers on the left
refer to the molecular weight marker lanes (M). 10T1/2 cells also show some E-cadherin
expression by RT-PCR, although not nearly as much as HC11 cells (not shown).
B: The Toxin B, Rho GTPase protein family inhibitor prevents the density-mediated, Stat3ptyr705 phosphorylation in fibroblasts.
Left: 10T1/2 cells were grown to different densities with daily media changes and treated with
50 pg/ml toxin B for 5 hrs (lanes 6-10), or not (lanes 1-5) and lysates probed for Stat3-ptyr705
(top panel), total Rac1 (middle panel) or Hsp90 as a loading control (bottom panel), as indicated.
Right: 10T1/2 cells were grown to 1 day post-confluence and treated with 50 pg/ml toxin B for 5
hours (lane 2) or the DMSO carrier alone (lane 1) and the activity of Rac1 or Cdc42 determined
(see Materials and Methods).
C: The NSC23766 Rac1 inhibitor prevents the density-mediated, Stat3-ptyr705
phosphorylation.
Left: For treatment with the NSC23766 Rac inhibitor (Calbiochem), 10T1/2 cells were allowed
to grow to various densities and were incubated with 100 μM inhibitor for 12 hours (lanes 9-16),
or not (lanes 1-8), as described (Gao et al., 2004), and lysates probed for Stat3-ptyr705 (top
panel), or Hsp90 as a loading control (bottom panel), as indicated.
Right: 10T1/2 cells were grown to 1 day post-confluence and treated with 100 μM NSC23766
for 12 hours (lane 2) or the DMSO carrier alone (lane 1) and the activity of Rac1 determined (see
Materials and Methods).
D: The dominant-negative Rac1N17 and Cdc42N17 mutants prevent the density-mediated,
Stat3-ptyr705 phosphorylation in 10T1/2 fibroblasts.
Left: 10T1/2 or 10T1/2 stably expressing Rac1N17 (top) or Cdc42N17 (bottom) were grown to
different densities, as indicated, and lysates were probed for Stat3-ptyr705 or Hsp90.
Right: 10T1/2 and 10T1/2 stably expressing Rac1N17or Cdc42N17 were grown to 1 day
postconfluence and probed with an antibody specific for the myc-tag. Hsp90 was used as a
loading control.
217
Figure A.10
Reduction of Cdc42 levels or activity prevents Stat3-tyr705 phosphorylation.
A: The dominant-negative Cdc42N17 mutant prevents the density-mediated, Stat3-ptyr705
phosphorylation.
Left: Increasing numbers of HC11 cells as indicated, before (lanes 1-4) or after (lanes 5-8)
transfection with the dominant-negative mutant Cdc42 N17 (a gift from Dr Graham Côté) were
plated in 3 cm dishes and lysates probed for Stat3-ptyr705, active Cdc42-GTP, total Cdc42, myc
or Hsp90, as indicated.
218
Right: Increasing numbers of HC11 cells as indicated, before or after transfection with Cdc42N17
and the Stat3-dependent pLucTKS3 and the Stat3-independent pRLSRE reporters, were lysed and
firefly (■) or Renilla ( ) luciferase activities determined (Vultur et al., 2004). Values shown
represent luciferase units expressed as a percentage of the highest value obtained, means ± s.e.m.
of at least 3 experiments, each performed in triplicate.
B: Cdc42 downregulation through expression of an shRNA prevents the density-mediated,
Stat3-tyr705 phosphorylation.
Top: For Cdc42 shRNA expression, cells were transfected with two iLenti-EGFPmCdc42
constructs (Applied Biological Materials), and selected for G418 resistance (see Supplementary
Table S3.4). HC11 cells before (lanes 1-4) or after (lanes 5-8) stable expression of a Cdc42shRNA were grown to different densities and lysates probed for Stat3-ptyr705, active Cdc42GTP (right), total Cdc42 or Hsp90 as a loading control, as indicated.
Bottom: HC11 cells before or after stable expression of a Cdc42-shRNA as indicated, were
transfected with the Stat3-dependent pLucTKS3 and the Stat3-independent pRLSRE reporters,
grown to different densities, then lysed and firefly (■) or Renilla ( ) luciferase activities
determined (Vultur et al., 2004). Values shown represent luciferase units expressed as a
percentage of the highest value obtained, means ± s.e.m. of at least 3 experiments, each
performed in triplicate.
219
Figure A.11
RacV12 and Cdc42V12 activate Stat3 in 10T1/2 cells at all densities
A: Both endogenous and mutant RacV12 protein levels and activity are dramatically
increased by cell density, while RacV12 expression increases Stat3-ptyr705 levels at all cell
densities
10T1/2 fibroblasts (lanes 1-6) or their counterparts stably expressing activated RacV12
(lanes 7-12) were grown to increasing densities, up to 4 days post-confluence. Detergent cell
lysates were probed for total Rac, active Rac-GTP, Stat3-ptyr705, or Hsp90 as a loading control,
as indicated (see Materials and Methods). Numbers at the left refer to molecular weight markers.
B: Both endogenous and mutant Cdc42V12 protein levels and activity are dramatically
increased by cell density, while Cdc42V12 expression increases Stat3-ptyr705 levels at all cell
densities
10T1/2 fibroblasts (lanes 1-7) or their counterparts stably expressing activated Cdc42V12
(lanes 8-14) were grown to increasing densities, up to 4 days post-confluence. Detergent cell
lysates were probed for total Cdc42, active Cdc42-GTP, Stat3-ptyr705, or Hsp90 as a loading
control, as indicated.
220
Figure A.12
Rac-dependent Stat3-ptyr705 phosphorylation requires JAK and NFκB in Balb/c
3T3 cells
A: The IKK inhibitor III inhibits the density-mediated, Stat3, tyr705 phosphorylation in
Balb/c 3T3 cells. Balb/c cells were trypsinised and plated at a high density (2.5x106 cells/3 cm
plate). Following attachment 30 min later, cells were treated for 48 hrs with the IKK inhibitor III
(lane 3), or the DMSO carrier (lane 2). Cell extracts were probed for Stat3-ptyr705, Rac or
Hsp90, as a loading control.
Lane 1: Balb/c cells immediately after attachment to the plastic.
B: JAK inhibitor 1 reduces the Rac-dependent, Stat3-tyr705 phosphorylation in Balb/c 3T3
cells.
Balb/c 3T3 cells were grown to a high density and treated with 0 (lane 1), or 10 (lane 2)
μg/ml JAK inhibitor 1. Cell lysates were probed for Stat3-ptyr705, Rac or Hsp90 as a loading
control, as indicated.
221
Figure A.13
Stat3 can be phosphorylated at critical tyrosine 705 in E-cadherin deficient cell lines
A, top panel: Normal mammary epithelial HC11 cells were grown to different densities and
lysates blotted and probed for E-cadherin (lanes 1-3). Chinese Hamster Ovary cells, while
deficient in E-cadherin (lanes 4-5), and others, do not undergo contact inhibition when they reach
confluence, as a consequence many cells detach from the substratum and remain floating with
apoptotic morphology [(Fiore and Degrassi, 1999); not shown)]. Mouse L fibroblasts do not
express E-cadherin (lanes 6-8), and appear to lack cadherin-based cell adhesion activity (Fukata
and Kaibuchi, 2001). The A431D cell line is a derivative of A431, originating from human
epidermoid carcinoma of the vulva, that has ceased to express E-cadherin (lanes 9-12), as well as
N- or P-cadherins (not shown) due to chronic treatment with dexamethasone (Lewis et al., 1997).
A, bottom panel, and B: E-cadherin expressing HC11 cells, as well as CHO-B1, L cells and
A431D were grown to difference densities and lysates blotted and probed for ptyr-705 Stat3 (A,
bottom, and B), with Hsp90 as a loading control.
Note the dramatic increase in ptyr-705 Stat3 with cell density in L cells, A431D and
CHO-B1, which lack E-cadherin expression. Our results suggest that these cell lines may contain
other, possibly unidentified, cadherins that may mediate the cell confluence dependent Stat3
activation, or that additional mechanisms exist to increase ptyr705-Stat3 by cell density. Full
222
time course experiments indicate that L cells, A431D and CHO-B1 reach their peak of tyr705
phosphorylation much later than HC11, possibly due to their slower growth rate. Interestingly,
total Rac1 levels were also found to increase with confluence in all cell lines (not shown).
A431D cells and L fibroblasts were a gift from Dr. Juliet Daniel, McMaster University.
Chinese Hamster Ovary cells expressing the B1 isoform of murine Fc receptor II (CHO-B1) were
from Dr. Bruce Elliott.
Figure A.14
N-cadherin expression increases Stat3 tyr705 phosphorylation in ES cells
Wild type, ES null cells and null cells where the N-cadherin gene was added back [Ncad-addback
cells (Larue et al., 1996)] were grown to different densities and total cell lysates blotted against
the tyr705, phosphorylated Stat3, and αtubulin as a loading control.
223
Figure A.15
E-cadherin levels increase with cell density
HC11 cells were grown to different densities, as indicated, and lysates blotted and probed for Ecadherin (top) or Hsp90 (bottom), as a loading control. Note the increased E-cadherin levels in
confluent cells, suggesting the possibility that E-cadherin may be degraded at low cell densities,
or that cell density could be increasing transcription from the E-cadherin promoter.
224
Table A.1
A. Quantitation of Rac1, Cdc42 and Stat3 levels in HC11 cells
and their derivatives with reduced Rac1 or Cdc42 protein levels.
confluence
Rac1.GTP
total Rac1
Rac1-GTP/
Cdc42-GTP
total Cdc42
Cdc42-GTP/
Stat3-705
total Stat3
705/total
Line
(%)*
(%)†
(%)‡
total Rac1
(%)§
(%)||
total Cdc42
(%)**
(%)††
Stat3 (%)
HC11
30
13±2
2±1
6.5±3.4
10±3
2±1
5.0±2.9
2±1
45±6
4±2
70
46±3
15±2
3.0±0.4
45±5
12±3
3.8±1.0
15±4
47±5
29±3
80
73±5
45±4
1.6±0.2
65±6
47±7
1.4±0.25
20±8
71±11
35±11
100
85±4
66±5
1.3±0.1
82±4
60±2
1.4±0.01
75±7
78±3
86±9
+3d
99±5
85±5
1.2±0.6
99±1
95±4
1.0±0.01
100±13
90±10
100±12
+7d
100±3
100±5
1.0±0.0
100±16
100±15
1.0±0.2
85±9
100±20
77±17
Line
confluence
Rac1-GTP
total Rac1
Rac1-GTP/
Stat3-705
total Stat3
705/total
(%)*
(%)†
(%)‡
total Rac1
(%)**
(%)††
Stat3 (%)
40
5±1
0±0
0.00±0.00
1±1
29±7
3±1
60
15±1
5±1
3.00±0.63
7±2
35±5
20±7
90
20±1
9±3
2.22±0.75
15±6
55±9
27±12
100
32±5
26±7
1.23±0.38
29±5
60±7
48±10
+3d
39±8
37±12
1.05±0.40
59±7
75±8
79±13
+7d
45±10
47±8
0.95±0.26
64±9
83±9
77±14
Cdc42-GTP/
Stat3-705
HshRac1
confluence
Line
Cdc42-
Total
§
GTP (%)
Cdc42 (%)
40
3±2
60
||
**
total Stat3
(%)
††
705/total
total Cdc42
(%)
0±0
0.00±0.00
1±1
31±3
3±1
14±5
7±2
1.56±0.71
9±3
34±5
26±10
90
24±3
17±6
1.41±0.97
20±5
51±4
39±10
100
35±5
29±8
1.21±0.38
31±7
66±4
47±11
+3d
41±5
45±5
0.91±0.15
60±9
79±10
76±15
+7d
46±7
53±8
0.87±0.55
71±14
80±8
89±20
(%)
HshCdc42
*
Stat3 (%)
B. Quantitation of Rac1, Cdc42 and Stat3 levels in HC11 cells
and their derivatives expressing dominant-negative mutants.
Line
HC11
cell number
Rac1-GTP
total Rac1
Rac1-GTP/
Cdc42-GTP
total Cdc42
Cdc42-GTP/
Stat3-705
total Stat3
705/total
(x106) ‡‡
(%)†
(%)‡
total Rac1
(%)§
(%)||
total Cdc42
(%)**
(%)††
Stat3 (%)
0.1
12±1
2±1
6.0±1.7
11±2
2±1
5.5±1.1
6±2
43±6
14±3
0.5
51±6
25±8
2.0±0.5
49±7
19±8
2.5±0.9
21±5
55±5
38±7
1.5
82±13
65±11
1.3±0.3
76±12
57±13
1.3±0.4
60±9
75±8
80±9
2.5
99±5
89±14
1.1±0.2
99±2
89±11
1.1±0.3
72±7
89±5
81±7
225
3.5
Line
N17
HRac1
Line
HCdc42N17
100±5
cell number
6 ‡‡
100±8
1.0±0.2
Rac1-GTP
†
100±5
Rac1
N17
(%)
‡
100±8
Rac1-GTP/
N17
1.0±0.1
Stat3-705
100±6
total Stat3
705/total
(%)
0.1
2±1
31±5
0.06±0.07
1±1
23±2
4±1
0.5
1±1
45±8
0.02±0.01
7±3
30±5
23±7
1.5
2±1
55±11
0.04±0.02
25±8
51±6
49±9
2.5
2±2
60±14
0.03±0.02
30±14
55±3
54±11
3.5
2±1
100±8
0.02±0.01
45±7
59±7
76±12
Cdc42-GTP/
Stat3-705
total Stat3
705/total
cell number
6 ‡‡
Cdc42-GTP
§
Cdc42
N17
||
N17
**
(%)
††
100±9
(x10 )
Rac1
(%)
**
100±5
(x10 )
(%)
0.1
1±1
35±3
0.03±0.01
2±1
28±3
7±2
0.8
2±1
47±6
0.04±0.02
10±3
31±6
32±8
1.5
2±1
56±9
0.03±0.01
23±9
49±8
47±12
2.5
3±2
86±8
0.03±0.02
29±8
52±4
56±15
3.5
3±1
99±5
0.03±0.01
54±8
62±7
87±11
(%)
Cdc42
(%)
(%)
††
Stat3 (%)
Stat3 (%)
Confluence was estimated visually and carefully quantitated by imaging analysis of live cells
under phase contrast as well as fixed cells stained with Coomasie blue, using a Leitz Diaplan
microscope (Rockleigh, NJ).
*
†
Active Rac1-GTP was quantitated by measuring the amount of Rac1 bound to its effector PAK,
with the levels for HC11 cells at 7 days post confluence taken as 100% (see Materials and
Methods). The results are averages of at least three independent experiments ± s.e.m.
‡
Total Rac1 or Rac1 N17 were measured by Western blotting. The results are averages of at least
three independent experiments ± s.e.m.
§
Active Cdc42-GTP was quantitated by measuring the amount of Cdc42 bound to its effector
PAK with the levels for HC11 cells at 7 days post confluence taken as 100% (see Materials and
Methods). The results are averages of at least three independent experiments ± s.e.m.
||
Total Cdc42 or Cdc42 N17 were measured by Western blotting. The results are averages of at
least three independent experiments ± s.e.m.
**
Stat3-ptyr705 levels were quantitated by fluorimager analysis and normalized to Hsp90 levels,
with the values for HC11 cells at 3 days post-confluence taken as 100%. The results are averages
of at least three independent experiments ± s.e.m. In all cases, the EMSA and transcriptional
activity values obtained paralleled the Stat3-705 phosphorylation levels indicated.
††
Total Stat3 levels were measured by Western blotting. The results are averages of at least three
independent experiments ± s.e.m.
‡‡
Cells were counted 48 hours after plating on 3cm dishes.
226
Table A.2
A. qRT-PCR array for cytokines secreted by densely-growing HC11 cells
Gene
Fold up- or
downregulation
confluent/sparse
Gene
Fold up- or
downregulation
confluent/sparse
Bone morphogenetic protein 1
1.67
Interleukin 1 family, member 5
2.08
Bone morphogenetic protein 10
2.71
Interleukin 1 family, member 6
1.97
Bone morphogenetic protein 2
1.94
Interleukin 1 family, member 8
1.19
Bone morphogenetic protein 3
1.94
Interleukin 1 family, member 9
-1.51
Bone morphogenetic protein 4
1.94
Interleukin 1 receptor antagonist
17.24
Bone morphogenetic protein 5
1.94
Interleukin 2
1.94
Bone morphogenetic protein 6
11.78
Interleukin 20
1.94
Bone morphogenetic protein 7
3.41
Interleukin 21
1.94
Bone morphogenetic protein 8b
3.20
Interleukin 24
-1.77
Colony stimulating factor 1
(macrophage)
7.05
Interleukin 27
8.50
Colony stimulating factor 2
(granulocyte-macrophage)
8.62
Interleukin 3
4.08
Cardiotrophin 1
2.08
Interleukin 4
5.61
Cardiotrophin 2
3.91
Interleukin 7
3.55
Fas ligand (TNF superfamily,
member 6)
1.94
Interleukin 9
3.83
Fibrosin
2.41
Inhibin alpha
2.96
Fibroblast growth factor 10
2.36
Inhibin beta-A
1.85
FMS-like tyrosine kinase 3 ligand
14.30
Left right determination factor 1
3.83
Growth differentiation factor 1
6.18
Leukemia inhibitory factor
9.63
Growth differentiation factor 10
2.13
Lymphotoxin A
5.69
Growth differentiation factor 11
1.31
Lymphotoxin B
4.11
Growth differentiation factor 15
7.35
Macrophage migration inhibitory
factor
2.22
Growth differentiation factor 2
1.94
Secretoglobin, family 3A, member 1
1.89
Growth differentiation factor 3
1.94
Small inducible cytokine subfamily
E, member 1
2.88
227
Growth differentiation factor 5
2.55
Tumour necrosis factor
4.25
Myostatin
1.94
Tumour necrosis factor receptor
superfamily member 11b
(osteoprotegerin)
2.22
Growth differentiation factor 9
4.62
Tumour necrosis factor (ligand)
superfamily, member 10
1.80
Interferon alpha 2
1.94
Tumour necrosis factor (ligand)
superfamily, member 11
51.91
Interferon alpha 4
1.59
Tumour necrosis factor (ligand)
superfamily, member 12
2.69
Interferon beta 1, fibroblast
5.42
Tumour necrosis factor (ligand)
superfamily, member 13
8.99
Interferon gamma
-1.10
Tumour necrosis factor (ligand)
superfamily, member 13b
3.24
Interleukin 10
-2.15
Tumour necrosis factor (ligand)
superfamily, member 14
3.50
Interleukin 11
2.22
Tumour necrosis factor (ligand)
superfamily, member 15
11.54
Interleukin 12b
2.39
Tumour necrosis factor (ligand)
superfamily, member 18
2.64
Interleukin 13
1.61
Tumour necrosis factor (ligand)
superfamily, member 4
1.94
Interleukin 15
6.44
CD40 ligand
1.94
Interleukin 16
9.18
CD70 antigen
10.99
Interleukin 17b
1.81
Tumour necrosis factor (ligand)
superfamily, member 8
1.94
Interleukin 17c
1.94
Tumour necrosis factor (ligand)
superfamily, member 9
2.11
Interleukin 25
1.94
Taxilin alpha
3.27
Interleukin 17f
2.86
Glucuronidase, beta
1.48
Interleukin 18
30.02
Hypoxanthine guanine
phosphoribosyl transferase 1
1.32
Interleukin 19
2.49
Heat shock protein 90 alpha
(cytosolic), class B member 1
-1.53
Interleukin 1 alpha
4.75
Glyceraldehyde-3-phosphate
dehydrogenase
1.52
228
Interleukin 1 beta
18.61
Actin, beta
-1.95
Interleukin 1 family, member 10
1.94
Interleukin 6
76.53
B. qRT-PCR array for cytokines secreted by densely-growing Balb/c 3T3 cells
Gene
Fold up- or
downregulation
confluent/sparse
Gene
Fold up- or
downregulation
confluent/sparse
Bone morphogenetic protein 1
4.53
Interleukin 1 family, member 5
2.58
Bone morphogenetic protein 10
2.23
Interleukin 1 family, member 6
40.60
Bone morphogenetic protein 2
2.23
Interleukin 1 family, member 8
2.23
Bone morphogenetic protein 3
2.23
Interleukin 1 family, member 9
2.48
Bone morphogenetic protein 4
2.95
Interleukin 1 receptor antagonist
44.02
Bone morphogenetic protein 5
3.23
Interleukin 2
2.23
Bone morphogenetic protein 6
3.56
Interleukin 20
2.23
Bone morphogenetic protein 7
2.23
Interleukin 21
2.23
Bone morphogenetic protein 8b
2.23
Interleukin 24
1.73
Colony stimulating factor 1
(macrophage)
1.43
Interleukin 27
2.38
Colony stimulating factor 2
(granulocyte-macrophage)
9.32
Interleukin 3
2.23
Cardiotrophin 1
-1.11
Interleukin 4
2.68
Cardiotrophin 2
1.22
Interleukin 7
2.23
Fas ligand (TNF superfamily,
member 6)
2.23
Interleukin 9
2.23
Fibrosin
1.56
Inhibin alpha
6.54
Fibroblast growth factor 10
2.23
Inhibin beta-A
1.69
FMS-like tyrosine kinase 3 ligand
6.15
Left right determination factor 1
4.20
Growth differentiation factor 1
-1.85
Leukemia inhibitory factor
11.96
Growth differentiation factor 10
2.23
Lymphotoxin A
2.23
Growth differentiation factor 11
2.43
Lymphotoxin B
2.23
Growth differentiation factor 15
-1.24
Macrophage migration inhibitory
factor
4.29
229
Growth differentiation factor 2
2.23
Secretoglobin, family 3A, member 1
1.88
Growth differentiation factor 3
2.23
Small inducible cytokine subfamily
E, member 1
1.03
Growth differentiation factor 5
-2.11
Tumour necrosis factor
2.23
Myostatin
2.23
Tumour necrosis factor receptor
superfamily member 11b
(osteoprotegerin)
1.30
Growth differentiation factor 9
3.12
Tumour necrosis factor (ligand)
superfamily, member 10
11.79
Interferon alpha 2
2.23
Tumour necrosis factor (ligand)
superfamily, member 11
2.23
Interferon alpha 4
2.23
Tumour necrosis factor (ligand)
superfamily, member 12
2.57
Interferon beta 1, fibroblast
2.25
Tumour necrosis factor (ligand)
superfamily, member 13
7.36
Interferon gamma
2.23
Tumour necrosis factor (ligand)
superfamily, member 13b
2.23
Interleukin 10
2.83
Tumour necrosis factor (ligand)
superfamily, member 14
6.87
Interleukin 11
3.71
Tumour necrosis factor (ligand)
superfamily, member 15
2.23
Interleukin 12b
2.23
Tumour necrosis factor (ligand)
superfamily, member 18
2.23
Interleukin 13
2.23
Tumour necrosis factor (ligand)
superfamily, member 4
2.23
Interleukin 15
6.59
CD40 ligand
2.23
Interleukin 16
6.87
CD70 antigen
3.14
Interleukin 17b
3.10
Tumour necrosis factor (ligand)
superfamily, member 8
2.23
Interleukin 17c
2.23
Tumour necrosis factor (ligand)
superfamily, member 9
1.91
Interleukin 25
2.23
Taxilin alpha
-1.13
Interleukin 17f
1.11
Glucuronidase, beta
-2.03
Interleukin 18
6,41
Hypoxanthine guanine
phosphoribosyl transferase 1
1.10
230
Interleukin 19
1.57
Heat shock protein 90 alpha
(cytosolic), class B member 1
-1.32
Interleukin 1 alpha
2.23
Glyceraldehyde-3-phosphate
dehydrogenase
3.48
Interleukin 1 beta
2.23
Actin, beta
-1.43
Interleukin 1 family, member 10
2.58
Interleukin 6
32.67
231
Table A.3
Stat3 tyr-705 phosphorylation and apoptosis following cadherin ablation
or inhibition with peptides*
Cell line E-cadherin levels (%)† Peptide‡ Confluence (%)§ Stat3-ptyr705 (%)|| Cells in subG1 (%)**
Ecad+/+
addback
null
HC11
100±7
-
90
66±6
2±1
“
-
100
73±5
5±2
“
-
100+2days
100±9
6±3
25±8
-
90
27±3
3±2
“
-
100
35±4
22±2
“
-
63±3
35±9
0
-
90
8±5
5±1
“
-
100
10±7
45±9
“
-
100+2 days
20±5
71±5
100±5
-
100+1 day
100±7
5±2
“
SHAVSA
100+1 day
9±3
65±5
“
SHAVSS
100+1 day
72±4
10±4
“
SHGVSA
100+1 day
97±4
7±8
100+1 day
263±33
3±1
-
100+2 days
HC11-Stat3C
“
HC11-Stat3C
“
SHAVSA
100+1 day
239±29
4±1
HC11-Stat3C
“
SHGVSA
100+1 day
270±24
4±1
Lysates from the indicated lines grown to different degrees of confluence and up to 2 days post
confluence as indicated, with or without treatment with the peptide inhibitors for 24 hours, were
resolved by gel electrophoresis. Western immunoblots were probed for the tyr705
phosphorylated form of Stat3 or for E-cadherin. Averages of at least three experiments ± s.e.m.
are shown (see Materials and Methods).
*
232
†
E-cadherin levels were quantitated by fluorimager analysis and normalized to Hsp90 levels.
The results are averages of at least three independent experiments ± s.e.m. with the value for
Ecad+/+ or untreated HC11 cells taken as 100%.
‡
HC11 cells were treated with 10 mM of the indicated peptides for 24 hours (see Materials and
Methods). The results are averages of at least three independent experiments ± s.e.m.
§
Confluence was estimated visually and carefully quantitated by imaging analysis of live cells
under phase contrast as well as fixed cells stained with Coomasie blue, using a Leitz Diaplan
microscope (Rockleigh, NJ).
||
Stat3-ptyr705 levels were quantitated by fluorimager analysis and normalized to Hsp90 levels,
with the values for Ecad+/+ cells or HC11 cells at 1 day post-confluence taken as 100%. The
results are averages of at least three independent experiments ± s.e.m. In all cases, the EMSA
and transcriptional activity values obtained paralleled the Stat3-705 phosphorylation levels
indicated.
**
To quantitate apoptosis, cells were stained with propidium iodide and their sub-G1 profile
analysed by FACS analysis. Numbers refer to cells in subG1 as a % of total cells. The results are
averages of at least three independent experiments ± s.e.m.
233
Table A.4
shRNA sequences
Target shRNA
OligoID
shRac1
Open Biosystems
V2MM_7967
shgp130-1
Open Biosystems
V2MM_219118
shgp130-2
Open Biosystems
V2MM_77377
shgp130-3
Open Biosystems
V2MM_70734
shgp130-4
Open Biosystems
V2MM71336
shCdc42-a
shCdc42-b
Applied Biological
Materials
mCdc42/i000053a
Applied Biological
Materials
mCdc42-404/i000053b
shCdh11-1
Open Biosystems
V2MM-66004
shCdh11-2
Open Biosystems
V2MM-75136
shCdh11-3
Open Biosystems
V2MM-81396
shCdh11-4
Open Biosystems
V2MM-66370
Sequence
TGCTGTTGACAGTGAGCGsense: AGGCGTTGAGTCCATATTTAAA
loop: TAGTGAAGCCACAGATGTA
antisense: TTTAAATATGGACTCAACGCCC
-TGCCTACTGCCTCGGA
TGCTGTTGACAGTGAGsense: CGCCCTCCCTTTGCTCTAAGAGAA
loop: TAGTGAAGCCACAGATGTA
antisense: TTCTCTTAGAGCAAAGGGAGGTT
-GCCTACTGCCTCGGA
TGCTGTTGACAGTGAGCGsense: CCCGACTGTTCGGACAAAGAAA
loop: TAGTGAAGCCACAGATGTA
antisense: TTTCTTTGTCCGAACAGTCGGTT
-GCCTACTGCCTCGGA
TGCTGTTGACAGTGAGCGsense: CCCAGAATGGCTTCATTAGAAA
loop: TAGTGAAGCCACAGATGTA
antisense:TTTCTAATGAAGCCATTCTGGT
-TGCCTACTGCCTCGGA
TGCTGTTGACAGTGAGCGsense: CGCTGCTTATTCTGTAGTGAAT
loop: TAGTGAAGCCACAGATGTA
antisense:ATTCACTACAGAATAAGCAGCT
-TGCCTACTGCCTCGGA
5’-GCAATGAGTGCTAGTTTTTTT-3’
5’-GCCTATTACTCCAGAGACT-3’
TGCTGTTGACAGTGAGCsense: GACGTGAGAACATCATAACCTAT
loop: TAGTGAAGCCACAGATGTA
antisense: ATAGGTTATGATGTTCTCACGG
-TGCCTACTGCCTCGGA
TGCTGTTGACAGTGAGCsense: GCGCCAACAGCCCAATAAGGTAT
loop: TAGTGAAGCCACAGATGTA
antisense: ATACCTTATTGGGCTGTTGGCA
-TGCCTACTGCCTCGGA
TGCTGTTGACAGTGAGCsense: GCGGTATTCAATTGATCGTCATA
loop: TAGTGAAGCCACAGATGTA
antisense: TATGACGATCAATTGAATACCT
-TGCCTACTGCCTCGGA
TGCTGTTGACAGTGAGCsense: GAGCACTCTCCAACCAGCCAATA
loop: TAGTGAAGCCACAGATGTA
antisense: TATTGGCTGGTTGGAGAGTGCCTGCC
-TACTGCCTCGGA
234
Table A.5
qRT-PCR array for cytokines secreted by H-Racv12cells
Fold Change
H-Racv12/HC11
Gene
Fold Change
HRacv12/HC11
Bone morphogenetic protein 1
1.54
Interleukin 1 family, member 5
0.42
Bone morphogenetic protein 10
1.40
Interleukin 1 family, member 6
9.33
Bone morphogenetic protein 2
0.90
Interleukin 1 family, member 8
0.17
Bone morphogenetic protein 3
1.40
Interleukin 1 family, member 9
3.21
Bone morphogenetic protein 4
0.17
Interleukin 1 receptor antagonist
7.07
Bone morphogenetic protein 5
0.73
Interleukin 2
0.20
Bone morphogenetic protein 6
0.94
Interleukin 20
0.33
Bone morphogenetic protein 7
1.44
Interleukin 21
1.40
Bone morphogenetic protein 8b
1.40
Interleukin 24
4.32
Colony stimulating factor 1
(macrophage)
0.87
Interleukin 27
0.69
Colony stimulating factor 2
(granulocyte-macrophage)
2.27
Interleukin 3
0.20
Cardiotrophin 1
1.04
Interleukin 4
0.41
Cardiotrophin 2
0.83
Interleukin 7
0.61
Fas ligand (TNF superfamily,
member 6)
1.40
Interleukin 9
0.21
Fibrosin
1.00
Inhibin alpha
0.25
Fibroblast growth factor 10
1.40
Inhibin beta-A
1.94
FMS-like tyrosine kinase 3 ligand
1.42
Left right determination factor 1
0.33
Growth differentiation factor 1
0.83
Leukemia inhibitory factor
9.92
Growth differentiation factor 10
0.46
Lymphotoxin A
0.35
Growth differentiation factor 11
1.30
Lymphotoxin B
1.15
Growth differentiation factor 15
0.88
Macrophage migration inhibitory
factor
1.57
Growth differentiation factor 2
1.40
Secretoglobin, family 3A, member 1
0.87
Growth differentiation factor 3
1.40
Small inducible cytokine subfamily E,
member 1
1.41
Gene
235
Growth differentiation factor 5
0.76
Tumour necrosis factor
2.31
Myostatin
0.93
Tumour necrosis factor receptor
superfamily member 11b
(osteoprotegerin)
0.59
Growth differentiation factor 9
1.91
Tumour necrosis factor (ligand)
superfamily, member 10
0.52
Interferon alpha 2
0.24
Tumour necrosis factor (ligand)
superfamily, member 11
1.40
Interferon alpha 4
2.33
Tumour necrosis factor (ligand)
superfamily, member 12
0.52
Interferon beta 1, fibroblast
2.32
Tumour necrosis factor (ligand)
superfamily, member 13
1.51
Interferon gamma
1.40
Tumour necrosis factor (ligand)
superfamily, member 13b
1.49
Interleukin 10
0.60
Tumour necrosis factor (ligand)
superfamily, member 14
0.37
Interleukin 11
0.70
Tumour necrosis factor (ligand)
superfamily, member 15
1.40
Interleukin 12b
0.22
Tumour necrosis factor (ligand)
superfamily, member 18
0.86
Interleukin 13
0.16
Tumour necrosis factor (ligand)
superfamily, member 4
1.40
Interleukin 15
1.42
CD40 ligand
1.17
Interleukin 16
0.93
CD70 antigen
0.31
Interleukin 17b
0.63
Tumour necrosis factor (ligand)
superfamily, member 8
1.37
Interleukin 17c
1.40
Tumour necrosis factor (ligand)
superfamily, member 9
1.39
Interleukin 25
0.68
Taxilin alpha
1.00
Interleukin 17f
1.06
Glucuronidase, beta
0.82
Interleukin 18
0.97
Hypoxanthine guanine
phosphoribosyl transferase 1
0.84
Interleukin 19
0.22
Heat shock protein 90 alpha
(cytosolic), class B member 1
1.24
Interleukin 1 alpha
30.56
Glyceraldehyde-3-phosphate
dehydrogenase
1.13
236
Interleukin 1 beta
9.73
Actin, beta
1.04
Interleukin 1 family, member 10
1.40
Interleukin 6
17.50
References
Fiore, M. and Degrassi, F. (1999). Dimethyl sulfoxide restores contact inhibition-induced growth
arrest and inhibits cell density-dependent apoptosis in hamster cells. Exp. Cell Res. 251, 102-110.
Fukata, M. and Kaibuchi, K. (2001). Rho-family GTPases in cadherin-mediated cell-cell
adhesion. Nat. Rev. Mol. Cell Biol. 2, 887-897.
Gao, Y., Dickerson, J. B., Guo, F., Zheng, J., and Zheng, Y. (2004). Rational design and
characterization of a Rac GTPase-specific small molecule inhibitor. Proc. Natl. Acad. Sci. U. S.
A 101, 7618-7623.
Larue, L., Antos, C., Butz, S., Huber, O., Delmas, V., Dominis, M., and Kemler, R. (1996). A
role for cadherins in tissue formation. Development 122, 3185-3194.
Lewis, J. E., Wahl, J. K., III, Sass, K. M., Jensen, P. J., Johnson, K. R., and Wheelock, M. J.
(1997). Cross-talk between adherens junctions and desmosomes depends on plakoglobin. J. Cell
Biol. 136, 919-934.
Vultur, A., Arulanandam, R., Turkson, J., Niu, G., Jove, R., and Raptis, L. (2005). Stat3 is
required for full neoplastic transformation by the Simian Virus 40 Large Tumor antigen.
Molecular Biology of the Cell 16, 3832-3846.
Vultur, A., Cao, J., Arulanandam, R., Turkson, J., Jove, R., Greer, P., Craig, A., Elliott, B. E., and
Raptis, L. (2004). Cell to cell adhesion modulates Stat3 activity in normal and breast carcinoma
cells. Oncogene 23, 2600-2616.
237