PRL-3 IS AN ONCOGENIC DRIVER IN TRIPLE-NEGATIVE BREAST CANCERS AND
A MEDIATOR OF THE NOVEL ANTICANCER COMPOUND AMPI-109
by
HAMID H. GARI
B.S., Colorado State University, 2009
M.Phil., University of Cambridge, 2010
A thesis submitted to the
Faculty of the Graduate School of the
University of Colorado in partial fulfillment
of the requirements for the degree of
Doctor of Philosophy
Cancer Biology Program
2016 This thesis for the Doctor of Philosophy degree by
Hamid H. Gari
has been approved for the
Cancer Biology Program
by
James DeGregori, Chair
M. Scott Lucia
Heide Ford
Jennifer Diamond
Bolin Liu
James Lambert, Advisor
Date: 05/20/2016
ii Gari, Hamid H. (Ph.D., Cancer Biology)
PRL-3 is an Oncogenic Driver in Triple-Negative Breast Cancers and a Mediator of the
Novel Anticancer Compound AMPI-109
Thesis directed by Associate Research Professor James Lambert
ABSTRACT
Triple-negative breast cancers (TNBCs) are among the most aggressive and heterogeneous cancers characterized by a high propensity to invade, metastasize and relapse.
AMPI-109 is a novel anticancer compound that is selectively efficacious in inhibiting proliferation and in inducing apoptosis of multiple TNBC subtype cell lines as assessed by activation of pro-apoptotic caspases-3 and 7, PARP cleavage and nucleosomal DNA fragmentation. Because AMPI-109 has little to no effect on growth in the majority of non-
TNBC cell lines, we utilized AMPI-109 in a genome-wide shRNA screen to investigate the utility of AMPI-109 as a tool in helping to identify molecular alterations unique to TNBC.
Our screen identified the phosphatase, PRL-3, as a putative intracellular target of
AMPI-109 and an important driver of TNBC growth, migration and invasion. Through stable lentiviral knock downs and transfection with catalytically impaired PRL-3 in TNBC cells, we show that loss of PRL-3 expression, or functionality, leads to substantial growth inhibition.
Mechanistically, we discovered that PRL-3 expression in TNBCs is transcriptionally regulated by the oncogenic NF-ĸB pathway and that PRL-3 ablation elicits a TNF-R1 feedback loop that results in TNBC cell cycle arrest and senescence followed by caspase-8 mediated apoptosis.
Additionally, AMPI-109 treatment, downregulation of PRL-3 expression or impairment of PRL-3 activity reduced TNBC cell migration and invasion by inactivating Src
iii and ERK signaling and downregulating downstream RhoA and Rac1/2/3 GTPase protein levels. This coincided with altered filamentous actin structures necessary for cell migration and invasion. Conversely, overexpression of PRL-3 promoted TNBC cell migration and invasion and promoted upregulation of a key matrix metalloproteinase,
MMP-10, which resulted in increased adherence to, and degradation of, the major basement membrane substrate, laminin.
Histological evaluation of human breast cancers revealed PRL- 3 expression was significantly associated with the TNBC subtype and correlated positively with the presence of regional and distant metastases, as well as poor relapse free survival.
This study is proof-of-concept that AMPI-109, a viable pre-clinical investigational therapeutic candidate for TNBCs, can be used as a molecular tool to uncover unique drivers of disease progression, such as PRL-3, which we show promotes oncogenic phenotypes in
TNBC cells.
The form and content of this abstract are approved. I recommend its publication.
Approved: James Lambert
iv I dedicate this work to my family, my wife Janie and daughter Layan, who have supported
me fully throughout the journey. Without them, my Ph.D. would not have been possible.
In particular, I dedicate this work to my father who continues to lovingly challenge and
encourage me to humbly become more than I thought I was capable of.
v ACKNOWLEDGEMENTS
I would like to thank Dr. Mary Reyland of the Cancer Biology Program and Dr. Ann
Thor and the research faculty of the Department of Pathology under the direction of Dr.
Steve Anderson for welcoming me into the cancer research community and allowing me to pursue my doctorate degree.
I owe a deep gratitude to my friend and mentor, Dr. James Lambert, for advising me every step of the way and giving me unconditional freedom to explore this thesis project. I’m also immensely grateful to have been welcomed as a colleague and as a friend by Drs. Scott
Lucia and Steve Nordeen for further inspiring me to challenge the status quo in every facet of this research project and the thoughtful discussions on developing new scientific concepts.
A special thank you is deserved for the members of my thesis committee, and my
Chair Dr. James DeGregori, who graciously donated their time and intellect despite their immense pre-existing commitments, to help me mature this important body of research. This work was partially funded by a research grant from the Colorado Cancer League and gift funds from David Paradice of Paradice Investment Management, to whom we’re indebted.
Finally, I would like to thank the late King Abdullah bin Abdul-Aziz, the Ministry of
Higher Education and the Saudi Arabian Cultural Mission in the USA for their generosity in funding my graduate studies and encouraging the Fellows to push for advancements in science as Ambassadors for the Kingdom of Saudi Arabia.
vi TABLE OF CONTENTS
CHAPTER
I. INTRODUCTION ...... 1
Molecular heterogeneity of breast cancer ...... 1
Therapeutic approaches to combating TNBC ...... 2
AMPI-109: Development history and mechanism of action in TNBC ...... 4
AMPI-109 impairs TNBC cell proliferation and induces apoptosis ...... 5
Ability of AMPI-109 to block TNBC growth is independent of the VDR ...... 8
Functional genomic screen identifies PRL-3 as a modifier of AMPI-109 action ...... 9
PTPs as potential molecular targets in cancer...... 1 3
The PTP4A family of phosphatases ...... 13
The role of PRL-3 in cancer ...... 14
Expression ...... 14
Substrates and control of metastatic signaling pathways ...... 15
Transcriptional regulation ...... 16
NF-ĸB in cancer development and metastasis ...... 16
Hypothesis and aims ...... 18
II. AMPI-109 TREATMENT AND PRL-3 ABLATION INHIBIT TNBC GROWTH, MIGRATION AND INVASION ...... 19
Introduction ...... 19
Materials and methods ...... 20
Materials ...... 20
Cell culture, immunoblot analysis and transfection ...... 21
vii In silico docking analysis ...... 21
Cellular proliferation, migration and invasion assays...... 2 1
Lentivirus ...... 22
In vitro enzymatic assays ...... 22
Site-directed mutagenesis ...... 23
Apoptosis assays ...... 23
Production of wild type and C104A FLAG-tagged recombinant PRL-3 ...... 23
Immunohistochemistry ...... 24
Microarray and statistical analysis ...... 24
Results ...... 25
AMPI-109 binds PRL-3 active site in silico and impairs its catalytic activity in vitro ...... 25
PRL-3 knock down and expression of catalytically impaired PRL-3 inhibits TNBC cell growth and confers partial resistance to AMPI-109 ...... 29
PRL-3 modulation and AMPI-109 treatment inhibit TNBC cell migration and invasion ...... 32
PRL-3 expression positively associates with the TNBC subtype, regional and distant metastases ...... 34
Discussion ...... 37
III. PRL-3 PARTICIPATES IN A PRO-INFLAMMATORY FEEDBACK LOOP THAT MODULATES TNF-R1 PLEIOTROPISM TO CONTROL TNBC GROWTH ...... 40
Introduction ...... 40
Materials and methods ...... 41
Materials ...... 41
Plasmids, transfection and viral transduction ...... 41
Luciferase assays ...... 42
viii qRT-PCR ...... 42
Cell culture, senescence and immunoblot analysis ...... 43
Cellular size, count and proliferation assays ...... 43
Chromatin immunoprecipitation (ChIP) assay ...... 44
Cell cycle and apoptosis analysis ...... 45
Gene co-expression and statistical analysis ...... 45
Results ...... 45
Knock down of PRL-3 induces senescence in TNBC cell lines ...... 45
Senescent PRL-3 knock down cells arrest in G1 despite checkpoint aberrations ...... 47
The NF-ĸB transcription factor subunit, p65, binds the PRL-3 promoter and regulates PRL-3 expression ...... 49
PRL-3 knock down cells upregulate TNFα to sustain senescence through NF-ĸB activation ...... 52
Continued suppression of PRL-3 results in a dynamic TNFα-associated apoptosis feedback loop in senescent TNBC cell lines ...... 53
Blockade of the TNFα-associated TNF-R1 extrinsic death pathway confers significant resistance to PRL-3 knock down cell death ...... 55
Inhibiting NF-ĸB activation following PRL-3 knock down sustains TNBC cell survival through c-Jun/AP-1 ...... 55
Discussion ...... 59
IV. PRL-3 ENGAGES THE FOCAL ADHESION PATHWAY IN TNBC CELLS TO ALTER ACTIN STRUCTURE AND SUBSTRATE ADHESION PROPERTIES CRITICAL FOR CONTROLLING CELL MIGRATION AND INVASION ...... 63
Introduction ...... 63
Materials and methods ...... 64
Materials ...... 64
ix Plasmids, transfection and viral transduction ...... 65
Cell culture and immunoblot analysis ...... 65
Immunofluorescence analysis ...... 65
MMP array ...... 66
Cell adhesion and spreading assay ...... 66
Cell invasion assay ...... 67
Results ...... 67
PRL-3 amplification correlates with high Src levels in basal human breast cancers ...... 67
AMPI-109 treatment and knock down of PRL-3 reduces Src, ERK activation ...... 70
Knock down of PRL-3 or treatment with AMPI-109 decreases RhoA and Rac1/2/3 GTPase levels and associates with remodeling of actin networks in TNBC cells ...... 71
PRL-3 overexpression increases secretion of MMP-10 facilitating cell adhesion, spreading and degradation of laminin for promoting TNBC cell invasion ...... 73
Discussion ...... 80
V. FUTURE DIRECTIONS ...... 83
REFERENCES ...... 94
x LIST OF TABLES
Table
1.1 AMPI-109 inhibits the growth of multiple TNBC subtype cell lines ...... 6
2.1 PRL-3 mRNA expression correlates with multiple clinicopathological variables of aggressive breast cancer ...... 36
4.1 Protein expression changes in invasive human breast carcinoma samples exhibiting PRL-3 copy number amplification (>2 copies) ...... 68
xi LIST OF FIGURES
Figure
1.1 Structure of AMPI-109 ...... 5
1.2 AMPI-109 mediated growth inhibition and induction of apoptosis is specific for TNBC cell lines ...... 7
1.3 AMPI-109 action is independent of VDR ...... 8
1.4 Functional genetic screen with AMPI-109 identifies PRL-3 amplification in invasive basal breast cancers ...... 12
2.1 AMPI-109 binds PRL-3 in silico ...... 25
2.2 Protein sequence alignment between PRL-3 and the most structurally similar phosphatases to PRL-3 ...... 26
2.3 AMPI-109 impairs PRL-3 activity, but not that of alkaline phosphatase ...... 27
2.4 Changing Cys104 to alanine results in catalytically impaired PRL-3 ...... 28
2.5 PRL-3 protein level in BT-20 cells after treatment with cycloheximide (CHX, 100 nM) or AMPI-109 (100 nM) ...... 29
2.6 PRL-3 knock down results in reduced growth of TNBC cells ...... 31
2.7 Catalytically impaired PRL-3 reduces growth of TNBC cells and confers partial resistance to AMPI-109 efficacy ...... 30
2.8 PRL-3 modulation and AMPI-109 inhibit TNBC migration ...... 32
2.9 PRL-3 modulation and AMPI-109 inhibit TNBC invasion ...... 33
2.10 PRL-3 expression is increased in TNBC ...... 35
3.1 Knock down of PRL-3 leads to enlargement of TNBC cell size ...... 46
3.2 Knock down of PRL-3 induces senescence in TNBC cell lines ...... 47
3.3 Senescent PRL-3 knock down cells arrest in G1 despite unique molecular checkpoint aberrations ...... 48
3.4 Heatmap generated with GENE-E (Broad Institute), depicting differential gene expression data across multiple breast cancer cell lines ...... 49
xii 3.5 Expression of the NF-ĸB subunit, p65, alters PRL-3 expression ...... 50
3.6 p65 binds to the PRL-3 promoter...... 5 1
3.7 Continued suppression of PRL-3 expression results in a dynamic TNFα–associated apoptosis feedback loop in TNBC cell lines ...... 54
3.8 Blockade of the TNFα-mediated TNF-R1 extrinsic death pathway confers significant resistance to PRL-3 knock down cell death, while inhibition of NF-ĸB following PRL- 3 knock down sustains TNBC cell survival through c-Jun/AP-1 ...... 57
3.9 Comprehensive model of PRL-3 action on the TNF-R1 pathway ...... 58
4.1 AMPI-109 treatment and PRL-3 knock down inactivates Src, ERK signaling in TNBC cells ...... 70
4.2 Knock down of PRL-3 and treatment with AMPI-109 alters RhoGTPase expression and filamentous actin structure ...... 72
4.3 Overexpression of PRL-3 alters MMP levels in BT-20 TNBC cells ...... 74
4.4 BT-20 TNBC cells preferentially adhere to- and spread on- elastin and laminin basement membrane substrates ...... 75
4.5 BT-20 TNBC cells overexpressing PRL-3 preferentially adhere to the laminin basement membrane substrate ...... 76
4.6 SUM159 TNBC cells preferentially adhere to- and spread on- elastin, fibronectin and laminin basement membrane substrates...... 77
4.7 PRL-3 knock down in SUM159 TNBC cells abrogates cell adhesion and spreading on laminin ...... 78
4.8 PRL-3 driven cell invasion is mediated through MMP-10 ...... 79
4.9 Cartoon model depicting the role of PRL-3 in the metastatic cascade at the primary TNBC tumor site ...... 82
5.1 Putative C-terminally truncated TNK2 splice variant expression is unique to TNBCs and interacts with PRL-3 by Co-immunoprecipitation analysis ...... 89
5.2 AMPI-109 treatment is effective at reducing tumor volume and mass in a PDX model of TNBC...... 9 1
5.3 Comprehensive model on the role of PRL-3 and AMPI-109 on TNBC growth, migration and invasion ...... 93
xiii LIST OF ABBREVIATIONS
Abbreviation
1,25D - 1α,25-dihydroxyvitamin D3
AMPI-109 - 1α,25-dihydroxyvitamin D3-3-bromoacetate
AR – Androgen receptor
BL1 – Basal-like 1
BL2 – Basal-like 2 cBio – Memorial sloan-kettering cancer center cancer genomics portal
CDKI – Cyclin-dependent kinase inhibitor
ChIP – Chromatin immunoprecipitation
CHX – Cycloheximide
CoIP – Co-immunoprecipitation
CRC – Colorectal cancer
DCIS – Ductal carcinoma in situ
DUSP – Dual specificity phosphatase
ECM – Extracellular matrix
EGFR – Epidermal growth factor receptor
ELISA – Enzyme-linked immunosorbent assay
EMT – Epithelial-to-mesenchymal transition
ER – Estrogen receptor
F-actin – Filamentous actin
FA – Focal adhesion
FDA – Food and drug administration
xiv FFPE – Formalin fixed paraffin embedded
HER2 – Human epidermal growth factor receptor 2
IHC – Immunohistochemistry
IKK – IĸB kinase
IM – Immunomodulatory
IP – Immunoprecipitation
IĸB – Inhibitors of kappa B
IRB – Institutional review board
LAR – Luminal androgen receptor
LPS – Lipopolysaccheride
M – Mesenchymal
M0 – Visceral metastases-free disease
M1 – Metastatic disease mCRPC – Metastatic castration-resistant prostate cancer
MDC – Monodansyl cadaverine
MEF – Mouse embryonic fibroblast
MMP – Matrix metalloproteinase
MS – Mass spectrometry
MSL – Mesenchymal stem-like
MW – Molecular weight
NF-ĸB – Nuclear factor-kappa B
NMR – Nuclear magnetic resonance p-NPP – para-nitrophenyl phosphatate
xv PARP – Poly-(ADP ribose) polymerase
PR – Progesterone receptor
PRL – Phosphatase of regenerating liver
PTP – Protein tyrosine phosphatase
PTP4A – Protein tyrosine phosphatase type IVA qRT-PCR – Quantitative real-time polymerase chain reaction
Rb – Retinoblastoma protein
RPPA – Reverse phase protein array
Rpt – Restriction point
SA-β-Gal – Senescence-associated beta-galactosidase
SASP – Senescence-associated secretory phenotype
Src – Proto-oncogene tyrosine-protein kinase c-src
TCGA – The cancer genome atlas
TK – Tyrosine kinase
TKI – Tyrosine kinase inhibitor
TNBC – Triple-negative breast cancer
TNFα – Tumor necrosis factor alpha
VDR – Nuclear vitamin D receptor
β-Gal – Beta-galactosidase
xvi CHAPTER I
INTRODUCTION1
Molecular heterogeneity of breast cancer
Breast cancer is the most commonly diagnosed cancer and principle cause of cancer- related mortality in women worldwide (1). Due to advancements in high-throughput gene expression profiling, breast cancer has been clustered into five major subtypes based on estrogen receptor (ER) expression, progesterone receptor (PR) expression and human epidermal growth factor receptor 2 (HER2) amplification (2). Several anti-hormonal therapies are FDA-approved for breast cancer patients with tumors expressing ER or PR, while targeted therapy with the monoclonal antibodies trastuzumab and pertuzumab, are indicated for patients with tumors exhibiting HER2 amplification. This categorization system, based on hormone receptor and HER2 status and the subsequent coupling of anti- hormonal and HER2 targeted therapy, is one of the first examples in modern oncology for molecular subtyping and personalized treatment that has resulted in significant decreases in disease burden and overall mortality. Unfortunately, triple-negative breast cancers (TNBCs), which comprise 15-20% of all newly diagnosed cases of breast cancer, lack expression of these three molecular markers and are rapidly progressive; typically, they are diagnosed as high grade tumors and are already invasive by the time of diagnosis (3). Metastatic TNBC tumors also have a higher risk of distant recurrence and death compared to other breast cancers (4). Paradoxically, patients with non-metastatic TNBC treated with neoadjuvant
1 Parts of this chapter were reprinted with permission from Oncotarget, Gari, H, et al., Genome-wide functional genetic screen with the anticancer agent AMPI-109 identifies PRL-3 as an oncogenic driver in triple-negative breast cancers. Oncotarget, 2016. 1 chemotherapy have better response rates than other breast cancer subtypes (5), suggesting a relative driver of poor outcome in TNBC is the development of metastatic disease.
In an attempt to stratify the disease, and towards the potential to identify molecular- based therapies for TNBC, extensive transcriptomic analyses of large TNBC datasets have been carried out (6-8). Gene clustering analysis has now defined at least six molecular subtypes of TNBC, including: two basal-like (BL1 and BL2), a mesenchymal (M), a mesenchymal stem-like (MSL), an immunomodulatory (IM), and a luminal androgen receptor (LAR) subtype (7). The identification of these subtypes may aid in elucidation of molecular signaling pathways that can potentially be targeted in drug discovery efforts and will aid in further dissecting the molecular mechanisms behind the aggressive nature of
TNBCs.
Therapeutic approaches to combating TNBC
Because TNBCs lack expression of ER, PR and HER2 amplification, cytotoxic chemotherapies are most frequently utilized (9, 10). However, these treatments are limited, particularly in the unselected metastatic population, by poor long-term therapeutic responses, non-selective toxicities, and clonal progression of disease with the development of resistance.
Although some TNBCs harbor overexpression of the epidermal growth factor receptor
(EGFR), anti-EGFR therapy has shown limited efficacy in clinical trials (11-12). Poly-(ADP ribose) polymerase (PARP) inhibitors which induce synthetic lethality in BRCA1-deficient cells have also demonstrated variable responses relative to standard of care (13), despite
TNBCs proclivity to BRCA1 deficiency. The limited success of these therapies may be due, in part, to a lack of clinical trial stratification based on the extensive histological and molecular heterogeneity of TNBCs. Thanks to recent insights gleaned from the six or more distinct
2 subtypes comprising TNBC, however, several actionable targets have now been identified and new therapies are currently under clinical investigation. Targeting of the androgen receptor
(AR) for example, in the LAR subtype of TNBC, is currently being studied in phase II trials exploring the efficacy of enzalutamide or abiraterone acetate - both of which are FDA- approved regimens for the treatment of men with metastatic castration-resistant prostate cancer
(mCRPC). Promisingly, a doubling in median progression free survival was observed at an interim analysis in patients with AR+ TNBC, as determined by immunohistochemical (IHC) detection of AR (14). However, enzalutamide remains under clinical investigation for only a small subset of TNBC patients, and the comparative value of enzalutamide therapy relative to standard of care remains unknown in this single-arm trial. Moreover, lessons from mCRPC patients on enzalutamide suggest the eventual development of resistance to therapy by the emergence of constitutively active AR splice variants and/or expression of AR mutants that alter enzalutamide properties. Nonetheless, the development of AR inhibitors for AR+ TNBC represents a significant step forward in combating TNBC, and confirmatory clinical trial results
– including utilization of a gene signature panel to predict responders a priori - are awaited.
Trials are also on-going to assess inhibition of tyrosine kinases (TK). However, targeting TKs in TNBC has been difficult. Phase II trials investigating the use of the multi- targeted receptor tyrosine kinase inhibitors (TKIs), sunitinib and sorafenib, as first or second- line therapy, were stopped early due to limited efficacy (13). Furthermore, partial responses and kinome reprogramming are emerging as limiting factors to TKI efficacy across different cancers. Taken together, despite recent advances in molecular subtyping of TNBC, there remains a clear unmet medical need for tolerable, efficacious and personalized treatment options for patients with TNBC.
3 AMPI-109: Development history and mechanism of action in TNBC
Numerous epidemiological studies have demonstrated the importance of the hormone, vitamin D, dietary or otherwise, in preventing epithelial cancers (15). Additionally, the therapeutic potential of 1α,25-dihydroxyvitamin D3 (1,25D; calcitriol), the biologically active metabolite of vitamin D, is well-documented for its effects on expression of cell cycle regulators, the apoptosis machinery, angiogenesis and metastatic potential (16). However, high-dose administration for increased bioavailability and the inherent calcemic toxicity of vitamin D and some of its analogues, particularly during daily dosing regimens, has prevented its use generally as an anticancer agent. Thus, the development of more potent vitamin D analogs exhibiting strong antiproliferative activity, but also reduced systemic toxicity, has become an active area of research.
Our laboratory has developed novel analogs of vitamin D and its pre-hormonal form,
25-hydroxyvitamin D3, that specifically bind and label the ligand-binding pocket of the nuclear receptor for vitamin D (vitamin D receptor, VDR) (17-21). Previously, our laboratory reported on the synthesis of a novel 1,25D derivative compound called AMPI-109
(1α,25-dihydroxyvitamin D3-3-bromoacetate) (Figure 1.1) (22-25).
AMPI-109 was synthesized as a biochemical tool to probe the residues of VDR that contact the hormone (19). More recently we have investigated the antiproliferative and mechanistic effects of AMPI-109 in a variety of cancer types including prostate, pancreatic and kidney cancer (22-25). In these studies, AMPI-109 exerted significant antiproliferative and pro-apoptotic activities and was more potent than 1,25D in doing so.
4 OH
O OH B r O
Figure 1.1 Structure of AMPI-109.
AMPI-109 impairs TNBC cell proliferation and induces apoptosis
Considering the heterogeneity of breast cancer, we also examined the ability of
AMPI-109 to inhibit the proliferation of breast cancer cells of various subtypes including
TNBC. A cohort of 12 cell lines, including representation from non-malignant (1 line), luminal A (2 lines), luminal B (1 line), HER2 (2 lines) and TNBC tumors (6 lines), was treated with AMPI-109 at its approximate IC50 value of 100 nM or vehicle control.
Importantly, the IC50 of AMPI-109 was pre-determined by cellular proliferation assays in response to escalating doses of AMPI-109. The approximate concentration of AMPI-109
(each cell line was different but values clustered around 100 nM) that resulted in approximately 50% cell killing was considered the IC50. Surprisingly, of the 7 cell lines that showed significant response to AMPI-109, 6 were TNBC cell lines representing 5 different molecular subtypes of TNBC (Table 1.1). In these experiments we also compared AMPI-109 to its parent compound, 1,25D. AMPI-109 was far superior to 1,25D in inhibiting the growth of all cell lines tested (Table 1.1), and in particular, TNBCs which were unaffected by 1,25D treatment. Importantly, AMPI-109 also had little to no effect on the proliferation of non- tumorigenic breast (MCF10A) or the majority of non-TNBC cell lines (Figure 1.2A).
5 Table 1.1 AMPI-109 inhibits the growth of multiple TNBC subtype cell lines. The indicated cell lines were treated for 36 hours with AMPI-109 (100 nM), 1,25D (100 nM) or ethanol control and MTS assays performed. P values represent significance between 1,25D and AMPI- 109 treatment on growth inhibition. TN, triple-negative; MSL, mesenchymal stem-like; BL1, basal-like 1; Unclass., unclassified; M, mesenchymal, BL2, basal-like 2.
AMPI-109 1,25D: AMPI-109 Cell Line Subtype 1,25D % Control % Control p-value
MCF10A Nonmalignant 102 99 0.23 MCF7 Luminal A 95 93 0.27 T47D Luminal A 98 94 0.40 BT-474 Luminal B 87 66 <0.0001 MDA-453 HER2 94 86 0.26 SKBR3 HER2 100 99 0.89 MDA-MB-231 TN (MSL) 96 44 <0.0001 MDA-MB-468 TN (BL1) 86 39 <0.0001 BT-20 TN (Unclass.) 82 29 <0.0001 BT-549 TN (M) 94 69 <0.0001 HCC70 TN (BL2) 83 49 <0.0001 SUM149 TN (BL2) 92 39 <0.0001
To examine the ability of AMPI-109 to induce apoptosis in the TNBC and non-
TNBC cell lines, we performed real time kinetic imaging of caspase-3/7 activity. We observed a strong induction of apoptosis in the TNBC cell lines, but not in the non-TNBC cell lines in response to AMPI-109 (Figure 1.2B). We also performed ELISA assays to detect cytoplasmic nucleosomal DNA and immunoblot analysis for PARP cleavage. Interestingly, paclitaxel, a commonly used drug in breast cancer, induced apoptosis and PARP cleavage in all cells whereas AMPI-109 specifically induced apoptosis in TNBCs (Figure 1.2C and D).
In concordance with our proliferation data, these results indicate AMPI-109 preferentially impairs the survival of TNBC cells and is more efficacious than 1,25D in doing so.
6 A B
C D
MCF10A MCF-7 BT-20 MDA-MB-231
p = 0.001 MCF10A MCF-7 BT-20 MDA-MB-231 p = 0.0004 C T D A C T D A C T D A C T D A PARP Cleaved PARP p = 0.010 p = 0.22
Figure 1.2 AMPI-109 mediated growth inhibition and induction of apoptosis is specific for TNBC cell lines. A. The response of the indicated TNBC (BT-20, MDA-MB-231 and MDA- MB-468) and non-TNBC (MCF7, MCF10A and T47D) cell lines to ethanol vehicle control (Ctl; blue) or 100 nM AMPI-109 (AMPI-109; red) was determined by real time kinetic monitoring of cellular proliferation. B. The same cell lines were analyzed for induction of apoptosis by real time kinetic monitoring of caspase-3/7 activity after treatment with ethanol vehicle control (Ctl; blue) or 100 nM AMPI-109 (AMPI-109; red). For A and B. * = p-value <0.05 as determined by Student t test on last time-point. C. The indicated cell lines were treated with ethanol control (C; purple), paclitaxel (10-7 M, (T; green)), 1,25D (10-7 M, (D; red)), or AMPI-109 (10-7 M, (A; blue)) for 24 hrs. Cytoplasmic nucleosomal DNA was measured by ELISA to assess apoptosis. D. The same cells as in C. were treated for 24 hrs and PARP cleavage determined by immunoblot.
7 Ability of AMPI-109 to block TNBC growth is independent of VDR
1,25D elicits its antiproliferative effects on cancer cells through binding to the ligand-activated transcription factor, VDR. Because AMPI-109 is an analog of 1,25D, a key mechanistic question is whether it, like 1,25D, exerts its antitumorigenic properties through binding to VDR. We used shRNAs directed against VDR in three TNBC cell lines and examined their response to AMPI-109. AMPI-109 was equally effective at inhibiting the growth of control cells and cells with VDR knock down demonstrating that AMPI-109 exerts its antiproliferative effects through a non-VDR dependent mechanism (Figure 1.3).
BT-20 MDA-MB-231 MDA-MB-468 pLKO pLKO pLKO VDR3 VDR8 VDR3 VDR8 VDR3 VDR8
VDR VDR VDR Actin Actin Actin
Figure 1.3 AMPI-109 action is independent of VDR. Top. Immunoblot of VDR level in BT-20, MDA-MB-231 and MDA-MB-468 cells with VDR shRNAs (VDR3 and VDR8). pLKO is non-silencing control. Bottom. Graphs depict response of cells to ethanol control (Ctl; blue) or 100 nM AMPI-109 (red) as determined by real time kinetic monitoring of cellular proliferation. * = p-value <0.05 as determined by Student t test on last time-point.
8 We emphasize that all of our data examining the role of VDR in AMPI-109 action, and in comparing AMPI-109 to 1,25D, discussed above, indicate that AMPI-109 is acting through a novel mechanism in TNBC cells, distinct from 1,25D and VDR.
Although a formal possibility exists that AMPI-109 targets a convergence of pathways downstream of the VDR, our data with VDR shRNAs strongly suggest that this will not be the case. Several lines of experimental evidence generated in our laboratory demonstrate that AMPI-109 exerts its actions independent of VDR signaling pathways.
Indeed, the lack of VDR requirement was our initial rationale for performing a genome-wide screen to determine the molecular mechanisms of AMPI-109 action in TNBC cells (discussed in the next section). As shown in Figure 1.3, we show that VDR is not required for AMPI-
109 cell killing in three different TNBC cell lines. We have also confirmed this result in two additional breast cancer cell lines (data not shown).
Furthermore, VDR or its downstream targets genes were not identified in our genome- wide screen indicating that VDR signaling is not involved in the cell killing effects of AMPI- 109. Although AMPI-109 was developed as an analog of vitamin D we emphasize that its ability to promote apoptosis of TNBC cells is unique. Indeed, we propose that this difference between AMPI-109 and the parent compound defines the novelty of AMPI-109.
Functional genomic screen identifies PRL-3 as a modifier of AMPI-109 action
To identify non-VDR proteins involved in the cellular response to AMPI-109, and towards identifying signaling patterns unique to TNBC, we performed a genome-wide, functional genetic screen in the TNBC cell line BT-20 (Figure 1.4A). BT-20 cells were chosen for the screen because they were the most sensitive cell line in response to AMPI-109
9 (Table 1.1). Furthermore, we conducted the screen using a dose of AMPI-109 that was chosen to be significantly higher than the approximate IC50 dose in order to maximize the stringency of the screen and reduce the potential for false positives. We detected 13,303 unique shRNAs representing 8,628 genes after filtering reads with low representation
(neither condition having at least 3 replicates with count values). The relative representation of the shRNAs were then compared, with the expectation that shRNAs under-represented after AMPI-109 treatment correspond to genes that confer chemosensitivity when inhibited
(synthetic lethal), while shRNAs over-represented correspond to genes that confer chemoresistance. To identify “hits” from the screen, we applied two statistical analysis strategies to identify candidate genes for subsequent analysis and validation: 1) a negative binomial model with correction for multiple comparisons (26) and 2) a model requiring redundancy of shRNAs targeting a gene with a minimum 3-fold change in relative representation (27). We surmised that genes common between the two analyses would be enriched for true positives and concentrated subsequent efforts on this overlapping list.
Unsupervised clustering by Spearman’s rank correlation with complete linkage demonstrated that replicates clustered based on treatment conditions (Figure 1.4B), indicating that enrichment or depletion of shRNAs was not stochastic, but due to AMPI-109 treatment. The analyses implicated 2,084 potential genes with 201 identified by both analysis methods
(Figure 1.4B).
We then queried the overlapping gene set obtained from the shRNA screen against the Memorial Sloan-Kettering Cancer Center cBio Cancer Genomics Portal (cBio) to determine the clinical significance of the genes identified. The highest ranking hit from our screen, PTP4A3 (encoding PRL-3) (Figure 1.4C), represents a protein tyrosine phosphatase
10 (PTP) that is amplified or up regulated in approximately 8-16% of all invasive breast cancers between two Cancer Genome Atlas (TCGA) datasets (8, 28) (Figure 1.4D). Amplification or up regulation of PRL-3 in invasive basal breast cancers, however, which includes TNBCs, ranged from 19-31% of cases based on the cohort examined (Figure 1.4D). Though triple- negative and basal breast cancers are not equivalent, there is considerable overlap. Up to 55% of basal-like breast cancers are triple-negative, and up to 65% of TNBCs are basal-like (8).
Significantly, other top ranking hits from our screen, such as CDK10, FGFBP1 and EGFR, have also been implicated as being overexpressed or pro-tumorigenic in the context of breast cancer, lending robust utility for our genome-wide screen with AMPI-109 in identifying aggressive, cancer-related markers.
Collectively, these data suggested to us that PRL-3 expression levels may be higher in the breast cancer subtypes where AMPI-109 shows growth inhibitory activity, and we therefore focused our studies on understanding the role of PRL-3 expression and activity on TNBC oncogenesis.
11 genes that modulate the effects of AMPI-109.effects the 201 of modulate genesidentifie genes that were tumors) and thosewithandinvasive basalascancer determined breast tumors) acrosslevel cohorts depicting of amplification TCGAPRL-3 graph two enrichednormalihighesthitstreatmentgene by by ranking AMPI-109 highlighting treateduntreated. andin representation tag ofchemoresistancemediators analyses chemos listsandof generating sequencing.deepquantifiedana by andwere Data recovered sequences were nM)and with orAMPI-109 left treatedshRNA (750 genome-wide library A. 1.4 Figure A C Schematic of the experimental work flow of the functional genomicexperimentalofthe Schematic workfunctional flow of the Functional genetic screen with AMPI-109Functionalscreen geneticamplificwith PRL-3 identifies Right panel : Twoimplicated analysisRFC) : 2,084 andstrategies (NB D B
(synthetic lethal) (synthetic AMPI-109 Untreated Depleted shRNA heatmap by the PAM50 signaturebyPAM50 the (PAM50 Basal). ensitivity to AMPI-109. to ensitivity d by combinedby analysisd strategies. screen. BT-20 cells were transducedBT-20cellswerescreen.a with for patients with invasivebreast patientswith for(all cancer (chemoresistant) untreated.5After passages shRNAthe tag Enriched lyzed by complementaryby lyzed computational ation ininvasive ation basalcancers.breast zed foldchangezed(nFC) B. HeatmapshRNA of Enriched in Treatment Depleted in Treatment in Depleted 275 NB 335 NB C. . 49 152 D. Table
Bar RFC 314 RFC 959
12 PTPs as potential molecular targets in cancer
Phosphatases are critical regulators of cell signaling. In coordination with kinases, phosphatases determine the appropriate response to physiological stimuli by modulating the phosphorylation of proteins involved in the signal cascade. Perhaps the most unfortunate generalization regarding phosphatases is that they serve as housekeeping enzymes to shut-off signaling pathways (29). In fact, accumulating evidence suggests an overall increase in PTP activity in a number of cancers compared to their normal tissue counterparts, indicating a much more complex role (30).
The majority of phosphatases are grouped as class I phosphatases, which can dephosphorylate tyrosine residues (classical) or dephosphorylate serine, threonine and tyrosine residues (dual specificity phosphatases; DUSPs) (30). Interestingly, overexpression of both classical phosphatases and DUSPs have been reported to execute oncogenic functions in human cancer cells. Proliferation, survival, apoptosis resistance, vesicular trafficking, loss of adhesion, migration and invasion have all been reportedly altered by aberrant functions of
PTPs during tumor development. In breast cancer, UBASH3B and Eya2 are examples of PTPs that have been shown to drive oncogenic events such as tumor growth, migration, invasion and metastasis (31, 32).
The PTP4A family of phosphatases
The PTP type IVA (PTP4A) family of phosphatases are considered DUSPs and were discovered as immediate-early genes activated following a partial hepatectomy (33), hence,
PTP4A phosphatases are also known as the phosphatases of regenerating liver (PRLs). There are three phosphatases in this family that share a high level of amino acid sequence homology;
13 PTP4A1 (PRL-1), PTP4A2 (PRL-2), and PTP4A3 (PRL-3). All three contain a COOH- terminal CAAX motif for post-translational farnesylation, suggesting an ability to localize with the plasma membrane or early endosome. Furthermore, all possess a conserved, cysteine-based (Cys104) active site for phosphate catalysis. In contrast to other phosphatases, the PRLs share a unique, shallow hydrophobic active site pocket that could be advantageous in the context of drug binding and drug selectivity.
In normal tissue, PRLs are predominately expressed in skeletal muscle, though PRL-1 is also expressed in the brain and PRL-3 is found at moderate levels in the heart (34).
Interestingly, all PRLs have been found to be overexpressed in cancer cells, but not in normal cells. PRL-1 overexpression has been found in melanoma (35) and pancreatic cancer (36),
PRL-2 in prostate cancer cell-lines and tissues (37), and PRL-3 in metastatic colorectal cancer
(CRC) (38, 39), gastric cancer as well as invasive breast vasculature (40). This diversity of associations suggests that despite their homology, the PRLs may execute tissue-specific tumorigenic processes.
The role of PRL-3 in cancer
Expression
PRL-3 is overexpressed in a number of cancers including CRC, gastric and breast cancer (41-43). In breast cancer, studies have retrospectively associated PRL-3 expression with poor survival in TNBC but not in any other subtypes (44-45), and no mechanistic evidence has supported a role for PRL-3 in driving aggressive breast cancer phenotypes.
Overexpression of PRL-3 is potentially significant because PRL-3 levels strongly correlate with presence of metastatic CRC. Gene expression profiling in Bert Vogelstein’s lab of
14 metastatic CRC samples revealed that among 144 up-regulated genes, PRL-3 was the only gene consistently overexpressed in all 18 of the cancer metastases and invasive primary tumors, but not in normal tissue (38).
Substrates and control of metastatic signaling pathways
In light of these phenotypic associations in CRC, PRL-3 has been mechanistically implicated in regulating Src function and integrin signaling involved with cell motility and invasion. However, only three substrates have been identified that could be attributed to these signaling outcomes – ezrin (46), and integrins α1 (47) and β1 (48, 49).
The biological role of ezrin in the cascade of tumor growth and metastasis is not yet understood, but it is thought that ezrin resides at the nexus of several pathways which regulate metastatic potential. One hypothesis suggests that ezrin bridges cell surface receptor signals with the actin cytoskeleton in order to execute cell restructuring, though this has not been demonstrated fully in cancer, especially with respect to PRL-3.
Integrins constitute a large family of heterodimeric cell-surface receptors that transduce signals which stimulate a variety of events, including regulation of cell adhesion, motility and invasion. Significantly, PRL-3’s interaction with both integrin α1 and β1 has been shown to enhance migration and invasion in CRC cells (48, 49). In addition, PRL-3 has been shown to downregulate the Src inhibitor, CSK, leading to activation of Src and ERK signaling,
STAT3, and p130Cas. PRL-3 has also been shown to downregulate PTEN and activate PI3K to promote epithelial-mesenchymal transition (EMT), further contributing to metastasis (50).
These data in CRC suggest that PRL-3 regulates, through complex mechanisms, the metastatic potential of cancer cells.
15 Transcriptional regulation
Much less is known about what regulates PRL-3 expression. Although PRL-3 has been reported to have p53 binding elements in its first intron, there has been discordance in literature with respect to the role p53 plays since its activity is frequently disrupted during the progression of many cancer types (51). Computational prediction tools have identified putative promoter binding sites for n-Myc, STAT3, and nuclear factor-kappa B (NF-ĸB), although the characterization and functional relevance of these factors on PRL-3 expression have not been determined. The potential for transcriptional regulation of PRL-3 by NF-ĸB, however, is of interest, because constitutive activation of the NF-ĸB pathway has been shown to promote oncogenesis in ER-negative and hormone-independent tumors, such as TNBC (52, 53).
NF-ĸB in cancer development and metastasis
NF-ĸB is a widely expressed transcription factor that regulates the expression of numerous genes involved in inflammatory responses, cell growth, differentiation and apoptosis
(54, 55). NF-ĸB was first identified as a factor that bound to immunoglobulin kappa light chain enhancers in B cells (56). Subsequently, NF-ĸB was found to be expressed in a variety of cell types and functions as a transcription factor regulating the expression of a diverse set of genes.
NF-ĸB proteins comprise a family of structurally-related proteins that share homology with the retroviral oncoprotein v-Rel (57). NF-ĸB is a dimer of members of the Rel family of proteins each containing a highly conserved amino-terminal DNA-binding/dimerization domain called the Rel homology domain. This region also contains a nuclear localization signal critical for NF-ĸB activity. The NF-ĸB /Rel family includes NF-ĸB1 (p50/p105), NF-
ĸB2 (p52/p100), RelA (p65), RelB and cRel (57). Due to the presence of a strong
16 transcriptional activation domain, p65 is responsible for most of NF-ĸB transcriptional activity with the most prevalent activated form of NF-ĸB being a heterodimer consisting of p50 and p65 (58). A number of exogenous and endogenous stimuli, including tumor necrosis factor α
(TNFα), have been shown to induce NF-ĸB activity (59).
In most cells, NF-ĸB exists in the cytoplasm in an inactive form associated with regulatory proteins termed inhibitors of ĸB (IĸB). In the absence of stimulation, the interaction between NF-ĸB and IĸBs masks the nuclear localization signal in NF-ĸB and prevents its nuclear localization. The majority of signals that lead to the activation of NF-ĸB result in activation of a high molecular weight complex containing a serine-specific IĸB kinase (IKK).
Phosphorylation of IĸB by IKK targets IĸBs for ubiquitin-dependent degradation by the proteasome (60). Degradation of IĸBs results in unmasking of the nuclear localization signal in NF-ĸB, translocation of NF-ĸB to the nucleus and activation of NF-ĸB target genes.
In addition to the key role NF-ĸB plays in normal homeostatic function, a role for NF-
ĸB in the initiation and progression of cancer may be due to the target genes of NF-ĸB. Along with the numerous proinflammatory genes, NF-ĸB target genes also include a number of genes involved in many aspects of tumorigenicity including cell growth and proliferation, anti- apoptosis, angiogenesis, cell adhesion and metastasis (61).
Since multiple consensus NF-ĸB elements exist in the PRL-3 promoter, we sought to understand whether a direct association could exist between oncogenic NF-ĸB signaling and
PRL-3 expression. We reasoned that studying this regulatory mechanism may help us identify pathways critical for PRL-3 expression, and pathways that may be reciprocally affected by changes in PRL-3 expression as part of a feedback loop, often characteristic of NF-ĸB regulated genes, to control TNBC growth outcomes.
17 Hypothesis and aims
Based on our preliminary data, I hypothesized that AMPI-109 may act, in part, through direct inhibition of PRL-3 to block aggressive TNBC attributes. The goals of my work are to characterize a potential interaction between AMPI-109 and PRL-3, and to uncover the molecular mechanism(s) by which PRL-3 alters TNBC growth, migration and invasion. Through the following aims, I proposed that we would uncover the pathways and/or molecular mechanisms behind PRL-3 mediated control of TNBC growth, migration and invasion:
• Aim 1: Determine a relationship between AMPI-109 and PRL-3 and identify how changes in PRL-3 expression phenocopy or synergize with AMPI-109 in controlling TNBC cell growth.
• A: Identify whether AMPI-109 can bind to PRL-3 in silico and inhibit its activity in vitro.
• A: Determine whether NF-ĸB binds the PRL-3 promoter region and if it is sufficient for transcriptional expression of PRL-3.
• B: Identify if PRL-3 knockdown exerts a reciprocal effect on the NF-ĸB pathway to control TNBC growth potential.
• Aim 2: Identify how PRL-3 mechanistically promotes the migratory and invasive capacity of TNBC cells.
• A: Determine if PRL-3 alters activation and/or expression of key players in the focal adhesion pathway as a mechanism for promoting TNBC cell migration.
• B: Determine if and how PRL-3 expression enhances TNBC cell invasive potential.
• Aim 3: Explore the potential therapeutic value of targeting PRL-3 in vivo.
• A: Interrogate archival breast tissue for PRL-3 expression using IHC and correlate expression levels with multiple clinicopathological variables.
• B: Determine effects of PRL-3 expression on TNBC xenograft growth and metastases, and assess PRL-3 protein levels as a biomarker of response to AMPI-109.
18 CHAPTER II
AMPI-109 TREATMENT AND PRL-3 ABLATION INHIBIT TNBC GROWTH,
MIGRATION AND INVASION2
Introduction
As mentioned previously, breast cancer is a heterogeneous disease exhibiting diverse biological characteristics and clinical responses. By definition, TNBCs fail to express three receptors shown to promote many breast cancers: ER, PR, and HER2. For many patients with breast cancer targeting of these molecules significantly improves outcome (62). Because
TNBCs lack expression of ER, PR and HER2 amplification, cytotoxic chemotherapies are most frequently utilized (9, 10). However, these treatments are limited, particularly in the unselected metastatic population, by poor long-term therapeutic response, non-selective toxicities, and clonal progression of disease with the development of resistance. On-going studies are currently underway in order to identify driver alterations in TNBC.
Our laboratory previously reported on the synthesis and antitumorigenic properties of a novel compound, AMPI-109 (Figure 1.1) (21-25). In the introductory chapter, I demonstrated that AMPI-109 potently induced apoptosis in TNBC cells of various molecular subtypes, yet had little to no effect on the majority of non-TNBC and non-tumorigenic cell lines examined. To gain a mechanistic understanding of AMPI-109 specific action in TNBC cells and towards identifying molecular alterations unique to TNBC, I also reviewed the design and results of a genome-wide functional shRNA screen in the TNBC cell line, BT-20, to identify genes that, when silenced, conferred resistance to AMPI-109. The highest
2 Parts of this chapter were reprinted with permission from Oncotarget, Gari, H, et al., Genome-wide functional genetic screen with the anticancer agent AMPI-109 identifies PRL-3 as an oncogenic driver in triple-negative breast cancers. Oncotarget, 2016. 19 ranking hit in our screen was the previously reported metastatis-promoting phosphatase,
PRL-3.
In this chapter, I establish a preliminary relationship between AMPI-109 and PRL-3 and investigate the phenotypic outcomes of modulating PRL-3 expression and activity on
TNBC growth, migration and invasion. We demonstrate that PRL-3 may indeed, be one direct target of AMPI-109 and show that reduction of PRL-3 expression or impairment of
PRL-3 catalytic activity leads to substantial growth inhibition and a reduction in the migratory and invasive ability of TNBC cells, partially phenocopying the effects of AMPI-
109. We also sought to determine whether the observed effects of PRL-3 on in vitro cell migration and invasion could have clinicopathological correlates in vivo. In a retrospective study, we show PRL-3 is more highly expressed in TNBC relative to other subtypes, and that
PRL-3 expression associates with the presence of regional disease and distant metastases.
Because the vast majority of TNBC deaths result as a consequence of metastatic disease to visceral organs, new therapies targeting the PRL-3 signaling axis could have a significant impact in stopping the progression of TNBC to metastatic disease or slowing the dissemination of already metastatic tumor cells. This chapter explores this theoretical framework.
Materials and methods
Materials
AMPI-109 was synthesized as previously described (17-19). PRL-3 cDNA expression vector was purchased from Origene (Cat. # SC308739). Purified, active recombinant human
PRL-3 protein was purchased from Sigma (Cat. # SRP0210, reported activity of hydrolyzing para-nitrophenyl phosphate (p-NPP) was >0.53 pmol/ min/ug).
20 Cell culture, immunoblot analysis and transfection
Cell lines were obtained from the University of Colorado Cancer Center Tissue
Culture Shared Resource. All lines were authenticated by short tandem repeat DNA profiling by the University of Colorado Cancer Center DNA Sequencing and Analysis Core. Western blot analysis was conducted as previously described (21). PRL-3 (Cat. # ab82568, Abcam),
β-actin (Cat. # A5441, Sigma Aldrich). Transfections were carried out using Mirus TransIT
LT1 transfection reagent according to manufacturer’s instructions (Mirus Bio).
In silico docking analysis
A 3-D model of AMPI-109 and its SMILES code were created using ChemDraw
(PerkinElmer) and converted to a .pdbqt file using OpenBabel. The NMR structure of PRL-3
(ID: 1V3A) was obtained from the RCSB Protein Data Bank and prepared for calculations using Autodock tools. AMPI-109 docking to PRL-3 was performed using Autodock Vina
(63). The Gibbs free energy of the drug-protein interaction was obtained using the Autodock
Vina algorithm. Calculations repeated 20 times using a different random number seed for each calculation, yielding 180 poses. In each calculation the highest scoring pose adopted the same conformation (pose 1). Results were visualized using PyMOL (Schrodinger, LLC).
Cellular proliferation, migration and invasion assays
Cell proliferation was assessed by IncuCyte Zoom Kinetic Live Cell Imaging (Essen
BioScience). Proliferation rates were determined using IncuCyte based on percentage of confluence over time. For migration, cells were plated in 96 well ImageLock Plates (Essen
Bioscience). A scratch was made using a 96-pin WoundMaker (Essen Bioscience). For
21 invasion, 96 well ImageLock Plates (Essen Bioscience) were coated overnight with 100
µg/mL Matrigel Basement Membrane Matrix (Cat. # 356231, Corning). The following day,
Matrigel was aspirated and 25,000 cells were plated per well and allowed to adhere prior to making a scratch with the 96-pin WoundMaker (Essen Bioscience). 50µL of Matrigel Matrix
(8 mg/mL) was then added to the wells, and covered in 100 µL media containing the appropriate treatments (refer to figure legend for details). Migration and invasion were quantified using the “Relative Wound Density” metric generated by IncuCyte software.
Lentivirus
Individual pLKO.1 lentiviral shRNA clones were purchased from the University of
Colorado Cancer Center Functional Genomics Shared Resource. The RNAi Consortium identifiers used in this study were TRCN0000010661 (PRL-3 sh1), TRCN0000355597
(PRL-3 sh2). Transduced cells were selected in medium containing 2.5 µg/mL puromycin.
In vitro enzymatic assay
Purified, recombinant PRL-3 protein (500 µg) was added to phosphatase assay buffer
(12.5 mM p-NPP, 50 mM Tris 7.4, 150 mM NaCl, and 5 mM DTT), in a final volume of 50
µL and incubated for 30 minutes at 30° C in 96 well plates. After 30 minutes, p-nitrophenol was determined at 405 nm using a Synergy 2 microplate reader (BioTek). Specific activity was calculated using an extinction coefficient of 1.78 x 104 for p-NPP.
22 Site-directed mutagenesis
Cys104 to alanine substitution in the PRL-3 cDNA was created using the Quickchange
Site-Directed mutagenesis kit (Agilent) and confirmed by DNA sequencing.
Oligonucleotides used were: 5’-GCGTGGCTGTGCACGCCGTGGCGGGCCTGGG-3’ and
5’-CCCAGGCCCGCCACGGCGTGCACAGCCACGC-3’.
Apoptosis assays
CellPlayer 96-well Kinetic Caspase-3/7 reagent was used according to manufacturer’s instructions at 5 µM (Essen Bioscience). Cells were plated overnight and the following day treated with AMPI-109 (100 nM) or vehicle control.
Production of wild type and C104A FLAG-tagged recombinant PRL-3
Wild type and C104A cDNAs were amplified from the respective mammalian expression vectors by PCR using the oligonucleotides: 5’-
GATAAGCTTATGGCTCGGATGAACCGCCCGGCCCG-3’ and 5’-
GATCTCGAGCTACATAACGCAGCACCGGGTCTTG-3’. Products were ligated into pCR2.1-TOPO (Invitrogen) for DNA sequencing. Following confirmation of sequences, wild type and C104A cDNAs were excised with HindIII and XhoI and ligated into bacterial expression vector pFLAG-MAC (Sigma) that had been digested with HindIII and XhoI.
Expression vectors were transformed into the E. coli strain BL21. Production and purification of FLAG-tagged proteins was carried out with manufacturer’s instructions (Sigma).
23 Immunohistochemistry
Archival formalin fixed paraffin embedded (FFPE) tumors from 117 TNBCs, 67
Luminal A, 43 Luminal B and 29 HER2 patients diagnosed at the Massachusetts General
Hospital between 1976 and 1993 were used. All specimens were covered by an institutional review board (IRB) approved protocol. Tumor size, patient age at diagnosis and nodal status was determined by review of pathology reports. IHC methods for ER, PR and HER2 are previously reported (64). FFPE sections were utilized for IHC using standard methods. PRL-
3 antibody previously validated for IHC (1:4000; Cat. # ab82568 (PTP4A3), Abcam) was used as a primary reagent at 4° C. Two reviewers (A.D.T. and S.M.E.) reviewed the entire slide and determined the percentage and intensity of tumor cells that were positive for cytoplasmic PRL-3 staining.
Microarray and statistical analysis
Microarray datasets from breast tumors and respective statistics were obtained from
Oncomine (www.oncomine.com, Oct 2015, Thermo Fisher Scientific, Ann Arbor, MI).
Significance was calculated using the Student t test in GraphPad Prism version 6.04 software.
Bioinformatic statistical analysis of the functional genomic screen was carried out as previously described (26, 27). To determine differences in PRL-3 expression in breast cancer subtypes, non-parametric Kruskal-Wallis ANOVA and post hoc Dunn’s test was conducted to determine which tumor types were significantly different from each other. Analyses were conducted using SAS (ver 9.4, SAS Institute) or GraphPad Prism. Tests were two-sided with significance set at p<0.05.
24 Results
AMPI-109 binds PRL-3 active site in silico and impairs its catalytic activity in vitro
To query the potential of PRL-3 as a target of AMPI-109, we carried out in silico docking experiments using a previously solved NMR structure representing the open conformation of PRL-3 (65) and a 3-D model of AMPI-109. We did not pre-define active or potential allosteric sites to avoid biasing the simulation towards a particular domain. After running 20 independent simulations, the 15 lowest favorable energy binding states of AMPI-
109 consistently occurred at the shallow PRL-3 active site, with Gibb’s free energy values ranging from -7.2 to -7.9 kcal/mol (Figure 2.1). The active site was defined as the shallow pocket spatially demarcated by Cys104, Asp72 and Asn142 residues, all of which have been implicated in either enzyme activation, activity or substrate specificity (66).
Figure 2.1 AMPI-109 binds PRL-3 in silico. Left: Space-filling model of NMR-solved PRL- 3 protein structure with top 15 predicted lowest energy binding states of AMPI-109 in the PRL-3 active site. Right: Ribbon model of PRL-3 depicting AMPI-109 docking in the active site. D72, C104, and N142 represent aspartic acid 72, cysteine 104, and asparagine 142, respectively.
25 Of note, we observed close proximity of the bromo-acetate moiety of AMPI-109 to
Asn142, a variant residue to the structurally most similar phosphatases to PRL-3, such as
VHR, CDC14, PTEN and KAP (Figure 2.2). Binding of AMPI-109 to the active sites of these phosphatases was less favorable than PRL-3 as determined by in silico docking analyses. We also performed docking on several other candidates identified in our shRNA screen whose NMR or crystal structures were available (CDK10, EGFR, and RPS6). PRL-3 was the only protein predicted to bind AMPI-109 in an energetically favorable manner.
*
Figure 2.2 Protein sequence alignment between PRL-3 and the most structurally similar phosphatases to PRL-3. * indicates positioning of the variant residue, asparagine 142.
26 To investigate the ability of AMPI-109 to inhibit the catalytic activity of PRL-3, we performed in vitro phosphatase assays under reducing conditions using purified, active recombinant PRL-3 and p-NPP as a substrate. It has been previously reported that a closely related PRL, PRL-1, is enzymatically active in hydrolyzing p-NPP to p-nitrophenol under reducing conditions (67). We found that AMPI-109 potently impaired the catalytic activity of
PRL-3 in a dose-dependent manner (Figure 2.3A).
Moreover, to demonstrate that the alkylating capacity of AMPI-109 is specific to the unique PRL-3 active site, and not any cysteine-containing phosphatase, we carried out the same enzymatic assay using alkaline phosphatase and found that AMPI-109, in molar excess to a positive control, sodium orthovanadate (Na3VO4•2H2O), was not successful in impairing enzymatic activity (Figure 2.3B).
A B
* *
Figure 2.3 AMPI-109 impairs PRL-3 activity, but not that of alkaline phosphatase. A. In vitro phosphatase assays of PRL-3 catalytic activity. The effect of increasing concentrations of AMPI-109 on the conversion of p-NPP to p-nitrophenol was measured at 405 nm. Ctl; ethanol vehicle control. * The ANOVA test was significant (p<0.0001) and a post-hoc Tukey test found that the 2 and 2.5 nM groups were significantly different from each other and from the control and 1 nM groups. B. In vitro phosphatase assays of alkaline phosphatase catalytic activity. The effect of AMPI-109 on the conversion of p-NPP to p-nitrophenol was measured -3 at 405 nm. EtOH; ethanol vehicle control (blue), AMPI-109; 10 M (red), Na3VO4•2H2O; 10-6 M (green).
27 To further establish the potential relationship between AMPI-109 and PRL-3 active site binding, we conducted site-directed mutagenesis and changed the coding sequence of
PRL-3 cDNA to result in a substitution of the enzymatic nucleophile Cys104 to an alanine
(C104A). It has been previously reported that mutation of the enzymatic nucleophile Cys104 to serine or alanine abolishes PRL-3 activity (66). To confirm that C104A results in a catalytically impaired phosphatase, we expressed and purified FLAG-tagged wild type
(FLAG-WT) and C104A (FLAG-C104A) from E. coli (Figure 2.4A) and performed phosphatase assays using p-NPP as substrate. Mutation of Cys104 to alanine significantly impaired, but did not completely abolish, the catalytic activity of PRL-3 (Figure 2.4B).
Moreover, the addition of AMPI-109 was unable to further impair PRL-3 catalytic activity suggesting Cys104 is a critical residue for PRL-3 enzymatic activity and that AMPI-109 may inhibit PRL-3 activity through a Cys104-dependent mechanism.
AB * MW * 188 98
62 n.s. 49 38 28 17 14
6
Figure 2.4 Changing Cys104 to alanine results in catalytically impaired PRL-3. A. SDS- PAGE analysis of FLAG-tagged recombinant wild type (FLAG-WT) and PRL-3 C104A (FLAG-C104A). Numbers indicate molecular weights (MW) in kilodaltons. B. The enzymatic activity of FLAG-tagged recombinant proteins was determined by in vitro phosphatase assay. * = p-value <0.05, n.s. = no significance.
28 Finally, we investigated whether AMPI-109 treatment affected degradation of PRL-3.
After treating with cycloheximide (CHX) to block new protein synthesis we observed no change in total PRL-3 protein levels over the course of 3 hours, but did observe a substantial decrease in PRL-3 after AMPI-109 treatment (Figure 2.5). This decrease was prevented by the proteasome inhibitor MG-132, suggesting that AMPI-109 may suppress PRL-3 activity through dual activities. In addition to blocking PRL-3 catalytic activity, AMPI-109 may suppress PRL-3 by promoting protein degradation through a proteasomal mechanism.
Figure 2.5 PRL-3 protein level in BT-20 cells after treatment with cycloheximide (CHX, 100 nM) or AMPI-109 (100 nM). Bottom panel: PRL-3 protein levels in BT-20 cells after treatment with AMPI-109 (100 nM) and MG-132 (10 µM) over the course of 3 hours. Numbers represent quantification of band intensities (PRL-3/Actin).
PRL-3 knock down and expression of catalytically impaired PRL-3 inhibit TNBC cell growth and confer partial resistance to AMPI-109
PRL-3 is a dual-specificity protein tyrosine phosphatase that has been reported to be overexpressed in a number of cancer types (38-51). Studies have reported roles for PRL-3 in
29 modulating the cell cycle and promoting survival but none to our knowledge have examined the phenotypic consequences of modulating PRL-3 in TNBC cell lines.
We examined the role of PRL-3 in proliferation of TNBC cells by knocking down
PRL-3. We used two shRNA sequences that were predicted by the Genetic Perturbation
Platform of the Broad Institute to specifically target PRL-3 transcripts, but not the closely related family members PRL-1 and PRL-2, and observed significant knock down of PRL-3 protein in two TNBC cell lines (Figure 2.6A). Importantly, we verified knock down specificity for the PRL-3 shRNAs against both PRL-1 and PRL-2 by qRT-PCR. Both PRL-3 shRNAs (sh1 and sh2) exerted specific knock down action on PRL-3 and did not reduce
RNA levels of either PRL-1 or PRL-2 (data not shown). In both lines, knock down of PRL-3 significantly impaired TNBC cellular proliferation (Figure 2.6B).
A BT-20 MDA-MB-468 h1 h2 h1 h2 s s s s pLKO pLKO PRL-3 PRL-3 β-Actin β-Actin
B BT-20 MDA-MB-468
Figure 2.6 PRL-3 knock down results in reduced growth of TNBC cells. A. Immunoblot analysis of PRL-3 levels following transduction of BT-20 (left) and MDA-MB-468 (right) cells with lentiviral vectors expressing PRL-3 shRNAs (sh1 and sh2). pLKO is non-silencing control shRNA. B. Real time kinetic monitoring of cellular proliferation of BT-20 (left) and MDA-MB-468 (right) TNBC cells with non-silencing control shRNA (pLKO; blue) and PRL-3 shRNAs (sh1; red, sh2: green).
30 We also sought to determine the impact of impairing PRL-3 phosphatase activity to determine whether loss of activity could phenocopy PRL-3 knock down. We examined the phenotypic consequences of expressing the catalytically impaired form of PRL-3 in vivo. We transiently transfected BT-20 and MDA-MB-468 cells with an expression vector containing the cDNA for C104A (pC104A). Interestingly, expression of C104A closely phenocopied
PRL-3 knock down (Figure 2.7), suggesting that forced expression of a catalytically impaired form of PRL-3 in TNBC cells may result in a dominant negative form of PRL-3.
In an effort to identify a preliminary in vivo relationship between PRL-3 and AMPI-
109 mediated growth inhibition, we examined consequences of treating pC104A cells with
AMPI-109, reasoning C104A would synergize with or confer resistance against AMPI-109.
We observed partial resistance to AMPI-109 mediated inhibition in cell lines transfected with pC104A (Figure 2.7), further suggesting PRL-3 is involved in a signaling axis partially required for AMPI-109 action. This effect appears to be specific for PRL-3 because knock down of PRL-1 or PRL-2 did not confer any degree of resistance to AMPI-109.
BT-20 MDA-MB-468
Figure 2.7 Catalytically impaired PRL-3 reduces growth of TNBC cells and confers partial resistance to AMPI-109 efficacy. BT-20 (left) and MDA-MB-468 (right) cells were transiently transfected with pC104A mammalian expression vector (pC104A; purple) or empty vector (Ctl; blue) 24 hrs prior to plating in IncuCyte. Cells were treated with ethanol vehicle control (EtOH; green) or 100 nM AMPI-109 (AMPI-109; black). Transfection control cells treated with ethanol vehicle (Ctl + EtOH; red) and pC104A cells treated with AMPI-109 (pC104A + AMPI-109; orange) are also depicted. Cellular proliferation measured by real time kinetic monitoring of plate confluence. * = p-value <0.05 as determined by Student t test on last time-point.
31 PRL-3 modulation and AMPI-109 treatment inhibit TNBC cell migration and invasion
We first measured migration of BT-20 and MDA-MB-468 cells following PRL-3 knock down. After adjusting for differences in cell proliferation by nuclear count, knock down of PRL-3 almost entirely blocked the ability of these cells to migrate as early as 4 hours in both cell lines (Figure 2.8A). AMPI-109 also strongly inhibited the migration of BT-
20 and MDA-MB-468 cells (Figure 2.8B). Additionally, forced expression of C104A in BT-
20 and MDA-MB-468 cells decreased their migratory rates, providing further evidence for a dominant negative function of C104A in TNBC cells (Figure 2.8C). Conversely, overexpression of wild type PRL-3 significantly increased migration (Figure 2.8C).
A BT-20 MDA-MB-468 Figure 2.8 PRL-3 modulation and AMPI- 109 inhibit TNBC migration. A. Quantification of wound closure in control (pLKO; blue) B and PRL-3 knock down MDA-MB-468 BT-20 (sh2; red) TNBC cells. B. TNBC cells treated with 100 nM AMPI-109 (AMPI-109; green) or ethanol vehicle control (Ctl; red) C. TNBC cells transfected with C mammalian expression BT-20 MDA-MB-468 vectors for wild type PRL-3 (pWT; blue), catalytically impaired PRL-3 (pC104A; purple) or empty vector (Ctl; red). Inset panels show PRL-3 by western
Ctl pWT pC104A Ctl pWT pC104A blot. * = p-value <0.05 by Student t test on last PRL-3 PRL-3 time-point. Actin Actin
32 Because MDA-MB-468 cells express high levels of PRL-3 but do not invade significantly in Matrigel in vitro, we chose to examine the consequences of treating with
AMPI-109 and knocking down PRL-3 in SUM159 TNBC cells, which exhibit high levels of
PRL-3 and are invasive in Matrigel (Figure 2.9A). Treatment with AMPI-109 or knock down of PRL-3 significantly impaired the ability of SUM159 cells to invade Matrigel (Figure
2.9B). Conversely, BT-20 cells have lower levels of PRL-3 compared to SUM159 cells. We therefore examined the consequence of overexpressing PRL-3 in promoting invasiveness of these cells. We observed that BT-20 cells exhibiting PRL-3 overexpression could invade through Matrigel more rapidly and to a higher degree relative to control cells (Figure 2.9B).
A 231 468 - - Figure 2.9 PRL-3 modulation MB MB - - and AMPI-109 inhibit TNBC 20 - HCC70 BT MDA SUM159 MDA invasion. A. Western blot comparing PRL-3 protein PRL-3 levels across several TNBC cell Actin lines. B. Real-time kinetic monitoring of cellular invasion B SUM159 through Matrigel Matrix in SUM159 and BT-20 cells treated with 100 nM AMPI-109 (top panel; red), stably expressing two different shRNA clones to knock down PRL-3 (middle panel; red and SUM159 purple) or BT-20 cells transiently transfected to overexpress wildtype PRL-3 Ctl sh1 sh2 (bottom panel; red). Invasive score was determined by the PRL-3 ability of cells to invade Actin through a Matrigel wound BT-20 (Relative Wound Density). Inset panels show PRL-3 levels by immunoblot analysis. * = p- Ctl pWT value <0.05 as determined by PRL-3 Student t test on last time-point. Actin
33 PRL-3 expression positively associates with the TNBC subtype, regional and distant metastases
Our data implicate an oncogenic role for PRL-3 in TNBC by 1) providing a growth advantage to cells expressing PRL-3 and 2) enhancing TNBC cell migratory and invasive ability. To examine the in vivo relevance of PRL-3 expression, we evaluated PRL-3 protein expression in human breast cancers and its relationship to several clinicopathological variables. IHC analysis of archival human breast tissue was carried out for PRL-3 (Figure
2.10A). Using the percentage of tumor cells positive for PRL-3 among the four subtypes, we determined that PRL-3 expression was significantly higher in TNBC versus luminal A breast cancer (p = 0.008, Figure 2.10B). We also observed a significantly higher percentage of
PRL-3 positive tumor cells in patients who had regional disease at the time of diagnosis (n =
48, 66%) as compared with patients whose nodes were tumor free (n = 69, 45%; p = 0.024).
We augmented our immunohistochemical analyses through mining of publicly available microarray tumor data using Oncomine. We corroborated higher PRL-3 mRNA expression in TNBC versus other breast cancer subtypes, including non-tumorigenic tissue, and PRL-3 association with regional disease in the Lu Breast cohort (Figure 2.10C) (68).
Multiple other cohorts also showed an association between PRL-3 and regional disease
(Table 2.1). We also determined a positive correlation with metastatic (M1) versus visceral metastases-free disease (M0) in the Schmidt Breast cohort (Figure 2.10D) (69). Finally, multiple datasets revealed significant associations between increased PRL-3 expression in cancerous versus normal tissue, invasive ductal breast carcinoma compared to non-invasive disease, increased tumor grade, stage and development of metastases at 1 and 3 year intervals
34 following diagnosis (Table 2.1). Collectively, these data support our in vitro findings that
PRL-3 acts as an oncogenic mediator to drive aggressive phenotypes in breast cancer.
ACPRL-3 immunohistochemistry
p = 0.009
Low High B
D
p = 0.039
Figure 2.10 PRL-3 expression is increased in TNBC. A. Examples of PRL-3 staining scored as low or high, 20 µm scale bars. B. Whisker plot and table of PRL-3 expression across different breast cancer subtypes. * = p-value (<0.008) based on post hoc Dunn’s test from Kruskal-Wallis non-parametric ANOVA. C and D. Differential PRL-3 mRNA expression in two human breast cancer datasets examining correlation to regional disease and visceral metastases.
35 Table 2.1 PRL-3 mRNA expression correlates with multiple clinicopathological variables of aggressive breast cancer. Left column, Oncomine dataset examined. N = Sample size in analysis. IBC = Invasive Breast Cancer. IDC = Invasive Ductal Carcinoma. TNBC = Triple- Negative Breast Cancer. FC = Fold Change in PRL-3 expression. P values computed by Pearson r test.
N Cancer vs. Normal FC P-value Richardson Breast 2 47 2.288 2.81e-6 N IBC vs. normal FC P-value Curtis Breast 2,136 1.369 1.13e-4 TCGA Breast 593 1.949 6.87e-15 Gluck Breast 158 1.491 4.00e-3 N IDC vs. normal FC P-value Curtis Breast 2,136 1.391 1.54e-37 TCGA Breast 2 1,602 1.27 3.60e-72 Turashvili Breast 30 2.441 2.00e-03 N TNBC vs. other subtypes FC P-value Richardson Breast 2 47 2.043 1.50e-2 Minn Breast 2 121 1.274 1.03e-04 Kao Breast 327 1.427 5.00e-3 Chin Breast 118 1.485 5.00e-3 Korde Breast 61 1.298 3.00e-2 Bonnefoi Breast 160 1.186 1.70e-2 N Grade 3 vs. 2 FC P-value Desmedt Breast 198 1.931 1.00e-2 Nik-Zainal Breast 21 1.378 4.00e-3 N Lymph node positive vs. negative FC P-value Stickeler Breast 57 1.901 3.10e-2 Lu Breast 129 2.028 9.00e-3 N 5 year recurrence FC P-value Finak Breast 59 1.24 2.30e-2 Loi Breast 3 77 1.343 3.20e-2 Ma Breast 3 60 1.325 2.30e-2 N Metastasis at 1 year FC P-value Schmidt Breast 200 1.509 3.90e-2 N Metastasis at 3 years FC P-value Symmans Breast 2 103 1.393 5.00e-2 Bos Breast 204 1.264 9.00e-3 Kao Breast 327 1.274 8.00e-3
36 Discussion
TNBCs are among the most aggressive breast cancer subtypes and are associated with a higher risk of metastasis and death when compared to other breast cancer subtypes (3-5).
While a molecular explanation for this higher risk associated with poor outcome remains unknown, it is likely due, in part, to signaling pathways governing TNBC cell growth, migration and invasion that enhance TNBC metastatic potential.
In this chapter, I demonstrate that PRL-3 knock down or impairment of PRL-3 activity phenocopies but does not synergize with AMPI-109 in impairing TNBC cell growth, migration and invasion. The exact mechanistic relationship between AMPI-109 and PRL-3, including the possibility that AMPI-109 directly targets PRL-3 as indicated by our in silico and in vitro studies appears plausible, but nonetheless remains an active area of investigation for our laboratory. In addition, we propose it unlikely that PRL-3 is the sole “modifier” of
AMPI-109 activity, as PRL-3 appears to be expressed in non-TNBC subtypes as evidenced by immunoblot in non-TNBC cell lines and our retrospective IHC analyses. Therefore, additional modifiers for AMPI-109 or other PRL-3 pathway associated effectors likely exist and the identification of these molecules is also actively ongoing in our laboratory.
Perhaps the most characterized role of PRL-3, extensively studied in CRC, is the promotion of cell migration, invasion and metastasis (70-74). One proposed model by which this is thought to occur is PRL-3 mediated downregulation of PTEN, leading to activation of
PI3K/Akt signaling and induction of EMT (50). Further studies propose direct interaction between PRL-3 and ezrin and/or integrin α1 and β1 to alter focal adhesion signaling pathways responsible for promoting cell restructuring and migration (46-49). However, we did not observe changes in ezrin phosphorylation after PRL-3 modulation in BT-20 and
37 MDA-MB-468 TNBC cells (data not shown). One possible explanation for this finding may be that ezrin is not a PRL-3 substrate in TNBC cells. We did, however, observe lower levels of Akt phosphorylation after PRL-3 knock down in MDA-MB-231 TNBC cells. To understand the molecular mechanism behind the observed oncogenic functions of PRL-3 in
TNBC, I report in chapters III and IV on cellular signaling pathways modulated by PRL-3 that promote the growth, migration and invasion of TNBC cells. Understanding these pathways is critical to expanding our knowledge of how PRL-3 promotes aggressive TNBC phenotypes. For example, MDA-MB-468 TNBC cells which express high levels of PRL-3, but do not invade through Matrigel in vitro, possess additional biomarkers, such as matrix metalloproteinases and focal adhesion components, which may be used to further stratify the invasive potential of TNBC cells beyond PRL-3. These data are reviewed in more detail in chapter IV.
Previous studies have associated high PRL-3 expression with poor prognoses in node- positive TNBC and other subtypes including HER2 overexpressing breast tumors (44, 45), but very little mechanistic evidence implicates a role for PRL-3 in driving breast cancer progression. Our finding that PRL-3 is overexpressed in invasive human basal tumors, which include TNBCs, in parallel with our evidence demonstrating knock down or catalytic impairment of PRL-3 mimic the effect of AMPI-109 on inducing growth, migratory and invasion blockade, imply that PRL-3 may be a critical component behind the aggressive nature of TNBCs. In concordance with the studies presented herein, another study has preliminarily reported that PRL-3 is critical for cell growth of TNBC (75). Interestingly, overexpression of PRL-3 in BT-20 and MDA-MB-468 TNBC cells does not lead to additional growth advantages. These data suggest that PRL-3 expression levels in these cells
38 already confers the maximal growth promoting capabilities of PRL-3. However, overexpression of PRL-3 does partially abrogate the antiproliferative properties of AMPI-109
(data not shown).
In the subsequent chapters of this thesis, I present data on the molecular mechanisms that we have identified as responsible for driving the observed oncogenic functions of PRL-3 in TNBC; the promotion of growth, migration and invasion. Understanding these pathways is critical to expanding our knowledge of how PRL-3 promotes aggressive TNBC phenotypes, and sheds new light on potential molecular vulnerabilities of TNBCs which may be therapeutically actionable.
39 CHAPTER III
PRL-3 PARTICIPATES IN A PRO-INFLAMMATORY FEEDBACK LOOP THAT
MODULATES TNF-R1 PLEIOTROPISM TO CONTROL TNBC GROWTH
Introduction
In chapter I, I reported on a genome-wide functional genomic shRNA screen conducted in our laboratory to identify genes that, when silenced, conferred resistance to the anticancer agent, AMPI-109. In chapter II, I demonstrated a relationship between AMPI-109 and PRL-3 and showed that AMPI-109 treatment and a reduction of PRL-3 expression leads to substantial growth inhibition and a reduction in the migratory and invasive ability of
TNBC cells. These studies established a strong case for the exploration of how PRL-3 inhibition or knock down affect TNBC growth potential. However, elucidation of the exact mechanism(s) by which loss of PRL-3 expression or activity impairs TNBC growth was not addressed by the exploratory phenotypic studies of chapter I and II. Moreover, very little evidence exists in literature for how PRL-3 alters the growth potential of cancer cells.
Stimulating cell senescence and apoptosis are proven methods for combating cancer.
However, senescence and apoptosis are conventionally viewed as parallel, not sequential, processes. In this chapter, I present data demonstrating that PRL-3 is transcriptionally regulated by the NF-ĸB pathway, and that PRL-3 knock down elicits a TNF-R1 feedback loop that results in TNBC cell senescence as a pre-determinant to engaging cell apoptosis.
In these studies, knock down of PRL-3 lead to rapid G1 cell cycle arrest and induction of a strong TNFα cytokine response that promoted a period of cellular senescence through TNF-R1 mediated activation of NF-ĸB. Interestingly, these senescent PRL-3 knock
40 down cells subsequently underwent apoptosis as a result of increased TNF-R1 signaling through the TNFα-associated extrinsic death pathway, shunting signaling away from the NF-
ĸB cascade. These data suggest that TNF-R1 signaling dynamically re-programs after PRL-3 knock down, from sustaining cell senescence through NF-ĸB, to promoting apoptosis through TNF-R1 internalization and caspase-8 activation.
The data presented in this chapter represent an important new discovery on the mechanisms of PRL-3 activity in TNBC, and shed new light on molecular mechanisms that arbitrate the survival-death balance of TNF-R1 signaling, which are poorly understood despite the fact that TNF-R1 has been extensively studied. We describe a direct association between pro-inflammatory NF-ĸB signaling and PRL-3 expression in TNBC, and report on a novel and dynamic feedback loop induced by PRL-3 ablation that affects TNF-R1 pleiotropism, leading to sequential TNBC cell senescence and apoptosis.
Materials and methods
Materials
Recombinant human TNFα was purchased from PeproTech (Cat. #300-01A), CHX from Sigma-Aldrich (Cat. #C-6255), Monodansyl cadaverine (MDC) from Sigma-Alridch
(Cat. #30432), caspases-8 and 9 inhibitors from BD Pharmingen (Cat. #51-69401U and 51-
69411U), and IKK inhibitor, PS 1145 (Cat. # 4569) from Tocris Bioscience.
Plasmids, transfection and viral transduction
PRL-3 cDNA expression vector was purchased from Origene (Cat. # SC308739).
Transfections were carried out using Mirus TransIT LT1 reagent according to manufacturer’s
41 instructions (Mirus Bio). Individual pLKO.1 lentiviral shRNA clones were purchased from the University of Colorado Cancer Center Functional Genomics Shared Resource. The RNAi
Consortium identifiers: TRCN0000010661 (shPRL-3, shPRL-3 #1), TRCN0000355597
(shPRL-3 #2), TRCN0000014683 (shp65 #1), TRCN0000014684 (shp65 #2). Transduced cells were selected in medium containing 2.5 ug/mL puromycin.
Luciferase assays
The luciferase reporter used to measure NF-ĸB activity contained four copies of the
NF-ĸB consensus sequence fused to a TATA-like promoter (PTAL) region from the Herpes simplex virus thymidine kinase promoter (pNF-ĸB-Luc, Clonetech Cat. # 6053-1). The reporter used to measure AP-1 activity (pAP1(PMA)-TA-Luc) was purchased from clonetech
(Cat. # 6056-1). Reporter plasmids were transfected into cells at 1 µg/well. Twenty-four hours after NF-ĸB-Luc transfection, cells were washed with PBS and lysed using luciferase harvest buffer, or, twenty-four hours after AP1-Luc transfection, cells were treated with 10 uM PS 1145 overnight and lysed in luciferase harvest buffer the following day. Buffer recipes and luciferase work-up was conducted as per our previous protocol (76).
qRT-PCR
Total RNA isolation, cDNA synthesis and qRT-PCR were performed according to our previous protocol (76) with the following oligonucleotides: PRL-3 5’-
AGTTGCCCGCTTTACTTTGGTTGG-3’ and 5’-AGGAAGCTGCCCACTGTTTGGATA-
3’, p65 5’-TATCAGTCAGCGCATCCAGACCAA-3’ and
AGAGTTTCGGTTCACTCGGCAGAT-3’, TNFα 5’-
42 AATCGGCCCGACTATCTCGACTTT-3’ and 5’-TTTGAGCCAGAAGAGGTTGAGGGT-
3’, Cyclin E1 5’-GTACTGAGCTGGGCAAATAGAG-3’ and 5’-
GAAGAGGGTGTTGCTCAAGAA-3’, Cyclin D1 5’-CCACTCCTACGATACGCTACTA-
3’ and 5’-CCAGCATCTCATAAACAGGTCA-3’, E2F1 5’-
GCTGGACCACCTGATGAATATC-3’ and 5’-GTCTGCAATGCTACGAAGGT-3’, p53
5’-AGGGATGTTTGGGAGATGTAAG-3’ and 5’-CCTGGTTAGTACGGTGAAGTG-3’, p27 5’-CTAACTCTGAGGACACGCATTT-3’ and 5’-TGCAGGTCGCTTCCTTATTC-3’, p16 5’-CGCTAAGTGCTCGGAGTTAATA-3’ and 5’-CGACCCTGTCCCTCAAATC-3’.
Cell culture, senescence and immunoblot analysis
Cell lines were obtained from the University of Colorado Cancer Center Tissue
Culture Shared Resource. All cells were authenticated by short tandem repeat DNA profiling performed by the UCCC DNA Sequencing and Analysis Core. β-galactosidase (β-Gal) staining was performed to manufacturer protocol (Cat. # 9860S, Cell Signaling). Western blot analysis was conducted according to our previous protocol (21). PRL-3 (Cat. # ab82568,
Abcam), p65 (Cat. #8242, Cell Signaling), TNFα (Cat. #8301, Santa Cruz Biotechnology),
Phospho-c-Jun S73 (Cat. #3270, Cell Signaling), c-Jun (Cat. #9165, Cell Signaling), β-actin
(Cat. # A5441, Sigma-Aldrich). Immunofluorescence staining was performed as previously described (77) using an antibody to p65 (Cat. #06-418), purchased from EMD Millipore.
Cellular size, count and proliferation assays
Cellular proliferation was assessed using the CellTiter 96 MTS cell Proliferation assay (Cat. # G109A, Promega). Samples were read at 490 nm in a Synergy 2 microplate
43 reader (BioTek). Cellular size, count and proliferation rates were also determined using the
IncuCyte Zoom Kinetic Live Cell Imaging (Essen BioScience). IncuCyte cell recognition software calculated values based on cellular size over time, cellular count per field of view or percentage of plate confluence over time.
Chromatin immunoprecipitation (ChIP) assay
p65 binding to the PRL-3 promoter was analyzed following 30 minutes of TNFα treatment in BT-20 and MDA-468 cells. Cells were treated with formaldehyde to crosslink proteins to DNA. Soluble chromatin was prepared by sonication and subjected to immunoprecipitation (IP) with an antibody specific to p65 (Cat. #8242, Cell Signaling).
Control IP samples using appropriate non-specific antibody were processed in parallel.
Extensive washing with buffers of increasing ionic strength were performed to remove unbound proteins and DNA. A phenol-chloroform precipitation reaction was utilized to further isolate DNA from protein. The purified DNA was then subjected to quantitative real- time PCR (qRT-PCR) using oligonucleotides specific for each of 5 putative p65 binding elements in the human P3 promoter [-7 kb, -6 kb, -4 kb, -1.3 kb, +0.4 kb]. Oligonucleotides specific to the NF-ĸB target gene promoter of IL-6 was used as a positive control.
Oligonucleotides: Site 3 5’-CTTAGCTTGAGCCCTCCCAC-3’ and 5’-
CCCCTCTCCTGAGCTCCCAG-3’, Site 4 5’-CTTAGCCAGGCCCAGTGGGC-3’ and 5’-
GCTGACTCCAGGGCAGGAAC-3’, Site 5 5’-CCGCTCGCTCCCGCTGTTAC-3’ and 5’-
GAAGGATCCCGGAACGCTCC-3’, Site 6 5’-GAGAGAACCCAGTTAACTGG-3’ and 5’-
CAAGCCCTGCAGAACCCTTC-3’, Site 7 5’-GTGACCCCTGTGGAGTGGAT-3’.
44 Cell cycle and apoptosis analysis
Cells were incubated in Krishan stain overnight and analyzed using a Beckman
Coulter FC500 flow cytometer. Doublets are excluded from the analysis using the peak vs. integral gating method. ModFit LT software (Verity Software House) was used for cell cycle analysis. CellPlayer 96-well Kinetic Caspase-3/7 reagent was used at 5 uM according to the manufacturer’s instructions (Essen Bioscience).
Gene co-expression and statistical analysis
RNA expression heatmaps were generated using GENE-E (Broad Institute, www.broadinstitute.org/cancer/software/GENE-E). Significance was calculated using the
Student t test in GraphPad Prism version 6.04 software. Tests were two-sided with significance set at p < 0.05.
Results
Knock down of PRL-3 induces senescence in TNBC cell lines
Previously, we reported on the role of PRL-3 on the growth of TNBC cells. Here, we analyzed changes in cellular morphology as a visual indicator of cell viability following stable lentiviral knock down of PRL-3 (shPRL-3) in two TNBC cell lines, BT-20 and MDA-
MB-468 in an effort to identify mechanisms of PRL-3 mediated growth inhibition. Knock down of PRL-3 was associated with substantial increases in cellular size, accompanied by flattened and rounded morphology in both TNBC cell lines (Figure 3.1A and B).
45 A B Parental shPRL-3 20 - BT 468 - MB - MDA
Figure 3.1 Knock down of PRL-3 leads to enlargement of TNBC cell size. A. Phase contrast images of BT-20 and MDA-MB-468 cells transduced with lentivirus to knock down PRL-3 (shPRL-3). B. Quantification of (A) as determined by ImageJ analysis.
Enlargement of cell size is one visual indicator that cells may be engaging a senescence program. To test the hypothesis that cells exhibiting PRL-3 knock down enter senescence, we profiled PRL-3 knock down cells for the expression of senescence associated-β-galactosidase (SA-β-Gal) at pH 6.0, a hallmark of senescent cells. Knock down of PRL-3 was associated with increased cytoplasmic SA-β-Gal staining in both TNBC cell lines (Figure 3.2A), with minimal SA-β-Gal staining observed in control TNBC cells.
We also strengthened our analysis by examining the ability of mitogenic stimulation to restore cellular growth in SA-β-Gal positive clones, as senescent cells should not respond to growth factor stimulation. As expected, PRL-3 knock down SA-β-Gal positive cells did not respond to increased mitogenic stimulation relative to control cells over 72 hours (Figure
3.2B).
Taken together, these data indicate that knock down of PRL-3 in TNBC cells leads to initiation of an irreversible cell senescence program.
46 ABParental shPRL-3 Gal β- 20 - BT Phase Gal β- 468 - MB - MDA Phase
Figure 3.2 Knock down of PRL-3 induces senescence in TNBC cell lines. A. Phase contrast images depicting SA-β-Gal positive cells after PRL-3 knock down in two TNBC cells. B. Effect of increased serum concentration (20% FBS vs. 10% FBS) on cellular growth as assessed by MTS assay. Growth of SA-β-Gal positive clones are depicted by red bars and control cell growth is depicted by blue bars. Images captured at 10x magnification. Data represented is the mean ± SD of three independent experiments. * = p-value <0.05 as determined by Student t test between initial and last data points.
Senescent PRL-3 knock down cells arrest in G1 despite checkpoint aberrations
Basak and colleagues previously reported that PRL-3 enforces G1 cell cycle arrest in triple knockout mouse embryonic fibroblasts (MEFs) deficient for all three retinoblastoma protein (Rb) family members (Rb1, p107, and p130) (78). To determine the phase of the cell cycle in which PRL-3 knock down induces arrest in TNBC cells, we quantitated DNA content using propidium iodide and flow cytometry. An absolute increase in the G1 population was observed following PRL-3 knock down in BT-20 cells (Figure 3.3A). To identify whether PRL-3 knock down results in G1 arrest pre- or post-restriction point (Rpt),
47 we carried out mRNA molecular marker analyses for several cyclin-dependent kinase
INK4A inhibitors (CDKIs), including p16 (pre-Rpt) and p27 (post- Rpt), and cyclins D1 (pre-Rpt) and E1 (post-Rpt). Following PRL-3 knock down using two shRNA clones (shPRL-3 #1 and shPRL-3 #2) in BT-20 cells, which possess a homozygous deletion for p16INK4A, we observed strong upregulation of p27 at the RNA level (Figure 3.3B). We also observed an increase in cyclin D1 levels in the same cells, indicating that BT-20 cells are capable of
INK4A progressing through the Rpt as a consequence of the p16 deletion, which relieves repression on the checkpoint. In contrast, MDA-MB-468 cells, which possess a partial homozygous deletion for Rb1, strongly upregulate p16INK4A to induce G1 arrest following
PRL-3 knock down (Figure 3.3C).
A Figure 3.3 Senescent PRL-3 pLKO shPRL-3 knock down cells arrest in G1 despite unique molecular G1: 43.25 % G1: 55.72 % S: 40.74 % S: 32.37 % checkpoint aberrations. A. G2M: 16.01 % G2M: 11.91 % Flow cytometry assessment of DNA quantitation using propidium iodide incorporation in BT-20 cells exhibiting PRL- 3 knock down. B and C. qRT- PCR results depicting percent 12.47 % absolute increase in G1 population RNA change of multiple cell cycle checkpoint regulators in B C BT-20 and MDA-MB-468 cells, respectively, transduced * with two shRNA clones to * PRL-3 (shRNA #1 (blue) and shRNA #2 (red)). Data are represented as percent change from control (pLKO) * expression. Data represented is * the mean ± SD of three * * independent experiments. * = p-value <0.05 as determined by Student t test.
48 These data suggest that, in contrast to BT-20 cells, cell cycle arrest in MDA-MB-468 cells occurs pre-Rpt, and thus, prevents the cell from capitalizing on relieved repression of
E2F1 as a result of the Rb1 deletion. Taken together, these data reveal that knock down of
PRL-3 induces G1 arrest in TNBC cells irrespective of pre-existing CDKI genetic aberrations that would normally create a permissive environment for G1/S transition.
The NF-ĸB transcription factor subunit, p65, binds the PRL-3 promoter and regulates
PRL-3 expression
We reasoned that if PRL-3 knock down associates with cell cycle arrest and senescence in TNBC cells, then understanding regulators of PRL-3 could provide clues toward PRL-3 function or pathway association and shed additional light on how PRL-3 modulation leads to TNBC growth arrest. To this end, we carried out microarray co- expression analyses of genes that clustered with PRL-3 expression across multiple breast cancer cell lines (Figure 3.4).
Figure 3.4 Heatmap generated with GENE- E (Broad Institute), depicting differential gene expression data across multiple breast cancer cell lines. Id references probes used to identify listed gene. Blue squares indicate lower levels of RNA expression relative to red squares. Hierarchical clustering of the dataset is depicted above the heatmap.
49 Through this analysis, we identified an upstream effector member of the NF-ĸB pathway, TRAPPC9, as a gene with expression levels that track positively with PRL-3 expression (r = 0.692). Because TRAPPC9 is transcriptionally regulated by p65, we presumed that PRL-3, may also be regulated by p65. We examined this by knocking down p65 using two shRNA clones (shp65 #1 and shp65 #2), and observed decreased PRL-3 RNA
(Figure 3.5A) and protein (Figure 3.5B), suggesting p65 may also be involved in regulating
PRL-3 expression. To test this hypothesis, we examined PRL-3 RNA levels following treatment with recombinant human TNFα to activate the NF-ĸB pathway, and observed a 3-4 fold increase in PRL-3 RNA in two TNBC cell lines (Figure 3.5C).
A B C
* * * *
p65
PRL-3
Actin
Figure 3.5 Expression of the NF-ĸB subunit, p65, alters PRL-3 expression. A. BT-20 RNA levels as assessed by qRT-PCR for p65 and PRL-3 following p65 knock down using two different shRNA clones (shp65 #1 and shp65 #2). B. Western blot analysis of p65 and PRL-3 protein levels following p65 knock down in BT-20 cells. C. Fold change in PRL-3 RNA in BT-20 (blue) and MDA-MB-468 (red) cells as determined by qRT-PCR after 6-hour treatment with TNFα (20 ng/mL) or TNFα (20 ng/mL) and cycloheximide (100 nM) to block new protein synthesis.
To determine if PRL-3 is a direct transcriptional target for p65, we co-treated cells with TNFα and cycloheximide to block de novo protein synthesis and observed a 2.5-3 fold increase in PRL-3 RNA (Figure 3.5C). These data suggest that in response to TNFα, p65 largely, and directly, regulates PRL-3 RNA.
50 In order to confirm whether p65 can directly bind the PRL-3 promoter, we scanned a
20 kb genomic region spanning the PRL-3 transcriptional start site and identified seven NF-
ĸB consensus binding elements (Figure 3.6A). We explored p65 binding in vivo by performing ChIP followed by qRT-PCR using oligo primers designed to recognize each of the five consensus sites closest to the transcriptional start site (Figure 3.6A, sites 3-7; blue arrows). We observed a 30-40 fold enrichment of p65 following TNFα treatment at consensus site 5, with minimal loading at sites 3, 6 and 7 in BT-20 cells (Figure 3.6B).
Similarly, we observed a 4-7 fold enrichment in the MDA-MB-468 cell line at site 5, and minimal loading at site 6 (Figure 3.6B).
A
B
Figure 3.6 p65 binds to the PRL-3 promoter. A. Schematic of the human PRL-3 promoter locus. Seven potential NF-ĸB binding sites indicated with hash marks and numbered, with the position of each site with respect to the transcription start site (red arrow) indicated. Blue arrows represent qRT-PCR oligonucleotide primers designed to recognize the respective binding sites. B. Results of three independent replicate ChIP experiments performed in BT- 20 and MDA-MB-468 cells following 30-minute treatment with 20 ng/mL TNFα. Data are represented as fold enrichment relative to IgG control. Data represented is the mean ± SD of three individual replicates. * = p-value <0.05 as determined by Student t test.
51 PRL-3 knock down cells upregulate TNFα to sustain senescence through NF-ĸB activation
p65 regulated genes, such as IĸBs, are known to exert reciprocal regulation on the NF-
ĸB pathway. In order to assess whether PRL-3 functions as part of a feedback loop, we determined the effect of PRL-3 knock down on the ability of p65 to translocate to the nucleus.
We hypothesized that if PRL-3 knock down impaired the growth of TNBC cells, p65 could be restricted from nuclear translocation and driving the expression of pro-survival genes.
Following PRL-3 knockdown in BT-20 cells, we surprisingly observed nuclear localization of p65 in the absence of TNFα treatment (Figure 3.7A), suggesting reactivation of NF-ĸB.
Furthermore, knock down of PRL-3 and treatment with exogenous TNFα was associated with diminished p65 staining, suggesting possible downregulation of the pathway. In order to determine whether nuclear p65, following PRL-3 knock down, was attributed to an increase in
TNFα, we profiled PRL-3 knock down BT-20 cells for TNFα expression. We observed a substantial increase in TNFα RNA (Figure 3.7B) and protein (Figure 3.7C). Moreover, this increase in TNFα was capable of functionally activating the NF-ĸB pathway in an autocrine fashion, as determined by transiently transfecting PRL-3 knock down cells with an NF-ĸB- luciferase reporter. We observed a ~2.5 fold induction of luciferase activity (Figure 3.7D), which was largely reversible using a TNFα neutralizing antibody. Considering knock down of
PRL-3 in TNBC cells invokes cell senescence, but also results in increased NF-ĸB signaling activity, is somewhat paradoxical. However, a resolution came after studying data on the roles of NF-ĸB in cancer. Indeed, accumulating evidence suggests a complex and context-dependent role for the NF-ĸB pathway in cancer, including the ability to maintain cellular senescence
(79-81).
52 Continued suppression of PRL-3 results in a dynamic TNFα-associated apoptosis feedback loop in senescent TNBC cell lines
TNFα is a pleiotropic cytokine that binds TNF-R1 and elicits diverse responses ranging from maintaining cell viability and proliferation, to activation of apoptosis (82,83). Upon
TNFα binding, TNF-R1 recruits the adaptor, TRADD, to its cytoplasmic death domain (84,
85). TRADD acts as a scaffolding platform to recruit both RIP-1 and TRAF-2 to activate the
NF-ĸB and AP-1 pathways (84), or recruits FADD and pro-caspase-8 to initiate apoptosis (86,
87). Very little is known about what regulates this survival-death balance of TNF-R1 signaling, despite the fact that it has been intensely studied since the late 1980s.
Since upregulation of TNFα and activation of NF-ĸB is associated with the senescence period in PRL-3 knock down cells, we hypothesized that eventually, senescent PRL-3 knock down cells may become sensitized to these high levels of TNFα, leading to apoptosis. We examined apoptosis by monitoring caspase-3/7 activity and observed that cells exhibiting PRL-
3 knock down underwent a higher rate of apoptosis relative to control cells 24 hours after transduction (Figure 3.7E). The timing of this higher apoptosis rate coincides with the approximate time TNFα levels are highest following PRL-3 knock down (Figure 3.7C; 24 hours). Prior to 24 hours, the rate of apoptosis is equivalent to control cells, suggesting PRL-3 knock down cells are still senescent and remain viable through NF-ĸB sustained signaling.
These data indicate that in the initial period following PRL-3 knock down, TNFα stimulates the TNF-R1 NF-ĸB program to invoke senescence. After 24 hours of increasing
TNFα signaling, however, the TNF-R1 pathway may re-program to initiate apoptosis. This hypothesis could support our IF observation of reduced p65 staining in PRL-3 knock down cells treated with TNFα, as these cells have likely downregulated NF-ĸB to promote apoptosis.
53
A -TNFα +TNFα -TNFα +TNFα D pLKO pLKO shPRL-3 shPRL-3 Composite p65 E
* Nuclei
B C
Hours 0 2 6 12 24 48 *
α shPRL-3
pLKO IB: TNF
Figure 3.7 Continued suppression of PRL-3 expression results in a dynamic TNFα- associated apoptosis feedback loop in TNBC cell lines. A. Composite immunofluorescence images (top row) of p65 (middle row) and DAPI-stained nuclei (bottom row) following PRL- 3 knock down and treatment with 20 ng/mL TNFα in BT-20 cells. Images were captured at 100x magnification. B. Expression level of TNFα RNA following PRL-3 knock down in BT- 20 cells as assessed by qRT-PCR. C. Western blot analysis of TNFα protein in conditioned media over the course of 48 hours following PRL-3 knock down in BT-20 cells. D. Fold- change of NF-ĸB activity in BT-20 (blue) and MDA-MB-468 (red) cells exhibiting PRL-3 knock down, as assessed by transfection with an NF-ĸB –luciferase reporter plasmid. Cells were either treated or untreated with a neutralizing TNFα antibody. E. Real-time kinetic monitoring of caspase 3/7 activity in control (blue) and PRL-3 knock down (red) BT-20 and MDA-MB-468 cells as quantified by the IncuCyte Zoom Live Cell Imaging System. Data represented is the mean ± SD of three independent replicates. * = p-value <0.05 as determined by Student t test.
55 Blockade of the TNFα-associated TNF-R1 extrinsic death pathway confers significant resistance to PRL-3 knock down cell death
In order to investigate whether increased levels of TNFα following PRL-3 knock down are responsible for promoting eventual cell death through TNF-R1, we blocked a critical step in the initiation of the extrinsic cell death pathway. TNF-R1 internalization is instrumental for recruitment of TRADD, FADD and caspase-8 and activation of TNFα- induced apoptosis (88). The transglutaminase inhibitor, monodansyl cadaverine (MDC), blocks TNF-R1 endocytosis and prevents TNF-induced apoptosis (89). As expected, we found that MDC treatment in PRL-3 knock down cells conferred significant resistance to
PRL-3 knock down-mediated cell death (Figure 3.8A). Moreover, inhibition of caspase-8, downstream of TNF-R1 endocytosis and responsible for activating extrinsic apoptosis, induced similar levels of resistance (Figure 3.8B) to PRL-3 knock down-mediated cell death.
In contrast, blockade of the intrinsic cell death pathway using a caspase-9 inhibitor showed no resistance to PRL-3 knock down-mediated cell death (Figure 3.8C), further validating that elevated TNFα levels following PRL-3 knock down, eventually induces cell death through an extrinsic TNF-R1 guided pathway.
Inhibiting NF-ĸB activation following PRL-3 knock down sustains TNBC cell survival through c-Jun/AP-1
Our studies suggest that the NF-ĸB senescence program may be a pre-requisite step to
TNFα-associated TNF-R1 cell death. Therefore, we hypothesized that blockade of the NF-ĸB signaling pathway after PRL-3 knock down might result in accelerated cell death or sustained
55 cell survival. We blocked NF-ĸB signaling using the IKK inhibitor, PS 1145, in BT-20 cells and observed sustained survival in cells exhibiting PRL-3 knock down (Figure 3.8D).
To understand whether AP-1, a heterodimeric transcription factor composed of c-Fos and c-Jun (90) and the third signaling cascade of the TNF-R1 pathway, is responsible for the observed level of cell survival, we immunoblotted for c-Jun activation. We observed substantial activation of c-Jun in cells exhibiting PRL-3 knock down and treated with the
IKK inhibitor as assessed by phosphorylation at serine 73 (Figure 3.8E). Moreover, to investigate the functional significance of c-Jun phosphorylation in the context of AP-1 transcriptional activity, we transiently transfected PRL-3 knock down cells with an AP1- luciferase construct and treated cell with the IKK inhibitor. As expected, we found enhanced
AP-1 transcriptional activity in PRL-3 knock down TNBC cells unable to signal through NF-
ĸB (Figure 3.8F). These data suggest that by-pass of the senescence program following PRL-
3 knock down does not accelerate cell death through TNF-R1, and is a pre-requisite event for inducing apoptosis.
Taken together, these data demonstrate that PRL-3 is a transcriptional product of pro- inflammatory NF-ĸB signaling in TNBC cells, and that PRL-3 ablation leads to a reciprocal
TNFα-driven feedback loop mediated by TNF-R1, to control TNBC cell senescence through activation of NF-ĸB and eventual apoptosis through TNF-R1 internalization and caspase-8 activation (Figure 3.9).
56 A D
* * * *
B E pLKO shPRL-3
IKK inhibitor - + - +
P-c-Jun (S73)
c-Jun * * PRL-3
Actin C F *
*
Figure 3.8 Blockade of the TNFα-mediated TNF-R1 extrinsic death pathway confers significant resistance to PRL-3 knock down cell death, while inhibition of NF-ĸB following PRL-3 knock down sustains TNBC cell survival through c-Jun/AP-1. Real-time kinetic monitoring of BT-20 cell proliferation was determined by the IncuCyte Zoom Live Cell Imaging System. Control cells (green) and PRL-3 knock down cells (red) were treated with vehicle (blue) or the following inhibitors (black): A. Monodansyl cadaverine (MDC) at 100 uM, B. Caspase 8 inhibitor at 20 uM, or C. Caspase 9 inhibitor at 20 uM. D. BT-20 cell proliferation as assessed in (A-C) with 10 uM IKK inhibitor. E. Western blot analysis depicting changes in activation of c-Jun as assessed by phosphorylation at serine 73 in control (pLKO) and PRL-3 knock down BT-20 cells (shPRL-3) treated with 10 uM IKK inhibitor. F. Fold change of AP-1 activity in PRL-3 knock down BT-20 cells transfected with an AP-1-luciferase reporter plasmid and treated with 10 uM IKK inhibitor. Data represented is the mean ± SD of three independent experiments. * = p-value <0.05 as determined by Student t test.
57 TNFα TNFα TNFα 1 2 ? TNF-R1 TNF-R1 TNF-R1
TRADD TRADD TRADD TRADD TRADD TRADD RIP-1 FADD RIP-1 FADD RIP-1 FADD TRAF2 TRAF2 TRAF2
IKK MAP3K Caspase-8 IKK MAP3K Caspase-8 IKK MAP3K Caspase-8
NF-kB C-Jun Caspases-3/7 NF-kB C-Jun Caspases-3/7 NF-kB C-Jun Caspases-3/7
PRL-3 AP-1 Apoptosis PRL-3 AP-1 Apoptosis PRL-3 AP-1 Apoptosis
Figure 3.9 Comprehensive model of PRL-3 action on the TNF-R1 pathway. The proposed pathways elicited by PRL-3 knock down in TNBC cells and PRL-3 reciprocal effects on TNF-R1 to induce TNBC cell death.
58 Discussion
Cancer cell survival hinges on complex cell-intrinsic and cell-extrinsic stimuli that modulate cell cycle arrest, senescence, apoptosis and immune surveillance of cancer cells
(79). In this chapter, I demonstrate that expression of the metastasis-promoting phosphatase,
PRL-3, is regulated by pro-inflammatory NF-ĸB signaling and that reduced PRL-3 expression results in an unexpected and dynamic feedback loop modulating both the intrinsic cell cycle machinery and extrinsic TNF-R1 signaling pathway to control TNBC cell growth.
The significance of these findings is underscored by a vital unmet need to understand molecular processes that promote the aggressive nature of TNBC, and the need to identify novel mechanisms for enhancing cancer cell death so that new therapeutic strategies may be explored.
In previous chapters, I reported that knock down of basal PRL-3 levels in two TNBC cell lines, BT-20 and MDA-MB-468, resulted in substantial growth inhibition. In this chapter, I report on the discovery that ablation of basal PRL-3 expression in the same TNBC cell lines results in activation of a cell cycle arrest and senescence program that cannot be rescued by mitogenic stimulation, thereby halting TNBC cell line growth potential. In light of these findings, it is reasonable to propose that our observed function for PRL-3, is consistent with direct or indirect regulation of growth response factors in TNBC cells. This hypothesis is also consistent with the expected role for a member of the PRL family of phosphatases, initially discovered as immediate-early growth response genes following mitogenic stimulation in the regenerating liver (33).
To explore an intrinsic mechanism for how PRL-3 knock down halts TNBC cell growth potential and promotes senescence, I profiled the effects of PRL-3 expression on the
59 cell cycle machinery. Interestingly, in normal, non-cancerous MEFs, both overexpression and knock down of PRL-3 have been shown to induce G1 cell cycle arrest (51). These data suggest that maintenance of basal PRL-3 levels is important for proper cell cycle progression in normal cells. In line with the presumed function of PRL-3 in normal cells, my data confirm that knock down of PRL-3 expression in TNBC cancer cells also potently triggers
G1 cell cycle arrest through upregulation of either p16INK4A or p27, depending on the mutational profile of the TNBC cell line. These data shed new light on how TNBC cells utilize alternative CDKIs as a failsafe mechanism to ensure G1 arrest following PRL-3 knock down. One intriguing theory for why TNBC cells may strongly upregulate G1 CDKIs after
PRL-3 knock down, is the activation or loss of negative feedback on an oncogenic signaling pathway, such as PI3K/Akt or MAPK. Indeed, it would be of interest to determine if PRL-3 knock down results in an oncogene-induced senescence program, which may shed additional light on PRL-3 biology. Collectively, these data could be important in the context of a heterogeneous tumor with multiple aberrations impinging on the cell cycle checkpoint machinery and lend further rationale for targeting PRL-3.
In contrast to normal cells, my data indicate that TNBC cells overexpressing PRL-3 have lost the ability to enforce cell cycle arrest or promote G1/S progression (data not shown). In agreement, Basak and colleagues also observed that PRL-3 overexpression in colon carcinoma and osteosarcoma cell lines failed to arrest cells in the cell cycle (51). These data suggest that the phenotypic outcome of PRL-3 overexpression, above basal levels in
TNBC cells, does not promote growth but may execute other oncogenic functions such as driving migration, invasion and metastasis as we and others, have observed. Indeed, more data on these mechanisms are presented in the next chapter.
60 In this chapter, I also sought to discover regulators of PRL-3 expression in TNBC cells lines in order to better understand the context in which cells utilize PRL-3. I identified the NF-ĸB pathway as a positive transcriptional regulator of PRL-3 expression in two TNBC cell lines. This finding builds upon existing data indicating that constitutive activation of the
NF-ĸB pathway promotes oncogenesis in ER-negative and hormone-independent tumors, such as TNBC (52, 53).
In order to understand if PRL-3 participates in a feedback loop with the NF-ĸB pathway, we knocked down PRL-3 and observed that senescence was accompanied by activation of NF-ĸB through TNFα upregulation. Interestingly, PRL-3 was found to activate the NF-ĸB pathway in a TNFα-independent manner by interacting with RAP1 (91), but how
PRL-3 ablation leads to increased TNFα levels in TNBC cells remains poorly understood and warrants further investigation. Interestingly, Amar and colleagues observed that administration of a PRL-3 peptide in mice can significantly attenuate adverse host responses to lipopolysaccharide (LPS) stimulation, providing complete resistance to lethal doses of LPS due to suppression of TNFα production (92). While the exact mechanism was not discovered, this study supports my observations between the inverse expression patterns of PRL-3 and
TNFα. One additional mechanistic theory is that NF-ĸB activity induces the so-called senescence-associated secretory phenotype (SASP), which is characterized by a strong cytokine response that acts in an autocrine feedback loop to achieve growth arrest (79-81, 93,
94). This theory is of particular interest because SASP has also been shown to attract immune cells that eliminate senescent cancer cells (52). Examination of whether SASP ensues following PRL-3 knock down and if this results in immune cell recruitment, activation and/or tumor cell clearance would be a valuable insight to the TNBC research community.
61 Because we observed a substantial increase in TNFα levels following PRL-3 knock down, we also sought to profile the longevity of the senescence program, reasoning that eventually, very high TNFα levels may be capable of activating cell death through an extrinsic, TNF-R1 mediated process. This line of reasoning reveals several potentially interesting paradigms: 1) That the presence of TNFα, at low to moderate levels, confers resistance to apoptosis in the TNBC cell lines and promotes PRL-3 expression and 2) That a molecular sensor exists downstream of the NF-ĸB pathway to regulate signaling flux and control pleiotropism of the TNF-R1 pathway. My findings robustly show that PRL-3 knock down results in TNBC cell sensitization to TNFα-induced apoptosis through TNFR-1.
Additionally, the TNFα-induced extrinsic cell death program cannot be accelerated by blocking the NF-ĸB senescence program that ensues immediately following PRL-3 knock down. This suggests that a sequential, temporally regulated process is required before TNFα can induce apoptosis in TNBC cells after impairing PRL-3 expression.
In summary, my data reveal a new, robust and important role for PRL-3 in TNBC biology. By engaging the adaptive signaling programs of the TNF-R1 pathway, loss of PRL-
3 expression induces cellular senescence and upregulates TNFα to convert senescent TNBC tumor cells into apoptotic cells. This is an impressive dynamic cellular response to reduced expression of a phosphatase, particularly in the chaotic background of genomic and proteomic alterations that plague TNBC cells. As a result, the totality of our data, from chapters I through III, continue to strongly support efforts geared towards targeting PRL-3 in vivo.
62 CHAPTER IV
PRL-3 ENGAGES THE FOCAL ADHESION PATHWAY IN TNBC CELLS TO
ALTER ACTIN STRUCTURE AND SUBSTRATE ADHESION PROPERTIES
CRITICAL FOR CONTROLLING CELL MIGRATION AND INVASION
Introduction
TNBCs have a high propensity to invade through the basement membrane and metastasize to distant visceral organs. To be successful at leaving the primary tumor site, cancer cells must develop an ability to undergo EMT, lose cell-cell contact, migrate, attach to, and invade through the basement membrane. This is a complex process primarily orchestrated through the formation, stabilization and remodeling of focal adhesion (FA) complexes composed of integrins, Src, FAK, ERK and numerous adaptor proteins and downstream pathways that collectively regulate EMT and cell migration and invasion – enabling hallmarks of metastasis. Assembly and disassembly of FA sites spatially and temporally control cell migration and invasion by mediating actin nucleation and assembly and extension, thereby affecting invadopodia formation and cell adhesion to extracellular substrates on which to migrate and invade through. Therefore, the identification of key regulators to FA site assembly/disassembly (either by formation or kinetics) or regulation of its downstream pathways is critical in order to prevent the maturation of these metastatic-enabling properties.
In chapter II, I examined the effect of AMPI-109 treatment and PRL-3 on altering the migratory capacity of TNBC cells and found that AMPI-109 and knock down or catalytic impairment of PRL-3 blocked the ability of TNBC cells to migrate, whereas overexpression 63 of PRL-3 markedly enhanced cell migration (Figure 2.8). Moreover, I presented data on the role of PRL-3 on TNBC cell invasion. These studies suggested that PRL-3 also significantly controls TNBC cell invasiveness through Matrigel in vitro (Figure 2.9). However, a detailed mechanistic analysis of how PRL-3 promotes the motility and invasiveness of TNBC cells is lacking. It is therefore crucial to understand whether AMPI-109 and PRL-3 mechanistically affect FA components or downstream effectors expression levels and/or activity so that a key mechanism of action for AMPI-109, and PRL-3 biology, is identified.
In this chapter, I describe a potential mechanism mediated through the FA pathway, by which PRL-3 appears to control these pro-metastatic phenotypes. Altering PRL-3 expression had a significant impact on protein expression and activation of a number of FA pathway effectors including the proto-oncogene tyrosine-protein kinase, c-src (Src), ERK, and RhoGTPases involved in cytoskeletal restructuring necessary for promoting motility and invasion. In addition, I present data on the role of the matrix metalloproteinase (MMP),
MMP-10, which is up-regulated during PRL-3 overexpression and coincides with TNBC cell attachment to- and degradation of- laminin, the major basement membrane component.
Collectively, this chapter presents newly discovered data on important pathways regulated by PRL-3 that promote cell migration and invasion as precursor events to metastasis – the major driver of TNBC-associated deaths.
Materials and methods
Materials
AMPI-109 was synthesized as previously described (17-19). PRL-3 cDNA expression vector was purchased from Origene (Cat. # SC308739).
64 Plasmids, transfection and viral transduction
PRL-3 cDNA expression vector was purchased from Origene (Cat. # SC308739).
Transfections were carried out using Mirus TransIT LT1 reagent according to manufacturer’s instructions (Mirus Bio). Individual pLKO.1 lentiviral shRNA clones were purchased from the University of Colorado Cancer Center Functional Genomics Shared Resource. The RNAi
Consortium identifiers: TRCN0000010661 (shPRL-3, sh1), TRCN0000355597 (sh2),
TRCN0000378843 (shMMP-10 #1), TRCN0000372935 (shMMP-10 #2). Transduced cells were selected in medium containing 2.5 ug/mL puromycin.
Cell culture and immunoblot analysis
Cell lines were obtained from the University of Colorado Cancer Center Tissue
Culture Shared Resource. All cell lines were authenticated by short tandem repeat DNA profiling performed by the UCCC DNA Sequencing and Analysis Core. Western blot analysis was conducted according to our previous protocol (21). PRL-3 (Cat. # ab82568,
Abcam), p-Src (Y416) (Cat. #2101, Cell Signaling), Src (36D10) (Cat. #2109, Cell
Signaling), p-ERK 1/2 (T202/Y204) (Cat. #4377, Cell Signaling), ERK 1/2 (44/42) (Cat.
#4695, Cell signaling), RhoA (67B9) (Cat. #2117, Cell Signaling), Rac1/2/3 (Cat. #2465,
Cell Signaling), β-actin (Cat. # A5441, Sigma-Aldrich).
Immunofluorescence analysis
Immunofluorescence staining was performed as previously described (77) using green
Alexa Fluor 488 phalloidin staining for F-actin (Cat. #A12379, Thermo Fisher), β-actin
65 antibody for both filamentous and monomer actin forms (Cat. # A5441, Sigma-Aldrich) and nuclear DAPI stain (Cat. #P-36931, Thermo Fisher).
MMP array
A human MMP antibody array kit was purchased from Abcam (Cat. # ab134004). BT-
20 cells were transiently transfected with PRL-3 cDNA expression vector 48 hours prior to cell lysis and work-up according to manufacturers suggested protocol. Membranes were detected by enhanced chemiluminescence (Perkin Elmer) and autoradiography.
Cell adhesion and spreading assay
We utilized the impedance-based xCELLigence Real-Time Cell Analysis system
(ACEA Biosciences) for the detection of BT-20 and SUM159 TNBC cell adhesion and spreading on the following basement membrane substrates: Laminin (Cat. #L4544, Sigma-
Aldrich), Elastin (Cat. #E1625-5G, Sigma-Aldrich), Fibronectin (Cat. #F1141, Signa-
Aldrich) and Collagen (Cat. #C2124, Sigma-Aldrich). Briefly, each substrate was diluted to
10 µg/mL in appropriate TNBC cell media and added to wells on the 96X E-Plate and incubated for 1 hour at 37 C. The coated plates were then washed with PBS and incubated in
0.5% BSA solution in PBS for 20 minutes at 37 C. Wells were washed again with PBS and
5,000 cells were added per well. Cell adhesion and spreading was measured as changes in impedance with the RT-CES system every 3 minutes for 3 hours. The assay expresses impedance in arbitrary cell index (AU) units. The cell index at each time point is defined as
(Rn-Rb)/15; where Rn is the cell-electrode impedance of the well when it contains cells and
Rb is the background impedance of the well with the media alone.
66 Cell invasion assay
Cell invasion was assessed by IncuCyte Zoom Kinetic Live Cell Imaging (Essen
BioScience). 96 well ImageLock Plates (Essen Bioscience) were coated overnight with 100
µg/mL Matrigel Basement Membrane Matrix (Cat. # 356231, Corning). The following day,
Matrigel was aspirated and 25,000 cells were plated per well and allowed to adhere prior to making a scratch with the 96-pin WoundMaker (Essen Bioscience). 50 µL of Matrigel
Matrix (8 mg/mL) was then added to the wells, and covered in 100 µL media containing the appropriate treatments (refer to figure 2.9 legend for details). Migration and invasion were quantified using the “Relative Wound Density” metric generated by IncuCyte software.
Results
PRL-3 amplification correlates with high Src levels in basal human breast cancers
In an effort to identify PRL-3 associated pathways driving gain of function migration and invasion phenotypes, I analyzed the provisional TCGA invasive breast carcinoma dataset and looked for proteins with expression patterns that significantly correlated with PRL-3 amplification. I reasoned that the identification of proteins upregulated with PRL-3 might lead to the discovery of pathways converging with PRL-3 signaling. We identified a number of proteins that were either downregulated or upregulated with PRL-3 amplification in human breast tumors, including Src (Table 4.1).
Src is the prototypic member of a family of non-receptor tyrosine kinases that play critical roles in signal transduction pathways involved with cell proliferation, differentiation, survival, motility and invasion (95-101). Considering the aberrant activation of Src in a number of cancer types including breast cancer, drug therapy aimed at inhibiting Src and/or
67 some of its downstream effectors is considered to be a potentially useful and promising clinical strategy (102-108). Therefore, it was similarly imperative to understand whether
AMPI-109 treatment, or PRL-3 knock down, alter Src expression or activity in TNBC cells.
Table 4.1 Protein expression changes in invasive human breast carcinoma samples exhibiting PRL-3 copy number amplification (>2 copies). Protein expression is indicated by the Z-score determined from reverse phase protein arrays (RPPA). Green indicates co-upregulation of proteins with PRL-3 and red indicates proteins downregulated during PRL-3 amplification.
Protein Expression Z- score (RPPA)
PRL-3 PRL-3 Protein P-value Unaltered Amplified
AR 0.15 -0.47 8.33E-08 GATA3 0.12 -0.37 3.15E-05 ESR1 0.11 -0.32 2.41E-04 PRKAA1 0.09 -0.28 3.48E-04 STK11 0.09 -0.28 0.003 BCL2 0.08 -0.25 0.003 MAPK9 0.08 -0.25 0.004 PGR 0.07 -0.22 0.007 SRC -0.08 0.23 0.006 CCNE1 -0.08 0.25 0.004 SYK -0.09 0.26 0.002 CCNB1 -0.10 0.29 3.35E-04
AMPI-109 treatment and knock down of PRL-3 reduces Src, ERK activation
Src is activated at FA sites following engagement by a variety of different cellular receptors including immune response receptors, integrins, receptor TKs, G protein-coupled receptors as well as cytokine receptors (95-101). The recruitment of Src to integrin complexes at FA sites, for example, is thought to lead to auto-phosphorylation of Src at tyrosine 416 (Y416). Activated Src plays a number of direct and indirect roles at activating
68 downstream pathways, including ERK, for propagating changes in cytoskeletal organization necessary for cell motility. Indeed, studies have demonstrated that Src-mediated migration and invasion of breast cancer and hepatocellular carcinoma cells is controlled through downstream activation of the ERK pathway (109-112). Therefore, we treated BT-20 and
MDA-MB-468 TNBC cells with AMPI-109 or transduced cells with two different shRNAs to PRL-3 (sh1 and sh2) and immunoblotted for Src and ERK activation.
We observed that AMPI-109 significantly blocked activation of Src at tyrosine 416
(Y416) and ERK at threonine 202 and tyrosine 204 (T202/Y204) (Figure 4.1), whereas the previously described PRL-3 inhibitor, BR-1 (PRL-3i) had no effect. We also observed decreases in total levels of Src and ERK with AMPI-109 treatment. Knock down of PRL-3 similarly resulted in deactivation of Src and ERK, but in contrast to AMPI-109 treatment, largely had no effect on total levels of Src or ERK in BT-20 cells (Figure 4.1).
69 BT-20 MDA-MB-468 109 109 - - 3i 3i - - DMSO DMSO PRL EtOH AMPI PRL Sh 2 Sh EtOH AMPI pLKO 1 Sh Sh 2 Sh pLKO 1 Sh
p-Src (Y416)
Src p-ERK (T202/Y204)
ERK
PRL-3
Actin
Figure 4.1 AMPI-109 treatment and PRL-3 knock down inactivates Src, ERK signaling in TNBC cells. Western blot analysis depicting changes in activation of Src and ERK proteins as assessed by phosphorylation of Src at tyrosine 416 and ERK at threonine 202 and tyrosine 204 in vehicle control (EtOH or DMSO), AMPI-109 treated or BR-1 treated (PRL-3i) BT-20 and MDA-MB-468 TNBC cells. Right panels for each cell line depict changes in Src and ERK activation in PRL-3 knock down (sh1 and sh2) or non-silencing control (pLKO) cells.
70 Knock down of PRL-3 or treatment with AMPI-109 decreases RhoA and Rac1/2/3
GTPase levels and associates with remodeling of actin networks in TNBC cells
We next determined whether AMPI-109 and PRL-3 act through common FA site transducers, downstream of Src and ERK. RhoGTPases function as downstream mediators of the FA site and are responsible for executing changes in the actin cytoskeleton that lead to cell restructuring and migration (113-118). We observed decreases in RhoA and Rac1/2/3
RhoGTPase levels following knock down of PRL-3 or AMPI-109 treatment (Figure 4.2A).
We did not observe changes in Cdc42, Rho B or Rho C (data not shown). Additionally, knock down of PRL-3 and AMPI-109 treatment led to an absence of filamentous actin (F- actin) networks as determined by undetectable phalloidin staining (Figure 4.2B). Conversely, overexpression of PRL-3 led to moderate cytoplasmic F-actin staining and pronounced clusters of actin that spatially coincided with regions of membrane extension or protrusion.
Collectively, these data, in addition to the data presented in the aforementioned section, indicate that AMPI-109 and PRL-3 impinge on FA members and RhoGTPase networks to alter physical cellular components, such as actin, that are necessary for cell migration. These data suggest that PRL-3 may also then play a key role in controlling the cells ability to adhere to the extracellular matrix (ECM) – as changes in the internal cytoskeleton would likely influence extracellular adhesion dynamics.
71 109 or transientlywith transfectedor(pC (Ctl)controlor PRL-3 DNA 109 treatcontrolshRNAnon-silcencing(pLKO), PRL-3 (shPRL-3), with antibody to ß-actin. DAPI=ß-actin. to blue. antibody byß-actinas(re phalloidindetermined actinstaining. filamentous PRL withcontrolorvehicle(EtOH)treatedAMPI- nM 100 (shPRL-3), or structure 4.2 Figure -3 protein, RhoA, Rac1/2/3 and Actinprotein,Rac1/2/3and protein RhoA, -3 levels. Rac1/2/3 B A PRL . RhoA Actin A. Knock downKnock treatment of andwithPRL-3 RhoGTPaseAMPI-109 alters Westernblotofcells BT-20 transducedTNBC cont non-silcencing with -3
pLKO
shPRL-3
EtOH
AMPI-109 β- F- β- F- actin actin actin actin B. shPRL Immunofluorescence images of BT-20cellsImmunofluorescence imagesof transduced pLKO d) represents the total actinan using total d) monomer representspool the -3 DNA- 109 for hours.Individualpanels109 6 for represent ed with vehicle with controled(EtOH), AMPI- nM 100 PRL -3. F-actin (green) representsF-actin(green)-3. AMPI EtOH expressionfilamentousand actin rol orPRL- shRNArol (pLKO) to - 109 pcDNA Ctl - PRL -3 3
72 PRL-3 overexpression increases secretion of MMP-10 facilitating cell adhesion, spreading and degradation of laminin for promoting TNBC cell invasion
Understanding tumor cell adhesion to ECM substrates is critical in order to identify how cancer cells breach the basement membrane and invade neighboring normal tissue to metastasize (118). In general, migrating cells rely on attachment to the ECM via formation and maintenance of integrin clusters at FA sites. Collectively, these FA sites mediate attachment to basement membrane substrates such as laminin, elastin, collagen and fibronectin and serve as anchor or attachment points for migrating cells – incorporating extracellular mechanical contact as trigger events for intracellular signaling networks, such as
Src, ERK and RhoGTPases (119). Cellular attachment to specific basement membrane substrates is mediated by complex intracellular and extracellular signals that regulate expression and activation of key FA receptors that preferentially bind to a given substrate
(120-125). Cell attachment to a specific substrate is important for spatial orientation of epithelial cells and serves as a general “sensor” for the cells local environment (126). In order for cells to invade through their specific substrate environment, key MMPs need to be expressed and activated to degrade the substrate (127). Because we have observed that
AMPI-109 and PRL-3 are involved in manipulating intracellular pathways activated by FA contact with the ECM, a key mechanistic question is whether AMPI-109 treatment and PRL-
3 expression similarly alter TNBC cell ability to adhere to and degrade particular ECM substrates.
To examine this, we overexpressed PRL-3 in BT-20 TNBC cells in order to identify key MMPs involved in the PRL-3 cascade. Utilizing a human MMP protein array, we identified a list of MMPs, whose total protein expression levels changed in response to PRL-
73 3 overexpression (Figure 4.3A). This data suggested that MMP-10, the most highly
upregulated MMP with PRL-3 overexpression, may be one key proteinase secreted to act
upon and degrade the ECM. Interestingly, the preferred substrates for MMP-10 include
laminin, fibronectin and elastin (Figure 4.3B) suggesting that BT-20 cells may preferentially
adhere to- and degrade- one of these three ECM components to facilitate cell invasion.
A B Figure 4.3 Overexpression of PRL-3 alters MMP Laminin (LN), levels in BT-20 TNBC fibronectin (FN) and elastin (EL) cells. A. BT-20 cells were transiently transfected with an expression vector encoding PRL-3 (pCDNA- Collagens, Gelatin P3) or scrambled control (pCDNA). Percent change in MMP expression from pCDNA control baseline depicted. B. Table highlighting molecular weight (MW) in kilodaltons (kDa) of key MMPs, location of activity (membrane or secreted) and associated substrates. B
Gene MW (kDa) Location Substrates , N) MMP1 45 secreted Col I, II, III, VII, VIII, X, gelatin L) MMP2 66 secreted Gelatin, Col I, II, III, IV, Vii, X MMP3 45 secreted Col II, IV, IX, X, XI, gelatin MMP7 19 secreted Fibronectin, laminin, Col IV, gelatin MMP8 58 secreted Col I, II, III, VII, VIII, X, aggrecan, gelatin MMP9 86 secreted Gelatin, Col IV, V MMP10 44 secreted Col IV, laminin, fibronectin, elastin MMP11 44 secreted Col IV, fibronectin, laminin, aggrecan MMP12 54 secreted Elastin, fibronectin, Col IV MMP13 53 secreted Col I, II, III, IV, IX, X, XIV, gelatin MMP14 66 membrane Gelatin, fibronectin, laminin MMP15 76 membrane Gelatin, fibronectin, laminin MMP16 69 membrane Gelatin, fibronectin, laminin MMP17 66 membrane Fibrinogen and fibrin
74 We therefore characterized the ability of BT-20 cells to adhere to and spread on each of these substrates. We pre-coated cell culture plates with elastin, laminin and fibronectin matrix and allowed BT-20 cells to adhere to each. We monitored cell attachment and extent of cell spreading on the various substrates by utilizing the real-time dynamic xCELLigence impedance-based adhesion and spreading platform. We observed that parental BT-20 TNBC cells exhibit a diverse adhesion profile to each of the three substrates (Figure 4.4).
Fibronectin was the least preferred substrate for BT-20 attachment, while elastin and laminin exhibited similar degrees of attachment at 2 hours. Interestingly, BT-20 cells contract or
“shrink” when first exposed to laminin as indicated by the negative cell index values between
0.5 and 1 hours, but then recover and spread on laminin at a higher rate relative to elastin
(Figure 4.4) –reaching equivalent impedance to elastin at 2 hours.
Figure 4.4 BT-20 TNBC cells preferentially adhere to- and spread on- elastin and laminin basement membrane substrates. BT-20 cells were plated in suspension and allowed to adhere to elastin (red), laminin (green) or fibronectin (blue). Each substrate was plated at 10 µg/mL and cell impedance was monitored for 2 hours after plating.
We next examined the effects of overexpressing PRL-3 on cell attachment to each of these substrates. By doing so, we reasoned that we may identify differences in binding to specific substrates; reflecting on key aspects of PRL-3 biology. We observed that overexpression of PRL-3 (pcDNA-PRL-3) in BT-20 cells had no effect on the extent of cell
75 spreading when exposed to fibronectin. However, these cells did appear to require more time sensing the local ECM environment before spreading (Figure 4.5A; 1 versus 0.5 hours). In contrast, BT-20 cells overexpressing PRL-3 exhibited a modest and improved ability to adhere to and spread on elastin (Figure 4.5B). Overexpression of PRL-3 in BT-20 cells exposed to laminin, however, significantly enhanced both the rate and magnitude of cell adhesion and spreading (Figure 4.5C), strongly suggesting that PRL-3 functions to enhance
BT-20 cell attachment to laminin, one of the most abundant ECM component in the basal lamina of human tissues.
Figure 4.5 BT-20 TNBC cells overexpressing PRL-3 preferentially adhere to the laminin basement membrane substrate. BT-20 cells transiently transfected with cDNA for PRL-3 (pcDNA- PRL-3; blue) or scrambled DNA (Ctl; red) were plated in suspension and allowed to adhere to A. fibronectin, B. elastin or C. laminin. Each substrate was plated at 10 µg/mL and cell impedance was monitored for 2 hours after plating.
76 To further study the relationship between PRL-3 and TNBC cell attachment to laminin, I chose to knock down PRL-3 in another TNBC cell line, SUM159, to reverse cell attachment to laminin. Interestingly, parental control SUM159 TNBC cells exhibited a similar preference pattern for ECM substrate binding – with collagen type IV being the least preferred substrate and elastin, fibronectin and laminin demonstrating favorable dynamics over the course of 3.5 hours (Figure 4.6).
Figure 4.6 SUM159 TNBC cells preferentially adhere to- and spread on- elastin, fibronectin and laminin basement membrane substrates. SUM159 cells were plated in suspension and allowed to adhere to collagen (blue), elastin (green), fibronectin (brown) or laminin (dark green). Each substrate was plated at 10 µg/mL and cell impedance was monitored for 3.5 hours after plating.
After knocking down PRL-3 in SUM159 cells, I observed no changes in binding profiles to collagen (Figure 4.7A), elastin (Figure 4.7B) or fibronectin (Figure 4.7C) but did observe, as hypothesized, a substantial impairment in the binding affinity and spreading on laminin (Figure 4.7D).
77 AB
CD
Figure 4.7 PRL-3 knock down in SUM159 TNBC cells abrogates cell adhesion and spreading on laminin. SUM159 cells were transduced with non-silencing control shRNA (pLKO) or shRNA to PRL-3 (shPRL-3) and selected for three days prior to being plated in suspension and allowed to adhere to collagen (A), elastin (B), fibronectin (C) or laminin (D). Each substrate was plated at 10 µg/mL and cell impedance was monitored for 3.5 hours after plating.
In chapter II, I demonstrated a role for PRL-3 in enhancing TNBC cell invasiveness through Matrigel (Figure 2.9). Here, we demonstrate a potentially important role for TNBC cell interaction with a key basement membrane substrate, laminin. By knocking down PRL-3 in SUM159 cells, invasive potential could be diminished, and conversely, by overexpressing
PRL-3 in BT-20 cells, invasion rate and absolute degree could be significantly enhanced
(Figure 2.9). However, these data do not take into account the role of MMP-10. Therefore, I chose to examine the effects of knocking down MMP-10 on SUM159 cell invasion. I observed a reduction in the ability of SUM159 cells to invade through Matrigel after knocking down MMP-10 using two different shRNAs (Figure 4.8A). These data are not entirely surprising; One preferred substrate for MMP-10 is laminin and the major component in Matrigel is also laminin - followed by collagen IV, heparin sulfate and proteoglycans.
78 However, since overexpression of PRL-3 upregulates MMP-10 (Figure 4.3), and MMP-10 exerts specific action on degrading laminin, the preferred substrate for TNBC cells expressing PRL-3 (Figures 4.4 - 4.7), a key question remaining is whether knock down of
MMP-10 could reverse the invasive potential driven by PRL-3 overexpression. After confirming that overexpression of PRL-3 enhances BT-20 cell invasion (Figure 4.8B), I chose to overexpress PRL-3 in BT-20 cells with MMP-10 knocked down. As expected, PRL-
3 was not able to drive cell invasion in the absence of MMP-10 (Figure 4.8B), suggesting that PRL-3 driven invasiveness is mediated, at least in part, through MMP-10 activity.
A SUM159
B BT-20
Figure 4.8 PRL-3 driven cell invasion is mediated through MMP-10. A. Real-time kinetic monitoring of invasion through Matrigel in SUM159 cells stably expressing two different shRNA clones to MMP-10 (red and green) or non-silencing control (blue). B. As in (A) using BT-20 control cells (Ctl; blue or Ctl + pLKO; green), transiently transfected to overexpress PRL-3 (pCDNA-PRL-3; red) or overexpressing PRL-3 in BT-20 cells with MMP-10 knock down (pCDNA-PRL-3 + shMMP-10 #1; purple). Invasive score was determined by the ability of cells to invade through a Matrigel wound (Relative Wound Density). * = p-value <0.05 as determined by Student t test on last time-point.
79 Discussion
There is a pressing need to understand which signaling pathways promote TNBC cell migration and invasion so that new strategies aimed at targeting key regulators of these processes may be identified for therapeutic exploitation. PRL-3 was first identified as a metastasis-associated phosphatase in the laboratory of Dr. Bert Vogelstein who showed that
PRL-3 levels were high in metastatic CRC samples, but low in non-metastatic carcinoma or normal epithelial cells (38). Similarly, our IHC evaluation of human breast cancers revealed that PRL-3 expression positively correlates with regional disease and distant metastases in
TNBC. However, a mechanistic explanation of how PRL-3 promotes cell acquisition of migration and invasion programs essential for TNBC metastasis – remained elusive.
In these studies, I describe a relationship between PRL-3 expression and AMPI-109 treatment on activation of the FA Src-ERK pathway. Our data augment existing reports that indicate Src is preferentially localized and active at the cell plasma membrane in TNBCs versus other breast cancer subtypes (119) and that basal and TNBC tumors exhibit sensitivity to dasatinib, an orally bioavailable inhibitor of Src and abl kinases (120). While the direct link between PRL-3 and the Src-ERK pathway remains unknown, these data strongly suggest there may be scientific rationale for exploring combinatorial inhibition of Src and PRL-3 in
TNBC models of migration and invasion to reduce metastatic potential.
We also discovered that changes in PRL-3 expression, or AMPI-109 treatment, exert an influence on protein expression levels of downstream members in the FA pathway, including Rho A and Rac1/2/3. These data are important because RhoGTPases are key players in the control of actin nucleation, filament assembly, stabilization and disassembly – which ultimately, mechanically control the cells ability to migrate and invade by forming cell
80 protrusions or invadopodia. Indeed, coinciding with the decrease in Rho A and Rac1/2/3 expression after PRL-3 knock down was the absence of a detectable filamentous actin network. These data suggest that one terminal effect of PRL-3 knock down or AMPI-109 treatment, mediated through the FA Src-ERK-RhoGTPase pathway may be the disruption of actin assembly, extension or stabilization – thereby eliminating the necessary “hardware” or infrastructure for TNBC cells to migrate and invade.
Because cancer cells must also degrade the local basement membrane to metastasize, we expanded our analysis on the role of PRL-3 in the metastatic cascade by studying the impact PRL-3 expression has on TNBC cell specificity for adhesion to- and degradation of-
ECM substrates. We identified that PRL-3 expression in TNBC cells significantly alters
TNBC cell adhesion and spreading on laminin, which is abundantly expressed in human breast tissue. Diaz et al. observed that out of 212 breast tumors, laminin protein staining was present in a linear pattern surrounding all normal ducts and lobules and around all ducts involved by ductal carcinoma in situ (DCIS). Importantly, extracellular deposition of laminin was also found immediately adjacent to invasive tumor cells and a negative correlation between the extent of extracellular laminin deposition and ER and HER2 positivity was found (121), suggesting that peri-cellular deposition of laminin may be higher in ER and
HER2 negative tumors, such as TNBCs.
Finally, we also observed that TNBC cells were more efficient in invading through
Matrigel, a laminin rich matrix, when PRL-3 was overexpressed and that this process was mediated by upregulation of MMP-10, a key metalloproteinase with specific action on laminin (Figure 4.9). Also, PRL-3 expression had no differential effect on the ability of
TNBC cells to degrade collagen as assessed by an in vitro zymogram assay (data not shown).
81 PRL-3 engages FA pathway
Enhance migration, attach to laminin (cytoskeletal changes)
Pro-MMP-10 Serine proteases in ECM MMP-10
Focal stromal degradation of Laminin
Invasion
Figure 4.9 Cartoon model depicting the role of PRL-3 in the metastatic cascade at the primary TNBC tumor site. PRL-3 expression engages the FA pathway to enhance migration through cytoskeletal changes that lead to cell attachment to laminin and upregulation of MMP-10, leading to focal stromal degradation of the basement membrane to facilitate invasion.
Taken together, these data provide a preliminary framework for further analyses on the exact mechanistic relationships between PRL-3 and the FA pathway (Src-ERK-
RhoGTPase) and actin dynamics, as well as the relationship between PRL-3 expression and the upregulation of MMP-10. Importantly, these data also suggest that PRL-3 ability to invoke a migration and invasion program may be limited by the surrounding levels of extracellular laminin. It is therefore conceivable that areas of “invasive heterogeneity” may exist in a tumor despite the levels of PRL-3 expression – which may not confer an invasive advantage in areas without laminin deposition.
82 CHAPTER V
FUTURE DIRECTIONS
Because metastatic disease is the ultimate cause of mortality from TNBC and is often refractory to standard of care therapy, our ability to treat TNBC is somewhat limited. We have identified a group of pathways and effector molecules regulated by the phosphatase,
PRL-3, that appear to be critical in controlling TNBC growth potential and initiating precursor events, such as migration and invasion, that drive tumor metastasis. Importantly, we have also developed a novel small molecule inhibitor, AMPI-109, that appears to impinge on and terminate activity of these oncogenic pathways. Collectively, the studies outlined herein constitute a significant and timely advancement in our understanding of
TNBC biology and set a solid framework for additional studies geared towards further dissecting PRL-3 biology and investigation of AMPI-109 as a potential new treatment option for patients with TNBC.
We have identified PRL-3 as a potential intracellular target of AMPI-109 as determined by in silico analyses, in vitro enzymatic inhibition assays and in vivo proteasomal degradation experiments. However, the exact mechanistic relationship between AMPI-109 and PRL-3, including the possibility that AMPI-109 directly targets PRL-3, or that AMPI-
109 alters its transcriptional or translational levels or affects PRL-3 protein stability, remains an active area of investigation. Validation studies using biophysical approaches such as
NMR, isothermal titration calorimetry or surface plasmon resonance should be deployed to help definitively answer this question. Unfortunately, initial studies aimed at analyzing
83 AMPI-109 binding to PRL-3 using mass spectrometry (MS) were unsuccessful due to AMPI-
109 solubility issues using the preferred carriers for MS analysis.
Moreover, studies focused on interpreting an in vivo read-out of PRL-3 activity, such as that of a substrate, should be used if possible. However, the lack of bone fide PRL-3 substrates has made this level of in vivo analysis difficult in all aspects of studying of PRL-3 biology to-date, but especially so, the examination of the ability of AMPI-109 to inhibit key de-phosphorylation events mediated by PRL-3 when studying PRL-3 as a putative target of
AMPI-109.
Our genome-wide functional genetic screen initially identified PRL-3 as a modifier of
AMPI-109 activity in the TNBC cell line, BT-20. The identification of a phosphatase as an oncogenic mediator was initially somewhat surprising. Phosphatases are typically regarded to be the “off” switches of molecular signaling pathways that drive cell growth and proliferation. As a result, much work in translational oncology has focused on understanding and targeting kinases, the “on” switches, with less attention being paid to phosphatases.
Recently, however, there have been extensive efforts to elucidate the roles of phosphatases in the development and progression of various cancers, including breast cancer.
While we have focused our attention on PRL-3, the highest ranking hit from the screen, our ability to examine the functional significance of the remaining hits identified in the screen has been limited. Validation of other top hits should be performed, particularly on those that may share signaling or pathway commonalities with known PRL-3 activities in order to solidify the true utility of AMPI-109 as a tool, the concept of the screen and the resulting data as a robust mechanism for identifying aggressive cancer-related markers.
Resource permitting, it would be similarly advantageous to perform another series of screens
84 in additional TNBC cell lines, and importantly, non-TNBC cell lines to compare differential gene expressions across AMPI-109 sensitive and non-sensitive cell lines. One potential avenue to explore this without conducting multiple screens is the utilization of a bioinformatics-guided approach. In this scenario, I propose conducting differential gene expression analysis based on publically available data between resistant and non-resistant cell lines, which can also be followed by a Pearson correlation for strength of gene expression to percent inhibition of AMPI-109. The resulting data could be further overlapped with the results of the BT-20 screen to help identify key pathways, or signaling nodes, that may be significant in the context of the mechanism of action for AMPI-109.
Nonetheless, we were able to identify and validate PRL-3 as one important driver of
TNBC. We observed that PRL-3 levels are enriched in patient tumors with an ER, PR and
HER negative basal gene signature, but importantly, a significant number of patient tumors in the TCGA datasets that were “unclassified” or “non-basal” but still ER, PR and HER negative, exhibited PRL-3 amplification. These data raise the possibility that PRL-3 overexpression may be even higher than what we observed in the basal cohorts of the TCGA breast datasets and individualized examination of the survival outcomes of those patients should also be investigated.
Mechanistically, we discovered that PRL-3 expression in TNBCs is transcriptionally regulated by the oncogenic NF-ĸB pathway and that PRL-3 ablation elicits a TNF-R1 feedback loop that results in TNBC cell cycle arrest and senescence followed by caspase-8 mediated apoptosis. These data are significant because constitutive activation of the NF-ĸB pathway has been reported in TNBCs as mentioned previously, and could be a main driver behind the increased expression of PRL-3 in addition to copy number amplification.
85 However, it remains unclear whether inflammatory or immunomodulatory-like tumors exhibit higher levels of PRL-3 expression versus other subtypes of TNBC and this should similarly be investigated.
Surprisingly, during our TCGA data mining of proteins differentially expressed with
PRL-3 amplification, we discovered that levels of cyclins E1 and B1 were upregulated, which strengthens our evidence in chapter II that PRL-3 first invokes the cell cycle machinery to modulate TNBC growth potential. These data augment our in vitro results and could indicate that co-upregulation is an occurrence in human tumors as well. Other proteins of interest were also identified as being downregulated with PRL-3 amplification including
AR, ER and PR. These data are very intriguing and lead us to hypothesize that a link between
PRL-3 and expression of these three molecular markers, that are considered classifiers of the major forms of breast cancer, may exist. A simple approach to examine this hypothesis could be knocking down PRL-3 to look for cell restoration of AR, ER or PR expression and conversely, overexpression of PRL-3 in an AR, ER or PR positive cell line to look for suppression of gene expression. This theory warrants urgent investigation.
In addition, while we observed that PRL-3 knock down leads to increased levels of secreted TNFα - responsible for maintaining cell senescence and eventual cell death - it is conceivable that other cytokines may also be up-regulated after PRL-3 knock down and identifying this cytokine signature may be important in understanding a system dynamics- based effect of knocking down PRL-3 beyond the intracellular environment. One conceivable advantage for invoking senescence and extrinsic cell death through a cytokine-driven autocrine loop is the enablement of a window of opportunity for immune cell recognition and clearing of the tumor cells. This could take the form of 1) recruiting immune cells to the
86 immediate tumor environment and/or 2) stimulating tumor infiltrating immune cells to clear the senescent cells prior to their ability to reach the TNFα signaling threshold necessary to induce TNF-R1 mediated apoptosis. These concepts should be similarly viewed with caution, as it is unclear whether the TNFα secreted would activate a pro- or anti-tumor response.
Because TNBCs are characterized by a high propensity to metastasize and relapse, we also examined the role of PRL-3 and AMPI-109 on cell migration and invasion. AMPI-109 treatment, downregulation of PRL-3 expression or impairment of PRL-3 activity reduced
TNBC cell migration and invasion by inactivating Src and ERK signaling and downregulating downstream RhoA and Rac1/2/3 GTPase protein levels, altering filamentous actin structures necessary for cell migration and invasion. Conversely, overexpression of
PRL-3 promoted TNBC cell migration and invasion by upregulating a key matrix metalloproteinase, MMP-10, which resulted in increased adherence to- and degradation of- the major basement membrane substrate, laminin.
While several groups have reported on the potential for ezrin to function as a PRL-3 substrate and activate the Src pathway, we did not observe changes in ezrin activity in TNBC cells. Instead, we took a new approach towards discovering a substrate(s) or pathway(s) that could be involved in PRL-3 mediated control of the FA pathway in TNBC cells. To this end,
I compared the gene lists between our functional genomic screen and a large phosphoproteome study conducted by Walls et al (122). Walls and colleagues performed an un-biased mass spectrometry-based screen to identify phosphoproteins exhibiting differences in tyrosine phosphorylation levels after PRL-3 overexpression (122). Walls and colleagues identified, consistent with our data and others, that Src is a major kinase responsible for
“hijacking” cells as a consequence of PRL-3 overexpression. In addition, the group identified
87 integrin receptors, RhoGTPases, PI3K, STATs and ERK as additional key players but no in vivo validation studies were performed. Therefore, our studies are the first to show the functional significance of altering PRL-3 expression on Src, ERK and RhoGTPase levels in vivo with measurable phenotypic endpoints.
Using data from our screen and Walls, I identified a gene, the so-called non-receptor tyrosine kinase activated Cdc42-associated kinase 1 (ACK1 or “TNK2”). TNK2 has gained recent interest as an oncogenic kinase in a number of cancers (123-132). Of note, TNK2 has been reported to modify KDM3A to regulate transcription of the HOXA1 locus to promote growth of tamoxifen-resistant breast cancer cells and has an established role in activating AR in a ligand-independent manner, contributing to hormone insensitive prostate cancer growth
(123). TNK2 has also been shown to inhibit the GTPase activity of Cdc42, which is upstream of Rac1/2/3 and among other roles, is involved in regulating cell adhesion, actin branching and protrusion and filament directionality. Considering these data, it was intriguing to investigate the activation levels of TNK2 in breast cancer.
I therefore decided to immunoblot for phospho-TNK2, pTNK2 (Y284) across multiple breast cancer cell lines. The western blot analysis revealed general antibody non- specificity but with a noticeable single band expression pattern exclusive to the HER2 positive cell line, MDA-MB-453 and all TNBC cell lines examined at approximately 60 kDa
(Figure 5.1A). By using the immunogen sequence for the antibody, we were able to identify one possible candidate for the band of interest as a C-terminally truncated splice variant of
TNK2 (isoform 2; identifier Q07912-2). Coincidentally, this variant form of TNK2 has lost its auto-inhibitory domain, inferring that a constitutively activated form of TNK2 may exist.
88 Importantly, the antibody used in the western blot was designed to recognize a tyrosine residue (Y284) in the N-terminal region of the kinase.
As a result of these findings, and the similarities between the reported functions of
TNK2 and our observed functions of PRL-3, I decided to immunopreciptate TNK2 in two
TNBC cell lines using a phospho- and total antibody specific to TNK2. After confirming successful pull-down of TNK2 using both antibodies, I observed that the predominant form of TNK2 in both cell lines was the activated, phospho-form (Figure 5.1B). Most surprising, however, was the detection of a co-immunoblot signal for PRL-3 after immunoprecipitating
TNK2 (Figure 5.1), indicating PRL-3 and TNK2 may physically interact.
A B -453 -231 -468 MDA-MB-231 MDA-MB-468 MB MB MB - - - -474 -20 -549 T47D MCF10A MCF7 BT MDA IgG IP MDA MDA BT BT IgG IP
FT: TNK2 FL-TNK2
FT: p-TNK2 Sv-TNK2
IP: TNK2
IP: p-TNK2
IB: PRL-3
Figure 5.1 Putative C-terminally truncated TNK2 splice variant expression is unique to TNBCs and interacts with PRL-3 by Co-immunoprecipitation analysis. A. Western blot analysis of TNK2 across multiple breast cancer cell lines using a phospho-TNK2 (Y284) antibody designed to recognize the N-terminal region. MDA-MB-453 cells are HER2 positive and MDA-MB-231, MDA-MB-468, BT-20 and BT-549 cells are TNBC. B. Results of the co-immunoprecipitation experiment after pull-down of total and phospho-TNK2 using respective antibodies in MDA-MB-231 and MDA-MB-468 TNBC cells. IgG depicts non- specific antibody control. IP = immunoprecipitation; FT = flow-through.
89 As a result of these data, it is clearly imperative that further investigations into the relationship between PRL-3 and TNK2 in TNBC be explored as these data suggest 1) the possible presence of a constitutively activated splice variant form a non-receptor tyrosine kinase in TNBC cells, and 2) that TNK2 may physically interact with PRL-3 as a possible substrate. Moving forward, it is critical to understand the functional significance of this variant on TNBC growth, migration and invasion by designing unique shRNAs to recognize full-length and variant forms, and if possible, cloning of the variant for additional phenotypic studies. Finally, a thorough examination on the dynamics between PRL-3 and AMPI-109 on
TNK2 expression and activity is warranted to help identify how PRL-3 and AMPI-109 mechanistically block TNBC growth, migration and invasion.
Considering all of the data outlined above with respect to the role of PRL-3 on enhancing migration and invasion by controlling actin-related pathways, it remains intriguing that PRL-3 is predominately expressed in normal tissues in the heart and skeletal muscle.
One commonality between these tissues is the high degree of cell contractility that is necessary for tissue functionality. Therefore, it is of interest to further investigate the role of
PRL-3 in normal tissues, so that its role in cancer may be better understood. It is conceivable that PRL-3, with its shallow active site pocket, may possess ATPase-like functions, which could aid in the actin-myosin cycle to enhance filament distension and ratcheting – aiding in cell contractility and allowing the ability to migrate.
Finally, in our aforementioned studies we performed histological evaluation of human breast cancers which revealed PRL- 3 expression was significantly associated with the TNBC subtype and correlated positively with the presence of regional and distant metastases, as well as poor relapse free survival rates at one and three years. These data, strongly supporting
90 our in vitro findings, offer a preliminary examination on the potential role of PRL-3 in human tumors. However, in order for PRL-3 blockade and/or AMPI-109 treatment to reach investigational status in humans, more robust mouse xenograft studies need to be completed.
In our initial studies with AMPI-109, we performed a patient-derived TNBC xenograft (PDX) study in NOD/SCID mice. The harvested tumor, an invasive ductal carcinoma with extensive necrosis, was implanted into the mammary fat pad of 19 mice and an intraperitoneal injection of AMPI-109 was administered every other day for three weeks at a potent 1 µg/kg dose. We observed a significant reduction in tumor volume corresponding with a 30% reduction in tumor mass at this low dose of AMPI-109 (Figure 5.2A and B).
A B 2000 30% reduction 1500 *
1000 p = .018 500 Tumor (mg) MassTumor 0 Vehicle AMPI-109 (1 ug/kg)
n = 9 n = 10
Figure 5.2 AMPI-109 treatment is effective at reducing tumor volume and mass in a PDX model of TNBC. A. Tumor volume response at the end of 3 weeks to 1 µg/kg AMPI-109 or sesame seed oil control (vehicle) treatment every other day on a PDX model of TNBC (tumor identifier: PK49). B. Assessment of tumor mass (mg) as in (A).
However, these studies preceded our knowledge of PRL-3 and its potential relationship to AMPI-109. Therefore, additional animal studies employing the experimental model of late metastasis using tail vein injection of TNBC cells with inducible PRL-3 expression, with and without AMPI-109, should be performed.
91 For these experiments, TNBC cells such as MDA-MB-231 cells, that have been engineered to express luciferase for intravital imaging should be utilized. This cell line, for example, is available in our laboratory and has also been engineered to have doxycycline
(dox) inducible expression of PRL-3 and been shown to form lung metastases in the tail vein injection model. Moreover, this cell line has been selected because it expresses relatively low levels of endogenous PRL-3 expression relative to other TNBC cell lines and thus allows for examination of the phenotypic consequences of elevated PRL-3 expression.
At the end of the experiment (6 weeks) major organs (lungs, livers) should be collected at necropsy and fixed in formalin for histochemical and IHC processing. Fixed tissue can be processed for hematoxylin and eosin staining to assess for tumor burden and examination of PRL-3 expression in tumor cells by IHC. Completion of the outlined in vivo studies, in addition to more PDX studies with AMPI-109, will lend further credibility for our hypothesis that AMPI-109 has the potential to be a viable investigational therapeutic candidate for TNBCs.
In summary, we have identified AMPI-109 as a pre-clinical agent active against
TNBC cell lines and have identified the oncogenic phosphatase PRL-3, as a modulator of several oncogenic pathways governing TNBC growth, migration and invasion. Our data warrant further investigation into the mechanistic relationships between AMPI-109 and PRL-
3 and merit additional studies exploring the oncogenic PRL-3 mechanisms of action in
TNBC in an effort to better understand the aggressive biological nature of this disease.
The future direction studies outlined in this chapter will likely foster discovery of additional targets and envisage new scientifically-guided therapeutic strategies for the prevention and treatment of metastatic TNBC. In addition, this work in its entirety, is proof-
92 of-concept that AMPI-109 can be used as a tool to uncover unique drivers of disease progression, such as PRL-3, which we show promotes oncogenic phenotypes in TNBC cells
(Figure 5.3).
PRL-3 blockade restores TNBC sensitivity to TNFα- induced death through TNF-R1 Constitutively active NF-kB in TNBCs
Growth G1 Cell cycle arrest Apoptosis disadvantage and Senescence ? AMPI-109 PRL-3 ACK1 Src, ERK and Actin Reduced RhoGTPases dynamics migration ? ?
Knockdown or MMPs Laminin Reduced Catalytic (MMP-10) attachment and invasion Inactivation degradation
Figure 5.3 Comprehensive model on the role of PRL-3 and AMPI-109 on TNBC growth, migration and invasion.
93 REFERENCES
1. Polyak K & Metzger Filho O (2012) SnapShot: breast cancer. Cancer Cell 22(4):562- 562 e561.
2. Perou CM, et al. (2000) Molecular portraits of human breast tumours. Nature 406(6797):747-752.
3. Berrada N, Delaloge S, & Andre F (2010) Treatment of triple-negative metastatic breast cancer: toward individualized targeted treatments or chemosensitization? Ann Oncol 21 Suppl 7:vii30-35.
4. Dent R, et al. (2007) Triple-negative breast cancer: clinical features and patterns of recurrence. Clin Cancer Res 13(15 Pt 1):4429-4434.
5. Liedtke C, et al. (2008) Response to neoadjuvant therapy and long-term survival in patients with triple-negative breast cancer. J Clin Oncol 26(8):1275-1281.
6. Sorlie T, et al. (2001) Gene expression patterns of breast carcinomas distinguish tumor subclasses with clinical implications. Proc Natl Acad Sci U S A 98(19):10869- 10874.
7. Lehmann BD, et al. (2011) Identification of human triple-negative breast cancer subtypes and preclinical models for selection of targeted therapies. J Clin Invest 121(7):2750-2767.
8. Cancer Genome Atlas N (2012) Comprehensive molecular portraits of human breast tumours. Nature 490(7418):61-70.
9. Villarreal-Garza C, et al. (2014) Platinum-based chemotherapy in triple-negative advanced breast cancer. Breast Cancer Res Treat 146(3):567-572.
10. Petrelli F, et al. (2014) The value of platinum agents as neoadjuvant chemotherapy in triple-negative breast cancers: a systematic review and meta-analysis. Breast Cancer Res Treat 144(2):223-232.
11. Corkery B, Crown J, Clynes M, & O'Donovan N (2009) Epidermal growth factor receptor as a potential therapeutic target in triple-negative breast cancer. Ann Oncol 20(5):862-867.
12. Peddi PF, Ellis MJ, & Ma C (2012) Molecular basis of triple negative breast cancer and implications for therapy. Int J Breast Cancer 2012:217185.
13. Herold CI & Anders CK (2013) New targets for triple-negative breast cancer. Oncology (Williston Park) 27(9):846-854.
94 14. Traina TA, et al. (2015) Abstract 1003: Results from a phase 2 study of enzalutamide (ENZA), an androgen receptor (AR) inhibitor, in advanced AR+ triple-negative breast cancer (TNBC). ASCO Annual Meeting 2015.
15. Giovannucci E, et al. (2005) The epidemiology of vitamin D and cancer incidence and mortality: a review (United States). Cancer Causes Control 16:83-95.
16. Masuda S & Jones G (2006) Promise of vitamin D analogues in the treatment of hyperproliferative conditions. Mol Cancer Ther 5(4):797-808.
17. Ray R, et al. (1996) Affinity labeling of the 1 alpha,25-dihydroxyvitamin D3 receptor. J Biol Chem 271(4):2012-2017.
18. Ray R, Ray S, Rose S, & Holick MF (1993) Photoaffinity labeling of chick intestinal 1 alpha,25-dihydroxyvitamin D3 receptor. Steroids 58(10):462-465.
19. Swamy N, et al. (2000) Molecular modeling, affinity labeling, and site-directed mutagenesis define the key points of interaction between the ligand-binding domain of the vitamin D nuclear receptor and 1 alpha,25-dihydroxyvitamin D3. Biochemistry 39(40):12162-12171.
20. Chen ML, Ray S, Swamy N, Holick MF, & Ray R (1999) Mechanistic studies to evaluate the enhanced antiproliferation of human keratinocytes by 1alpha,25- dihydroxyvitamin D(3)-3-bromoacetate, a covalent modifier of vitamin D receptor, compared to 1alpha,25-dihydroxyvitamin D(3). Arch Biochem Biophys 370(1):34-44.
21. Ray R & Lambert JR (2011) 1alpha,25-Dihydroxyvitamin D3-3beta-bromoacetate, a potential cancer therapeutic agent: synthesis and molecular mechanism of action. Bioorg Med Chem Lett 21(8):2537-2540.
22. Lambert JR, et al. (2010) A vitamin D receptor-alkylating derivative of 1alpha,25- dihydroxyvitamin D3 inhibits growth of human kidney cancer cells and suppresses tumor growth. Cancer Prev Res (Phila) 3(12):1596-1607.
23. Lambert JR, Young CD, Persons KS, & Ray R (2007) Mechanistic and pharmacodynamic studies of a 25-hydroxyvitamin D3 derivative in prostate cancer cells. Biochem Biophys Res Commun 361(1):189-195.
24. Swamy N, Persons KS, Chen TC, & Ray R (2003) 1alpha,25-Dihydroxyvitamin D3- 3beta-(2)-bromoacetate, an affinity labeling derivative of 1alpha,25-dihydroxyvitamin D3 displays strong antiproliferative and cytotoxic behavior in prostate cancer cells. J Cell Biochem 89(5):909-916.
95 25. Persons KS, et al. (2010) Anti-growth effect of 1,25-dihydroxyvitamin D3-3- bromoacetate alone or in combination with 5-amino-imidazole-4-carboxamide-1- beta-4-ribofuranoside in pancreatic cancer cells. Anticancer Res 30(6):1875-1880.
26. Porter CC, et al. (2012) Integrated genomic analyses identify WEE1 as a critical mediator of cell fate and a novel therapeutic target in acute myeloid leukemia. Leukemia 26(6):1266-1276.
27. Kim J & Tan AC (2012) BiNGS!SL-seq: a bioinformatics pipeline for the analysis and interpretation of deep sequencing genome-wide synthetic lethal screen. Methods Mol Biol 802:389-398.
28. Ciriello G, et al. (2015) Comprehensive Molecular Portraits of Invasive Lobular Breast Cancer. Cell 163(2):506-519.
29. Tonks NK (2013) Protein tyrosine phosphatases--from housekeeping enzymes to master regulators of signal transduction. FEBS J 280(2):346-378.
30. Nunes-Xavier CE, Martin-Perez J, Elson A, & Pulido R (2013) Protein tyrosine phosphatases as novel targets in breast cancer therapy. Biochim Biophys Acta 1836(2):211-226.
31. Lee ST, et al. (2013) Protein tyrosine phosphatase UBASH3B is overexpressed in triple-negative breast cancer and promotes invasion and metastasis. Proc Natl Acad Sci U S A 110(27):11121-11126.
32. Farabaugh SM, Micalizzi DS, Jedlicka P, Zhao R, & Ford HL (2012) Eya2 is required to mediate the pro-metastatic functions of Six1 via the induction of TGF-beta signaling, epithelial-mesenchymal transition, and cancer stem cell properties. Oncogene 31(5):552-562.
33. Mohn KL, et al. (1991) The immediate-early growth response in regenerating liver and insulin-stimulated H-35 cells: comparison with serum-stimulated 3T3 cells and identification of 41 novel immediate-early genes. Mol Cell Biol 11(1):381-390.
34. Stephens BJ, Han H, Gokhale V, & Von Hoff DD (2005) PRL phosphatases as potential molecular targets in cancer. Mol Cancer Ther 4(11):1653-1661.
35. Wang J, Kirby CE, & Herbst R (2002) The tyrosine phosphatase PRL-1 localizes to the endoplasmic reticulum and the mitotic spindle and is required for normal mitosis. J Biol Chem 277(48):46659-46668.
36. Han H, et al. (2002) Identification of differentially expressed genes in pancreatic cancer cells using cDNA microarray. Cancer Res 62(10):2890-2896.
96 37. Wang Q, Holmes DI, Powell SM, Lu QL, & Waxman J (2002) Analysis of stromal- epithelial interactions in prostate cancer identifies PTPCAAX2 as a potential oncogene. Cancer Lett 175(1):63-69.
38. Saha S, et al. (2001) A phosphatase associated with metastasis of colorectal cancer. Science 294(5545):1343-1346.
39. Bardelli A, et al. (2003) PRL-3 expression in metastatic cancers. Clin Cancer Res 9(15):5607-5615.
40. Parker BS, et al. (2004) Alterations in vascular gene expression in invasive breast carcinoma. Cancer Res 64(21):7857-7866.
41. Xing X, et al. (2009) Prognostic value of PRL-3 overexpression in early stages of colonic cancer. Histopathology 54(3):309-318.
42. Ooki A, et al. (2009) Phosphatase of regenerating liver-3 as a prognostic biomarker in histologically node-negative gastric cancer. Oncol Rep 21(6):1467-1475.
43. Ren T, et al. (2009) Prognostic significance of phosphatase of regenerating liver-3 expression in ovarian cancer. Pathol Oncol Res 15(4):555-560.
44. Ustaalioglu BB, et al. (2012) Clinical importance of phosphatase of regenerating liver-3 expression in breast cancer. Clin Transl Oncol 14(12):911-922.
45. Wang L, et al. (2006) Overexpression of phosphatase of regenerating liver-3 in breast cancer: association with a poor clinical outcome. Ann Oncol 17(10):1517-1522.
46. Forte E, et al. (2008) Ezrin is a specific and direct target of protein tyrosine phosphatase PRL-3. Biochim Biophys Acta 1783(2):334-344.
47. Peng L, et al. (2006) Identification of integrin alpha1 as an interacting protein of protein tyrosine phosphatase PRL-3. Biochem Biophys Res Commun 342(1):179- 183.
48. Peng L, et al. (2009) PRL-3 promotes the motility, invasion, and metastasis of LoVo colon cancer cells through PRL-3-integrin beta1-ERK1/2 and-MMP2 signaling. Mol Cancer 8:110.
49. Tian W, et al. (2012) Phosphatase of regenerating liver-3 directly interacts with integrin beta1 and regulates its phosphorylation at tyrosine 783. BMC Biochem 13:22.
50. Wang H, et al. (2007) PRL-3 down-regulates PTEN expression and signals through PI3K to promote epithelial-mesenchymal transition. Cancer Res 67(7):2922-2926.
97 51. Basak S, et al. (2008) The metastasis-associated gene Prl-3 is a p53 target involved in cell-cycle regulation. Mol Cell 30(3):303-314.
52. Nakshatri H, Bhat-Nakshatri P, Martin DA, Goulet RJ, Jr., & Sledge GW, Jr. (1997) Constitutive activation of NF-kappaB during progression of breast cancer to hormone-independent growth. Mol Cell Biol 17(7):3629-3639.
53. Klein U & Ghosh S (2011) The Two Faces of NF-kappaB Signaling in Cancer Development and Therapy. Cancer Cell 20(5):556-558.
54. Barnes PJ & Karin M (1997) Nuclear factor-kappaB: a pivotal transcription factor in chronic inflammatory diseases. N Engl J Med 336(15):1066-1071.
55. Li X & Stark GR (2002) NFkappaB-dependent signaling pathways. Exp Hematol 30(4):285-296.
56. Sen R & Baltimore D (1986) Multiple nuclear factors interact with the immunoglobulin enhancer sequences. Cell 46(5):705-716.
57. Ghosh S, May MJ, & Kopp EB (1998) NF-kappa B and Rel proteins: evolutionarily conserved mediators of immune responses. Annu Rev Immunol 16:225-260.
58. Kopp EB & Ghosh S (1995) NF-kappa B and rel proteins in innate immunity. Adv Immunol 58:1-27.
59. Pahl HL (1999) Activators and target genes of Rel/NF-kappaB transcription factors. Oncogene 18(49):6853-6866.
60. Bonizzi G & Karin M (2004) The two NF-kappaB activation pathways and their role in innate and adaptive immunity. Trends Immunol 25(6):280-288.
61. Karin M, Cao Y, Greten FR, & Li ZW (2002) NF-kappaB in cancer: from innocent bystander to major culprit. Nat Rev Cancer 2(4):301-310.
62. Hudis CA & Gianni L (2011) Triple-negative breast cancer: an unmet medical need. Oncologist 16 Suppl 1:1-11.
63. Trott O & Olson AJ (2010) AutoDock Vina: improving the speed and accuracy of docking with a new scoring function, efficient optimization, and multithreading. J Comput Chem 31(2):455-461.
64. Thor AD, Edgerton SM, Liu S, Moore DH, 2nd, & Kwiatkowski DJ (2001) Gelsolin as a negative prognostic factor and effector of motility in erbB-2-positive epidermal growth factor receptor-positive breast cancers. Clin Cancer Res 7(8):2415-2424.
98 65. Skinner AL, Vartia AA, Williams TD, & Laurence JS (2009) Enzyme activity of phosphatase of regenerating liver is controlled by the redox environment and its C- terminal residues. Biochemistry 48(20):4262-4272.
66. Andersen JN, et al. (2001) Structural and evolutionary relationships among protein tyrosine phosphatase domains. Mol Cell Biol 21(21):7117-7136.
67. Diamond RH, Cressman DE, Laz TM, Abrams CS, & Taub R (1994) PRL-1, a unique nuclear protein tyrosine phosphatase, affects cell growth. Mol Cell Biol 14(6):3752-3762.
68. Lu X, et al. (2008) Predicting features of breast cancer with gene expression patterns. Breast Cancer Res Treat 108(2):191-201.
69. Schmidt M, et al. (2008) The humoral immune system has a key prognostic impact in node-negative breast cancer. Cancer Res 68(13):5405-5413.
70. Cates CA, et al. (1996) Prenylation of oncogenic human PTP(CAAX) protein tyrosine phosphatases. Cancer Lett 110(1-2):49-55.
71. Matter WF, et al. (2001) Role of PRL-3, a human muscle-specific tyrosine phosphatase, in angiotensin-II signaling. Biochem Biophys Res Commun 283(5):1061-1068.
72. Werner SR, et al. (2003) Enhanced cell cycle progression and down regulation of p21(Cip1/Waf1) by PRL tyrosine phosphatases. Cancer Lett 202(2):201-211.
73. Ming J, Liu N, Gu Y, Qiu X, & Wang EH (2009) PRL-3 facilitates angiogenesis and metastasis by increasing ERK phosphorylation and up-regulating the levels and activities of Rho-A/C in lung cancer. Pathology 41(2):118-126.
74. Guo K, et al. (2006) PRL-3 initiates tumor angiogenesis by recruiting endothelial cells in vitro and in vivo. Cancer Res 66(19):9625-9635.
75. Den Hollander P, et al. (2014) Abstract 971: Phosphatase PTP4A3 is critical for cell growth of triple-negative breast cancer. AACR Annual Meeting.
76. Lambert JR, et al. (2015) Reduced expression of GDF-15 is associated with atrophic inflammatory lesions of the prostate. Prostate 75(3):255-265.
77. Jacamo R, et al. (2014) Reciprocal leukemia-stroma VCAM-1/VLA-4-dependent activation of NF-kappaB mediates chemoresistance. Blood 123(17):2691-2702.
78. Basak S, et al. (2008) The metastasis-associated gene Prl-3 is a p53 target involved in cell-cycle regulation. Mol Cell 30(3):303-314.
99 79. Chien Y, et al. (2011) Control of the senescence-associated secretory phenotype by NF-kappaB promotes senescence and enhances chemosensitivity. Genes Dev 25(20):2125-2136.
80. Salminen A, Kauppinen A, & Kaarniranta K (2012) Emerging role of NF-kappaB signaling in the induction of senescence-associated secretory phenotype (SASP). Cell Signal 24(4):835-845.
81. Wang J, et al. (2009) RelA/p65 functions to maintain cellular senescence by regulating genomic stability and DNA repair. EMBO Rep 10(11):1272-1278.
82. Wajant H, Pfizenmaier K, & Scheurich P (2003) Tumor necrosis factor signaling. Cell Death Differ 10(1):45-65.
83. McPherson PS, Kay BK, & Hussain NK (2001) Signaling on the endocytic pathway. Traffic 2(6):375-384.
84. Hsu H, Shu HB, Pan MG, & Goeddel DV (1996) TRADD-TRAF2 and TRADD- FADD interactions define two distinct TNF receptor 1 signal transduction pathways. Cell 84(2):299-308.
85. Micheau O & Tschopp J (2003) Induction of TNF receptor I-mediated apoptosis via two sequential signaling complexes. Cell 114(2):181-190.
86. Boldin MP, Goncharov TM, Goltsev YV, & Wallach D (1996) Involvement of MACH, a novel MORT1/FADD-interacting protease, in Fas/APO-1- and TNF receptor-induced cell death. Cell 85(6):803-815.
87. Muzio M, et al. (1996) FLICE, a novel FADD-homologous ICE/CED-3-like protease, is recruited to the CD95 (Fas/APO-1) death--inducing signaling complex. Cell 85(6):817-827.
88. Schneider-Brachert W, et al. (2004) Compartmentalization of TNF receptor 1 signaling: internalized TNF receptosomes as death signaling vesicles. Immunity 21(3):415-428.
89. Schutze S, et al. (1999) Inhibition of receptor internalization by monodansylcadaverine selectively blocks p55 tumor necrosis factor receptor death domain signaling. J Biol Chem 274(15):10203-10212.
90. Kappelmann M, Bosserhoff A, & Kuphal S (2014) AP-1/c-Jun transcription factors: regulation and function in malignant melanoma. Eur J Cell Biol 93(1-2):76-81.
91. Lian S, et al. (2013) PRL-3 activates NF-kappaB signaling pathway by interacting with RAP1. Biochem Biophys Res Commun 430(1):196-201.
100 92. Tang X, Woodward T, & Amar S (2010) A PTP4A3 peptide PIMAP39 modulates TNF-alpha levels and endotoxic shock. J Innate Immun 2(1):43-55.
93. Coppe JP, et al. (2008) Senescence-associated secretory phenotypes reveal cell- nonautonomous functions of oncogenic RAS and the p53 tumor suppressor. PLoS Biol 6(12):2853-2868.
94. Kuilman T, Michaloglou C, Mooi WJ, & Peeper DS (2010) The essence of senescence. Genes Dev 24(22):2463-2479.
95. Irby RB & Yeatman TJ (2000) Role of Src expression and activation in human cancer. Oncogene 19(49):5636-5642.
96. Ishizawar R & Parsons SJ (2004) c-Src and cooperating partners in human cancer. Cancer Cell 6(3):209-214.
97. Bromann PA, Korkaya H, & Courtneidge SA (2004) The interplay between Src family kinases and receptor tyrosine kinases. Oncogene 23(48):7957-7968.
98. Shupnik MA (2004) Crosstalk between steroid receptors and the c-Src-receptor tyrosine kinase pathways: implications for cell proliferation. Oncogene 23(48):7979- 7989.
99. Burgess DJ (2011) Breast cancer: SRC hits the mark. Nat Rev Cancer 11(5):314-315.
100. Elsberger B (2014) Translational evidence on the role of Src kinase and activated Src kinase in invasive breast cancer. Crit Rev Oncol Hematol 89(3):343-351.
101. Guarino M, et al. (2009) Src signaling in cancer invasion. J Cell Physiol n. pag. Web.
102. Kim LC, Song L, & Haura EB (2009) Src kinases as therapeutic targets for cancer. Nat Rev Clin Oncol 6(10):587-595.
103. Wheeler DL, Iida M, & Dunn EF (2009) The role of Src in solid tumors. Oncologist 14(7):667-678.
104. Finn RS (2008) Targeting Src in breast cancer. Ann Oncol 19(8):1379-1386.
105. Martin GS (2006) Fly Src: the Yin and Yang of tumor invasion and tumor suppression. Cancer Cell 9(1):4-6.
106. McCarthy N (2012) Signalling: SRC and survival. Nat Rev Cancer 12(2):80-81.
107. Sgroi DC (2009) Breast cancer SRC activity: bad to the bone. Cancer Cell 16(1):1-2.
101 108. Tan M, et al. (2005) ErbB2 promotes Src synthesis and stability: novel mechanisms of Src activation that confer breast cancer metastasis. Cancer Res 65(5):1858-1867.
109. Zhang L, et al. (2012) C-Src-mediated RANKL-induced breast cancer cell migration by activation of the ERK and Akt pathway. Oncol Lett 3(2):395-400.
110. Sun CK, et al. (2008) Proline-rich tyrosine kinase 2 (Pyk2) promotes proliferation and invasiveness of hepatocellular carcinoma cells through c-Src/ERK activation. Carcinogenesis 29(11):2096-2105.
111. Tang ZN, Zhang F, Tang P, Qi XW, & Jiang J (2011) RANKL-induced migration of MDA-MB-231 human breast cancer cells via Src and MAPK activation. Oncol Rep 26(5):1243-1250.
112. Liu Y, et al. (2011) C-Src mediated RANKL-induced breast cancer cell migration by activation of the ERK and Akt pathway. Oncol Lett n. pag. Web.
113. Hodgson L (2014) Forword. Regulation of RhoGTPases in motility: A fine balancing act. Cell Adh Migr 8(6):525.
114. Jimenez-Dalmaroni, et al. (2011) Control of cortical actin assembly and cadherin- catenin localization by RhoGTPases. Dev Biol 356.1:137.
115. Lane J, et al. (2011) HGF and RhoGTPases in cancer cell motility. Current Signal Transduction Therapy 6.2:173-179.
116. Lazer G & Katzav S (2011) Guanine nucleotide exchange factors for RhoGTPases: good therapeutic targets for cancer therapy? Cell Signal 23(6):969-979.
117. Moissoglu K, McRoberts KS, Meier JA, Theodorescu D, & Schwartz MA (2009) Rho GDP dissociation inhibitor 2 suppresses metastasis via unconventional regulation of RhoGTPases. Cancer Res 69(7):2838-2844.
118. Mokady D & Meiri D (2015) RhoGTPases - A novel link between cytoskeleton organization and cisplatin resistance. Drug Resist Updat 19:22-32.
119. Tryfonopoulos D, et al. (2011) Src: a potential target for the treatment of triple- negative breast cancer. Ann Oncol 22(10):2234-2240.
120. Finn RS, et al. (2007) Dasatinib, an orally active small molecule inhibitor of both the src and abl kinases, selectively inhibits growth of basal-type/"triple-negative" breast cancer cell lines growing in vitro. Breast Cancer Res Treat 105(3):319-326.
121. Diaz LK, et al. (2005) B4 integrin subunit gene expression correlates with tumor size and nuclear grade in early breast cancer. Modern Pathology 18:1165-1175.
102 122. Walls CD, et al. (2013) Phosphatase of regenerating liver 3 (PRL3) provokes a tyrosine phosphoproteome to drive prometastatic signal transduction. Mol Cell Proteomics 12(12):3759-3777.
123. Mahajan K & Mahajan NP (2015) ACK1/TNK2 tyrosine kinase: molecular signaling and evolving role in cancers. Oncogene 34(32):4162-4167.
124. Gelman IH (2014) Androgen receptor activation in castration-recurrent prostate cancer: the role of Src-family and Ack1 tyrosine kinases. Int J Biol Sci 10(6):620- 626.
125. Howlin J, Rosenkvist J, & Andersson T (2008) TNK2 preserves epidermal growth factor receptor expression on the cell surface and enhances migration and invasion of human breast cancer cells. Breast Cancer Res 10(2):R36.
126. Mahajan K & Mahajan NP (2014) ACK1/TNK2 tyrosine kinase: An emerging target for cancer therapeutics. AACR Educational Book 2014.1:97-102.
127. Mahajan K, et al. (2012) Ack1-mediated androgen receptor phosphorylation modulates radiation resistance in castration-resistant prostate cancer. J Biol Chem 287(26):22112-22122.
128. Mahajan K & Mahajan NP (2013) ACK1 tyrosine kinase: targeted inhibition to block cancer cell proliferation. Cancer Lett 338(2):185-192.
129. Mahajan NP, et al. (2007) Activated Cdc42-associated kinase Ack1 promotes prostate cancer progression via androgen receptor tyrosine phosphorylation. Proc Natl Acad Sci U S A 104(20):8438-8443.
130. Mahajan NP, Whang YE, Mohler JL, & Earp HS (2005) Activated tyrosine kinase Ack1 promotes prostate tumorigenesis: role of Ack1 in polyubiquitination of tumor suppressor Wwox. Cancer Res 65(22):10514-10523.
131. Mahajan K & Mahajan NP (2012) PI3K-independent AKT activation in cancers: a treasure trove for novel therapeutics. J Cell Physiol 227(9):3178-3184.
132. Xu SH, et al. (2015) ACK1 promotes gastric cancer epithelial-mesenchymal transition and metastasis through AKT-POU2F1-ECD signalling. J Pathol 236(2):175-185.
103