Role of PTPRT in Obesity and Its Substrate Paxillin Tyrosine-88 in Colorectal Cancer� ��

ROLE OF PTPRT IN OBESITY AND ITS SUBSTRATE PAXILLIN

TYROSINE-88 IN COLORECTAL CANCER

ANTHONY SCOTT

Submitted in partial fulfillment of the requirements for the degree of

Doctor of Philosophy

Department of Genetics and Genome Sciences

CASE WESTERN RESERVE UNIVERSITY

January 2015 CASE WESTERN RESERVE UNIVERSITY

SCHOOL OF GRADUATE STUDIES

We hereby approve the thesis/dissertation of

Anthony Scott

candidate for the degree of Doctor of Philosophy*.

Committee Chair

Hua Lou

Committee Member (Advisor)

Zhenghe John Wang

Committee Member

Sanford Markowitz

Committee Member

Clark Distelhorst

Committee Member

Alex Huang

Date of Defense

7/17/2014

*We also certify that written approval has been obtained for any proprietary material contained therein.

2 Table of Contents

List of Tables 8

List of Figures 9

Acknowledgements 11

List of Abbreviations 12

Abstract 15

Chapter 1: Background and Significance 17

Colorectal Cancer 18

Etiology, staging, and therapy 18

Molecular basis of colorectal cancer 20

Genomics of colorectal cancer 21

PTPRT: Structure and Function 24

PTPRT’s Domain Structure 25

PTPRT’s role in neurological development 26

PTPRT as a tumor suppressor 26

Other Type IIB RPTPs as tumor suppressors 28

3 PTPRT’s role in cell adhesion 29

PTPRT in cancer signaling pathways 30

Paxillin, a substrate of PTPRT 31

Summary of PTPRT structure and function 34

Obesity 36

Diet and obesity 36

Obesity-related pathology 37

Central nervous system and obesity 38

Summary 39

Chapter 2: Identification of paxillin Y88 as a direct target of Src kinase 41

Abstract 42

Introduction 43

Results 45

Src regulates paxillin Y88 phosphorylation 45

Src directly phosphorylates paxillin Y88 47

PY88 paxillin regulates PI3-Kinase activation 48

4 PY88 paxillin does not correlate with clinical 50

characteristics of CRC

Paxillin serves as a predictive factor for dasatinib sensitivity 52

Discussion 53

Materials and Methods 56

Chapter 3: PTPRT regulates high-fat diet-induced 59

obesity and insulin resistance

Abstract 60

Introduction 61

Results 63

Ptprt-/- mice are resistant to high-fat diet-induced obesity 63

Ptprt-/- mice have less body fat by percentage than 64

wild type littermates

Ptprt-/- reduces food intake 66

Ptprt-/- mice have reduced energy expenditure than WT mice 68

Ptprt-/- mice resist high-fat diet-induced hyperglycemia 69

and insulin resistance

5 Metabolic differences between Ptprt+/+ and Ptprt-/- littermates 72

Phospho-STAT3 increased in the hypothalamus of Ptprt-/- mice 72

Discussion 74

Materials and Methods 76

Chapter 4: Discussion and Future Directions 81

Summary 82

PTPRT and obesity: Future Directions 82

What is the role of PTPRT in glucose and lipid metabolism? 82

How does PTPRT affect the relationship between NPY, 90

stress and obesity?

What are other implications of hypothalamic 91

phospho-STAT3?

What is the in vivo role of PTPRT? 94

Paxillin and cancer: Future Directions 96

Why is pY88 paxillin important therapeutically? 96

How does pY88 paxillin affect p130cas phosphorylation? 98

6 What are alternative pathways through which 100

pY88 paxillin can act?

Bibliography 103

7 List of Tables

Chapter 2

Table 2-1. Clinical characteristics of colorectal carcinoma samples. 50

8 List of Figures

Chapter 1

Figure 1-1. A multiple-stage colorectal progression model. 21

Figure 1-2. Effect of PTPRT mutations on its function. 24

Figure 1-3. Schematic of neurohormonal/nutrient feedback loop. 39

Chapter 2

Figure 2-1. Src phosphorylates paxillin at Y88 in cell lines. 46

Figure 2-2. Src directly phosphorylates paxillin at Y88. 47

Figure 2-3. PY88 paxillin regulates p130CAS-p85 interaction. 49

Figure 2-4. PY88 paxillin is upregulated in colorectal cancer tissues 51

Figure 2-5. PY88 paxillin levels predict sensitivity to dasatinib. 53

Chapter 3

Figure 3-1. PTPRT KO mice demonstrate slightly lower body weight than wild type littermates on normal chow diet. 63

Figure 3-2. PTPRT KO mice are resistant to high-fat diet- induced body composition changes. 65

Figure 3-3. PTPRT KO mice eat less but do not absorb dietary fats differently. 67

9 Figure 3-4. PTPRT KO mice do not have different circulating levels of leptin. 67

Figure 3-5. PTPRT KO mice have decreased NPY levels before high-fat diet. 67

Figure 3-6. PTPRT KO mice utilize more glucose and expend less energy than wild type mice. 69

Figure 3-7. PTPRT KO mice have less insulin resistance than wild type mice after high-fat diet. 71

Figure 3-8. PTPRT KO mice demonstrate better insulin regulation than wild type mice. 73

Figure 3-9. PTPRT KO mice have different blood chemistry values after high-fat diet. 73

Figure 3-10. PTPRT regulates STAT3 phosphorylation in mouse hypothalamus. 74

Chapter 4

Figure 4-1. Summary of PTPRT KO phenotypes after

14 weeks on a high-fat diet. 83

Figure 4-2. Diagram of hepatic insulin signaling. 85

Figure 4-3. Effect of PTPRT KO on end organs. 95

Figure 4-4. Extended pY88 paxillin signaling model. 101

10 Acknowledgements

First, I would like to acknowledge the hard work that my adviser Dr. Zhenghe

John Wang put in to train me. Specifically, a key scientific skill I needed to work on when I entered the lab was the organization and presentation of my research. Regardless of whether it was a lab journal club or departmental seminar, Dr. Wang emphasized the importance of presenting research in a succinct and easily digestible manner, poring over my and my labmates' presentations to continually improve them. I attribute my three

CWRU Biomedical Graduate Student Symposium poster awards, which are indicative of my progress to these ends, to his diligence in training me.

I also would like to thank all of my labmates past and present, especially Dr.

Yujun Hao and Dr. Xiujing Feng who were with me for all four years. They were both very gracious in providing me with help in lab and Xiujing particularly with conducting mouse studies for our obesity work.

Next, I very much appreciate the time put in by my thesis committee, Dr. Hua

Lou (chair), Dr. Sanford Markowitz, Dr. Clark Distelhorst and Dr. Alex Huang in preparing me to produce this body of work.

Also, I would like to thank my parents and my girlfriend, Dr. Katie Linder, for their support as well.

11 List of Abbreviations

ACTH Adrenocorticotropic Hormone

AOM Azoxymethane

ATP Adenosine Triphosphate

BMI Body Mass Index

CRC Colorectal Carcinoma

Csk C-terminal Src Kinase

DMEM Dulbecco's Modified Essential Medium

DNA Deoxyribonucleic Acid

EDTA Ethylenediaminetetraacetic acid

EE Energy Expenditure

FAP Familial Adenomatous Polyposis

FBS Fetal Bovine Serum

FFA/NEFA Free Fatty Acids/Non-Essential Fatty Acids

FN Fibronectin

G6P Glucose 6-Phosphate

GAPDH Glyceraldehyde 3-phosphate dehydrogenase

GTT Glucose Tolerance Test

HNPCC Hereditary Non-Polyposis Colorectcal Carcinoma

HOMA-IR Homestatic Model Assessment-Insulin Resistance

HRP Horseradish Peroxidase

Ig Immunoglobulin

IPTG Isopropyl β-D-1-thiogalactopyranoside

12 ITT Insulin Tolerance Test

MAM Meprin/A5-protein/PTPmu

NPY Neuropeptide Y p Phospho- p130cas p130-Crk Associated Substrate

PAGE Polyacrylamide Gel Electrophoresis

PDK1 Phosphoinositide-Dependent Kinase

PI3K Phosphoinositide 3-kinase

PIP2 Phosphatidylinositol-4,5-bisphosphate

PIP3 Phosphatidylinositol-3,4,5-triphosphate

PKC Protein Kinase C

POMC Proopiomelanocortin

PTK Protein Tyrosine Kinase

PTP Protein Tyrosine Phosphatase

PTPN14 Protein Tyrosine Phosphatase, Non-receptor type 14

PTPRT Protein Tyrosine Phosphatase, Receptor type, T

PTPRT KO Protein Tyrosine Phosphatase, Receptor type, T homozygous knockout mice PXN Paxillin pY Phosphotyrosine pY88 Phosphotyrosine-88 (paxillin)

RIPA Radioimuniprecipitation assay

RPTP Receptor-type Protein Tyrosine Phosphatase

RQ Respiratory Quotient

13 SDS Sodium Dodecyl Sulfate

SFK Src Family Kinase

SH Src Homology

SHP-2 Src Homology-2 domain containing phosphatase-2 shRNA Short Hairpin RNA

SNP Single Nucleotide Polymorphism

STAT3 Signal Transducer and Activator of Transcription 3

VLDL Very Low Density Lipoprotein

Y88F Paxillin Tyrosine-88-to-Phenylalanine knock-in mutation

14 Role of PTPRT in Obesity and its Substrate Paxillin Tyrosine-88 in Colorectal Cancer

Abstract

ANTHONY SCOTT

Regulation of protein tyrosine phosphorylation is important in maintaining appropriate cellular homeostasis. Accordingly, protein tyrosine phosphatases are frequently mutated in cancer. The most commonly mutated protein tyrosine phosphatase in colorectal cancer is PTPRT. Follow-up studies validated it as a tumor suppressor, especially through its activity on its substrates STAT3 phosphotyrosine-705 and paxillin phosphotyrosine-88.

While the latter substrate is well characterized, further study is needed into pY88 paxillin.

Since previous studies show that PTPRT is inactivated in colorectal carcinoma, understanding what kinase directly phosphorylates its substrates is an important question to investigate. Here, we show that paxillin Y88 is directly targeted by Src kinase for phosphorylation. Consequently, this finding has implications for cells that express high levels of pY88 paxillin, as they become sensitive to dasatinib treatment. Moreover, although prior work demonstrated that pY88 paxillin impacts Akt signaling, how this signal was transduced was not immediately clear. We show that pY88 paxillin promotes

15 interaction between p130Cas and the p85alpha regulatory subunit of PI3K. Therefore,

we shed further light into how PTPRT affects colorectal cancer tumorigenesis. Another

important aspect of this and many other cancers is the role of obesity in tumor

development. One of PTPRT’s substrates, pY705 STAT3, plays a crucial role in energy

homeostasis. Like PTPRT, there are other neuronally-expressed tyrosine phosphatases that target STAT3 for dephosphorylation. These proteins have a dramatic impact on obesity development, making PTPRT a good candidate for a similar study.

Appropriately, the loss of PTPRT via a mouse knockout has a dramatic impact on the development of obesity. In the genetic background of a mouse strain that is sensitive to high-fat diet-induced obesity, PTPRT knockout mice resist the development of obesity by

decreased food intake. Accordingly, they also avoid many deleterious side effects, such

as increased adiposity and increased peripheral insulin resistance. Consistent with

PTPRT’s tyrosine phosphatase function, we attribute these findings to a change in

phospho-STAT3 in the mouse hypothalamus.

Chapter 1

Background and Significance

A modified version of this chapter was previously published as:

Scott A, Wang Z. Tumour suppressor function of protein tyrosine phosphatase receptor-

T. Biosci Rep. 2011. 31(5):303-7.

Scott, A, Wang Z. “Colon cancer” in Systems Biology of Cancer, ed. Sam Thiagalingam.

New York: Cambridge University Press. In press.

COLORECTAL CANCER

The colon and rectum are the final part of the human digestive tract. Cancers derived from these parts are called colorectal cancers. Colorectal cancer is the third most commonly diagnosed cancer with over one million cases reported world-wide annually

(Markowitz and Bertagnolli 2009). Although advances in early detection and chemotherapy have decreased the mortality of this disease, colorectal cancer remains to be the second leading cause of cancer death in the United States with an estimated 57,000 deaths annually (Markowitz and Bertagnolli 2009).

Etiology, staging, screening and therapy

Etiology

Most colorectal cancers are derived from predominantly epithelial cells (i.e., adenomas or adenocarcinomas) (Kinzler and Vogelstein 1996). The causes of colorectal cancer can largely be divided into environmental or molecular factors. Diet has long been hypothesized to affect one’s risk of developing the disease, with red meats and specific fats thought to be deleterious. Similarly, obese individuals with a high BMI and sedentary lifestyle are at increased risk for CRC. (Compton 2008)

Familial vs sporadic colorectal cancer

About 15-20% of colorectal cancer patients are positive for a family history of the disease, and a sizable proportion of these cases can be classified as having a familial

18 syndrome (Kinzler and Vogelstein 1996) (Lynch and de la Chapelle 2003). Taken as a whole, these relatively rare syndromes account for 5% of all colorectal cancer cases overall, but only account for a quarter of the patients that have a first-degree relative with cancer. Therefore, further studies are currently being done by clinical geneticists to determine the inheritance patterns of the many familial cancers that do not fit into an identified colorectal cancer syndrome.

Staging

Colorectal cancer progression occurs in well-recognized clinical and histopathological stages beginning with dysplastic epithelium and ending with metastatic disease.

Pathologically, colorectal cancers are often divided in four stages. Stage I cancers are confined to the muscular layer. Stage II cancers breach the muscular layer but are limited to the adjacent soft tissues. Cancers metastatic to regional lymph nodes are designated as

Stage III. Cancers metastatic to distant organ sites such as the liver are designated as

Stage IV. Surgical resection is a highly effective treatment for early stage colon cancers, providing cure rates of over 90% in Stage I and over 80% in Stage II disease. The presence of nodal involvement (Stage III) predicts for 60% likelihood for recurrence, with virtually all recurrences being ultimately lethal. Treatment of these high-risk individuals with post-operative chemotherapy reduces the recurrence rate to 40%, increasing overall survival to 60%. On the contrary, the five year survival rate for the last

Stage IV colorectal cancer patients is only 8%. (Compton 2008)

Therapy

Surgical resection is the standard of care for most cases of disease, and adjuvant therapy

is often indicated upon nodal involvement and/or metastasis. 5-Fluorouracil, an anti-

metabolite, is used alongside leuvocorin for adjuvant therapy; this modality may be used

in combination with oxaliplatin (a platinum-based compound used to disrupt DNA) or

irinotecan (an alkaloid that stabilizes topoisomerase I and causes DNA double-stranded

breaks). Compounds that target specific molecules, such as bevacizumab or cetuximab

(VEGF and EGFR, respectively) are also possible therapies depending on the tumor

profile. However, cytotoxic therapy is still the standard of care for most colorectal cancer

cases, highlighting the importance of further study into targeted compounds for specific

oncogenes. (Compton 2008)

Molecular basis of colorectal cancer

It is now widely accepted that cancer is a genetic disease, mediated by alterations in

specific genes (Vogelstein and Kinzler 2004). Tumor progression form normal epithelia to the final metastatic disease occurs over decades and is accompanied by a series of genetic changes that affect at least one oncogene and several tumor suppressor genes

(Figure 1-1). One of the first events in this process is inactivation of the APC tumor suppressor gene or activation of its downstream target β-catenin (Vogelstein and Kinzler

2004), resulting in the formation of pre-cancerous adenomas. Mutations that activate

KRAS or BRAF oncogenes are thought to follow APC mutations and are observed in at least half of adenomas greater than 1 cm in size as well as in carcinomas (Rajagopalan,

Figure 1-1. A multiple-stage colorectal progression model.

Colorectal cancer progression form normal epithelia to the final metastatic disease occurs over decades and is accompanied by a series of genetic changes affecting multiple pathways. Genetic instabilities including microsatellite instability and chromosomal instability are thought to be the driving forces of tumor progression.

Bardelli et al. 2002). As shown in Figure 1-1, subsequent waves of clonal expansion

driven by mutations in the PIK3CA/PTEN (Samuels, Wang et al. 2004); TP53/BAX

(Vogelstein, Lane et al. 2000); TGF-βRII/SMAD (Markowitz and Bertagnolli 2009) pathways and probably other pathways are responsible for the transition of a large adenoma (benign) to an early carcinoma (malignant). In addition to mutation in oncogenes and tumor suppressor genes, genomic instability and aberrant DNA methylation also drive the development of colorectal cancers.

Genomics of colorectal cancer

Recently, genome-wide mutational analyses of colorectal cancers present a complex picture of the heterogeneous molecular pathogenesis of this disease (Sjoblom, Jones et al.

2006, Wood, Parsons et al. 2007). Despite the presence of a few frequently mutated genes

(mountains) including APC, p53, RAS and PIK3CA, most of the genes are mutated in

<5% of colorectal cancers (hills). The mutated genes in any given two tumors only overlap to a small extent, indicating a complex genetic heterogeneity of colorectal cancers.

The initial high-throughput mutational analyses of colorectal cancer focused on gene families that encode protein tyrosine kinases (PTKs), protein tyrosine phosphatases

(PTPs) and lipid kinases (Bardelli, Parsons et al. 2003, Samuels, Wang et al. 2004, Wang,

Shen et al. 2004). Seven PTKs (MLK4, NTRK3, FES, KDR, EPHA3, NTRK2 and

GUCY2F) were identified to be mutated somatically in colorectal cancers. These tumor- derived mutations are likely to be activation mutations (Bardelli, Parsons et al. 2003).

Similarly, Samuels et al. discovered that PIK3CA was mutated in over 30% of colorectal cancers in the mutational analysis of PI3K gene family (Samuels, Wang et al. 2004). In subsequent studies, PIK3CA was shown to be frequently mutated in hepatocellular carcinomas (Lee, Soung et al. 2005), endometrial carcinomas (Oda, Stokoe et al. 2005), breast carcinomas (Bachman, Argani et al. 2004), gastric carcinomas (Li, Wong et al.

2005), ovarian carcinomas (Campbell, Russell et al. 2004) and a smaller fraction of lung carcinomas (Samuels, Wang et al. 2004), medullablastomas and anaplastic astrocytomas

(Broderick, Di et al. 2004). Therefore, PIK3CA is one of the most frequently mutated oncogenes in human cancer. Follow-up studies have verified that these mutations will also aberrantly activate this protein’s lipid kinase activity, precipitating tumor formation.

In contrast to tyrosine and lipid kinases, mutations found in PTPs are frequently

22 inactivation mutations (Wang, Shen et al. 2004). Among the six mutated PTPs identified

(PTPRT, PTPRG, PTPRF, PTPN13, PTPN14 and PTPN3), PTPRT is the most frequently mutated one.

These studies underscore the importance of protein tyrosine phosphorylation in cancer development (Cohen 2002, Arena, Benvenuti et al. 2005, Ostman, Hellberg et al. 2006,

Tonks 2006, Julien, Dube et al. 2011). Although protein tyrosine kinases have been long known to serve as oncogenes, such as JAK, BCR-Abl and Src (Blume-Jensen and Hunter

2001, Yeatman 2004, Motiwala and Jacob 2006), the role that their cellular counterpart – protein tyrosine phosphatases (PTPs) – could play in cancer progression had not been extensively explored. The protein tyrosine phosphatase superfamily, which regulates a multitude of signaling pathways, is divided into either “receptor-like” (RPTP) or “non- transmembrane” (Alonso, Sasin et al. 2004, Tonks 2006, Barr, Ugochukwu et al. 2009).

Among them, protein tyrosine phosphatase receptor-T (PTPRT), also known as PTPρ, is the most frequently mutated tyrosine phosphatase.

PTPRT: Structure and Function

Figure 1-2. Effect of PTPRT mutations on its function.

In normal cells, PTPRT engages in homophilic cell-cell adhesion via its extracellular domains and dephosphorylates target proteins -- paxillin and STAT3 -- via its intracellular phosphatase domain. In cancer cells, mutations affect its cell-cell adhesion capability and phosphatase activity, resulting in failed cell contacts and aberrantly phosphorylated proteins. (Adapted from Ostman, Hellberg et al. 2006)

PTPRT’s Domain Structure

PTPRT, encoded by one of the largest genes in the human genome (Scherer 2010), belongs to the type IIB RPTP subfamily which consists of PTPRK, PTPRM, PTPRT and

PTPRU (PTPRU is frequently referred to as PCP-2). All four type IIB RPTPs share common domain architecture: an extracelluar domain, a transmembrane domain, a juxtamembrane region and two phosphatase domains. The extracellular domains of type

II RPTPs have high sequence identities (Tonks 2006): all consist of a MAM

(meprin/A5/PTPμ) domain, an Ig domain and four fibronectin type III (FNIII) repeats.

The MAM domain is suggested to play a role in protein dimerization. The Ig domain is a disulfide structure that is found in many cell surface proteins and has been shown to mediate homophilic and heterophilic interactions between cell adhesion molecules. The

FNIII motif was originally identified in the extracellular matrix protein fibronectin and later found to be present in many cell adhesion molecules (Tonks 2006). Intracellular domains include the juxtamembrane domain, which has similarity to cadherin domains

(Besco, Popesco et al. 2004, Tonks 2006), and two phosphatase domains, D1 and D2.

The juxtamembrane domain was specifically identified as being responsible for many of the sequence differences between the four type IIb family members, providing specificity in signal transduction (Besco, Popesco et al. 2004) and potentially needed for catalytic activity in some family members (Besco, Frostholm et al. 2001, Brady-Kalnay 2001).

The phosphatase domain D1 is actively responsible for the phosphatase activity of type

IIb members, while the “pseudophosphatase” domain D2, though catalytically inactive, is thought to be important for regulation (Brady-Kalnay 2001).

PTPRT’s role in neurological development

McAndrew et al. discovered PTPRT due in part to the role that many other RPTPs play

in neural physiology (McAndrew, Frostholm et al. 1998, Ensslen-Craig and Brady-

Kalnay 2004). These studies demonstrated that expression of PTPRT was largely limited to the brain and spinal cord. In situ hybridization indicated the hippocampus and dentate gyrus as specific areas of the brain, especially in postmigratory cells instead of the dividing and mobile neurons of prenatal nervous system development. As such, PTPRT is generally more involved in synaptic regulation rather than nervous system growth and development. Specifically, PTPRT fosters synaptic formation by interacting with a series of synaptic-related proteins (including neuroligin family members and neurexin) and binding with other PTPRT molecules on other cells; both interactions occur via PTPRT’s extracellular domains (Lim, Kwon et al. 2009). The intracellular domain is also crucial for synaptic regulation. When Fyn phosphorylates PTPRT in its catalytic domain, it inactivates interactions between PTPRT molecules on the same cell, a loss of phosphatase activity and ultimately an inability for PTPRT to interact with its target proteins important for synaptic transmission (Lim, Kwon et al. 2009). Genome-wide studies into autism (Wittkowski, Sonakya et al. 2014) and amyotrophic lateral sclerosis

(Dunckley, Huentelman et al. 2007, Carter, Anklesaria et al. 2009) associate PTPRT with these neurological diseases, further connecting it with nervous system function.

PTPRT as a tumor suppressor

PTPRT is mutated in head & neck, colon, lung, stomach and skin (melanoma) cancers.

Several lines of genetic evidence suggest that PTPRT normally functions as a tumor suppressor gene: First, a significant fraction of the mutations found in tumor samples are nonsense and insertion and deletion mutations, resulting in premature truncation of

PTPRT and therefore loss of its function (Wang, Shen et al. 2004). Second, consistent with Knudson’s Two-Hit Hypothesis for the inactivation of tumor suppressor activity

(Knudson 1971, Jacob and Motiwala 2005, Motiwala and Jacob 2006), approximately one-third of the tumors with PTPRT mutations harbor two different mutations in one tumor (Wang, Shen et al. 2004). Third, many of the tumor-derived missense mutations in either the D1 or the D2 phosphatase domains lead to reduced phosphatase activity of

PTPRT (Wang, Shen et al. 2004, Lui, Peyser et al. 2014); similarly, such mutations in the extracellular domain result in defective cell adhesion (Yu, Becka et al. 2008, Zhang,

Becka et al. 2009). Fourth, overexpression of PTPRT activity inhibited cell growth, acting as a putative tumor suppressor in cancer cell culture (Wang, Shen et al. 2004).

Fifth, a screen for methylated genes in colorectal cancer recovered PTPRT as being heavily methylated in tumor samples versus normal controls (Laczmanska, Karpinski et al. 2013). Most importantly, knockout mice lacking both alleles of PTPRT, grown in a strain normally resistant to azoxymethane (AOM)-induced colon cancer, became highly susceptible to AOM-induced adenomas (Zhao, Zhang et al. 2010). Another large-scale screen also independently identified PTPRT as a putative tumor suppressor: after mobilizing transposable elements in mouse somatic cells, the authors documented which insertion sites led to tumor formation, ultimately including PTPRT in its list (Collier,

Carlson et al. 2005).

Other Type IIB RPTPs as tumor suppressors

PTPRT is not the only type IIB family member implicated as a tumor suppressor, as

PTPRK, PTPRM and PTPRU are also implicated in tumorigenesis (Jacob and Motiwala

2005). Chromosomal deletions recovered from tumors (Zhang, Siebert et al. 1998) and

microsatellite studies (Nakamura, Kishi et al. 2003) implicate PTPRK as a possible tumor

suppressor, especially in Hodgkin Lymphoma related to the Epstein-Barr Virus (Flavell,

Baumforth et al. 2008). This tumor suppressor role can be extended to skin cancers, as

transcriptional levels of PTPRK and PTPRU were decreased in melanoma tumor samples

according to data generated from RT-PCR (McArdle, Rafferty et al. 2001) and microarray (McArdle, Rafferty et al. 2005). PTPRU might also be a tumor suppressor based on its mutational status in colorectal carcinoma (Sjoblom, Jones et al. 2006), hypothesized to act through Beta-catenin antagonism (Yan, Yang et al. 2006).

PTPRM is of particular interest because of its potential utility as a cancer diagnostic tool.

Its expression is decreased in glioblastoma multiforme (GBM), causing these cells to increase their migration capacity (Burgoyne, Palomo et al. 2009, Burgoyne, Phillips-

Mason et al. 2009). While other RPTPs are known to be cleaved proteolytically, PTPRM cleavage products were identified especially in GBM cell lines. One of these fragments translocates to the nucleus and affects GBM cell proliferation and movement (Burgoyne,

Phillips-Mason et al. 2009). A subsequent study demonstrated that these fragments are

useful diagnostically, as probes against the extracellular fragment were used to detect

GBM tumors in vivo (Burden-Gulley, Gates et al. 2010).

PTPRT’s role in cell adhesion

Similar to PTPRM and PTPRK, the extracellular domain of PTPRT also mediated

homophilic cell-cell adhesion (Yu, Becka et al. 2008). This homophilic interaction is very

specific, as the extracellular part of PTPRT does not interact with PTPRM, PTPRK or

PTPRU (Yu, Becka et al. 2008, Zhang, Becka et al. 2009). The specificity of the homophilic cell-cell adhesion seems to be determined by the MAM and Ig domains, because chimeric PTPRT that substituted its own MAM or Ig domains for those of

PTPRU would not aggregate with cells expressing wild-type PTPRT (Zhang, Becka et al.

2009). Consistent with the notion that the MAM and Ig domains are critical for PTPRT’s cell adhesion function, deletion mutants devoid of either domain abolish its ability to mediate cell aggregation (Yu, Becka et al. 2008). Surprisingly, the four FN III domains of

PTPRT are also required for its cell-cell adhesion function (Zhang, Becka et al. 2009).

Importantly, all the tumor-derived mutations located in the extracelluar domains of

PTPRT show defects in cell-cell adhesion (Yu, Becka et al. 2008, Zhang, Becka et al.

2009).

The crystal structures of PTPRM revealed some clues to how PTPRT mutations affect cell-cell adhesion, as these mutated residues in PTPRT were able to be correlated to key amino acids in PTPRM (Aricescu, Hon et al. 2006, Aricescu, Siebold et al. 2007).

According to these data, some of these mutations were predicted to disturb the structure

as a whole, affecting overall folding patterns, the stability of the extracellular portion of

the protein or the ability for the protein to correctly localize to the membrane (Aricescu,

Siebold et al. 2007). However, one mutated residue in the Ig domain was thought to be specifically responsible for changing the capability for protein-protein interaction,

underscoring the importance of this function in normal physiology (Aricescu, Hon et al.

2006). Moreover, crystal structures also inform the importance of the ectodomain in

dimerization as well as ensuring correct spacing between adjacent cells at adherens

junctions so that interactions with substrates can occur properly (Aricescu, Hon et al.

2006).

Consistent with the role of PTPRT in cell-cell adhesion, PTPRT interacts with E-cadherin

(Besco, Hooft van Huijsduijnen et al. 2006). This interaction leads to E-cadherin

dephosphorylation, affecting the stability of junctional complexes (Besco, Hooft van

Huijsduijnen et al. 2006).

PTPRT in cancer signaling pathways

Identification of the substrates of PTPRT is an important step to understand its tumor suppressor function. Since PTPRT’s phosphatase activity is often attenuated in cancers, proteins that retain phosphorylated tyrosine residues when PTPRT is mutated in the crucial D1 and D2 domains may be involved in tumorigenesis. Studies to this end ultimately identified proteins that had been previously speculated to be proto-oncogenes,

30 specifically STAT3 and paxillin (Zhang, Guo et al. 2007, Zhao, Zhang et al. 2010). A phosphopeptide profile was collected from wild type cancer cells and cells from the same lineage that express either the intracellular fragment of PTPRT (which contain both of the phosphatase domains needed for enzymatic activity) or the extracellular fragment (which lacks such activity). Those phosphopeptides that still had tyrosine phosphorylation in the absence of a functional copy of PTPRT warranted further review as possible substrates.

One such substrate is STAT3 phosphotyrosine-705, a modification associated with

STAT3 activation (Zhang, Guo et al. 2007). Substrate-trapping assays reinforced this finding. Then, PTPRT’s phosphatase activity was crucial in regulating STAT3’s translocation into the nucleus and activation of target genes, many of which are involved in cell survival and tumorigenesis (especially BCL-XL and SOCS3) (Zhang, Guo et al.

2007). Inactivating phosphatase domain mutations identified in head and neck cancers confirm these findings, as they were associated with increased STAT3 pY705 as well

(Lui, Peyser et al. 2014).

Paxillin, a substrate of PTPRT

Another target for PTPRT dephosphorylation found by proteomic methods is paxillin tyrosine-88. This finding complements the canonical function of the type IIB family members, which often interact with adherens junction proteins (Besco, Hooft van

Huijsduijnen et al. 2006), as paxillin is known to be involved in cell adhesion pathways and is also heavily tyrosine-phosphorylated. Paxillin was initially discovered via a screen

for Src kinase substrates, pulled down by phosphotyrosine immunoprecipitation (Glenney

and Zokas 1989). A follow-up study gave this protein its name – an adaptation of the

Latin “paxillus,” meaning peg – when it became apparent that this protein localized to

contacts with the extracellular matrix and anchored the cell’s cytoskeleton (Turner,

Glenney et al. 1990). Eventually, its role expanded to include regulating signal transduction from these focal adhesions through protein-protein interactions (Deakin and

Turner 2008). Important paxillin interactants include focal adhesion kinase, vinculin and small GTPase regulatory factors. Taken as a whole, these interacting proteins allow paxillin to coordinate focal adhesion turnover with the actin cytoskeleton (Nakamura,

Yano et al. 2000), permitting the cell to adequately migrate in a coordinated fashion.

As a testament to the method by which it was discovered, paxillin is heavily regulated by tyrosine phosphorylation. These phosphorylation events occur in response to extracellular cues such as growth factors and interaction with extracellular matrix components like fibronectin (Bellis, Miller et al. 1995, Schaller and Schaefer 2001).

While there are multiple possible Src phospho-tyrosine substrates on paxillin, much attention has been paid to two specific tyrosines, tyrosines 31 and 118 (Schaller and

Schaefer 2001, Deakin and Turner 2008). When phosphorylated, the latter site serves as a canonical substrate for Src Homology 2-domain containing proteins, reinforcing its role in Src signal transduction. One such SH2-domain containing protein is Crk. Phospho- paxillin interacts with this protein involved in Src signaling to impact cell migration capacity (Petit, Boyer et al. 2000, Deakin and Turner 2008).

This protein’s importance in cellular processes is encapsulated in the mouse paxillin

knockout, as these homozygous knockout mice die in utero before day nine (Hagel,

George et al. 2002). From a developmental standpoint, these mice have defects in heart formation, consistent with its effect on cell migration. At a cellular level, fibroblasts

from these mice did not have correct associations between focal adhesion proteins like

FAK and p130cas, consistent with paxillin’s role as an adapter protein.

Paxillin’s normal function ultimately informs its role in tumorigenic processes (Deakin,

Pignatelli et al. 2012). For example, its importance in migration and cell shape can

impact cancer cell metastasis (Young, Liu et al. 2003, Azuma, Tanaka et al. 2005, Deakin

and Turner 2011, Deakin, Pignatelli et al. 2012, Deramaudt, Dujardin et al. 2014).

Additionally, paxillin’s ability to act as a docking protein is easily dysregulated as well,

encouraging anchorage-independent FAK phosphorylation and ras transformation of

mouse embryonic fibroblasts (Wade, Brimer et al. 2011) and cell survival pathways

(Sorenson 2004, Wu, Wu et al. 2013). Accordingly, paxillin is overexpressed and/or

hyperphoshorylated in metastatic lung cancers (Jagadeeswaran, Surawska et al. 2008),

ultimately correlating with overall and recurrence-free survival (Chen and Gallo 2012).

Many of these studies involve paxillin as an effector for Src-mediated transformation

(Yeatman 2004). However, an equally important and complementary avenue of research

is understanding phospho-paxillin as a target of phosphatases.

PTPRT is responsible for dephosphorylating paxillin tyrosine-88 (Zhao, Zhang et al.

2010). The importance of Y88 phosphorylation was especially seen in xenograft models,

where cancer cells with mutated Y88 residues (a “knock-in” cell line with phenylalanine substituted for tyrosine-88 – Y88F – prevented phosphorylation from occurring) were not able to grow xenografts. Similarly, the Y88F mutants in cell lines experienced reduction in anchorage-independent growth and cell migration, as well as changes in Akt, SHP-2 and p130CAS signaling due to alterations in their respective phosphorylation levels.

Consistent with the premise that pY88 paxillin plays an oncogenic role, this phosphorylated residue is up-regulated in majority of human colon cancer tissues in comparison with the matched adjacent normal tissues (Zhao, Zhang et al. 2010).

However, the corresponding kinase had not been identified. Given the difficulty of drugging a mutated or silenced tumor suppressor phosphatase, identifying the kinase responsible for pY88 paxillin is of great therapeutic import.

Summary of PTPRT structure and function

Recent studies have elaborated on the role that Receptor Protein Tyrosine Phosphatase type IIB family members play in suppressing tumor growth, especially PTPRT.

Although originally discovered as a primarily neurological protein, PTPRT has been shown to play integral roles in cell adhesion and intracellular signaling, due in part to extensive studies done on its extracellular and intracellular domains, respectively. When this normal physiologic role becomes abrogated, however, cellular transformation often occurs, demonstrating its importance as a tumor suppressor (Figure 1-2). Further studies

need to be done on PTPRT and its type IIB family members to identify additional

phosphatase targets to obtain a better understanding of progression to cancer.

Additionally, not much is known about the in vivo purpose of PTPRT, as it has been

largely studied in the context of in vitro studies and cancer. Therefore, the PTPRT

knockout mouse discussed here provides an invaluable tool in understanding the normal

function of this protein, which may ultimately inform its role as a tumor suppressor.

The importance of PTPRT in cancer development underscores its role in pathologic

cellular signaling. Similar to cancer, obesity is caused by dysregulation of important

cellular pathways involved in growth and homeostasis (Schwartz and Porte 2005,

Gerozissis 2008, Khandekar, Cohen et al. 2011). For example, inflammatory, insulin and adipokine pathways all overlap in their stimulating effects on obesity and cancer

(Khandekar, Cohen et al. 2011). Accordingly, cancer is frequently caused by obesity

(Calle and Thun 2004), and these diseases become intertwined as diets high in fat often cause cancer (Khandekar, Cohen et al. 2011). Much like in cancer, many phosphatases have been implicated in the development of obesity and diabetes, especially via their antagonism of tyrosine phosphorylation related to insulin signaling (Andersen, Jansen et al. 2004, Bento, Palmer et al. 2004). In order to better understand its function in normal settings as well as in cancer, PTPRT should be studied in the context of obesity.

OBESITY

Nearly one-third of adults in the United States classify as obese according to the body- mass index calculation of body habitus (Ogden et al. 2014). Since it increases mortality due to a variety of causes, such as cancer, diabetes and cardiovascular disease

(Berrington de Gonzalez, Hartge et al. 2010), understanding and treating obesity represents an unmet clinical need.

Diet and obesity

Diet plays a crucial role in obesity, specifically those high in fats and sugar (Bray 2010,

Ahima 2011). These begin at the biochemical level – fat is an efficient form of energy.

Dietary fat affects organism physiology as a whole by influencing satiety signals.

Additionally, fats are stored instead of used as an energy source after feeding (Jequier

2002, Hariri and Thibault 2010). While decreases in dietary fat are largely associated with reductions in body weight, a recent meta-analysis demonstrates that the extent of this effect can vary from study to study (Hooper, Abdelhamid et al. 2012). The strong genetic contribution to diet-induced obesity partially explains why some individuals will not lose weight on low-fat diets (Maes, Neale et al. 1997, Barsh, Farooqi et al. 2000).

High-fat diets can also blunt the central nervous response to insulin (Gerozissis 2008).

From an experimental perspective, the complicated relationship between genetics and high fat diet seen in humans is best modeled in laboratory mice. Though certain mouse strains largely resist diet-induced obesity, others – namely C57BL/6 – quickly gain weight when fed a high-fat diet (West, Boozer et al. 1992).

Obesity-related pathology

Adipocytes, which increase in size and number during obesity, actively engage in cell

signaling; as more of these cells exist in the body, they can dramatically influence a

variety of metabolic processes by disturbing normal homeostatic signals. Specifically,

adipocytes elaborate hormones and cytokines that contribute to insulin resistance and/or

abnormal liver function and secrete various paracrine-acting factors that can cause hypertension (Haslam and James 2005, Rosen and Spiegelman 2014). Excessive fat consumption also results in cellular lipotoxicity: cells can not adequately process the excess of dietary fatty acids, and their accumulation can dysregulate intracellular signaling pathways (Berne 2003). As such, understanding the mechanisms behind the development of obesity can help decrease the morbidity and mortality from this disease.

Obesity induced by a high-fat diet often results in hyperglycemia and type-II diabetes

mellitus (Surwit, Seldin et al. 1991, Schreyer, Wilson et al. 1998, Rossmeisl, Rim et al.

2003, Hussain 2010). Increased blood glucose has many negative effects on the body, ranging from macrovascular complications like atherosclerosis to microvascular dysfunction that causes nephropathy, retinopathy and neuropathy (Brownlee 2001,

Groop, Forsblom et al. 2005). High levels of glucose can modify proteins, known as advanced glycosylation endproducts, leading to protein dysfunction and reactive oxygen species production. Pathology also stems from abnormal PKC activation, as excessive amounts of glucose inadvertently increase diacylglycerol, which in turn leads to increases

in active PKC. As such, hyperglycemia – often secondary to increased insulin resistance

– causes many of the complications of obesity.

Central nervous system and obesity

The central nervous system regulates energy homeostasis at multiple levels, and therefore

contributes greatly to the development of obesity (Figure 1-3). Circulating nutrients and

hormones act at the hypothalamus, where neurons integrate a panoply of systemic signals

regarding energy status, especially the hormones leptin and insulin (Elmquist and Marcus

2003, Gerozissis 2008, Morton and Schwartz 2011) (Schwartz and Porte 2005). These

neurons then signal to higher order brain structures and to the peripheral nervous system

to affect a variety of processes, such as food seeking behavior, energy expenditure

(Haque, Minokoshi et al. 1999) and glucose homeostasis (Elmquist and Marcus 2003).

Environmental and genetic factors can dysregulate leptin and insulin signaling in the

brain; they pervert a negative feedback loop between body fat stores and the brain that

ensures homeostasis (Morton and Schwartz 2011), and instead act as a positive feedback

loop leading to the development of obesity and diabetes (Gerozissis 2008). Moreover, dysfunctional stress responses and obesity share a molecular underpinning, further underscoring the link between the nervous system and metabolic disturbances (Bornstein,

Schuppenies et al. 2006).

Figure 1-3. Schematic of neurohormonal/nutrient feedback loop.

Energy stores and blood glucose levels result in hormonal and nutrient changes that are registered by the hypothalamus. The hypothalamus can then signal to other parts of the brain that either change behavior, regulate the autonomic nervous system and/or secrete neurohormonal signals internally to and externally from the brain. Ultimately, these signals will affect glucose handling, food intake and energy expenditure, affecting energy stores and blood glucose. Obesity acts by blunting these signals at a variety of levels, preventing this feedback loop from adequately closing. (Adapted from Elmquist and Marcus 2003)

Summary

Aberrant protein tyrosine phosphorylation has a dramatic effect on cellular signaling, and

therefore impacts diseases such as obesity and cancer. Accordingly, PTPRT was

recovered as the most commonly mutated phosphatase in colorectal cancer. Follow-up studies validated it as a bona-fide tumor suppressor and identified its substrates, most

notably paxillin at tyrosine-88. However, since this phosphorylation event has been

poorly studied, the kinase that targets it is unknown. Further study is also needed to

understand how it impacts certain tumorigenic pathways, especially Akt. Therefore, we

address the question of what kinase phosphorylates paxillin at Y88, and how does it

affect Akt signaling.

Additionally, in vivo functions of PTPRT other than as a tumor suppressor are not well

known. Expression studies and in vitro experiments indicate this protein may regulate

neuronal function. Accordingly, many of the pathways in the central nervous system

control of obesity are mediated by phosphatases. This finding is validated in vivo, as

mouse knockouts of neuronally-expressed phosphatases, namely TCPTP become resistant to high-fat diet-induced obesity and its attendant metabolic changes. Therefore, we interrogate the role of PTPRT in obesity and glucose regulation.

Chapter 2

Identification of paxillin tyrosine-88 as a direct target of Src kinase

A modified version of this chapter was previously published as:

Scott A, Zhao Y, Zhang P, Hao Y, Wang Z, Willis J. Identification of paxillin tyrosine-88 as a direct target of Src kinase. Submitted.

Abstract

Paxillin is an adapter protein involved in tumor migration, invasion and metastasis via

mediation of growth factor signaling and adhesion to the extracellular matrix. Recent

studies from our laboratory demonstrate that paxillin tyrosine-88 phosphorylation (pY88) plays a critical role in colorectal tumorigenesis. An important downstream effector of pY88 paxillin is the oncogene Akt. Here, we show that pY88 paxillin impacts the Akt pathway by regulating the interaction between p130cas and the p85 regulatory subunit of

PI3-Kinase. Additionally, while pY88 paxillin is a substrate of the tumor suppressor phosphatase PTPRT, the corresponding kinase has not been identified. In this study, we demonstrate that the oncogenic kinase Src directly phosphorylates paxillin at Y88.

Moreover, colorectal cancer cells that express high levels of pY88 paxillin are sensitive to dasatinib treatment, suggesting that pY88 paxillin may serve as a predictive biomarker for Src family kinase inhibitors.

Introduction

Protein tyrosine phosphorylation plays a critical role in virtually all human cellular

processes that involve tumorigenesis. Addition of phosphate to tyrosine of proteins by

protein tyrosine kinases (PTKs) and its removal by protein tyrosine phosphatases (PTPs)

are part of a collaborative process that controls functionally significant phospho-

modification of important proteins that determine cancer cell phenotypes (Tonks 2006).

While the roles of PTKs in tumorigenesis are well documented, increasing evidence suggests that PTPs also play critical roles in cancer development. In the first comprehensive mutational analysis of the entire PTP family in human cancers, we identified protein tyrosine phosphatase receptor-T (PTPRT) as the most frequently mutated PTP in colorectal cancers (CRC) (Wang, Shen et al. 2004). A follow-up mutational analysis of PTP family genes in head and neck squamous cell carcinomas

(HNSCC) also found that PTPRT is frequently mutated in this tumor type (Lui, Peyser et al. 2014). Moreover, recent whole-exome sequencing of various human cancers revealed that PTPRT is frequently mutated in bladder (Guo, Sun et al. 2013), endometrial (Le

Gallo, O'Hara et al. 2012), esophageal (Dulak, Stojanov et al. 2013), lung (Liu, Morrison et al. 2012), skin (Wei, Walia et al. 2011) and stomach cancers (Zang, Cutcutache et al.

2012) as well. The tumor-derived mutations of PTPRT distribute throughout the protein.

A portion of PTPRT mutations are nonsense, insertion and deletion mutations resulting in premature truncation of the protein, whereas most of PTPRT mutations are missense. Our studies showed that the missense mutations located in the phosphatase domains reduce the phosphatase activity of PTPRT and that extracellular domain mutations impair its

cell-cell adhesion ability (Yu, Becka et al. 2008, Zhang, Becka et al. 2009, Becka, Zhang

et al. 2010). Together, those data suggest PTPRT may normally function as a tumor

suppressor. This notion is further supported by the study showing that PTPRT knockout

mice are susceptible to azoxymethane (AOM)-induced colon tumor formation (Zhao,

Zhang et al. 2010).

Given the compelling evidence showing that PTPRT functions as a tumor suppressor, it is

important to understand the signaling pathways regulated by this phosphatase. To this

end, using a phospho-proteomics approach, we identified and validated paxillin and

STAT3 as the substrates of PTPRT. While PTPRT dephosphorylates the well-studied

Y705 residue of STAT3 (Zhang, Guo et al. 2008), the PTPRT target site on paxillin is a

previously uncharacterized tyrosine-88 residue. We showed that PTPRT directly

dephosphorylates pY88 paxillin (Zhao, Zhang et al. 2010). Paxillin is an adapter protein

involved in tumor growth, focal adhesion turnover, cell migration and metastasis

(Schaller, Hildebrand et al. 1999, Webb, Donais et al. 2004, Deakin and Turner 2008,

Deakin, Pignatelli et al. 2012, Reynolds, Kanner et al. 2013). It becomes tyrosine

phosphorylated in response to external stimuli such as growth factors and adhesion to the

extracellular matrix (Bellis, Miller et al. 1995, Schaller and Schaefer 2001, Young, Liu et

al. 2003). Paxillin Y88 is phosphorylated in response to PDGF-AA stimulation. Paxillin

Y88 phosphorylation in turn transduces a signal to activate Akt in CRC cells (Zhao,

Zhang et al. 2010). Using a knock-in method that targets endogenous gene loci (Zhang,

Guo et al. 2008), we changed the Y88 residue to phenylalanine (Y88F), a mutant that can not be phosphorylated at this site. Paxillin Y88F mutant CRC cells displayed attenuated

tumorigenicity, forming fewer colonies in soft agar and failing to form xenograft tumors

in nude mice. Moreover, compared to matched normal colon tissues, pY88 paxillin is up-

regulated in a majority of human colon cancer specimens. In aggregate, these data

suggest that pY88 paxillin plays an oncogenic role in colorectal tumorigenesis.

Given that PTPRT is a tumor suppressor, it would be challenging to target PTPRT

mutations in cancers. However, the corresponding kinases of PTPRT substrates are

potential therapeutic targets for cancer patients harboring PTPRT mutations. Here, we

demonstrate that Src kinase directly phosphorylates Y88 paxillin. Furthermore, CRC cell

lines expressing higher levels of pY88 paxillin are more sensitive to killing by Src

inhibitor dasatinib, suggesting that pY88 paxillin may be exploited as a predictive

biomarker for drugs targeting Src.

Results

Src regulates paxillin Y88 phosphorylation

PTPRT dephosphorylates phospho-paxillin at Y88, but the kinase directly responsible for this phosphorylation event was not identified. Kinase prediction software suggests that the oncogenic kinase Src and PDGF receptor are potential kinases that phosphorylate the

paxillin Y88 residue (Xue, Ren et al. 2008). Therefore, we set out to test whether inhibitors against the two tyrosine kinases could reduce pY88 paxillin levels in CRC cells. As shown in Figure 2-1A, saracatinib, a Src-specific inhibitor, decreased pY88 paxillin phosphorylation in a dose dependent manner in CRC cells. A PDGFR inhibitor,

45 imatinib (Gleevec), had no effect on paxillin Y88 phosphorylation (Figure 2-1A).

Consistently, Src overexpression causes pY88 paxillin levels to increase dramatically

(Figure 2-1B). Conversely, knocking down Src using shRNA attenuated pY88 paxillin levels (Figure 2-1C). Taken together, our data suggest that Src is a kinase that phosphorylates paxillin Y88.

Figure 2-1. Src phosphorylates paxillin at Y88 in cell lines. A) Small molecule inhibitors of Src decrease pY88 paxillin levels. CRC cells were treated with saracatinib or Gleevec (control). Cells were lysed and the lysates blotted with the indicated antibodies. B) Overexpression of Src increases pY88 paxillin levels. 293AAV cells were transfected with a vector overexpressing Src or an empty control vector. Cells were lysed and the lysates blotted with the indicated antibodies. C and D) Knockdown of Src decreases pY88 paxillin levels. CRC cells were transfected with a shRNA vector against Src or a scrambled control vector. Cells were lysed and the lysates blotted with the indicated antibodies. Effect of knockdown on pY88 paxillin was quantified in (D).

Src directly phosphorylates paxillin at Y88

Figure 2-2. Src directly phosphorylates paxillin at Y88. A) Recombinant paxillin was expressed in bacteria, purified and then incubated with recombinant Src, ATP and a buffer. The reaction mixture was blotted with the indicated antibodies. B) Recombinant paxillin was purified and resolved using SDS-PAGE. Equal amounts of paxillin and paxillin Y88F protein was mixed with recombinant Src for the in vitro kinase assay.

To demonstrate that Src phosphorylates paxillin Y88 directly, we set out to determine if

Src phosphorylates paxillin Y88 in vitro. A 6xHistidine tagged N-terminus paxillin fragment (amino acids 1 to 165) was expressed in E. coli and purified to near homogeneity (Figure 2-2). The purified recombinant paxillin proteins were incubated

with purified and active recombinant Src proteins in the presence of ATP. Phospho-Y88

paxillin was detected by Western blot analyses using a phospho-specific antibody

recognizing pY88 paxillin. Figure 2-2 shows that Src indeed phosphorylated paxillin at the Y88 residue in vitro. In contrast, the anti-pY88 paxillin antibody failed to detect any signal in a control experiment using a recombinant paxillin Y88F mutant protein as a

47 substrate, demonstrating the specificity of the antibody. Notably, incubation of Src with paxillin Y88F mutant protein induced a mobility shift of the paxillin fragment, suggesting that Src is capable of phosphorylating other tyrosine residues in the N-terminal fragment of paxillin. Nonetheless, our data demonstrate that Src directly phosphorylates paxillin

Y88.

pY88 paxillin regulates PI3-Kinase activation

Previous studies show that phospho-Akt is a major downstream effector of pY88 paxillin, as Akt phosphorylation induced by PDGF is significantly attenuated in paxillin Y88F mutants (Zhao, Zhang et al. 2010). How this signal is transduced has not yet been determined. It is well documented that AKTs are activated by phosphatidylinositol 3- kinase (PI3K) (Liu, Cheng et al. 2009). PI3K converts phosphatidylinositol 4,5- bisphosphate (PIP2) to phosphatidylinositol 3,4,5-trisphosphate (PIP3). PIP3 then recruits

AKT and PDK1 to the plasma membrane. There PDK1 phosphorylates and activates

AKT. PI3K consists of a p85 regulatory subunit and a p110 catalytic subunit. PI3K becomes activated when it is recruited to the membrane by interaction between p85 and phospho-tyrosine residues on membrane-bound receptors or adapter proteins (Okkenhaug and Vanhaesebroeck 2001). Several studies demonstrated that membrane-associated p130cas activates PI3K through interaction with p85 and that this interaction is dependent on tyrosine phosphorylation of the substrate-binding domain on p130cas (Li, Stupack et al. 2000, Riggins, DeBerry et al. 2003, Zhang, Guo et al. 2013). We previously showed that pY88 paxillin regulates p130cas tyrosine phosphorylation at tyrosine-165 (Y165), located in the substrate domain that will engage p85 (Zhao, Zhang et al. 2010). Recent

Figure 2-3. pY88 paxillin regulates p130cas-p85 interaction. A) Parental (WT) and paxillin Y88F cell lines were starved overnight and stimulated with PDGF-AA for the indicated times. Cell lysates were blotted with the indicated antibodies. B) HCT parental cells or HCT paxillin Y88F cell lines were lysed and the lysates were immunoprecipitated with anti-p130cas antibody. The immunoprecipitates and lysates were blotted with the indicated antibodies. C) Quantified p85 signal intensity was averaged over three independent replicates of this experiment using ImageJ software (National Institutes of Health, Bethesda, MD, USA). D) Paxillin becomes phosphorylated by Src at Y88. In turn, p130cas phosphorylation is upregulated, potentiating its interaction with p85 (dotted line indicates undetermined steps). p130cas-p85 interaction activates PI3-kinase, which then leads to cellular transformation. evidence has also implicated phosphorylation of p130cas tyrosine-128 (pY128), another

tyrosine residue located in the substrate-binding domain on p130cas, in PI3K-Akt regulation (Zhang, Guo et al. 2013). Consistent with pY165 p130cas, parental cells demonstrate robust phosphorylation of pY128 p130cas post-PDGF stimulation.

However, this phosphorylation event is dramatically attenuated in paxillin Y88F cells

(Figure 2-3A). We thus hypothesized that p130cas protein in paxillin Y88F mutant cells fails to interact with p85, or it does so to a lesser extent, thereby resulting in reduced

PI3K and AKT activation. As shown in Figure 2-3 B and C, p130cas binds p85 more tightly in parental cells versus paxillin Y88F cell lines. As such, pY88 paxillin affects

PI3K signaling via upregulating the p130cas-p85 interaction (Figure 2-3D).

pY88 paxillin does not correlate with clinical characteristics of CRC

We have shown previously that pY88 paxillin is up-regulated in a majority of human colon cancer specimens compared to matched normal colon tissues. To determine if varying levels of pY88 paxillin are associated with tumor prognosis, we stained a colorectal carcinoma tissue microarray with pY88 paxillin antibody for

Figure 2-4. pY88 paxillin is upregulated in colorectal cancer tissues. A) A representative image of strong pY88 paxillin immunohistochemical staining of human colorectal carcinoma. B) A representative image of low pY88 paxillin immunohistochemical staining of human colorectal carcinoma. C) A representative image of pY88 paxillin immunohistochemical staining of human colorectal carcinoma with signet ring morphology. D) A representative image of no pY88 paxillin immunohistochemical staining of human colorectal carcinoma. E) Proportion of high, intermediate and low pY88 paxillin immunohistochemical staining in a tissue microarray sorted by tumor stage (n.s. – chi-squared test between Stage I/II and Stage IV, low vs. high staining). F) Proportion of high and low pY88 paxillin immunohistochemical staining in a tissue microarray sorted by Stage IV tumor site (primary versus liver) (p = 0.0003 – chi-squared test between primary tumor and liver metastasis, low vs. high staining).

immunohistochemistry. Recorded patient characteristics include stage, age, right-

sidedness of primary tumor, peritoneal involvement, good outcome, race, sex and

microsatellite status (Table 2-1). Consistent with our initial report (Zhao, Zhang et al.

2010), we again saw strong pY88 paxillin staining in tumor tissue (Figure 2-4). However,

we were unable to correlate the staining intensity with stage. We also tested tumors with

signet ring morphology, as this subtype of colorectal carcinoma – although rare – is

associated with worse outcomes (O'Connell, Maggard et al. 2004); pY88 paxillin staining

did not correlate with this subtype. Intriguingly, there appears to be a relationship

between Stage IV tumor location and pY88 paxilin staining. Tumors with strong pY88 paxillin staining tend to be primary Stage IV tumors. Conversely, there are a greater proportion of liver metastases in the low pY88 paxillin cohort.

Paxillin serves as a predictive factor for dasatinib sensitivity

Although pY88 paxillin immunohistochemistry does not have prognostic value, we

wanted to see if it can predict dasatinib sensitivity. Dasatinib, used to treat acute

lymphoblastic and chronic myelogenous leukemias (Talpaz, Shah et al. 2006), is

currently in clinical trials in conjunction with standard treatment protocols for advanced

metastatic colorectal cancer (Strickler, McCall et al. 2014) (Montero, Seoane et al. 2011).

Given that we demonstrated that Src is the kinase that phosphorylates paxillin Y88, we

tested the Src family kinase inhibitor dasatinib on a panel of CRC cell lines with varying

levels of pY88 paxillin. Cell lines with low levels of pY88 paxillin were resistant to

dasatinib treatment, indicated by a high IC50 value. Conversely, cell lines with

progressively higher levels of pY88 paxillin levels had low IC50 values and thus are more

52 sensitive to dasatinib (Figure 2-5). Therefore, our data suggest that pY88 paxillin may be exploited as a predictive marker for response of colorectal cancer patients to dasatinib.

Figure 2-5. pY88 paxillin levels predict sensitivity to dasatinib. A panel of CRC cell lines was treated with varying concentrations of dasatinib and the IC50 of these lines was calculated. These cells were lysed and the lysates blotted with the indicated antibodies.

Discussion

This study identifies Src as the kinase that phosphorylates paxillin at Y88. Colorectal cancer cells that express high levels of pY88 paxillin are sensitive to dasatinib, therefore suggesting the use of pY88 paxillin as a predictive biomarker for Src inhibitor treatment.

Additionally, we demonstrate that paxillin Y88 phosphorylation promotes interaction

between p130cas and the p85 regulatory subunit of PI3K, thereby activating PI3K/AKT

signaling.

Although a previous study identified Y88 on paxillin as a possible site for

phosphorylation, it was largely thought to be a minor event and of no physiologic

importance (Schaller and Schaefer 2001). In contrast, prior work from our lab and this study strongly suggest otherwise. The decreased tumorigenicity of paxillin Y88F mutants described previously (7) indicates this phosphorylation event is important in cancer development. Moreover, down-regulation of pY88 paxillin by Src inhibitor data and shRNA suggests that paxillin Y88 is a physiologic substrate of Src. Using a specific pY88 paxillin antibody, we prove that Src directly phosphorylates pY88 paxillin by an in vitro kinase assay. Taken together, our study demonstrates unequivocally that Src is a kinase that directly phosphorylates paxillin at Y88.

The tumor suppressor PTPRT is mutated in a variety of cancers and as such is an appealing potential therapeutic target. However, given that PTPRT normally functions as a tumor suppressor and that its function is lost in cancer, it would be extremely difficult to reactivate PTPRT function in tumors. Therefore, it is crucial to identify the agent responsible for the corresponding oncogenic event; in the case of PTPRT, the kinase that phosphorylates its phospho-tyrosine substrate. Here we show that Src kinase phosphorylates paxillin Y88, the target site of PTPRT. Most importantly, we showed that

CRC cells expressing high levels of pY88 paxillin are sensitive to Src kinase inhibition, suggesting that this phosphorylation event may be exploited as a predictive biomarker for

Src family kinase inhibitors. However, further in vivo study in a xenograft and/or human

clinical trials is needed to validate this observation.

Additionally, we investigated further the in vivo pattern of pY88 paxillin levels. We

showed previously that it is upregulated in tumors versus normal matched controls (Zhao,

Zhang et al. 2010), but it is not associated with disease prognosis. However, in metastatic liver tumors, pY88 paxillin levels are dramatically reduced. This finding suggests that phosphorylation of paxillin Y88 is strongly affected by the tumor microenvironment.

In this study, we also show how pY88 paxillin transduces a signal to activate Akt, an

important mechanism for oncogenic growth. PI3K is activated by engagement with

phospho-tyrosine substrates. One such SH2-domain containing protein is the p85 subunit

of PI3K, as its docking with tyrosine-phosphorylated p130cas activates the p110alpha

subunit (Li, Stupack et al. 2000, Riggins, DeBerry et al. 2003, Zhang, Guo et al. 2013).

This interaction is attenuated in paxillin Y88F mutants, suggesting pY88 paxillin potentiates PI3K via p130cas-p85 interaction. However, an important outstanding question from this study is how pY88 paxillin upregulates the phosphorylation of p130cas. Specifically, tyrosine-165 and tyrosine-128 on p130cas both are phosphorylated to a greater extent in parental versus paxillin Y88F mutant cells, but other phosphorylation targets on p130cas (such as tyrosine-249 and tyrosine-410) are unaffected by the paxillin Y88F mutation (Zhao, Zhang et al. 2010). Given this lead,

further research in how pY88 paxillin affects kinases, phosphatases and signaling

molecules will improve our knowledge how it impacts in vivo tumorigenesis.

Materials and methods

Cell culture and reagents

HCT116, DLD1, HT29, SW480, RKO and HEK 293AAV cells were obtained from the

American Type Culture collection (Manassas, VA, USA). HEK 293AAV cells were grown in DMEM + 10% FBS while the other cell lines were grown in McCoy’s 5A

+10% FBS. Paxillin Y88F homozygous knock-in cell lines were generated in DLD1 and

HCT116 parental cells as described previously (Zhao, Zhang et al. 2010). For co- immunoprecipitation studies, cells were starved overnight and stimulated with 30ng/ml of platelet-derived growth factor-AA (PDGF-AA) for one hour.

Transfection

Cells were transfected using Lipofectamine transfection reagent (Invitrogen) as previously described (Hao, Wang et al. 2013). After 48 hours, cells were either lysed

(transient transfection) or put into selective media (stable transfection). The Src overexpression insert used a pCMV-3Tag-2A-3xmyc backbone (Aglient Technologies,

Santa Clara, CA, USA). Src shRNA knockdown used the Mission pLKO.1-puro system

(Sigma-Aldrich, St. Louis, MO, USA).

In vitro kinase

The N-terminal 165 amino acids of paxillin were cloned into the pET28 vector and

expressed in BL21 competent bacteria. Protein was purified using the Qiaexpressionist

(Qiagen, Valencia, CA, USA) protocol. Briefly, expression was induced overnight using

0.1mM IPTG, cells were lysed and protein was purified from lysate using Ni-NTA

agarose. Purified protein was then dialyzed using a slide-a-lyzer dialysis cassette

(Thermo Fisher Scientific, Waltham, MA, USA). Dialyzed protein was added to a reaction mixture containing ATP, buffer and recombinant Src (Promega, Fitchburg, WI,

USA) and the reaction proceeded for 1h at room temperature.

Cell lysis

Cells were either lysed using a urea lysis buffer (10mM Tris-HCl, pH 8; 100mM

NaH2PO4; 8M urea; 1mM Na3VO4; 20mM NaF; 80mcM beta-glycerophosphate; 20mM sodium pyrophosphate) or a Flag lysis buffer (50mM Tris-HCl pH 7.5; 1mM EDTA;

150mM NaCl; 1% Triton X-100; 1mM Na3VO4; 20mM NaF; Protease Inhibitor Cocktail

(Roche, Penzberg, GER) for co-immunoprecipitation. Western blots were performed as described previously (Zhao, Zhang et al. 2010). Antibodies include anti-pY88 paxillin, anti-PI3K p85 (Millipore, Temecula, CA, USA) anti-Src (Cell Signaling, Danvers, MA,

USA), anti-paxillin, anti-p130cas (BD Biosciences, San Jose, CA, USA), anti-GAPDH

(Santa Cruz Biotechnology, Santa Cruz, CA, USA).

Immunohistochemistry

Paraffin-embedded human tissue was deparaffinized in xylene and boiled for 20 minutes for antigen retrieval. Samples were incubated in primary pY88 paxillin antibody

57 overnight. The sections were stained with secondary antibody for 30 min at room temperature then visualized with EnVision-HRP kit (Dako).

Statistical analysis

Western blot quantification results were assessed using two-tailed paired Student’s t-test with significance set at p < 0.05. For immunohistochemistry, chi-square tests were used.

Chapter 3

PTPRT regulates high-fat diet-induced obesity and insulin resistance

A modified version of this chapter was previously published as:

Feng, X*, Scott A*, Wang Y, Wang L, Zhao Y, Doerner S, Satake M, Croniger CM,

Wang Z. PTPRT regulates high-fat diet-induced obesity and insulin resistance. PLOS

ONE.

Abstract

Obesity is a risk factor for many human diseases. However, the underlying molecular

causes of obesity are not well understood. Here, we report that protein tyrosine

phosphatase receptor T (PTPRT) knockout mice are resistant to high-fat diet-induced obesity. Those mice avoid many deleterious side effects of high-fat diet-induced obesity, displaying improved peripheral insulin sensitivity, lower blood glucose and insulin levels.

Compared to wild type littermates, PTPRT knockout mice show reduced food intake.

Consistently, STAT3 phosphorylation is up-regulated in the hypothalamus of PTPRT knockout mice. These studies implicate PTPRT-modulated STAT3 signaling in the regulation of high-fat diet-induced obesity.

Introduction

Numerous studies have shown the deleterious effects of obesity on health, increasing all-

cause mortality (Berrington de Gonzalez, Hartge et al. 2010) and predisposing

individuals to cardiovascular disease, diabetes and cancer (Flegal, Graubard et al. 2007).

Diet plays a crucial role in obesity, specifically those high in fats and sugar that increase

body fat (Bray 2010, Ahima 2011). Adipocytes, which increase in size and number

during obesity, can dramatically influence a variety of metabolic processes by disturbing

normal homeostatic signals (Haslam and James 2005). Chief among these disturbances is

insulin resistance, leading to hyperglycemia and diabetes (Surwit, Seldin et al. 1991,

Schreyer, Wilson et al. 1998, Rossmeisl, Rim et al. 2003, Hussain 2010).

Energy imbalance – essentially a combination of increased food intake with decreased

energy expenditure – causes obesity (Jequier 2002, Ahima 2011). Circulating hormones, such as insulin and leptin, are readouts of the body’s energy state and act at the hypothalamus to affect food intake (Schwartz, Woods et al. 2000, Beck 2006, Hariri and

Thibault 2010, Ahima 2011, Bi, Kim et al. 2012, Williams and Elmquist 2012). Ideally,

energy intake is equal to energy expenditure, leading to weight homeostasis. However, if

not enough energy is released proportional to calories consumed, the excess energy is

stored as lipid in adipocytes and weight gain ensues (Hariri and Thibault 2010). For

example, dietary fat consumption affects both sides of the energy imbalance equation.

Since it releases less satiety signals in comparison to protein and carbohydrate, it leads to

increased food intake (Jequier 2002). Conversely, since fats are an efficient form of

energy and because they are stored instead of used as an energy source after feeding,

dietary lipids also contribute to decreased energy expenditure (Jequier 2002, Hariri and

Thibault 2010). Therefore, from both biochemical and physiologic perspectives of energy

homeostasis, an excess of food intake over what is expended leads to weight gain.

Protein tyrosine phosphatases (PTPs) modulate signaling pathways that regulate a variety

of metabolic processes through de-phosphorylating tyrosine residues on proteins (Xu,

Schwab et al. 2013). Increasing evidence suggests that PTPs play a crucial role in obesity

and metabolic disease (Xu, Schwab et al. 2013). It has long been known that PTP1B is

implicated in obesity, insulin resistance and type-2 diabetes mellitus by regulating insulin signaling (Bence, Delibegovic et al. 2006). A recent study showed that TCPTP is also involved in obesity through modulating leptin signaling (Loh, Fukushima et al. 2011).

TCPTP dephosphorylates STAT3 at the tyrosine 705 (Y705) residue. STAT3 Y705 phosphorylation is a key mediator of leptin signaling in the hypothalamus (Vaisse, Halaas et al. 1996, Bates, Stearns et al. 2003). Leptin-STAT3 signaling suppresses the drive for food intake by increasing the expression of anorectic neuropeptides and repress those favoring orexigenic responses (Elmquist, Maratos-Flier et al. 1998, Schwartz, Woods et al. 2000, Munzberg, Bjornholm et al. 2005, Bence, Delibegovic et al. 2006, Loh,

Fukushima et al. 2011).

Because we previously showed that STAT3 is a substrate of protein tyrosine phosphatase

receptor T (PTPRT) (Zhang, Guo et al. 2007), we investigate here whether PTPRT

regulates food intake and obesity in mice.

Results

Ptprt-/- mice are resistant to high-fat diet-induced obesity

Figure 3-1. PTPRT KO mice demonstrate slightly lower body weight than wild type littermates on normal chow diet.

Eight week-old male mice of Ptprt+/+ (n=13), Ptprt+/- (n=13) and Ptprt-/- (n=13) genotypes were maintained on a normal chow diet for 29 weeks. Body weight of the three genotypes was assessed weekly. (* p < 0.05; t-test comparing Ptprt+/+ and Ptprt-/- genotypes) As described previously (Zhao, Zhang et al. 2010), we bred the Ptprt knockout allele into

the C57BL/6 strain for over 15 generations. When mouse body weights were measured

from 8-week-old to 36-week-old mice, we observed that the body weight of Ptprt-/- mice were slightly and consistently lower than those of Ptprt+/+ littermates on chow diet

(Figure 3-1). However, these mice were not obese. It is well documented that high-fat diet induces obesity and insulin resistance in C57BL/6 male mice (West, Boozer et al.

1992). To interrogate if PTPRT plays a role in obesity, five-week-old male Ptprt+/+,

Ptprt+/- and Ptprt-/- mice were fed with a high-fat diet for 14 weeks. Although Ptprt-/-

mice were largely indistinguishable from their Ptprt+/+ and Ptprt+/- littermates in terms of

body weight on a normal diet at a baseline of five weeks (Figure 3-2A, Time 0 weeks),

the body weights of Ptprt-/- mice are significantly lower than those of Ptprt+/+ littermates

through the course of the high-fat diet (Figure 3-2A). Consistent with previous reports

(West, Boozer et al. 1992), the Ptprt +/+ male mice were obese at the end of 14 weeks

(average body weight = 46.5g). In contrast, the Ptprt-/- male mice remained lean (average

body weight = 33.3g) after being fed with a high-fat diet for 14 weeks, suggesting that

knocking out of Ptprt renders male mice resistant to high-fat diet-induced obesity.

Ptprt-/- mice have less body fat by percentage than wild type littermates

Obesity and its associated co-morbidities are caused by excess amounts of body fat.

Therefore, we set out to determine body composition of the high-fat diet fed mice using

quantitative magnetic resonance to ensure that the difference in weight gain can be

attributed to increased obesity (Tinsley, Taicher et al. 2004). As expected, the percentages of body fat of Ptprt+/+ and Ptprt+/- mice were significantly higher than that of

Ptprt-/- mice (Figure 3-2B). Consistently, the lean body mass and water in Ptprt+/+ mice were lower than the Ptprt-/- mice (Figure 3-2 C and D). Since the body weight and body fat percentage of Ptprt+/- mice were not significantly different from those of Ptprt+/+ mice

(Figure 3-2), we focused on Ptprt+/+ versus Ptprt-/- mice for in depth analyses in this

study.

Figure 3-2. PTPRT KO mice are resistant to high-fat diet-induced body composition changes.

A) Five week-old male mice of Ptprt+/+ (n=8), Ptprt+/- (n=14) and Ptprt-/- (n=11) genotypes were fed with high-fat diet for 14 weeks. Body weight of the three genotypes was assessed weekly. (* p < 0.05; t-test comparing Ptprt+/+ and Ptprt-/- genotypes).

B-D) Body composition was analyzed by quantitative magnetic resonance after 14 weeks on a high-fat diet (Fat % – B, Lean % – C, Water % – D; * p < 0.05; t-test comparing Ptprt+/+ and Ptprt-/- genotypes; t-test comparing Ptprt+/- and Ptprt-/- genotypes was n.s.).

Ptprt-/- reduces food intake

Next, we set out to interrogate the variety of mechanisms by which Ptprt-/- mice are

resistant to high-fat diet-induced obesity. Given that food intake is one of the major

factors that impact body weight, we measured food intake of Ptprt+/+ and Ptprt-/- mice

both at the beginning and at the end of the high-fat diet. For a ten-day period, food intake was measured daily and average values were calculated. As shown in Figure 3-3A, Ptprt-

/- mice ate significantly less than their Ptprt+/+ counterparts at the beginning of the high-

fat diet period. However, at the end of the high-fat diet period, Ptprt-/- mice again trended

toward lower food intake, but the difference did not reach statistical significance (p =

0.17).

To determine the reason behind the increased food intake in Ptprt+/+ versus Ptprt-/- mice,

we inferred that neurohormonal signals may play a role. Chief among these signals are

leptin and neuropeptide Y, serving in an anorectic or orexigenic fashion, respectively.

While plasma levels of leptin did not show significant difference between Ptprt+/+ versus

Ptprt-/- mice (Figure 3-4), Ptprt+/+ mice had a significantly higher plasma level of

neuropeptide Y versus Ptprt-/- mice at the beginning of the high-fat diet period (Figure 3-

5A). This difference disappeared by the end of the high-fat diet period (Figure 3-5B).

The plasma NPY appears to correlate with the food intake patterns at the beginning and

end of the high-fat diet period.

Figure 3-3. PTPRT KO mice eat less but do not absorb dietary fats differently.

A) Average daily food intake of Ptprt+/+ (n=8), and Ptprt-/- (n=12) mice was assessed at beginning and end of the high-fat diet (* p < 0.05; t-test).

B) Dietary lipid absorption was assessed using fecal samples of Ptprt+/+ and Ptprt-/- mice after being fed a butter oil and sucrose behenate diet.

Figure 3-4. PTPRT KO mice do not have different circulating levels of leptin.

A-B) Fasting plasma leptin levels of Ptprt+/+ and Ptprt-/- mice were assessed before (A) and after (B) high-fat diet.

Figure 3-5. PTPRT KO mice have decreased NPY levels before high-fat diet.

A-B) Fasting plasma neuropeptide Y levels of Ptprt+/+ and Ptprt-/- mice were assessed before (A) and after (B) high-fat diet (** p < 0.01; t-test).

Since dietary lipid absorption also impacts body weight and because PTPRT is expressed

in the small intestine and colon (Zhao, Zhang et al. 2010), we set out to assess the lipid

absorption capacity of Ptprt+/+ and Ptprt-/- mice using non-invasive fecal analysis

(Jandacek, Heubi et al. 2004). However, these two cohorts did not show any difference in

absorbing dietary fats (Figure 3-3B). Taken together, our data suggest that the reduced

body weight of Ptprt-/- mice may be due to less food consumption.

Ptprt-/- mice have reduced energy expenditure than wild type mice

Since reduced energy expenditure and differences in nutrient utilization could also contribute to obesity, we assessed the energy expenditure and respiratory quotient of

Ptprt+/+ and Ptprt-/- mice via indirect calorimetry in both the fed (Figure 3-6A) and fasted

(Figure 3-6B) state. According to the energy expenditure values, Ptprt-/- mice had

reduced energy expenditure than Ptprt+/+ mice in both the fed and fasted state (Figure 3-

6, second row). However, reduced energy expenditure in Ptprt-/- mice does not explain

the weight difference between Ptprt+/+ and Ptprt-/- mice. The respiratory quotient of

Ptprt+/+ and Ptprt-/- mice was indistinguishable in the fasted state (Figure 3-6B, top

panel). In the fed state, Ptprt-/- mice had a higher respiratory quotient than Ptprt+/+ mice

during the dark phase (Figure 3-6A, top panel), indicating they preferentially use glucose,

but this difference was not sustained through the whole 24-hour period of testing. Taken

as a whole, indirect calorimetry demonstrates that the metabolic phenotype of Ptprt-/- mice does not explain their reduced body weight versus their Ptprt+/+ littermates.

Figure 3-6. PTPRT KO mice utilize more glucose and expend less energy than wild type mice.

Energy expenditure (Heat Released and Respiratory Quotient) was assessed through indirect calorimetry for mice in fed (A) and fasted (B) state through 24 hours (green segments: p < 0.05; t-test). Ptprt+/+ (n=8); Ptprt-/- (n = 11).

Ptprt-/- mice resist high-fat diet-induced hyperglycemia and insulin resistance

Given that obesity often causes metabolic syndrome, such as hyperglycemia and

peripheral insulin resistance, we measured fasting glucose and insulin levels in Ptprt+/+

and Ptprt-/- littermates. At the end of the high-fat diet treatment, the average blood

glucose levels in Ptprt+/+ mice reached 220.6mg/dL; therefore, these mice were

hyperglycemic. In contrast, the blood glucose levels in Ptprt-/- were within normal range

at 152.6mg/dL (Figure 3-7A). Accordingly, Ptprt+/+ mice also had higher insulin levels than the Ptprt-/- counterparts after the high-fat diet (Figure 3-7B). These blood glucose and blood insulin values can be used to estimate peripheral insulin resistance using the

HOMA-IR model (Matthews, Hosker et al. 1985). By this calculation, Ptprt-/- had much

lower HOMA-IR values than their Ptprt+/+ littermates (Figure 3-7C), indicating a

prediction of insulin sensitivity versus Ptprt+/+ mice. It is worth noting that the Ptprt+/+

mice were neither hyperglycemic nor insulin resistant before high-fat diet treatment,

although the blood glucose levels in the Ptprt+/+ mice were slightly higher than that of

Ptprt-/- littermates (Figure 3-7D). No blood insulin difference was observed before high-

fat diet treatment among the Ptprt+/+ and Ptprt-/- mice (data not shown). Before the high-

fat diet, Ptprt-/- mice also secreted more insulin in response to a glucose bolus (Figure 3-

7E).

We then further measured peripheral insulin resistance in mice fed the high-fat diet via an insulin tolerance test. Consistent with the HOMA-IR calculation, Ptprt-/- mice had a

lower glucose value in response to an insulin bolus after 60 minutes (Figure 3-8A).

However, Ptprt-/- and Ptprt+/+ mice did not have different glucose clearance in response to

a glucose tolerance test (Figure 3-8B). Taken together, our data suggest that loss of

PTPRT function attenuates the development of peripheral insulin resistance after a high- fat diet.

Figure 3-7. PTPRT KO mice have less insulin resistance than wild type mice after HFD.

A) Fasting blood glucose levels were assessed after 14 weeks on HFD (** p < 0.01; t-test). B) Fasting insulin levels were assessed after 14 weeks on HFD (** p < 0.01; t-test). C) HOMA-IR calculations of Ptprt+/+ and Ptprt-/- mice after HFD (** p < 0.01; t-test). D) Fasting blood glucose levels were assessed before HFD (* p < 0.05). E) Glucose stimulated insulin secretion test between Ptprt+/+ (n=5) and Ptprt-/- (n=5) mice before HFD (* p < 0.05; t-test).

Metabolic differences between Ptprt+/+ and Ptprt-/- littermates

Given the deviation seen between Ptprt+/+ and Ptprt-/- mice as it relates to glucose and

insulin metabolism, we decided to interrogate their blood plasma for differences in other

nutrients. These blood metabolites will shed additional light onto the metabolic

disturbances that Ptprt+/+ mice are experiencing. Interestingly, Ptprt-/- mice had lower

cholesterol and higher free-fatty acids than Ptprt+/+ mice, but we did not observe an

increase in triglycerides or β-hydroxybutyrate in these mice (Figure 3-9). Once fatty acids

are oxidized, the acetyl CoA produced is used to generate ketone bodies such as β- hydroxybutyrate. As such, PTPRT may also regulate the utilization and storage of dietary fats.

Phospho-STAT3 increased in the hypothalamus of Ptprt-/- mice

To elucidate the molecular mechanisms by which Ptprt-/- mice resist high-fat diet-induced obesity, we assessed PTPRT expression in tissues implicated in metabolic regulation.

Consistent with a previous report that PTPRT is expressed in the hypothalamus (Visel,

Carson et al. 2007), we detected PTPRT protein in the hypothalamus using Western blot analyses (Figure 3-10A). However, it is not expressed in the liver, adipose or muscle.

Since STAT3 is a substrate of PTPRT, we reasoned that phospho-STAT3 levels may be increased in the hypothalamus of Ptprt-/- mice. As expected, we found that pY705 STAT3

is up-regulated in hypothalamus of Ptprt-/- mice compared to Ptprt+/+ mice (Figure 3-

10B), suggesting that PTPRT modulates food intake by affecting phospho-STAT3 levels in the hypothalamus (Figure 3-10C).

Figure 3-8. PTPRT KO mice demonstrate better insulin regulation than wild type mice.

A) Insulin tolerance test of Ptprt+/+ (n=8) and Ptprt-/- (n=11) mice after high-fat diet (* p < 0.05; t-test). B) Glucose tolerance test of Ptprt+/+ (n=8) and Ptprt-/- (n=11) mice after high-fat diet.

Figure 3-9. PTPRT KO mice have different blood chemistry values after high-fat diet.

Fasted plasma concentrations of cholesterol (A), non-esterified fatty acids (B), triglycerides (C) and beta-hydroxybutyrate (D) of Ptprt +/+ (n=8) and Ptprt-/- (n=11) mice after high-fat diet (** p < 0.01; t-test).

Figure 3-10. PTPRT regulates STAT3 phosphorylation in mouse hypothalamus.

A) Tissue lysates from Ptprt+/+ and Ptprt-/- mice were blotted with indicated antibodies.

B) Hypothalamic lysates from Ptprt+/+ and Ptprt+/- mice were blotted with indicated antibodies.

C) Proposed model for the effect of PTPRT on food intake.

Discussion

Our study reveals that PTPRT regulates metabolism and body weight. Our data suggest

that PTPRT could be a drug target for obesity, because Ptprt-/- mice resist many key

effects of a high-fat diet, including increased body mass, hyperglycemia,

hypercholesterolemia, insulin resistance and increased adiposity. Consistent with this

notion, several recent human genetic studies linked obesity to chromosome 20q12-13

(Borecki, Rice et al. 1994, Lembertas, Perusse et al. 1997, Lee, Reed et al. 1999), the genomic region in which PTPRT is located.

The decreased food intake in Ptprt-/- mice suggests a behavioral mechanism as to why

they weigh less than their Ptprt+/+ littermates. Food intake is primarily decreased by

leptin signaling pathway (Elmquist, Maratos-Flier et al. 1998, Schwartz, Woods et al.

2000, Munzberg, Bjornholm et al. 2005) and increased by neuropeptide Y (Beck 2006,

Brothers and Wahlestedt 2010, Bi, Kim et al. 2012). Leptin suppresses food intake by

activating STAT3 phosphorylation in the hypothalamus (Vaisse, Halaas et al. 1996,

Bates, Stearns et al. 2003); our previous study shows that PTPRT dephosphorylates

STAT3 in colorectal cancers (Zhang, Guo et al. 2007). The decreased food intake in

Ptprt-/- mice in the absence of increased circulating leptin levels suggests that Ptprt-/-

mice have increased phospho-STAT3 independent of leptin activity in the hypothalamus

(Figure 3-10C). Our data indicate that STAT3 hyper-phosphorylation in the

hypothalamus represses food intake in Ptprt-/- mice. As such, we propose that the central

nervous system plays a dominant role in the phenotype of Ptprt-/- mice.

Ptprt-/- mice demonstrate decreased peripheral insulin resistance as well as lower levels of blood insulin and glucose. Although a human study shows that PTPRT expression levels in adipose tissue are much higher in insulin-resistant individuals compared to insulin-sensitive individuals (Elbein, Kern et al. 2011), we failed to detect PTPRT protein in mouse adipose (Figure 3-10A). Neither could we detect PTPRT protein in liver or muscle (Figure 3-10A). Our data indicate that PTPRT does not directly modulate insulin sensitivity in peripheral tissues. Instead, PTPRT may indirectly impact peripheral insulin resistance through affecting the nervous system control of energy homeostasis. It is well documented that increased plasma NPY from autonomic nervous system sources is associated with greater adiposity and increased insulin resistance (Kuo, Kitlinska et al.

2007, Warne and Dallman 2007, Ruohonen, Pesonen et al. 2008, Bagherian, Kalhori et al. 2010, Han, Li et al. 2012, Ruohonen, Vahatalo et al. 2012). Consistently, our data show increased NPY secretion in Ptprt+/+ mice that go on to develop obesity and insulin resistance. As such, the decrease in NPY in Ptprt-/- mice further suggests the role of

PTPRT in nervous system regulating obesity and peripheral insulin resistance (Obici,

Zhang et al. 2002, Gerozissis 2008, Koch, Wunderlich et al. 2008, Pagotto 2009, Lam,

Chari et al. 2011, Carey, Kehlenbrink et al. 2013).

Materials and Methods

Animals and Diet

Treatment of experimental mice and related protocols were done in accordance with the

Institutional Animal Care and Use Committee at Case Western Reserve University

(Protocol Number 2010-0125). Male and female PTPRT heterozygous and homozygous

knockout mice in a C57BL/6 background were generated as described previously (Zhao,

Zhang et al. 2010); referred to as Ptprt+/- and Ptprt-/-, respectively. Colonies of these mice

were maintained on a normal chow diet (#5010 (4.5% fat by weight, 12.7% fat by

calorie), LabDiet St. Louis, MO). Five weeks after birth, male Ptprt+/+, Ptprt+/- and Ptprt-

/- mice were put on a high-fat diet (#D12331: 33% Hydrogenated Coconut Oil (35% fat

by weight; 58% fat by calorie), Research Diets, Inc. New Brunswick, NJ) for 14 weeks.

Body composition was analyzed by quantitative magnetic resonance as described

previously (Tinsley, Taicher et al. 2004).

Glucose and insulin tolerance test

Tests were done as described previously (Berry, Jacobs et al. 2013, Marwarha, Berry et

al. 2013). Mice were deprived of food overnight and then injected intraperitoneally with

glucose (2 g/Kg) or insulin (0.9 g/Kg). Tail vein blood was sampled for glucose levels at

0, 15, 30 and 60 minutes after insulin or 0, 15, 30 60 and 120 minutes after glucose using

an UltraTouch meter. GTT was performed at the Mouse Metabolic Phenotyping Center

of CWRU.

HOMA-IR

Insulin resistance was estimated for Ptprt+/+ and Ptprt-/- mice after high-fat diet using the

homeostatic model assessment (Matthews, Hosker et al. 1985). Formula = (Insulin

(mcU/L) x Glucose (mg/dl) / 405).

Insulin, neuropeptide Y and blood chemistry measurements

Tests were done as described previously (Berry, Jacobs et al. 2013). Mouse blood plasma

was isolated using Microtainer plasma separator tubes (BD Biosciences). Insulin was

measured using an insulin enzyme-linked immunosorbent assay (Mercodia, Inc.,

Uppsala, SWE). Neuropeptide Y was measured using neuropeptide Y insulin enzyme-

linked immunosorbent assay (EMD Millipore, Billerica, Massachusetts, USA). Mouse

plasma was sent to Marshfield Laboratories to assess β-hydroxybutyrate, triglycerides,

non-esterified fatty acids and total cholesterol.

Lipid absorption

Tests were done as described previously (Buchner, Geisinger et al. 2012). Mice were fed a diet consisting of butter oil and 5% sucrose polybehenate for three days. Fecal pellets from these mice were collected and analyzed via gas chromatography of fatty acid methyl esters. The ratio of behenic acid to other fatty acids then was used to determine intestinal lipid absorption (Jandacek, Heubi et al. 2004). The lipid absorption studies were performed at the Cincinnati Mouse Metabolic Phenotyping Center.

Glucose-stimulated insulin secretion test

Tests were done as described previously (Millward, Desantis et al. 2010). For the glucose-stimulated insulin secretion test, mice were starved overnight and 2g/kg of glucose was injected intraperitoneally into the mice. Tail vein blood was collected and plasma insulin concentrations were measured at 0 and 30 minutes after glucose injection.

Indirect calorimetry

Tests were done as described previously (Prince, Zhang et al. 2013). Indirect calorimetry

(IDC) was performed using the Oxymax system (Columbus Instruments' Comprehensive

Lab Animal Monitoring System (CLAMS), Columbus, OH). VO2, VCO2, respiratory quotient (RQ) and heat (energy expenditure – EE) were determined. Energy expenditure was normalized to body mass. IDC was performed on mice either after an overnight fast and water (Fasted) or with ad lib food and water (Fed). The experiments ran for 22 hours on a 12 hour dark cycle (6pm to 6am).

Food intake

Mice with ad lib access to food and water were placed in a clean cage and the food was weighed. The remaining food after 24 hours was weighed and the average food intake per mouse was calculated (Marwarha, Berry et al. 2013).

Western Blot

Tissues were lysed as described previously (Marwarha, Berry et al. 2014). Total brain, hypothalamus, hind leg muscle, perigonadal adipose or liver tissue were lysed in RIPA lysis buffer (150mM NaCl, 10mM Tris-HCl (pH 7.5), 0.1% SDS, 1% Triton X-100, 1%

Deoxycholate, 0.5M 5mM EDTA) supplemented with protease (Roche, Penzberg, GER) and phosphatase inhibitors (1mM NaVO4, 50mM NaF). Western blots were performed as described previously (Zhao, Zhang et al. 2010). Antibodies for pSTAT3Y705 and

STAT3 were from Cell Signaling (Danvers, MA, USA), tubulin from Sigma-Aldrich (St.

Louis, MO, USA) and PTPRT from Biovendor (Asheville, NC, USA).

Statistical analysis

Results were assessed using two-tailed unpaired Student’s t-test with significance set at p

< 0.05.

Chapter 4

Discussion and Future Directions

Summary

Protein tyrosine phosphorylation plays an important role in transducing cellular signals.

Here we discuss the implications of this protein modification in obesity and in cancer.

Using a knockout mouse for the protein tyrosine phosphatase superfamily member

PTPRT, we demonstrate its importance in energy homeostasis. Specifically PTPRT KO mice have reduced weight, either on normal chow or a high-fat diet, and have reduced fat

burden after a high-fat diet. This has ramifications in regard to glucose metabolism, as

these mice are less insulin resistant. We can tie this phenotype back to tyrosine

phosphorylation, as PTPRT KO mice have increased phospho-STAT3 in the

hypothalamus. The chief effect of phospho-STAT3 is decreased food intake, and also

impacts other aspects of energy homeostasis as well. Another important PTPRT

substrate is paxillin at tyrosine-88. We study this phosphorylation event using a cell line

with an unphosphorylatable mutant of paxillin tyrosine-88 (Y88F), which is less

tumorigenic versus parental cell lines. As such, this phosphorylation event has

ramifications for tumorigenic signaling pathways, affecting activation of PI3K.

Ultimately, pY88 paxillin is a phosphorylation event at the crossroads of both kinase

overexpression and phosphatase inactivation, representing the end result of genetic

alterations that silence tumor suppressors and activate oncogenes. It underscores novel

approaches for therapy and belies synergy between parallel oncongenic pathways.

PTPRT and Obesity: Future Directions

What is the role of PTPRT in glucose and lipid metabolism?

Many of the experimental findings seen in PTPRT KO mice after a high-fat diet are consistent with the established relationship between obesity and diabetes (Figure 4-1).

Their reduced food intake leads to reduced adiposity versus their wild-type counterparts, leading to improved insulin sensitivity as measured by homeostatic modeling (HOMA-IR value), an insulin tolerance test and glucose utilization (insulin stimulates glycolysis, consistent with the increased respiratory exchange ratio in KO mice).

Figure 4-1. Summary of PTPRT KO phenotypes after 14 weeks on a high-fat diet.

A combination of decreased food intake and decreased plasma NPY decreased both adiposity and insulin resistance. As a result, PTPRT KO mice were able to use glucose and clear it effectively from the blood. However, these mice may not reflect certain aspects of insulin sensitivity (marked by red X), such as decreased plasma free fatty acids, increased cholesterol production and decreased hepatic glucose production

This effect on glucose homeostasis makes PTPRT a good candidate for drug development

in the treatment of diabetes. The PTP1B knockout mouse generated much interest in

drugging the PTP family of proteins in the treatment of diabetes, and so many

compounds were derived to target this protein to attenuate insulin resistance (Johnson,

Ermolieff et al. 2002, Thareja, Aggarwal et al. 2012). Although many of these compounds did not reach market, recent research has delved into the stabilization of the inactive, oxidized form of PTPs (Haque, Andersen et al. 2011).

However, the glucose tolerance test results and fatty acid metabolism data suggest a more

complicated phenotype for PTPRT KO (Figure 4-1, red X). In relation to free fatty acid

metabolism, improved insulin signaling in PTPRT KO mice should either stimulate free

fatty acid (FFA) storage in the adipose tissue and/or stimulate triglyceride and cholesterol

synthesis in the liver. This expectation is opposite to experimental results. Either by

impaired lipogenesis or dysregulated lipolysis, PTPRT KO mice are unable to effectively

store free fatty acids; the lack of increase in beta-hydroxybutyrate and the respiratory quotient value also implies they are not being used as energy. As such, insulin is overridden by some other signaling pathway that mobilizes FFA from the adipose, which may explain the decreased adiposity in PTPRT KO mice. Additionally, the inability for the liver to make cholesterol despite the surplus of its constituent parts (triglycerides and free fatty acids that are catabolized for the sterol backbone of cholesterol) leads to the hypothesis of hepatic insulin resistance. Similar triglyceride levels in wild-type and

PTPRT KO may represent normal liver function in the latter, as nutrient excess will lead to endogenous triglyceride production by the liver (Berne 2003). Nevertheless, the low

84 cholesterol levels are still suspicious, since triglycerides should be packaged in cholesterol for plasma transport.

Figure 4-2. Diagram of hepatic insulin signaling.

Hepatic insulin sensitivity - insulin downregulates the transcription of genes involved in gluconeogenesis and glycogen breakdown, shunting glucose through glycolysis or into glycogen storage. Insulin also upregulates the transcription of genes involved in cholesterol and triglycerol synthesis.

Hepatic insulin resistance - the liver fails to blunt glucose production despite the presence of insulin. Additionally, fatty acids are not formed into cholesterol and not incorporated to triglycerides. (Brown and Goldstein, 2008)

(Pocai, Morgan et al. 2005)

The indistinguishable glucose tolerance test represents another manifestation of hepatic

insulin resistance (Figure 4-2), as an insulin sensitive liver should help decrease blood

glucose. As seen in the glucose-stimulated insulin secretion test and the insulin tolerance test, PTPRT KO mice are superior to their wild-type littermates as it relates to the secretion of insulin in response to glucose and the ability for peripheral tissues to clear

glucose from the blood stream in response to insulin. We can rule out a defect in glucose

homeostasis at these two levels. However, the liver has a slightly different role than

other peripheral tissues in regard to glucose handling. In response to insulin, it does not

directly increase glucose uptake (via insulin-dependent glucose transporters) but instead affects key enzymes involved in glycogen synthesis and gluconeogenesis at both the transcriptional and post-translational levels (Meshkani and Adeli 2009) to indirectly decrease blood glucose. As such, the PTPRT KO liver ostensibly has some defect in response to a bolus of glucose.

Given the crucial role of the liver in anabolic processes, an attractive hypothesis is that

PTPRT modulates hepatic insulin resistance distinctly from its role in adipose and muscle

(which seemingly remain insulin sensitive as it relates to glucose). Additional experiments are needed to explain how PTPRT can affect nutrient metabolism.

Experiments that would implicate the liver in this process include the glucose clamp

(which directly measures glucose uptake by tissues) as well as metabolic MRI and deuterated water studies (which follow catabolic and anabolic processes through the

86 body). Also, although we know PTPRT KO mice have lower cholesterol levels, understanding the composition of the plasma cholesterol will better characterize the liver function in these mice: fast protein lipid chromatography separates the deleterious Very-

Low Density Lipoprotein from High Density Lipoprotein and can also be used to identify triglyceride incorporation into cholesterol molecules (Biddinger, Hernandez-Ono et al.

2008). Hepatic insulin resistance is associated with increased VLDL (Biddinger,

Hernandez-Ono et al. 2008). Moreover, the experiments should be done before and after a high-fat diet to determine if the nutrient milieu affects hepatic insulin signaling or if there is an inborn error in PTPRT KO mice.

Interrogating other important hormones involved in energy expenditure, such as epinephrine, glucagon and somatostatin, will better determine the level at which this dysfunction is occurring. Inappropriately high glucagon could potentially explain the increased plasma NEFA and failed glucose tolerance test (Champe et al. 2007).

However, increased plasma glucagon should have increased the beta-hydroxybutyrate and resulted in preferential use of fatty acids instead of glucose per the respiratory quotient, inconsistent with our PTPRT KO findings. Sympathetic outflow mediated by epinephrine is another mechanism by which the body responds to bolus of glucose

(Haque, Minokoshi et al. 1999, Esler, Rumantir et al. 2001, Straznicky, Lambert et al.

2009). As such, repeating the glucose tolerance test with beta2-adrenergic agonists, which cause vasodilation in the musculature (Harvey and Champe 2008) will address the aspects of sympathetic outflow that could affect glucose uptake.

Additional PTP knockout mouse models provide important clues as to the potential mechanism for PTPRT’s activity on high-fat diet-induced obesity. Brain-specific knockout of PTP1B makes mice resistant to high-fat diet-induced obesity (to a similar extent as the PTPRT knockout) by improved leptin sensitivity (Bence, Delibegovic et al.

2006); this effect was not recapitulated if PTP1B was knocked out in adipose or other peripheral tissues. Similarly, TCPTP brain-specific knockout mice are resistant to high-

fat diet-induced weight gain and insulin resistance, specifically by affecting STAT3

phosphorylation (Loh, Fukushima et al. 2011); glucose metabolism phenotypes are also different on the normal diet. However, they are different from PTPRT knockout mice in that their body-fat percentage is no different than their wild-type littermates, and the weight difference is not as dramatic (but they are started on a high-fat diet approximately three weeks later than PTPRT KO mice). The more dramatic difference in weight gain in the PTPRT KO mice may be due to a baseline difference in insulin metabolism in the

PTPRT KO mouse. Consistent with PTPRT’s expression in pancreatic beta cells (data not shown), PTPRT KO mice released more insulin in response to a glucose bolus, implicating PTPRT in glucose regulation at the level of the pancrease. Additionally, the authors demonstrate that TCPTP expression increases in the brain after a high-fat, implicating it in high-fat diet-induced metabolic changes. In opposition to both above

PTP knockout models, PTPRT KO mice actually had less energy expenditure.

Using these other PTP knockout as models, additional experiments can be done to further

investigate PTPRT’s mechanism of high-fat diet-induced obesity. For example, a pathogenic role for PTPRT could be investigated by assessing its expression level after a high-fat diet feeding period. Similar to the PTP1B experiments, tissue-specific PTPRT

KO could assess whether its effect on high-fat diet-induced obesity is centrally mediated.

To verify that phospho-STAT3 is downstream of PTPRT, the PTPRT KO should be crossed with brain- or adipose-specific STAT3 knockouts, which have been shown to have a dramatic effect on obesity development (Cui, Huang et al. 2004, Cernkovich,

Deng et al. 2008). Finally, given that many of the phenotypes of the TCPTP KO mouse were present even before high-fat diet period, many of the experiments done on PTPRT

KO mice should be repeated on normal chow, and also paying close attention to the relative weights between the mice. Doing so will tease apart the effects of PTPRT KO on body weight from a possible independent effect on glucose homeostasis.

Despite the improved insulin tolerance test and lower HOMA-IR value, certain aspects of the PTPRT KO mice suggest it may serve as a model for a pre-diabetes. Pre-diabetes is a clinical finding of impaired glucose tolerance and/or elevated fasting glucose levels, but not at the level of hyperglycemia (American Diabetes 2005). This constellation of findings is defined because it comes with it an increased risk of cardiovascular disease as well as the development of diabetes itself. The PTPRT KO may fall under this category because of its average fasting blood glucose value (roughly at the upper bound of normal glucose levels for C57BL/6 mice, 150mg/dL) and because of its indistinguishable glucose

89 tolerance test with their wild-type littermates. As such, the PTPRT KO mice may serve as an effective tool into the study of pre-diabetes.

Taken together, the complex and enigmatic phenotype of PTPRT KO mice after high-fat diet consumption may ultimately point toward the in vivo role of this protein.

How does PTPRT affect the relationship between NPY, stress and obesity?

PTPRT’s effect on circulating NPY will help determine why PTPRT KO are resistant to diet-induced obesity, given that NPY levels were higher at the beginning of the high-fat diet period. Consistent with this finding, circulating NPY potentiates fat accretion and peripheral insulin resistance. In order to validate NPY’s effect on the PTPRT KO resistance to obesity, future studies should feed the high-fat diet to these mice in the presence of an NPY receptor agonist, many of which are commercially available (Zhang,

Bijker et al. 2011), to compensate for the baseline differences in NPY levels.

Certain post-ganglionic neurons release NPY alongside norepinephrine and canonically it is associated with the stress response. Systemic sources mainly include sympathetic nerves, the adrenal medulla and platelets (Hirsch and Zukowska 2012). NPY signaling promotes angiogenesis (Kuo, Kitlinska et al. 2007), also regulating lipid oxidation, lipogensis and adipogenesis; peripheral NPY receptor knockout mice become resistant to diet induced obesity (Shi and Baldock 2012) (Geerling, Boon et al. 2014). Our PTPRT

KO phenotype is consistent with these studies. Additionally, NPY is implicated in vasoconstriction (Zukowska-Grojec 1995) and inflammation, the latter of which is especially salient to the peripheral insulin resistance phenotype seen with increased levels of NPY (Kuo, Kitlinska et al. 2007, Hirsch and Zukowska 2012). Further study needs to determine if the difference in NPY release is centrally mediated at the level of the hypothalamus or if the end-organs (e.g., adrenal medulla and sympathetic ganglia) are dysfunctional and not adequately secreting it. Ultimately, the relationship between

PTPRT and NPY may inform stress-related mechanisms of obesity in wild-type vs.

PTPRT KO mice (de Kloet, Joels et al. 2005, Bornstein, Schuppenies et al. 2006).

What are other implications of hypothalamic phospho-STAT3?

The hypothalamus represents an important central regulator of nutrient and energy homeostasis (Schwartz, Woods et al. 2000, Elmquist and Marcus 2003, Munzberg,

Bjornholm et al. 2005). Accordingly, an important experimental finding in the PTPRT

KO mice is the increased phospho-STAT3 in the hypothalamus versus wild-type counterparts. Specific manipulations of hypothalamic STAT3 underscore its importance in energy balance: its knockout in certain hypothalamic neuron populations variously results in increases in food intake, hyperinsulinemia and hyperleptinemia (Xu, Ste-Marie et al. 2007, Gong, Yao et al. 2008), while constitutively active STAT3 results in lower body weight, leaner body habitus and increased locomotor activity (Mesaros, Koralov et al. 2008).

Chiefly, increased phospho-STAT3 explains the food intake difference seen between

PTPRT KO and wild-type mice, acting to increase POMC neuron activity to affect satiety

(Bates, Stearns et al. 2003, Xu, Ste-Marie et al. 2007). Food intake as a mechanism for weight gain can occur on a couple different levels. For example, mice could fail to match energy intake with energy expenditure, modulate food intake in response to environmental stressors/perturbations, inappropriately enjoy food consumption and/or fail to adequately sense satiation within a meal (Ellacott, Morton et al. 2010). More sensitive assays that assess real-time food consumption could address whether PTPRT KO mice eat less at a given meal or have a different feeding reaction to handling.

Ultimately, subjecting wild-type and PTPRT KO mice to a pair-feeding experiment – in which both mice strains are fed the same amount of food – would decisively implicate food intake differences as causing the weight difference between these two strains

(Ellacott, Morton et al. 2010). Additionally, since normal mice become hyperphagic after exposure to a high-fat diet (Ellacott, Morton et al. 2010), repeating the food intake study on a normal diet will remove this confounding factor. However, even if pair-feeding results were inconclusive, there also are other important effects of hypothalamic STAT3 phosphorylation.

For example, increased POMC neuron activity via upregulated phospho-STAT3 will also increase ACTH secretion (Bousquet, Zatelli et al. 2000); inflammatory cytokines such as

IL-6 especially activate STAT3 in the anterior pituitary also to upregulate POMC and

ACTH signaling (Berne 2003). Increased ACTH will increase sex hormone precursor, mineralcorticoid and glucocorticoid secretion by the adrenal cortex. Most salient to the

PTPRT KO phenotype would presumably be an increase in cortisol in these mice.

Cortisol’s effect on end tissues varies whether it is secreted acutely or chronically. Over the short-term, increased cortisol will increase epinephrine/norepinephrine release, lipolysis in adipose tissue, and glycogenolysis with gluconeogenesis in the liver.

Conversely, long-term cortisol release will have the exact opposite effect (Berne 2003).

Many of the short-term effects of cortisol would be consistent with the PTPRT KO phenotype. Measuring the daily variance cortisol levels in wild-type and PTPRT KO

mice will shed light into its possible role.

Another implication of increased hypothalamic STAT3 is increased hepatic insulin

sensitivity. In leptin-knockout mice, restoration of leptin potentiates phospho-STAT3 and reverts hepatic insulin resistance (Pocai, Morgan et al. 2005). Conversely, inhibiting

the activation of hypothalamic STAT3 increases hepatic insulin resistance (Buettner,

Pocai et al. 2006). The vagus nerve mediates the effect of leptin on hepatic insulin

sensitivity (German, Kim et al. 2009). As such, the disconnect between increased hypothalamic phospho-STAT3 and hepatic insulin sensitivity in the PTPRT KO mouse

calls into question the effectiveness of neuroendocrine signals between the nervous

system and the end organ in these mice, potentially implicating a defect in vagal-liver contacts.

What is the in vivo role of PTPRT?

In order to fully explain the relationship between PTPRT and nutrient metabolism, one

must reconcile this protein’s independent effect on a variety of tissues without being

significantly expressed in them. As verified by our study, PTPRT expression is largely

restricted to the central nervous system (McAndrew, Frostholm et al. 1998, Ensslen-Craig

and Brady-Kalnay 2004), but also has been identified in the colon (Zhao, Zhang et al.

2010).

Given the complex relationship between the brain and glucose homeostasis, PTPRT KO could possibly have a selective effect on some aspects of the brain’s response to high-fat diet and glucose homeostasis but not others (Schwartz and Porte 2005), reconciling the mixed insulin resistance phenotype.

Additionally, considering its effect on NPY secretion and hepatic insulin resistance,

PTPRT could modulate how the nervous system interfaces with end organs, being expressed in only a subset of cells that directly interface with neurons, which then disseminate a signal to neighboring cells (Figure 4-3). The hypothesis that PTPRT affects neuronal-end organ communication may extend PTPRT’s tumor suppressive role to include how the nervous system can promote tumors. Adrenergic signaling mediated by the sympathetic nervous system causes increased angiogenesis and metastasis along with decreased immune cell activity (Cole and Sood 2012). As it relates to colon cancer, enteric neurons normally influence growth of colonic epithelial cells, especially in

94 relation to adequate intestinal epithelial barrier function (Neunlist, Van Landeghem et al.

2013). Moreover, some studies have correlated an increase in norepinephrine levels

(Tutton and Barkla 1977, Tatsuta, Iishi et al. 1989, Tatsuta, Iishi et al. 1990, Iishi, Tatsuta

Figure 4-3. Effect of PTPRT KO on end organs.

PTPRT KO has varied effects on organs in regard to the high-fat diet phenotype and azoxymethane-induced cancer. Question marks indicate ambiguity if these are centrally- mediated effects or if PTPRT mediates cell-cell contacts between neurons and end organs. (Pictured: brain, adipose tissue, adrenal medulla, liver and intestines; brain and intestines taken from (Berne 2003)) et al. 1992, Tatsuta, Iishi et al. 1992, Tatsuta, Iishi et al. 1992, Iishi, Tatsuta et al. 1994),

dopamine antagonists (Iishi, Baba et al. 1991) and/or adrenergic signaling (Tatsuta, Iishi

et al. 1988) with azoxymethane-induced colon tumors in rats. Taken as a whole, PTPRT may help regulate proper proliferation of colonic epithelium in response to nervous system signaling, therefore preventing adenoma formation.

Paxillin and cancer: Future Directions

Why is pY88 paxillin important therapeutically?

The combination of cell line dasatinib sensitivity data and patient immunohistochemistry data demonstrate an in vivo application for pY88 paxillin. Since cell lines with increased pY88 paxillin have increased sensitivity to dasatinib, human clinical trials should assess if tumors with high pY88 paxillin are similarly sensitive to these drugs as well. Given the side effect profile of dasatinib (Talpaz, Shah et al. 2006), it will be crucial to understand what CRC patients will benefit most from SFK inhibition (Baselga 2006).

pY88 paxillin can serve as a biomarker to these ends, serving as a readout of overactive

Src and loss of PTPRT inhibition.

pY88 paxillin also helps transduce signals from PI3K, an important oncogenic pathway.

A therapeutic drawback in inhibiting PI3K is the various p110 isoforms that have normal

in vivo function (Liu, Cheng et al. 2009). Instead, drugging pY88 paxillin will indirectly

attenuate those p85alpha-stimulated isoforms that are most likely acting in an oncogenic

fashion while sparing the other ones necessary for normal functioning. Given the genetic

background of DLD1 and HCT116 (Samuels, Diaz et al. 2005), pY88 paxillin seemingly affects p110alpha isoforms that are constitutively activated via hotspot mutations

(Ikenoue, Kanai et al. 2005). However, p85alpha binding to tyrosine phosphorylation ligands still impacts oncogenic PI3K signaling, so drugs that target aberrant tyrosine phosphorylation also can influence PI3K (Blume-Jensen and Hunter 2001, Brognard and

Hunter 2011).

The Src kinase inhibitor data may point to additional kinases that could phosphorylate paxillin at Y88. Saracatinib is a much more specific Src family kinase (SFK) inhibitor than the archetype compound dasatinib, which is important in reducing side effects given the role of SFK members in immune cell development (Chen, Elfiky et al. 2014).

Combined with the kinase prediction software, the saracatinib data suggest Abl and Lck as other putative kinases for pY88 paxillin (Hennequin, Allen et al. 2006). While many

SFKs are implicated in a variety of cancers, especially Abl in leukemias (Kim, Song et al.

2009), Src is the predominantly acting SFK in colorectal cancer, meaning it is most likely to be the SFK that phosphorylates pY88 paxillin in vivo (Bolen, Veillette et al. 1987,

Cartwright, Kamps et al. 1989, Blume-Jensen and Hunter 2001, Kim, Song et al. 2009,

Sirvent, Benistant et al. 2012, Chen, Elfiky et al. 2014)).

An intriguing finding from the tumor microarray is the greater proportion of metastatic liver tumors that demonstrated low pY88 paxillin staining when compared to primary

Stage IV tumors. Not all every cancer cell necessarily has the requisite changes that are

required for successful metastasis, meaning that the pro-growth role of pY88 paxillin could be distinct from (and potentially incongruous with) metastatic dissemination of disease (Fidler 2003). Many possible hypotheses could explain this finding. First, it suggests that pY88 paxillin could be responsive to the differences between colon and

liver as it relates to the growth factor, stromal cell and/or nutrient milieu (Boman and

Huang 2008). Also, Src activity levels have been shown to correlate with tumor cell

differentiation (Yeatman 2004), meaning that poorly differentiated cells (which are more

likely to metastasize) have less active Src and therefore less pY88 paxillin. Additionally,

given the role of paxillin phosphorylation in response to the extracellular matrix (Schaller

and Schaefer 2001) and the differences between the collagen content of the colon and

liver (Paschos et al. 2014, Yoshimura et al. 2009, Nystrom et al. 2012), the tumor

microenvironment may play a crucial role in the regulation of pY88 paxillin. Taken as a

whole, clinical data from CRC patients may ultimately inform the true in vivo role of

pY88 paxillin.

How does pY88 paxillin affect p130cas phosphorylation?

We hypothesize that pY88 paxillin’s effect on tumorigenesis mainly involves the

upregulation of p130cas phosphorylation. Therefore, another line of investigation into

rescuing the paxillin Y88F poor growth phenotype should address the specific process by

which p130cas becomes phosphorylated at tyrosine-165 and tyrosine-128, but not at other tyrosine residues on the protein (Zhao, Zhang et al. 2010). Src heavily phosphorylates

p130cas (Goldberg, Alexander et al. 2003, Patwardhan, Shen et al. 2006, Reynolds,

Kanner et al. 2013) so how pY88 paxillin specifically affects certain residues presents an intriguing question. This disparity can be the result of kinase action or phosphatase inactivation, especially because these two tyrosines have the same consensus sequence

(YQVP, compared to the Src consensus YxxP at other residues). Using kinase prediction software (Kim, Song et al. 2009), Csk is predicted to phosphorylate these sites but not other tyrosines to the same extent. However, a problematic aspect of implicating Csk is its inactivation of Src at tyrosine-527, which should ultimately attenuate tumorigenic processes (Kim, Song et al. 2009, Chen, Elfiky et al. 2014). Given the frequency in CRC

of either Src overexpression or Src truncation mutants (Irby, Mao et al. 1999) that render

tyrosine-527 useless, Csk inactivation of Src could potentially be a non-issue.

Conversely, the homology between these two residue sites implicates phosphatase activity as well. PTPN14 dephosphorylates p130cas at Y128 (Zhang, Guo et al. 2013), an important phosphorylation event in tumorigenesis; pY128 could act in concert with pY165. Either hypothesis is consistent with the known function of paxillin. Csk contains

Src homology domains, meaning it can easily dock with paxillin in close proximity to the

Y88 residue. Similarly, PTPN14 is predicted to bind cytoskeletal elements and localizes to cell-cell contacts, putting it in close approximation with the actin regulator and focal adhesion protein paxillin. Further experiments should either overexpress Csk or knock down PTPN14 in the paxillin Y88F cell lines to see if signaling defects and/or the slow-

growing phenotype are rescued.

What are alternative pathways through which pY88 paxillin can act?

While we focused largely on the implications of PI3K signaling via p130cas, pY88

paxillin affects additional proteins involved in tumorigenesis (Figure 4-4). Constitutively active Akt (Manning and Cantley 2007) had difficulty in rescuing the paxillin Y88F growth defect (data not shown), suggesting another pathway involved in mediating pY88 paxillin’s effect on tumorigenicity. The p130cas-p85 interaction provides a solid lead for investigating this matter further. First, it may implicate Akt-independent mechanisms of

PI3K tumorigenesis. For example, there are instances of activating mutations in p110alpha without overactive Akt; tumorigenic flux then travels through membrane- localized PDK1 (Vasudevan, Barbie et al. 2009). Additionally, p85alpha can act independently from PI3K catalytic activity to affect cytokinesis (Garcia, Silio et al.

2006): cells without p85alpha accumulate in telophase, fail to localize important components of cytokinesis and have an increased population of binucleated cells.

Curiously, HCT Y88F cells have a similar phenotype, but these findings were not recapitulated in DLD Y88F counterparts (data not shown).

Another important oncogene implicated in pY88 paxillin signaling is SHP-2, which is activated in parental cells more than in paxillin Y88F mutant cells. While many PTPs act as tumor suppressors, this phosphatase frequently serves as an oncogene via Ras/Erk pathway activation (Chan, Kalaitzidis et al. 2008) and other growth factor pathways (Qu

2000). Activating mutations in this protein are associated with Noonan syndrome, leukemias and colon cancer (Bentires-Alj, Paez et al. 2004). Salient to pY88 paxillin

100

signaling, SHP-2 also upregulates Akt1 phosphorylation via PI3K activity (Hakak, Hsu et

al. 2000, Wu, O'Rourke et al. 2001). Therefore, increasing SHP-2 expression in paxillin

Y88F mutants may rescue the slow-growth phenotype.

Figure 4-4. Extended pY88 paxillin signaling model.

Additional options for Akt1 activation and/or tumorigenesis in response to pY88 paxillin are outlined here. A hypothetical protein tyrosine kinase (PTK) can act on either p130cas, SHP-2 and/or Akt1 directly. SHP-2 and p130cas activate PI3K to effect Akt1 activation or contribute to tumorigenesis independently of Akt1.

Our model suggests that pY88 paxillin promotes Akt1 activation via regulating PI3K.

However, another intriguing possibility is that pY88 paxillin could also affect Akt1 phosphorylation through non-canonical methods. Specifically, many PI3K-independent

Akt1 activation mechanisms act via tyrosine phosphorylation to upregulate the canonical

101

Thr308/Ser473 phosphorylation residues (Chen, Kim et al. 2001, Jiang and Qiu 2003,

Mahajan, Coppola et al. 2010, Mahajan and Mahajan 2012). Given the differential

tyrosine phosphorylation of p130cas and SHP-2, pY88 paxillin could also be a requirement for Akt1 tyrosine phosphorylation to maximize its activation.

102

Bibliography

1. Ahima RS. Digging deeper into obesity. J Clin Invest 2011; 121: 2076-9.

2. Aligayer H, Boyd DD, Heiss MM, Abdalla EK, Curley SA, Gallick GE. Activation of Src kinase in primary colorectal carcinoma: an indicator of poor clinical prognosis. Cancer 2002; 94: 344-51.

3. Alonso A, Sasin J, Bottini N, Friedberg I, Friedberg I, Osterman A et al. Protein tyrosine phosphatases in the human genome. Cell 2004; 117: 699-711.

4. Altomare DA, Khaled AR. Homeostasis and the importance for a balance between AKT/mTOR activity and intracellular signaling. Curr Med Chem 2012; 19: 3748-62.

5. American Diabetes A. Diagnosis and classification of diabetes mellitus. Diabetes Care 2005; 28 Suppl 1: S37-42.

6. Andersen JN, Jansen PG, Echwald SM, Mortensen OH, Fukada T, Del Vecchio R et al. A genomic perspective on protein tyrosine phosphatases: gene structure, pseudogenes, and genetic disease linkage. FASEB J 2004; 18: 8-30.

7. Arena S, Benvenuti S, Bardelli A. Genetic analysis of the kinome and phosphatome in cancer. Cell Mol Life Sci 2005; 62: 2092-9.

8. Aricescu AR, Hon WC, Siebold C, Lu W, van der Merwe PA, Jones EY. Molecular analysis of receptor protein tyrosine phosphatase mu-mediated cell adhesion. EMBO J 2006; 25: 701-12.

9. Aricescu AR, Siebold C, Choudhuri K, Chang VT, Lu W, Davis SJ et al. Structure of a tyrosine phosphatase adhesive interaction reveals a spacer-clamp mechanism. Science 2007; 317: 1217-20.

10. Azuma K, Tanaka M, Uekita T, Inoue S, Yokota J, Ouchi Y et al. Tyrosine phosphorylation of paxillin affects the metastatic potential of human osteosarcoma. Oncogene 2005; 24: 4754-64.

11. Bachman KE, Argani P, Samuels Y, Silliman N, Ptak J, Szabo S et al. The PIK3CA gene is mutated with high frequency in human breast cancers. Cancer Biol Ther 2004; 3: 772-5.

12. Bader AG, Kang S, Zhao L, Vogt PK. Oncogenic PI3K deregulates transcription and translation. Nat Rev Cancer 2005; 5: 921-9.

103

13. Bagherian A, Kalhori KA, Sadeghi M, Mirhosseini F, Parisay I. An in vitro study of root and canal morphology of human deciduous molars in an Iranian population. J Oral Sci 2010; 52: 397-403.

14. Bai Y, Zhang S, Kim KS, Lee JK, Kim KH. Obese gene expression alters the ability of 30A5 preadipocytes to respond to lipogenic hormones. J Biol Chem 1996; 271: 13939-42.

15. Bardelli A, Parsons DW, Silliman N, Ptak J, Szabo S, Saha S et al. Mutational analysis of the tyrosine kinome in colorectal cancers. Science 2003; 300: 949.

16. Bardelli A, Parsons DW, Silliman N, Ptak J, Szabo S, Saha S et al. Mutational Analysis of the Tyrosine Kinome in Colorectal Cancers. Science 2003; 300: 949- 949.

17. Barr AJ, Ugochukwu E, Lee WH, King ON, Filippakopoulos P, Alfano I et al. Large-scale structural analysis of the classical human protein tyrosine phosphatome. Cell 2009; 136: 352-63.

18. Barsh GS, Farooqi IS, O'Rahilly S. Genetics of body-weight regulation. Nature 2000; 404: 644-51.

19. Baselga J. Targeting tyrosine kinases in cancer: the second wave. Science 2006; 312: 1175-8.

20. Bates SH, Stearns WH, Dundon TA, Schubert M, Tso AW, Wang Y et al. STAT3 signalling is required for leptin regulation of energy balance but not reproduction. Nature 2003; 421: 856-9.

21. Beck B. Neuropeptide Y in normal eating and in genetic and dietary-induced obesity. Philos Trans R Soc Lond B Biol Sci 2006; 361: 1159-85.

22. Becka S, Zhang P, Craig SE, Lodowski DT, Wang Z, Brady-Kalnay SM. Characterization of the adhesive properties of the type IIb subfamily receptor protein tyrosine phosphatases. Cell Commun Adhes 2010; 17: 34-47.

23. Bellis SL, Miller JT, Turner CE. Characterization of tyrosine phosphorylation of paxillin in vitro by focal adhesion kinase. J Biol Chem 1995; 270: 17437-41.

24. Bence KK, Delibegovic M, Xue B, Gorgun CZ, Hotamisligil GS, Neel BG et al. Neuronal PTP1B regulates body weight, adiposity and leptin action. Nat Med 2006; 12: 917-24.

25. Bentires-Alj M, Paez JG, David FS, Keilhack H, Halmos B, Naoki K et al. Activating mutations of the noonan syndrome-associated SHP2/PTPN11 gene in

104

human solid tumors and adult acute myelogenous leukemia. Cancer Res 2004; 64: 8816-20.

26. Bento JL, Palmer ND, Mychaleckyj JC, Lange LA, Langefeld CD, Rich SS et al. Association of protein tyrosine phosphatase 1B gene polymorphisms with type 2 diabetes. Diabetes 2004; 53: 3007-12.

27. Berne RL, MN; Koeppen, BM; Stanton, BA. Physiology, 5th edn Mosby: Maryland Hts, MO, 2003.

28. Berrington de Gonzalez A, Hartge P, Cerhan JR, Flint AJ, Hannan L, MacInnis RJ et al. Body-mass index and mortality among 1.46 million white adults. N Engl J Med 2010; 363: 2211-9.

29. Berry DC, Jacobs H, Marwarha G, Gely-Pernot A, O'Byrne SM, DeSantis D et al. The STRA6 receptor is essential for retinol-binding protein-induced insulin resistance but not for maintaining vitamin A homeostasis in tissues other than the eye. J Biol Chem 2013; 288: 24528-39.

30. Besco J, Popesco MC, Davuluri RV, Frostholm A, Rotter A. Genomic structure and alternative splicing of murine R2B receptor protein tyrosine phosphatases (PTPkappa, mu, rho and PCP-2). BMC Genomics 2004; 5: 14.

31. Besco JA, Frostholm A, Popesco MC, Burghes AH, Rotter A. Genomic organization and alternative splicing of the human and mouse RPTPrho genes. BMC Genomics 2001; 2: 1.

32. Besco JA, Hooft van Huijsduijnen R, Frostholm A, Rotter A. Intracellular substrates of brain-enriched receptor protein tyrosine phosphatase rho (RPTPrho/PTPRT). Brain Res 2006; 1116: 50-7.

33. Bi S, Kim YJ, Zheng F. Dorsomedial hypothalamic NPY and energy balance control. Neuropeptides 2012; 46: 309-14.

34. Biddinger SB, Hernandez-Ono A, Rask-Madsen C, Haas JT, Aleman JO, Suzuki R et al. Hepatic insulin resistance is sufficient to produce dyslipidemia and susceptibility to atherosclerosis. Cell Metab 2008; 7: 125-34.

35. Blume-Jensen P, Hunter T. Oncogenic kinase signalling. Nature 2001; 411: 355- 65.

36. Bolen JB, Veillette A, Schwartz AM, DeSeau V, Rosen N. Activation of pp60c- src protein kinase activity in human colon carcinoma. Proc Natl Acad Sci U S A 1987; 84: 2251-5.

105

37. Boman BM, Huang E. Human colon cancer stem cells: a new paradigm in gastrointestinal oncology. J Clin Oncol 2008; 26: 2828-38.

38. Borecki IB, Rice T, Perusse L, Bouchard C, Rao DC. An exploratory investigation of genetic linkage with body composition and fatness phenotypes: the Quebec Family Study. Obes Res 1994; 2: 213-9.

39. Bornstein SR, Schuppenies A, Wong ML, Licinio J. Approaching the shared biology of obesity and depression: the stress axis as the locus of gene- environment interactions. Mol Psychiatry 2006; 11: 892-902.

40. Bousquet C, Zatelli MC, Melmed S. Direct regulation of pituitary proopiomelanocortin by STAT3 provides a novel mechanism for immuno- neuroendocrine interfacing. J Clin Invest 2000; 106: 1417-25.

41. Brady-Kalnay SM. Protein tyrosine phosphatases. In: Beckerle M (ed) Cell Adhesion: Frontiers in Molecular Biology. Oxford University Press: Oxford, UK, 2001.

42. Bray GA. Soft drink consumption and obesity: it is all about fructose. Curr Opin Lipidol 2010; 21: 51-7.

43. Broderick DK, Di C, Parrett TJ, Samuels YR, Cummins JM, McLendon RE et al. Mutations of PIK3CA in anaplastic oligodendrogliomas, high-grade astrocytomas, and medulloblastomas. Cancer Res 2004; 64: 5048-50.

44. Brognard J, Hunter T. Protein kinase signaling networks in cancer. Curr Opin Genet Dev 2011; 21: 4-11.

45. Brothers SP, Wahlestedt C. Therapeutic potential of neuropeptide Y (NPY) receptor ligands. EMBO Mol Med 2010; 2: 429-39.

46. Brown MS, Goldstein JL. Selective versus total insulin resistance: a pathogenic paradox. Cell Metab 2008; 7: 95-6.

47. Brownlee M. Biochemistry and molecular cell biology of diabetic complications. Nature 2001; 414: 813-20.

48. Buchner DA, Geisinger JM, Glazebrook PA, Morgan MG, Spiezio SH, Kaiyala KJ et al. The juxtaparanodal proteins CNTNAP2 and TAG1 regulate diet-induced obesity. Mamm Genome 2012; 23: 431-42.

49. Buettner C, Pocai A, Muse ED, Etgen AM, Myers MG, Jr., Rossetti L. Critical role of STAT3 in leptin's metabolic actions. Cell Metab 2006; 4: 49-60.

106

50. Burden-Gulley SM, Gates TJ, Burgoyne AM, Cutter JL, Lodowski DT, Robinson S et al. A novel molecular diagnostic of glioblastomas: detection of an extracellular fragment of protein tyrosine phosphatase mu. Neoplasia 2010; 12: 305-16.

51. Burgoyne AM, Palomo JM, Phillips-Mason PJ, Burden-Gulley SM, Major DL, Zaremba A et al. PTPmu suppresses glioma cell migration and dispersal. Neuro Oncol 2009; 11: 767-78.

52. Burgoyne AM, Phillips-Mason PJ, Burden-Gulley SM, Robinson S, Sloan AE, Miller RH et al. Proteolytic cleavage of protein tyrosine phosphatase mu regulates glioblastoma cell migration. Cancer Res 2009; 69: 6960-8.

53. Calle EE, Thun MJ. Obesity and cancer. Oncogene 2004; 23: 6365-78.

54. Campbell IG, Russell SE, Choong DY, Montgomery KG, Ciavarella ML, Hooi CS et al. Mutation of the PIK3CA gene in ovarian and breast cancer. Cancer Res 2004; 64: 7678-81.

55. Carey M, Kehlenbrink S, Hawkins M. Evidence for central regulation of glucose metabolism. J Biol Chem 2013; 288: 34981-8.

56. Carter BJ, Anklesaria P, Choi S, Engelhardt JF. Redox modifier genes and pathways in amyotrophic lateral sclerosis. Antioxid Redox Signal 2009; 11: 1569- 86.

57. Cartwright CA, Kamps MP, Meisler AI, Pipas JM, Eckhart W. pp60c-src activation in human colon carcinoma. J Clin Invest 1989; 83: 2025-33.

58. Cernkovich ER, Deng J, Bond MC, Combs TP, Harp JB. Adipose-specific disruption of signal transducer and activator of transcription 3 increases body weight and adiposity. Endocrinology 2008; 149: 1581-90.

59. Cernkovich ER, Deng J, Bond MC, Combs TP, Harp JB. Adipose-specific disruption of signal transducer and activator of transcription 3 increases body weight and adiposity. Endocrinology 2008; 149: 1581-90.

60. Champe PH, RA; Ferrier, DR. Lippincott's Illustrated Reviews: Biochemistry, 4th edn Lippincott Williams & Wilkins: Philadelpha, PA, 2007.

61. Chan G, Kalaitzidis D, Neel BG. The tyrosine phosphatase Shp2 (PTPN11) in cancer. Cancer Metastasis Rev 2008; 27: 179-92.

62. Chen J, Elfiky A, Han M, Chen C, Saif MW. The role of Src in colon cancer and its therapeutic implications. Clin Colorectal Cancer 2014; 13: 5-13.

107

63. Chen J, Gallo KA. MLK3 regulates paxillin phosphorylation in chemokine- mediated breast cancer cell migration and invasion to drive metastasis. Cancer Res 2012; 72: 4130-40.

64. Chen R, Kim O, Yang J, Sato K, Eisenmann KM, McCarthy J et al. Regulation of Akt/PKB activation by tyrosine phosphorylation. J Biol Chem 2001; 276: 31858- 62.

65. Cohen P. The origins of protein phosphorylation. Nat Cell Biol 2002; 4: E127-30.

66. Cole SW, Sood AK. Molecular pathways: beta-adrenergic signaling in cancer. Clin Cancer Res 2012; 18: 1201-6.

67. Collier LS, Carlson CM, Ravimohan S, Dupuy AJ, Largaespada DA. Cancer gene discovery in solid tumours using transposon-based somatic mutagenesis in the mouse. Nature 2005; 436: 272-6.

68. Compton CH, E; Grochow, L; Lee, F; Ritter, M; Neiderhuber, JE. Colon Cancer. . In: Abelloff MD AJ, Niederhuber JE, Kastan MB, McKenna WG (ed) Abeloff's Clinical Oncology. Churchill Livingstone: Philadelphia (PA), 2008, pp 1477- 1534.

69. Cui Y, Huang L, Elefteriou F, Yang G, Shelton JM, Giles JE et al. Essential role of STAT3 in body weight and glucose homeostasis. Mol Cell Biol 2004; 24: 258- 69.

70. de Kloet ER, Joels M, Holsboer F. Stress and the brain: from adaptation to disease. Nat Rev Neurosci 2005; 6: 463-75.

71. Deakin NO, Pignatelli J, Turner CE. Diverse roles for the paxillin family of proteins in cancer. Genes Cancer 2012; 3: 362-70.

72. Deakin NO, Turner CE. Paxillin comes of age. J Cell Sci 2008; 121: 2435-44.

73. Deakin NO, Turner CE. Distinct roles for paxillin and Hic-5 in regulating breast cancer cell morphology, invasion, and metastasis. Mol Biol Cell 2011; 22: 327- 41.

74. Deramaudt TB, Dujardin D, Noulet F, Martin S, Vauchelles R, Takeda K et al. Altering FAK-paxillin interactions reduces adhesion, migration and invasion processes. PLoS One 2014; 9: e92059.

75. Dulak AM, Stojanov P, Peng S, Lawrence MS, Fox C, Stewart C et al. Exome and whole-genome sequencing of esophageal adenocarcinoma identifies recurrent driver events and mutational complexity. Nat Genet 2013; 45: 478-86.

108

76. Dunckley T, Huentelman MJ, Craig DW, Pearson JV, Szelinger S, Joshipura K et al. Whole-genome analysis of sporadic amyotrophic lateral sclerosis. N Engl J Med 2007; 357: 775-88.

77. Elbein SC, Kern PA, Rasouli N, Yao-Borengasser A, Sharma NK, Das SK. Global gene expression profiles of subcutaneous adipose and muscle from glucose-tolerant, insulin-sensitive, and insulin-resistant individuals matched for BMI. Diabetes 2011; 60: 1019-29.

78. Ellacott KL, Morton GJ, Woods SC, Tso P, Schwartz MW. Assessment of feeding behavior in laboratory mice. Cell Metab 2010; 12: 10-7.

79. Elmquist JK, Maratos-Flier E, Saper CB, Flier JS. Unraveling the central nervous system pathways underlying responses to leptin. Nat Neurosci 1998; 1: 445-50.

80. Elmquist JK, Marcus JN. Rethinking the central causes of diabetes. Nat Med 2003; 9: 645-7.

81. Ensslen-Craig SE, Brady-Kalnay SM. Receptor protein tyrosine phosphatases regulate neural development and axon guidance. Dev Biol 2004; 275: 12-22.

82. Esler M, Rumantir M, Wiesner G, Kaye D, Hastings J, Lambert G. Sympathetic nervous system and insulin resistance: from obesity to diabetes. Am J Hypertens 2001; 14: 304S-309S.

83. Fearon ER, Pardoll DM, Itaya T, Golumbek P, Levitsky HI, Simons JW et al. Interleukin-2 production by tumor cells bypasses T helper function in the generation of an antitumor response. Cell 1990; 60: 397-403.

84. Fidler IJ. The pathogenesis of cancer metastasis: the 'seed and soil' hypothesis revisited. Nat Rev Cancer 2003; 3: 453-8.

85. Fishel R, Lescoe MK, Rao MR, Copeland NG, Jenkins NA, Garber J et al. The human mutator gene homolog MSH2 and its association with hereditary nonpolyposis colon cancer. Cell 1993; 75: 1027-38.

86. Flavell JR, Baumforth KR, Wood VH, Davies GL, Wei W, Reynolds GM et al. Down-regulation of the TGF-beta target gene, PTPRK, by the Epstein-Barr virus encoded EBNA1 contributes to the growth and survival of Hodgkin lymphoma cells. Blood 2008; 111: 292-301.

87. Flegal KM, Graubard BI, Williamson DF, Gail MH. Cause-specific excess deaths associated with underweight, overweight, and obesity. JAMA 2007; 298: 2028-37.

109

88. Garcia Z, Silio V, Marques M, Cortes I, Kumar A, Hernandez C et al. A PI3K activity-independent function of p85 regulatory subunit in control of mammalian cytokinesis. EMBO J 2006; 25: 4740-51.

89. Geerling JJ, Boon MR, Kooijman S, Parlevliet ET, Havekes LM, Romijn JA et al. Sympathetic nervous system control of triglyceride metabolism: novel concepts derived from recent studies. J Lipid Res 2014; 55: 180-9.

90. German J, Kim F, Schwartz GJ, Havel PJ, Rhodes CJ, Schwartz MW et al. Hypothalamic leptin signaling regulates hepatic insulin sensitivity via a neurocircuit involving the vagus nerve. Endocrinology 2009; 150: 4502-11.

91. Gerozissis K. Brain insulin, energy and glucose homeostasis; genes, environment and metabolic pathologies. Eur J Pharmacol 2008; 585: 38-49.

92. Gilmore AP. Anoikis. Cell Death Differ 2005; 12 Suppl 2: 1473-7.

93. Glenney JR, Jr., Zokas L. Novel tyrosine kinase substrates from Rous sarcoma virus-transformed cells are present in the membrane skeleton. J Cell Biol 1989; 108: 2401-8.

94. Goldberg GS, Alexander DB, Pellicena P, Zhang ZY, Tsuda H, Miller WT. Src phosphorylates Cas on tyrosine 253 to promote migration of transformed cells. J Biol Chem 2003; 278: 46533-40.

95. Gong L, Yao F, Hockman K, Heng HH, Morton GJ, Takeda K et al. Signal transducer and activator of transcription-3 is required in hypothalamic agouti- related protein/neuropeptide Y neurons for normal energy homeostasis. Endocrinology 2008; 149: 3346-54.

96. Grady WM, Carethers JM. Genomic and epigenetic instability in colorectal cancer pathogenesis. Gastroenterology 2008; 135: 1079-99.

97. Groop PH, Forsblom C, Thomas MC. Mechanisms of disease: Pathway-selective insulin resistance and microvascular complications of diabetes. Nat Clin Pract Endocrinol Metab 2005; 1: 100-10.

98. Gryfe R. Inherited colorectal cancer syndromes. Clin Colon Rectal Surg 2009; 22: 198-208.

99. Guo G, Sun X, Chen C, Wu S, Huang P, Li Z et al. Whole-genome and whole- exome sequencing of bladder cancer identifies frequent alterations in genes involved in sister chromatid cohesion and segregation. Nat Genet 2013; 45: 1459- 63.

110

100. Hagel M, George EL, Kim A, Tamimi R, Opitz SL, Turner CE et al. The adaptor protein paxillin is essential for normal development in the mouse and is a critical transducer of fibronectin signaling. Mol Cell Biol 2002; 22: 901-15.

101. Hakak Y, Hsu YS, Martin GS. Shp-2 mediates v-Src-induced morphological changes and activation of the anti-apoptotic protein kinase Akt. Oncogene 2000; 19: 3164-71.

102. Han NM, Curley SA, Gallick GE. Differential activation of pp60(c-src) and pp62(c-yes) in human colorectal carcinoma liver metastases. Clin Cancer Res 1996; 2: 1397-404.

103. Han R, Li A, Li L, Kitlinska JB, Zukowska Z. Maternal low-protein diet upregulates the neuropeptide Y system in visceral fat and leads to abdominal obesity and glucose intolerance in a sex- and time-specific manner. FASEB J 2012; 26: 3528-36.

104. Hanahan D, Weinberg Robert A. Hallmarks of Cancer: The Next Generation. Cell 2011; 144: 646-674.

105. Hao Y, Wang C, Cao B, Hirsch BM, Song J, Markowitz SD et al. Gain of interaction with IRS1 by p110alpha-helical domain mutants is crucial for their oncogenic functions. Cancer Cell 2013; 23: 583-93.

106. Haque A, Andersen JN, Salmeen A, Barford D, Tonks NK. Conformation-sensing antibodies stabilize the oxidized form of PTP1B and inhibit its phosphatase activity. Cell 2011; 147: 185-98.

107. Haque MS, Minokoshi Y, Hamai M, Iwai M, Horiuchi M, Shimazu T. Role of the sympathetic nervous system and insulin in enhancing glucose uptake in peripheral tissues after intrahypothalamic injection of leptin in rats. Diabetes 1999; 48: 1706-12.

108. Hariri N, Thibault L. High-fat diet-induced obesity in animal models. Nutr Res Rev 2010; 23: 270-99.

109. Harvey RAC, PC. Lippincott's Illustrated Reviews: Pharmacology, 4th edn Lippincott Williams & Wilkins: Philadelphia, PA, 2008.

110. Haslam DW, James WP. Obesity. Lancet 2005; 366: 1197-209.

111. Hennequin LF, Allen J, Breed J, Curwen J, Fennell M, Green TP et al. N-(5- chloro-1,3-benzodioxol-4-yl)-7-[2-(4-methylpiperazin-1-yl)ethoxy]-5- (tetrahydro-2H-pyran-4-yloxy)quinazolin-4-amine, a novel, highly selective, orally available, dual-specific c-Src/Abl kinase inhibitor. J Med Chem 2006; 49: 6465-88.

111

112. Hers I, Vincent EE, Tavare JM. Akt signalling in health and disease. Cell Signal 2011; 23: 1515-27.

113. Hirsch D, Zukowska Z. NPY and stress 30 years later: the peripheral view. Cell Mol Neurobiol 2012; 32: 645-59.

114. Hong J, Stubbins RE, Smith RR, Harvey AE, Nunez NP. Differential susceptibility to obesity between male, female and ovariectomized female mice. Nutr J 2009; 8: 11.

115. Hooper L, Abdelhamid A, Moore HJ, Douthwaite W, Skeaff CM, Summerbell CD. Effect of reducing total fat intake on body weight: systematic review and meta-analysis of randomised controlled trials and cohort studies. BMJ 2012; 345: e7666.

116. Hussain AH, M.Z.I.; Claussen, B.; Asghar, S. Type 2 Diabetes and obesity: A review. Journal of Diabetology 2010; 2.

117. Iishi H, Baba M, Tatsuta M, Okuda S, Taniguchi H. Inhibition by the dopamine antagonist haloperidol of experimental carcinogenesis induced by azoxymethane in rat colon. Cancer Res 1991; 51: 6150-2.

118. Iishi H, Tatsuta M, Baba M, Taniguchi H. Monoamine oxidase B inhibitor enhances experimental carcinogenesis in rat colon induced by azoxymethane. Cancer Lett 1994; 76: 177-83.

119. Iishi H, Tatsuta M, Baba M, Yamamoto R, Taniguchi H. Enhancement by bombesin of colon carcinogenesis and metastasis induced by azoxymethane in Wistar rats. Int J Cancer 1992; 50: 834-9.

120. Ikenoue T, Kanai F, Hikiba Y, Obata T, Tanaka Y, Imamura J et al. Functional analysis of PIK3CA gene mutations in human colorectal cancer. Cancer Res 2005; 65: 4562-7.

121. Irby RB, Mao W, Coppola D, Kang J, Loubeau JM, Trudeau W et al. Activating SRC mutation in a subset of advanced human colon cancers. Nat Genet 1999; 21: 187-90.

122. Jacob ST, Motiwala T. Epigenetic regulation of protein tyrosine phosphatases: potential molecular targets for cancer therapy. Cancer Gene Ther 2005; 12: 665- 72.

123. Jagadeeswaran R, Surawska H, Krishnaswamy S, Janamanchi V, Mackinnon AC, Seiwert TY et al. Paxillin is a target for somatic mutations in lung cancer: implications for cell growth and invasion. Cancer Res 2008; 68: 132-42.

112

124. Jagadeeswaran R, Surawska H, Krishnaswamy S, Janamanchi V, Mackinnon AC, Seiwert TY et al. Paxillin Is a Target for Somatic Mutations in Lung Cancer: Implications for Cell Growth and Invasion. Cancer research 2008; 68: 132-142.

125. Jandacek RJ, Heubi JE, Tso P. A novel, noninvasive method for the measurement of intestinal fat absorption. Gastroenterology 2004; 127: 139-44.

126. Jequier E. Pathways to obesity. Int J Obes Relat Metab Disord 2002; 26 Suppl 2: S12-7.

127. Jiang T, Qiu Y. Interaction between Src and a C-terminal proline-rich motif of Akt is required for Akt activation. J Biol Chem 2003; 278: 15789-93.

128. Johnson TO, Ermolieff J, Jirousek MR. Protein tyrosine phosphatase 1B inhibitors for diabetes. Nat Rev Drug Discov 2002; 1: 696-709.

129. Julien SG, Dube N, Hardy S, Tremblay ML. Inside the human cancer tyrosine phosphatome. Nat Rev Cancer 2011; 11: 35-49.

130. Khandekar MJ, Cohen P, Spiegelman BM. Molecular mechanisms of cancer development in obesity. Nat Rev Cancer 2011; 11: 886-95.

131. Kim LC, Song L, Haura EB. Src kinases as therapeutic targets for cancer. Nat Rev Clin Oncol 2009; 6: 587-95.

132. Kinzler KW, Vogelstein B. Lessons from hereditary colorectal cancer. Cell 1996; 87: 159-70.

133. Knudson AG, Jr. Mutation and cancer: statistical study of retinoblastoma. Proc Natl Acad Sci U S A 1971; 68: 820-3.

134. Koch L, Wunderlich FT, Seibler J, Konner AC, Hampel B, Irlenbusch S et al. Central insulin action regulates peripheral glucose and fat metabolism in mice. J Clin Invest 2008; 118: 2132-47.

135. Kumar VF, N; Abbas, A. Robbins and Cotran Pathologic Basis of Disease, 7th edn Saunders: Philadelphia, PA, 2004.

136. Kuo LE, Kitlinska JB, Tilan JU, Li L, Baker SB, Johnson MD et al. Neuropeptide Y acts directly in the periphery on fat tissue and mediates stress-induced obesity and metabolic syndrome. Nat Med 2007; 13: 803-11.

137. Laczmanska I, Karpinski P, Bebenek M, Sedziak T, Ramsey D, Szmida E et al. Protein tyrosine phosphatase receptor-like genes are frequently hypermethylated in sporadic colorectal cancer. J Hum Genet 2013; 58: 11-5.

113

138. Lam CK, Chari M, Rutter GA, Lam TK. Hypothalamic nutrient sensing activates a forebrain-hindbrain neuronal circuit to regulate glucose production in vivo. Diabetes 2011; 60: 107-13.

139. Le Gallo M, O'Hara AJ, Rudd ML, Urick ME, Hansen NF, O'Neil NJ et al. Exome sequencing of serous endometrial tumors identifies recurrent somatic mutations in chromatin-remodeling and ubiquitin ligase complex genes. Nat Genet 2012; 44: 1310-5.

140. Leary RJ, Lin JC, Cummins J, Boca S, Wood LD, Parsons DW et al. Integrated analysis of homozygous deletions, focal amplifications, and sequence alterations in breast and colorectal cancers. Proc Natl Acad Sci U S A 2008; 105: 16224-9.

141. Lee JH, Reed DR, Li WD, Xu W, Joo EJ, Kilker RL et al. Genome scan for human obesity and linkage to markers in 20q13. Am J Hum Genet 1999; 64: 196- 209.

142. Lee JW, Soung YH, Kim SY, Lee HW, Park WS, Nam SW et al. PIK3CA gene is frequently mutated in breast carcinomas and hepatocellular carcinomas. Oncogene 2005; 24: 1477-80.

143. Lembertas AV, Perusse L, Chagnon YC, Fisler JS, Warden CH, Purcell-Huynh DA et al. Identification of an obesity quantitative trait locus on mouse chromosome 2 and evidence of linkage to body fat and insulin on the human homologous region 20q. J Clin Invest 1997; 100: 1240-7.

144. Li E, Stupack DG, Brown SL, Klemke R, Schlaepfer DD, Nemerow GR. Association of p130CAS with phosphatidylinositol-3-OH kinase mediates adenovirus cell entry. J Biol Chem 2000; 275: 14729-35.

145. Li VS, Wong CW, Chan TL, Chan AS, Zhao W, Chu KM et al. Mutations of PIK3CA in gastric adenocarcinoma. BMC Cancer 2005; 5: 29.

146. Li X, Earp HS. Paxillin is tyrosine-phosphorylated by and preferentially associates with the calcium-dependent tyrosine kinase in rat liver epithelial cells. J Biol Chem 1997; 272: 14341-8.

147. Lim SH, Kwon SK, Lee MK, Moon J, Jeong DG, Park E et al. Synapse formation regulated by protein tyrosine phosphatase receptor T through interaction with cell adhesion molecules and Fyn. EMBO J 2009; 28: 3564-78.

148. Liu P, Cheng H, Roberts TM, Zhao JJ. Targeting the phosphoinositide 3-kinase pathway in cancer. Nat Rev Drug Discov 2009; 8: 627-44.

114

149. Liu P, Morrison C, Wang L, Xiong D, Vedell P, Cui P et al. Identification of somatic mutations in non-small cell lung carcinomas using whole-exome sequencing. Carcinogenesis 2012; 33: 1270-6.

150. Loh K, Fukushima A, Zhang X, Galic S, Briggs D, Enriori PJ et al. Elevated hypothalamic TCPTP in obesity contributes to cellular leptin resistance. Cell Metab 2011; 14: 684-99.

151. Lui VW, Peyser ND, Ng PK, Hritz J, Zeng Y, Lu Y et al. Frequent mutation of receptor protein tyrosine phosphatases provides a mechanism for STAT3 hyperactivation in head and neck cancer. Proc Natl Acad Sci U S A 2014; 111: 1114-9.

152. Lynch HT, de la Chapelle A. Hereditary colorectal cancer. N Engl J Med 2003; 348: 919-32.

153. Maes HH, Neale MC, Eaves LJ. Genetic and environmental factors in relative body weight and human adiposity. Behav Genet 1997; 27: 325-51.

154. Mahajan K, Coppola D, Challa S, Fang B, Chen YA, Zhu W et al. Ack1 mediated AKT/PKB tyrosine 176 phosphorylation regulates its activation. PLoS One 2010; 5: e9646.

155. Mahajan K, Mahajan NP. PI3K-independent AKT activation in cancers: a treasure trove for novel therapeutics. J Cell Physiol 2012; 227: 3178-84.

156. Manning BD, Cantley LC. AKT/PKB signaling: navigating downstream. Cell 2007; 129: 1261-74.

157. Markowitz SD, Bertagnolli MM. Molecular origins of cancer: Molecular basis of colorectal cancer. N Engl J Med 2009; 361: 2449-60.

158. Markowitz SD, Bertagnolli MM. Molecular Basis of Colorectal Cancer. New England Journal of Medicine 2009; 361: 2449-2460.

159. Marwarha G, Berry DC, Croniger CM, Noy N. The retinol esterifying enzyme LRAT supports cell signaling by retinol-binding protein and its receptor STRA6. FASEB J 2013.

160. Marwarha G, Berry DC, Croniger CM, Noy N. The retinol esterifying enzyme LRAT supports cell signaling by retinol-binding protein and its receptor STRA6. FASEB J 2014; 28: 26-34.

161. Matthews DR, Hosker JP, Rudenski AS, Naylor BA, Treacher DF, Turner RC. Homeostasis model assessment: insulin resistance and beta-cell function from

115

fasting plasma glucose and insulin concentrations in man. Diabetologia 1985; 28: 412-9.

162. McAndrew PE, Frostholm A, White RA, Rotter A, Burghes AH. Identification and characterization of RPTP rho, a novel RPTP mu/kappa-like receptor protein tyrosine phosphatase whose expression is restricted to the central nervous system. Brain Res Mol Brain Res 1998; 56: 9-21.

163. McArdle L, Rafferty M, Maelandsmo GM, Bergin O, Farr CJ, Dervan PA et al. Protein tyrosine phosphatase genes downregulated in melanoma. J Invest Dermatol 2001; 117: 1255-60.

164. McArdle L, Rafferty MM, Satyamoorthy K, Maelandsmo GM, Dervan PA, Herlyn M et al. Microarray analysis of phosphatase gene expression in human melanoma. Br J Dermatol 2005; 152: 925-30.

165. Mesaros A, Koralov SB, Rother E, Wunderlich FT, Ernst MB, Barsh GS et al. Activation of Stat3 signaling in AgRP neurons promotes locomotor activity. Cell Metab 2008; 7: 236-48.

166. Meshkani R, Adeli K. Hepatic insulin resistance, metabolic syndrome and cardiovascular disease. Clin Biochem 2009; 42: 1331-46.

167. Millward CA, Desantis D, Hsieh CW, Heaney JD, Pisano S, Olswang Y et al. Phosphoenolpyruvate carboxykinase (Pck1) helps regulate the triglyceride/fatty acid cycle and development of insulin resistance in mice. J Lipid Res 2010; 51: 1452-63.

168. Montero JC, Seoane S, Ocana A, Pandiella A. Inhibition of SRC family kinases and receptor tyrosine kinases by dasatinib: possible combinations in solid tumors. Clin Cancer Res 2011; 17: 5546-52.

169. Morton GJ, Schwartz MW. Leptin and the central nervous system control of glucose metabolism. Physiol Rev 2011; 91: 389-411.

170. Motiwala T, Jacob ST. Role of protein tyrosine phosphatases in cancer. Prog Nucleic Acid Res Mol Biol 2006; 81: 297-329.

171. Munzberg H, Bjornholm M, Bates SH, Myers MG, Jr. Leptin receptor action and mechanisms of leptin resistance. Cell Mol Life Sci 2005; 62: 642-52.

172. Nakamura K, Yano H, Uchida H, Hashimoto S, Schaefer E, Sabe H. Tyrosine phosphorylation of paxillin alpha is involved in temporospatial regulation of paxillin-containing focal adhesion formation and F-actin organization in motile cells. J Biol Chem 2000; 275: 27155-64.

116

173. Nakamura M, Kishi M, Sakaki T, Hashimoto H, Nakase H, Shimada K et al. Novel tumor suppressor loci on 6q22-23 in primary central nervous system lymphomas. Cancer Res 2003; 63: 737-41.

174. Neunlist M, Van Landeghem L, Mahe MM, Derkinderen P, des Varannes SB, Rolli-Derkinderen M. The digestive neuronal-glial-epithelial unit: a new actor in gut health and disease. Nat Rev Gastroenterol Hepatol 2013; 10: 90-100.

175. Nystrom H, Naredi P, Berglund A, Palmqvist R, Tavelin B, Sund M. Liver- metastatic potential of colorectal cancer is related to the stromal composition of the tumour. Anticancer Res 2012; 32: 5183-91.

176. Obici S, Zhang BB, Karkanias G, Rossetti L. Hypothalamic insulin signaling is required for inhibition of glucose production. Nat Med 2002; 8: 1376-82.

177. O'Connell JB, Maggard MA, Ko CY. Colon cancer survival rates with the new American Joint Committee on Cancer sixth edition staging. J Natl Cancer Inst 2004; 96: 1420-5.

178. Oda K, Stokoe D, Taketani Y, McCormick F. High frequency of coexistent mutations of PIK3CA and PTEN genes in endometrial carcinoma. Cancer Res 2005; 65: 10669-73.

179. Ogden CL, Carroll MD, Kit BK, Flegal KM. Prevalence of childhood and adult obesity in the United States, 2011-2012. JAMA 2014; 311: 806-14.

180. Okkenhaug K, Vanhaesebroeck B. New responsibilities for the PI3K regulatory subunit p85 alpha. Sci STKE 2001; 2001: pe1.

181. Ostman A, Hellberg C, Bohmer FD. Protein-tyrosine phosphatases and cancer. Nat Rev Cancer 2006; 6: 307-20.

182. Pagotto U. Where does insulin resistance start? The brain. Diabetes Care 2009; 32 Suppl 2: S174-7.

183. Papadopoulos N, Nicolaides NC, Wei YF, Ruben SM, Carter KC, Rosen CA et al. Mutation of a mutL homolog in hereditary colon cancer. Science 1994; 263: 1625-9.

184. Paschos KA, Majeed AW, Bird NC. Natural history of hepatic metastases from colorectal cancer--pathobiological pathways with clinical significance. World J Gastroenterol 2014; 20: 3719-37.

185. Patwardhan P, Shen Y, Goldberg GS, Miller WT. Individual Cas phosphorylation sites are dispensable for processive phosphorylation by Src and anchorage- independent cell growth. J Biol Chem 2006; 281: 20689-97.

117

186. Petit V, Boyer B, Lentz D, Turner CE, Thiery JP, Valles AM. Phosphorylation of tyrosine residues 31 and 118 on paxillin regulates cell migration through an association with CRK in NBT-II cells. J Cell Biol 2000; 148: 957-70.

187. Pocai A, Morgan K, Buettner C, Gutierrez-Juarez R, Obici S, Rossetti L. Central leptin acutely reverses diet-induced hepatic insulin resistance. Diabetes 2005; 54: 3182-9.

188. Prince A, Zhang Y, Croniger C, Puchowicz M. Oxidative metabolism: glucose versus ketones. Adv Exp Med Biol 2013; 789: 323-8.

189. Qu CK. The SHP-2 tyrosine phosphatase: signaling mechanisms and biological functions. Cell Res 2000; 10: 279-88.

190. Rajagopalan H, Bardelli A, Lengauer C, Kinzler KW, Vogelstein B, Velculescu VE. Tumorigenesis: RAF/RAS oncogenes and mismatch-repair status. Nature 2002; 418: 934.

191. Reynolds AB, Kanner SB, Bouton AH, Schaller MD, Weed SA, Flynn DC et al. SRChing for the substrates of Src. Oncogene 2013.

192. Riant E, Waget A, Cogo H, Arnal JF, Burcelin R, Gourdy P. Estrogens protect against high-fat diet-induced insulin resistance and glucose intolerance in mice. Endocrinology 2009; 150: 2109-17.

193. Riggins RB, DeBerry RM, Toosarvandani MD, Bouton AH. Src-dependent association of Cas and p85 phosphatidylinositol 3'-kinase in v-crk-transformed cells. Mol Cancer Res 2003; 1: 428-37.

194. Rosen ED, Spiegelman BM. What we talk about when we talk about fat. Cell 2014; 156: 20-44.

195. Rossmeisl M, Rim JS, Koza RA, Kozak LP. Variation in type 2 diabetes--related traits in mouse strains susceptible to diet-induced obesity. Diabetes 2003; 52: 1958-66.

196. Ruohonen ST, Pesonen U, Moritz N, Kaipio K, Roytta M, Koulu M et al. Transgenic mice overexpressing neuropeptide Y in noradrenergic neurons: a novel model of increased adiposity and impaired glucose tolerance. Diabetes 2008; 57: 1517-25.

197. Ruohonen ST, Vahatalo LH, Savontaus E. Diet-induced obesity in mice overexpressing neuropeptide y in noradrenergic neurons. Int J Pept 2012; 2012: 452524.

118

198. Samuels Y, Diaz LA, Jr., Schmidt-Kittler O, Cummins JM, Delong L, Cheong I et al. Mutant PIK3CA promotes cell growth and invasion of human cancer cells. Cancer Cell 2005; 7: 561-73.

199. Samuels Y, Wang Z, Bardelli A, Silliman N, Ptak J, Szabo S et al. High frequency of mutations of the PIK3CA gene in human cancers. Science 2004; 304: 554.

200. Schaller MD, Hildebrand JD, Parsons JT. Complex formation with focal adhesion kinase: A mechanism to regulate activity and subcellular localization of Src kinases. Mol Biol Cell 1999; 10: 3489-505.

201. Schaller MD, Schaefer EM. Multiple stimuli induce tyrosine phosphorylation of the Crk-binding sites of paxillin. Biochem J 2001; 360: 57-66.

202. Scheid MP, Woodgett JR. PKB/AKT: functional insights from genetic models. Nat Rev Mol Cell Biol 2001; 2: 760-8.

203. Scherer S. Guide to the human genome, Cold Spring Harbor Laboratory Press: Cold Spring Harbor, N.Y., 2010.

204. Schreyer SA, Wilson DL, LeBoeuf RC. C57BL/6 mice fed high fat diets as models for diabetes-accelerated atherosclerosis. Atherosclerosis 1998; 136: 17- 24.

205. Schwartz MW, Porte D, Jr. Diabetes, obesity, and the brain. Science 2005; 307: 375-9.

206. Schwartz MW, Woods SC, Porte D, Jr., Seeley RJ, Baskin DG. Central nervous system control of food intake. Nature 2000; 404: 661-71.

207. Scott A, Wang Z. Tumour suppressor function of protein tyrosine phosphatase receptor-T. Biosci Rep 2011; 31: 303-7.

208. Shi YC, Baldock PA. Central and peripheral mechanisms of the NPY system in the regulation of bone and adipose tissue. Bone 2012; 50: 430-6.

209. Sirvent A, Benistant C, Roche S. Oncogenic signaling by tyrosine kinases of the SRC family in advanced colorectal cancer. Am J Cancer Res 2012; 2: 357-71.

210. Sjoblom T, Jones S, Wood LD, Parsons DW, Lin J, Barber TD et al. The consensus coding sequences of human breast and colorectal cancers. Science 2006; 314: 268-74.

211. Sorenson CM. Interaction of bcl-2 with Paxillin through its BH4 domain is important during ureteric bud branching. J Biol Chem 2004; 279: 11368-74.

119

212. Straznicky NE, Lambert GW, Masuo K, Dawood T, Eikelis N, Nestel PJ et al. Blunted sympathetic neural response to oral glucose in obese subjects with the insulin-resistant metabolic syndrome. Am J Clin Nutr 2009; 89: 27-36.

213. Strickler JH, McCall S, Nixon AB, Brady JC, Pang H, Rushing C et al. Phase I study of dasatinib in combination with capecitabine, oxaliplatin and bevacizumab followed by an expanded cohort in previously untreated metastatic colorectal cancer. Invest New Drugs 2014; 32: 330-9.

214. Stubbins RE, Najjar K, Holcomb VB, Hong J, Nunez NP. Oestrogen alters adipocyte biology and protects female mice from adipocyte inflammation and insulin resistance. Diabetes Obes Metab 2012; 14: 58-66.

215. Surwit RS, Seldin MF, Kuhn CM, Cochrane C, Feinglos MN. Control of expression of insulin resistance and hyperglycemia by different genetic factors in diabetic C57BL/6J mice. Diabetes 1991; 40: 82-7.

216. Talamonti MS, Roh MS, Curley SA, Gallick GE. Increase in activity and level of pp60c-src in progressive stages of human colorectal cancer. J Clin Invest 1993; 91: 53-60.

217. Talpaz M, Shah NP, Kantarjian H, Donato N, Nicoll J, Paquette R et al. Dasatinib in imatinib-resistant Philadelphia chromosome-positive leukemias. N Engl J Med 2006; 354: 2531-41.

218. Tatsuta M, Iishi H, Baba M, Taniguchi H. Tissue norepinephrine depletion as a mechanism for cysteamine inhibition of colon carcinogenesis induced by azoxymethane in Wistar rats. Int J Cancer 1989; 44: 1008-11.

219. Tatsuta M, Iishi H, Baba M, Taniguchi H. Enhancement of azoxymethane- induced colon carcinogenesis in spontaneously hypertensive rats. Int J Cancer 1990; 45: 957-60.

220. Tatsuta M, Iishi H, Baba M, Taniguchi H. Enhanced induction of colon carcinogenesis by azoxymethane in Wistar rats fed a low-protein diet. Int J Cancer 1992; 50: 108-11.

221. Tatsuta M, Iishi H, Baba M, Taniguchi H. Inhibition of azoxymethane-induced experimental colon carcinogenesis in Wistar rats by 6-hydroxydopamine. Int J Cancer 1992; 50: 298-301.

222. Tatsuta M, Iishi H, Yamamura H, Baba M, Taniguchi H. Inhibition by isoproterenol and neostigmine of experimental carcinogenesis in rat colon by azoxymethane. Br J Cancer 1988; 58: 619-20.

120

223. Thareja S, Aggarwal S, Bhardwaj TR, Kumar M. Protein tyrosine phosphatase 1B inhibitors: a molecular level legitimate approach for the management of diabetes mellitus. Med Res Rev 2012; 32: 459-517.

224. Thelemann A, Petti F, Griffin G, Iwata K, Hunt T, Settinari T et al. Phosphotyrosine Signaling Networks in Epidermal Growth Factor Receptor Overexpressing Squamous Carcinoma Cells. Molecular & Cellular Proteomics 2005; 4: 356-376.

225. Tinsley FC, Taicher GZ, Heiman ML. Evaluation of a quantitative magnetic resonance method for mouse whole body composition analysis. Obes Res 2004; 12: 150-60.

226. Tonks NK. Protein tyrosine phosphatases: from genes, to function, to disease. Nat Rev Mol Cell Biol 2006; 7: 833-46.

227. Turner CE, Glenney JR, Jr., Burridge K. Paxillin: a new vinculin-binding protein present in focal adhesions. J Cell Biol 1990; 111: 1059-68.

228. Tutton PJ, Barkla DH. The influence of adrenoceptor activity on cell proliferation in colonic crypt ipithelium and in colonic adenocarcinomata. Virchows Arch B Cell Pathol 1977; 24: 139-46.

229. Vaisse C, Halaas JL, Horvath CM, Darnell JE, Jr., Stoffel M, Friedman JM. Leptin activation of Stat3 in the hypothalamus of wild-type and ob/ob mice but not db/db mice. Nat Genet 1996; 14: 95-7.

230. Vasudevan KM, Barbie DA, Davies MA, Rabinovsky R, McNear CJ, Kim JJ et al. AKT-independent signaling downstream of oncogenic PIK3CA mutations in human cancer. Cancer Cell 2009; 16: 21-32.

231. Visel A, Carson J, Oldekamp J, Warnecke M, Jakubcakova V, Zhou X et al. Regulatory pathway analysis by high-throughput in situ hybridization. PLoS Genet 2007; 3: 1867-83.

232. Vivanco I, Sawyers CL. The phosphatidylinositol 3-Kinase AKT pathway in human cancer. Nat Rev Cancer 2002; 2: 489-501.

233. Vogelstein B, Kinzler KW. Cancer genes and the pathways they control. Nat Med 2004; 10: 789-99.

234. Vogelstein B, Lane D, Levine AJ. Surfing the p53 network. Nature 2000; 408: 307-10.

121

235. Wade R, Brimer N, Lyons C, Vande Pol S. Paxillin enables attachment- independent tyrosine phosphorylation of focal adhesion kinase and transformation by RAS. J Biol Chem 2011; 286: 37932-44.

236. Wang MY, Lee Y, Unger RH. Novel form of lipolysis induced by leptin. J Biol Chem 1999; 274: 17541-4.

237. Wang Z, Shen D, Parsons DW, Bardelli A, Sager J, Szabo S et al. Mutational analysis of the tyrosine phosphatome in colorectal cancers. Science 2004; 304: 1164-6.

238. Wang Z, Shen D, Parsons DW, Bardelli A, Sager J, Szabo S et al. Mutational Analysis of the Tyrosine Phosphatome in Colorectal Cancers. Science 2004; 304: 1164-1166.

239. Warne JP, Dallman MF. Stress, diet and abdominal obesity: Y? Nat Med 2007; 13: 781-3.

240. Webb DJ, Donais K, Whitmore LA, Thomas SM, Turner CE, Parsons JT et al. FAK-Src signalling through paxillin, ERK and MLCK regulates adhesion disassembly. Nat Cell Biol 2004; 6: 154-61.

241. Wei X, Walia V, Lin JC, Teer JK, Prickett TD, Gartner J et al. Exome sequencing identifies GRIN2A as frequently mutated in melanoma. Nat Genet 2011; 43: 442- 6.

242. West DB, Boozer CN, Moody DL, Atkinson RL. Dietary obesity in nine inbred mouse strains. Am J Physiol 1992; 262: R1025-32.

243. Williams KW, Elmquist JK. From neuroanatomy to behavior: central integration of peripheral signals regulating feeding behavior. Nat Neurosci 2012; 15: 1350-5.

244. Wittkowski KM, Sonakya V, Bigio B, Tonn MK, Shic F, Ascano M et al. A novel computational biostatistics approach implies impaired dephosphorylation of growth factor receptors as associated with severity of autism. Transl Psychiatry 2014; 4: e354.

245. Wood LD, Parsons DW, Jones S, Lin J, Sjoblom T, Leary RJ et al. The genomic landscapes of human breast and colorectal cancers. Science 2007; 318: 1108-13.

246. Wu CJ, O'Rourke DM, Feng GS, Johnson GR, Wang Q, Greene MI. The tyrosine phosphatase SHP-2 is required for mediating phosphatidylinositol 3-kinase/Akt activation by growth factors. Oncogene 2001; 20: 6018-25.

122

247. Wu DW, Wu TC, Wu JY, Cheng YW, Chen YC, Lee MC et al. Phosphorylation of paxillin confers cisplatin resistance in non-small cell lung cancer via activating ERK-mediated Bcl-2 expression. Oncogene 2013.

248. Xu AW, Ste-Marie L, Kaelin CB, Barsh GS. Inactivation of signal transducer and activator of transcription 3 in proopiomelanocortin (Pomc) neurons causes decreased pomc expression, mild obesity, and defects in compensatory refeeding. Endocrinology 2007; 148: 72-80.

249. Xu E, Schwab M, Marette A. Role of protein tyrosine phosphatases in the modulation of insulin signaling and their implication in the pathogenesis of obesity-linked insulin resistance. Rev Endocr Metab Disord 2013.

250. Xue Y, Ren J, Gao X, Jin C, Wen L, Yao X. GPS 2.0, a tool to predict kinase- specific phosphorylation sites in hierarchy. Mol Cell Proteomics 2008; 7: 1598- 608.

251. Yan HX, Yang W, Zhang R, Chen L, Tang L, Zhai B et al. Protein-tyrosine phosphatase PCP-2 inhibits beta-catenin signaling and increases E-cadherin- dependent cell adhesion. J Biol Chem 2006; 281: 15423-33.

252. Yeatman TJ. A renaissance for SRC. Nat Rev Cancer 2004; 4: 470-80.

253. Yoshimura K, Meckel KF, Laird LS, Chia CY, Park JJ, Olino KL et al. Integrin alpha2 mediates selective metastasis to the liver. Cancer Res 2009; 69: 7320-8.

254. Young MR, Liu SW, Meisinger J. Protein phosphatase-2A restricts migration of Lewis lung carcinoma cells by modulating the phosphorylation of focal adhesion proteins. Int J Cancer 2003; 103: 38-44.

255. Young MRI, Liu SW, Meisinger J. Protein phosphatase-2A restricts migration of Lewis lung carcinoma cells by modulating the phosphorylation of focal adhesion proteins. International Journal of Cancer 2003; 103: 38-44.

256. Yu J, Becka S, Zhang P, Zhang X, Brady-Kalnay SM, Wang Z. Tumor-derived extracellular mutations of PTPRT /PTPrho are defective in cell adhesion. Mol Cancer Res 2008; 6: 1106-13.

257. Zang ZJ, Cutcutache I, Poon SL, Zhang SL, McPherson JR, Tao J et al. Exome sequencing of gastric adenocarcinoma identifies recurrent somatic mutations in cell adhesion and chromatin remodeling genes. Nat Genet 2012; 44: 570-4.

258. Zhang L, Bijker MS, Herzog H. The neuropeptide Y system: pathophysiological and therapeutic implications in obesity and cancer. Pharmacol Ther 2011; 131: 91-113.

123

259. Zhang P, Becka S, Craig SE, Lodowski DT, Brady-Kalnay SM, Wang Z. Cancer- derived mutations in the fibronectin III repeats of PTPRT/PTPrho inhibit cell-cell aggregation. Cell Commun Adhes 2009; 16: 146-53.

260. Zhang P, Guo A, Possemato A, Wang C, Beard L, Carlin C et al. Identification and functional characterization of p130Cas as a substrate of protein tyrosine phosphatase nonreceptor 14. Oncogene 2013; 32: 2087-95.

261. Zhang X, Guo A, Yu J, Possemato A, Chen Y, Zheng W et al. Identification of STAT3 as a substrate of receptor protein tyrosine phosphatase T. Proc Natl Acad Sci U S A 2007; 104: 4060-4.

262. Zhang X, Guo C, Chen Y, Shulha HP, Schnetz MP, LaFramboise T et al. Epitope tagging of endogenous proteins for genome-wide ChIP-chip studies. Nat Meth 2008; 5: 163-165.

263. Zhang Y, Siebert R, Matthiesen P, Yang Y, Ha H, Schlegelberger B. Cytogenetical assignment and physical mapping of the human R-PTP-kappa gene (PTPRK) to the putative tumor suppressor gene region 6q22.2-q22.3. Genomics 1998; 51: 309-11.

264. Zhao Y, Zhang X, Guda K, Lawrence E, Sun Q, Watanabe T et al. Identification and functional characterization of paxillin as a target of protein tyrosine phosphatase receptor T. Proc Natl Acad Sci U S A 2010; 107: 2592-7.

265. Zukowska-Grojec Z. Neuropeptide Y. A novel sympathetic stress hormone and more. Ann N Y Acad Sci 1995; 771: 219-33.

266. Zukowska-Grojec Z, Karwatowska-Prokopczuk E, Rose W, Rone J, Movafagh S, Ji H et al. Neuropeptide Y: a novel angiogenic factor from the sympathetic nerves and endothelium. Circ Res 1998; 83: 187-95.

124

Role of PTPRT in Obesity and Its Substrate Paxillin Tyrosine-88 in Colorectal Cancer� ������� ��� ��� ��������� ��������

Role of PTPRT in Obesity and Its Substrate Paxillin Tyrosine-88 in Colorectal Cancer� ��