sign in sign up
<<
Home , CDC42, DLC1, GNA13, Rac3, Radixin, Ras GTPase

A Dissertation

entitled

A Mechanistic Investigation of the Role of RhoA and RKIP in Breast Invasion

and

by

Gardiyawasam Kalpana

Submitted to the Graduate Faculty as partial fulfillment of the requirements for the

Doctor of Philosophy Degree in

Cancer Biology

______Dr. Kam C. Yeung, Committee Chair

______Dr. Kathryn Eisenmann, Committee Member

______Dr. Saori Furuta, Committee Member

______Dr. Ivana de la Serna, Committee Member

______Dr. Rafael Garcia-Mata, Committee Member

______Dr. Cyndee Gruden, Dean College of Graduate Studies

The University of Toledo

August 2019

Copyright 2019 Gardiyawasam Kalpana

This document is copyrighted material. Under copyright law, no parts of this document may be reproduced without the expressed permission of the author.

An Abstract of

A Mechanistic Investigation for the Role of RhoA and RKIP in Invasion

and Metastasis

by

Gardiyawasam Kalpana

Submitted to the Graduate Faculty as partial fulfillment of the requirements for the Doctor of Philosophy Degree in Cancer Biology

The University of Toledo August 2019

Tumor metastasis suppressors impede secondary tumor formation by inhibiting one or more steps of the metastasis cascade without stimulating primary tumor growth. Raf-1 inhibitor (RKIP) is a metastasis suppressor that inhibits breast and prostate cancer metastasis. The molecular mechanism through which RKIP executes its anti- metastasis effects is not yet completely defined. The objective of the current study is to understand how RKIP inhibits breast cancer invasion and metastasis at the molecular level. Given their primary functions in dynamics and cell movement regulation, Rho were studied as possible downstream effectors of RKIP. Among all Rho GTPases, RhoA has both pro-metastatic and anti-metastatic cell-context dependent functions. In this study, we demonstrate that the increased anti-metastatic activity of RhoA is one of the causes of RKIP-mediated suppression of breast cancer metastasis. Upon

RNAi-mediated RhoA knockdown, more tumor cells were detected in the sentinel lymph node and more colonies in the lungs, suggesting that RhoA suppresses breast cancer lymph node and lung metastasis. Further characterizations showed that RhoA suppresses

iii breast cancer lung metastasis partly due to upstream RKIP-dependent signaling and this is mediated through downstream E-cadherin and CCL5. Additionally, RhoA’s effect on tumor chemokine expression and infiltration of SMA+ cells are independent of upstream RKIP signaling and might be mediated by RKIP-independent upstream RhoA regulators.

iv

Education is the only asset that will remain with you forever.

I dedicate my dissertation to my dear husband, Suneth Watthage and my loving parents,

Wilson and Shanthi!

v

Acknowledgments

First, I would like to express my sincere gratitude and appreciation to my mentor,

Dr. Kam Yeung, for his continuous encouragement and guidance, prudent advice and sustenance throughout my graduate research work. He has been greatly supportive and patient during obstacles and challenges in my research life as well as in personal life. I could not have hoped for a better advisor, and I sincerely recognize him for that.

I am sincerely grateful to my thesis committee, Dr. Ivana de la Serna, Dr. Kathryn

Eisenmann, Dr. Saori Furuta and Dr. Rafael Garcia-Mata for their valuable advice, support, and feedback throughout this journey. My sincere thanks go to Dr. Kandace Williams for her support and guidance. I would like to convey my sincere thanks to all the current members and the alumni of the Yeung group for their support throughout the years.

In addition, I would like to thank Dr. Rafael Garcia-Mata lab and the Dr. Daya

Raman lab for their contributions to my research work. I would also like to express my immense gratitude to my dearest husband and my loving parents for their unconditional love and support, without which this would be an impossible goal to achieve. Additionally,

I would like to thank all the faculty members, staff, and students in the Department of

Cancer Biology.

vi

Table of Contents

Abstract ...... iii

Acknowledgments...... vi

Table of Contents ...... vii

List of Figures ...... xi

List of Abbreviations ...... xiii

List of Symbols ...... xv

1 An Introduction to Rho GTPases, RKIP, Breast Cancer and Tumor

microenvironment ...... 1

1.1 of GTPases ...... 1

1.1.1 Rho GTPase Family ...... 3

1.1.1.1 RhoA GTPase and RhoA-like GTPases sub-family ...... 4

1.1.1.2 Rac1 GTPase and Rac1-like GTPase sub-family ...... 5

1.1.1.3 Cdc42 GTPase and Cdc42-like GTPase sub-family ...... 6

1.1.2 Regulation of Rho GTPases ...... 6

1.1.2.1 Regulation by RhoGEFs ...... 7

1.1.2.2 Regulation by RhoGAPs ...... 8

1.1.2.3 Regulation by RhoGDIs ...... 8

1.1.2.4 Transcriptional and Post-transcriptional Regulatory

Mechanisms ...... 8 vii 1.1.2.5 Regulation by Post-translational Modifications ...... 10

1.1.3 Rho GTPase Effectors and their Functions ...... 12

1.1.3.1 Actin Cytoskeleton Dynamics ...... 12

1.1.3.2 Cytoskeleton Dynamics ...... 15

1.1.3.3 Rho GTPases as Regulators of ...... 15

1.1.3.4 Rho GTPases as Regulators of Progression .....17

1.1.3.5 Rho GTPases as Regulators of and Invasion

...... 18

1.1.3.6 Rho GTPases as Regulators of Cell-cell Adhesions ...... 19

1.1.4 Rho GTPases in Cancer ...... 22

1.1.4.1 RhoA as an ...... 23

1.1.4.2 RhoA as a Tumor Suppressor ...... 26

1.2 Raf-1 Kinase Inhibitor Protein (RKIP) ...... 29

1.2.1 RKIP as a Signaling ...... 30

1.2.2 Metastasis Cascade and RKIP as a Metastasis Suppressor ...... 32

1.3 Breast Cancer ...... 34

1.3.1 Molecular Sub-types ...... 34

1.3.2 Breast Cancer Metastasis ...... 36

1.4 Tumor Microenvironment ...... 38

1.4.1 Cancer-associated Fibroblasts and Tumor Metastasis ...... 38

2 Reduced RhoA Expression Enhances Breast Cancer Metastasis with a

Concomitant Increase in CCR5 and CXCR4 Chemokines Signaling ...... 41

2.1 Abstract ...... 42

viii 2.2 Introduction ...... 42

2.3 Materials and Methods ...... 44

2.4 Results ...... 49

2.4.1 RhoA Suppresses Breast Cancer Cell Invasion in vitro ...... 49

2.4.2 RhoA Suppresses Breast Cancer Lung Metastasis Burden in Mice .52

2.4.3 RhoA Suppresses Breast Cancer Sentinel Lymph Node Metastasis in

Mice ...... 55

2.4.4 RhoA Suppresses Breast Cancer Cell Invasion by Modulating the

Tumor Microenvironment ...... 58

2.5 Discussion ...... 63

2.6 Supplementary Figures ...... 67

3 Anti-metastatic Role of RhoA and RKIP in Breast Cancer ...... 72

3.1 Abstract ...... 72

3.2 Introduction ...... 72

3.3 Material and Methods ...... 74

3.4 Results ...... 80

3.4.1 RKIP Promotes RhoA GTPase Activation in Breast Cancer Cells ..80

3.4.2 RKIP Required Downstream RhoA to Suppress Breast Cancer Cell

Invasion and Metastasis ...... 81

3.4.3 RKIP Suppresses Breast Cancer Cell Invasion Through RhoA-

mediated Regulation of E-cadherin ...... 88

3.4.4 RKIP suppresses Breast Cancer Metastasis Partially Through

Downstream RhoA-regulated Mechanisms ...... 97

ix 3.4.5 RKIP Activates RhoA Through ERK2 and GEFH1 ...... 101

3.5 Discussion ...... 106

4 Summary ...... 109

References ...... 113

x

List of Figures

1 – 1 The Rho GTPases cycle ...... 2

1 – 2 Phylogenetic tree showing the mammalian Ras GTPase superfamily ...... 4

1 – 3 Schematic representation of the basic structural components of the adherens

junctions ...... 22

1 – 4 Schematic model of organ-specific metastatic extravasation of breast cancer cells 39

2 – 1 RhoA suppresses breast cancer cell invasion in vitro ...... 52

2 – 2 RhoA suppresses breast cancer lung metastasis burden in mice ...... 54

2 – 3 RhoA suppresses breast cancer sentinel lymph node metastasis in mice ...... 57

2 – 4 RhoA suppresses breast cancer cell invasion by modulating the tumor

microenvironment ...... 62

3 – 1 RKIP promotes RhoA activation in breast cancer cells ...... 86

3 – 2 RKIP required downstream RhoA to suppresses breast cancer cell invasion ...... 87

3 – 3 RKIP required downstream RhoA to suppresses breast cancer metastasis ...... 90

3 – 4 RKIP suppresses breast cancer cell invasion through RhoA-mediated regulation of

E-cadherin ...... 92

3 – 5 RKIP’s effect on E-cadherin is breast cancer cell-type specific ...... 94

3 – 6 RKIP promotes E-cadherin membrane localization through downstream RhoA ....96

3 – 7 RKIP promotes E-cadherin membrane association with β-catenin and p120-catenin

in adherens junctions...... 97 xi 3 – 8 RKIP-mediated inhibition of invasion depends on the membrane E-cadherin

localization ...... 98

3 – 9 RKIP suppresses breast cancer metastasis partially through downstream E-cadherin

...... 101

3 – 10 RKIP suppresses breast cancer metastasis partially through downstream CCL5 and

F4/80 ...... 103

3 – 11 RKIP activates RhoA through ERK2 ...... 105

3 – 12 RKIP activates RhoA through GEFH1 ...... 108

4 – 1 Schematic representation of the research findings ...... 115

xii

List of Abbreviations

α-SMA ...... Alpha Smooth Muscle Actin

AITL ...... Angioimmunoblastic T-cell lymphoma AJ ...... Adherens junction Arp2/3 ...... Actin-Related Protein 2/3

CAF ...... Cancer-Associated Fibroblasts CCL ...... C-C motif CXCL ...... C-X-C motif ligand

Dbl...... diffuse B-cell lymphoma DH ...... Dbl homology DHR2 ...... DOCK homology region 2 DOCK ...... Dedicator of family DRF ...... Diaphanous-related formin

E-cadherin ...... Epithelial Cadherin ECT2 ...... Epithelial cell transforming sequence 2 EGF ...... Epidermal EMT ...... Epithelial to Mesenchymal Transition ER ...... Estrogen receptor ERK...... Extracellular signal-regulated kinase

FAK...... Focal adhesion kinase

GAP...... GTPase activating protein GDI ...... Guanine dissociation inhibitor GDP...... diphosphate GEF ...... Guanine nucleotide exchange factor GPCR ...... G-protein-coupled receptor GTP ...... GTPase ...... Guanosine

LIMK ......

MAPK ...... activated xiii mDia ...... mammalian Diaphanous Formin MIM ...... Missing in metastasis MLC ...... light chain

PAK...... -activated kinase PH ...... Pleckstrin homology PKA...... PR ...... receptor

RNA ...... Ribonucleic acid ROCK ...... Rho-associated protein kinase Rop ...... Rho of plants ROS ...... Reactive oxygen species

SRF ...... Serum response factor STAT...... Signal transducers and activators of

TGFβ ...... Transforming growth factor beta

VSMC ...... Vascular smooth muscle

WASP ...... Wiskott-Aldrich syndrome protein WAVE...... WASP-family verprolin-homology protein

xiv

List of Symbols

α ...... Alpha β ...... Beta γ ...... Gamma κ...... Kappa

µ ...... Micro sign

xv

Chapter 1

An Introduction to Rho GTPases, RKIP, Breast Cancer and Tumor Microenvironment

1.1 Ras superfamily of GTPases

In humans, the Ras superfamily of guanosine triphosphatases (GTPases) consists of over 150 and include evolutionarily conserved orthologs in all eukaryotes from

S. cerevisiae to plants. Depending on their sequence and functional similarities they are further divided into five major branches; Ras, Rho, , , and Arf[1].

One common feature of Ras superfamily proteins is their ability to bind GTP/GDP and the intrinsic GTP hydrolysis ability. This is characterized by the structurally and mechanistically conserved N-terminal G box motifs, G1-G5, comprising the G domain[1].

The functionality of these GTPases depends on the bound nucleotide status, hence they act as molecular switches, which shift between active and inactive conformations. (Fig 1.1)

The conformational changes are primarily confined to two loop regions named switch I and switch II. In the GTP-bound state, the conformational changes in these regions are 1

sensed and subsequently bound by proteins, thus activating the downstream signaling events. But these effector-bound conformations are transient in nature. Upon hydrolysis of the GTP and release of the gamma-, the conformation changes, thus releasing the effector and attenuating the downstream signaling. In addition to the slow intrinsic GTP hydrolysis ability of GTPases, a group of proteins called GTPase activating proteins (GAPs) are involved to enhance the GTP hydrolysis process. The exchange of bound GDP to GTP is the rate-limiting step of the Ras signaling cycle. Another class of regulatory proteins named guanine nucleotide exchange factors (GEFs) catalyze the substitution of GDP with GTP, thereby regulating the Ras signaling in a spatial and temporal manner. Each Ras protein can be recognized by several GEF proteins, while a specific GEF can stimulate several GTPases, further complexing the Ras signaling events[1, 2].

Figure 1-1: The Rho GTPase cycle. (Reprinted by permission from [Springer Nature

Customer Service Centre GmbH]: [Springer Nature] [Rho GTPases in cell biology]

2

[REFERENCE CITATION (Rho GTPases in cell biology, Sandrine Etienne-

Manneville et al), [COPYRIGHT] (2002)

Another group of proteins named guanine nucleotide dissociation inhibitors (GDIs) function differently from GEFs to regulate the GTPase cycle. They bind to the GDP-bound

GTPase conformation and prevent the release of the nucleotide, thereby acting as negative regulators of GTPase signaling. Additionally, they act as chaperons, by binding and shielding the hydrophobic C- terminal phenyl modifications of GTPases, thereby making it possible to sequester and maintain an inactive soluble reservoir of GTPases in the , and shuttle them between different sub-cellular and plasma membranes upon demand[2,

3].

1.1.1 Rho GTPase family

Ras homology (Rho) GTPases similar to other Ras superfamily proteins, function as major signaling nodes, which transduce signals from upstream cell surface receptors to downstream signaling pathways[4]. The first Rho gene was isolated from the marine snail,

Aplysia californica in 1985, and found to have a 35% to classical Ras proteins[5]. Soon after, this sequence information was used to identify three Rho homolog in mammals, RhoA, RhoB and RhoC. Since then, about twenty Rho GTPases have been identified in mammals, and these evolutionarily conserved proteins are classified into eight different sub-families depend on their sequence, functional and biochemical characteristics. (Fig 1.2) In addition to these conserved Rho proteins, RhoS expresses in rodent testis,[6] and Rop proteins (Rho of plants) distinctly expresses in plants[7]. 3

Among these twenty GTPases, RhoA, Rac1 and Cdc42 have been extensively studied in terms of their physiological functions, and mechanisms and regulation.

Figure 1-2: Phylogenetic tree showing the mammalian Ras GTPase superfamily.

(Reprinted by permission from [John Wiley and Sons]: [John Wiley and Sons] [Rho

GTPases in cancer cell biology] [REFERENCE CITATION (Rho GTPases in cancer cell biology, Francisco M. Vega, Anne J. Ridley), [COPYRIGHT] (2008)

1.1.1.1 RhoA GTPase and RhoA-like GTPases sub-family

4

RhoA-like GTPase sub-family includes three highly conserved GTPases, RhoA,

RhoB and RhoC. These isoforms share about 85% of sequence homology with some minor sequence diversity in the “insert loop” sequence, a short helix that is only present in some of the Rho proteins but not in any other Ras superfamily members, and some major diversity in the C-terminus. The C-terminus sequence determines the localization of the

GTPase in the cell, hence sequence divergence in this region indicates differential regulation of their localization[8]. RhoA and RhoC are mainly located on the plasma membrane and the cytoplasm, while RhoB is mainly localized on late endosomes and lysosomes[9]. Similar to other Ras proteins, these Rho GTPases are post-translationally modified by at the C-terminus. The specific type of prenyl moiety differs between Rho proteins, and they are important in determining the specific localization of these proteins. Prenylation is essential for the stability and the correct functionality of Rho proteins as well[8].

In addition to the differences in the sequence and the localization, the three Rho

GTPase proteins are different in terms of their physiological functions. According to over- expression studies, all three proteins can induce stress fibers in cells. Yet, RhoC is preferentially upregulated in cancer and is believed to have an oncogenic effect, while

RhoB is mostly down-regulated in cancer and is suggested to function as a tumor suppressor. Similarly, apart from its common effects on actin organization and cell motility with RhoA and RhoC, RhoB has a unique function in endosome trafficking. Hitherto,

RhoA has a tumor-type and cell-type specific effect on tumorigenesis and tumor

5

progression, and both tumor promoting and tumor suppressive roles for RhoA has been reported[10, 11].

1.1.1.2 Rac1 GTPase and Rac1-like GTPase sub-family

Rac1-like sub-family consists of Rac1, Rac2, and RhoG GTPases. Among these, Rac1 is the most studied member of the group. It was first identified in 1989 as Ras- related C3 botulinum toxin 1[12]. Since then, a variety of physiological functions for Rac1 has been identified in actin cytoskeleton rearrangements, cell migration and adhesion, cell transformation, reactive oxygen species (ROS) production, and axonal growth[13, 14].

1.1.1.3 Cdc42 GTPase and Cdc42-like GTPase sub-family

Cdc42-like GTPase sub-family includes Cdc42 and closely related proteins

RhoJ/TCL and RhoQ/TC10[15]. Cdc42 plays a major cellular function in maintaining epithelial as well as polarity during cell migration. Additionally, Cdc42 and other Cdc42-like proteins are important for formation, segregation during and vesicle trafficking, and suggest a functional redundancy between

Cdc42-like family proteins[15, 16].

1.1.2 Regulation of Rho GTPases

Rho GTPases act as major nodes in pathways and transduce signals from upstream cell-surface receptors including GPCR, cytokine receptors, tyrosine , and adhesion receptors, to downstream targets. Hence, their activities need to be 6

tightly regulated to maintain cellular homeostasis, and this is achieved by several layers of regulatory mechanisms. Most of the Rho GTPases switch between an active and inactive conformation, and regulation of this cycle by Rho GTPase regulators as RhoGEFs,

RhoGAPs, and RhoGDIs is a major point of control. Yet, some “atypical” Rho GTPases, such as the members of the Rnd subfamily (, , RhoE, RhoH), exist in a constitutively active GTP-bound state, either due to the low intrinsic GTPase activity or due to the high intrinsic nucleotide exchange activity, and therefore are not regulated by

Rho regulators. Hence additional regulation mechanisms exist to modulate the expression and functionality of Rho GTPases as well as Rho GTPase regulators at the post- transcriptional and post-translational levels[17].

1.1.2.1 Regulation by RhoGEFs

The first member of the mammalian RhoGEF family was identified as a transforming oncogene in diffuse B-cell lymphoma cells, hence named as Dbl[18, 19].

Since then about 69 more RhoGEFs in the Dbl sub-family has been identified, and they all are characterized by two sequential, conserved domains, a DH (Dbl homology) catalytic

RhoGEF domain and a PH (pleckstrin homology) membrane-binding domain. The DH-PH domains are flanked by a diverse array of other domain sequences that determine the intracellular localization and interactive partners of these GEFs and regulate the intrinsic catalytic activities of the GEF proteins[20]. The other major sub-family of RhoGEFs is the

DOCK (dedicator of cytokinesis) family of proteins. There are 11 members identified in the group so far, and they all are structurally and mechanistically different from the Dbl

7

family proteins and regulate only Rac and Cdc42 GTPases[20, 21]. DOCK GEFs are characterized by the presence of two conserved domains, a DHR2 (DOCK homology region 2) catalytic domain and a DHR1 phospholipid-binding domain that localizes them to the plasma membrane[22].

For 20 Rho GTPases, there are 80 RhoGEFs, suggesting some redundancy in their activity. Any given RhoGEF can regulate more than one GTPase, and one GTPase can be regulated by several RhoGEFs. These specificities are conferred by the domain sequences flanking the RhoGEF domains. This intricate regulatory network helps to fine-tune the spatial and temporal activation of Rho GTPases[23].

1.1.2.2 Regulation by RhoGAPs

RhoGAPs stimulate the intrinsic GTPase activity of Rho proteins, and therefore regulate the transient signaling through Rho GTPases. Using in silico and biochemical approaches, around 66 RhoGAP proteins have been identified in the [24].

They are characterized by a conserved 150 bp RhoGAP catalytic domain, and the flanking sequences in the RhoGAPs determine their localization and specificities[17]. Like

RhoGEFs, several RhoGAPs can work on a single Rho GTPase, while one RhoGAP can work on several Rho GTPases[25].

1.1.2.3 Regulation by RhoGDIs

RhoGDIs function as negative regulators of Rho GTPases activity by binding and inhibiting the release of GDP from the GDP-bound inactive form of Rho, thereby 8

preventing their subsequent activation. Additionally, they can sequester GTPases away from their sites of action and maintain an inactive pool of GTPases in the cytoplasm[3].

1.1.2.4 Transcriptional and post-transcriptional regulatory mechanisms

Expression and cellular activities of Rho GTPases and Rho regulators are modulated at the transcription level by transcription factors and post-transcription level by miRNAs[26]. The first evidence for the transcriptional regulation of Rho GTPases was the identification of RhoE as a direct downstream regulator of the -mediated stress response[27]. This study showed that under genotoxic stress, p53 directly binds to the

RhoE promoter and enhances its expression, and this induced RhoE then promoted actin depolymerization and inhibited . Since this study, more pieces of evidence for transcriptional regulation of Rho GTPases were reported and include p53 as a direct regulator of RhoC,[26] p38 MAPK modulation of RhoB through c-Jun,[28] TGFβ regulation of RhoB,[29] and as a regulator of RhoA expression[30].

Post-transcriptional regulation of Rho proteins is modulated by miRNAs. These short, non-coding RNA molecules can silence target genes by degrading mRNAs or by suppressing their translation. Numerous studies have reported Rho GTPase regulation by miRNAs, and their implications in cancer and other physiological conditions[31].

MicroRNAs play a crucial role in cardiac development and cardiac hypertrophy. In murine models of cardiac hypertrophy, it was found that miR-133 and miR-1 are significantly downregulated, and further studies identified RhoA and Cdc42 as miR-133 specific downstream targets that regulate cardiac hypertrophy in these models[32]. Later, it was shown that Cdc42 is the miR-1 specific downstream target implicated in cardiac 9

development[33]. Also in developing and adult , miR-124 promotes outgrowth by suppressing Cdc42 expression and by affecting the subcellular localization of Rac1[34]. During tumorigenesis and tumor progression, miRNAs play a significant role by modulating the expression of and tumor suppressors, including Rho

GTPases[31]. RhoA GTPase was reported to be modulated by miR-31, miR-155 and miR-185 in breast and glioblastoma, colorectal and breast, and colorectal respectively[35-37]. Additionally, Cdc42 was found to be a target of miR-29, miR-137 and miR-185 in breast and lung, colorectal and glioblastoma, and colorectal cancers respectively[35, 38, 39]. Furthermore, RhoB was reported to be targeted by miR-21 in breast and colorectal cancers[40].

1.1.2.5 Regulation by post-translational modifications

Rho proteins are post-translationally modified with lipid moieties that assist their targeting to distinct membranes for localization and direct their interactions with specific

Rho regulators. Prenylation at the C-terminus is the most frequent modification that directs the membrane localization and involves the addition of a farnesyl or a geranylgeranyl moiety to a Cys residue in the CAAX motif[17]. In addition, is a reversible lipid modification that allows dynamic interactions of Rho proteins with membranes and

Rho regulators[41].

Phosphorylation is another post-translational modification that regulates Rho

GTPases localization and dynamic interactions with Rho regulators and effectors. RhoA is the first GTPase protein identified to be regulated by phosphorylation[42]. RhoA 10

phosphorylation at Ser188 by PKA or PKG inhibits RhoA signaling by reducing its affinity for downstream effectors including ROCK,[43, 44] or by enhancing its interactions with

RhoGDIs thereby extracting them from functional membranes[44]. This inhibitory mechanism leads to rapid regulation of actin protrusions at the leading edge of migrating cells[45]. In vascular smooth muscle (VSMC) cells, the interaction of phosphorylated

RhoA with RhoGDIs releases Rac1 from RhoGDI complexes thereby inducing Rac1- mediated migration and adhesion of these cells[46]. In addition, the interaction with

RhoGDIs prevents RhoA degradation by ubiquitylation in VSMC cells[47].

Rac1 phosphorylation at Tyr64 by focal adhesion kinase (FAK) or SRC kinase inhibits its activity and reduces endothelial cell spreading[48]. Additionally, phosphorylation at Ser71 by /AKT also reduces GTP binding and activity[49]. Also, phosphorylation at Thr108 by Extracellular signal-regulated kinase

(ERK) induces its translocation and sequestering it in the nucleus thus preventing Rac1- mediated responses in the cytoplasm[50].

Cdc42 phosphorylation at Tyr64 by SRC kinase or Ser185 by PKA enhances its interaction with RhoGDIs[51]. In addition to these examples, several other Rho GTPases including RhoB,[52] RhoC,[53] RhoH,[54] RhoQ,[55] and RhoU [56] are regulated by phosphorylation.

Ubiquitylation is another major post-translational modification that regulates protein stability and turnover of Rho GTPases. Covalent addition of a chain of four or more ubiquitin monomers linked through their Lys48 residue marks target proteins for degradation through the 26S proteasome[17]. Ubiquitylation of RhoA has been well

11

studied and three different E3 complexes that drives these modifications have been identified and includes SKP1–CUL1–F‑box (SCF)FBXL19 complex [57], the SMURF1

(SMAD-specific E3 ubiquitin protein ligase 1) complex [58] and the BTB/POZ domain- containing adaptor for CUL3‑mediated RhoA degradation (BACURD)–CUL3–RING complex[59]. Rac1 is also regulated by ubiquitylation and HACE1,[60]

IAP,[61] and SCFFBXL19 [62] are the major E3 that ubiquitylate Rac1.

Sumoylation is another post-translational modification that regulates GTPase activity and Rac1 is the only GTPase that is reported to be sumoylated so far[63]. Rac1 sumoylation by E3 SUMO ligase, PIAS3, is not essential for GTP binding and Rac1 activation, yet it helps to maintain and stabilize the GTP-bound active state.

1.1.3 Rho GTPase effectors and their functions

With the binding of GTP, the conformational changes in the GTPases allow them to interact with target/effector proteins in a conformational-specific manner and initiate downstream signaling events. More than 50 effectors have been identified for Rho proteins, using biochemical approaches as affinity chromatography and yeast two-hybrid assay, and among them are tyrosine kinases, serine/threonine kinases, lipid kinases, oxidases, lipases, actin regulators and scaffold proteins[64]. Each Rho protein can interact with more than one effector, in a spatially and temporally regulated manner, to initiate different downstream signaling cascades leading to its biochemical functions in regulating actin and microtubule cytoskeleton dynamics, and gene expression, which leads to biological

12

functions in regulating cell cycle progression, cell shape and polarity, , and cell migration and invasion[64].

1.1.3.1 Actin cytoskeleton dynamics

One of the major functions of all Rho proteins is to regulate the actin cytoskeleton.

The role of Rho proteins in actin dynamics was initially identified using dominant negative mutants in Swiss 3T3 fibroblasts[65]. Addition of lysophosphatidic acid (LPA) induced contractile actin-myosin stress fibers and focal adhesions at the tip of the stress fibers, and this induction was blocked by C3 , hence implicated a function for RhoA

GTPases family in stress fibers formation[66]. Growth factors such as platelet-derived growth factor (PDGF) and epidermal growth factor (EGF), induced protrusive membrane ruffles called lamellipodia, and a dominant negative Rac mutant specifically inhibited this response[67]. Bradykinin induced the formation of microspikes and finger-like membrane protrusions called filopodia, and the dominant negative mutant of Cdc42 blocked this response[68]. These initial works identified the role of Rho proteins in regulating dynamic actin structures. These dynamic structures are achieved by tightly controlling the actin polymerization and actin filament organization through diverse downstream effectors.

Though Rac and Cdc42 form distinct membrane structures, they both initiate peripheral actin polymerization through either Arp2/3 complex or Diaphanous-related formin proteins (DRF). The Arp2/3 heptameric actin-nucleating complex is found in all eukaryotes and binds to the sides of existing actin filaments and induces polymerization of actin to form a branching actin filament network. Both GTPases indirectly activate Arp2/3

13

complex through Wiskott-Aldrich syndrome protein (WASP) family proteins. GTP bound

Cdc42 directly binds to WASP and induce their activation of the Arp2/3 complex.

Activated Rac indirectly activates WAVE protein, which in turn activates the Arp2/3 complex[64].

DRF family includes Dia1, Dia2, and Dia3, and they stimulate the nucleation and extension of non-branching actin filaments by binding as dimers at the plus end of an actin filament. Each Dia protein is activated by a subset of Rho proteins. During filopodia formation, Dia2 activated by Cdc42 plays a major role in polymerizing non-branch actin filaments. RhoA like GTPases induce actin polymerization through directly activating

Dia1 during stress fiber formation.

Actin polymerization is also affected by cofilin, which is an actin filament-severing and actin depolymerization protein. Filament severing by cofilin increases the availability of uncapped plus ends that serve as new sites for actin polymerization. The activity of cofilin is tightly regulated by phosphorylation, PIP2 binding, intracellular pH and protein- protein interactions[64]. Cofilin is directly phosphorylated and inactivated by LIM kinase

(LIMK), which is activated by activated RhoA-like GTPases through ROCK effector or activated Rac/Cdc42 through PAK kinases[69].

ROCK is a serine/threonine kinase and has many substrates, myosin light chain

(MLC) and MLC phosphatase being the major ones. ROCK-mediated phosphorylation inactivates MLC phosphatase leading to an increase in MLC phosphorylation. ROCK also directly phosphorylates and activates MLC as well. Activated MLC promotes the actin filament crosslinking activity of myosin II and increases the contractility of actin fibers.

14

ROCK has two isoforms in eukaryotes that have redundant as well as distinct functions on the actin cytoskeleton. ROCK1 is essential for actin-myosin contractile stress fiber formation and focal adhesions whereas ROCK II is important for and cell contraction[70].

Additionally, Rho proteins regulate the actin cytoskeleton by inducing the transcription of the actin gene through the serum response factor (SRF) which binds to the serum response element (SRE) in the actin promoter[71]. This mechanism is mediated by

LIMK, which gets activated upon phosphorylation by ROCK, then, in turn, phosphorylates and inactivates cofilin. Due to cofilin inactivation, F-actin gets stabilized which in turn reduces the G-actin pool in the cell. This reduction is “sensed” by SRF and gets activated[72].

1.1.3.2 Microtubule cytoskeleton dynamics

Microtubules are highly dynamic polymer structures composed of α- and β- tubulin heterodimers arranged into protofilaments. In most cells, there are two major pools of microtubule filaments: dynamic and stabilized microtubules. Dynamic microtubules contain tyrosinated tubulin and are responsible for dynamic instability, a response of fast assembly and disassembly of microtubules during mitosis. Stabilized microtubules contain detyrosinated and determine cell polarity and distribution of intracellular organelles such as Golgi[73].

Microtubules are very tightly regulated, especially at the growing plus end. One major plus end-binding protein is the Op18/ proteins, and they promote

15

disassembly by either binding and inducing catastrophic disassembly at the plus end or interacting with tubulin dimers to inhibit polymerization[74]. Phosphorylation of

Op18/stathmin proteins at Ser16 by Cdc42/Rac1 activated-PAK kinases inactivates these proteins, hence promotes microtubule assembly[75]. In neurons, collapsin response mediator protein-2 (CRMP-2) binds tubulin heterodimers and promotes microtubule assembly. RhoA-like GTPases-activated ROCK phosphorylates CRMP-2 at Thr555 and inactivates its activity on microtubules[76]. Yet, in NIH 3T3 fibroblasts, RhoA-like

GTPases-activated mDia promotes the formation of stabilized microtubules[73].

Therefore, the effect of Rho proteins on microtubule cytoskeleton is likely to be context dependent.

1.1.3.3 Rho GTPases as regulators of gene expression

Rho proteins promote normal and aberrant and the transforming potential of these proteins are mainly due to its ability to regulate nuclear gene expression[77]. The first evidence came from a study that showed activated Rac and Cdc42 small GTPases can transduce upstream stimuli to activate c-Jun amino-terminal kinases

(JNKs)[77]. Activated JNK phosphorylates the transactivation domain of ternary complex factor (TCF), and activated TCF binds to the SRE of c-fos promoter to induce transcription of this AP-1 family [78]. JNK also phosphorylates and activates c-Jun and ATF2 transcription factors that bind to the AP1 response element on the c-jun promoter as heterodimers to activate its transcription[79]. In addition to Rac/Cdc42 GTPases, RhoA

GTPase also activates JNK through ROCK effector[79].

16

Additionally, all major Rho GTPases regulate the activation of nuclear factor-κB

(NF-κB), transcription factors,[80] which regulates a broad range of biological processes including innate and adaptive immunity, inflammation, stress responses, B-cell development, apoptosis, and cytokine and chemokine expression[81]. NF-κB dimers, p50 and p65, are sequestered in the cytoplasm by a group of regulatory proteins called IκBs, and IκB phosphorylation by a large multiprotein complex containing two catalytic subunits, IKKα and IKKβ, induce the release of IκB and subsequent ubiquitylation dependent degradation. NF-κB regulation by Rac GTPase was reported to be dependent on the IKKβ activation, whereas Cdc42 and RhoA-like GTPases regulate NF-κB through an

IKKβ-independent pathway[82]. In diabetic kidneys and several human cancer cell lines,

RhoA activates NF-κB through ROCK activation[83].

Signal transducers and activators of transcription (Stats) are another group of latent transcription factors that are stimulated by many cytokines and growth factors. Upon a simultaneous tyrosine and serine phosphorylation, these get activated and translocate to the nucleus. RhoA GTPase activates Stat3 by coordinated phosphorylation of serine-727 by

JNK kinases and tyrosine-705 by Src kinases, and this is mediated by ROCK effector[84].

RhoA is reported to activate Stat5a as well[85].

In addition to these transcriptional regulators, Rho GTPases also regulate GATA and CREB transcription factors. In cardiomyocytes, RhoA GTPase mediates the phosphorylation of the GATA-4 activation domain through a p38 MAPK-dependent pathway, and this activation of GATA-4, a key regulator of cardiac genes, is important for

17

sarcomere reorganization[86]. Rho regulator, p190-B RhoGAP, enhances the activity of

CREB transcription factor through negatively regulating the Rho GTPases activity[87].

1.1.3.4 Rho GTPases as regulators of cell cycle progression

The eukaryotic cell cycle includes a synthesis (S) phase and a nuclear division (M) phase separated by two gap phases (G1 and G2). The timely progression through G1 commits the cell to go through the whole cycle, and this is mediated by two types of - dependent kinases (Cdks), Cdk4/6 and Cdk 2, which are activated upon binding to cyclin

D and , respectively. The inhibition of these Cdks is mediated by INK4A and

Cip/Kip family of proteins[64]. The timely progression through G1 is entirely dependent on the temporal regulation of cellular levels of these cyclin activators and inhibitors. The

Cdk4/ interact through the mid- and requires the sustained expression of cyclin D, whose expression is tightly regulated by the Ras/ERK pathway and [88].

Upon ectopic expression, Rac and Cdc42 GTPases are reported to stimulate the cyclin D expression in cells[89]. The Cdk2/cyclin E complex interacts towards the end of the G1 phase, and cyclin E expression is stimulated by Cdc42 through its p70 S6 kinase effector[90]. In addition to promote expression of activators, Rho proteins also inhibit the expression of inhibitors. Cdk2 inhibitor, p21cip1 is inhibited by Rho, either by suppressing transcription or by promoting protein degradation,[91] while p27kip1 is inhibited at the post- transcription level[92].

18

Rho GTPases also play important roles during mitosis and cytokinesis. ROCK- mediated actomyosin filaments in the cell cortex are important for proper positioning of the centrosomes during mitosis[93]. Kinetochore, a complex of around 50 proteins, includes mDia3, a Cdc42 specific effector, and inhibition of either Cdc42 or mDia3 induces mitotic arrest where many are not properly attached to the spindle microtubules[94]. During cytokinesis, Rho GTPases are localized at the cleavage furrow with its effectors including ROCK, mDia and , and affects the function of the contractile ring[95].

1.1.3.5 Rho GTPases as regulators of cell migration and invasion

Cell migration is an important physiological response during embryonic development and associated with pathological conditions as inflammation and cancer metastasis. Cells can migrate either as single cells, using mesenchymal movements or ameboid movements, or as groups of cells known as collective cell migration[96]. The directionality of a moving cell as a response to an extracellular cue is mediated by actin- based structures formed at the leading edge and rear end of the cell due to the activity of specific Rho GTPase proteins. The leading edge is often defined by Rac-mediated lamellipodia and Cdc42-mediated filopodia structures formation, and the rear end is associated with RhoA-mediated focal adhesions and stress fibers[97]. Localized actomyosin contractile forces generated by RhoA-like GTPases at the front and the rear of the cell also needed during migration. Rapid assembly and disassembly of these actin-based structures couples with fast turn-over of cell-cell and cell- adhesions

19

mediate the directionality and effective cell migration[98]. Alternative to lamellipodia- based migration, certain cells migrate using bleb-based forward protrusions. Membrane blebbing formation is associated with high levels of actomyosin contractile forces, and therefore regulated by RhoA-like GTPases[99].

Cell migration through tissues in vivo requires the degradation of extracellular matrix (ECM) components, and this is mediated by specialized actin-based structures called invadopodia and podosomes[100]. These Cdc42 activity-based transient structures are enriched in matrix metalloproteinases (MMPs), that degrade most matrix proteins, and are regulated by actin-regulatory proteins as N-WASP, cortactin, and cofilin[98].

1.1.3.6 Rho GTPases as regulators of cell-cell adhesions

Cell-cell adhesions are a critical component of an epithelium. There are three major types of intercellular adhesive structures that have distinct molecular components, cytoskeletal interactions, and functions. They are adherens junctions, tight junctions, and desmosomes. Adherens junctions, which are formed by classical E- or P-cadherin receptors, and desmosome junctions, which are formed by desmosomal cadherins, provide mechanical strength and tight adhesion between neighboring cells[101]. Tight junctions, which are formed by claudins and occludins, are responsible for maintaining an impermeable barrier for ions and small solutes across epithelial cells[102].

Regulation of adherens junctions (AJs) by Rho GTPases

20

The adherens junction (AJ) consists of two major adhesive units: the classical cadherin-catenin complex and a secondary nectin-afadin complex. (Fig 1.3) Cadherins are ubiquitously expressed receptors that mediate calcium-dependent cell-cell adhesions.

Epithelial cells express at least two types of classical cadherins, E-cadherin and P-cadherin.

In E-cadherin, the extracellular region has five cadherin domains that homophilically interact with cadherin domains in neighboring cells, whereas intracellular cytoplasmic tail directly binds p120 catenin and β-catenin through conserved binding sites, and indirectly bind α-catenin through β-catenin[103]. Plakoglobin or γ-catenin is closely related to β- catenin and can substitute for it in AJs. The catenins link the AJs to actin and microtubule and can mediate the interactions of AJs with other proteins as well. α-catenin is a known binding partner of several actin-binding proteins including vinculin, α-actinin, formin, ZO-1, and afadin[104]. Additionally, β-catenin is the major mediator of the , and furthermore, it can directly bind to other proteins such as EGF receptor and tyrosine phosphatases[105].

21

Figure 1-3: Schematic representation of the basic structural components of the adherens junctions. Shown are the cadherin-catenin complex and the nectin–afadin complex and their potential interactions with actin. (Reprinted by permission from

[Elsevier]: [Elsevier] [Tight Junctions/Adherens Junctions: Basic Structure and

Function] [REFERENCE CITATION (Tight Junctions/Adherens Junctions: Basic

Structure and Function, Carien M. Niessen), [COPYRIGHT] (2007)

The nectin-afadin complex can also be an integral part of the AJs. Nectins are immunoglobulin superfamily adhesion molecules that can interact homophilically or heterophilically with other nectins to mediate adhesion. They bind to the cytoplasmic

22

adaptor protein called afadin, which links the nectin-afadin complex to the actin cytoskeleton.

Rho GTPase signaling plays an important role during AJ clustering and maintaining of junctions. It was shown that inhibition of RhoA-like family protein activity by the bacterial toxin C3 transferase removes cadherin receptors from the newly formed and mature AJs, and this can be rescued by introducing a mutant RhoA construct that is insensitive to the C3 effect[106].

During the initial cell contacts between adjacent cells, the cell protrusions initiate cadherin-catenin clustering, which in turn promotes Rac activation at the protrusive ends by recruiting a RacGEF, T lymphoma invasion and metastasis-inducing protein 1

(TIAM1)[107]. The activated Rac promotes further actin protrusions for expanding AJs by activating the actin nucleator Arp2/3 through Rac effector WAVE[108]. During this stage,

Rac also suppresses Rho activity through recruiting a RhoGAP, p190RhoGAP to p120 catenin[109]. Once cadherin-catenin clusters have expanded, this inhibits the Rac activation and Arp2/3 activity, and as a result, RhoA signaling is promoted at the AJs[110].

Rho signaling act together with α-catenin-based actin remodeling to promote the formation of contractile actin bundles for AJ maturation. In addition to this, through the activity of mDia1, RhoA regulates and stabilizes AJs in several epithelial cell types[111].

1.1.4 Rho GTPases in cancer

Being one of the major regulators of actin and microtubule dynamics, Rho GTPases play a significant role in regulating cell cycle progression, cell-cell adhesions, gene 23

expression, and cell migration and invasion, all of which are key biological processes that are deregulated during tumorigenesis and tumor metastasis. As a result, a plethora of research has been done on Rho GTPases and their implications in and metastasis.

1.1.4.1 RhoA as an oncogene

Early studies with RhoA mutants

During early work, RhoA has been shown to play a role in stress fiber formation,[66] signal transduction,[71] and [112]. In 1993, a study showed that overexpression of wildtype and activated mutant RhoA in fibroblasts was sufficient to confer anchorage and serum-independent growth in vitro, and induced well- differentiated fibrosarcomas upon injection into nude mice[113]. This was the first direct evidence to suggest an oncogenic role for RhoA. Later, it was shown that expression of these activated RhoA mutants induced tumor lung metastasis in nude mice as well[114].

In 1995, it was shown that co-expression of an activated RhoA with activated Raf mutants caused a dramatic synergistic enhancement in mutant Raf-mediated transformation, while dominant negative RhoA mutant co-expressed with an activated Raf mutant abrogated the oncogenic Ras transforming activity[112]. Parallel to this, another study showed that during NIH 3T3 focus formation assay, activated RhoA mutant does not show any transformative potential by itself, but strongly cooperates with an activated Raf mutant in focus formation[115]. These observations suggested that oncogenic Rho mutants function

24

downstream of oncogenic Ras/Raf mediated transformation. However, it should be noted that these specific dominant negative and constitutively active RhoA mutants used for early experiments were not naturally detected in human tumors and generated solely considering the analogy to Ras mutants.

RhoA expression in human tumor tissues

In the following years, many studies reported that RhoA is overexpressed in human tumor tissue samples. One study showed that RhoA is overexpressed in breast tumors compared to its matching benign tissues, and this enhancement is not regulated at the mRNA level. Further, it showed that RhoA is not altered by mutations in breast tumors[116]. However, a study in ovarian carcinoma demonstrated that RhoA mRNA level and protein expression are progressively increased from benign tissues to localized primary ovarian tumors, to high-grade serous carcinoma to distant metastases[117]. Even in gastric cancer tissues samples, RhoA mRNA level was shown to be significantly higher than its matched non-tumor tissues and associated with a poorly differentiated histological type[118]. Similarly, in tissues, RhoA protein expression was shown to be more abundant in tumor tissues and metastatic lymph nodes than in paired non-tumor bladder tissues and uninvolved lymph nodes. Also, in these samples, high RhoA expression was significantly correlated with poor tumor differentiation[119]. Later, it was shown that in testicular germ cell tumor patients, RhoA protein expression was significantly enriched in tumor tissues than paired non-tumor tissues, and increased with tumor grade[120].

25

Likewise, in head and neck squamous cell carcinoma tissues, RhoA expression was shown to be significantly enriched compared to non-tumor tissues, during an IHC analysis[121].

In vivo studies with RhoA

Several in vivo studies with small animals also demonstrated the oncogenic activity of RhoA in tumorigenesis and metastasis. MIM (Missing in metastasis) is a potential tumor metastasis suppressor protein in breast cancer, and it was shown that MIM suppresses breast cancer cell invasion and tumor metastasis in mice through suppressing the activity of RhoA GTPases[122, 123]. Using the siRNA-mediated RhoA gene knockdown, one group showed that reduced RhoA inhibits cell proliferation and invasion in MDA-MB231 cells, and inhibits tumor growth and angiogenesis in MDA-MB231 xenografts in nude mice[124].

Oncogenic activity of RhoGEFs

RhoGEFs promote RhoA activation, hence increased activity of these regulators leads to increased RhoA activity. Chromosome rearrangement or N-terminal truncation are the primary methods of mutation that leads to activation of RhoGEF proteins[23]. These activated RhoGEF proteins are reported to drive cell transformation and tumorigenesis, hence supporting an oncogenic role for RhoA. The first mutated RhoGEF to be identified was the Dbl, a RhoGEF specific for Rac1, that was reported to drive diffuse B-cell lymphoma[18]. Ect2, epithelial cell transforming sequence 2, is an oncogenic RhoGEF, that primarily activates RhoA, but also activates Rac1 and Cdc42. Activated Ect2, due to

26

gene amplification has been reported in various human cancer cell lines and tissue samples including lung, glioblastoma and esophageal squamous cell carcinoma[125, 126]. LARG, leukemia-associated RhoGEF, is another RhoA activating RhoGEF that rearranged to form the truncated chimeric oncoprotein MLL-LARG, which is associated with acute myelogenous leukemia[127]. BCR-ABL1 oncoprotein is another example of a rearranged oncogenic RhoGEF. BCR (breakpoint cluster region) is a RhoGEF that can activate RhoA, and ABL1 is a protein tyrosine kinase. This fusion protein is encoded by the translocation that is associated with the Philadelphia chromosome, a condition found in 90% of chronic myelogenous leukemia[128].

Tumor-suppressive activity of RhoGAPs

Similar to RhoGEFs, RhoGAPs also another regulator of RhoA activity, that induces the intrinsic GTPases activity of these proteins, hence promotes their inactivation.

Therefore, reduced expression or activity of RhoGAPs can support an oncogenic role for

Rho GTPases. One example of this situation is the DLC1 (deleted in liver cancer 1) protein, which is a RhoGAP for RhoA and Cdc42. It is located on chromosome 8p region, which is frequently deleted in human cancers including liver cancer. It has been shown that knockdown of DLC1 cooperates with Myc to promote liver cancer formation in mice[129].

Additionally, DLC2 was also shown to exhibit its tumor suppressor function in vivo by reducing the RhoA activity[130].

1.1.4.2 RhoA as a tumor suppressor 27

Somatic mutations of RhoA in human tumors

Unlike other GTPases, mutations in the RhoA gene were not detected until recently.

In 2014, the first pieces of evidence for the presence of somatic mutations in RhoA were reported by several groups. By performing a whole-exome sequencing of a cohort of diffuse-type gastric carcinoma (DGC) samples, one study discovered several recurrent

RhoA mutations in 25.3% of DGC samples[131]. The hotspot mutations were at Tyr42,

Arg5, and Gly17, and upon biological characterization determined to be gain-of-function in nature. Another group identified loss-of-function recurrent RhoA mutations in 14.3 % of DGC tumors by a whole-genome sequencing approach[132]. These studies suggested

RhoA as a novel driver in DGC. The most common alteration, Y42C in RhoA, lies in the effector binding region, and even though not previously identified in human tumors, this mutation has been biochemically characterized to attenuate the binding of RhoA effector, protein kinase N. Therefore, it is considered biochemically loss-of-function in nature[133].

This discrepancy in biochemically loss-of-function and biologically gain-of-function natures in RhoA gene hotspot mutation has been thoroughly investigated later[134]. During the same period, an independent whole-genome sequencing effort in angioimmunoblastic

T-cell lymphoma (AITL) patients, identified G17V as a recurrent mutation in the RhoA gene and characterized to be loss-of-function in nature[135, 136]. Additionally, a loss-of- function recurrent mutation in RhoA also identified by a genome-wide sequencing in

Burkitt’s lymphoma patient samples[137]. These discoveries of RhoA gene hotspot mutations using whole-genome/-exome sequencing suggested a novel tumor suppressor function for wild type RhoA GTPase in a certain type of tumors.

28

In-vivo studies with RhoA

In addition to the mutational studies, several recent in vivo studies using mouse models and human cancer tissue samples also suggested a possible tumor suppressor function for RhoA. In one study, using microarray analysis of freshly frozen colorectal tumor samples, significant gene expression differences in colorectal cancers were identified and RhoA was one of the most differentially expressed genes. The prognostic potential of RhoA in colorectal cancer was further validated with tissue microarrays and

IHC which showed that reduced RhoA expression was significantly associated with shorter patient survival[138]. Using mouse models of genetic and carcinogen-induced intestinal carcinogenesis, one study showed that reduced RhoA activity in the intestine significantly accelerates the tumorigenic process and leads to increased tumor size and shorter animal survival. Using paired human colorectal samples, they also showed that RhoA activity is significantly reduced in metastases compared to primary tumors[139]. They further showed that reduced RhoA leads to a redistribution of membrane β-catenin to the cytoplasm, which leads to an enhancement in the Wnt/β-catenin signaling pathway. The reduced expression of RhoA in colorectal cancer tissues was shown to be mostly due to a copy number loss, and partially due to a high expression of negative regulators of RhoA expression such as miR-200a and c-Myc[140].

Oncogenic activity of RhoA GAPs

29

As negative regulators of Rho GTPases, increased activity of RhoGAPs can suppress Rho activity. Therefore, tumors driven by increased expression of RhoGAPs support a tumor suppressor role for Rho GTPases. The first study to support this idea used

RNA sequencing and showed that in human basal-like breast cancer tissues, ARHGAP11A and RACGAP1 genes are highly enriched. Further characterizations showed that these two

GAPs are required for basal-like breast cancer cell proliferation, and suppression of

ArhGAP11A induces cell cycle arrest, while suppression of RacGAP1 induces senescence and defects in cytokinesis[141]. Similarly, another study showed that in highly migratory triple-negative breast cancer cells, ArhGAP18 is highly enriched compared to other subtypes. Deletion of the ARHGAP18 gene in these cells increased RhoA activation but reduced their growth, migration and metastatic capacity. Also, in human breast tumors, higher ArhGAP18 levels were shown to be associated with the worst overall survival[142].

1.2 Raf-1 kinase inhibitor protein (RKIP)

Raf-1 kinase inhibitor protein (RKIP), also known as phosphatidylethanolamine- binding protein 1 (PEBP1), is a small kinase inhibitor protein, that was initially identified during a yeast two-hybrid assay as a physiologically relevant binding partner of Raf-1 kinase[143]. It was the first endogenous kinase inhibitor of Raf-1, and the Ras-Raf-MEK-

ERK signaling pathway, which is a ubiquitously expressed signaling pathway that transduces mitogenic, differentiation, and apoptosis signals from the to the nucleus. RKIP binds and inhibits the kinase activity of Raf-1, preventing the subsequent

30

phosphorylation and activation of MEK and ERK kinases. In vitro, RKIP can bind to Raf-

1, MEK, and ERK, but not to Ras. RKIP is not a substrate for Raf-1 or MEK, but rather act as a competitive inhibitor that disrupts the Raf-1 and MEK complex. The Raf-1 and

MEK binding sites on RKIP are partially overlapping, hence RKIP binding to Raf-1 and

MEK is mutually exclusive[143, 144]. Activated Ras or Raf-1 mediated transformation, and AP-1-dependent reporter gene expression can be abrogated by RKIP overexpression and silencing of endogenous RKIP with siRNA or neutralizing antibodies induces the activation of ERK and AP-1 mediated transcription.

1.2.1 RKIP as a signaling mediator

RKIP was initially identified as a physiologically relevant inhibitor of the MAPK signaling pathway[143]. Following this discovery, several other studies identified RKIP as a regulator of other important signaling pathways as well. The NF-κB signaling pathway activates genes related to inflammation, infections and stress conditions. RKIP was reported to act as a negative regulator of NF-κB activation by interacting with NF-κB- inducing kinase and transforming growth factor beta-activated kinase 1 (TAK1),[145] and suppressing cell invasion[146]. RKIP was also shown to interact with B-Raf and to inhibit its activation in melanoma cells[147]. Cancer cells are most sensitive to chemotherapeutic agent-induced apoptosis. It was shown that drug-induced apoptotic signaling mediates through RKIP, and knockdown of endogenous RKIP expression with small interfering

31

RNAs confers resistance to chemotherapeutic drugs in human prostate and breast cancer cells[148].

RKIP was shown to suppress breast cancer cell invasion and metastasis in an orthotopic murine model, through upregulating the let-7 miRNA expression[149]. This miRNA targets BACH1 transcription factor and the HMGA2 chromatin remodeler for degradation, thereby suppressing the expression of pro-invasive and pro-metastatic genes including Snail, chemokine (C-X-C motif) receptor 4 (CXCR4), -binding glycoprotein osteopontin (OPN), and matrix metalloproteinase1 (MMP1)[150]. In addition to that, RKIP was reported to suppress breast cancer local invasion and metastasis by downregulating the MMP13 and CCL5 expressions respectively[151, 152].

G-protein-coupled receptors (GPCR) are the largest and most diverse group of cell- surface receptors, and their activity is regulated by feedback regulation mechanisms to allow rapid adaptation to different stimuli. G-protein-coupled receptor kinase 2 (GRK-2) is the major feedback inhibitor for GPCRs, and RKIP was reported to bind and act as a physiological inhibitor of GSK-2 activity. This target switch from Raf-1 to GSK-2 is stimulated by RKIP phosphorylation at ser153 by protein kinase C, which causes RKIP dimerization[153, 154].

RKIP was also reported to block the activation of STAT3 in breast and prostate cancer cells, by inhibiting the IL-6-, JAK1-, JAK2- and Raf-1-mediated STAT3 tyrosine and serine phosphorylation[155]. During a pathogen invasion, the innate immunity is stimulated by various classes of host pattern recognition receptors (PRRs), including Toll- like receptors (TLRs), who identify conserved molecular motifs or the pathogen-associated

32

molecular patterns (PAMPs) on the pathogen. The activation of PRRs triggers the subsequent activation of TBK1 and IRF3-mediated type I interferon production. It was recently reported that RKIP acts as a positive regulator of type I interferon production by promoting TBK1 activation[156]. Glycogen synthase kinase-3β (GSK3β) suppresses tumor progression by downregulating multiple oncogenic pathways such as Wnt signaling, and activation. RKIP was reported to act as a physiological binding partner of

GSK3β, and maintain the activity of the protein[157].

Additionally, RKIP expression and activity were reported to be regulated by several upstream regulators. Enhancer of zeste homology 2 (EZH2) is a histone in the polycomb repressive complex 2 (PRC2) and highly expressed in breast and prostate cancer cell lines. EZH2 was reported to act as a negative regulator of RKIP transcription through repression-associated histone modifications[158]. RKIP was also reported to be repressed at the transcription initiation level by the zinc-transcriptional repressor,

Snail[159]. It was also reported that in the prostate, androgens induce RKIP expression through androgen receptor-mediated transcriptional modulation[160].

1.2.2 Metastatic cascade and RKIP as a suppressor of metastasis

Invasion and Metastatic Cascade

Metastasis is loosely defined as the formation of progressively growing secondary tumors in distant organ tissues. It is a complex process with a series of sequential steps, all of which must be completed in order to form a metastatic nodule. The metastatic progression is initiated with epithelial carcinoma cells acquiring mesenchymal-like traits 33

through a process called epithelial-to-mesenchymal transition or EMT, which is a highly coordinated cell biological program, mediated by EMT-inducing transcription factors as

Slug, Snail, Twist, and Zeb1, that concomitantly suppress epithelial cell markers while promoting mesenchymal markers[161]. Cells undergoing EMT, detach their epithelial junctions and cell-cell adhesions with neighboring cells and separate from the primary tumor mass. These dissociated tumor cells infiltrate the surrounding stroma, as individual cells or as cell clusters, and locally invade through the underlying basement membrane and the endothelium to intravasate into the blood and/or lymphatic vessels. These circulating tumor cells (CTCs) that intravasate the vasculature must survive the high hemodynamic forces and the immune surveillance inside the vessels until they reach appropriate molecular signals or a capillary bed to actively extravasate through the endothelium. Upon exiting the vasculature, they colonize to form micro-metastases and remain as dormant cells until the surrounding environment favors their proliferation and formation of macro- metastases[162]. This metastasis process is highly inefficient, and only a tiny fraction of cells that dissociate from the primary tumor will end up as macro-metastatic nodules.

RKIP as a Suppressor of Metastasis

Certain proteins can suppress the metastatic process, by regulating one or more steps, without affecting the primary tumor initiation or progression. They are known as metastasis suppressor proteins (MSPs), and around 25 MSPs have identified in the human genome to date[163]. RKIP’s ability to function as an MSP was first demonstrated in 2003, in an orthotopic prostate cancer mouse model[164]. This study showed that metastatic prostate cancer cells have reduced RKIP expression compared to non-metastatic prostate

34

cancer cells and invade more efficiently by activating the MAPK pathway. Upon the orthotopic injection, RKIP-expressed metastatic prostate cancer cells established primary tumors with similar kinetics as the control metastatic prostate cancer cells, yet RKIP- expressed metastatic prostate cancer-bearing mice showed significantly reduced lung metastatic burden compared to the control mice[164]. However, orthotopic implantation of

RKIP knockdown prostate cancer cells did not affect the mice lung metastatic burden, although the loss of RKIP in the transgenic adenocarcinoma of the mouse prostate

(TRAMP) genetic mouse model of prostate cancer, decreased latency of tumorigenesis and promoted distant metastases formation[165]. Subsequently, RKIP was shown to suppress breast cancer angiogenesis, local invasion, intravasation, bone metastasis, and lung metastasis burden upon orthotopic expression in mice[149, 151, 152, 166]. Additionally, using clinical primary tumor samples and paired metastases, it was shown that RKIP loss or depletion is associated with metastatic disease in an increasing number of solid tumors, including colorectal,[167] pancreatic,[168, 169] thyroid,[170] liver,[171] and renal carcinoma[172]. Further, loss of RKIP was identified as a significant indicator of poor prognosis associated with several carcinomas including colorectal,[173] prostate,[174] and breast[150].

1.3 Breast Cancer

Breast cancer will be the most diagnosed cancer among US women in 2019, with an estimate of 268,600 of new cases of invasive breast carcinoma, corresponding to 30% of all estimated new cancer incidences. In addition to that, there will be an estimated 2,670

35

cases diagnosed in men, and 62,930 cases of in situ breast lesions (ductal carcinoma in situ or lobular carcinoma in situ) diagnosed in women. It will also be the second leading cause of estimated cancer deaths among US women after , with an estimate of 41,670 incidences[175].

1.3.1 Molecular sub-types

Breast cancer is vastly heterogeneous with regards to receptor expression, histopathological characteristics, mutational profile, comprehensive gene expression profile, response to treatments, and prognosis. Classically, breast cancer is divided into four therapeutic subtypes depending on the receptor expression: estrogen receptor (ER), progesterone receptor (PR), epidermal growth factor receptor (ErbB2/HER2), and triple- negative form that doesn’t have any receptor expression.

Based on several comprehensive gene expression profiling studies, breast cancers are classified into five major intrinsic subtypes: luminal, HER2-enriched (HER2+), basal- like, Claudin-low and normal-like[176, 177]. Each of these subtypes shows differential responses to treatments and prognosis and have different risk factors for disease incidence and progression. Luminal breast cancers can be further divided into three subgroups: luminal A, luminal B, and luminal-HER2. All luminal cancers are positive for either ER or

PR, and only luminal-HER2 is also positive for HER2. Luminal A and B subtypes are largely disparate in the expression of two major biological processes, proliferation/cell cycle-related such as Ki67 and luminal/hormone-regulated such as PR and FOXA1[178].

Compared to luminal A tumors, luminal B tumors have higher expression of Ki67 and

36

lower expression of PR and FOXA1. At the level of DNA changes, luminal A tumors show a lower frequency of mutations and chromosomal copy number changes than the luminal

B type. Luminal A subtype corresponds to the most common breast cancer subtype and has the best disease outcome from the rest of the subtypes[179]. Luminal type cancers respond well to hormone therapy, but poorly to conventional chemotherapy. Other targeted therapies such as anti-angiogenic therapies, can be effective for luminal tumors as well.

Yet, the exact response differs among the group[178].

HER2+ enriched tumors are characterized by the amplification and overexpression of ERBB2 oncogene, intermediate expression of luminal-related genes, and low expression of basal-related genes such as keratin 5 and FOXC1. At the level of DNA, these tumors show the highest frequency of mutations across the genome as well. These tumors can be effectively targeted with available anti-HER2 therapies[178].

The basal-like subtype is characterized by the high expression of proliferation- related markers as Ki67, and keratins, intermediate expression of HER2-related genes, and very low expression of luminal-related genes. At the DNA level, these tumors have the second highest frequency of mutation across the genome. BRCA1-mutated breast cancer is associated with basal-like disease[178]. The majority of basal-like tumors lack either hormone receptors and HER2 and are considered as triple-negative (TNBC) tumors.

Currently, no molecular targeted therapy is available for treating TNBC.

Claudin-low subtype tumors are characterized by the low to absent expression of luminal-related genes and junctional proteins, claudin 3,4,7 and E-cadherin, and high enrichment of epithelial-to-mesenchymal-related genes, immune response genes and

37

cancer stem cell-like features. Clinically, the majority of claudin-low tumors are invasive ductal carcinomas with a negative expression of ER, PR, and HER2, and hence

TNBC[181].

1.3.2 Breast Cancer Metastasis

More than 90% of breast cancer-related deaths are due to the distant metastases formed by highly invasive cells disseminated from the primary tumor into specific organs including lungs, bones, liver, and brain. Clinically, evaluation of prognostic markers predicts metastatic disease and includes well-established markers such as primary tumor size, axillary lymph node status, angioinvasion, and histological grade, and case-specific prognostic markers as ERBB2 gene amplification, steroid receptor expression, and uPA/PAI1 protein level[182].

The proclivity of metastatic breast cancer cells towards specific organs has been extensively researched. (Fig 1.4) Whole transcriptome microarray analysis of parental

MDA-MB231 breast cancer cells against the lung-tropic MDA-MB231_4175, brain-tropic

BrM2 and bone-tropic MDA-MB231_1833 cell lines led to the identification of lung, brain, and bone metastatic signature profiles for breast cancer[183-185].

38

Figure 1-4: Schematic model of organ-specific metastatic extravasation of breast cancer cells. (Reprinted by permission from [Springer Nature]: [Springer Nature]

[Genes that mediate breast cancer metastasis to the brain] [REFERENCE

CITATION (Genes that mediate breast cancer metastasis to the brain, Paula D. Bos,

Xiang H.-F. Zhang, Cristina Nadal, Weiping Shu, Roger R. Gomis et al.),

[COPYRIGHT] (2009)

1.4 Tumor Microenvironment

Primary tumors are heterogeneous in nature and are composed of two major discrete yet highly dynamic and interactive components; tumor parenchyma and tumor stroma/tumor microenvironment. The constant bidirectional communication between these two components is an essential factor for efficient tumor progression and metastatic dissemination. The stromal compartment consists of bone marrow-derived cells (BMDCs), endothelial cells, fibroblasts, fat cells, and ECM proteins. Tumor stromal features can vary

39

depending on the tumor tissue-of-origin, and in the case of breast cancer, the stroma is mostly dominated by fibroblasts.

1.4.1 Cancer-associated fibroblasts and tumor metastasis

Fibroblasts form the framework of the connective tissues and get activated during a wound healing or fibrosis phenotype. Due to the chronic inflammatory phenotype of tumor tissue, fibroblasts get activated and these activated fibroblasts are called cancer- associated fibroblasts or CAFs. In addition to fibroblasts, CAFs can be differentiated from bone marrow-derived mesenchymal stem cells (MSCs), adipose tissue-derived stem cells

(ASC), endothelial cells, and tumor cells themselves, under appropriate stimuli. Hence, a single marker for CAFs cannot be defined and generally, CAFs are enriched for α-smooth muscle actin (αSMA), vimentin, fibroblast-specific protein-1 (FSP-1) and fibroblast activation protein (FAP)[186]. CAFs are drastically different from resting fibroblasts in regard to their proliferative capacity, migratory behavior, and secretory and synthetic phenotypes. Their enhanced secretory phenotype mostly consists of IFNγ, CXCL12, IL-6,

IL-8, HGF, PDGF, MMPs and TNF[187]. Fibroblast activation into CAFs is believed to be initiated and maintained by two major autocrine signaling loop mechanisms mediated by TGF-β and CXCL12/SDF-1 cytokines[188].

CAFs are positive regulators of tumorigenesis and tumor metastasis. The ability of

CAFs to promote tumor growth can be partly attributable to enhanced angiogenesis by

CAF-derived CXCL12, and recruitment of bone marrow-derived endothelial cells[189].

Additionally, CAF-derived CXCL12 is shown to promote invasive breast cancer growth

40

and metastasis through paracrine activation of CXCR4-expressing tumor cells[189, 190].

Chemokine receptor CXCR4 is highly expressed in human breast cancer cells, malignant breast tumors, and metastases, and activated CXCR4 signaling mediates actin polymerization, formation and subsequent chemotactic and invasive behavior in breast cancer cells[191]. Neutralizing the CXCR4/CXCL12 interactions in vivo, significantly impairs regional lymph node and lung metastases formation[192]. Hence, the

CXCR4/CXCL12 axis represents a major pro-metastatic mechanism in breast cancer cells.

In breast tumors, CAFs can be derived from bone marrow-derived stem cells

(MSCs). Once differentiated they express CAF markers as αSMA, FAP, and tenascin-

C[193]. The tumor infiltrated MSCs secrete CCL5,[194] another pro-tumor chemokine, and this tumor-derived CCL5 act on CCR5-expressing breast tumor cells in a paracrine manner to promote their motility, invasion, and metastasis[195]. Breast tumor tissues were shown to be enriched for CCL5 and CCR5, and CCR5 antagonist, maraviroc, inhibits breast cancer lung metastasis in a preclinical mouse model[196]. Additionally, tumor- derived CCL5 was shown to recruit monocytes and promote their differentiation into macrophages leading to a pro-metastatic stroma[152, 197].

41

Chapter 2

Reduced RhoA Expression Enhances Breast Cancer Metastasis with a Concomitant Increase in CCR5 and CXCR4 Chemokines Signaling

Gardiyawasam Kalpana1, Christopher Figy1, Miranda Yeung1, Kam C. Yeung1*

1 Department of Cancer Biology, College of Medicine and Life Sciences, University of

Toledo, Health Science Campus, Toledo, Ohio 43614, USA

*Corresponding author

K. C. Yeung, Department of Cancer Biology, College of Medicine and Life Sciences,

University of Toledo, Health Science Campus, Toledo, OH 43614, USA.

E-mail: [email protected]

Competing Interests

The authors declare no competing interests.

42

2.1 Abstract

The role of RhoA GTPases in breast cancer tumorigenesis and metastasis is unclear.

Early studies within which mutations in RhoA were designed based on cancer-associated mutations in Ras supported an oncogene role for RhoA. However, recent whole-genome sequencing studies of cancers raised the possibility that RhoA may have a tumor suppression function. Here, using a syngeneic triple negative breast cancer murine model we investigated the physiological effects of reduced RhoA expression on breast cancer tumorigenesis and metastasis. RhoA knockdown had no effect on primary tumor formation and tumor proliferation, concurring with our in vitro findings where reduced RhoA had no effect on breast cancer cell proliferation and clonogenic growth. In contrast, primary tumors with RhoA knockdown efficiently invaded sentinel lymph nodes and significantly metastasized to lungs compared to control tumors. Mechanistically, the current study demonstrated that this is achieved by promoting a pro-tumor microenvironment, with increased cancer-associated fibroblasts and macrophage infiltration, and by modulating the

CCL5-CCR5 and CXCL12-CXCR4 chemokine axes in the primary tumor. To our knowledge, this is the first such mechanistic study in breast cancer showing the ability of

RhoA to suppress chemokine receptor expression in breast tumor cells. Our work suggests a physiological lung and lymph node metastasis suppressor role for RhoA GTPase in breast cancer.

2.2 Introduction

43

Rho GTPases belong to the Ras GTPase superfamily and in humans consists of 20 members. Among them, RhoA, Rac1 and Cdc42 are the most studied Rho GTPases. These

GTPases act as intermediate signal transducers for a diverse array of cell surface receptors such as cytokine receptors, cadherins, integrins, GPCR, and tyrosine kinases, and therefore regulate major cellular events such as cell polarity, cell migration, cell cycle progression, cytoskeleton rearrangements, vesicular trafficking, and gene expression. Aberrant Rho

GTPase signaling results in neurological and immunological abnormalities, and malignant cell transformation[17].

Like most other GTPases, RhoA and their activating proteins, guanine nucleotide exchange factors (GEFs), are commonly believed to be pro-proliferative and act as cancer oncogenes. The evidence for the oncogenic activity of RhoA was first reported using dominant negative and constitutively active RhoA mutants in fibroblasts. While activating

RhoA mutants enhanced the transformation activity of a weakly oncogenic RAS mutant, a dominant negative RhoA mutant abrogated the oncogenic Ras-mediated transformation[112, 115]. Later, RhoA was reported to be overexpressed in several epithelial human cancer tissue samples including, breast,[116] testicular,[120] liver, colorectal,[198] ovarian,[117] and gastric carcinoma[118].

Recent advances in whole-genome sequencing have identified hitherto undiscovered mutations in cancers and identified recurrent loss-of-function and gain-of- dominant-negative function mutations in RhoA in lymphoma, leukemia and several solid tumors. These latest discoveries, therefore, suggested a possible tumor suppressor function for RhoA in cancers[131, 132, 135, 137]. Indeed, studies with mouse models have

44

demonstrated the tumorigenic and metastatic functions of the identified RhoA dominant- negative alleles in colorectal cancer[139].

In addition, analysis of the RNA-seq data generated from The Cancer Genome

Atlas (TCGA) project revealed high expression of several RhoGAP genes, which are the upstream negative regulators of RhoA, in basal-like breast cancer tumors raising the possibility that RhoGAPs are oncogenic[141]. The oncogenic role of a RhoGAP named

ARHGAP18 in breast cancer was subsequently proven with mouse transplantation model.

Importantly, ARHGAP18 was shown to enhance metastasis partly by inactivating

RhoA[142].

Although these recent advances in the field support a tumor suppressive role for

RhoA, a comprehensive mechanistic analysis on altered RhoA expression in primary tumors and its subsequent effects on tumor microenvironment and metastasis is lacking. In this study, using the highly metastatic 4T1 orthotopic mouse model, which closely resembles the human triple negative breast cancer, we show that downregulation of RhoA expression increases breast cancer lung metastasis. Mechanistically our study suggests that by increasing the expression of chemokine receptors, CXCR4 and CCR5, decreased RhoA expression increases the proclivity of cancer cells for the sentinel lymph node and enables the cancer cells to have access to the circulation and metastasize.

2.3 Materials and Methods

Cell culture

45

BT20 and MDA-MB231 human breast cancer cell lines were cultured in

Dulbecco’s modified Eagle’s medium (DMEM, HyClone, UT, USA) with 10% FBS

(Atlanta Biologicals, GA, USA) and 1% penicillin-streptomycin (HyClone), and the 4T1 mouse breast cancer cell line was cultured in DMEM with 5% FBS, 5% calf serum (Lonza,

Switzerland) and 1% penicillin-streptomycin. Cells were grown in a humidified tissue culture incubator at 37°C in 5% CO2. 4T1 cells were kindly provided by Dr. Fred Miller

(Karmanos Cancer Institute, MI), and MDA-MB231 cells and BT20 cells were purchased from ATCC.

Clonogenic growth assay

BT20 (1500 cells/well) and MDA-MB231 (2500 cells/well) cells were plated on a

6-well plate in triplicates and grown for 10 days before staining. For PD0325901 (CAS#

391210-10-9) treatment, BT20 and MDA-MB231 cells were plated with 2 µM and 10 µM of PD0325901 respectively, for 10 days before staining. Cells were fixed with a 4% formaldehyde solution and stained in a 0.5% crystal violet for 24 hours before being washed with 1x PBS and imaged. The analysis was done with ImageJ software (NIH).

Immunofluorescence with Ki67

Cells were plated on laminin-coated (Sigma, MO, USA) glass coverslips and grown until the desired confluency. Next, they were fixed with 100% methanol for 10 min, and incubated with anti-Ki67 antibody (SC-15402, Santa Cruz, CA, USA), Alexa Fluor® 546 secondary antibody (1:5000, ThermoFisher, MA, USA), and DRAQ5® (1:2000, Cell

Signaling, MA, USA) nuclear stain. Fluorescence images were captured with a Leica TCS

SP5 multiphoton laser scanning confocal microscope and analyzed with ImageJ software.

46

Immunohistochemistry

FFPE sections were deparaffinized and rehydrated with an ethanol gradient.

Antigen retrieval was performed with citrate buffer (pH 6) and tissues were blocked with

10% normal goat serum (Vector Laboratories, CA, USA) for 2 hours. Overnight primary antibody incubations were carried out with Ki67 (SC-15402, Santa Cruz, CA, USA), CD-

19 (bs-4755R, Bioss, MA, USA), CK8 (TROMA-I, deposited to DSHB by Brulet, P./

Kemler, R.), E-cadherin (3195, Cell Signaling, MA, USA), GFP (ab290, Abcam, MA,

USA), CD31 (ab28364, Abcam), F4/80 (ThermoFisher, MA, USA), SMA (19245, Cell

Signaling), CCR5 (ab65850, Abcam), CXCR4 (MAB21651, R&D systems, MN, USA),

CCL5 (AP-20618, Acris, MD, USA) and CXCL12 (MAB350, R&D Systems). Sections were incubated with biotinylated secondary antibodies (Vector Laboratories) and ABC reagent (Vector Laboratories) and developed with DAB reagent (Sigma). Finally, they were counterstained with hematoxylin (Fisher Scientific, NH, USA), dehydrated with an ethanol gradient, and mounted with Permount™ (Fisher Scientific). Whole slides were scanned with an Olympus slide scanner and analyzed with ImageJ software. For co- localization studies, after primary antibody incubation, sections were incubated with corresponding Alexa Fluor® secondary antibodies (1:2000, 1 hour, ThermoFisher) and

DRAQ5® (1:2000, 10 minutes, Cell Signaling, MA, USA) nuclear stain. Fluorescence images were captured with a Leica TCS SP5 multiphoton laser scanning confocal microscope.

In vitro Matrigel invasion assay

47

The PET membranes (8µm pore size) of FluoroBlok™ cell culture inserts (351152,

Corning, NY, USA) were coated with 90 µL of diluted Matrigel (0.3mg/ml) (356234,

Corning), and incubated at 370C for 2-3 hours until solidified. Next, 1x104 of 4T1, BT20 or MDA-MB231 cells suspended in serum-free DMEM media were seeded on the solidified Matrigel layer. Then, 700µL of chemo-attractant medium (DMEM, 1% P/S and

10% FBS) was added to the lower chambers (353504, multiwell 24 well companion plate,

Corning), and the plate was incubated in a 370C incubator. After 24 hours of incubation, the insert bottoms were dipped in 1X PBS and stained in diluted Calcein AM (354217,

Corning) in PBS for 10 min. Fluorescence images of invaded cells were captured with an

EVOS inverted microscope, and analysis was done with ImageJ software.

Mammary fat pad injection for spontaneous metastasis assay

The detailed procedure for mammary fat pad injection was reported before[199].

For all our studies, BALB/c female mice (Taconic, NY, USA) of 5 weeks old were used and 1x105 of 4T1 GFP-LUC control knockdown and RhoA knockdown cells were injected.

After 29 days of injections, whole animals were imaged for luciferase bioluminescence

(BLI) with a Xenogen IVIS system as follows: Fresh D-luciferin (LUCK-1G, GoldBio,

MO, USA) working solution of 15 mg/ml was prepared in 1x PBS, and it was filter sterilized. Mice were inoculated with a 10 µl of luciferin solution per gram of body weight as an IP injection. After 10 min, they were anesthetized with Isoflurane and imaged. Images were processed and quantified with a Living Image 3.2 software.

48

After 30 days of injections, mice were euthanized, and blood and tissues were collected and processed for RNA or FFPE histology analysis. Macro-metastases on the lung surface were imaged and counted.

The Department of Laboratory Animal Resources at the University of Toledo

Health Science Campus is accredited by the Association for the Assessment and

Accreditation of Laboratory Animal Care International (AAALAC) and operates in full compliance with the OLAW/PHS policy on the Humane Care and Use of Laboratory

Animals and the USDA Animal Welfare Act. All animal protocols used in this study were approved by the University of Toledo Institutional Animal Care and Use Committee

(IACUC) and all experiments were performed in accordance with relevant guidelines and regulations in the approved protocols.

Experimental metastasis assay

Tail vein (IV) injections for experimental metastasis were performed as reported before[199]. For our studies, 2x105 of 4T1 GFP-LUC control knockdown and RhoA knockdown cells were tail vein injected. Whole animals were imaged on days 1, 7 and 14 for BLI as described above. After 19 days of injections, mice were euthanized, and lungs were harvested and ex vivo BLI were performed as follows: Fresh D-luciferin working solution of 300 µg/ml was prepared in 1x PBS. Lungs were soaked in 1 ml of luciferin solution for 5 min in a 24 well plate and imaged with a Xenogen IVIS system. Images were processed and quantified with a Living Image 3.2 software.

Real-time PCR analysis

49

Total RNA samples were prepared from cultured cells and mice tumor tissue samples with Qiagen™ RNA preparation kits. These RNAs were used to synthesize cDNA

(M-MLV , Invitrogen, MA, USA). Real-time PCR reactions were performed with SyBR green reagent (Qiagen, MD, USA) and appropriate RT-PCR primers. Primer details are as follows; mCXCR4 F-GACTGGCATAGTCGGCAATG, R-

AGAAGGGGAGTGTGATGACAAA; mCCR5 F-TTTTCAAGGGTCAGTTCCGAC, R-

GGAAGACCATCATGTTACCCAC; mCXCL12 F-TGCATCAGTGACGGTAAACCA,

R-TTCTTCAGCCGTGCAACAATC; mCCL5 F-GCTGCTTTGCCTACCTCTCC, R-

TCGAGTGACAAACACGACTGC; mCYCLOPHILLINA F-

GAGCTGTTTGCAGACAAAGTTC, R-CCCTGGCACATGAATCCTGG; mβ-ACTIN

F-ATCTGGCACCAGACCTTCTACAATGAGCTGCG, R-

CGTCATACTCCTGCTTGCTGATCCACATCTGC.

Western blotting

Total cell lysates were prepared with the lysis buffer (20 mM Tris, pH 7.4, 150mM

NaCl, 2mM EDTA, and 1% Triton X-100), and proteins were separated with sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE). Then, separated proteins were electrophoretically transferred from the gel to a polyvinylidene fluoride membrane

(Millipore, MA, USA). Membranes were blocked and incubated overnight with primary antibodies diluted in PBS (pH 7.4), 0.2% Tween-20, 5% bovine serum albumin and 0.002% sodium azide. Primary antibodies used for western analyses were RhoA (1:1000, Cell

Signaling) and tubulin (1:2000, Sigma). Then, they were incubated with the appropriate horseradish peroxidase-conjugated secondary antibodies (1:2000, Cell Signaling) and

50

developed with Immobilon western chemiluminescence substrate (Millipore) in a Bio-Rad

ChemiDoc EQ system.

Statistical analysis

Statistical calculations with two-tailed Student’s t-test were done using GraphPad

Prism software. All the data are presented as means with error bars representing standard error.

2.4 Results

2.4.1 RhoA suppresses breast cancer cell invasion in vitro

The role of RhoA in breast cancer tumorigenesis and tumor progression have long been a topic of debate[200, 201]. Although early work suggested a potential oncogenic role for RhoA, numerous recent studies start to challenge this notion[137, 141, 142, 202]. To further investigate the role of RhoA in breast cancer, we stably down-regulated RhoA expression by lentiviral transduction of specific shRNAs (Fig. 2.1a) and measured the effects of the RhoA knockdown on cancer cells proliferation and invasion. Two different triple-negative breast cancer cell lines (TNBC) were used to ensure the observed effects were not cell-type specific. While reduced RhoA expression significantly increased breast cancer cells invasion through Matrigel (Fig. 2.1b), it did not have a noticeable effect on cell proliferation as measured by two different assays (Fig. 2.1c and Supplementary Fig.

1). As a positive control for our proliferation assay, we treated breast cancer cells with a selective MEK inhibitor, PD0325901, and observed an expected near-complete suppression on colony formation in clonogenic growth assay (Fig. 2.1c).

51

We reasoned that if the effect of RhoA knockdown on breast cancer cell invasion is RhoA specific and physiologically relevant, we would expect to observe the opposite effect when RhoA expression or activity is increased. Indeed, stable expression of a constitutively active RhoA variant, RhoAQ63L, in both BT20 and MDA-MB231 considerably decreased their proclivity to invade (Fig. 2.1d) Together, these results suggested that in breast cancer cells, RhoA suppresses cell invasiveness.

52

Figure 2-1: RhoA suppresses breast cancer cell invasion in vitro. a Representative western blots of the RhoA expression in BT20 (left) and MDA-MB231 (right) cell lines after lentiviral transduction of indicated constructs. Numbers are shown for quantified bands normalized with tubulin. b Number of invaded cells through Matrigel in indicated

BT20 (top) and MDA-MB231 (bottom) cell lines (mean ± SE) and representative images of the stained invaded cells in indicated cell lines. Representative results of two independent assays. c Number of colonies for indicated BT20 (top) and MDA-MB231

(bottom) cell lines (mean ± SE) and representative images of the stained colonies in indicated cell lines. Representative results of two independent assays. PD0325901

(selective MEK inhibitor) as a positive control. d Number of invaded cells through Matrigel in indicated BT20 (left) and MDA-MB231 (right) cell lines (mean ± SE) and representative images of the stained invaded cells in indicated cell lines. *P<0.05, ns- not significant, unpaired Student's t-test (two-tailed).

2.4.2 RhoA suppresses breast cancer lung metastasis burden in mice

The effects of altered RhoA expression on breast cancer invasion in vitro raises the possibility that RhoA may have a causal role in suppressing breast cancer cells metastasis.

Metastasis is a complex multimodal activity that involves both host and tumor cells and can only be adequately addressed with an animal model. To investigate the possible role of RhoA in breast cancer metastasis, we exploited the well-established 4T1 syngeneic murine breast cancer model[203, 204]. The 4T1 TNBC cell line was originally derived from a spontaneous mammary tumor in a Balb/c mouse. Hence, it is widely utilized for syngeneic orthotopic mammary tumor allograft experiments. We generated a panel of 53

lentiviral-mediated stable RhoA knockdowns in 4T1 cells with different levels of knockdowns, and used two of them, with a 30% (shRhoA 34) and a 50% (shRhoA 32) knockdown, for subsequent experiments (Fig. 2.2a). Concurring with human breast cancer cells results, reduced RhoA expression significantly increased 4T1 cells invasion through

Matrigel in a dose-dependent manner (Fig. 2.2b). Significantly, similar effects were observed with stable expression of a dominant negative RhoAT19N, but the opposite effect was observed when the constitutively active RhoAQ63L was expressed in 4T1 cells (Fig.

2.2c). Collectively, these findings strongly suggested that the mouse 4T1 breast cancer cells could be used to investigate the role of RhoA in metastasis.

54

Figure 2-2: RhoA suppresses breast cancer lung metastasis burden in mice. a

Representative western blots of the RhoA expression in 4T1 cell lines after lentiviral transduction of indicated constructs. Numbers are shown for quantified bands normalized with tubulin. b Number of invaded cells through Matrigel in 4T1 shRhoA 32 (left) and 4T1 shRhoA 34 (right) cell lines relative to their control cells lines (mean ± SE) and representative images of the stained invaded cells in indicated cell lines. Representative results of two or more independent assays. c Number of invaded cells through Matrigel in

4T1 RhoATN (top) and 4T1 RhoAQL (bottom) cell lines relative to their controls (mean ±

SE) and representative images of the stained invaded cells in indicated cell lines.

Representative results of two independent assays. d Representative BLI images of Balb/c mice orthotopically injected with 4T1 gfp-luc cells carrying indicated lentiviral modifications, imaged 29 days after the implantation (left). Total photon flux quantification

(mean ± SE) of BLI images shown on left for the signals in primary tumors (top) and metastases (bottom) (right). N=4 for all groups. e Representative gross lung images of the mice orthotopically injected with 4T1 gfp-luc cells carrying indicated lentiviral modifications, showing visible metastatic nodules (left). The number of lung metastatic nodules (mean ± SE) of these mice. N=4. f Representative hematoxylin and eosin (H&E) staining of lung cross-sections from above-harvested lungs showing metastases highlighted with black (left). Total metastases tumor area per total lung area (mean ± SE) quantification of the H&E images (right). N=4, ns- not significant *P<0.05, unpaired Student's t-test (two- tailed).

55

To address this question, we first engineered 4T1 cells to express a GFP-LUC hybrid protein to allow the growth and metastasis of the tumor to be monitored by luciferase bioluminescent imaging (BLI). Upon orthotopic implantation of 4T1 cells, all mice in RhoA knockdown and control groups developed primary tumors. In accord with our in vitro findings where RhoA knockdown did not influence breast cancer cell proliferation, at 30 days post-implantation we did not observe any statistically significant changes in weights or expression levels of proliferation antigen ki67 in RhoA knockdown primary tumors when compared with knockdown control. (Supplementary Fig. 2). We did not observe any statistically significant changes in expression levels of cleaved caspase-3 in RhoA knockdown primary tumors as well (Supplementary Fig. 2). On the contrary, bioluminescent imaging of mice at 29 days post-implantation revealed possible distant metastases in mice implanted with the highest RhoA knockdown (shRhoA 32) 4T1 cells

(Fig. 2.2d) showing a dose-dependent effect of RhoA knockdown on cancer metastasis.

Indeed, mice implanted with shRhoA 32 knockdown 4T1 cells exhibited a significantly higher lung metastasis burden compared to other groups during histology analysis (Fig.

2.2e-f). In a complementary approach, we also downregulated the activity of RhoA in 4T1 cells by stably expressing the dominant negative RhoAT19N allele. Significantly, expression of RhoAT19N had the same effect on lung metastasis as the knockdown of

RhoA expression in 4T1 cells and with a minimal effect on tumor weights (Supplementary

Fig. 2). Collectively, these findings suggested that RhoA could act as a physiological suppressor of breast cancer lung metastasis in mice.

56

2.4.3 RhoA suppresses breast cancer sentinel lymph node metastasis in mice

It has been thought that invasive breast cancer cells from the primary tumor gain access to the systemic circulation through intra-tumoral and peripheral blood vasculature.

However, with novel and more sensitive research technologies, scientists have identified that for breast cancer, the major systemic metastasis route is through the sentinel lymph node (SLN). Once tumor cells reach the SLN, they can access the systemic circulation directly without first passing down to the next lymph node[205, 206].

Figure 2-3: RhoA suppresses breast cancer sentinel lymph node metastasis in mice. a

Representative immunohistochemical (IHC) staining images of right axillary lymph node

57

sections of mice orthotopically injected with 4T1 gfp-luc cells carrying indicated lentiviral modifications, stained with antibodies for CK8 (top, right) and GFP (bottom, right).

Percentage of CK8+ area (top, left) and GFP+ area (bottom, left) per axillary lymph node area (mean ± SE) quantification of IHC images. N=4. b Representative IHC staining images of left inguinal lymph node sections of above mice, stained with antibodies for CK8

(top, right) and GFP (bottom, right). Percentage of CK8+ area (top, left) and GFP+ area

(bottom, left) per inguinal lymph node area (mean ± SE) quantification of IHC images.

N=4. c Representative BLI images of mice tail-vein injected with 4T1 gfp-luc cells carrying indicated lentiviral modifications, imaged 1, 7 and 14 days after the injections showing the kinetics of lung tumor formation (left). Total photon flux quantification (mean

± SE) of BLI images shown on left (right). N=3 for both groups. d Representative ex vivo

BLI images of lungs harvested from above mice imaged 19 days after the tail-vein injections showing the total lung tumor burden (bottom). Total photon flux quantification

(mean ± SE) of ex vivo lung BLI images shown below (up). N=3, ns- not significant

*P<0.05, unpaired Student's t-test (two-tailed).

It is possible that RhoA suppresses the lung metastasis of 4T1 breast cancer cells by interfering with the access of the cancer cells to the SLN. To examine this possibility, four weeks after implantation, we harvested the draining/sentinel lymph nodes (right axillary LN) from the cancer cells injected sites and processed them for immunohistochemical (IHC) analysis. As a control, we also harvested lymph nodes (left inguinal LN) on the contralateral side of the primary tumor. IHC analysis of the SLNs with

58

an epithelial cell-specific CK8 antibody revealed that the shRhoA 32 knockdown group mice have higher CK8 staining compared to the control mice, while no statistical significance in CK8 staining in the inguinal lymph nodes was observed between RhoA knockdown and control knockdown mice (Fig. 2.3a). Similar results were observed with a

GFP antibody, which specifically stained the 4T1 cancer cells (Fig. 2.3b). Collectively, these observations suggested that when the RhoA expression is reduced, invasive 4T1 cells reached the axillary sentinel lymph node more efficiently than the control cells and readily gain access to the systemic circulation for spreading into other organs including the lungs.

Metastasis is a complex multiple-step process. In addition to invading the draining lymph node and entering the circulation, the circulating cancer cells must survive the hostile environmental conditions inside the blood vessels until extravasating into a distant tissue for metastatic colonization. To investigate whether RhoA also suppresses the later steps of the metastasis cascade, we tail-vein injected RhoA knockdown or control knockdown 4T1 GFP-LUC cells in Balb/c mice to bypass the early stages of metastasis.

Mice were monitored at days 1, 7 and 14 post-injection for lung tumor burden by bioluminescent imaging. As shown in Fig 2.3c-d, both the kinetics of lung tumor formation and extent of the tumor burden showed no significant difference between RhoA knockdown and control group mice suggesting that RhoA may not have a role in cancer cell extravasation and colonization.

2.4.4 RhoA suppresses breast cancer cell invasion by modulating the tumor microenvironment

59

Constant bidirectional communication between cancer cells and tumor microenvironment (TME) plays a major impact on cancer growth and metastasis[207]. It is possible that reduced RhoA expression promotes cancer cell dissemination into sentinel lymph nodes by modulating this dynamic interplay. To investigate this hypothesis, we assessed the major TME components in the control and RhoA knockdown primary tumors harvested from the Balb/c mice 30 days after the orthotopic implantation. TME is mainly composed of activated fibroblasts called cancer-associated fibroblasts (CAF), adaptive and innate immune cells, and vascular endothelial cells. We examined the presence of vascular endothelial cells, CAFs, B-lymphocytes and macrophages in these primary tumors by IHC staining with CD31, SMA, CD19, and F4/80 specific antibodies, respectively. Though no differences were detected in CD31 and CD19 expression between the two groups, a significant elevation was detected in SMA and F4/80 expression in RhoA knockdown tumors compared to control tumors, suggesting an increased infiltration of CAFs and macrophages in RhoA knockdown tumors, respectively (Fig 2.4a and Supplementary Fig.

3).

60

61

Figure 2-4: RhoA suppresses breast cancer cell invasion by modulating the tumor microenvironment. a Representative immunohistochemical (IHC) staining images of breast primary tumor sections of mice orthotopically injected with 4T1 gfp-luc cells carrying indicated lentiviral modifications, stained with antibodies for CD31 (top, left),

F4/80 (middle, left) and SMA (bottom, left). CD31+ area (top, right) and SMA+ area

(bottom, right) per tumor field of view (FOV) (mean ± SE) quantification and F4/80+ cells per tumor section (middle, right) (mean ± SE) quantification of IHC images. N=4. b

Representative IHC staining images of breast primary tumor sections of above mice stained 62

with CXCL12 antibody (left). CXCL12+ area per tumor FOV (mean ± SE) quantification of IHC images. N=4. c Representative immunofluorescent images of breast primary tumor sections of above mice co-stained with SMA, CXCL12, and DRAQ5 nuclear stain, showing the co-localization of SMA and CXCL12 signals. Scale bar:50 µm d

Representative IHC staining images of breast primary tumor sections of above mice stained with CXCR4 antibody (left). CXCR4+ area per tumor FOV (mean ± SE) quantification of

IHC images. N=4. e Relative mCXCR4 mRNA levels normalized with mActin (mean ±

SE), as quantified by qRT-PCR in primary breast tumors of above mice (left) and in 4T1 gfp-luc cells carrying indicated lentiviral modifications (right). N=4. f Representative IHC staining images of breast primary tumor sections of above mice stained with CCL5 antibody (left). CCL5+ area per tumor FOV (mean ± SE) quantification of IHC images.

N=4. g Representative immunofluorescent images of breast primary tumor sections of above mice co-stained with SMA, CCL5, and DRAQ5 nuclear stain showing the co- localization of SMA and CCL5 signals. Scale bar:25 µm h Representative IHC staining images of breast primary tumor sections of above mice stained with CCR5 antibody (left).

CCR5+ area per tumor section (mean ± SE) quantification of IHC images. N=4. i Relative mCCR5 mRNA levels normalized with mActin (mean ± SE), as quantified by qRT-PCR in primary breast tumors of above mice (top) and in 4T1 gfp-luc cells carrying indicated lentiviral modifications (bottom). N=4, ns- not significant *P<0.05, unpaired Student's t- test (two-tailed).

63

CAFs are known to drive the tumorigenesis and metastasis phenotype by secreting pro-tumorigenic chemokines, and by enhancing tumor stiffness[208]. We analyzed the major chemokines receptors expression in 4T1 control primary tumors by RT PCR.

(Supplementary Fig.3) While we did not detect any significant differences in stiffness between control and RhoA knockdown tumors as measured by Masson’s trichrome staining, IHC staining with pro-tumorigenic chemokine CXCL12 specific antibody revealed that RhoA knockdown tumors are significantly enriched in CXCL12 expression

(Fig. 2.4b and Supplementary Fig.4). To identify the source of CXCL12 in the TME, we double-stained the RhoA knockdown tumors with SMA and CXCL12 antibodies. We observed that the CXCL12 staining was predominantly co-localized with the staining of

SMA suggesting that CXCL12 detected in the TME was mainly secreted by the infiltrated

CAFs (Fig. 2.4c). In agreement with this line of thinking, we failed to detect the expression of CXCL12 in 4T1 cells by qRT-PCR (data not shown). The predominant receptor for

CXCL12 is CXCR4. As expected, the expression levels of CXCR4 were also enhanced in

RhoA knockdown tumors as measured by IHC staining and qRT-PCR RNA analysis (Fig.

2.4d-e). Apparently, the enhanced CXCR4 expression is an autonomous effect of the downregulation of RhoA expression in 4T1 cancer cells (Fig. 2.4e).

Tumor-associated macrophages or TAMs were shown to promote tumor progression, angiogenesis and lung metastasis in several breast cancer in vivo studies,[209,

210] and associated with poor patient survival[211]. During tumor progression, of tumor-derived chemoattractant chemokines, mainly CCL5, leads to the recruitment of circulating monocytes into breast tumors. The infiltrated monocytes will subsequently

64

differentiate into tumor associated-macrophages[212]. It is possible that the increased number of TAMs is a consequence of increased expression of CCL5 in RhoA knockdown tumors. Truly, RhoA knockdown tumors were significantly enriched in CCL5 expression

(Fig 2.4f). In addition to secreting CXCL12, CAFs have also been shown to be a key source of CCL5 in tumors[213]. Indeed, CAFs but not the cancer cells, are the major contributors of CCL5 in RhoA knockdown tumors as assessed by immunofluorescent staining and qRT-

PCR (Fig. 2.4g).

Paracrine activities of intra-tumoral CCL5 through its cognate receptor CCR5 are reported to promote breast tumor progression, and LN and thoracic metastasis[195, 196,

214]. It is possible that the observed increase in LN and lung metastases in RhoA knockdown mice is caused by activated CCL5-CCR5 signaling in primary tumors. In agreement with this line of reasoning, we observed the expression levels of CCR5 were increased in RhoA knockdown tumors and the increased CCR5 expression levels are partly due to an increase in CCR5 transcripts in cancer cells (Fig 2.4h-i).

2.5 Discussion

In this study, we investigated the role of RhoA in breast cancer cell proliferation, invasion, and metastasis. Stable knockdown of RhoA expression had no effect on breast cancer cell proliferation but increased in vitro cell invasion and metastasis. At present, the mechanism of how RhoA inhibits breast cancer cell invasion is not known. RhoA GTPase has several downstream effectors. Among them, mDia and ROCK are two best-described effectors that are important for RhoA-mediated regulation of cancer cell

65

migration/invasion by stabilizing adherens junctions and regulating focal adhesion dynamics[215-218]. It remains to be determined if RhoA inhibits breast cancer cell metastasis by targeting adherens junctions and/or focal adhesions.

In addition to cancer cell-autonomous functions, continuous heterotypic communications between cancer cells, ECM, and their surrounding non-malignant cells are also vital for disease progression and metastatic dissemination[207]. Fibroblasts are the most common non-malignant cells in the tumor. In normal tissues, they usually exist in a resting/quiescent state and are activated in response to a plethora of signaling cues.

Activated fibroblasts are highly proliferative and metabolically active and identified by their expression of markers like α-SMA and fibroblast activation protein (FAP). Activated fibroblasts in tumor stroma are termed cancer-associated fibroblasts (CAFs), which are generally pro-tumorigenic with a dynamic and complex secretory phenotype including

ECM molecules (tenascin C, collagen), growth factors (VEGFA, PDGF, HGF), cytokines, and chemokines (TNF, IFNγ, CXCL12, IL-6, IL-8)[208]. Using the 4T1 triple negative breast cancer mouse model, we observed an inverse correlation between RhoA expression and infiltration of α-SMA positive CAFs. Consistent with increased infiltrated CAFs,

RhoA knockdown tumors were also highly enriched with CAFs-secreted CXCL12. The reduced RhoA activity resulted in increased expression of the cognate receptor, CXCR4, for CXCL12 in 4T1 cancer cells.

CAFs-secreted CXCL12 drives tumor progression and metastasis through two possible mechanisms. It mediates the chemotactic recruitment of endothelial progenitor cells (EPC) to promote tumor angiogenesis. It may also enhance the invasive capacity of

66

CXCR4-expressing breast cancer cells through paracrine stimulation[189]. Since RhoA knockdown had no observable effect on tumor angiogenesis, it is possible that the increased

CXCL12 expression might drive the paracrine stimulation of the RhoA knockdown cancer cells’ invasiveness. Therefore, our study suggests that RhoA suppresses breast cancer cell metastatic dissemination by suppressing the invasiveness of tumor cells through reduced expression of the tumor CXCR4 receptor with a subsequent reduction of CAFs in the TME.

Similarly, RhoA knockdown primary tumors displayed a significant F4/80 positive macrophage infiltration. This was correlated with an increased expression of stromal- derived CCL5, a well-established chemokine important in recruiting and converting CCR5 positive monocytes into TAMs. The expression of CCR5, the cognate receptor for CCL5 was likewise found to be upregulated in RhoA knockdown 4T1 cancer cells. In addition to CXCL12, CAFs were also a source of increased CCL5 expression in RhoA knockdown tumors. It is therefore likely that the increased co-expression of CXCR4 and CCR5 in

RhoA knockdown cancer cells facilitates the recruitment of CXCL12/CCL5 secreting

CAFs and later the TAMs into the TME. Like CAFs, TAMs have been shown to be pro- tumorigenic and pro-metastatic. Since RhoA knockdown tumors with significant CAFs and

TAMs infiltration showed relatively increased lymph node and lung metastases formation, it is therefore possible that reduced RhoA activity in tumor cells promotes its metastatic dissemination by activating both the CXCL12-CXCR4 and CCL5-CCR5 signaling pathways.

Metastatic dissemination and distant colonization are inefficient processes with multiple sequential steps. The systemic metastasis of breast cancer was long thought to be

67

through the intratumoral and peritumoral blood vasculature, even though lymph node metastasis was often associated with tumor aggressiveness and poorer prognosis[219]. It was long believed that LN metastases enter systemic circulation by traveling from SLN to the next through efferent lymphatics until they drain into the thoracic duct and through the neck lympho-vascular anastomosis into the internal jugular vein[219]. However, recent studies using intravital imaging, photo-switchable protein, and cancer cell micro-infusion showed that cancer cells in the SLN are a significant source for distant lung metastases and lymph node-residing tumor cells disseminate by efficiently invading LN blood vasculature rather than by transiting through efferent lymphatic vessels[205, 206]. In the present study, we detected more cancer cells in the SLNs but not in the contralateral lymph nodes of

RhoA knockdown mice compared to the control group. Cancer cells tail-vein injection experiments designed to bypass the early steps revealed no difference in cancer cells extravasation and lung colonization between RhoA knockdown and control groups. Taken together, our study suggests that RhoA suppresses the cancer cells’ initial steps of local invasion into SLNs, and subsequent intravasation into LN blood vasculature while has no observable role in the later steps of the metastasis cascade.

Knockdown of RhoA expression increases CCR5 and CXCR4 transcripts in 4T1 cancer cells suggesting a possible transcription-regulatory function of RhoA. RhoA

GTPases are known for its functions in cytoskeleton dynamics, regulating cell polarity and cell migration and cell cycle progression[220]. However, a lesser-known function of RhoA is its ability to regulate several transcriptional signaling pathways such as NFκB signaling

[221], GATA-4 signaling [86], CREB signaling [222], STAT5a signaling [223] and AP-1

68

signaling[79]. Rho regulates the transcriptional activity of NFκB by a mechanism involving phosphorylation of IκBα [80], and in breast and prostate cancer cells, NFκB directly upregulates CXCR4 mRNA expression and stability[224, 225]. Similarly, it was also reported that Rho GTPases could modulate the activity of CREB, and CCR5 is a well- established transcriptional target of the cAMP/PKA/CREB pathway[222, 226]. Therefore, it is possible that RhoA suppresses CCR5 and CXCR4 expression by impinging on one or more of these transcriptional signaling pathways.

In summary, using a triple negative breast cancer mouse model, we demonstrated that reduced RhoA expression increases breast cancer lymph node and lung metastasis.

Potential tumor and metastasis suppressive ability of RhoA in breast cancer have been suggested before[141, 142]. But the current study demonstrated that this is achieved by promoting a pro-tumor microenvironment, with increased CAF and macrophage infiltration, and by modulating the CCL5-CCR5 and CXCL12-CXCR4 chemokine axes at the primary tumor. To our knowledge, this is the first such mechanistic study in breast cancer, shows the ability of RhoA to suppresses chemokine receptor expression in breast tumor cells, in turn suppressing the breast cancer LN and lung metastasis.

2.6 Supplementary figures

69

Sup Fig 1. Representative immunofluorescent images of BT20 (top, right) and MDA-

MB231 (bottom, right) cells co-stained with Ki67 and DRAQ5 nuclear stain, showing the co-localization of Ki67 and nuclei. Percentage of Ki67+ cells per field of view (FOV)

(mean ± SE) quantification of immunofluorescent images of BT20 (top, left) and MDA-

MB231 (bottom, left). ns- not significant *P<0.05, unpaired Student's t-test (two-tailed).

70

Sup Fig 2. a Weight of the breast primary tumors (mean ± SE) of mice orthotopically injected with 4T1 gfp-luc cells carrying indicated lentiviral modifications, in grams. N=4. b Representative immunohistochemical (IHC) staining images of breast primary tumor sections of above mice, stained with Ki67 antibody (bottom). Percentage of Ki67+ cells per field of view (FOV) (mean ± SE) quantification of IHC images (top). N=4. c Weight of the breast primary tumors (mean ± SE) of mice orthotopically injected with 4T1 gfp-

71

luc cells carrying indicated lentiviral modifications, in grams. N=4. d Number of lung metastatic nodules (mean ± SE) of mice orthotopically injected with 4T1 gfp-luc cells carrying indicated lentiviral modifications. N=3 (top). Representative hematoxylin and eosin (H&E) staining of lung cross-sections from above lungs showing metastases highlighted with black (bottom, right) and total metastases tumor area (mean ± SE) quantification of the H&E images (bottom, left). N=3, ns- not significant *P<0.05, unpaired Student's t-test (two-tailed).

Sup Fig 3. The expression profile for major chemokine receptor expression in 4T1 primary tumors analyzed by RT-PCR. N=4

72

Sup Fig 4. a Representative immunohistochemical (IHC) staining images of breast primary tumor sections of mice orthotopically injected with 4T1 gfp-luc cells carrying indicated lentiviral modifications, stained with CD19 (right). CD19+ area per tumor field of view (FOV) (mean ± SE) quantification of IHC images. N=4. b Representative trichrome staining images of breast primary tumor sections of above mice (right).

Collagen+ area per tumor field of view (FOV) (mean ± SE) quantification of trichrome images. N=4, ns- not significant *P<0.05, unpaired Student's t-test (two-tailed).

73

Chapter 3

RKIP Represses Breast Cancer Invasion and Metastasis

Partially Through Activation of RhoA GTPase

Gardiyawasam Kalpana1, Clair Tipton1, Julius De Castro1, Vu Bach1, Clariza Borile1,

Augustus Tilley1, Christopher Figy1, Miranda Yeung1, Sahezeel Awadia2, Dayanidhi

Raman1, Rafael Garcia-Mata2, Kam C. Yeung1*

1 Department of Cancer Biology, College of Medicine and Life Sciences, University of

Toledo, Health Science Campus, Toledo, Ohio 43614, USA

2 Department of Biological Sciences, College of Natural Sciences, University of Toledo,

Toledo, Ohio 43606, USA

*Corresponding author

K. C. Yeung, Department of Cancer Biology, College of Medicine and Life Sciences,

University of Toledo, Health Science Campus, Toledo, OH 43614, USA.

74

E-mail: [email protected]

3.1 Abstract

Raf-1 kinase inhibitor protein was initially discovered as a physiological kinase inhibitor of the MAPK signaling pathway and later was shown to suppress tumor metastasis in prostate and breast cancer murine models. Yet, the molecular mechanism through which

RKIP executes its anti-metastasis effects is not completely defined. During this study, Rho

GTPases were studied as possible downstream effectors of RKIP, given their primary functions in actin cytoskeleton dynamics and cell movement regulation. Among Rho

GTPases, RhoA has both pro-metastatic and anti-metastatic cell-context dependent functions. During the current study, we demonstrated that RKIP specifically activates downstream RhoA GTPase, and this activation is required for RKIP-mediated suppression of breast cancer cell invasion and in vivo lung metastasis. Further, we were able to identify that RKIP’s anti-metastatic effects in breast cancer can be partially attributable to the

RhoA-mediated membrane E-cadherin localization into adherens junctions and inhibition of tumor CCL5 expression. We also showed that mechanistically, RKIP activates RhoA in an ERK2 and GEF-H1 dependent manner.

3.2 Introduction

Raf-1 kinase inhibitor protein or RKIP was initially identified as a physiologically relevant kinase inhibitor of the Ras-Raf-MEK-ERK signaling pathway[143]. Shortly following its discovery, RKIP was shown to suppress metastasis in an orthotopic prostate 75

cancer mouse model, opening a new era for RKIP research[164]. Subsequently, RKIP was shown to suppress breast cancer metastasis in xenograft and orthotopic mouse models[149,

152, 166]. Additionally, it was shown that RKIP loss or depletion is associated with metastatic disease in an increasing number of solid tumors, including colorectal,[167] thyroid,[170] pancreatic,[168, 169] liver,[171] and renal carcinoma[172]. Further, loss of

RKIP was identified as a significant indicator of poor prognosis associated with several carcinomas including colorectal,[173] prostate,[174] and breast[150].

Metastasis is defined as the formation of progressively growing secondary tumors in distant organs and is a multi-step complex process. RKIP was reported to suppress tumor angiogenesis, local invasion, intravasation, bone and lung metastasis in murine models[149, 151, 152, 166]. Several molecular mechanisms have been reported to explain

RKIP’s role in metastasis, yet a comprehensive representation is still lacking. In our present study, we explore the effect of RKIP on Rho small GTPases. During early research, RhoA

GTPases were reported to be overexpressed in cancers and suggested an oncogenic role in tumor progression[118, 119]. Yet recently, several high-throughput sequencing efforts have identified loss-of-function mutations in RhoA gene,[132, 134-136] and oncogenic roles for negative RhoA regulators, RhoGAPs, have been reported as well[141, 142]. These recent findings indicate a context-dependent tumor suppressive role for RhoA in cancer.

One of the possible mechanisms how RhoA mediates its tumor suppressive activity is through stabilizing the E-cadherin based adherens junctions through mDia effector[111].

Adherens junctions (AJs) consists of membrane-spanning E-cadherin, whose cytoplasmic domains interact with p120 and β-catenin[103]. α-catenin is recruited through

76

its β-catenin binding and connects AJs to the cytoskeleton through actin-binding proteins such as α-actinin and vinculin[104]. AJs connect adjacent cell cytoskeletons through these proteins and confer mechanical strength. E-cadherin is the basic cell-cell adhesion unit in

AJs, and loss of E-cadherin is considered as a hallmark of epithelial-to-mesenchymal transition (EMT) and associated with poorly differentiated breast tumors and poor prognosis[227, 228]. Expression of E-cadherin is transcriptionally repressed by zinc-finger transcription factors, Snail and Slug,[229, 230] and post-transcriptionally regulated by miR-221 and miR-200a[231, 232]. Previously, E-cadherin had been reported to suppress cancer cell invasion and metastasis in vivo[233, 234].

During the current study, we demonstrated that RKIP specifically activates RhoA

GTPase in an ERK2- and GEFH1-dependent manner and this activation is partially required for RKIP-mediated suppression of breast cancer cell invasion and in vivo lung metastasis. Further, we were able to identify that RKIP’s anti-metastatic effects in breast cancer can be partially attributable to the RhoA-mediated membrane E-cadherin localization into adherens junctions and inhibition of tumor CCL5 expression. This study elucidates a novel mechanistic explanation for RKIP's role in breast cancer lung metastasis suppression and broadens our current understanding of downstream RKIP signaling.

3.3 Materials and Methods

Cell culture

BT20 and MDA-MB231 human breast cancer cell lines were cultured in

Dulbecco’s modified Eagle’s medium (DMEM, HyClone, UT, USA) with 10% FBS

(Atlanta Biologicals, GA, USA) and 1% penicillin-streptomycin (HyClone), and the 4T1 77

mouse breast cancer cell line was cultured in DMEM with 5% FBS, 5% calf serum (Lonza,

Switzerland) and 1% penicillin-streptomycin. Cells were grown in a humidified tissue culture incubator at 37°C in 5% CO2. 4T1 cells were kindly provided by Dr. Fred Miller

(Karmanos Cancer Institute, MI), and MDA-MB231 cells and BT20 cells were purchased from ATCC.

Immunofluorescence

Cells were plated on laminin-coated (Sigma, MO, USA) glass coverslips and grown until the desired confluency. Next, they were fixed with 100% methanol for 10 min, and incubated with anti-E-cadherin (3195, Cell Signaling, MA, USA), Alexa Fluor® 546 secondary antibody (1:5000, ThermoFisher, MA, USA), and DRAQ5® (1:2000, Cell

Signaling, MA, USA) nuclear stain. Fluorescence images were captured with a Leica TCS

SP5 multiphoton laser scanning confocal microscope and E-cadherin intensity on cell-cell membranes were analyzed with MetaMorph software (Molecular Devices).

Immunohistochemistry

FFPE sections were deparaffinized and rehydrated with an ethanol gradient.

Antigen retrieval was performed with citrate buffer (pH 6) and tissues were blocked with

10% normal goat serum (Vector Laboratories, CA, USA) for 2 hours. Overnight primary antibody incubations were carried out with Ki67 (SC-15402, Santa Cruz, CA, USA), CD-

19 (bs-4755R, Bioss, MA, USA), CK8 (TROMA-I, deposited to DSHB by Brulet, P./

Kemler, R.), E-cadherin (3195, Cell Signaling, MA, USA), GFP (ab290, Abcam, MA,

USA), CD31 (ab28364, Abcam), F4/80 (ThermoFisher, MA, USA), SMA (19245, Cell

Signaling), CCR5 (ab65850, Abcam), CXCR4 (MAB21651, R&D systems, MN, USA),

78

CCL5 (AP-20618, Acris, MD, USA) and CXCL12 (MAB350, R&D Systems). Sections were incubated with biotinylated secondary antibodies (Vector Laboratories) and ABC reagent (Vector Laboratories) and developed with DAB reagent (Sigma). Finally, they were counterstained with hematoxylin (Fisher Scientific, NH, USA), dehydrated with an ethanol gradient, and mounted with Permount™ (Fisher Scientific). Whole slides were scanned with an Olympus slide scanner and analyzed with ImageJ software. For co- localization studies, after primary antibody incubation, sections were incubated with corresponding Alexa Fluor® secondary antibodies (1:2000, 1 hour, ThermoFisher) and

DRAQ5® (1:2000, 10 minutes, Cell Signaling, MA, USA) nuclear stain. Fluorescence images were captured with a Leica TCS SP5 multiphoton laser scanning confocal microscope.

In vitro Matrigel invasion assay

The PET membranes (8µm pore size) of FluoroBlok™ cell culture inserts (351152,

Corning, NY, USA) were coated with 90 µL of diluted Matrigel (0.3mg/ml) (356234,

Corning), and incubated at 370C for 2-3 hours until solidified. Next, 1x104 of 4T1, BT20 or MDA-MB231 cells suspended in serum-free DMEM media were seeded on the solidified Matrigel layer. Then, 700µL of chemo-attractant medium (DMEM, 1% P/S and

10% FBS) was added to the lower chambers (353504, multiwell 24 well companion plate,

Corning), and the plate was incubated in a 370C incubator. After 24 hours of incubation, the insert bottoms were dipped in 1X PBS and stained in diluted Calcein AM (354217,

Corning) in PBS for 10 min. Fluorescence images of invaded cells were captured with an

EVOS inverted microscope, and analysis was done with ImageJ software.

79

Mammary fat pad injection for spontaneous metastasis assay

The detailed procedure for mammary fat pad injection was reported before[199].

For all our studies, BALB/c female mice (Taconic, NY, USA) of 5 weeks old were used and 1x105 of 4T1 GFP-LUC control knockdown and RhoA knockdown cells were injected.

After 29 days of injections, whole animals were imaged for luciferase bioluminescence

(BLI) with a Xenogen IVIS system as follows: Fresh D-luciferin (LUCK-1G, GoldBio,

MO, USA) working solution of 15 mg/ml was prepared in 1x PBS, and it was filter sterilized. Mice were inoculated with a 10 µl of luciferin solution per gram of body weight as an IP injection. After 10 min, they were anesthetized with Isoflurane and imaged. Images were processed and quantified with a Living Image 3.2 software.

After 30 days of injections, mice were euthanized, and blood and tissues were collected and processed for RNA or FFPE histology analysis. lungs were harvested and ex vivo BLI was performed as follows: Fresh D-luciferin working solution of 300 µg/ml was prepared in 1x PBS. Lungs were soaked in 1 ml of luciferin solution for 5 min in a 24 well plate and imaged with a Xenogen IVIS system. Images were processed and quantified with a Living Image 3.2 software. After BLI imaging, macro-metastases on the lung surface were imaged and manually counted and processed for FFPE.

The Department of Laboratory Animal Resources at the University of Toledo

Health Science Campus is accredited by the Association for the Assessment and

Accreditation of Laboratory Animal Care International (AAALAC) and operates in full compliance with the OLAW/PHS policy on the Humane Care and Use of Laboratory

Animals and the USDA Animal Welfare Act. All animal protocols used in this study were

80

approved by the University of Toledo Institutional Animal Care and Use Committee

(IACUC) and all experiments were performed in accordance with relevant guidelines and regulations in the approved protocols.

Proximity ligation assay

Cells were plated on laminin-coated (Sigma, MO, USA) glass coverslips and grown until the desired confluency. Next, they were fixed with 100% methanol for 10 min. After fixation, the cells were blocked using Background Sniper (Biocare Medical, Ref# BS966H) for 15’ at room temperature. Coverslips were washed in sterile filtered 1X PBS-Tween20 prior to primary antibody incubation. Primary antibodies E-cadherin (3195, Cell Signaling) and β-catenin (Cell Signaling) or p120 catenin (BD Sciences) were diluted 1:50 in the antibody dilution reagent provided in the Duolink In Situ PLA kit (Sigma-Aldrich, Cat#

DUO92101), and incubated overnight at 4oC. After the primary, we followed the recommended steps detailed in the Duolink PLA protocol available at the Sigma-Aldrich website. After elongation, the coverslips were washed in 1X PLA wash buffer B, followed by nuclear staining with DRAQ5 (Thermo Scientific, Cat# 62251) at a 1:3000 dilution in

1x PBS for 5’ at room temperature. Post-staining the coverslips were again washed in 1X

PLA wash buffer B. Lastly, a 10’ Phalloidin stain at a 1:300 dilution in 1X PBS was performed at 37oC in order to detail the outline and shape of the cells. Coverslips were then washed, dried in the dark, and mounted on glass slides using Prolong Gold anti-fade mountant (Life Technologies, Ref# P36934). Imaging of the coverslips was performed by using the Leica TCS SP5 multiphoton laser scanning confocal microscope provided by the

University of Toledo Advanced Microscopy and Imaging Center.

81

RhoA activity pulldown assay

The detailed protocol has been published before[235]. Briefly, media was aspirated and washed with 10ml of TBS buffer (with 1 mM MgCl2). 1 ml of buffer A (50 mM Tris pH 7.6, 500 mM NaCl, 0.1 % SDS, 0.5 % deoxycholate, 1% Triton X-100, 10 mM MgCl2,

Protease inhibitors) was added, and cells were scraped and collected to a tube. The lysate was sonicated for 10 sec at 40%, and spined down at max for 5 min. The protein concentration was measured with the BioRad DC assay. The pull-down tubes were set up to equalize both the total protein and total volume (1mg/1ml) and the lysate and the buffer was added into tubes with RBD/PBD beads (30 µg). Tubes were rotated for 45min at 4 C.

Beads were washed with buffer B (50 mM Tris pH 7.6, 150 mM NaCl, 1% Triton X-100,

10 mM MgCl2) 3 times (invert once, spin at 13200 for 30 sec). Beads were resuspended in

25ul of 2x SDS loading buffer, boiled for 5 min, and proceed with the western analysis.

RhoGEF activity pulldown assay

The detailed protocol has been published before[235]. Briefly, media was aspirated and washed with 10ml of HBS buffer (with 1 mM MgCl2). 1 ml of GEF buffer (20 mM

HEPES, 150 mM NaCl, 1% Triton X-100, 1mM DTT, 5 mM MgCl2, Protease inhibitors) was added, and cells were scraped and collected to a tube. The lysate was sonicated for 10 sec at 40%, and spined down at max for 5 min. The protein concentration was measured with the BioRad DC assay. The pull-down tubes were set up to equalize both the total protein and total volume (1mg/1ml) and the lysate and the buffer was added into tubes with

RhoAG17A beads (30 µg). Tubes were rotated for 45min at 4 C. Beads were washed with

82

GEF buffer 3 times (invert once, spin at 13200 for 30 sec). Beads were resuspended in 25ul of 2x SDS loading buffer, boiled for 5 min, and proceed with the western analysis.

Western blotting

Total cell lysates were prepared with the lysis buffer (20 mM Tris, pH 7.4, 150mM

NaCl, 2mM EDTA, and 1% Triton X-100), and proteins were separated with sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE). Then, separated proteins were electrophoretically transferred from the gel to a polyvinylidene fluoride membrane

(Millipore, MA, USA). Membranes were blocked and incubated overnight with primary antibodies diluted in PBS (pH 7.4), 0.2% Tween-20, 5% bovine serum albumin and 0.002% sodium azide. Primary antibodies used for western analyses were RhoA (1:1000, Cell

Signaling) and tubulin (1:2000, Sigma). Then, they were incubated with the appropriate horseradish peroxidase-conjugated secondary antibodies (1:2000, Cell Signaling) and developed with Immobilon western chemiluminescence substrate (Millipore) in a Bio-Rad

ChemiDoc EQ system.

Statistical analysis

Statistical calculations with two-tailed Student’s t-test were done using GraphPad

Prism software. All the data are presented as means with error bars representing standard error.

3.4 Results

3.4.1 RKIP promotes RhoA GTPase activation in breast cancer cells

83

Previously, RKIP has been shown to suppress breast cancer cell in-vitro invasion and metastasis in breast cancer xenograft and orthotopic murine models. Multiple studies have reported several independent mechanistic pathways regulating this phenotype[149,

150, 152, 236]. Being an upstream regulatory node of several major signaling pathways,

RKIP-mediated regulation of downstream molecular processes consists of a diverse array of signal transducers with constant cross-talks. Indeed, many of these identified mechanisms for RKIP-mediated regulation of metastasis cannot completely explain the

RKIP’s effect on metastasis, suggesting the existence of an intricate network of signaling that works cumulatively to affect metastasis downstream of RKIP.

During an effort to identify signaling pathways and targets that are differentially affected by RKIP expression/knockdown using microarray, our lab previously identified and reported downstream targets of RKIP that belong to chemokines and matrix metalloproteinases families[151, 152]. Yet, using a microarray approach, only the differences in transcription initiation and mRNA stability would be detectable, which is a major limitation associated with this assay. Certain signaling pathways are mainly regulated at the protein activity level, and Rho GTPases signaling is one such example.

These small GTPase proteins exist in two conformations, GTP-bound active, and the GDP- bound inactive forms, and the switch between two conformations is regulated by GEF and

GAP regulatory proteins. Being a major signal transducer in the actin and tubulin cytoskeletal regulation, Rho GTPases affect several physiological processes related to cell migration and invasion including, the formation of lamellipodia, filopodia, stress fibers, invadopodia, and focal adhesions. Therefore, RKIP might be affecting cell invasion and

84

metastasis through Rho GTPases. To investigate the role of RKIP in Rho GTPase signaling, we stably downregulated the RKIP expression by lentiviral transduction of specific shRNAs and measured the effect of RKIP knockdown on major Rho proteins activation in

168 FARN breast cancer cells. While reduced RKIP significantly reduced the activity of

RhoA, it did not affect the activation of Rac1 and Cdc42 GTPases (Fig 3.1a,b). We further validated RKIP’s effect on RhoA activation in several other breast cancer cell lines to eliminate possible off-target and cell line specific effects and observed a similar effect (Fig

3.1c,d).

3.4.2 RKIP required downstream RhoA to suppress breast cancer cell invasion and metastasis

RKIP has been previously reported to suppress breast cancer cell invasion. To investigate whether increased downstream RhoA activity is required for RKIP-mediated suppression of cell invasion, we knocked down RKIP expression in BT20 cells expressed with suboptimal levels of activated-RhoAQ63L construct. RKIP knockdown cells that have reduced RhoA activity significantly invaded through Matrigel, but the co-expression of the suboptimal levels of activated-RhoAQ63L suppressed the invasiveness of RKIP knockdown cells (Fig 3.2a). During an alternative approach, RKIP re-expression in 4T1 cells reduced their invasiveness, while suboptimal RhoA knockdown in RKIP expressed cells rescued the invasiveness of these cells (Fig 3.2b). These observations suggested a possible RhoA-dependency for RKIP to suppress breast cancer cell invasion.

85

Figure 3-1: RKIP promotes RhoA activation in breast cancer cells. a Representative western blots for the GTPase pull-down assay, after lentiviral transduction of indicated constructs, showing the calculated activity below the westerns. b RhoA activity with RKIP knockdown in 168 FARN cells. Representative results of more than three independent assays, error bars showing (mean ± SE). c Representative western blots for the GTPase pull-down assay, after lentiviral transduction of indicated constructs, showing the 86

calculated activity below the westerns. d Representative western blots for the GTPase pull- down assay, after adenoviral transduction of indicated constructs, showing the calculated activity below the westerns. *P<0.05, unpaired Student's t-test (two-tailed) for all analyses.

Figure 3-2: RKIP required downstream RhoA to suppress breast cancer cell invasion. a (left) Number of invaded cells through Matrigel in indicated BT20 cell lines (mean ± SE)

Representative results of two independent assays (right) Representative western blots for

RhoA and RKIP expression after lentiviral transduction of indicated constructs. b (left)

87

Number of invaded cells through Matrigel in indicated 4T1 cell lines (mean ± SE)

Representative results of two independent assays (right) Representative western blots for

RhoA and RKIP expression after lentiviral transduction of indicated constructs. *p<0.05, unpaired Student's t-test (two-tailed) for all analyses.

Using the 4T1 syngeneic orthotopic mouse model, we have previously shown that injection of stably RKIP re-expressed 4T1 cells suppressed the lung metastasis,[152] while stable RhoA knockdown 4T1 cell injections significantly elevated the metastatic burden in lungs (manuscript submitted). In line with that, by injecting a stable RKIP knockdown 4T1 cells, we detected a significant lung metastatic burden during the current study (Fig 3.3a,b).

To determine whether RKIP depends on downstream RhoA to suppress lung metastasis, we stably co-expressed a suboptimal RhoA knockdown construct, shRhoA 34, in 4T1 GFP-

LUC RKIP expressing cells, and used for orthotopic injections (Fig 3.3a). All the injected mice developed primary tumors, and 30 days after implantation, no significant differences were detected among tumors either in whole-animal bioluminescence imaging signal or in the expression of proliferation antigen Ki67, upon RKIP re-expression or RhoA knockdown (Fig 3.3c). On the contrary, ex-vivo bioluminescence imaging (BLI) of freshly harvested lungs revealed a significant difference in lung BLI signal in between groups suggesting differences in metastatic burden (Fig 3.3d). This was further confirmed by counting the number of macro nodules on the lung surface and quantifying the tumor metastases area in an H&E-stained lung section (Fig 3.3e). As expected, RKIP expressing

4T1 cells-injected mice showed the least metastatic burden. Yet, the concurrent suboptimal

RhoA knockdown in RKIP-expressed cells, caused a significant enhancement in the 88

metastatic signal in the mice compared to the RKIP expressing 4T1 cells-injected mice.

Collectively, these results suggested that downstream RhoA activity is required for RKIP- mediated suppression of breast cancer cell invasion and metastasis.

89

Figure 3-3: RKIP required downstream RhoA to suppress breast cancer metastasis. a Representative western blots for RhoA and RKIP expression after lentiviral transduction of indicated constructs. Numbers are shown for quantified bands normalized with tubulin. b (left) primary tumor weight, (right) number of lung metastatic nodules, after 30 days of implantation of Balb/c mice orthotopically injected with 4T1 gfp-luc cells carrying indicated lentiviral modifications. N=6. c (left) Representative BLI images of tumors,

(middle) percentage of Ki67+ cells per tumor section, (right) whole animal photon flux quantification (mean ± SE) of BLI images shown on left for the signals in primary tumors.

N=4. d (top) Representative ex vivo BLI images of lungs of the mice orthotopically injected with 4T1 gfp-luc cells carrying indicated lentiviral modifications, (bottom) and their corresponding representative hematoxylin and eosin (H&E) staining of lung cross-sections showing metastases highlighted with black. e (left) ex vivo lung photon flux quantification

(mean ± SE) of BLI images shown in d, (middle) the number of lung metastatic nodules

(mean ± SE) of these mice, (right) total tumor metastases area calculated from the H&E sections shown in d. N=4. *P<0.05, unpaired Student's t-test (two-tailed) for all analyses. 90

3.4.3 RKIP suppresses breast cancer cell invasion through RhoA-mediated regulation of E-cadherin

The findings so far suggested that RKIP depends on downstream RhoA activity to suppress cell invasion. Previously, RhoA has been reported to stabilize E-cadherin on to adherens junction complexes (AJs)[106]. To investigate this possibility, we stained for E- cadherin in BT20 cells grown on coverslips using a specific antibody. Control BT20 cells expressed E-cadherin-rich AJs along the cell-cell boundaries in confluent cultures, and upon the stable downregulation of RhoA expression by lentiviral transduction of specific shRNAs, there was a significant reduction in the membrane E-cadherin localization (Fig

3.4a,b). In contrast, stable expression of the activated-RhoAQ63L construct in BT20 further increased the E-cadherin localization on AJs significantly (Fig 3.4a,c). Yet, no significant difference in the E-cadherin protein expression was detected upon RhoA knockdown or RhoAQL expression by the western blotting (Fig 3.4a). Since RKIP promotes RhoA activation, we expected RKIP to mimic the RhoA’s effect on E-cadherin localization as well. Indeed, the stable downregulation of RKIP expression by lentiviral transduction of specific shRNAs showed a significant loss in membrane E-cadherin localization, while stable expression of a Flag-tagged RKIP construct enhanced the localization of E-cadherin on junctions (Fig 3.4d). Interestingly, RKIP’s effect on E- cadherin localization was only observed in the triple-negative epithelial-like subset of breast cancer cell lines as BT20, SUM149, and MDA-MB468, and not in other epithelial- like breast cancer cell lines such as MCF7, and T47D (Fig 3.5a-c). In order to determine whether RKIP’s ability to regulate E-cadherin localization depends on the downstream

91

RhoA, BT20 cells grown on coverslips were immunostained for E-cadherin. Upon stable knockdown of RKIP, E-cadherin was less localized on the membrane, and this effect could be rescued by the stable co-expression of RhoAQ63L in RKIP knockdown cells (Fig 3.6).

Collectively, these results indicated that, in BT20 cells, RKIP mimics and functions through downstream RhoA to promote E-cadherin localization and stabilization into AJs.

92

Figure 3-4: RKIP suppresses breast cancer cell invasion through RhoA-mediated regulation of E-cadherin. a Representative western blots for RhoA and E-cadherin expression after lentiviral transduction of indicated constructs. Numbers are shown for quantified bands normalized with tubulin. b (left) Representative immunofluorescence images for BT20 after lentiviral transduction of indicated constructs, (right) average E- cadherin quantification on cell-cell membranes using MetaMorph analysis with mean ±

SE. c (top) Representative immunofluorescence images for BT20 after lentiviral transduction of indicated constructs, (bottom) average E-cadherin quantification on cell- cell membranes using MetaMorph analysis with mean ± SE. d (left) Representative immunofluorescence images for BT20 after lentiviral transduction of indicated constructs,

(right) average E-cadherin quantification on cell-cell membranes using MetaMorph analysis with mean ± SE. e (left) Representative western blots for RhoA and E-cadherin expression after lentiviral transduction of indicated constructs. Numbers shown for quantified bands normalized with tubulin, (right) Relative RKIP and E-cadherin mRNA levels normalized with hGAPDH (mean ± SE), as quantified by qRT-PCR BT20 cells transduced with indicated lentiviral particles. *P<0.05, unpaired Student's t-test (two- tailed) for all analyses.

93

Figure 3-5: RKIP’s effect on E-cadherin is breast cancer cell-type specific. a (left)

Representative immunofluorescence images for MCF7 after lentiviral transduction of indicated constructs, (middle) representative immunofluorescence images for T47D after lentiviral transduction of indicated constructs, (right) representative western blots for RKIP expression after lentiviral transduction of indicated constructs. b (top) Representative

94

immunofluorescence images for SUM 149 after lentiviral transduction of indicated constructs, (bottom) representative immunofluorescence images for MDA-MB468 after lentiviral transduction of indicated constructs, (right) representative western blots for RKIP expression after lentiviral transduction of indicated constructs. c Average E-cadherin quantification on cell-cell membranes using MetaMorph analysis with mean ± SE for indicated cell lines with modifications. *P<0.05, unpaired Student's t-test (two-tailed) for all analyses.

Adherens junction complex is composed of several proteins. E-cadherin, the major functional unit in the AJs, has an extracellular region with five cadherin domains that homophilically interact with cadherin domains in neighboring cells, and an intracellular cytoplasmic tail that directly binds p120 catenin and β-catenin through conserved binding sites, and indirectly bind α-catenin through β-catenin[103]. In order to further validate the

RKIP’s ability to stabilize E-cadherin on to AJ complex, a proximity ligation assay (PLA) was performed with specific antibody pairs for E-cadherin/β-catenin and E-cadherin/p120 catenin. During this assay, a physiologically adjacent localization of E-cadherin with other protein will give a specific PLA signal. Upon stable RKIP knockdown in BT20, PLA signal for E-cadherin and β-catenin localization, and E-cadherin and p120 catenin localization significantly reduced confirming that the membrane E-cadherin is in fact associated in the adherens junctional complex (Fig 3.7).

95

Figure 3-6: RKIP promotes E-cadherin membrane localization through downstream

RhoA. (top) Representative immunofluorescence images for BT20 after lentiviral transduction of indicated constructs, (bottom) average E-cadherin quantification on cell- cell membranes using MetaMorph analysis with mean ± SE for indicated cell lines with modifications. *P<0.05, unpaired Student's t-test (two-tailed) for all analyses.

Previously, E-cadherin was reported to suppress breast cancer cell invasion and metastasis in vivo[233, 234]. In both 4T1 and BT20 cells, E-cadherin suppresses in vitro cell invasion in a dose-dependent manner (Fig 3.8a). To investigate whether RKIP- mediated junctional E-cadherin localization causes the suppression of breast cancer cell invasion, we stably downregulated the E-cadherin expression in suboptimal levels in

RKIP-expressed BT20 cells and performed an invasion assay (Fig 3.8b). RKIP expression 96

suppressed BT20 cell invasion, while concurrent E-cadherin suboptimal stable knockdown in RKIP-expressed cells rescued the invasiveness of these cells, indicating that RKIP- mediated junctional E-cadherin localization is required for the suppression of breast cancer cell in vitro invasion.

Figure 3-7: RKIP promotes E-cadherin membrane association with β-catenin and p120-catenin in adherens junctions. a Representative top view images for immunofluorescence images during a PLA assay with E-cadherin/β-catenin and E- cadherin/p120-catenin in BT20 after lentiviral transduction of indicated constructs. b A combined side view of a one representative immunofluorescence image during the corresponding PLA assay, where all the possible single section side views were stacked to 97

generate a one combined side-view map of the image showing PLA channel (yellow) and nuclear channel (blue).

Figure 3-8: RKIP-mediated inhibition of invasion depends on the membrane E- cadherin localization. a (left) (top) Number of invaded cells through Matrigel in 4T1 cells transduced with indicated constructs (mean ± SE) and (bottom) their representative western blots for E-cadherin expression, (right) (top) number of invaded cells through Matrigel in

BT20 cells transduced with indicated constructs (mean ± SE) and (bottom) their 98

representative western blots for E-cadherin expression. b Number of invaded cells through

Matrigel in indicated BT20 cell lines (mean ± SE). *P<0.05, unpaired Student's t-test (two- tailed) for all analyses.

3.4.4 RKIP suppresses breast cancer metastasis partially through downstream RhoA- regulated mechanisms

E-cadherin is a well-established tumor metastasis suppressor. Hence RKIP- mediated suppression in breast cancer metastasis might be partially attributed to RKIP’s ability to enhance E-cadherin junctional localization and stabilization of adherens junctions through RhoA in primary tumors. To investigate this possibility, we examined the E- cadherin expression in primary tumors harvested from orthotopically injected mice with stable RKIP knockdown or RhoA knockdown 4T1 cells. As expected, the total E-cadherin expression of RKIP knockdown and RhoA knockdown primary tumors was significantly reduced compared to the knockdown control tumors (Fig 3.9a). Interestingly, a closer look at the E-cadherin distribution in the tumor tissue revealed a mostly membrane localization pattern for E-cadherin in knockdown control tumors, while RhoA and RKIP knockdown tumors showed mostly a cytoplasmic expression of E-cadherin (Fig 3.9b). Collectively, these observations suggested a physiologically relevant correlation between RKIP/RhoA expression and membrane E-cadherin expression and localization, and their subsequent metastatic potential in 4T1 primary breast tumors.

In order to validate this correlation, we stained E-cadherin in primary tumors from mice injected with RhoA knockdown RKIP expressed 4T1 cells. As expected, RKIP-

99

expressed primary tumors showed a significantly higher membrane E-cadherin expression, while concurrent RhoA knockdown in RKIP-expressed tumors caused a decrease in membrane E-cadherin expression (Fig 3.9c). These observations further suggested that

RKIP-mediated suppression in breast cancer metastasis can be partially attributed to

RKIP’s ability to enhance junctional E-cadherin localization through RhoA in primary tumors.

100

Figure 3-9: RKIP suppresses breast cancer metastasis partially through downstream

E-cadherin. a (left) E-cadherin+ area per tumor field of view (FOV) (mean ± SE) for immunohistochemical (IHC) staining sections of breast primary tumor of mice orthotopically injected with 4T1 gfp-luc cells carrying indicated lentiviral modifications

101

and (right) their corresponding representative IHC stained primary tumors. b

Representative zoomed-in images of the primary tumors stained with E-cadherin and hematoxylin. c E-cadherin+ area per tumor field of view (FOV) (mean ± SE) for immunohistochemical (IHC) staining sections of breast primary tumor of mice orthotopically injected with 4T1 gfp-luc cells carrying indicated lentiviral modifications.

N=4, *P<0.05, unpaired Student's t-test (two-tailed) for all analyses.

We have also previously reported that RKIP partially suppresses breast cancer lung metastasis and F4/80+ tumor macrophage infiltration by inhibiting CCL5 expression by the

4T1 tumor cells[152]. In accord with that, the present study showed increased tumor CCL5 expression and F4/80+ macrophage infiltration in the primary tumors harvested from the mice injected with stably RKIP knockdown 4T1 cells (Fig 3.10a). During another independent study, we reported a negative correlation between primary tumor RhoA and

CCL5 expression (manuscript submitted). Since RKIP promotes RhoA, RKIP might be inhibiting the CCL5 expression through activating RhoA. To test this possibility, we immunostained for CCL5 expression in primary tumors harvested from mice injected with

RKIP-expressed and RhoA knockdown 4T1 cells. As expected, RKIP-expressed tumors have significantly reduced CCL5 expression compared to control tumors (Fig 3.10b). Yet, the concurrent RhoA knockdown in RKIP-expressed tumors rescued the tumor CCL5 expression comparable to control tumor levels, indicating that RKIP-mediated inhibition of CCL5 expression is facilitated through downstream RhoA.

102

Figure 3-10: RKIP suppresses breast cancer metastasis partially through downstream CCL5 and F4/80. a (left) CCL5+ area per tumor field of view (FOV) (mean

± SE) for immunohistochemical (IHC) staining sections of breast primary tumor of mice orthotopically injected with 4T1 gfp-luc cells carrying indicated lentiviral modifications,

(right) F4/80+ area per tumor field of view (FOV) (mean ± SE) for immunohistochemical

(IHC) staining sections of breast primary tumor of mice orthotopically injected with 4T1 gfp-luc cells carrying indicated lentiviral modifications. b (left) CCL5+ area per tumor field of view (FOV) (mean ± SE) for immunohistochemical (IHC) staining sections of breast 103

primary tumor of mice orthotopically injected with 4T1 gfp-luc cells carrying indicated lentiviral modifications,(right) F4/80+ area per tumor field of view (FOV) (mean ± SE) for immunohistochemical (IHC) staining sections of breast primary tumor of mice orthotopically injected with 4T1 gfp-luc cells carrying indicated lentiviral modifications.

*P<0.05, unpaired Student's t-test (two-tailed) for all analyses.

3.4.5 RKIP activates RhoA through ERK2 and GEFH1

As an attempt to understand the intermediate signal transducers between RKIP and

RhoA in regulating E-cadherin localization and breast cancer cell invasion/metastasis, we examined the effect of ERK1/2 on E-cadherin localization in BT20 cells. ERK is a downstream kinase of the MAPK pathway where RKIP acts as a negative regulator by binding and inhibiting the kinase activity of Raf-1 kinase (Fig 3.11a). We stably downregulated the ERK1/2 expression by lentiviral transduction of specific shRNAs for

ERK1 and ERK2 and stained these cells with E-cadherin antibody. ERK1 knockdown did not have a substantial effect on the E-cadherin expression and the membrane localization in BT20, while specific ERK2 knockdown significantly elevated the membrane E-cadherin localization compared to the knockdown control cells, indicating an ERK2-specific negative regulatory effect on membrane E-cadherin (Fig 3.11b). In order to examine whether RKIP regulates E-cadherin and subsequent cell invasion through ERK2, we stably co-expressed RKIP- and ERK2-specific shRNAs through lentiviral transduction and studied the effect on E-cadherin localization and the cell invasion in BT20 (Fig 3.11c).

Reduced expression of RKIP decreased the membrane E-cadherin localization and 104

significantly increased their invasiveness, while concurrent inhibition of ERK2 in RKIP knockdown cells restored the membrane E-cadherin and reversed their invasiveness, indicating that RKIP regulates E-cadherin and subsequent cell invasion through ERK2.

105

Figure 3-11: RKIP activates RhoA through ERK2. a Representative western blots for phopho-ERK and total ERK expression after lentiviral transduction of indicated constructs.

Numbers are shown for quantified bands normalized with tubulin. b (top) Representative immunofluorescence images for BT20 after lentiviral transduction of indicated constructs,

(bottom) (left) representative western blots for total ERK expression after lentiviral transduction of indicated constructs, (bottom) (right) average E-cadherin quantification on cell-cell membranes using MetaMorph analysis with mean ± SE. c (left) Representative immunofluorescence images for BT20 after lentiviral transduction of indicated constructs,

(right) number of invaded cells through Matrigel in BT20 cells transduced with indicated constructs (mean ± SE). *P<0.05, unpaired Student's t-test (two-tailed) for all analyses.

Cellular activity of RhoA GTPase is primarily regulated by RhoGEF positive regulators and RhoGAP negative regulators. Vastly dynamic spatiotemporal regulation through these regulators determines the RhoA-mediated downstream signaling effects.

ERK has been previously reported to inhibit RhoA activation through negatively regulating

GEFH1[237]. Another RhoGEF, Vav2, has been reported to activate RhoA downstream of growth factor receptor signaling[238]. Moreover, Vav2 gets activated upon the expression of RKIP in BT20 cells (Fig 3.12a). Hence, we examined the effect of GEFH1 and Vav2 on membrane E-cadherin localization by stably knocking down their expression through lentiviral transduction of specific shRNAs. Reduced expression of Vav2 had no effect on the E-cadherin, whereas reduced GEFH1 expression caused a reduced localization of E- cadherin on cell membranes indicating GEFH1 as a positive regulator of E-cadherin localization (Fig 3.12b). To test whether RKIP regulates E-cadherin and subsequent cell 106

invasion through GEFH1, we stably knockdown GEFH1 expression in RKIP expressing

BT20 cells and studied the effect on E-cadherin localization and the cell invasion. RKIP expression increased the membrane E-cadherin and repressed cell invasion, while concurrent knockdown of GEFH1 reversed the E-cadherin localization and restored their invasiveness (Fig 3.12c). This indicated that GEFH1 is a mediator of RKIP’s effects on E- cadherin and cell invasion. Collectively, these results suggested the existence of a signal transduction from RKIP to breast cancer invasion/metastasis through ERK2, GEFH1,

RhoA and E-cadherin.

107

Figure 3-12: RKIP activates RhoA through GEFH1. a Representative western blots for

Vav2 GEF activity pull-down in BT20 after adenoviral transduction of indicated constructs. b (top) Representative western blots for GEFH1 and Vav2 expression inBT20

108

after lentiviral transduction of indicated constructs, (bottom) representative immunofluorescence images for BT20 after lentiviral transduction of indicated constructs. c (top) Representative immunofluorescence images for BT20 after lentiviral transduction of indicated constructs, (bottom) (left) number of invaded cells through Matrigel in BT20 cells transduced with indicated constructs (mean ± SE), (bottom) (right) average E- cadherin quantification on cell-cell membranes using MetaMorph analysis with mean ±

SE. *P<0.05, unpaired Student's t-test (two-tailed) for all analyses.

3.5 Discussion

During this current study, we mechanistically investigated RKIP as a metastasis suppressor in breast cancer. Previously, it has been shown to suppress breast cancer cell in-vitro invasion and metastasis in breast cancer xenograft and orthotopic murine models, and several possible mechanisms have been proposed to explain the metastasis suppresser behavior of RKIP in breast cancer[149-152, 238]. Even though these have been depicted as linear signaling pathways, physiologically they are part of a larger and complex network linking feedback regulatory mechanisms, crosstalk, and other RKIP targets. Indeed, many of these proposed mechanisms for RKIP-mediated regulation of metastasis cannot themselves entirely explain RKIP’s effect on metastasis, suggesting a cumulative regulation of the phenotype. Using a triple negative basal epithelial-like breast cancer cell line, BT20, we showed that RKIP suppresses cell invasion by upregulating the junctional localization of E-cadherin into adherens junctions. Similarly, with the 4T1 orthotopic murine model of triple negative basal breast cancer, we demonstrated a positive correlation

109

between RKIP expression, and expression and membrane localization of E-cadherin in primary breast tumors, and a subsequent negative correlation with their metastatic potential. However, RKIP has been previously reported to have no effect on E-cadherin expression in triple negative basal mesenchymal-like breast cancer cell MDA-MB231 and luminal breast cancer cell MCF7[149]. But, it should be noted that in MDA-MB231, the

E-cadherin gene is epigenetically silenced through promoter methylations, while MCF7 has a functional E-cadherin gene[239]. In accord with the previous reports, stable RKIP knockdown in luminal breast cancer cells, MCF7 and T47D, showed no effect on E- cadherin localization, while stable RKIP knockdown in triple negative basal epithelial-like cells, SUM149 and MDA-MB468, significantly reduced the membrane E-cadherin localization. Thus, our data implicate a breast cancer subtype-specific regulation of E- cadherin expression and membrane localization by RKIP.

The immunostaining data revealed less membrane localization of E-cadherin with

RKIP knockdown, yet the functional consequences of this phenotype in terms of the stability of adherens junctions was unknown. Hence, a proximity ligation assay was performed to examine the effect of stable RKIP knockdown on adherens junction complex, and a reduced physical association between E-cadherin and β-catenin, and E-cadherin and p120 catenin was detected, which indicates a loss of adherens junction complex formation with RKIP downregulation.

Loss of E-cadherin is a hallmark of epithelial-to-mesenchymal transition (EMT), a highly coordinated cell biological program that activated during tumor malignant progression. EMT transition is also associated with upregulation of mesenchymal cell traits

110

as vimentin and N-cadherin and drives by EMT-inducing transcription factors such as

Snail, Slug, Twist and Zeb1. Previously RKIP was reported to repress Snail through signaling mediated by Myc, let-7 and HMGA-2 in MDA-MB231 cells,[149] and Snail was reported to negatively regulate RKIP transcription through a feedback loop in metastatic prostate cancer cells[159]. During the current study with BT20 cells, RKIP downregulation was associated with reduced E-cadherin expression and membrane localization. But the expression of vimentin and N-cadherin was not affected by reduced RKIP. Hence, in this case, reduced RKIP might be initiating a partial EMT-like phenotype.

Previously we identified a mechanism where RKIP inhibits tumor angiogenesis,

F4/80+ macrophage infiltration, and lung metastasis by downregulating CCL5 expression in the tumor[152]. But the signal intermediates between RKIP and CCL5 were unknown.

Recent findings from our lab identified tumor RhoA expression is negatively correlated with tumor CCL5 expression, and concurrent RhoA knockdown in the RKIP expressed tumors rescued the RKIP-mediated CCL5 suppression. Hence RKIP appears to inhibit

CCL5 through RhoA GTPases.

In summary, the present study identified a physiological mechanism where RKIP suppresses breast cancer lung metastasis through ERK2, GEFH1, RhoA, and E-cadherin.

Some of these signaling components were known to affect breast cancer invasion and metastasis, but the current study elucidated a complete signaling pathway.

111

Chapter 4

Summary

Breast cancer metastasis accounts for nearly 90% of breast cancer-related deaths.

Breast cancer is prone to metastasize to lungs, bones, brain, and liver, and tumor dissemination into local lymph nodes initiates the malignant phenotype. RKIP was first identified as a metastasis suppressor of prostate cancer but later was shown to suppress breast cancer lung and bone metastasis as well. Several mechanisms have been identified to explain the anti-metastatic effects of RKIP, yet a comprehensive understanding is lacking. This study was initiated as an attempt to identify additional physiological downstream targets of RKIP in mediating breast cancer lung metastasis suppression. Small

Rho GTPases were studied given their role in regulating cytoskeleton dynamics, cell motility and invasion, cell adhesion, gene expression, and cell cycle progression.

Preliminary cell-based studies discovered RKIP as a positive regulator of RhoA GTPase activity, and the effect was RhoA-specific. Also, RKIP-mediated inhibition of breast cancer cell invasion was found to be RhoA-dependent. These initial findings suggested a physiological anti-invasive role for RhoA in breast cancer cells. However, given its 112

common acceptance as a pro-invasive tumor promoter, these results were rather unexpected and need to be further validated.

Hence, we undertook the task to examine the role of RhoA in breast cancer progression and metastasis. RhoA was found to have no effect on breast cancer cell proliferation and colony formation in vitro. Yet, during the 4T1 orthotopic breast cancer mouse model, RhoA knockdown increased sentinel lymph node dissemination and lung metastatic nodule formation, indicating an anti-metastatic role for RhoA. In an effort to dissect out RhoA’s effect on early and late steps of the metastatic cascade, an experimental metastasis assay was performed. Both RhoA knockdown and control knockdown 4T1 cells were able to colonize lungs with similar kinetics, differentiating the effect of RhoA on early steps of the metastatic cascade, possibly the initial dissemination of invasive tumor cells from the primary tumor. To test this possibility, we characterized the primary tumors and the tumor microenvironment and identified several distinguishing phenotypes in the

RhoA knockdown primary tumors. Upon reduced RhoA expression, tumor E-cadherin expression and the membrane localization were reduced. Additionally, RhoA knockdown primary tumors expressed elevated levels of CXCR4 and CCR5 pro-tumor chemokine receptors, and CXCL12 and CCL5 chemo-attractive ligands, and increased infiltration of

SMA+ cells and F4/80+ macrophages. These observations suggested that the anti-metastatic behavior of RhoA in breast cancer might be attributable to increased E-cadherin expression and/or reduced expression of CXCR4/CXCL12 and CCR5/CCL5 pro-tumor chemotactic axes.

113

Extensive cell-based studies identified RKIP as a positive regulator of E-cadherin expression and localization into adherens junctions in triple negative basal epithelial-like breast cancer subtype cells. Further, this phenotype was shown to be driven by ERK2,

GEFH1 and RhoA-dependent manner. RhoA-mediated junctional E-cadherin localization was found to be one mechanism how RKIP inhibits breast cancer cell invasion.

Furthermore, our study demonstrated the RhoA-dependency of the anti-metastatic effect of RKIP in 4T1 orthotopic breast cancer mouse model. Mechanistically, this anti-metastatic phenotype of RKIP was shown to be mediated by the cumulative effects of positively regulated E-cadherin and negatively regulated CCL5 by RhoA.

Collectively, the present study recognized RhoA as a novel and physiological suppressor of the breast cancer lymph node and lung metastasis. To our understanding, this is the first study to propose a mechanistic explanation for the anti-metastatic role of RhoA in breast cancer. Further, we identified RhoA as a downstream regulator of RKIP in suppressing breast cancer cell invasion and metastasis. Mechanistically, our findings suggest that RhoA is a signaling intermediate for RKIP and other RKIP-independent upstream signaling pathways, since RhoA downstream effects on CXCR4/CXCL12 and

CCR5 expression, and SMA+ cell infiltration were RKIP-independent.

114

Figure 4-1: Schematic representation of the summary of the research findings showing the signaling mechanisms affect breast cancer invasion/metastasis. Each node corresponds to a target studied. Blue arrows show the RKIP-dependent signaling, while red arrows are RKIP-independent signaling.

115

References

1 Wennerberg K, Rossman KL, Der CJ. The Ras superfamily at a glance. Journal of

cell science 2005; 118: 843-846.

2 Colicelli J. Human RAS superfamily proteins and related GTPases. Science's

STKE : signal transduction knowledge environment 2004; 2004: Re13.

3 Garcia-Mata R, Boulter E, Burridge K. The 'invisible hand': regulation of RHO

GTPases by RHOGDIs. Nat Rev Mol Cell Biol 2011; 12: 493-504.

4 Ridley AJ. Historical overview of Rho GTPases. Methods in molecular biology

(Clifton, NJ) 2012; 827: 3-12.

5 Madaule P, Axel R. A novel ras-related gene family. Cell 1985; 41: 31-40.

6 Zhang N, Liang J, Tian Y, Yuan L, Wu L, Miao S et al. A novel testis-specific

GTPase serves as a link to proteasome biogenesis: functional characterization of

RhoS/RSA-14-44 in spermatogenesis. Mol Biol Cell 2010; 21: 4312-4324.

116

7 Valster AH, Hepler PK, Chernoff J. Plant GTPases: the Rhos in bloom. Trends

Cell Biol 2000; 10: 141-146.

8 Wheeler AP, Ridley AJ. Why three Rho proteins? RhoA, RhoB, RhoC, and cell

motility. Exp Cell Res 2004; 301: 43-49.

9 Adamson P, Paterson HF, Hall A. Intracellular localization of the P21rho

proteins. The Journal of cell biology 1992; 119: 617-627.

10 Rathinam R, Berrier A, Alahari SK. Role of Rho GTPases and their regulators in

cancer progression. Frontiers in bioscience (Landmark edition) 2011; 16: 2561-

2571.

11 Vega FM, Ridley AJ. The RhoB small GTPase in physiology and disease. Small

GTPases 2018; 9: 384-393.

12 Didsbury J, Weber RF, Bokoch GM, Evans T, Snyderman R. , a novel ras-

related family of proteins that are botulinum toxin substrates. The Journal of

biological chemistry 1989; 264: 16378-16382.

117

13 Bosco EE, Mulloy JC, Zheng Y. Rac1 GTPase: a "Rac" of all trades. Cellular and

molecular life sciences : CMLS 2009; 66: 370-374.

14 Bid HK, Roberts RD, Manchanda PK, Houghton PJ. RAC1: an emerging

therapeutic option for targeting cancer angiogenesis and metastasis. Molecular

cancer therapeutics 2013; 12: 1925-1934.

15 Vega FM, Ridley AJ. Rho GTPases in cancer cell biology. FEBS letters 2008;

582: 2093-2101.

16 Burridge K, Wennerberg K. Rho and Rac take center stage. Cell 2004; 116: 167-

179.

17 Hodge RG, Ridley AJ. Regulating Rho GTPases and their regulators. Nat Rev

Mol Cell Biol 2016; 17: 496-510.

18 Eva A, Vecchio G, Rao CD, Tronick SR, Aaronson SA. The predicted DBL

oncogene defines a distinct class of transforming proteins. Proc Natl

Acad Sci U S A 1988; 85: 2061-2065.

118

19 Hart MJ, Eva A, Evans T, Aaronson SA, Cerione RA. Catalysis of guanine

nucleotide exchange on the CDC42Hs protein by the dbl oncogene product.

Nature 1991; 354: 311-314.

20 Goicoechea SM, Awadia S, Garcia-Mata R. I'm coming to GEF you: Regulation

of RhoGEFs during cell migration. Cell adhesion & migration 2014; 8: 535-549.

21 Meller N, Merlot S, Guda C. CZH proteins: a new family of Rho-GEFs. Journal

of cell science 2005; 118: 4937-4946.

22 Cote JF, Motoyama AB, Bush JA, Vuori K. A novel and evolutionarily conserved

PtdIns(3,4,5)P3-binding domain is necessary for signalling. Nat Cell

Biol 2005; 7: 797-807.

23 Cook DR, Rossman KL, Der CJ. Rho guanine nucleotide exchange factors:

regulators of Rho GTPase activity in development and disease. Oncogene 2014;

33: 4021-4035.

24 Lander ES, Linton LM, Birren B, Nusbaum C, Zody MC, Baldwin J et al. Initial

sequencing and analysis of the human genome. Nature 2001; 409: 860-921.

119

25 Jacobs T, Hall C. Rho GAPs — Regulators of Rho GTPases and More. In:

Manser E (ed). RHO Family GTPases. Springer Netherlands: Dordrecht, 2005, pp

93-112.

26 Croft DR, Olson MF. Transcriptional regulation of Rho GTPase signaling.

Transcription 2011; 2: 211-215.

27 Ongusaha PP, Kim HG, Boswell SA, Ridley AJ, Der CJ, Dotto GP et al. RhoE is

a pro-survival p53 target gene that inhibits ROCK I-mediated apoptosis in

response to genotoxic stress. Current biology : CB 2006; 16: 2466-2472.

28 Ahn J, Choi JH, Won M, Kang CM, Gyun MR, Park HM et al. The activation of

p38 MAPK primarily contributes to UV-induced RhoB expression by recruiting

the c-Jun and p300 to the distal CCAAT box of the RhoB promoter. Biochemical

and biophysical research communications 2011; 409: 211-216.

29 Vasilaki E, Papadimitriou E, Tajadura V, Ridley AJ, Stournaras C, Kardassis D.

Transcriptional regulation of the small GTPase RhoB gene by TGF{beta}-

induced signaling pathways. FASEB journal : official publication of the

Federation of American Societies for Experimental Biology 2010; 24: 891-905.

120

30 Chan CH, Lee SW, Li CF, Wang J, Yang WL, Wu CY et al. Deciphering the

transcriptional complex critical for RhoA gene expression and cancer metastasis.

Nat Cell Biol 2010; 12: 457-467.

31 Liu M, Bi F, Zhou X, Zheng Y. Rho GTPase regulation by miRNAs and covalent

modifications. Trends in cell biology 2012; 22: 365-373.

32 Care A, Catalucci D, Felicetti F, Bonci D, Addario A, Gallo P et al. MicroRNA-

133 controls cardiac hypertrophy. Nature medicine 2007; 13: 613-618.

33 Qian L, Wythe JD, Liu J, Cartry J, Vogler G, Mohapatra B et al. Tinman/Nkx2-5

acts via miR-1 and upstream of Cdc42 to regulate heart function across species. J

Cell Biol 2011; 193: 1181-1196.

34 Yu JY, Chung KH, Deo M, Thompson RC, Turner DL. MicroRNA miR-124

regulates neurite outgrowth during neuronal differentiation. Exp Cell Res 2008;

314: 2618-2633.

35 Kong W, Yang H, He L, Zhao JJ, Coppola D, Dalton WS et al. MicroRNA-155 is

regulated by the transforming growth factor beta/Smad pathway and contributes

to epithelial cell plasticity by targeting RhoA. Mol Cell Biol 2008; 28: 6773-6784.

121

36 Liu M, Lang N, Chen X, Tang Q, Liu S, Huang J et al. miR-185 targets RhoA and

Cdc42 expression and inhibits the proliferation potential of human colorectal

cells. Cancer letters 2011; 301: 151-160.

37 Valastyan S, Chang A, Benaich N, Reinhardt F, Weinberg RA. Concurrent

Suppression of Integrin α<sub>5</sub>, , and RhoA

Phenocopies the Effects of miR-31 on Metastasis. Cancer Research 2010; 70:

5147.

38 Park SY, Lee JH, Ha M, Nam JW, Kim VN. miR-29 miRNAs activate p53 by

targeting p85 alpha and CDC42. Nature structural & molecular biology 2009; 16:

23-29.

39 Liu M, Lang N, Qiu M, Xu F, Li Q, Tang Q et al. miR-137 targets Cdc42

expression, induces cell cycle G1 arrest and inhibits invasion in colorectal cancer

cells. International journal of cancer 2011; 128: 1269-1279.

40 Connolly EC, Van Doorslaer K, Rogler LE, Rogler CE. Overexpression of miR-

21 promotes an in vitro metastatic phenotype by targeting the tumor suppressor

RHOB. Molecular cancer research : MCR 2010; 8: 691-700.

122

41 Rocks O, Gerauer M, Vartak N, Koch S, Huang ZP, Pechlivanis M et al. The

palmitoylation machinery is a spatially organizing system for peripheral

membrane proteins. Cell 2010; 141: 458-471.

42 Lang P, Gesbert F, Delespine-Carmagnat M, Stancou R, Pouchelet M, Bertoglio

J. Protein kinase A phosphorylation of RhoA mediates the morphological and

functional effects of cyclic AMP in cytotoxic lymphocytes. Embo j 1996; 15: 510-

519.

43 Takemoto K, Ishihara S, Mizutani T, Kawabata K, Haga H. Compressive stress

induces dephosphorylation of the myosin regulatory light chain via RhoA

phosphorylation by the /protein kinase A signaling pathway.

PLoS One 2015; 10: e0117937.

44 Tkachenko E, Sabouri-Ghomi M, Pertz O, Kim C, Gutierrez E, Machacek M et

al. Protein kinase A governs a RhoA-RhoGDI protrusion-retraction pacemaker in

migrating cells. Nat Cell Biol 2011; 13: 660-667.

45 Ellerbroek SM, Wennerberg K, Burridge K. Serine phosphorylation negatively

regulates RhoA in vivo. The Journal of biological chemistry 2003; 278: 19023-

19031.

123

46 Rolli-Derkinderen M, Toumaniantz G, Pacaud P, Loirand G. RhoA

phosphorylation induces Rac1 release from guanine dissociation inhibitor alpha

and stimulation of vascular smooth muscle cell migration. Mol Cell Biol 2010; 30:

4786-4796.

47 Rolli-Derkinderen M, Sauzeau V, Boyer L, Lemichez E, Baron C, Henrion D et

al. Phosphorylation of serine 188 protects RhoA from ubiquitin/proteasome-

mediated degradation in vascular smooth muscle cells. Circulation research 2005;

96: 1152-1160.

48 Chang F, Lemmon C, Lietha D, Eck M, Romer L. of

Rac1: A Role in Regulation of Cell Spreading. PLOS ONE 2011; 6: e28587.

49 Kwon T, Kwon DY, Chun J, Kim JH, Kang SS. Akt Protein Kinase Inhibits

Rac1-GTP Binding through Phosphorylation at Serine 71 of Rac1. Journal of

Biological Chemistry 2000; 275: 423-428.

50 Tong J, Li L, Ballermann B, Wang Z. Phosphorylation of Rac1 T108 by

Extracellular Signal-Regulated Kinase in Response to Epidermal Growth Factor:

a Novel Mechanism To Regulate Rac1 Function. Molecular and Cellular Biology

2013; 33: 4538.

124

51 Tu S, Wu WJ, Wang J, Cerione RA. Epidermal Growth Factor-dependent

Regulation of Cdc42 Is Mediated by the Src Tyrosine Kinase. Journal of

Biological Chemistry 2003; 278: 49293-49300.

52 Tillement V, Lajoie-Mazenc I, Casanova A, Froment C, Penary M, Tovar D et al.

Phosphorylation of RhoB by CK1 impedes actin stress fiber organization and

epidermal growth factor receptor stabilization. Experimental Cell Research 2008;

314: 2811-2821.

53 Lehman HL, Van Laere SJ, van Golen CM, Vermeulen PB, Dirix LY, van Golen

KL. Regulation of Inflammatory Breast Cancer Cell Invasion through Akt1/PKBα

Phosphorylation of RhoC GTPase. Molecular Cancer Research 2012; 10: 1306.

54 Gu Y, Chae H-D, Siefring JE, Jasti AC, Hildeman DA, Williams DA. RhoH

GTPase recruits and activates Zap70 required for T cell receptor signaling and

thymocyte development. Nature Immunology 2006; 7: 1182-1190.

55 Okada S, Yamada E, Saito T, Ohshima K, Hashimoto K, Yamada M et al. CDK5-

dependent Phosphorylation of the Rho Family GTPase TC10α Regulates Insulin-

stimulated GLUT4 Translocation. Journal of Biological Chemistry 2008; 283:

35455-35463.

125

56 Alan JK, Berzat AC, Dewar BJ, Graves LM, Cox AD. Regulation of the Rho

Family Small GTPase Wrch-1/RhoU by C-Terminal Tyrosine Phosphorylation

Requires Src. Molecular and Cellular Biology 2010; 30: 4324.

57 Wei J, Mialki RK, Dong S, Khoo A, Mallampalli RK, Zhao Y et al. A new

mechanism of RhoA ubiquitination and degradation: Roles of SCFFBXL19 E3

ligase and Erk2. Biochimica et Biophysica Acta (BBA) - Molecular Cell Research

2013; 1833: 2757-2764.

58 Wang H-R, Zhang Y, Ozdamar B, Ogunjimi AA, Alexandrova E, Thomsen GH et

al. Regulation of Cell Polarity and Protrusion Formation by Targeting RhoA for

Degradation. Science 2003; 302: 1775.

59 Chen Y, Yang Z, Meng M, Zhao Y, Dong N, Yan H et al. Mediates

Degradation of RhoA through Evolutionarily Conserved BTB Adaptors to

Control Actin Cytoskeleton Structure and Cell Movement. Molecular Cell 2009;

35: 841-855.

60 Torrino S, Visvikis O, Doye A, Boyer L, Stefani C, Munro P et al. The

E3 Ubiquitin-Ligase HACE1 Catalyzes the Ubiquitylation of Active Rac1.

Developmental Cell 2011; 21: 959-965.

126

61 Oberoi TK, Dogan T, Hocking JC, Scholz RP, Mooz J, Anderson CL et al. IAPs

regulate the plasticity of cell migration by directly targeting Rac1 for degradation.

The EMBO Journal 2012; 31: 14.

62 Zhao J, Mialki RK, Wei J, Coon TA, Zou C, Chen BB et al. SCF E3 ligase F-box

protein complex SCFFBXL19 regulates cell migration by mediating Rac1

ubiquitination and degradation. The FASEB Journal 2013; 27: 2611-2619.

63 Castillo-Lluva S, Tatham MH, Jones RC, Jaffray EG, Edmondson RD, Hay RT et

al. SUMOylation of the GTPase Rac1 is required for optimal cell migration.

Nature Cell Biology 2010; 12: 1078.

64 Jaffe AB, Hall A. RHO GTPASES: Biochemistry and Biology. Annual Review of

Cell and Developmental Biology 2005; 21: 247-269.

65 Hanna S, El-Sibai M. Signaling networks of Rho GTPases in cell motility.

Cellular Signalling 2013; 25: 1955-1961.

66 Ridley AJ, Hall A. The small GTP-binding protein rho regulates the assembly of

focal adhesions and actin stress fibers in response to growth factors. Cell 1992;

70: 389-399.

127

67 Ridley AJ, Paterson HF, Johnston CL, Diekmann D, Hall A. The small GTP-

binding protein rac regulates growth factor-induced membrane ruffling. Cell

1992; 70: 401-410.

68 Kozma R, Ahmed S, Best A, Lim L. The Ras-related protein Cdc42Hs and

bradykinin promote formation of peripheral actin microspikes and filopodia in

Swiss 3T3 fibroblasts. Molecular and cellular biology 1995; 15: 1942-1952.

69 Ridley AJ. Rho GTPases and actin dynamics in membrane protrusions and vesicle

trafficking. Trends in Cell Biology 2006; 16: 522-529.

70 Yoneda A, Multhaupt HAB, Couchman JR. The Rho kinases I and II regulate

different aspects of myosin II activity. The Journal of Cell Biology 2005; 170:

443.

71 Hill CS, Wynne J, Treisman R. The Rho family GTPases RhoA, Racl , and

CDC42Hsregulate transcriptional activation by SRF. Cell 1995; 81: 1159-1170.

72 Sotiropoulos A, Gineitis D, Copeland J, Treisman R. Signal-Regulated Activation

of Serum Response Factor Is Mediated by Changes in Actin Dynamics. Cell

1999; 98: 159-169.

128

73 Palazzo AF, Cook TA, Alberts AS, Gundersen GG. mDia mediates Rho-regulated

formation and orientation of stable microtubules. Nature Cell Biology 2001; 3:

723-729.

74 Cassimeris L. The oncoprotein 18/stathmin family of microtubule destabilizers.

Current Opinion in Cell Biology 2002; 14: 18-24.

75 Daub H, Gevaert K, Vandekerckhove J, Sobel A, Hall A. Rac/Cdc42 and

p65PAK Regulate the Microtubule-destabilizing Protein Stathmin through

Phosphorylation at Serine 16. Journal of Biological Chemistry 2001; 276: 1677-

1680.

76 Arimura N, Inagaki N, Chihara K, Ménager C, Nakamura N, Amano M et al.

Phosphorylation of Collapsin Response Mediator Protein-2 by Rho-kinase:

EVIDENCE FOR TWO SEPARATE SIGNALING PATHWAYS FOR

GROWTH CONE COLLAPSE. Journal of Biological Chemistry 2000; 275:

23973-23980.

77 Coso OA, Chiariello M, Yu J-C, Teramoto H, Crespo P, Xu N et al. The small

GTP-binding proteins Rac1 and Cdc42regulate the activity of the JNK/SAPK

signaling pathway. Cell 1995; 81: 1137-1146.

129

78 Whitmarsh AJ, Yang SH, Su MS, Sharrocks AD, Davis RJ. Role of p38 and JNK

mitogen-activated protein kinases in the activation of ternary complex factors.

Molecular and Cellular Biology 1997; 17: 2360.

79 Marinissen MJ, Chiariello M, Tanos T, Bernard O, Narumiya S, Gutkind JS. The

Small GTP-Binding Protein RhoA Regulates c-Jun by a ROCK-JNK Signaling

Axis. Molecular Cell 2004; 14: 29-41.

80 Perona R, Montaner S, Saniger L, Sánchez-Pérez I, Bravo R, Lacal JC. Activation

of the nuclear factor-kappaB by Rho, CDC42, and Rac-1 proteins. Genes &

Development 1997; 11: 463-475.

81 Sun S-C, Ley SC. New insights into NF-κB regulation and function. Trends in

Immunology 2008; 29: 469-478.

82 Cammarano MS, Minden A. Dbl and the Rho GTPases activate NF kappa B by I

kappa B kinase (IKK)-dependent and IKK-independent pathways. The Journal of

biological chemistry 2001; 276: 25876-25882.

83 Xie X, Peng J, Chang X, Huang K, Huang J, Wang S et al. Activation of

RhoA/ROCK regulates NF-κB signaling pathway in experimental diabetic

nephropathy. Molecular and Cellular Endocrinology 2013; 369: 86-97.

130

84 Aznar S, Valerón PF, del Rincon SV, Pérez LF, Perona R, Lacal JC.

Simultaneous Tyrosine and Serine Phosphorylation of STAT3 Transcription

Factor Is Involved in Rho A GTPase Oncogenic Transformation. Molecular

Biology of the Cell 2001; 12: 3282-3294.

85 Benitah SA, Valeron PF, Rui H, Lacal JC. STAT5a activation mediates the

epithelial to mesenchymal transition induced by oncogenic RhoA. Mol Biol Cell

2003; 14: 40-53.

86 Charron F, Tsimiklis G, Arcand M, Robitaille L, Liang Q, Molkentin JD et al.

Tissue-specific GATA factors are transcriptional effectors of the small GTPase

RhoA. Genes & development 2001; 15: 2702-2719.

87 Sordella R, Classon M, Hu KQ, Matheson SF, Brouns MR, Fine B et al.

Modulation of CREB activity by the Rho GTPase regulates cell and organism size

during mouse embryonic development. Dev Cell 2002; 2: 553-565.

88 Coleman ML, Marshall CJ, Olson MF. RAS and RHO GTPases in G1-phase cell-

cycle regulation. Nature Reviews Molecular Cell Biology 2004; 5: 355-366.

131

89 Westwick JK, Lambert QT, Clark GJ, Symons M, Van Aelst L, Pestell RG et al.

Rac regulation of transformation, gene expression, and actin organization by

multiple, PAK-independent pathways. Mol Cell Biol 1997; 17: 1324-1335.

90 Chou MM, Masuda-Robens JM, Gupta ML. Cdc42 promotes G1 progression

through p70 S6 kinase-mediated induction of cyclin E expression. The Journal of

biological chemistry 2003; 278: 35241-35247.

91 Olson MF, Paterson HF, Marshall CJ. Signals from Ras and Rho GTPases interact

to regulate expression of p21Waf1/Cip1. Nature 1998; 394: 295-299.

92 Weber JD, Hu W, Jefcoat SC, Jr., Raben DM, Baldassare JJ. Ras-stimulated

extracellular signal-related kinase 1 and RhoA activities coordinate platelet-

derived growth factor-induced G1 progression through the independent regulation

of cyclin D1 and p27. The Journal of biological chemistry 1997; 272: 32966-

32971.

93 Rosenblatt J, Cramer LP, Baum B, McGee KM. Myosin II-dependent cortical

movement is required for centrosome separation and positioning during mitotic

spindle assembly. Cell 2004; 117: 361-372.

132

94 Yasuda S, Oceguera-Yanez F, Kato T, Okamoto M, Yonemura S, Terada Y et al.

Cdc42 and mDia3 regulate microtubule attachment to kinetochores. Nature 2004;

428: 767-771.

95 Glotzer M. Animal cell cytokinesis. Annu Rev Cell Dev Biol 2001; 17: 351-386.

96 De Pascalis C, Etienne-Manneville S. Single and collective cell migration: the

mechanics of adhesions. Mol Biol Cell 2017; 28: 1833-1846.

97 Ridley AJ. Rho GTPase signalling in cell migration. Curr Opin Cell Biol 2015;

36: 103-112.

98 Lawson CD, Ridley AJ. Rho GTPase signaling complexes in cell migration and

invasion. The Journal of Cell Biology 2018; 217: 447.

99 Paluch EK, Raz E. The role and regulation of blebs in cell migration. Curr Opin

Cell Biol 2013; 25: 582-590.

100 Paterson EK, Courtneidge SA. Invadosomes are coming: new insights into

function and disease relevance. The FEBS journal 2018; 285: 8-27.

133

101 Green KJ, Getsios S, Troyanovsky S, Godsel LM. Intercellular Junction

Assembly, Dynamics, and Homeostasis. Cold Spring Harbor Perspectives in

Biology 2010; 2.

102 Terry S, Nie M, Matter K, Balda MS. Rho Signaling and Tight Junction

Functions. Physiology 2010; 25: 16-26.

103 Hulpiau P, van Roy F. Molecular evolution of the cadherin superfamily. The

international journal of biochemistry & cell biology 2009; 41: 349-369.

104 Harris TJC, Tepass U. Adherens junctions: from molecules to morphogenesis.

Nature Reviews Molecular Cell Biology (Review Article) 2010; 11: 502.

105 Niessen CM. Tight junctions/adherens junctions: basic structure and function. The

Journal of investigative dermatology 2007; 127: 2525-2532.

106 Braga VM, Machesky LM, Hall A, Hotchin NA. The small GTPases Rho and Rac

are required for the establishment of cadherin-dependent cell-cell contacts. The

Journal of cell biology 1997; 137: 1421-1431.

134

107 Malliri A, van Es S, Huveneers S, Collard JG. The Rac exchange factor Tiam1 is

required for the establishment and maintenance of cadherin-based adhesions. The

Journal of biological chemistry 2004; 279: 30092-30098.

108 Yamazaki D, Oikawa T, Takenawa T. Rac-WAVE-mediated actin reorganization

is required for organization and maintenance of cell-cell adhesion. Journal of cell

science 2007; 120: 86-100.

109 Wildenberg GA, Dohn MR, Carnahan RH, Davis MA, Lobdell NA, Settleman J

et al. p120-catenin and p190RhoGAP regulate cell-cell adhesion by coordinating

antagonism between Rac and Rho. Cell 2006; 127: 1027-1039.

110 Chen X, Macara IG. Par-3 controls tight junction assembly through the Rac

exchange factor Tiam1. Nat Cell Biol 2005; 7: 262-269.

111 Sahai E, Marshall CJ. ROCK and Dia have opposing effects on adherens

junctions downstream of Rho. Nat Cell Biol 2002; 4: 408-415.

112 Khosravi-Far R, Solski PA, Clark GJ, Kinch MS, Der CJ. Activation of Rac1,

RhoA, and mitogen-activated protein kinases is required for Ras transformation.

Molecular and Cellular Biology 1995; 15: 6443.

135

113 Perona R, Esteve P, Jimenez B, Ballestero RP, Ramon y Cajal S, Lacal JC.

Tumorigenic activity of rho genes from Aplysia californica. Oncogene 1993; 8:

1285-1292.

114 del Peso L, Hernandez-Alcoceba R, Embade N, Carnero A, Esteve P, Paje C et al.

Rho proteins induce metastatic properties in vivo. Oncogene 1997; 15: 3047-

3057.

115 Qiu RG, Chen J, McCormick F, Symons M. A role for Rho in Ras transformation.

Proceedings of the National Academy of Sciences 1995; 92: 11781.

116 Fritz G, Brachetti C, Bahlmann F, Schmidt M, Kaina B. Rho GTPases in human

breast tumours: expression and mutation analyses and correlation with clinical

parameters. British Journal of Cancer 2002; 87: 635-644.

117 Horiuchi A, Imai T, Wang C, Ohira S, Feng Y, Nikaido T et al. Up-Regulation of

Small GTPases, RhoA and RhoC, Is Associated with Tumor Progression in

Ovarian Carcinoma. Laboratory Investigation (Article) 2003; 83: 861.

118 Pan Y, Bi F, Liu N, Xue Y, Yao X, Zheng Y et al. Expression of seven main Rho

family members in gastric carcinoma. Biochemical and Biophysical Research

Communications 2004; 315: 686-691.

136

119 Kamai T, Tsujii T, Arai K, Takagi K, Asami H, Ito Y et al. Significant association

of Rho/ROCK pathway with invasion and metastasis of bladder cancer. Clinical

cancer research : an official journal of the American Association for Cancer

Research 2003; 9: 2632-2641.

120 Kamai T, Yamanishi T, Shirataki H, Takagi K, Asami H, Ito Y et al.

Overexpression of RhoA, Rac1, and Cdc42 GTPases Is Associated with

Progression in Testicular Cancer. Clinical Cancer Research 2004; 10: 4799.

121 Abraham MT, Kuriakose MA, Sacks PG, Yee H, Chiriboga L, Bearer EL et al.

Motility-related proteins as markers for head and neck squamous cell cancer. The

Laryngoscope 2001; 111: 1285-1289.

122 Lei R, Tang J, Zhuang X, Deng R, Li G, Yu J et al. Suppression of MIM by

microRNA-182 activates RhoA and promotes breast cancer metastasis. Oncogene

2014; 33: 1287-1296.

123 Parr C, Jiang WG. Metastasis suppressor 1 (MTSS1) demonstrates prognostic

value and anti-metastatic properties in breast cancer. European journal of cancer

(Oxford, England : 1990) 2009; 45: 1673-1683.

137

124 Pillé JY, Denoyelle C, Varet J, Bertrand JR, Soria J, Opolon P et al. Anti-RhoA

and Anti-RhoC siRNAs Inhibit the Proliferation and Invasiveness of MDA-MB-

231 Breast Cancer Cells in Vitro and in Vivo. Molecular Therapy 2005; 11: 267-

274.

125 Hirata D, Yamabuki T, Miki D, Ito T, Tsuchiya E, Fujita M et al. Involvement of

epithelial cell transforming sequence-2 oncoantigen in lung and esophageal

cancer progression. Clinical cancer research : an official journal of the American

Association for Cancer Research 2009; 15: 256-266.

126 Sano M, Genkai N, Yajima N, Tsuchiya N, Homma J, Tanaka R et al. Expression

level of ECT2 proto-oncogene correlates with prognosis in glioma patients.

Oncology reports 2006; 16: 1093-1098.

127 Kourlas PJ, Strout MP, Becknell B, Veronese ML, Croce CM, Theil KS et al.

Identification of a gene at 11q23 encoding a guanine nucleotide exchange factor:

evidence for its fusion with MLL in acute myeloid leukemia. Proc Natl Acad Sci

U S A 2000; 97: 2145-2150.

128 Sahay S, Pannucci NL, Mahon GM, Rodriguez PL, Megjugorac NJ, Kostenko EV

et al. The RhoGEF domain of p210 Bcr-Abl activates RhoA and is required for

transformation. Oncogene 2008; 27: 2064-2071.

138

129 Xue W, Krasnitz A, Lucito R, Sordella R, VanAelst L, Cordon-Cardo C et al.

DLC1 is a chromosome 8p tumor suppressor whose loss promotes hepatocellular

carcinoma. Genes & Development 2008; 22: 1439-1444.

130 Leung TH-Y, Ching Y-P, Yam JWP, Wong C-M, Yau T-O, Jin D-Y et al.

<em>Deleted in liver cancer 2</em> (<em>DLC2</em>)

suppresses cell transformation by means of inhibition of RhoA activity. Proc Natl

Acad Sci U S A 2005; 102: 15207.

131 Kakiuchi M, Nishizawa T, Ueda H, Gotoh K, Tanaka A, Hayashi A et al.

Recurrent gain-of-function mutations of RHOA in diffuse-type gastric carcinoma.

Nature 2014; 46: 583.

132 Wang K, Yuen ST, Xu J, Lee SP, Yan HHN, Shi ST et al. Whole-genome

sequencing and comprehensive molecular profiling identify new driver mutations

in gastric cancer. Nature Genetics (Article) 2014; 46: 573.

133 Sahai E, Alberts AS, Treisman R. RhoA effector mutants reveal distinct effector

pathways for cytoskeletal reorganization, SRF activation and transformation. The

EMBO journal 1998; 17: 1350-1361.

139

134 Maeda M, Ushijima T. RHOA mutation may be associated with diffuse-type

gastric cancer progression, but is it gain or loss? Gastric cancer : official journal

of the International Gastric Cancer Association and the Japanese Gastric Cancer

Association 2016; 19: 326-328.

135 Palomero T, Couronne L, Khiabanian H, Kim MY, Ambesi-Impiombato A,

Perez-Garcia A et al. Recurrent mutations in epigenetic regulators, RHOA and

FYN kinase in peripheral T cell lymphomas. Nature genetics 2014; 46: 166-170.

136 Yoo HY, Sung MK, Lee SH, Kim S, Lee H, Park S et al. A recurrent inactivating

mutation in RHOA GTPase in angioimmunoblastic T cell lymphoma. Nature

genetics 2014; 46: 371-375.

137 O'Hayre M, Inoue A, Kufareva I, Wang Z, Mikelis CM, Drummond RA et al.

Inactivating mutations in GNA13 and RHOA in Burkitt’s lymphoma and diffuse

large B-cell lymphoma: a tumor suppressor function for the Gα13/RhoA axis in B

cells. Oncogene (Original Article) 2015; 35: 3771.

138 Arango D, Laiho P, Kokko A, Alhopuro P, Sammalkorpi H, Salovaara R et al.

Gene-expression profiling predicts recurrence in Dukes' C colorectal cancer.

Gastroenterology 2005; 129: 874-884.

140

139 Rodrigues P, Macaya I, Bazzocco S, Mazzolini R, Andretta E, Dopeso H et al.

RHOA inactivation enhances Wnt signalling and promotes colorectal cancer.

Nature Communications (Article) 2014; 5: 5458.

140 Dopeso H, Rodrigues P, Bilic J, Bazzocco S, Carton-Garcia F, Macaya I et al.

Mechanisms of inactivation of the tumour suppressor gene RHOA in colorectal

cancer. British journal of cancer 2018; 118: 106-116.

141 Lawson CD, Fan C, Mitin N, Baker NM, George SD, Graham DM et al. Rho

GTPase Transcriptome Analysis Reveals Oncogenic Roles for Rho GTPase-

Activating Proteins in Basal-like Breast Cancers. Cancer Research 2016; 76:

3826.

142 Humphries B, Wang Z, Li Y, Jhan J-R, Jiang Y, Yang C. ARHGAP18

Downregulation by miR-200b Suppresses Metastasis of Triple-Negative Breast

Cancer by Enhancing Activation of RhoA. Cancer Research 2017; 77: 4051.

143 Yeung K, Seitz T, Li S, Janosch P, McFerran B, Kaiser C et al. Suppression of

Raf-1 kinase activity and MAP kinase signalling by RKIP. Nature 1999; 401:

173-177.

141

144 Yeung K, Janosch P, McFerran B, Rose DW, Mischak H, Sedivy JM et al.

Mechanism of suppression of the Raf/MEK/extracellular signal-regulated kinase

pathway by the inhibitor protein. Mol Cell Biol 2000; 20: 3079-3085.

145 Yeung KC, Rose DW, Dhillon AS, Yaros D, Gustafsson M, Chatterjee D et al.

Raf kinase inhibitor protein interacts with NF-kappaB-inducing kinase and TAK1

and inhibits NF-kappaB activation. Mol Cell Biol 2001; 21: 7207-7217.

146 Beshir AB, Ren G, Magpusao AN, Barone LM, Yeung KC, Fenteany G. Raf

kinase inhibitor protein suppresses nuclear factor-κB-dependent cancer cell

invasion through negative regulation of matrix metalloproteinase expression.

Cancer letters 2010; 299: 137-149.

147 Park S, Yeung ML, Beach S, Shields JM, Yeung KC. RKIP downregulates B-Raf

kinase activity in melanoma cancer cells. Oncogene 2005; 24: 3535-3540.

148 Chatterjee D, Bai Y, Wang Z, Beach S, Mott S, Roy R et al. RKIP sensitizes

prostate and breast cancer cells to drug-induced apoptosis. The Journal of

biological chemistry 2004; 279: 17515-17523.

142

149 Dangi-Garimella S, Yun J, Eves EM, Newman M, Erkeland SJ, Hammond SM et

al. Raf kinase inhibitory protein suppresses a metastasis signalling cascade

involving LIN28 and let-7. Embo j 2009; 28: 347-358.

150 Yun J, Frankenberger CA, Kuo WL, Boelens MC, Eves EM, Cheng N et al.

Signalling pathway for RKIP and Let-7 regulates and predicts metastatic breast

cancer. Embo j 2011; 30: 4500-4514.

151 Datar I, Feng J, Qiu X, Lewandowski J, Yeung M, Ren G et al. RKIP Inhibits

Local Breast Cancer Invasion by Antagonizing the Transcriptional Activation of

MMP13. PLoS One 2015; 10: e0134494.

152 Datar I, Qiu X, Ma HZ, Yeung M, Aras S, de la Serna I et al. RKIP regulates

CCL5 expression to inhibit breast cancer invasion and metastasis by controlling

macrophage infiltration. Oncotarget 2015; 6: 39050-39061.

153 Deiss K, Kisker C, Lohse MJ, Lorenz K. Raf kinase inhibitor protein (RKIP)

dimer formation controls its target switch from Raf1 to -coupled

receptor kinase (GRK) 2. The Journal of biological chemistry 2012; 287: 23407-

23417.

143

154 Lorenz K, Lohse MJ, Quitterer U. Protein kinase C switches the Raf kinase

inhibitor from Raf-1 to GRK-2. Nature 2003; 426: 574-579.

155 Yousuf S, Duan M, Moen EL, Cross-Knorr S, Brilliant K, Bonavida B et al. Raf

kinase inhibitor protein (RKIP) blocks signal transducer and activator of

transcription 3 (STAT3) activation in breast and prostate cancer. PLoS One 2014;

9: e92478.

156 Gu M, Liu Z, Lai R, Liu S, Lin W, Ouyang C et al. RKIP and TBK1 form a

positive feedback loop to promote type I interferon production in innate

immunity. Embo j 2016; 35: 2553-2565.

157 Al-Mulla F, Bitar MS, Al-Maghrebi M, Behbehani AI, Al-Ali W, Rath O et al.

Raf Kinase Inhibitor Protein RKIP Enhances Signaling by Glycogen Synthase

Kinase-3β. Cancer Research 2011; 71: 1334.

158 Ren G, Baritaki S, Marathe H, Feng J, Park S, Beach S et al. Polycomb protein

EZH2 regulates tumor invasion via the transcriptional repression of the metastasis

suppressor RKIP in breast and prostate cancer. Cancer Res 2012; 72: 3091-3104.

144

159 Beach S, Tang H, Park S, Dhillon AS, Keller ET, Kolch W et al. Snail is a

repressor of RKIP transcription in metastatic prostate cancer cells. Oncogene

2008; 27: 2243-2248.

160 Zhang H, Wu J, Keller JM, Yeung K, Keller ET, Fu Z. Transcriptional regulation

of RKIP expression by androgen in prostate cells. Cellular physiology and

biochemistry : international journal of experimental cellular physiology,

biochemistry, and pharmacology 2012; 30: 1340-1350.

161 Chaffer CL, San Juan BP, Lim E, Weinberg RA. EMT, cell plasticity and

metastasis. Cancer and Metastasis Reviews 2016; 35: 645-654.

162 Alizadeh AM, Shiri S, Farsinejad S. Metastasis review: from bench to bedside.

Tumour biology : the journal of the International Society for Oncodevelopmental

Biology and Medicine 2014; 35: 8483-8523.

163 Yan J, Yang Q, Huang Q. Metastasis suppressor genes. Histol Histopathol 2013;

28: 285-292.

164 Fu Z, Smith PC, Zhang L, Rubin MA, Dunn RL, Yao Z et al. Effects of raf kinase

inhibitor protein expression on suppression of prostate cancer metastasis. Journal

of the National Cancer Institute 2003; 95: 878-889.

145

165 Escara-Wilke J, Keller JM, Ignatoski KM, Dai J, Shelley G, Mizokami A et al.

Raf kinase inhibitor protein (RKIP) deficiency decreases latency of tumorigenesis

and increases metastasis in a murine genetic model of prostate cancer. The

Prostate 2015; 75: 292-302.

166 Li HZ, Gao Y, Zhao XL, Liu YX, Sun BC, Yang J et al. Effects of raf kinase

inhibitor protein expression on metastasis and progression of human breast

cancer. Molecular cancer research : MCR 2009; 7: 832-840.

167 Minoo P, Zlobec I, Baker K, Tornillo L, Terracciano L, Jass JR et al. Loss of raf-

1 kinase inhibitor protein expression is associated with tumor progression and

metastasis in colorectal cancer. American journal of clinical pathology 2007; 127:

820-827.

168 Zhang L, Fu Z, Binkley C, Giordano T, Burant CF, Logsdon CD et al. Raf kinase

inhibitory protein inhibits beta-cell proliferation. Surgery 2004; 136: 708-715.

169 Karamitopoulou E, Zlobec I, Gloor B, Kondi-Pafiti A, Lugli A, Perren A. Loss of

Raf-1 kinase inhibitor protein (RKIP) is strongly associated with high-grade

tumor budding and correlates with an aggressive phenotype in pancreatic ductal

adenocarcinoma (PDAC). Journal of translational medicine 2013; 11: 311.

146

170 Akaishi J, Onda M, Asaka S, Okamoto J, Miyamoto S, Nagahama M et al.

Growth-suppressive function of phosphatidylethanolamine-binding protein in

anaplastic thyroid cancer. Anticancer research 2006; 26: 4437-4442.

171 Lee HC, Tian B, Sedivy JM, Wands JR, Kim M. Loss of Raf kinase inhibitor

protein promotes cell proliferation and migration of human hepatoma cells.

Gastroenterology 2006; 131: 1208-1217.

172 Hill B, De Melo J, Yan J, Kapoor A, He L, Cutz JC et al. Common reduction of

the Raf kinase inhibitory protein in clear cell renal cell carcinoma. Oncotarget

2014; 5: 7406-7419.

173 Al-Mulla F, Hagan S, Behbehani AI, Bitar MS, George SS, Going JJ et al. Raf

kinase inhibitor protein expression in a survival analysis of colorectal cancer

patients. Journal of clinical oncology : official journal of the American Society of

Clinical Oncology 2006; 24: 5672-5679.

174 Fu Z, Kitagawa Y, Shen R, Shah R, Mehra R, Rhodes D et al. Metastasis

suppressor gene Raf kinase inhibitor protein (RKIP) is a novel prognostic marker

in prostate cancer. The Prostate 2006; 66: 248-256.

147

175 Siegel RL, Miller KD, Jemal A. Cancer statistics, 2019. CA: A Cancer Journal for

Clinicians 2019; 69: 7-34.

176 Sorlie T, Perou CM, Tibshirani R, Aas T, Geisler S, Johnsen H et al. Gene

expression patterns of breast carcinomas distinguish tumor subclasses with

clinical implications. Proc Natl Acad Sci U S A 2001; 98: 10869-10874.

177 Perou CM, Sorlie T, Eisen MB, van de Rijn M, Jeffrey SS, Rees CA et al.

Molecular portraits of human breast tumours. Nature 2000; 406: 747-752.

178 Dai X, Li T, Bai Z, Yang Y, Liu X, Zhan J et al. Breast cancer intrinsic subtype

classification, clinical use and future trends. American journal of cancer research

2015; 5: 2929-2943.

179 Voduc KD, Cheang MC, Tyldesley S, Gelmon K, Nielsen TO, Kennecke H.

Breast cancer subtypes and the risk of local and regional relapse. Journal of

clinical oncology : official journal of the American Society of Clinical Oncology

2010; 28: 1684-1691.

180 Polyak K. Heterogeneity in breast cancer. The Journal of Clinical Investigation

2011; 121: 3786-3788.

148

181 Prat A, Parker JS, Karginova O, Fan C, Livasy C, Herschkowitz JI et al.

Phenotypic and molecular characterization of the claudin-low intrinsic subtype of

breast cancer. Breast Cancer Research 2010; 12: R68.

182 Weigelt B, Peterse JL, van 't Veer LJ. Breast cancer metastasis: markers and

models. Nature reviews Cancer 2005; 5: 591-602.

183 Minn AJ, Gupta GP, Siegel PM, Bos PD, Shu W, Giri DD et al. Genes that

mediate breast cancer metastasis to lung. Nature 2005; 436: 518-524.

184 Kang Y, Siegel PM, Shu W, Drobnjak M, Kakonen SM, Cordon-Cardo C et al. A

multigenic program mediating breast cancer metastasis to bone. Cancer cell 2003;

3: 537-549.

185 Bos PD, Zhang XHF, Nadal C, Shu W, Gomis RR, Nguyen DX et al. Genes that

mediate breast cancer metastasis to the brain. Nature 2009; 459: 1005.

186 Mao Y, Keller ET, Garfield DH, Shen K, Wang J. Stromal cells in tumor

microenvironment and breast cancer. Cancer and Metastasis Reviews 2013; 32:

303-315.

149

187 Kalluri R. The biology and function of fibroblasts in cancer. Nature reviews

Cancer 2016; 16: 582-598.

188 Kojima Y, Acar A, Eaton EN, Mellody KT, Scheel C, Ben-Porath I et al.

Autocrine TGF-beta and stromal cell-derived factor-1 (SDF-1) signaling drives

the evolution of tumor-promoting mammary stromal myofibroblasts. Proc Natl

Acad Sci U S A 2010; 107: 20009-20014.

189 Orimo A, Gupta PB, Sgroi DC, Arenzana-Seisdedos F, Delaunay T, Naeem R et

al. Stromal Fibroblasts Present in Invasive Human Breast Carcinomas Promote

Tumor Growth and Angiogenesis through Elevated SDF-1/CXCL12 Secretion.

Cell 2005; 121: 335-348.

190 Dvorak KM, Pettee KM, Rubinic-Minotti K, Su R, Nestor-Kalinoski A,

Eisenmann KM. Carcinoma associated fibroblasts (CAFs) promote breast cancer

motility by suppressing mammalian Diaphanous-related formin-2 (mDia2). PLoS

One 2018; 13: e0195278.

191 Muller A, Homey B, Soto H, Ge N, Catron D, Buchanan ME et al. Involvement

of chemokine receptors in breast cancer metastasis. Nature 2001; 410: 50-56.

150

192 Liang Z, Yoon Y, Votaw J, Goodman MM, Williams L, Shim H. Silencing of

CXCR4 blocks breast cancer metastasis. Cancer Res 2005; 65: 967-971.

193 Spaeth EL, Dembinski JL, Sasser AK, Watson K, Klopp A, Hall B et al.

Mesenchymal stem cell transition to tumor-associated fibroblasts contributes to

fibrovascular network expansion and tumor progression. PLoS One 2009; 4:

e4992-e4992.

194 Karnoub AE, Dash AB, Vo AP, Sullivan A, Brooks MW, Bell GW et al.

Mesenchymal stem cells within tumour stroma promote breast cancer metastasis.

Nature (Article) 2007; 449: 557.

195 Lee E, Fertig EJ, Jin K, Sukumar S, Pandey NB, Popel AS. Breast cancer cells

condition lymphatic endothelial cells within pre-metastatic niches to promote

metastasis. Nature Communications (Article) 2014; 5: 4715.

196 Velasco-Velázquez M, Jiao X, De La Fuente M, Pestell TG, Ertel A, Lisanti MP

et al. CCR5 Antagonist Blocks Metastasis of Basal Breast Cancer Cells. Cancer

Research 2012; 72: 3839.

151

197 Läubli H, Spanaus K-S, Borsig L. Selectin-mediated activation of endothelial

cells induces expression of CCL5 and promotes metastasis through recruitment of

monocytes. Blood 2009; 114: 4583.

198 del Pulgar TG, Benitah SA, Valerón PF, Espina C, Lacal JC. Rho GTPase

expression in tumourigenesis: Evidence for a significant link. BioEssays 2005; 27:

602-613.

199 Kang Y. Analysis of Cancer Stem Cell Metastasis in Xenograft Animal Models.

In: Yu JS (ed). Cancer Stem Cells: Methods and Protocols. Humana Press:

Totowa, NJ, 2009, pp 7-19.

200 Orgaz J, Herraiz C, Sanz-Moreno V. Rho GTPases modulate malignant

transformation of tumor cells, vol. 5, 2014.

201 Ellenbroek SIJ, Collard JG. Rho GTPases: functions and association with cancer.

Clinical & Experimental Metastasis 2007; 24: 657-672.

202 Zandvakili I, Lin Y, C Morris J, Zheng Y. Rho GTPases: Anti- or Pro-neoplastic

Targets?, vol. 36, 2016.

152

203 Aslakson CJ, Miller FR. Selective Events in the Metastatic Process Defined by

Analysis of the Sequential Dissemination of Subpopulations of a Mouse

Mammary Tumor. Cancer Research 1992; 52: 1399.

204 Pulaski BA, Ostrand-Rosenberg S. Mouse 4T1 Breast Tumor Model. Current

Protocols in Immunology 2000; 39: 20.22.21-20.22.16.

205 Brown M, Assen FP, Leithner A, Abe J, Schachner H, Asfour G et al. Lymph

node blood vessels provide exit routes for metastatic tumor cell dissemination in

mice. Science 2018; 359: 1408.

206 Pereira ER, Kedrin D, Seano G, Gautier O, Meijer EFJ, Jones D et al. Lymph

node metastases can invade local blood vessels, exit the node, and colonize distant

organs in mice. Science 2018; 359: 1403.

207 Quail DF, Joyce JA. Microenvironmental regulation of tumor progression and

metastasis. Nature Medicine (Review Article) 2013; 19: 1423.

208 Kalluri R. The biology and function of fibroblasts in cancer. Nature Reviews

Cancer (Review Article) 2016; 16: 582.

153

209 Lin EY, Nguyen AV, Russell RG, Pollard JW. Colony-Stimulating Factor 1

Promotes Progression of Mammary Tumors to Malignancy. The Journal of

Experimental Medicine 2001; 193: 727.

210 Murdoch C, Giannoudis A, Lewis CE. Mechanisms regulating the recruitment of

macrophages into hypoxic areas of tumors and other ischemic tissues. Blood

2004; 104: 2224.

211 Leek RD, Landers RJ, Harris AL, Lewis CE. Necrosis correlates with high

vascular density and focal macrophage infiltration in invasive carcinoma of the

breast. British Journal of Cancer 1999; 79: 991-995.

212 Weitzenfeld P, Ben-Baruch A. The chemokine system, and its CCR5 and CXCR4

receptors, as potential targets for personalized therapy in cancer. Cancer Letters

2014; 352: 36-53.

213 Su X, Ye J, Hsueh EC, Zhang Y, Hoft DF, Peng G. Tumor Microenvironments

Direct the Recruitment and Expansion of Human Th17 Cells. The Journal of

Immunology 2010; 184: 1630.

154

214 Jin K, Pandey NB, Popel AS. Simultaneous blockade of IL-6 and CCL5 signaling

for synergistic inhibition of triple-negative breast cancer growth and metastasis.

Breast Cancer Research 2018; 20: 54.

215 Sahai E, Marshall CJ. ROCK and Dia have opposing effects on adherens

junctions downstream of Rho. Nature Cell Biology 2002; 4: 408-415.

216 Carramusa L, Ballestrem C, Zilberman Y, Bershadsky AD. Mammalian

diaphanous-related formin Dia1 controls the organization of E-cadherin-mediated

cell-cell junctions. Journal of Cell Science 2007; 120: 3870.

217 Liu S, Goldstein RH, Scepansky EM, Rosenblatt M. Inhibition of Rho-Associated

Kinase Signaling Prevents Breast Cancer Metastasis to Human Bone. Cancer

Research 2009; 69: 8742.

218 Gupton SL, Eisenmann K, Alberts AS, Waterman-Storer CM. mDia2 regulates

actin and focal adhesion dynamics and organization in the lamella for efficient

epithelial cell migration. Journal of Cell Science 2007; 120: 3475.

219 Nathanson SD, Krag D, Kuerer HM, Newman LA, Brown M, Kerjaschki D et al.

Breast cancer metastasis through the lympho-vascular system. Clinical &

Experimental Metastasis 2018; 35: 443-454.

155

220 Hodge RG, Ridley AJ. Regulating Rho GTPases and their regulators. Nature

Reviews Molecular Cell Biology (Review Article) 2016; 17: 496.

221 Cammarano MS, Minden A. Dbl and the Rho GTPases Activate NFκB by IκB

Kinase (IKK)-dependent and IKK-independent Pathways. Journal of Biological

Chemistry 2001; 276: 25876-25882.

222 Sordella R, Classon M, Hu K-Q, Matheson SF, Brouns MR, Fine B et al.

Modulation of CREB Activity by the Rho GTPase Regulates Cell and Organism

Size during Mouse Embryonic Development. Developmental Cell 2002; 2: 553-

565.

223 Benitah SA, Valerón PF, Rui H, Lacal JC. STAT5a Activation Mediates the

Epithelial to Mesenchymal Transition Induced by Oncogenic RhoA. Molecular

Biology of the Cell 2002; 14: 40-53.

224 Helbig G, Christopherson KW, Bhat-Nakshatri P, Kumar S, Kishimoto H, Miller

KD et al. NF-κ B Promotes Breast Cancer Cell Migration and Metastasis by

Inducing the Expression of the Chemokine Receptor CXCR4. Journal of

Biological Chemistry 2003; 278: 21631-21638.

156

225 Kukreja P, Abdel-Mageed AB, Mondal D, Liu K, Agrawal KC. Up-regulation of

CXCR4 Expression in PC-3 Cells by Stromal-Derived Factor-1α (CXCL12)

Increases Endothelial Adhesion and Transendothelial Migration: Role of

MEK/ERK Signaling Pathway–Dependent NF-κB Activation. Cancer Research

2005; 65: 9891.

226 Banerjee A, Pirrone V, Wigdahl B, Nonnemacher MR. Transcriptional regulation

of the chemokine co-receptor CCR5 by the cAMP/PKA/CREB pathway.

Biomedicine & Pharmacotherapy 2011; 65: 293-297.

227 Siitonen SM, Kononen JT, Helin HJ, Rantala IS, Holli KA, Isola JJ. Reduced E-

cadherin expression is associated with invasiveness and unfavorable prognosis in

breast cancer. American journal of clinical pathology 1996; 105: 394-402.

228 Singhai R, Patil VW, Jaiswal SR, Patil SD, Tayade MB, Patil AV. E-Cadherin as

a diagnostic biomarker in breast cancer. North American journal of medical

sciences 2011; 3: 227-233.

229 Hajra KM, Chen DY, Fearon ER. The SLUG zinc-finger protein represses E-

cadherin in breast cancer. Cancer Res 2002; 62: 1613-1618.

157

230 Batlle E, Sancho E, Franci C, Dominguez D, Monfar M, Baulida J et al. The

transcription factor snail is a repressor of E-cadherin gene expression in epithelial

tumour cells. Nat Cell Biol 2000; 2: 84-89.

231 Pan Y, Li J, Zhang Y, Wang N, Liang H, Liu Y et al. Slug-upregulated miR-221

promotes breast cancer progression through suppressing E-cadherin expression.

Scientific Reports (Article) 2016; 6: 25798.

232 Nagaoka K, Zhang H, Watanabe G, Taya K. Epithelial cell differentiation

regulated by MicroRNA-200a in mammary glands. PLoS One 2013; 8: e65127.

233 Christofori G, Semb H. The role of the cell-adhesion molecule E-cadherin as a

tumour-suppressor gene. Trends in Biochemical Sciences 1999; 24: 73-76.

234 Vleminckx K, Vakaet L, Mareel M, Fiers W, Van Roy F. Genetic manipulation of

E-cadherin expression by epithelial tumor cells reveals an invasion suppressor

role. Cell 1991; 66: 107-119.

235 Guilluy C, Dubash AD, García-Mata R. Analysis of RhoA and Rho GEF activity

in whole cells and the . Nature Protocols 2011; 6: 2050.

158

236 von Thun A, Preisinger C, Rath O, Schwarz JP, Ward C, Monsefi N et al.

Extracellular Signal-Regulated Kinase Regulates RhoA Activation and Tumor

Cell Plasticity by Inhibiting Guanine Exchange Factor H1 Activity. Molecular

and Cellular Biology 2013; 33: 4526.

237 Liu BP, Burridge K. Vav2 Activates Rac1, Cdc42, and RhoA Downstream from

Growth Factor Receptors but Not β1 Integrins. Molecular and Cellular Biology

2000; 20: 7160.

238 Sun M, Gomes S, Chen P, Frankenberger CA, Sankarasharma D, Chung CH et al.

RKIP and HMGA2 regulate breast tumor survival and metastasis through lysyl

oxidase and syndecan-2. Oncogene 2014; 33: 3528-3537.

239 Bajrami I, Marlow R, van de Ven M, Brough R, Pemberton HN, Frankum J et al.

E-Cadherin/ROS1 Inhibitor Synthetic Lethality in Breast Cancer. Cancer

discovery 2018; 8: 498-515.

159