A TARGETABLE GATA2-IGF2 AXIS CONFERS AGGRESSIVENESS IN

CHEMOTHERAPY RESISTANT PROSTATE CANCER

Samuel J. Vidal

Submitted in partial fulfillment of the

requirements for the degree of

Doctor of Philosophy

under the Executive Committee

of the Graduate School of Arts and Sciences

COLUMBIA UNIVERSITY

2015

©2014

Samuel J. Vidal

All rights reserved

ABSTRACT

A Targetable GATA2-IGF2 Axis Confers Aggressiveness in

Chemotherapy Resistant Prostate Cancer

Samuel J. Vidal

Prostate cancer is a common malignancy with nearly one million annual diagnoses

worldwide. Among a subset of patients, primary disease eventually progresses to disseminated

castration resistant prostate cancer (CRPC). In recent years, treatment modalities that improve

survival in CRPC have emerged including taxane chemotherapy and second generation androgen

signaling inhibitors, among others. Indeed, today the first line chemotherapeutic docetaxel as

well as the second line agent cabazitaxel are mainstays of treatment. However, CRPC inexorably

progresses to a chemotherapy resistant state that ultimately precedes lethality. Elucidating the

molecular determinants of aggressiveness in chemotherapy resistant CRPC may therefore

stimulate new therapeutic strategies that improve clinical outcomes. We used laboratory models

and clinical databases to identify GATA2 as a regulator of chemotherapy resistance and

tumorigenicity in this context. Whole genome expression profiling, clinical validation and

genetic screening approaches revealed that GATA2 regulates a signature of cancer progression

associated genes. Mechanistically, direct upregulation of the growth hormone IGF2 emerged as a

significant mediator of the aggressive properties regulated by GATA2. IGF2 in turn activated

IGF1R and INSR as well as a downstream polykinase program. The characterization of this

regulatory axis prompted a combination strategy whereby dual IGF1R/INSR inhibition restored

the efficacy of chemotherapy and improved survival in preclinical models. These studies reveal a

GATA2-IGF2 aggressiveness axis in chemotherapy resistant prostate cancer and identify a

therapeutic opportunity in this challenging disease.

i

TABLE OF CONTENTS

LIST OF FIGURES iv

ACKNOWLEDGEMENTS vii

CHAPTER 1: Introduction 1

1.1: The natural history and management of prostate cancer 2

1.1.1: The development of metastatic prostate cancer 2

1.1.2: The changing therapeutic landscape of metastatic prostate cancer 6

1.1.3: The role of taxane chemotherapy in metastatic CRPC 9

1.1.4: Taxane resistance in metastatic CRPC 12

1.2: The role of GATA2 in physiology and disease 15

1.2.1: The role of GATA2 in mammalian physiology 15

1.2.2: The role of GATA2 in hematopoietic malignancy 20

1.2.3: The role of GATA2 in solid tumors 21

1.3: The role of IGF2 in physiology and disease 23

1.3.1: The role of IGF2 in mammalian physiology 23

1.3.2: The regulation of IGF2 in physiology and disease 27

1.3.3: The role of IGF2 in malignancy 29

1.3.4: IGF signaling as a therapeutic target 31

CHAPTER 2: GATA2 confers aggressiveness in CRPC 35

2.1: GATA2 is upregulated in chemotherapy resistant CRPC in laboratory models

and clinical databases 36

2.1.1: Generation of a docetaxel resistant ARCaPM subline 36

2.1.2: Characterization of cross resistance to cabazitaxel in docetaxel resistant cells 37

ii

2.1.3: GATA2 is upregulated in chemotherapy resistant CRPC in laboratory

models and clinical databases 39

2.2: GATA2 regulates chemotherapy resistance and tumorigenicity 41

2.3: GATA2 regulates a signature of cancer progression associated genes 43

2.3.1: Identification and characterization of GATA2 regulated genes 43

2.3.2: Genetic screen of a subset of clinically salient GATA2 regulated genes 45

2.3.3: The consensus signature is unrelated to AR biology 49

CHAPTER 3: IGF2 significantly mediates GATA2 biology in CRPC 52

3.1: Functional validation of IGF2 as a downstream effector of GATA2 53

3.2: GATA2 directly upregulates IGF2 59

3.2.1: GATA2 binds to and activates the IGF2 P4 promoter 59

3.2.2: GATA2 and IGF2 are co-expressed during prostate cancer progression 61

3.3: IGF2 activates a polykinase program downstream of GATA2 62

3.3.1: Characterization of dephosphorylation patterns in knockdown models 62

3.3.2: Chemical inhibition of the polykinase program is functionally significant 64

3.4: Dual IGF1R/INSR inhibition improves the efficacy of chemotherapy and

survival in preclinical models 67

3.4.1: Dual IGF1R/INSR inhibition improves the efficacy of chemotherapy

in vitro 67

3.4.2: OSI-906 improves the efficacy of chemotherapy in vivo 69

3.4.3: OSI-906 improves the efficacy of chemotherapy and survival in preclinical

models of disseminated CRPC without added general drug toxicity 71

iii

CHAPTER 4: Discussion 73

4.1: Chemotherapy resistant CRPC remains an intractable clinical entity 74

4.2: The emerging role of GATA2 in cancer biology 75

4.2.1: Previous characterization of GATA2 in prostate cancer and other

malignancies 75

4.2.2: GATA2 confers aggressiveness in CRPC through a consensus signature 77

4.2.3: The increasingly complex role of GATA2 in prostate cancer 81

4.3: A novel role for IGF2 in prostate cancer 83

4.4: Therapeutic implications of the GATA2-IGF2 axis 85

MATERIALS AND METHODS 89

REFERENCES 104

iv

LIST OF FIGURES

Figure 1: Characterization of the ARCaPM-DR subline 37

Figure 2: Docetaxel resistant sublines exhibit varying degrees of cross resistance to

cabazitaxel 38

Figure 3: GATA2 is upregulated in three models of chemotherapy resistant CRPC 39

Figure 4: GATA2 is upregulated in chemotherapy treated prostate cancer tissues from

multiple clinical databases 40

Figure 5: Characterization of two independent GATA2 shRNAs in chemotherapy

resistant CRPC sublines 41

Figure 6: GATA2 regulates chemotherapy resistance in CRPC cells 42

Figure 7: GATA2 regulates tumorigenicity in CRPC cells 43

Figure 8: GATA2 regulates a consensus 28 member signature of cancer associated

genes 44

Figure 9: The consensus signature is strongly expressed in clinical prostate cancer

tissues treated with chemotherapy 45

Figure 10: A subset of 7 genes is regulated by GATA2 in the chemotherapy resistant

sublines and exhibits a strongly complementary expression profile in clinical prostate

cancer tissues 46

Figure 11: Characterization of RNA interference reagents and expression constructs

for a focused genetic screen 47

Figure 12: Focused genetic screen of clinically salient GATA2 regulated genes 48

v

Figure 13: The consensus GATA2 regulated genes are unrelated to AR biology 49

Figure 14: GATA2 regulates a subset of AR targets that are not deregulated in

chemotherapy resistant CRPC cells or prostate cancer tissues 51

Figure 15: Characterization of two independent IGF2 shRNAs in chemotherapy

resistant CRPC sublines 53

Figure 16: IGF2 knockdown reduces the chemotherapy resistance and tumorigenicity

of CRPC cells 54

Figure 17: An IGF2 neutralizing antibody reduces the chemotherapy resistance and

tumorigenicity of CRPC cells 55

Figure 18: Characterization of an IGF2 vector in GATA2 shRNA expressing cells 56

Figure 19: IGF2 expression rescues the defects instigated by GATA2 knockdown on

chemotherapy resistance and tumorigenicity in CRPC cells 57

Figure 20: Exogenous IGF2 rescues the defects instigated by GATA2 knockdown on

chemotherapy resistance and tumorigenicity in CRPC cells 58

Figure 21: IGF2 protein levels are elevated and responsive to GATA2 knockdown in

chemotherapy resistant CRPC sublines 59

Figure 22: GATA2 occupies predicted binding elements in the IGF2 P4 promoter 60

Figure 23: GATA2 activates an IGF2 P4 luciferase reporter 60

Figure 24: IGF2 is co-expressed with GATA2 during prostate cancer progression 61

Figure 25: IGF2 activates a polykinase program through IGF1R and INSR downstream

of GATA2 63

vi

Figure 26: Chemicals inhibitors of the polykinase program reduce chemotherapy

resistance and soft agar growth 65

Figure 27: Alternative chemicals inhibitors of the polykinase program reduce

chemotherapy resistance and soft agar growth 66

Figure 28: Dual IGF1R/INSR inhibition improves the efficacy of chemotherapy in vitro 68

Figure 29: Monotherapy with OSI-906 does not significantly constrain the growth of

established CRPC xenografts 69

Figure 30: OSI-906 restores the efficacy of chemotherapy in vivo 70

Figure 31: OSI-906 improves the efficacy of chemotherapy and survival in preclinical

models of disseminated CRPC without added general drug toxicity 72

vii

ACKNOWLEDGEMENTS

As I conclude this chapter of my career and prepare for the next, I feel a profound sense

of gratitude and indebtedness to the extraordinary network of mentors, colleagues, family and

friends that has nurtured me throughout these last twenty eight years. It is beyond any doubt that

I would not be where I am today without their continuous support.

First and foremost, I wish to thank my thesis advisor Dr. Carlos Cordon-Cardo. From a

young age I dreamed of becoming a physician scientist and using the laboratory to contribute to

the field of medicine. In Carlos I found a mentor and role model deeply committed to this

endeavor and willing to take me on despite my limited knowledge and skills. The intellectual

freedom, enthusiasm for investigation, limitless resources and scientific mentorship afforded by

Carlos’ laboratory constituted an ideal environment for me to pursue translational research

projects. My time in Carlos’ laboratory was a period of immense personal and scientific growth

for which I will be everlastingly grateful.

I wish to thank the Columbia University faculty members who generously agreed to serve

on my thesis committee. I am particularly grateful to Dr. Cathy Mendelsohn, the chairwoman of

my committee, as well as Drs. Andrea Califano, Stephen Emerson and Ramon Parsons. These

faculty members patiently and insightfully guided my research projects during the last four years,

and I feel greatly honored to have stood on the shoulders of such a distinguished group of

scientists and mentors.

I wish to thank the members of the Cordon-Cardo laboratory. In particular, I thank Dr.

Josep Domingo-Domenech, whose highly skilled and passionate dedication to the improvement

of the oncology field has been a great source of inspiration and mentorship since my first days in

viii

Carlos’ laboratory. I thank Dennis Bonal, Janis de la Iglesia-Vicente, Nataliya Gladoun,

Elizabeth Charytonowicz, Mireia Castillo-Martin and Estrelania Williams for their scientific

support and camaraderie. I also wish to recognize all of the support staffers who worked behind

the scenes to make the scientific endeavor possible, including administrators, security officers,

housekeepers, and facility technicians. In particular, I thank Uncle Manu, Jorge Miranda,

Anthony Green, William Parkinson, Yu Zhou and Kenson Roberts.

I wish to thank the leaders of the MD/PhD program at Columbia University for their

institutional support, guidance and friendship over the years: Drs. Michael Shelanski, Ron Liem

and Patrice Spitalnik. Their generous invitation to the MD/PhD program has been one of the

most rewarding and exciting experiences in my life. I thank Stacy Warren, Zaia Sivo and Jeffrey

Brandt for their administrative support. I also thank the students of the MD/PhD program who

have been my peers, colleagues, mentors and friends during these last six years.

I thank my many scientific collaborators. I thank Dr. Veronica Rodriguez-Bravo from the

Memorial Sloan Kettering Comprehensive Cancer Center for assistance with

immunoprecipitation protocols. I thank Drs. Amaia Lumabio and Scott Lowe from the Memorial

Sloan Kettering Comprehensive Cancer Center for assistance with custom shRNA design and

cloning. I thank S. Aidan Quinn, Albert Lee and Dr. Raul Rabadan from Columbia University as

well as Xiaochen Sun, Ben Readhead, Xintong Chen and Drs. Joel Dudley and Yujin Hoshida

from the Mount Sinai Hospital for assistance with computational analyses. I thank Drs. Ruth

Rodriguez-Barrueco and Jose Silva from Columbia University for assistance with chromatin

immunoprecipitation and luciferase reporter assay protocols.

ix

Finally, I wish to thank my family and friends. I am everlastingly indebted to my mother

and father, Judy Stewart Vidal and Pierre-Paul Vidal, who in addition to making me from scratch

endowed me with a thirst for knowledge, a strong work ethic and a sense of commitment to the

uplift of humanity. I thank my brother William Vidal for his companionship and support over the

years. Among my friends, I am particularly grateful to Jason Long and Tsega Gebreyesus.

I feel grateful to all of these individuals. It is my fervent hope and expectation that in the

end my character and my contributions will prove worthy of their support.

1

CHAPTER 1:

Introduction

2

1.1: The natural history and management of prostate cancer

1.1.1: The development of metastatic prostate cancer

Prostate cancer is one of the most common malignant neoplasms in humans. Worldwide,

prostate cancer is the second leading cause of cancer incidence and the sixth leading cause of

cancer mortality in males, including more than 900,000 diagnoses and 250,000 fatalities in 2008

(Jemal et al., 2011). In the United States, prostate cancer is the leading cause of cancer incidence

and the second leading cause of cancer mortality in males, including more than 200,000

diagnoses and nearly 30,000 fatalities estimated in 2014 (Siegel et al., 2014).

Mortality from prostate cancer occurs as a result of complications associated with

metastatic disease. Accordingly, the disparity between prostate cancer incidence and mortality is

largely attributable to the fact that primary prostate cancer is highly heterogeneous with only a

subset patients progressing to metastatic disease within the adult lifetime. For example, in a

Swedish cohort of primary prostate cancer patients, 17.5% of initially untreated tumors

progressed to metastatic disease following a mean observation period of 21 years (Johansson et

al., 2004). In an American cohort of primary and locally advanced prostate cancer patients, 5.1%

progressed to metastatic disease within 15 years of radical prostatectomy (Pound et al., 1999).

This profound heterogeneity has stimulated intense interest in the classification and

treatment of patients with primary prostate cancer. Although there is considerable variability in

the management of primary prostate cancer today, decades of investigation have resulted in the

identification of several widely accepted practices. Classification is performed using

clinicopathological criteria, and treatment options include active surveillance, hormone therapy,

3

radiotherapy and surgery (Chang et al., 2014; Klotz and Emberton, 2014; Thompson et al.,

2007).

The salient clinicopathological criteria are tumor stage, Gleason grade and circulating

prostate specific antigen (PSA, encoded by the KLK3 gene) levels. Extensive clinical evidence

suggests that primary tumors with a definitive Gleason score ≤6 and a PSA <10ng/ml rarely

mestastasize and these cases are consequently termed low risk (Thompson et al., 2007). For

example, in a retrospective study of a large American radical prostatectomy cohort, only 22 of

14,123 patients with Gleason ≤6 disease exhibited lymph node metastasis at the time of surgery,

and 19 of these showed higher grade than originally assigned by pathologists upon secondary

review (Ross et al., 2012). Similarly, in a second retrospective study of a large American radical

prostatectomy cohort, only 3 of 9,557 patients with Gleason ≤6 disease experienced fatality from

prostate cancer over a mean of 15 years (Eggener et al., 2011).

These clinical observations cause many physicians to advocate active surveillance in

which patients undergo serial PSA and biopsy examinations as the appropriate management for

low risk disease (Klotz and Emberton, 2014). During active surveillance, local treatment is

deferred until the development of high risk disease. However, the excellent prognosis of low risk

disease is complicated by the fact that a substantial number of Gleason ≤6 diagnoses made by

needle biopsy harbor higher grade disease that is detectable post-operatively in prostatectomy

samples. For example, in the large Scandinavian Prostate Cancer Group Study Number 4 trial,

patients who underwent prostatectomy exhibited postoperative Gleason scores one step higher in

30% and two or more steps higher in 28% of cases (Bill-Axelson et al., 2011). Patient anxiety

may also limit the applicability of active surveillance (Klotz and Emberton, 2014).

Unfortunately, to date there are no published randomized trials comparing active surveillance

4

with radiotherapy or surgery, and as a result there is considerable variability in the management

of low risk primary prostate cancer.

Prostate cancer is termed high risk by the American Urological Association when it

exhibits spread to both lobes of the prostate, a Gleason score ≥8 or PSA >20 ng/ml (Thompson et

al., 2007). High risk patients exhibit a significant chance of local or metastatic progression as

well as prostate cancer related morbidity and mortality. For example, in a large French cohort

treated with prostatectomy and post-operative radiotherapy, 39.4% of high risk patients

experienced biochemical or clinical progression or death after a median follow up of 10.6 years

(Bolla et al., 2012). Similarly, in a retrospective study of a large German cohort, 43.8% of high

risk patients treated with prostatectomy experienced biochemical recurrence after 10 years

(Boorjian et al., 2011).

The literature on the management of high risk prostate cancer is limited as a result of

variations in definition, lack of randomized trials and underpowered studies (Chang et al., 2014).

Nonetheless, radiotherapy, hormone therapy and surgery are mainstays of management. Notably,

two large randomized trials demonstrated that a combination of radiotherapy and androgen

deprivation therapy (ADT) is superior to either alone (Bolla et al., 2002; Pilepich et al., 2005). In

contrast, the relative efficacy of radical prostatectomy versus a combination of radiotherapy and

ADT, as well as the benefits of more extensive surgeries and post-operative radiotherapy remain

unclear (Chang et al., 2014).

After local therapy both low and high risk patients are monitored by serial PSA testing.

Biochemical recurrence is heterogeneous and may signal either local recurrence or metastatic

disease (Pound et al., 1999). These outcomes are often distinguished by imaging studies and are

5

characterized by greatly differing prognoses. For example, in one American retrospective study

of patients who experienced biochemical recurrence following prostatectomy, prostate cancer

specific mortality occurred as rapidly as one year after recurrence but median overall survival for

the entire cohort was not reached after more than 15 years (Freedland et al., 2005). Although

accepted guidelines are limited, local recurrence is primarily treated with salvage radiotherapy,

ADT, surgery or a combination (Freedland and Moul, 2007).

Metastatic prostate cancer cells exhibit a strong tropism for bone and accordingly 90% of

advanced patients exhibit skeletal involvement (Gartrell and Saad, 2014). Although prostate

cancer may also colonize the liver, lungs and brain, its signature tropism for bone has garnered

the most scientific interest. The mechanisms of hematogenous dissemination in prostate cancer

include hallmarks described for other epithelial malignancies, including deregulation of cell

adhesion molecules, integrin and focal adhesion signaling and matrix metalloproteases, among

others (Jin et al., 2011). Prostate cancer tropism for bone is also multifactorial. For example,

bone marrow derived SDF1 may act as a chemoattractant for CXCR4 expressing prostate cancer

cells (Taichman et al., 2002). Moreover, prostate cancer cells may bind more avidly to bone

marrow endothelial cells than other endothelial cell types (Lehr and Pienta, 1998). In addition,

prostate cancer cells may commonly express integrins that facilitate adhesion to the bone marrow

extracellular matrix (Cooper et al., 2002). After prostate cancer cells enter the bone marrow, a

rich cellular and molecular microenvironment that includes osteoblasts and various growth

factors stimulates their growth into overt metastases. For example, co-culture with primary

mouse osteoblasts may increase the proliferation of bone metastatic prostate cancer cells (Fizazi

et al., 2003). In addition, bone marrow derived insulin-like growth factors may stimulate the

proliferation of prostate cancer cells (Ritchie et al., 1997).

6

1.1.2: The changing therapeutic landscape of metastatic prostate cancer

Although the management of metastastic prostate cancer is rapidly evolving, hormonal

manipulations remain first line therapy (Wong et al., 2014). Serum androgen levels in males are

primarily controlled by the hypothalamus–pituitary axis. Gonadotropin releasing hormone

(GnRH) secreted by the hypothalamus stimulates the anterior pituitary to secrete leutenizing

hormone (LH). LH in turn stimulates Leydig cells in the testes to secrete the vast majority of

circulating androgens. Finally, androgens regulate the development and maintenance of male

characteristics through binding to androgen receptor (AR).

Pioneering work published by Charles Huggins in 1941 demonstrated for the first time

that reduction of circulating androgen levels by surgical castration could induce significant

improvements in the progression of metastatic prostate cancer (Huggins et al., 1941). Moreover,

improved characterization of the hypothalamus–pituitary axis in the following decades resulted

in the discovery of medical castration with estrogens and GnRH agonists (Group, 1967; Schally

et al., 2000). Finally, discovery of AR led to the development of non-steroidal anti-androgens

that directly inhibit the receptor (Liao et al., 1974).

Today medical castration with GnRH agonists has largely replaced surgical castration in

developed countries, although medical and surgical castration are equally effective at reducing

circulating androgens to castrate levels and controlling the progression of metastatic disease

(Seidenfeld et al., 2000). Importantly, adrenal steroidogenesis results in residual circulating

androgens despite the 90-95% reductions achieved though castration (Gomella, 2009). Indeed,

this observation stimulated combination strategies with castration and anti-androgens to produce

combined androgen blockade (CAB). However, the modest survival benefit of CAB is mitigated

7

by added toxicity, and as a consequence CAB is only implemented in subsets of patients (Group,

2000).

ADT elicits clinically significant responses in 80-90% metastatic prostate cancer patients,

resulting in a median progression free survival benefit of 12-33 months (Hellerstedt and Pienta,

2002). However, ADT invariably precedes progression to CRPC, an aggressive disease

characterized by limited survival. The mechanisms of resistance to ADT have accordingly

attracted considerable interest and are generally classified into AR dependent and independent

categories (Zong and Goldstein, 2013).

For many years, the mechanisms of progression to CRPC were thought to be primarily

AR independent, a notion supported by considerable laboratory and clinical evidence. For

example, epithelial mesenchymal transition (EMT) may play a role in the emergence of CRPC.

Several groups reported that in vivo castration of well characterized androgen dependent

xenograft models results in the emergence of EMT markers including CDH2 and ZEB1, and

similar observations were made in clinical prostate cancer tissues managed with ADT

(Jennbacken et al., 2010; Sun et al., 2012; Tanaka et al., 2010). Altered apoptosis and survival

signaling may also promote CRPC independently of AR. For example, the anti-apoptotic factor

BCL2 is upregulated by castration in laboratory models and human prostate cancer tissues

(McDonnell et al., 1992). In addition, a number of groups reported that PTEN loss and AKT

activation contribute to castration resistant growth (Carver et al., 2011; Gao et al., 2006; Jiao et

al., 2007; Mulholland et al., 2011). Finally, several well characterized CRPC cell lines are AR

negative and form aggressive castration resistant tumors in vivo (van Bokhoven et al., 2003).

8

Nonetheless, a wealth of laboratory and clinical data suggest that AR dependent

mechanisms also contribute ADT resistance. Several groups initially reported that AR is both

necessary and sufficient for progression of androgen dependent cell lines to CRPC in vitro and in

vivo (Chen et al., 2004; Cheng et al., 2006; Snoek et al., 2009; Zegarra-Moro et al., 2002).

Moreover, a multitude of studies identified ligand dependent and independent mechanisms

whereby AR signaling remains active in CRPC. For example, intratumoral testosterone levels

may remain elevated during ADT through conversion of residual adrenal androgens or de novo

synthesis by prostate cancer cells (Chang et al., 2011; Locke et al., 2008; Montgomery et al.,

2008; Stanbrough et al., 2006). In addition, the AR gene is frequently amplified or mutated in

metastatic CRPC, and these alterations enable AR activation in the setting of reduced circulating

androgens during ADT (Grasso et al., 2012; Taylor et al., 2010; Visakorpi et al., 1995). Finally,

alternative splicing may result in the expression of transcripts that encode a constitutively active

AR (Dehm et al., 2008; Guo et al., 2009; Hu et al., 2009; Sun et al., 2010).

Perhaps the most unequivocal data that AR dependent mechanisms contribute to

progression to CRPC emerged from clinical trials in which second generation androgen signaling

inhibitors improved survival in patients with CRPC. 17-α-hydroxylase/17,20-lyase (encoded by

the CYP17A1 gene) is a critical enzyme required for androgen synthesis in the adrenal glands,

testes and prostate cancer. Abiraterone acetate is a CYP17A1 inhibitor with clinical activity in

CRPC both before and after chemotherapy (de Bono et al., 2011; Ryan et al., 2013). Moreover,

enzalutamide is a second generation anti-androgen with greater affinity for AR than bicalutamide

(Tran et al., 2009), and enzalutamide also exhibits clinical activity in CRPC both before and after

chemotherapy (Beer et al., 2014; Scher et al., 2012).

9

Apart from second generation androgen signaling inhibitors and chemotherapy, two other

treatments exhibit clinical activity in CRPC. Sipuleucel-T is a dendritic cell vaccine prepared

from peripheral blood mononuclear cells obtained by leukapheresis and exposed ex vivo to a

recombinant immunogenic protein. Sipuleucel-T improves survival in CRPC patients with

minimally symptomatic metastatic disease, although its clinical activity is impossible to monitor

with imaging or PSA (Kantoff et al., 2010). This limitation in clinical management as well as the

cost of Sipuleucel-T have greatly restricted its widespread application (Bishr and Saad, 2013).

Moreover, radium-223 is an alpha particle emitting radiopharmaceutical with tropism for bone.

Radium-223 also improves survival in CRPC and reduces skeletal events, although its use is

limited to patients lacking symptomatic visceral metastases (Parker et al., 2013).

1.1.3: The role of taxane chemotherapy in metastatic CRPC

Until recently metastatic CRPC was considered a chemotherapy resistant disease (Schutz

et al., 2014). Prior to the turn of the century, a large number of chemotherapy agents including

anthracyclines, alkylating agents, antimetabolites, platinums and topoisomerase inhibitors had

been evaluated in phase II clinical trials. Remarkably, a review of 26 trials conducted between

1987 and 1991 found that the average response rate to cytotoxic chemotherapy was less than

10% (Yagoda and Petrylak, 1993). In the following decade, the topoisomerase inhibitor

mitoxantrone was found in several large trials to provide palliative benefit in CRPC (Berry et al.,

2002; Kantoff et al., 1999; Tannock et al., 1996). Although mitoxantrone conferred no survival

benefit in these trials, the grim therapeutic landscape of CRPC resulted in the adoption of this

agent as standard chemotherapy (Schutz et al., 2014).

10

The taxane family of cytotoxic chemotherapy agents consists of paclitaxel (also called

taxol), docetaxel (also called taxotere) and cabazitaxel. The United States Congress created the

National Cancer Chemotherapy Service Center at the National Cancer Institute in 1955, and

under its auspices paclitaxel was discovered as a natural isolate from the bark of the Pacific Yew

tree in the 1960s (Chabner and Roberts, 2005). Docetaxel and cabazitaxel were later developed

by the pharmaceutical company Sanofi-Aventis as semi-synthetic derivatives of paclitaxel

(Schutz et al., 2014). The taxane chemotherapeutics bind to polymerized tubulin dimers and

function as microtubule stabilizing agents (Dumontet and Sikic, 1999). Although the

mechanisms are incompletely characterized, microtubule stabilization interferes with mitosis and

intracellular transport and activates apoptosis through the mitochondrial pathway. Some of the

molecular events that accompany this process include G2/M arrest (Fabbri et al., 2006),

activation of the spindle assembly checkpoint (Sudo et al., 2004) and inactivation of BCL2 by

phosphorylation (Haldar et al., 1996; Wang et al., 1999).

Following the clinical introduction of paclitaxel in the 1980s and docetaxel in the 1990s,

results from phase II trials began to emerge on their activity in metastatic CRPC. While

paclitaxel showed modest activity (Roth et al., 1993), two phase II trials suggested docetaxel

may be more effective (Beer et al., 2001; Berry et al., 2001). On the basis of these results, two

randomized phase III studies were initiated, TAX-327 and SWOG-9916, which included more

than 1,500 CRPC patients (Petrylak et al., 2004; Tannock et al., 2004). In both trials docetaxel

instigated PSA responses of approximately 50% and conferred an overall survival benefit.

Moreover, the results from the TAX-327 trial were later reaffirmed after longer follow up

(Berthold et al., 2008). Remarkably, in 2004 single agent docetaxel thereby became the first

11

known therapeutic intervention to prolong survival in CRPC and presently remains a mainstay of

treatment (Bishr and Saad, 2013).

Docetaxel remained the only approved treatment for CRPC for several years until the

introduction of cabazitaxel in 2010. Cabazitaxel showed activity in preclinical models of

docetaxel resistant cells and the ability to cross the blood-brain barrier, an attractive property for

the management of patients with brain metastasis (Cisternino et al., 2003; Schutz et al., 2014).

Moreover, promising results from a phase II clinical trial (Pivot et al., 2008) laid the groundwork

for the randomized phase III TROPIC study (de Bono et al., 2010). The TROPIC study showed

that cabazitaxel confers a survival benefit in patients who progress with docetaxel, and

cabazitaxel was subsequently widely approved as a second line chemotherapy agent for patients

with CRPC (Bishr and Saad, 2013).

Finally, although ADT has historically been the first line therapy for patients with

metastatic prostate cancer, recent evidence suggests that taxane chemotherapy may soon play a

major role in this context. An initial phase III trial showed that ADT combined with docetaxel

confers a measurable but not statistically significant survival benefit in men with untreated

metastatic disease (Gravis et al., 2013). However, the overwhelming majority (>75%) of patients

in this trial exhibited low or intermediate risk disease. In contrast, the larger phase III

CHAARTED trial enrolled a majority (>60%) of high risk patients, and an interim analysis

revealed an unprecedented survival benefit of 17 months in this subgroup (Sweeney, 2014).

While continued follow up is required to provide definitive results from the CHAARTED trial,

this preliminary analysis suggests that taxane chemotherapy may soon play a significant role in

the setting of untreated metastatic prostate cancer.

12

1.1.4: Taxane resistance in metastatic CRPC

Despite the advances of the last decade—first and second line chemotherapy,

immunotherapy, second generation androgen signaling inhibitors, a bone targeting

radiopharmaceutical and combination strategies—CRPC remains a uniformly lethal diagnosis

(Bishr and Saad, 2013). This harsh reality is due in part to the fact that CRPC exhibits

mechanisms of resistance to currently available therapeutic modalities. For example, although

taxanes instigate high response rates, treatment is never curative because patients universally

relapse with chemotherapy resistant disease (Seruga et al., 2011). The mechanisms of resistance

to taxane chemotherapy in metastatic CRPC are therefore of considerable clinical significance.

Mechanisms of resistance to chemotherapy are generally divided into distinct

pharmacokinetic and cellular categories (Holohan et al., 2013). Pharmacokinetic mechanisms

include systemic distribution, metabolism and elimination and are primarily determined by

chemical structure. Consequently, pharmacokinetic considerations figure primarily in drug

development and medicinal chemistry rather than the study of drug resistance. Indeed, the best

characterized mechanisms of taxane resistance in CRPC are cellular in nature (Chien and

Moasser, 2008; Murray et al., 2012; Seruga et al., 2011).

Perhaps the most clinically significant mechanism of docetaxel resistance in CRPC to

date is the molecular chaperone clusterin (Zoubeidi et al., 2010). Although the precise

mechanisms are poorly characterized, clusterin is thought to inhibit stress induced protein

aggregation and apoptosis (Yerbury et al., 2007). Interestingly, an initial study reported that

clusterin suppression with an antisense oligonucleotide could improve the efficacy of docetaxel

in CRPC cell line xenograft models (Springate et al., 2005). Although poorly characterized, the

13

increased efficacy of docetaxel in this context appears to be reduced resistance to apoptosis

(Sallman et al., 2007), a classical mechanism of chemotherapy resistance (Holohan et al., 2013).

Moreover, the second generation clusterin antisense oligonucleotide OGX-011 was well

tolerated in a phase I clinical trial in combination with docetaxel (Chi et al., 2008). Notably, two

subsequent randomized phase II trials showed that addition of OGX-011 to docetaxel may confer

a substantial survival benefit, and phase III trials are ongoing (Chi et al., 2010a; Saad et al.,

2011).

A number of studies have linked other apoptosis and survival signaling pathways to

docetaxel resistance in CRPC (Seruga et al., 2011). For example, chemical inhibition of the

PI3K/AKT (Domingo-Domenech et al., 2012; Qian et al., 2010; Wu et al., 2005), NFκB/IL6

(Domingo-Domenech et al., 2006) and EGFR (Mimeault et al., 2007) survival pathways

improves the efficacy of docetaxel in CRPC cells. Likewise, expression of the antiapoptotic

molecules SPHK1 (Pchejetski et al., 2008) and BCL2 (Domingo-Domenech et al., 2012) reduces

the efficacy of docetaxel in CRPC cells.

Alteration of drug targets is another classical mechanism of chemotherapy resistance

(Holohan et al., 2013). Microtubules are composed of tubulin subunits that belong to six distinct

classes based on their C-terminal sequence variability, and purified preparations of these distinct

classes exhibit differing dynamics (Chien and Moasser, 2008; Sullivan and Cleveland, 1986).

Interestingly, a widely reported mechanism of resistance to paclitaxel is increased composition

of class III beta tubulin (TUBB3) in microtubules. Microtubules composed of class III tubulin

are less sensitive to the effects of paclitaxel on microtubule dynamics (Derry et al., 1997),

possibly as a result of the increased microtubule instability (Banerjee et al., 1990; Panda et al.,

1994). In one study, genetic manipulation of TUBB3 in CRPC cell lines modestly regulated

14

docetaxel resistance, and elevated TUBB3 protein levels predicted poor survival in metastatic

CRPC patients receiving docetaxel (Ploussard et al., 2010). However, the reproducibility and

clinical significance of these findings remain unclear, and evidence implicating TUBB3 in

docetaxel resistance in CRPC is therefore limited at this time.

Several other mechanisms of taxane resistance have been widely reported, although their

relevance to docetaxel and CRPC are unclear. Perhaps foremost among these mechanisms is the

family of ATP-binding cassette (ABC) proteins that are energy dependent transporters for a wide

range of substrates (Holohan et al., 2013). Paclitaxel and docetaxel are known substrates for

ABCB1 (also known as MDR1 or P-gp), the best characterized ABC transporter (Gottesman et

al., 2002). However, data from cell lines and in particular from cancer tissues are conflicting,

with a substantial number of studies finding no association between taxane resistance and

ABCB1 (Chien and Moasser, 2008; Murray et al., 2012). Moreover, the failure of three

generations of ABCB1 inhibitors over a period of two decades to improve the efficacy of

chemotherapy in clinical trials has further cast doubt on its widespread significance (Tamaki et

al., 2011). In prostate cancer, one study reported that a docetaxel resistant C4-2B subline

exhibited increased ABCB1 expression and could be sensitized by genetic or pharmacological

inhibition of that transporter (Zhu et al., 2013). However, the reproducibility and clinical

significance of this finding remains unknown.

Other possible mechanisms of taxane resistance include tubulin mutations (Giannakakou

et al., 1997; Gonzalez-Garay et al., 1999), microtubule associations with other cytoskeletal

elements (Verrills et al., 2006) and limited tissue penetration (Kyle et al., 2007). However, the

evidence for these mechanisms of resistance is limited, and their relevance to docetaxel

resistance in CRPC is unknown (Chien and Moasser, 2008; Murray et al., 2012). Therefore, to

15

date the mechanisms of resistance to chemotherapy in CRPC remain unclear. Moreover, the

limited understanding of these mechanisms has failed to translate to therapeutic strategies with

clinical benefit in a phase III clinical trial. The continued investigation of docetaxel resistance in

CRPC is thus both biologically and clinically meritorious.

1.2: The role of GATA2 in physiology and disease

1.2.1: The role of GATA2 in mammalian physiology

The GATA genes are a family of evolutionarily conserved transcription factors that

regulate development and differentiation in eukaryotic organisms (Vicente et al., 2012a). In

mammals, the GATA family contains six members termed GATA1-6 and is a prototype of

lineage restricted transcription factors that regulate cell fate determination. GATA1 is the

founding member of the GATA family and was initially discovered as an erythroid nuclear

protein involved in the regulation of globin genes (Evans and Felsenfeld, 1989; Tsai et al., 1989).

Although the gene family owes its name to the consensus DNA binding motif

(A/T)GATA(A/G), GATA binding elements are now known to vary considerably in their

sequences (Kobayashi-Osaki et al., 2005; May et al., 2013; Merika and Orkin, 1993).

GATA1 is extensively characterized in the hematopoietic system where it regulates the

maturation of erythroid (Fujiwara et al., 1996) and megakaryocytic (Shivdasani et al., 1997)

lineages, among others. GATA3 regulates brain (Pandolfi et al., 1995), breast (Kouros-Mehr et

al., 2006) and kidney (Grote et al., 2008) development, among others. Interestingly, GATA3 is

also a luminal differentiation marker in breast cancer that inhibits progression and signals

favorable prognosis (Dydensborg et al., 2009; Kouros-Mehr et al., 2008; Mehra et al., 2005).

GATA4, GATA5 and GATA6 are less well characterized but known to play important roles in

16

the development of mesodermal and endodermal lineages including heart, liver, lung, gut and

gonads (Molkentin, 2000). Notably, null mutations in all GATA family genes with the exception

of GATA5 are embryonic lethal in mice (Vicente et al., 2012a).

The vertebrate GATA factors contain a conserved DNA binding domain that is

comprised of two multifunctional zinc finger domains (Trainor et al., 2000). The N-terminal and

C-terminal zinc finger domains of GATA1 are essential for its protein-DNA and protein-protein

interactions and have been studied in considerable detail. However, little is known about the

structural properties of GATA2 (Vicente et al., 2012a). Like the other GATA family members,

the two zinc finger domains of GATA2 are required for protein-DNA interactions (Evans and

Felsenfeld, 1989). In addition, the GATA2 zinc finger domains are known to be required for

protein-protein interactions. For example, the zinc finger domains are required for binding to the

transcriptional co-activator FOG1 (Chang et al., 2002), the transcription factor PU.1 (Zhang et

al., 1999) and the chromatin modifying enzyme HDAC3 (Ozawa et al., 2001). Finally, GATA2

contains a number of other poorly characterized domains, including two transactivation domains,

a nuclear localization signal and a negative regulatory domain (Vicente et al., 2012a).

The GATA2 protein undergoes a number of post-translational modifications. For

example, phosphorylation of GATA2 by AKT reduces its localization to the nucleus and its

DNA binding activity (Menghini et al., 2005), while phosphorylation by cyclin dependent

kinases regulates its protein levels during the cell cycle (Koga et al., 2007). In addition,

acetylation of at least six lysine residues by EP300 results in increased transcriptional activity

(Hayakawa et al., 2004). Moreover, sumoylation of GATA2 reduces its transcriptional activity at

the ET1 promoter of endothelial cells (Chun et al., 2003). Finally, three ubiquination sites

regulate GATA2 protein levels through the proteasome pathway (Minegishi et al., 2005).

17

GATA2 developmental biology is mostly characterized in the hematopoietic system,

although a role in the vascular system is also emerging (Vicente et al., 2012a). Hematopoietic

and vascular development are complex processes that involve multiple tissues, and efficient

blood circulation commences at E8.5 in the mouse (McGrath et al., 2003). As the site of

primitive hematopoiesis, the extraembryonic yolk sac produces hematopoietic cells beginning on

E7.5 (Ueno and Weissman, 2010). Definitive hematopoietic stem cells (HSCs) that will give rise

to all the hematopoietic cells of the adult emerge primarily from the hemogenic endothelium of

the large vessels of the intraembryonic aorta-gonad-mesonephros (AGM) region beginning on

E10.5. Interestingly, although vascular development is less well characterized, primitive

vasculogenesis also occurs in the yolk sac (Udan et al., 2013).

The first strong evidence that GATA2 regulates the development of the hematopoietic

system emerged from genetically engineered mice bearing null mutations at the Gata2 locus

(Tsai et al., 1994). Gata2-/- embryos succumb to severe anemia at E10.5, presumably as a result

of defective primitive hematopoiesis. Moreover, the profound loss or absence of Gata2-/- cells in

every hematopoietic compartment in Gata2+/+/Gata2-/- chimeric embryos after E13.5 and adults

indicates a defect in the ontogeny definitive HSCs. Indeed, later studies confirmed that GATA2

loss causes a dramatic reduction in the ontogeny of HSCs from both cultured embryonic stem

cells (Tsai and Orkin, 1997) and the AGM (de Pater et al., 2013; Gao et al., 2013; Ling et al.,

2004). Interestingly, the emergence of HSCs from the AGM is dependent on several

transcriptional activators of GATA2, including NOTCH1 (Robert-Moreno et al., 2005) and EVI1

(Yuasa et al., 2005), among others.

In addition to its widely reported role in hematopoietic development, several studies

suggest that GATA2 regulates vascular development. GATA2 mRNA and protein are strongly

18

expressed in the lymphatic vasculature of developing and adult mice, and GATA2 regulates

genes associated with vascular development in cultured primary lymphatic cells isolated from

mouse embryos (Kazenwadel et al., 2012; Lim et al., 2012). Moreover, second generation

genetically engineered mice that survive several days longer than the originally described null

mutants provide strong in vivo evidence that GATA2 regulates vascular development in the

developing embryo (Johnson et al., 2012; Lim et al., 2012). Mechanistically, GATA2 may

regulate the migration and sprouting of the developing vasculature through the VEGF co-

receptor NRP2 (Coma et al., 2013).

The severe hematopoietic defects and consequent embryonic lethality of null mice were

later found to unexpectedly mask a second facet of GATA2 developmental biology. In one study,

a yeast artificial chromosome (YAC) containing the Gata2 locus as well as >150kbp of 5’

sequence and >80kbp of 3’ sequence was found to completely rescue Gata2-/- embryos,

suggesting that all the regulatory elements required for GATA2 expression in the hematopoietic

system are contained in this genomic area (Zhou et al., 1998). Importantly, however, all the

resulting newborns rapidly succumbed to severe hydronephrosis. More detailed analysis revealed

that GATA2 is widely expressed in embryonic structures that give rise to the urogenital system,

including the uretic bud, Wolffian duct and urogenital sinus. This embryonic expression pattern

is linked to numerous postnatal abnormalities, including defects in the ureters, seminal vesicles

and vas deferens, among others (Zhou et al., 1998). Intriguingly, the authors did not comment on

the development of the prostate, a structure which emerges from the GATA2 expressing cells

that line the urogenital sinus. Nonetheless, these studies with a Gata2 YAC in Gata2-/- embryos

revealed a significant role for GATA2 in urogenital development. Moreover, a later study found

19

that GATA2 is expressed in the developing mouse prostate as well as the adult mouse and human

prostate (Perez-Stable et al., 2000).

Similar to its role in embryonic development, the role of GATA2 in adult physiology is

best characterized in the hematopoietic system (Vicente et al., 2012a). For example, GATA2 is

required for the survival and proliferation of adult HSCs. First, GATA2 is highly expressed in

adult HSCs (Orlic et al., 1995; Suzuki et al., 2006). Moreover, conditional homozygous

inactivation of Gata2 in adult mice results in a profound reduction in the number of bone marrow

HSCs (Lim et al., 2012). Likewise, adult Gata2+/- bone marrow HSCs perform poorly in

competitive transplantation experiments and exhibit a proliferation defect following cytotoxic

insult with 5-fluorouracil in vivo (Ling et al., 2004; Rodrigues et al., 2005).

In addition to its role in adult HSCs, GATA2 plays a complex role in adult hematopoiesis

(Vicente et al., 2012a). A dynamic reciprocal regulation pattern at several genomic loci called

GATA factor switching primarily involves GATA1 and GATA2 and is essential for erythroid

and myeloid differentiation. For example, disruption of GATA2 binding to its own promoter and

subsequent silencing by GATA1 is required for erythroid differentiation (Grass et al., 2003;

Ikonomi et al., 2000). Likewise, GATA factor switching involving GATA2 also regulates mast

cell differentiation (Cantor et al., 2008). In addition, GATA2 is highly expressed during

megakaryocyte differentiation, promotes megakaryocytic progenitor cell proliferation and drives

megakaryocytic lineage gene expression (Huang et al., 2009; Terui et al., 2000).

In recent years the laboratory findings implicating GATA2 in hematopoietic and vascular

biology have been complemented by a growing number of genetic studies in humans exhibiting

rare heritable diseases in which these tissues are dysfunctional. For example, two closely related

20

immunodeficiency syndromes known as monocytopenia with Mycobacterium avium complex

(MonoMAC) and dendritic cell, monocyte, B and NK lymphoid (DCML) deficiency are

characterized by mutations in GATA2 (Dickinson et al., 2011; Hsu et al., 2011). Interestingly,

patients with MonoMAC/DCML and Emberger disease, another heritable immunodeficiency

syndrome associated with GATA2 mutations, may exhibit primary lymphedema (Kazenwadel et

al., 2012; Ostergaard et al., 2011), complementing the laboratory studies linking GATA2 to

vascular biology. Finally, recent large retrospective studies have confirmed the association

between GATA2 mutation and hematological and vascular defects in human adults (Dickinson et

al., 2014; Spinner et al., 2014).

1.2.2: The role of GATA2 in hematopoietic malignancy

The heritable immunodeficiency syndromes characterized by GATA2 mutation are also

associated with malignant transformation. For example, MonoMAC/DCML deficiency and

Emberger disease are both associated with myelodysplastic syndrome (MDS) and AML

(Dickinson et al., 2011; Hsu et al., 2011; Ostergaard et al., 2011). In addition, familial GATA2

mutations may result in MDS/AML without previous hematological or vascular symptoms

(Hahn et al., 2011). Although the majority of the GATA2 mutations occurring these families have

not been characterized mechanistically, at least some of them are known to exhibit loss-of-

function and possibly dominant negative properties (Hahn et al., 2011). These observations have

led to speculation that MDS arises in these patients as a consequence of stem cell exhaustion and

bone marrow stress (Migliaccio and Bieker, 2011). In contrast, high GATA2 mRNA expression

is associated with poor prognosis in sporadic MDS (Fadilah et al., 2002) and AML (Luesink et

al., 2012; Vicente et al., 2012b), although the mechanistic role of GATA2 in this context is

unknown.

21

Chronic myeloid leukemia (CML) is characterized by an indolent chronic phase (CP) that

often progresses to a lethal blast crisis (BC) (Goldman and Melo, 2003). Although tyrosine

kinase inhibitors have dramatically altered the clinical progression of CML, a subset of patients

develops resistance and succumbs to BC (Cortes et al., 2011). CP is associated with an expansion

of differentiated myeloid cells, while BC is characterized by a reduction in differentiation and

the appearance of myeloblasts as well as lymphoblasts. The transcription factor PU.1 is a critical

regulator of myeloid differentiation (Kastner and Chan, 2008) that is repressed though a protein-

protein interaction with the C-terminal zinc finger domain of GATA2 (Zhang et al., 1999).

Interestingly, in one study 8/85 blast crisis patients exhibited a single gain-of-function mutation

in the C-terminal zinc finger domain of GATA2 (Zhang et al., 2008). The mutation increased the

transactivation potential of GATA2, accentuated repressive association of GATA2 with PU.1

and reduced the differentiation potential of human leukemia cells. Thus, gain-of-function

mutations in GATA2 may contribute to the development of BC in CML.

1.2.3: The role of GATA2 in solid tumors

Lung cancer is the most common cause of cancer incidence and mortality worldwide, and

NSCLC is the most common histological category (Ferlay et al., 2010). Although the RAS

family of GTPases and upstream receptor tyrosine kinases are among the most commonly

mutated genes in NSCLC, decades of investigation have failed to produce RAS pathway targeted

agents that provoke durable responses (Vasan et al., 2014). Consequently, synthetic lethal

interactions with RAS pathway mutation in NSCLC have attracted considerable laboratory

investigation. Notably, in one study an RNA interference screen identified GATA2 as a synthetic

lethal partner with RAS pathway mutation in NSCLC (Kumar et al., 2012). GATA2 knockdown

in cultured RAS pathway mutated NSCLC cells reduced their viability and tumorigenicity, and

22

Gata2 knockout in genetically engineered mouse models reduced lung tumorigenesis and

provoked regression of established tumors. Moreover, integrated transcriptome and genome

occupancy analyses revealed that GATA2 transcriptionally activates proteasome, IL1/NFκB and

RHO signaling, leading to a novel preclinical combination strategy. Surprisingly, this extensive

in vitro and in vivo characterization was recently challenged by a report that GATA2 expression

is strongly reduced in lung cancer versus normal tissues and that GATA2 knockdown does not

reduce the viability of cultured RAS pathway mutated NSCLC cells (Tessema et al., 2014).

Therefore, further investigation is needed to clarify the biology and therapeutic significance of

GATA2 in NSCLC.

In addition to its possible role in NSCLC, GATA2 has been studied by several groups in

prostate cancer. These studies have predominantly focused on the role of GATA2 in castration

naive disease, in particular the widely characterized AR expressing and androgen dependent

LNCaP cell line. First and foremost, GATA2 has emerged as a pioneer factor for AR.

An initial study exploring the role of GATA family proteins in the regulation of prostate

specific genes revealed that multiple GATA2 binding elements (GBEs) flank an AR responsive

site in an enhancer of KLK3 (Perez-Stable et al., 2000). Moreover, the GBEs were required for

androgen stimulated induction of PSA. A second study showed that GATA2 regulates androgen

stimulated induction of TMPRSS2, another canonical AR target, again through GBEs in an AR

regulated enhancer (Wang et al., 2007). Furthermore, the latter study showed that GATA2

binding to the KLK3 and TMPRSS2 enhancers is required for AR recruitment to these sites,

revealing a regulatory hierarchy whereby GATA2 is a pioneer factor for AR at these loci. This

regulatory hierarchy has been confirmed at several other AR regulated loci, including UBE2C

(Wang et al., 2009) as well as ABCC4 and ADPGK (Wu et al., 2014). Moreover, high

23

throughput data from both LNCaP cells (Wu et al., 2014) and genetically engineered mouse

models (Chen et al., 2013) suggest that this hierarchy operates at a genome wide level.

Finally, a recent study suggests that GATA2 may promote metastasis in the hormone

dependent setting. GATA2 knockdown in the LNCaP cell line resulted in reduced wound healing

and matrigel invasion (Chiang et al., 2014). Moreover, mechanistic data suggested that GATA2

regulates migration and invasion through focal adhesion disassembly. Therefore, the current

literature related to GATA2 in prostate cancer is limited to the castration naïve context,

advocates an important function as a pioneer factor for AR regulated genes and supports a

possible role in metastatic progression.

1.3: The role of IGF2 in physiology and disease

1.3.1: The role of IGF2 in mammalian physiology

More than half a century ago, the observation that growth hormone (GH) fails to directly

stimulate uptake of sulphate into cartilage led to the proposed existence of an intermediate

sulphation factor (SF) (Salmon and Daughaday, 1957). Shortly afterward an independent group

described a serum factor that exhibited insulin-like activities but could not be neutralized by an

insulin antibody and was accordingly termed non-suppressible insulin-like activity (NSILA)

(Froesch et al., 1963). Interestingly, SF and NSILA were later shown to regulate similar

properties, suggesting that they may in fact be the same factor (Underwood et al., 1972; Zingg

and Froesch, 1973). This factor, which had been renamed somatomedin by investigators in the

SF field (Daughaday et al., 1972), was later shown by investigators in the NSILA field to in fact

consist of two peptides (Rinderknecht and Humbel, 1976b). These peptides were soon found to

exhibit amino acid sequence homology with insulin and were consequently renamed IGF1 and

24

IGF2 (Rinderknecht and Humbel, 1976a). These early findings laid the groundwork for decades

of research that have predominantly focused on the biology of IGF1.

The IGFs bind to a family of homologous receptor tyrosine kinases that includes IGF1R

and insulin receptor (INSR) and is thought to have evolved from a common ancestral gene

(Pollak, 2008). Both IGF1 and IGF2 bind to IGF1R with very high affinity that is comparable to

that of insulin binding to INSR (Pandini et al., 2002). The INSR gene encodes two isoforms

termed A and B, and the A isoform is predominantly expressed in embryonic tissues (Giddings

and Carnaghi, 1992). In addition, the INSR A isoform is widely expressed by cancer cells and

regulates a range of malignant properties (Belfiore and Malaguarnera, 2011). Importantly, IGF2

binds with high affinity to the A isoform (Frasca et al., 1999). Moreover, IGF2 binds with high

affinity to IGF2R, which is structurally unrelated to IGF1R and INSR, lacks tyrosine kinase

activity and is thought to primarily function by sequestering IGF2 (Brown et al., 2009). In

contrast, IGF1 binds weakly to both INSR and IGF2R (Pollak, 2008).

The human IGF2 gene contains at least four promoters arranged among several exons of

which the last three are protein coding (Bergman et al., 2013; Nordin et al., 2014). The

promoters give rise to differing transcripts in a context dependent manner, although the protein

coding segment is always the same. This segment is initially translated as a 180 amino acid pre-

propeptide that contains five domains as well as a signal sequence (O'Dell and Day, 1998). The

signal sequence is cleaved to yield the 156 amino acid propeptide, which itself undergoes

sequential proteolysis to yield a mature 67 amino acid peptide that lacks the large E domain.

A wealth of evidence implicates IGF2 as a key regulator of fetal growth (Harris and

Westwood, 2012). Early RNA in situ hybridization studies revealed that IGF2 is widely

25

expressed in both human (Han et al., 1987) and rodent (Stylianopoulou et al., 1988) embryos.

However, perhaps the most decisive data originated from genetically engineered mouse models.

In an initial study, germline transmission of a null allele from male chimeras resulted in

heterozygous progeny that were approximately half the size but otherwise phenotypically

indistinguishable relative to their wild type littermates (DeChiara et al., 1990; DeChiara et al.,

1991). Reciprocally, null mutations in Igf2r result in fetal overgrowth (Lau et al., 1994; Wang et

al., 1994). In addition, IGF2 increases fetal weight when directly infused into the circulation of

pregnant guinea pigs (Sferruzzi-Perri et al., 2006). Finally, evidence from human disease also

supports a role of IGF2 in fetal development. For example, Beckwith-Wiedemann syndrome

(BWS) is characterized by fetal overgrowth and cancer predisposition and specifically involves

alterations in a genomic locus that emcompasses the IGF2 gene (Cooper et al., 2005).

At the tissue level, both Igf2 null mice and mice lacking Igf2 specifically in placental

tissues exhibit reduced placental and fetal growth (Constancia et al., 2002). Indeed, tissue

explant experiments showed that IGF2 stimulates the survival and proliferation of the

cytotrophoblast progenitors that give rise to the syncytiotrophoblast layer (Forbes et al., 2008).

Interestingly, IGF2 is also highly expressed in the trophoblastic columns that invade the maternal

endometrium (Han et al., 1996), and IGF2 regulates the migration (Irving and Lala, 1995) and

invasion (Hamilton et al., 1998) properties of these cells. IGF2 may also promote fetal

development by regulating the glucose and amino acid transport of nutrients through placental

trophoblasts (Karl, 1995; Kniss et al., 1994). Finally, IGF2 is expressed in the embryonic mouse

epicardium during midgestation heart development, and Igf2 null mice exhibit reduced

ventricular wall proliferation and ventricular wall hypoplasia (Li et al., 2011).

26

IGF2 expression in rodents declines sharply after birth with the exception of the brain

(Murphy et al., 1987; Soares et al., 1985). In postnatal humans, IGF2 is expressed form the liver

specific P1 promoter and is secreted abundantly into the circulation (Blum et al., 1988;

Sussenbach et al., 1992; Zapf et al., 1981). However, little is known about the function of this

phenomenon in adult physiology, and in recent years the biology of IGF2 has been investigated

almost exclusively in the context of malignancy (Livingstone, 2013). Nonetheless, the potential

consequences of IGF2 expression in postnatal life have been studied in several genetically

engineered mouse models (Wolf et al., 1998).

Mirroring the role of IGF2 in fetal development, overexpression also causes the

overgrowth of several tissues during adult life. In an initial study, IGF2 expression driven by a

bovine keratin 10 promoter resulted in overgrowth of the skin as well as the colon, appendix and

uterus (Ward et al., 1994). In another study, IGF2 expression driven by a rat

phosphoenolpyruvate carboxykinase promoter resulted in increased size of the kidneys, adrenal

glands and testes (Wolf et al., 1994). In a third study, IGF2 expression driven by a mouse H-2K

promoter caused increased weight of the thymus as a result of increased numbers of CD4 T cells

(Kooijman et al., 1995; van Buul-Offers et al., 1995).

Some transgenic mouse lines expressing elevated IGF2 also exhibit metabolic alterations.

In one study, IGF2 expression driven by a mouse major urinary protein promoter resulted in

animals with hypoglycemia and hypoinsulinemia, suggesting IGF2 may exert insulin-like effects

(Rogler et al., 1994). Indeed, in another study the authors confirmed that IGF2 overexpressing

animals exhibit physiological properties consistent with markedly increased peripheral insulin

action, primarily in skeletal muscle (Rossetti et al., 1996). Interestingly, both the bovine kertain

10 and mouse major urinary protein driven models exhibit markedly reduced levels body fat,

27

likely as a result of increased dietary lipid oxidation and reduced lipid incorporation into body fat

(Da Costa et al., 1994). Therefore, elevated serum IGF2 levels in human adults may regulate

tissue growth as well as glucose and lipid metabolism.

1.3.2: The regulation of IGF2 in physiology and disease

Genomic imprinting is a complex epigenetic phenomenon whereby genes are expressed

from a single allele according to parental origin (Ferguson-Smith, 2011). Although originally

described in plants four decades ago, imprinting is now known to broadly regulate development,

adult physiology and disease in mammals (Peters, 2014). Notably, the first three genes shown to

be imprinted in mice were the IGF2 receptor Igf2r (Barlow et al., 1991), Igf2 itself (DeChiara et

al., 1991) and the neighboring H19 gene (Bartolomei et al., 1991). Moreover, IGF2 was soon

confirmed as an imprinted gene in humans (Giannoukakis et al., 1993; Ogawa et al., 1993;

Ohlsson et al., 1993; Rainier et al., 1993).

Prior to the decisive data showing that Igf2 is an imprinted gene in mice, an initial clue

was the observation that the heterozygous progeny of male germline chimeras exhibited IGF2

mRNA levels ten times lower than their wild type littermates (DeChiara et al., 1990). Moreover,

the authors soon demonstrated that the growth phenotype observed in heterozygotes was never

observed in the progeny of wild type males and that the heterozygous progeny of mutant males

were phenotypically indistinguishable from homozygous mutants (DeChiara et al., 1991).

Moreover, RNAase protection assays provided definitive molecular evidence that only the

paternal Igf2 allele is expressed (DeChiara et al., 1991). The mechanism of maternal IGF2 allele

silencing in mice and humans has since attracted considerable interest and is now known to

28

involve well characterized differentially methylated regions (also called imprinting control

regions) located in and around the IGF2 locus (Nordin et al., 2014).

The finding that the maternal IGF2 allele is silenced through imprinting was quickly

followed by the discovery that LOI occurs in human cancer. Two landmark studies showed that

the maternal IGF2 allele may become reactivated in Wilms tumors (Ogawa et al., 1993; Rainier

et al., 1993). In these studies, IGF2 was shown to be exclusively expressed form the paternal

allele in embryonic and normal adult kidneys, while a subset of tumors exhibited biallelic

expression. Subsequent reports showed that maternal allele reactivation is associated with

methylation changes in the imprinting control regions (Moulton et al., 1994; Steenman et al.,

1994; Sullivan et al., 1999; Taniguchi et al., 1995) and results in substantially increased mRNA

levels of IGF2 (Ravenel et al., 2001).

The discovery of IGF2 LOI in Wilms tumors stimulated numerous complementary

studies in other tumor types. Notably, these studies have repeatedly shown that IGF2 LOI is

widespread and characterizes a large subset of patients in a vast array of malignancies (Cui,

2007). For example, an initial study in prostate cancer reported biallelic IGF2 expression in a

small fraction (2/11) of benign prostatic hyperplasia samples and in a large fraction (8/10) of

primary prostate cancers (Jarrard et al., 1995). Interestingly, further studies showed that IGF2

LOI is a consistent feature of aging prostates in both mice and humans, and in mice LOI is

associated with alterations in the well characterized differentially methylated regions that

regulate IGF2 imprinting (Fu et al., 2008).

IGF2 overexpression is extensively documented in human cancer, and a variety of

mechanisms in addition to LOI are known to underlie this phenomenon (Livingstone, 2013).

29

Indeed, early studies commented on the complexity of the regulatory sequences that comprise the

various IGF2 promoters (Holthuizen et al., 1993). For example, the transcription factors C/EBP

(van Dijk et al., 1992), EGR1 (Bae et al., 1999) and SP1 (Lee et al., 1998) were found to bind to

and activate the P1, P3 and P4 promoters, respectively, in human liver cells. Notably, the well

characterized tumor suppressor WT1 was also found to bind to the P3 promoter in human cells

and repress the expression of IGF2 (Drummond et al., 1992).

More recently, a number of additional transcription factors and a microRNA have been

found to promote IGF2 expression in cancer. For example, the transcription factor ID1 promotes

the growth and metastasis of esophageal cancer cells through upregulation of IGF2 (Li et al.,

2014), and the transcription factor ZFP57 promotes the tumorigenicity of fibrosarcoma cells

through upregulation of IGF2 (Tada et al., 2014). In addition, downregulation of the microRNA

miR-100 in breast cancer promotes tumor growth through derepression of IGF2 (Gebeshuber and

Martinez, 2013). Finally, transcription factors may drive IGF2 expression in prostate cancer.

Indeed, the transcription factor E2F3 binds to and activates the Igf2 P2 promoter in mouse

tissues, and the genes are co-expressed in clinical prostate cancer samples (Lui and Baron, 2013).

1.3.3: The role of IGF2 in malignancy

Although IGF1 has attracted considerably more interest over the last four decades, the

role of IGF2 in cancer has also been the subject of investigation (Livingstone, 2013). IGF2 binds

to canonical multimeric receptor tyrosine kinase complexes composed of IGF1R and INSR alone

or in combination as hydrid receptors (Belfiore et al., 2009). Extracellular ligand binding results

in autophosphorylation of the intracellular tyrosine kinase domains. These activated domains in

turn phosphorylate scaffolding intermediaries from the phosphoinositide-3-kinase (PI3K) and

30

mitogen activated protein kinase (MAPK) signaling pathways, primarily members of the insulin

receptor substrate (IRS) and Src homology 2 domain containing (SHC) protein families,

respectively. The activated IRS intermediaries principally activate AKT, while the SHC

intermediaries may activate JNK, ERK and P38 alone or in combination. These effector kinases

in turn regulate canonical cancer properties including survival, proliferation and metastasis in a

context dependent manner (Weroha and Haluska, 2012).

Observations from genetically engineered mouse models suggest that IGF2 may play an

important role in the pathogenesis of a number of tumor types (Livingstone, 2013). In an initial

study, expression of IGF2 in mammary tissue with a sheep beta-lactoglobulin promoter resulted

in transgenic animals that developed breast tumors beginning at five months of age (Bates et al.,

1995), and in a second study IGF2 overexpression was sufficient to initiate metastatic breast

tumors (Pravtcheva and Wise, 1998). In addition, after longer latencies IGF2 is sufficient to

initiate lung cancer (Moorehead et al., 2003) as well as liver, lymphoid, skin, soft tissue and

thyroid cancer (Rogler et al., 1994). Similarly, IGF2 modulates the penetrance of large T antigen

induced islet cell tumors (Christofori et al., 1994) and PTEN deficient breast tumors (Church et

al., 2012), and IGF2 is indispensable for the formation of PTCH deficient medulloblastoma and

rhabdomyosarcoma (Hahn et al., 2000).

In addition to this role in tumorigenesis, IGF2 may also regulate several aspects of tumor

progression. Interestingly, hypoxia instigates a rapid upregulation IGF2 in human liver cancer

cells in an EGR1 dependent manner (Bae et al., 1999). Moreover, IGF2 stimulates the secretion

of VEGF in this context, suggesting that IGF2 may regulate angiogenesis (Kim et al., 1998; Yao

et al., 2012). Furthermore, as alluded to earlier IGF2 stimulates the invasion and migration of

human placental trophoblasts (Hamilton et al., 1998; Irving and Lala, 1995), and IGF2

31

expression is sufficient to initiate metastatic breast tumors in transgenic mice (Pravtcheva and

Wise, 1998). Indeed, IGF2 confers invasive properties in leiomyosarcoma (Sciacca et al., 2002)

and choriocarcinoma (Diaz et al., 2007) cells, supporting a role for IGF2 in metastatic

progression. Finally, consistent with a large literature that modifiers of apoptosis and survival

signaling regulate the efficacy of chemotherapy (Pommier et al., 2004), IGF2 may regulate

cisplatin resistance in head and neck cancer (Ogawa et al., 2010).

Today knowledge about the role of IGF2 in prostate cancer is exceedingly scarce. As

alluded to earlier, LOI may contribute to increased IGF2 expression in primary prostate cancer

(Jarrard et al., 1995). Moreover, in one study exposure of LNCaP cells to IGF2 increased the

expression of steroidogenic enzymes, the production and secretion of androgens and the

expression of AR regulated genes (Lubik et al., 2013). Notably, the authors also reported that

IGF2 expression increases in patients undergoing ADT, suggesting that IGF2 may regulate

steroidogenesis and castration resistance in prostate cancer cells.

1.3.4: IGF signaling as a therapeutic target

The success of trastuzumab and the extensive laboratory research implicating IGF1 and

IGF1R in cancer biology stimulated the development of multiple IGF1R monoclonal antibodies

(Pollak, 2008). Unfortunately, these antibodies uniformly yielded disappointing results in large

randomized clinical trials. For example, CP-751,871 (also called figitumumab) failed to confer a

survival benefit when combined with paclitatexl and carboplatin for lung cancer in a phase III

trial (Langer et al., 2014). Similarly, R1507 is a monoclonal antibody that failed to improve

survival in combination with erlotinib for lung cancer in a large randomized phase II trial

(Ramalingam et al., 2011). Likewise, AMG479 (also called ganitumab) failed to improve

32

survival in large randomized phase II trials in combination with endocrine therapy for breast

cancer (Robertson et al., 2013) and gemcitabine for pancreatic cancer (Cohn et al., 2013).

Finally, figitumumab failed to improve survival in combination with docetaxel for CRPC in a

large randomized phase II study (de Bono et al., 2014).

These results raised considerable interest in understanding why IGF1R blockade fails to

confer clinically meaningful results, and several possible explanations have emerged (Pollak,

2012; Yee, 2012). The hypothalamus stimulates GH secretion by the anterior pituitary, which in

turn causes the liver to produce IGF1. Importantly, this endocrine system is regulated by

negative feedback at the level of both the hypothalamus and pituitary. IGF1R blockade therefore

predictably causes large compensatory increases in GH and IGF1 in humans (Atzori et al., 2011;

Haluska et al., 2007; Tolcher et al., 2009). In the setting of supraphysiological IGF1, weak

binding of that ligand to INSR may activate downstream signaling in spite of IGF1R blockade.

Moreover, GH excess is known in the endocrinology field to promote insulin resistance (Moller

and Jorgensen, 2009). Indeed, hyperglycemia and compensatory hyperinsulinemia are well

documented consequences of IGF1R blockade (Atzori et al., 2011; Haluska et al., 2007; Tolcher

et al., 2009), and hyperinsulinemia would be expected to activate INSR expressed by tumor cells.

The finding that IGF1R knockdown causes INSR hypersensitivity to insulin is particularly

relevant in this context (Dinchuk et al., 2010). In addition, the high serum IGF2 levels

physiologically present in humans would be expected to efficiently activate INSR and

downstream signaling irrespective of IGF1R blockade. Finally, laboratory models suggest that

IGF1R antibodies may cause compensatory INSR activation in the absence of changes in ligand

concentrations (Buck et al., 2010). Therefore, a number of mechanisms are thought to limit the

efficacy of IGF1R blockade, and these mechanisms generally involve activation of INSR.

33

The data emphasizing the likely significant role of INSR in IGF1R blockade resistance is

particularly relevant to the recent development of dual IGF1R/INSR tyrosine kinase inhibitors

and anti-ligand therapeutics. For example, OSI-906 is a potent, selective and orally bioavailable

small molecule inhibitor of IGF1R and INSR (Mulvihill et al., 2009). OSI-906 shows therapeutic

activity in an array of in vivo models (Flanigan et al., 2010; Flanigan et al., 2013; Fox et al.,

2011; Kuhn et al., 2012; Mulvihill et al., 2009; Zeng et al., 2012; Zhao et al., 2012; Zinn et al.,

2013), with some studies reporting abrogation of IGF1R induced activation of INSR (Buck et al.,

2010; Zhao et al., 2012). Moreover, OSI-906 was well tolerated in phase I clinical trials (Jones et

al., 2014; Puzanov et al., 2014), and a phase III study is underway in the United States for

adrenocortical carcinoma (NCT00924989). Like OSI-906, BMS-754807 is a potent, selective

and orally bioavailable small molecule inhibitor of IGF1R and INSR (Carboni et al., 2009;

Wittman et al., 2009). BMS-754807 also exhibits therapeutic activity in animal models and is

currently in clinical development (Awasthi et al., 2012; Heilmann et al., 2014; Hou et al., 2011;

Litzenburger et al., 2011; Tognon et al., 2011).

In addition to dual IGF1R/INSR small molecule inhibitors, antibodies that target the IGF

ligands have been developed. For example, MEDI-573 binds to IGF1 and IGF2 with high

affinity and reverses IGF ligand induced signaling (Gao et al., 2011). Moreover, MEDI-573

exhibits therapeutic activity in vivo (Gao et al., 2011; Zhong et al., 2014) and was well tolerated

in a phase I clinical trial (Haluska et al., 2014). Similarly, BI-836845 binds to IGF1 and IGF2

with high affinity, reverses IGF ligand induced signaling, exhibits therapeutic activity in vivo

and is currently in clinical development (Friedbichler et al., 2014). Notably, both MEDI-573 and

BI-836845 inhibit signaling downstream of IGF2 and INSR. Therefore, disappointing results

from the first generation of IGF1R antibodies have given rise to a second generation of small

34

molecule inhibitors and antibodies. These second generation therapeutics are more aligned with

the molecular pathophysiology of IGF signaling in cancer and therefore may instigate more

favorable outcomes (Pollak, 2012; Yee, 2012).

35

CHAPTER 2:

GATA2 confers aggressiveness in CRPC

36

2.1: GATA2 is upregulated in chemotherapy resistant CRPC in laboratory models and clinical

databases

2.1.1: Generation of a docetaxel resistant ARCaPM subline

We previously generated two in vitro models of docetaxel resistance using the well

characterized CRPC cell lines DU145 and 22Rv1 (Domingo-Domenech et al., 2012). The

resulting docetaxel resistant sublines were termed DU145-DR and 22Rv1-DR. In addition to

docetaxel resistance, DU145-DR and 22Rv1-DR were characterized by increased tumorigenicity

in immunocompromised mice.

To further investigate the mechanisms underlying docetaxel resistance in CRPC, we

generated a third model using the CRPC cell line ARCaPM. ARCaPM is a derivative of a cell

line generated from the ascites fluid of a patient with metastatic CRPC (Zhau et al., 1996).

ARCaPM cells were grown to 50% confluence, treated with 50nM docetaxel for 72 hours and

the emergence of docetaxel resistant clones was monitored microscopically over a period of 10-

20 days. This selection process was repeated serially over a period of six months with gradually

increasing concentrations of docetaxel until a dose of 500nM was reached. The resulting subline

was termed ARCaPM-DR and exhibited no reduction in two dimensional colony formation

capacity following 72 hour exposure to 125nM docetaxel, a concentration uniformly lethal in the

parental cell line (Figure 1(a)). Moreover, immunoblot studies confirmed that there was no

increase in the induction of apoptosis following 72 hour exposure to 125nM docetaxel (Figure

1(b)). Finally, consistent with our previous characterization of DU145-DR and 22Rv1-DR, we

observed that ARCaPM-DR cells are more tumorigenic than their parental counterparts

following subcutaneous implantation in immunocompromised mice (Figure 1(c)).

37

Figure 1: Characterization of the ARCaPM-DR subline. (a) Representative colony formation assays and quantifications of ARCaPM parental and ARCaPM-DR cells following indicated docetaxel treatment for 72 hours. Data represent the mean ± SD. (b) Representative immunoblots of cleaved PARP protein levels in ARCaPM parental and ARCaPM-DR cells following treatment with DMSO or docetaxel (125nM) for 72 hours. (c) Tumor incidence and latency of ARCaPM parental and ARCaPM-DR cells following subcutaneous implantation in immunocompromised mice. Data represent the mean ± SD. *P<0.05.

2.1.2: Characterization of cross resistance to cabazitaxel in docetaxel resistant cells

Cabazitaxel is a second generation semi-synthetic taxane agent that improves survival in

CRPC patients who progress with docetaxel (de Bono et al., 2010). However, while docetaxel

elicits responses of approximately 50% (Petrylak et al., 2004; Tannock et al., 2004), cabazitaxel

does so only in a small subset of docetaxel resistant patients (de Bono et al., 2010). Interestingly,

cell viability assays showed that the three docetaxel resistant sublines exhibit varying degrees of

cross resistance to cabazitaxel (Figure 2(a)). Colony formation assays confirmed the relative

resistance of the docetaxel resistant sublines to cabazitaxel (Figure 2(b)). Moreover, immunblot

(Figure 2(c)) and flow cytometry (Figure 2(d)) studies showed that cabazitaxel efficiently

38

induces apoptosis in the parental cell lines, but that this effect is abolished in the docetaxel

resistant sublines. These results suggest that the molecular landscape of the docetaxel resistant

models confers varying degrees of resistance to the second line agent cabazitaxel.

Figure 2: Docetaxel resistant sublines exhibit varying degrees of cross resistance to cabazitaxel. (a) MTS assays of parental and docetaxel resistant cells treated for 72 hours with DMSO or indicated cabazitaxel concentrations. Data represent the mean ± SD. *P<0.05. (b) Representative colony formation assays and quantifications of parental and docetaxel resistant cells following indicated treatments for 72 hours. Data represent the mean ± SD. (c) Representative immunoblots of cleaved PARP protein levels in the cells and 72 hour treatments described in (b). (d) Flow cytometry detection of Annexin V and propidium iodide (PI) in the cells and 72 hour treatments described in (b). Data represent the mean ± SD. *P<0.05.

39

2.1.3: GATA2 is upregulated in chemotherapy resistant CRPC in laboratory models and clinical

databases

To investigate whether GATA2 contributes to the properties of the chemotherapy

resistant sublines, we first evaluated whether its expression levels vary between the sublines and

their parental counterparts. Indeed, we observed that GATA2 mRNA and protein are

substantially elevated in the previously characterized DU145-DR and 22Rv1-DR models (Figure

3(a)). Moreover, we observed that GATA2 mRNA and protein levels are also increased in the

third independent ARCaPM-DR model (Figure 3(b)).

Figure 3: GATA2 is upregulated in three models of chemotherapy resistant CRPC. (a) qRT-PCR and immunoblot analyses of GATA2 mRNA and protein levels, respectively, in DU145-DR and 22Rv1-DR cells relative to their parental counterparts. Data represent the mean ± SD. *P<0.05. (b) qRT-PCR and immunoblot analyses of GATA2 mRNA and protein levels, respectively, in the ARCAPM-DR subline relative to its parental counterpart. Data represent the mean ± SD. *P<0.05.

To assess the clinical relevance of these findings, we interrogated published gene

expression datasets derived from clinical prostate cancer tissues. Specifically, we investigated

whether GATA2 is significantly deregulated during the progression from primary disease to both

heavily treated lethal prostate cancer in the Grasso et al. study (Grasso et al., 2012) and

disseminated chemotherapy treated disease in the Taylor et al. study (Taylor et al., 2010).

40

Indeed, we observed that GATA2 mRNA increases during the progression to chemotherapy

treated disease in both published datasets (Figure 4(a)).

To further characterize the expression of GATA2 in clinical prostate cancer, we

performed an immunohistochemistry analysis in a series of 124 human paraffin embedded

tissues. We observed that GATA2 protein levels increase during the progression from primary

prostate cancer to disseminated CRPC, while the subset of patients treated with taxane

chemotherapy exhibit the highest levels (Figure 4(b)).

Figure 4: GATA2 is upregulated in chemotherapy treated prostate cancer tissues from multiple clinical databases. (a) Box plots of GATA2 mRNA levels during disease progression in the indicated clinical transcriptome datasets. Line, median; box, 25th to 75th percentiles; bars, 1.5 times interquartile range (IQR); dots, outliers. *P<0.05. (b) Representative immunohistochemistry images and quantifications of GATA2 protein levels during disease progression in a series of human paraffin embedded clinical prostate cancer tissues. Line, median; box, 25th to 75th percentiles; bars, 1.5 times IQR; dots, outliers. n=124. Bar, 100µm. *P<0.05.

41

2.2: GATA2 regulates chemotherapy resistance and tumorigenicity

Our data from laboratory models and clinical databases supported a possible association

between GATA2, chemotherapy resistance and tumorigenicity in CRPC. To further investigate

this association, we screened a panel of custom short hairpin RNAs (shRNAs) and characterized

two that efficiently reduced the mRNA (Figure 5(a)) and protein (Figure 5(b)) levels of GATA2

in the three chemotherapy resistant sublines.

Figure 5: Characterization of two independent GATA2 shRNAs in chemotherapy resistant CRPC sublines. (a) qRT-PCR of GATA2 mRNA levels in DU145-DR, 22Rv1-DR and ARCaPM-DR cells stably expressing two independent GATA2 shRNAs or a control shRNA. Data represent the mean ± SD. (b) Representative immunoblots of GATA2 protein levels in cells from (a).

Colony formation assays revealed that GATA2 knockdown potently sensitizes the three

chemotherapy resistant sublines to docetaxel and cabazitaxel (Figure 6(a)). Moreover,

immunoblot (Figure 6(b)) and flow cytometry (Figure 6(c)) experiments showed that

sensitization is accompanied by increased induction of an apoptotic response.

42

Figure 6: GATA2 regulates chemotherapy resistance in CRPC cells. (a) Representative colony formation assays and quantifications of DU145-DR, 22Rv1-DR and ARCaPM-DR cells stably expressing indicated control or GATA2 shRNAs and following 72 hour treatment with DMSO, docetaxel (125nM) or cabazitaxel (DU145-DR and 22Rv1-DR, 25nM; ARCaPM-DR, 5nM). Data represent the mean ± SD. (b) Representative immunoblots of cleaved PARP protein levels in the cells and 72 hour treatments described in (b). (c) Flow cytometry detection of Annexin V and PI in the cells and 72 hour treatments described in (b). Data represent the mean ± SD. *P<0.05.

In addition, detailed transplantation studies showed that GATA2 knockdown reduces the

capacity of CRPC cells to form tumors in immunocompromised mice (Figure 7(a)). For example,

following implantation of 100 cells GATA2 knockdown completely abrogated the

tumorigenicity of the chemotherapy resistant sublines. Moreover, while infrequent tumors

formed at larger inoculums, they did so after long latencies and likely as a result of escape from

GATA2 knockdown (Figure 7(b)).

43

Figure 7: GATA2 regulates tumorigenicity in CRPC cells. (a) Tumor incidence and latency of DU145-DR, 22Rv1-DR and ARCaPM-DR cells stably expressing indicated control or GATA2 shRNAs. Data represent the mean ± SD. *P<0.05. (b) qRT-PCR of GATA2 mRNA levels in preimplanted DU145-DR, 22Rv1-DR and ARCaPM-DR cells stably expressing indicated control or GATA2 shRNAs as well as infrequent explanted tumors (103 cell inoculum). Adjoining representative immunohistochemistry images of GATA2 protein levels in the explanted tumors. Data represent the mean ± SD. Bar, 100µm.

2.3: GATA2 regulates a signature of cancer progression associated genes

2.3.1: Identification and characterization of GATA2 regulated genes

To elucidate mechanisms whereby GATA2 regulates chemotherapy resistance and

tumorigenicity, we performed expression profiling and knowledge based computational studies

comparing control and GATA2 knockdown conditions in the chemotherapy resistant sublines.

Prompted by the observation that GATA2 regulates similar biological properties in DU145-DR,

44

22Rv1-DR and ARCaPM-DR, we focused an initial series of studies on mechanisms overlapping

among the three models. RNA sequencing followed by differential expression analysis

(Padj<0.05, fold change (FC)≥1.5) resulted in the identification of a consensus 28 member

signature of GATA2 regulated genes (Figures 8(a) and 8(b)). In addition, gene ontology analysis

showed that ‘Cancer’, ‘Cell Death and Survival’ and ‘Cellular Growth and Proliferation’ are the

most commonly deregulated biological categories (Figure 8(c)).

Figure 8: GATA2 regulates a consensus 28 member signature of cancer associated genes. (a) Venn diagram of differentially expressed genes (Padj<0.05 and FC≥1.5) identified by RNA sequencing after shRNA mediated GATA2 knockdown in DU145-DR, 22Rv1-DR and ARCaPM-DR cells. (b) Heatmap of the consensus 28 member signature of GATA2 regulated genes from (a). (c) Ingenuity knowledge based molecular and cellular functions analysis of GATA2 regulated genes in DU145-DR, 22Rv1-DR and ARCaPM-DR cells from (a).

To assess the clinical relevance of the consensus signature, we interrogated its

representation in patient derived datasets. Specifically, we performed Pearson correlation

coefficient calculations comparing the average expression of the signature and the expression

data of patients as previously described (Liu et al., 2007). These calculations showed that the 28

45

gene signature is remarkably enriched in the heavily treated lethal prostate cancer population

from the large Grasso et al. study (Figure 9). Similarly, the signature is also significantly

enriched in the disseminated chemotherapy treated population from the independent Taylor et al.

study (Figure 9). Therefore, the collective expression pattern of the consensus signature genes

during prostate cancer progression is strongly consistent with regulation by GATA2.

Figure 9: The consensus signature is strongly expressed in clinical prostate cancer tissues treated with chemotherapy. Box plots of the Pearson coefficients for the correlation between the expression data of each patient in the indicated clinical transcriptome datasets and the average expression of the consensus GATA2 signature during prostate cancer progression. Line, median; box, 25th to 75th percentiles; bars, 1.5 times IQR; dots, outliers.

2.3.2: Genetic screen of a subset of clinically salient GATA2 regulated genes

To further distill the consensus signature to its clinically salient components, we focused

on genes that exhibited the most robust differential expression pattern consistent with regulation

by GATA2 in patient derived datasets. We thereby identified a core subset of 7 genes that

demonstrate both uniform deregulation by GATA2 knockdown in the three chemotherapy

resistant sublines (Figure 10(a)) as well as a strongly complementary (FDR<0.1) expression

pattern during disease progression in the Grasso et al. and Taylor et al. studies (Figure 10(b)).

46

Figure 10: A subset of 7 genes is regulated by GATA2 in the chemotherapy resistant sublines and exhibits a strongly complementary expression profile in clinical prostate cancer tissues. (a) qRT-PCR of mRNA levels of a subset of cancer progression associated genes in DU145-DR, 22Rv1-DR and ARCaPM-DR cells stably expressing indicated control or GATA2 shRNAs. Data represent the mean ± SD. (b) Z scores of mRNA levels of genes from (a) in the indicated clinical transcriptome datasets. Line, median; box, 25th to 75th percentiles; bars, 1.5 times IQR. (c) qRT-PCR of mRNA levels of genes from (a) and (b) in DU145-DR, 22Rv1-DR and ARCaPM-DR cells relative to their parental counterparts. Data represent the mean ± SD. *P<0.05.

Interestingly, all 7 genes also exhibited corresponding expression patterns in the

chemotherapy resistant cultures relative to their parental counterparts (Figure 10(c)), suggesting

possible roles in chemotherapy resistance, tumorigenicity or both. Moreover, complementing the

knowledge based identification of ‘Cancer’, ‘Death and Survival’ and ‘Growth and Proliferation’

as relevant GATA2 regulated biological categories (Figure 8(c)), this core subset was rich in

genes known to regulate malignant properties. For example, known cancer progression

promoting genes (e.g. IGF2, PAK4, FOXM1) were repressed by GATA2 knockdown and

upregulated during prostate cancer progression, while known cancer progression inhibiting genes

47

(e.g. GALNT7, ARRDC3) were derepressed by GATA2 knockdown and downregulated during

prostate cancer progression (Figures 10(a) and 10(b)).

Figure 11: Characterization of RNA interference reagents and expression constructs for a focused genetic screen. qRT-PCR of mRNA levels of a subset of clinically salient GATA2 regulated genes in DU145-DR, 22Rv1-DR and ARCaPM-DR following the indicated genetic manipulations. Data represent the mean ± SD.

In order to functionally characterize the 7 gene signature, we performed a genetic screen.

Specifically, we performed RNA interference (genes repressed by GATA2 knockdown) and

overexpression (genes derepressed by GATA2 knockdown) studies using the three chemotherapy

resistant sublines (Figure 2.11). We monitored both chemotherapy resistance and soft agar

colony formation as a surrogate for in vivo tumorigenicity. Notably, 5 of the 7 genes consistently

regulated chemotherapy resistance (Figure 12(a)), soft agar growth (Figure 12(b)) or both,

revealing a discrete network of functionally validated genes. For example, IGF2, PAK4 and

FOXM1 suppression diminished chemotherapy resistance and soft agar growth, while re-

expression of ARRDC3 primarily impinged on soft agar growth. These results suggest that

GATA2 regulates a core subset of clinically and biologically significant genes during the

progression to chemotherapy resistant CRPC.

48

Figure 12. Focused genetic screen of clinically salient GATA2 regulated genes. (a) Colony formation assay quantifications of DU145-DR, 22Rv1-DR and ARCaPM-DR cells transfected with indicated siRNAs or stable expression vectors and following 72 hour treatment with DMSO, docetaxel (125nM) or cabazitaxel (DU145-DR and 22Rv1-DR, 25nM; ARCaPM-DR, 5nM). Data represent the mean ± SD. *P<0.05. (c) Soft agar colony formation assay quantifications of DU145-DR, 22Rv1-DR and ARCaPM-DR transfected with indicated siRNAs or stable expression vectors. Data represent the mean ± SD. *P<0.05.

49

2.3.3: The consensus signature is unrelated to AR biology

Finally, mindful of previous reports establishing GATA2 as a pioneer factor for AR

regulated genes, we focused a second series of studies on this association. Consistent with their

parental lines, DU145-DR and ARCaPM-DR lack detectable AR protein (Figure 13(a)),

indicating that AR is dispensable for regulation of the consensus 28 gene signature by GATA2.

This observation was further credentialed by the near complete failure of the 28 gene signature to

overlap with >1,000 known AR regulated genes from two published (Sharma et al., 2013; Wang

et al., 2009) CRPC studies (Figure 2.13(b)). Moreover, the core GATA2 regulated genes

identified in our studies are unresponsive to AR knockdown in the AR expressing 22Rv1-DR

culture (Figure 13(c)).

Figure 13: The consensus GATA2 regulated genes are unrelated to AR biology. (a) Representative immunoblots of AR protein levels in DU145-DR, 22Rv1-DR and ARCaPM-DR cells. (b) Venn diagrams of the consensus GATA2 signature and two signatures of known AR regulated genes in CRPC. (c) qRT-PCR of mRNA levels of a subset of clinically salient GATA2 regulated genes in 22Rv1-DR cells transfected with two independent AR siRNAs or a control siRNA. Data represent the mean ± SD.

50

Our data suggests that the consensus GATA2 signature is unrelated to AR biology.

Nonetheless, we observed that subsets of AR regulated genes from the Sharma et al. and Wang et

al. studies are significantly deregulated by GATA2 knockdown in the AR expressing 22Rv1-DR

culture (Figure 14(a)). Notably, these GATA2 and AR regulated genes include the canonical AR

targets KLK3 and TMPRSS2 (Figure 14(b)), corroborating previous reports (Perez-Stable et al.,

2000; Wang et al., 2007) that GATA2 functions as a pioneer factor for AR at these loci.

However, the GATA2 and AR regulated signatures are not consistently deregulated in 22Rv1-

DR relative to its parental line (Figure 14(c)), suggesting that they do not contribute to

chemotherapy resistance or tumorigenicity. Complementing this hypothesis, the GATA2 and AR

regulated signatures are poorly correlated with progression to heavily treated disease in clinical

datasets (Figure 14(d)). Therefore, our data collectively suggest that in addition to its established

pioneer function at androgen regulated sites, GATA2 regulates clinically relevant AR

independent genes that confer chemotherapy resistant and tumorigenic properties in CRPC.

51

Figure 14: GATA2 regulates a subset of AR targets that are not deregulated in chemotherapy resistant CRPC cells or prostate cancer tissues. (a) Venn diagrams of GATA2 regulated genes in the AR expressing 22Rv1-DR subline and two independent signatures of known AR regulated genes in CRPC. (b) qRT-PCR of mRNA levels of two canonical AR regulated genes in 22Rv1-DR cells stably expressing control or GATA2 shRNAs. Data represent the mean ± SD. *P<0.05. (c) Heat maps of mRNA levels of GATA2 and AR regulated genes in 22Rv1 parental versus 22Rv1-DR cells. (d) Box plots of the Pearson coefficients for the correlation between the expression data of each patient in the indicated clinical transcriptome datasets and the average expression of the overlapping genes from (a) during prostate cancer progression. Line, median; box, 25th to 75th percentiles; bars, 1.5 times IQR; dots, outliers.

52

CHAPTER 3:

IGF2 significantly mediates GATA2 biology in CRPC

53

3.1: Functional validation of IGF2 as a downstream effector of GATA2

We next investigated whether GATA2 regulates actionable targets with the potential to

confer therapeutic benefit. Among the functionally validated GATA2 regulated genes (Figure

12), IGF2 was distinctive through its strong and consistent regulation of both chemotherapy

resistance and soft agar colony formation. Indeed, as a canonical activator of IGF signaling,

IGF2 promotes aggressive properties in a multitude of malignancies (Livingstone, 2013). In

addition, we noted that several IGF pathway inhibitors are currently in advanced clinical

development, highlighting a possible therapeutic opportunity. Together, these observations

credentialed IGF2 as a biologically and therapeutically meritorious candidate downstream of

GATA2, and we therefore selected it for more detailed analysis.

Figure 15: Characterization of two independent IGF2 shRNAs in chemotherapy resistant CRPC sublines. (a) qRT-PCR of IGF2 mRNA levels in DU145-DR, 22Rv1-DR and ARCaPM-DR cells stably expressing two independent IGF2 shRNAs or a control shRNA. Data represent the mean ± SD. (b) Representative immunoblots of cells from (a).

We first performed a series of studies to clearly establish IGF2 as a functional mediator

of GATA2 biology. Two independent shRNAs efficiently reduced the mRNA and protein levels

of IGF2 in the three chemotherapy resistant sublines (Figure 15). Notably, colony formation

54

(Figure 16(a)) and xenotransplantation (Figure 16(b)) studies showed that IGF2 knockdown

reduces the chemotherapy resistant and tumorigenic properties of these cells, respectively.

Figure 16: IGF2 knockdown reduces the chemotherapy resistance and tumorigenicity of CRPC cells. (a) Representative colony formation assays and quantifications of DU145-DR, 22Rv1-DR and ARCaPM-DR cells stably expressing indicated control or IGF2 shRNAs and following 72 hour treatment with DMSO, docetaxel or cabazitaxel. Data represent the mean ± SD. (b) Tumor incidence and latency of DU145-DR, 22Rv1-DR and ARCaPM-DR cells stably expressing indicated control or IGF2 shRNAs. Data represent the mean ± SD. *P<0.05.

55

Moreover, we observed that IGF2 is abundantly secreted into the culture medium of the

chemotherapy resistant sublines (Figure 17(a)), and addition of a neutralizing IGF2 antibody

reduces their chemotherapy resistance (Figure 17(b)) and growth in soft agar (Figure 17(c)).

Figure 17: An IGF2 neutralizing antibody reduces the chemotherapy resistance and tumorigenicity of CRPC cells. (a) Enzyme-linked immunosorbent assay (ELISA) of IGF2 protein levels in the culture medium of DU145-DR, 22Rv1-DR and ARCaPM-DR cells relative to their parental cell lines. Data represent the mean ± SD. *P<0.05. (b) Representative colony formation assays and quantifications of DU145-DR, 22Rv1-DR and ARCaPM-DR cells following 72 hour treatment with isotype control or anti-IGF2 immunoglobulin (10µg/ml) as well as DMSO, docetaxel or cabazitaxel. Data represent the mean ± SD. *P<0.05. (c) Representative soft agar colony formation assays and quantifications of DU145-DR, 22Rv1-DR and ARCaPM-DR cells treated continuously with isotype control or anti-IGF2 immunoglobulin (10µg/ml). Data represent the mean ± SD. *P<0.05.

56

Next, we transduced control and GATA2 shRNA expressing chemotherapy resistant cells

with a control or IGF2 expression vector (Figure 18). Notably, colony formation (Figure 19(a))

and xenotransplantation (Figure 19(b)) studies showed that IGF2 expression significantly rescues

the biological defects instigated by GATA2 knockdown in these models.

Figure 18: Characterization of an IGF2 vector in GATA2 shRNA expressing cells. (a) qRT-PCR of IGF2 mRNA levels in DU145-DR, 22Rv1-DR and ARCaPM-DR cells stably expressing control or GATA2 shRNAs as well as a control or IGF2 expression vector. Data represents the mean ± SD. (b) Representative immunoblots of cells from (a).

Moreover, we observed that addition of recombinant IGF2 (rIGF2) to the culture medium

of the chemotherapy resistant sublines rescues the defects instigated by GATA2 knockdown on

taxane resistance (Figure 20(a)) and soft agar growth (Figure 20(b)). Collectively, these results

suggest that IGF2 significantly mediates chemotherapy resistant and tumorigenic properties of

CRPC cells downstream of GATA2.

57

Figure 19: IGF2 expression rescues the defects instigated by GATA2 knockdown on chemotherapy resistance and tumorigenicity in CRPC cells. (a) Representative colony formation assays and quantifications of DU145-DR, 22Rv1-DR and ARCaPM-DR cells stably expressing indicated control or GATA2 shRNAs as well as a control or IGF2 expression vector and following 72 hour treatment with DMSO, docetaxel or cabazitaxel. Data represent the mean ± SD. (b) Tumor incidence and latency of DU145-DR, 22Rv1-DR and ARCaPM-DR cells stably expressing indicated control or GATA2 shRNAs as well as a control or IGF2 expression vector. Data represent the mean ± SD.

58

Figure 20: Exogenous IGF2 rescues the defects instigated by GATA2 knockdown on chemotherapy resistance and tumorigenicity in CRPC cells. (a) Representative colony formation assays and quantifications of DU145-DR, 22Rv1-DR and ARCaPM-DR cells stably expressing indicated control or GATA2 shRNAs and following 72 hour treatment with vehicle or rIGF2 (100ng/ml) as well as DMSO, docetaxel or cabazitaxel. Data represent the mean ± SD. (b) Representative soft agar colony formation assays and quantifications of DU145-DR, 22Rv1-DR and ARCaPM-DR cells stably expressing indicated control or GATA2 shRNAs and continuously treated with vehicle or rIGF2 (100ng/ml). Data represent the mean ± SD. *P<0.05.

59

3.2: GATA2 directly upregulates IGF2

3.2.1: GATA2 binds to and activates the IGF2 P4 promoter

We previously observed that IGF2 mRNA levels are strongly upregulated and responsive

to GATA2 knockdown in the chemotherapy resistant sublines (Figure 10). We first confirmed

these observations at the protein level (Figure 21), and we next sought to investigate the

mechanism whereby GATA2 regulates IGF2 expression in our models.

Figure 21: IGF2 protein levels are elevated and responsive to GATA2 knockdown in chemotherapy resistant CRPC sublines. (a) Representative immunoblots of IGF2 protein levels in DU145-DR, 22Rv1-DR and ARCaPM-DR relative to their parental counterparts. (b) Representative immunoblots of IGF2 protein levels in DU145-DR, 22Rv1-DR and ARCaPM-DR cells stably expressing indicated control or GATA2 shRNAs.

The human IGF2 gene is transcribed from as many as five promoters termed P0-P4

(Livingstone, 2013). Interestingly, probing the TRANSFAC database with these promoter

sequences indicated that P4 contains two predicted GATA2 binding sites that we termed GBE1

and GBE2 (Figure 22(a)). Indeed, chromatin immunoprecipitation followed by qPCR (ChIP-

qPCR) studies showed that GATA2 occupies the GBEs but not adjacent control regions in the

three chemotherapy resistant sublines (Figure 22(b)).

60

Figure 22: GATA2 occupies predicted binding elements in the IGF2 P4 promoter. (a) Schematic representation of two predicted GBEs in the IGF2 P4 region identified computationally using the TRANSFAC database. (b) ChIP-qPCR of GATA2 occupancy at GBE1, GBE2 and flanking control regions in DU145-DR, 22Rv1-DR and ARCaPM-DR cells. Data represent the mean ± SD. *P<0.05.

Figure 23: GATA2 activates an IGF2 P4 luciferase reporter. (a) Luciferase luminescence of parental CRPC cells following co-transfection with a control or GATA2 expression vector, a P4 luciferase reporter (wild type (WT), mutated GBE1 (mutGBE1) or mutated GBE2 (mutGBE2)) and a renilla transfection efficiency control. Data represent the mean ± SD. *P<0.05 relative to WT. **P<0.05 relative to control vector. (b) Luciferase luminescence of NIH3T3 cells following the experimental procedures described in (a). Data represent the mean ± SD. *P<0.05 relative to WT. **P<0.05 relative to control vector.

61

In addition, co-transfection assays showed that GATA2 expression can activate luciferase

transcription from a P4 reporter in the three parental cell lines, while mutation of the GBEs

reverses this effect (Figure 23(a)). Moreover, these observations were independently validated in

embryonic fibroblast cells (Figure 23(b)). Thus, we observed that GATA2 may occupy predicted

binding elements in the P4 sequence and thereby activate IGF2 transcription.

3.2.2: GATA2 and IGF2 are co-expressed during prostate cancer progression

Figure 24: IGF2 is co-expressed with GATA2 during prostate cancer progression. Representative immunohistochemistry images and quantifications of IGF2 protein levels during disease progression in a series of human paraffin embedded prostate cancer tissues. Line, median; box, 25th to 75th percentiles; bars, 1.5 times IQR; dots, outliers. n=124. Bar, 100µm. *P<0.05.

Finally, a corollary of our hypothesis that GATA2 transcriptionally upregulates IGF2 is

that these genes are co-expressed in clinical prostate cancer tissues. Indeed, we previously

observed that IGF2 is upregulated during disease progression in the two clinical transcriptome

datasets (Figure 10(b)). To confirm and extend this finding, we performed an

immunohistochemistry analysis with our cohort of 124 human paraffin embedded prostate cancer

tissues. We observed that IGF2 protein expression increases during the progression from primary

prostate cancer to disseminated CRPC, while the subset of patients treated with taxane

chemotherapy exhibits the highest levels (Figure 24).

62

3.3: IGF2 activates a polykinase program downstream of GATA2

3.3.1: Characterization of dephosphorylation patterns in knockdown models

We next investigated possible downstream mechanisms underlying the GATA2-IGF2

axis. IGF2 exhibits homology with insulin, and IGF2 ligand binding may activate both the

IGF1R and INSR tyrosine kinases (Livingstone, 2013). The receptors phosphorylate scaffolding

intermediates from the IRS and SHC families that in turn activate PI3K and MAPK signaling,

respectively. Accordingly, the downstream effectors of IGF pathway signaling principally

include AKT, JNK, ERK1/2 and p38 and are context dependent.

Immunoprecipitation studies revealed that IGF2 knockdown in the chemotherapy

resistant cultures results in dephosphorylation of both the IGF1R and INSR tyrosine kinases

(Figure 25(a)). We also noted dephosphorylation of IRS and SHC scaffolding intermediaries,

suggesting that IGF2 may signal through both the PI3K and MAPK pathways, respectively

(Figure 25(b)). Indeed, further studies showed that IGF2 knockdown consistently results in

dephosphorylation of AKT and JNK, but not p38, in the three chemotherapy resistant sublines.

In addition, ERK1/2 is dephosphorylated in DU145-DR, but not 22Rv1-DR or ARCaPM-DR.

Moreover, GATA2 knockdown in the chemotherapy resistant sublines recapitulates this

phosphorylation pattern, while an IGF2 expression vector results in a complete rescue (Figure

25(c)). These results suggest that GATA2 regulates a context dependent polykinase program

through IGF2.

63

Figure 25: IGF2 activates a polykinase program through IGF1R and INSR downstream of GATA2. (a) Representative immunoblots of phospho-tyrosine (pTyr) protein levels following immunoprecipitation of IGF1R or INSR in DU145-DR, 22Rv1-DR and ARCaPM-DR cells stably expressing control or IGF2 shRNAs in serum free conditions. (b) Representative immunoblots of relevant IGF signaling protein levels in DU145-DR, 22Rv1-DR and ARCaPM-DR cells stably expressing indicated control or IGF2 shRNAs in serum free conditions. (c) Representative immunoblots of relevant IGF signaling protein levels in DU145-DR, 22Rv1-DR and ARCaPM-DR cells stably expressing indicated control or GATA2 shRNAs as well as a control or IGF2 expression vector in serum free conditions.

64

3.3.2: Chemical inhibition of the polykinase program is functionally significant

We next investigated whether the polykinase program regulated by GATA2 and IGF2 is

functionally significant. To this end, we evaluated the effect of two independent combinations of

well characterized chemical inhibitors of PI3K/AKT (LY294002 and MK2206), JNK (SP600125

and AS601245) and MEK (U0126 and PD98059) on chemotherapy resistance and soft agar

growth as a surrogate assay for tumorigenicity in our three in vitro models. We observed that

combined inhibition of PI3K/AKT and JNK signaling strongly diminishes the chemotherapy

resistance (Figures 26(a) and 27(a)) and soft agar growth (Figures 26(b) and 27(b)) of the three

models. Moreover, as predicted by our immunoblot results (Figure 25), additional MEK/ERK

signaling inhibition in the DU145-DR model further attenuates these properties. Therefore, our

data collectively suggest that GATA2 confers aggressive properties in part through a network of

kinases downstream of IGF2.

65

Figure 26: Chemicals inhibitors of the polykinase program reduce chemotherapy resistance and soft agar growth. (a) Representative colony formation assays and quantifications of DU145-DR, 22Rv1-DR and ARCaPM-DR cells following 72 hour treatment with PI3K (LY294002, 10µM), JNK (SP600125, 10µM) and MEK/ERK (U0126, 1µM) pathway inhibitors alone or in combination with DMSO, docetaxel (125nM) or cabazitaxel (DU145-DR and 22Rv1-DR, 25nM; ARCaPM-DR, 5nM). Data represent the mean ± SD. *P<0.05. (b) Representative soft agar colony formation assays and quantifications of DU145-DR, 22Rv1-DR and ARCaPM-DR cells treated continuously with the cells and conditions described in (a). Data represent the mean ± SD. *P<0.05.

66

Figure 27: Alternative chemicals inhibitors of the polykinase program reduce chemotherapy resistance and soft agar growth. (a) Representative colony formation assays and quantifications of DU145-DR, 22Rv1-DR and ARCaPM-DR cells following 72 hour treatment with AKT (MK2206, 1µM), JNK (AS601245, 5µM) and MEK/ERK (PD98059, 10µM) pathway inhibitors alone or in combination with DMSO, docetaxel (125nM) or cabazitaxel (DU145-DR and 22Rv1-DR, 25nM; ARCaPM-DR, 5nM). Data represent the mean ± SD. *P<0.05. (b) Representative soft agar colony formation assays and quantifications of DU145-DR, 22Rv1-DR and ARCaPM-DR cells treated continuously with the cells and conditions described in (a). Data represent the mean ± SD. *P<0.05.

67

3.4: Dual IGF1R/INSR inhibition improves the efficacy of chemotherapy and survival in

preclinical models

3.4.1: Dual IGF1R/INSR inhibitors improve the efficacy of chemotherapy in vitro

Prompted by our data implicating the GATA2-IGF2 axis in chemotherapy resistant

prostate cancer, we speculated that pharmacological disruption of this pathway may confer

therapeutic benefit. We studied dual inhibitors of the IGF1R and INSR tyrosine kinases, since

this class of drugs is in advanced clinical development. In particular, we focused on OSI-906, a

potent, selective and orally bioavailable dual IGF1R/INSR inhibitor that is currently in phase III

clinical testing for refractory adrenocortical carcinoma (Mulvihill et al., 2009).

We first performed colony formation and flow cytometry experiments with the

chemotherapy resistant cultures in vitro. We observed that monotherapy with OSI-906 or two

other dual IGF1R/INSR inhibitors exhibits negligible activity, while the inhibitors potently

sensitize the three chemotherapy resistant sublines to both docetaxel and cabazitaxel (Figures

28(a) and 28(b)). Moreover, immunoblots showed that exposure to OSI-906 results in the same

context dependent dephosphorylation pattern caused by IGF2 knockdown (Figure 28(c)).

68

Figure 28: Dual IGF1R/INSR inhibition improves the efficacy of chemotherapy in vitro. (a) Representative colony formation assays and quantifications of DU145-DR, 22Rv1-DR and ARCaPM-DR cells following 72 hour treatment with DMSO, docetaxel (125nM) or cabazitaxel (DU145-DR and 22Rv1-DR, 25nM; ARCaPM-DR, 5nM) as well as the dual IGF1R/INSR inhibitors BMS-536924 (1µM), GSK-1904529A (1µM) and OSI-906 (1µM) or vehicle control. Data represent the mean ± SD. (b) Flow cytometry detection of Annexin V and PI levels in cells and conditions described in (a). Data represent the mean ± SD. *P<0.05. (c) Representative immunoblots of relevant IGF signaling pathway protein levels in DU145-DR, 22Rv1-DR and ARCaPM-DR cells following 72 hour treatment with DMSO or OSI-906 (1µM).

69

3.4.2: OSI-906 improves the efficacy of chemotherapy in vivo

For in vivo studies, we established subcutaneous tumors using two patient derived the

xenograft derived (Sramkoski et al., 1999) 22Rv1-DR subline as well as two patient derived

xenograft models termed LPC1 and LPC2. Consistent with our in vitro data and an ongoing

phase II clinical trial in the United States (NCT01533246), OSI-906 exhibits modest activity as a

single agent in our in vivo models (Figure 29). This observation suggests that IGF signaling

inhibition alone is insufficient to significantly constrain the growth of established CRPC tumors.

Figure 29: Monotherapy with OSI-906 does not significantly constrain the growth of established CRPC xenografts. Volumes of subcutaneous 22Rv1-DR, LPC1 and LPC2 xenografts during 28 days of monotherapy with OSI-906 (50mg/kg orally daily) or vehicle control. Data represent the mean ± SD.

In contrast, we observed that OSI-906 strongly restores the efficacy of both docetaxel and

cabazitaxel, as evidenced by tumor volume (Figure 30(a)) and cleaved caspase expression

(Figure 30(b)) during four weeks of combination therapy in vivo. Moreover, pharmacodynamic

studies confirm that OSI-906 reduced phosphorylation of the IGF1R and INSR tyrosine kinases

in these studies (Figure 30(c)).

70

Figure 30: OSI-906 restores the efficacy of chemotherapy in vivo. (a) Volumes of subcutaneous 22Rv1-DR, LPC1 and LPC2 xenografts during 28 days of combination therapy with vehicle, docetaxel (10mg/kg ip. weekly) or cabazitaxel (10mg/kg ip. weekly) as well as OSI-906 (25mg/kg orally four days weekly) or vehicle control. Data represent the mean ± SD. *P<0.05. (b) Representative immunohistochemistry images and quantifications of cleaved caspase 3 levels in subcutaneous 22Rv1-DR, LPC1, and LPC2 xenografts after 14 days of the combination therapy described in (a). Bar, 100µm. Data represent the mean ± SD. *P<0.05. (c) Representative immunohistochemistry images of pIGF1R and pINSR protein levels in subcutaneous 22Rv1-DR, LPC1, and LPC2 xenografts after 14 days of the combination therapy described in (a). Bar, 100µm.

71

3.4.3: OSI-906 improves the efficacy of chemotherapy and survival in preclinical models of

disseminated CRPC without added general drug toxicity

To extend the therapeutic benefit of combination therapy to preclinical models of

disseminated chemotherapy resistant disease, we performed intracardiac (IC) injections of

22Rv1-DR, LPC1 and LPC2 cells as well as toxicity profiles and survival analyses. Similar to a

previous report (Drake et al., 2005), IC injection of 22Rv1-DR, LPC1 and LPC2 cells resulted in

metastatic colonization of bone, liver and other organs, recapitulating disseminated human

disease in vivo (data not shown). We also exploited the in vitro nature of the 22Rv1-DR model to

generate a luciferase labeled subline suitable for in vivo bioluminescent imaging. Our studies

showed that addition of OSI-906 to standard chemotherapy confers therapeutic benefit, as

evidenced by reduced photon flux in 22Rv1-DR (Figure 31(a)) and improved overall survival in

all three independent preclinical models (Figure 31(b)). Notably, combining OSI-906 with

weekly taxane chemotherapy did not instigate a significant increase in general drug toxicity

(Figure 31(c)).

72

Figure 31: OSI-906 improves the efficacy of chemotherapy and survival in preclinical models of disseminated CRPC without added general drug toxicity. (a) Bioluminescence of IC luciferase expressing and IC administered 22Rv1-DR cells during 28 days of combination therapy with vehicle, docetaxel (10mg/kg ip. weekly) or cabazitaxel (10mg/kg ip. weekly) as well as OSI-906 (25mg/kg orally four days weekly) or vehicle control. Data represent the mean ± SD. *P<0.05. (b) Kaplan-Meier survival analysis following IC administration of 22Rv1-DR, LPC1 and LPC2 cells and the combination strategy described (a). *P<0.05. (c) Weight loss in mice following IC administration of 22Rv1-DR, LPC1 and LPC2 cells and the combination strategy described in (a). Data represent the mean ± SD.

73

CHAPTER 4:

Discussion

74

4.1: Chemotherapy resistant CRPC remains an intractable clinical entity

Cancer is a leading cause of death in humans (Jemal et al., 2011). Among the estimated

7-8 million deaths worldwide in 2008, more than 250,000 resulted from prostate cancer. In the

United States, prostate cancer is the most common cancer diagnosis and the second leading cause

of cancer death in males (Siegel et al., 2014). Therefore, significant improvements in the

management of prostate cancer exercise a considerable demographic impact in the United States

and around the world.

The pioneering work of Charles Huggins in 1941 was the first breakthrough in the

management of prostate cancer (Huggins et al., 1941). Notably, Huggins demonstrated that

orchiectomy causes durable remissions in a large majority of prostate cancer patients nearly a

decade before Sydney Farber reported transient responses with antifolates for acute

lymphoblastic leukemia (Farber and Diamond, 1948). Moreover, the elaboration of the

hypothalamus-pituitary axis and the discovery of AR in the following decades led to the

development of GnRH agonists and AR antagonists, respectively, as medical alternatives to

surgical castration (Group, 1967; Liao et al., 1974; Schally et al., 2000). However, the

development of medical castration was marginally significant with respect to prostate cancer

survival (Seidenfeld et al., 2000). Indeed, both medical and surgical castration universally evolve

to CRPC within 1-3 years of therapy (Hellerstedt and Pienta, 2002).

Remarkably, the more than six decades spanning 1941 and 2004 produced no

improvement in the survival of patients with CRPC. During this time, the modern chemotherapy

age transformed the therapeutic landscape for an array of malignancies including leukemia,

lymphoma and testicular cancer, among others (Chabner and Roberts, 2005). However,

75

chemotherapy resistance in CRPC remained an intractable challenge. For example, a review of

twenty six trials conducted between 1987 and 1991 found that the average response rate to

cytotoxic chemotherapy was less than ten percent (Yagoda and Petrylak, 1993). It was therefore

a breakthrough when the two independent phase III clinical trials TAX-327 and SWOG-9916

demonstrated a survival benefit with the second generation taxane docetaxel in 2004 (Petrylak et

al., 2004; Tannock et al., 2004).

The decade since the pivotal docetaxel trials has introduced additional therapeutic options

for CRPC. The phase III TROPIC trial showed that the third generation taxane cabazitaxel

improves survival in CRPC patients progressing with docetaxel (de Bono et al., 2010). In

addition, the second generation androgen signaling inhibitors abiraterone and enzalutamide have

entered widespread clinical practice (Beer et al., 2014; de Bono et al., 2011; Ryan et al., 2013;

Scher et al., 2012). Finally, Sipuleucel-T and radium-223 improved survival in phase III trials

(Kantoff et al., 2010; Parker et al., 2013), although their widespread clinical application is

limited by several considerations including cost, pharmacodynamic markers and exclusion

criteria (Bishr and Saad, 2013). Despite these important advances of the last decade, CRPC

invariably progresses to a therapy resistant state that precedes lethality (Bishr and Saad, 2013).

Indeed, today lethal prostate cancer is a widespread phenomenon that is characterized by

resistance to chemotherapy as well as other treatments. Accordingly, the characterization of

mechanisms of resistance to these modalities is of considerable clinical significance.

4.2: The emerging role of GATA2 in cancer biology

4.2.1: Previous characterization of GATA2 in prostate cancer and other malignancies

76

The role of GATA2 in cancer biology received little interest until recently. A number of

studies have now shown that GATA2 mutations cause heritable disorders characterized

hematopoietic and vascular defects as well as predisposition to MDS and AML (Dickinson et al.,

2011; Hahn et al., 2011; Hsu et al., 2011; Ostergaard et al., 2011). Although sparingly studied,

these mutations most likely represent loss-of-function events (Hahn et al., 2011), leading to

speculation that stem cell exhaustion leads to bone marrow stress and transformation (Migliaccio

and Bieker, 2011). In contrast, gain-of-function mutations that repress differentiation occur in

BC CML (Zhang et al., 2008). Thus, the role of GATA2 in hematopoietic malignancy is

complex and merits further investigation.

A comprehensive study later showed that GATA2 is broadly required for RAS pathway

mutated NSCLC tumorigenesis, in vivo growth and overall survival (Kumar et al., 2012).

Moreover, detailed expression profiling and genome occupancy experiments showed that

GATA2 unexpectedly regulates the proteasome as well as IL1/NFκB and RHO signaling.

Surprisingly, this extensive in vitro and in vivo characterization was later questioned by a report

that GATA2 expression is strongly reduced in lung cancer relative normal tissues and that

GATA2 knockdown does not reduce the viability of cultured RAS pathway mutated NSCLC

cells (Tessema et al., 2014). Therefore, data regarding the role of GATA2 in RAS pathway

mutated NSCLC are conflicting at this time.

In prostate cancer, GATA2 has primarily been studied in the castration naïve context. An

initial study implicated GATA2 in androgen regulated expression of PSA (Perez-Stable et al.,

2000). Subsequent studies clearly established GATA2 as a pioneer factor for AR at several other

androgen regulated loci (Wang et al., 2007; Wang et al., 2009) as well as at the whole genome

level (Chen et al., 2013; Wu et al., 2014). Finally, a recent report suggests GATA2 may be

77

involved in metastatic progression (Chiang et al., 2014).

4.2.2: GATA2 confers aggressiveness in CRPC through a consensus signature

Our data both corroborate the emerging role of GATA2 in cancer and reveal new

insights. On the functional level, our finding that GATA2 is required for the tumorigenicity of

CRPC cells is highly reminiscent of the finding that GATA2 knockdown completely abrogates

the tumorigenicity of RAS pathway mutated A549 and PC9 NSCLC cells (Kumar et al., 2012).

On the other hand, our finding that GATA2 regulates chemotherapy resistance has not been

previously reported. Nonetheless, modifiers of survival signaling are known to broadly regulate

the efficacy of chemotherapeutics (Pommier et al., 2004), and our data therefore complement the

finding that GATA2 regulates survival in RAS pathway mutated NSCLC cells (Kumar et al.,

2012). Finally, a previous study of CML suggested that GATA2 contributes to BC by repressing

the differentiation potential of leukemic cells (Zhang et al., 2008). In this context, it is of interest

that we previously showed that DU145-DR and 22Rv1-DR exhibit loss of epithelial

differentiation markers (Domingo-Domenech et al., 2012), suggesting a possible conserved role

whereby GATA2 promotes a primitive cellular state in cancer.

The implementation of three independent cell line models combined with efficacious

shRNAs and whole genome expression profiling endowed our study with considerable power to

identify bona fide GATA2 regulated genes in CRPC. Indeed, these factors enabled the

identification of a consensus 28 member signature of GATA2 regulated genes that is potently

deregulated in patients with chemotherapy resistant CRPC. Moreover, the signature is rich in

cancer progression associated genes, many of which had not been previously associated with

prostate cancer. For example, TBC1D16 is a small G protein that was recently shown to promote

melanoma proliferation (Akavia et al., 2010) and is upregulated by GATA2 in our CRPC

78

models. Similarly, ADCY9 is an adenylate cyclase gene that was recently shown to promote

MAPK pathway inhibition resistance in melanoma (Johannessen et al., 2013) and is upregulated

by GATA2 in our CRPC models. On the other hand, PPP2R1B is a known lung and colorectal

cancer tumor suppressor gene (Wang et al., 1998) whose expression is derepressed by GATA2

knockdown in our CRPC models. Finally, DPYSL3 is a known metastasis suppressor gene in

prostate cancer (Gao et al., 2010) that is derepressed by GATA2 knockdown in our models.

Together with IGF2, these genes comprise a strong basis to explain the prominent biological

defects instigated by GATA2 knockdown in CRPC cells.

In order to distill the consensus signature to its clinically salient components, we focused

on genes that exhibited the most robust differential expression pattern consistent with regulation

by GATA2 in patient derived datasets. We thereby identified a core subset of 7 genes that

demonstrate both uniform deregulation by GATA2 knockdown in the three chemotherapy

resistant sublines as well as strongly complementary expression patterns during progression to

chemotherapy resistant CRPC in the Grasso et al. and Taylor et al. datasets.

In addition to its highly relevant expression pattern in our models and clinical prostate

cancer tissues, the seven member signature is also strongly enriched in cancer progression

associated genes. For example, PAK4 and FOXM1 are upregulated by GATA2 in our models

and upregulated during progression to chemotherapy resistant CRPC in clinical datasets. PAK4

is a serine/threonine kinase that promotes a number of cancer properties in a range of tumor

types (Radu et al., 2014). In prostate cancer, PAK4 promotes cellular motility (Ahmed et al.,

2008; Wells et al., 2010; Whale et al., 2013) as well as chemotherapy resistance and

tumorigenicity (Park et al., 2013). FOXM1 is a forkhead box transcription factor that promotes a

multitude of cancer properties in a large array of tumor types (Koo et al., 2012). Notably, the role

79

of FOXM1 in prostate cancer has attracted considerable interest. For example, FOXM1 regulates

the initiation and progression of prostate cancer in genetically engineered mouse models (Cai et

al., 2013; Kalin et al., 2006), and FOXM1 cooperates with AR to accelerate DNA replication and

proliferation (Liu et al., 2014). Moreover, a recent study showed that FOXM1 and CENPF

coordinately promote tumor growth and that co-expression of these genes is a robust predictor of

metastasis and survival in clinical databases (Aytes et al., 2014). Finally, our data are

complemented by a recent report that GATA2 regulates FOXM1 in the AR dependent LNCaP

cell line (Chiang et al., 2014).

Likewise, the seven member signature contains genes that are downregulated by GATA2

in our models, are downregulated during progression to chemotherapy resistant CRPC in clinical

datasets and are known to exhibit cancer inhibitory properties. For example, GALNT7 is a

galactosaminyltransferase that represses hepatocellular tumorigenesis (Shan et al., 2013) and

melanoma metastasis (Gaziel-Sovran et al., 2011). Similarly, ARRDC3 represses proliferation,

migration, invasion, soft agar growth and tumorigenicity in breast cancer (Draheim et al., 2010).

Importantly, a focused genetic screen of the seven member signature validated the roles

of PAK4, FOXM1 and ARRDC3 in chemotherapy resistance, soft agar growth or both in three

independent in vitro models. Moreover, GALNT7 regulated soft agar growth in 22Rv1-DR cells,

and the poorly characterized nuclear pore complex protein POM121 regulated both

chemotherapy resistance and soft agar growth in the three models. Finally, IGF2 regulated both

chemotherapy resistance and soft agar growth in the three models and was selected for detailed

mechanistic analysis owing to its potential therapeutic implications. Overall, our genetic screen

contributed to the identification of a discrete network of clinically and biologically validated

GATA2 effectors in CRPC.

80

Until now the investigation of GATA2 biology in prostate cancer has focused almost

exclusively on its role as a pioneer factor for androgen regulated genes. Indeed, studies have

shown that GATA2 is required for the recruitment of AR to several well characterized androgen

regulated genes (Wang et al., 2007; Wang et al., 2009), and this hierarchy likely operates at the

genome wide level (Chen et al., 2013; Wu et al., 2014). In this context, it is of considerable

interest that the DU145-DR and ARCaPM-DR cell lines do not exhibit detectable levels of AR

protein. Moreover, the consensus signature overlapped poorly with published signatures of AR

regulated genes in CRPC, and the core GATA2 regulated genes identified in our studies were

unresponsive to AR knockdown in the AR expressing 22Rv1-DR subline. Therefore, our data

support a prominent AR independent role for GATA2 in prostate cancer for the first time.

Nonetheless, we observed that GATA2 regulates subsets of androgen regulated genes in

the AR expressing 22Rv1-DR subline. Indeed, GATA2 knockdown in 22Rv1-DR cells resulted

in reduced levels of the canonical androgen regulated genes PSA and TMPRSS2. Moreover, our

finding that GATA2 regulates a multitude of cell cycle genes in 22Rv1-DR cells (e.g. CDK1,

CENPF, NUSAP1, PBK, PTTG1 and WEE1) is highly reminiscent of a recent report that AR

preferentially regulates cell cycle genes in an in vitro castrated LNCaP model (Wang et al.,

2009). Therefore, overall our data credential the hypothesis that GATA2 regulates gene

expression in prostate cancer through both AR dependent and independent mechanisms.

Interestingly, this hypothesis is consistent with a recent report conducted with the androgen

dependent LNCaP cell line wherein approximately half of genome wide GATA2 binding sites

did not overlap with those of AR (Wu et al., 2014).

81

4.2.3: The increasingly complex role of GATA2 in prostate cancer

Although the literature related to GATA2 in prostate physiology and disease remains

sparse, the currently published studies as well as ours strongly credential an increasingly central

and multifaceted role. For example, a constellation of data now strongly supports a significant

role for GATA2 in metastatic progression. In an initial study, protein levels of GATA2 in

prostatectomy samples were significantly predictive of biochemical relapse and distant

metastasis (Bohm et al., 2009). Moreover, a second study showed that GATA2 regulates focal

adhesion disassembly and cellular motility in the androgen dependent LNCaP cell line (Chiang

et al., 2014). Notably, our data complement these findings in at least two respects. First, we

found that GATA2 mRNA levels are elevated in disseminated chemotherapy resistant CRPC

tissues from the Grasso et al. and Taylor et al. datasets. Although this observation may be

explained in part by the castration and chemotherapy resistant nature of these tissues, the data are

nonetheless consistent with a role in metastatic progression. This notion is further supported by

our observation that GATA2 protein levels are elevated is paraffin embedded disseminated

CRPC tissues without previous chemotherapy. Second, our expression profiling studies showed

that GATA2 regulates a multitude of metastogenic genes including ARRDC3, DPYSL3,

FOXM1, GALNT7 and PAK4. Therefore, in vitro invasion assays and in vivo metastasis

experiments are strongly warranted at this time in order to definitively establish GATA2 as a

regulator of metastatic progression in prostate cancer.

In addition, several observations suggest that GATA2 promotes castration resistance.

Extensive laboratory and clinical evidence now supports the hypothesis that continued AR

signaling is a critical feature of castration resistance in prostate cancer (Zong and Goldstein,

2013). Moreover, it is now firmly established that GATA2 functions as a pioneer factor that

82

recruits AR to androgen regulated genes (Wang et al., 2009; Wu et al., 2014). These two

observations strongly support a likely role for GATA2 in castration resistance. Indeed, this

notion was indirectly supported by a rigorous study dissecting the role of AR in an in vitro model

of castrated LNCaP cells (Wang et al., 2009). In that study, the transcriptional program regulated

by AR changed significantly during the process of castration. Specifically, the authors reported

that AR predominantly regulates cell cycle related genes in the castrated setting and that AR

regulation of the anaphase promoting complex associated gene UBE2C in particular promotes

castration resistance. Notably, GATA2 binding to an androgen regulated enhancer was required

for AR mediated transcription of UBE2C. Interestingly, although GATA2 did not regulate

UBE2C in our 22Rv1-DR model, GATA2 did impinge on a multitude of cell cycle related genes

that are regulated by AR in castrated LNCaP cells (e.g. CDK1, CENPF, NUSAP1, PBK, PTTG1

and WEE1). Therefore, the published literature as well as our results invite a detailed

investigation of the likely role of GATA2 in castration resistance.

Finally, it is noteworthy that GATA2 has been implicated in both urogenital development

and adult prostate physiology. Although it did not specifically comment on the prostate, one

study used a Gata2 YAC in Gata2-/- embryos to demonstrate that GATA2 is indispensable for

normal urogenital development (Zhou et al., 1998). Furthermore, the authors showed that

GATA2 is expressed in the urogenital sinus, the embryonic antecedent of the adult prostate. In

addition, a recent study showed that GATA2 is expressed in adult prostate and may play a role in

adult prostate physiology (Rosa-Ribeiro et al., 2014).

Collectively, the published literature as well as our data highlight an intriguing

constellation of observations: GATA2 is likely involved in prostate development and physiology

as well as nearly every aspect of prostate cancer biology. These observations are highly

83

reminiscent of so called “lineage oncogene” transcription factors that have been characterized in

other tumor types. For example, ETV1 is expressed in and required for the development of

interstitial cells of Cajal, the putative cell-of-origin for gastrointestinal stromal tumors (GISTs),

and ETV1 is required for the in vivo growth of GIST cells (Chi et al., 2010b). Similarly, MITF is

required for development of the melanocytic lineage (Steingrimsson et al., 2004), and MITF

regulates the soft agar growth and chemotherapy resistance of melanoma cells (Garraway et al.,

2005). Studies in genetically engineered mouse models are needed to definitively establish the

role of GATA2 in prostate development and malignant transformation. Nonetheless, current

evidence strongly credentials the notion that GATA2 is a lineage specific transcription factor that

regulates prostate development, physiology and many facets of malignancy.

4.3: A novel role for IGF2 in prostate cancer

IGF2 has attracted considerable investigation over the decades, and today several aspects

of IGF2 biology in cancer are well described (Livingstone, 2013). First, IGF2 has emerged as a

classical oncofetal gene. Early studies showed that IGF2 is widely expressed in developing

embryos in rodents and humans (Han et al., 1987; Stylianopoulou et al., 1988), and the advent of

genetically engineered mouse models revealed its role as an important fetal growth hormone

(DeChiara et al., 1990). The discovery of imprinting stimulated a surge of interest in IGF2 and

the identification of IGF2 LOI in a wide range of malignancies (Cui, 2007). Interestingly, an

early whole genome expression profiling study found IGF2 to be the single most overexpressed

transcript in colorectal cancer (Zhang et al., 1997).

The discovery of widespread IGF2 overexpression in cancer stimulated interest in its

functional properties. Indeed, a multitude of genetically engineered mouse models revealed a

84

significant role in tumorigenesis. For example, IGF2 overexpression is sufficient to initiate breast

tumors (Bates et al., 1995; Pravtcheva and Wise, 1998) as well as several other malignancies in

mice (Moorehead et al., 2003; Rogler et al., 1994). Similarly, IGF2 modulates the penetrance of

large T antigen induced islet cell tumors (Christofori et al., 1994) and PTEN deficient breast

tumors (Church et al., 2012), and IGF2 is indispensable for the formation of PTCH deficient

medulloblastoma and rhabdomyosarcoma (Hahn et al., 2000). In addition, several studies

suggest that IGF2 may regulate angiogenesis (Kim et al., 1998; Yao et al., 2012) and metastasis

(Diaz et al., 2007; Pravtcheva and Wise, 1998; Sciacca et al., 2002). Finally, although modifiers

of survival signaling are known to broadly impinge on the efficacy of chemotherapeutics

(Pommier et al., 2004), studies implicating IGF2 in chemotherapy resistance have been

exceedingly scarce (Ogawa et al., 2010).

The role of IGF2 in prostate cancer has received little attention. An early study

implicated LOI in primary prostate cancer (Jarrard et al., 1995), and more recently it was shown

that IGF2 may contribute to steroidogenesis (Lubik et al., 2013). Therefore, our data break

considerable ground in characterizing the role of IGF2 in prostate cancer. First, we showed that

chemotherapy resistant CRPC models exhibit a strong and functionally significant IGF2

autocrine loop. Although IGF2 has scarcely been implicated in chemotherapy resistance, this

finding is highly consistent with a large literature implicating survival factors in this property.

Interestingly, soft agar growth and tumorigenicity are generally accepted as complementary

measures of cellular transformation (Shin et al., 1975). Our finding that IGF2 regulates these

properties is therefore consistent with genetically engineered mouse models that showed a role

for IGF2 in tumorigenesis. In addition, ChIP studies, reporter assays and functional genetic

experiments supported IGF2 as a mediator of GATA2 biology. Notably, our laboratory models

85

were complemented by results from both public domain datasets and paraffin embedded tissues,

which showed that IGF2 expression increases during prostate cancer progression and reaches its

highest levels in disseminated CRPC patients treated with chemotherapy. Therefore, overall our

data support a greatly expanded role for IGF2 in prostate cancer.

4.4: Therapeutic implications of the GATA2-IGF2 axis

Our studies revealed that GATA2 confers aggressiveness in CRPC through increased

chemotherapy resistance and tumorigenicity. Moreover, other reports as well data form our

studies further support significant roles in metastasis, castration resistance, and possibly

tumorigenesis. Overall, these data strongly credential GATA2 as a promising therapeutic target

in prostate cancer. In this context, it is noteworthy that whole body inducible knockout of Gata2

in adult mice provokes no detectable toxicity (Kumar et al., 2012).

Although the therapeutic targeting of transcription factors is generally considered

impracticable, is not without precedent. This notion is perhaps most strongly evidenced by

decades of clinical experience with breast and prostate cancer. For example, tamoxifen (Shiau et

al., 1998) and bicalutamide (Bohl et al., 2005) bind to the ligand binding domains of estrogen

receptor and AR, respectively, and both agents are mainstays of therapy today. Moreover,

mounting evidence suggests that inhibition of several other transcription factors, including the

STATs, NFκB and NOTCH1, among others, may enter clinical practice (Yeh et al., 2013). In

addition, characterizing in detail the protein complex that facilitates GATA2 transcription may

yield clinically relevant insights. For example, the histone acetyltransferase (HAT) EP300

acetylates GATA2 and thereby increases its transcriptional activity (Hayakawa et al., 2004), and

HATs are potentially viable therapeutic targets (Manzo et al., 2009).

86

Alternatively, focusing on the effectors of GATA2 biology may be a practicable strategy.

In this context, the serine/threonine kinase PAK4 is of particular interest. PAK4 is upregulated

and responsive to GATA2 knockdown in our chemotherapy resistant CRPC models. Moreover,

PAK4 is upregulated in clinical CRPC tissues treated with chemotherapy and PAK4 knockdown

in our models reduced chemotherapy resistance and soft agar growth. These results are highly

consistent with a recent study wherein stable PAK4 knockdown reduced the tumroigenicity and

chemotherapy resistance of PC3 cells (Park et al., 2013). Moreover, the small molecule PAK4

inhibitor PF-3758309 exhibits therapeutic activity with breast, colon, GIST and melanoma

xenografts (Murray et al., 2010). Although PF-3758309 did not enter advanced clinical

development owing to pharmacokinetic limitations (Dart and Wells, 2013), these data comprise

proof of principle that targeting PAK4 downstream of GATA2 may confer therapeutic benefit.

IGF2 was the most biologically and therapeutically meritorious candidate among the

GATA2 consensus signature genes. IGF2 was strikingly upregulated and responsive to GATA2

knockdown in our models and it exhibited a complementary expression pattern during

progression to chemotherapy resistant CRPC in multiple clinical databases. Moreover, functional

genetic experiments supported IGF2 as a significant mediator of GATA2 biology. Notably, these

data stimulated the preclinical characterization of a combination strategy involving OSI-906, a

dual IGF1R/INSR inhibitor currently in phase III testing. These findings have important

implications for both the interpretation of past clinical trials and the design of future studies.

The first generation of IGF signaling inhibitors to enter clinical development was IGF1R

antibodies, and these therapeutics unfortunately met with consistent failure in phase II and phase

III studies. For example, CP-751,871 was investigated for lung cancer in a phase III trial (Langer

et al., 2014), R1507 was investigated for lung cancer in a large randomized phase II trial

87

(Ramalingam et al., 2011) and AMG479 was investigated for breast (Robertson et al., 2013) and

pancreatic (Cohn et al., 2013) cancer in large randomized phase II trials. Notably, CP-751,871

failed to improve survival in combination with docetaxel for CRPC in a large randomized phase

II study (de Bono et al., 2014).

The disappointing results from these trials stimulated considerable interest in

understanding the mechanisms whereby cancer cells persist under IGF1R blockade (Pollak,

2012; Yee, 2012). Interestingly, IGF1R blockade in humans provokes increased levels of GH,

IGF1 and insulin (Atzori et al., 2011; Haluska et al., 2007; Tolcher et al., 2009). Elevated levels

of IGF1 may activate residual IGF1R and bind weakly to INSR, while elevated levels of insulin

in particular would be expected to robustly drive PI3K and MAPK signaling downstream of

INSR. Moreover, the high serum IGF2 levels physiologically present in humans would be

expected to efficiently activate INSR and downstream signaling irrespective of IGF1R blockade

(Pollak, 2012; Yee, 2012).

In addition to these general considerations, our data provide a further possible

explanation for the failure of IGF1R blockade, particularly in the setting of CRPC. Exposure to

docetaxel provoked a powerful IGF2 autocrine loop in our in vitro models. Moreover, we

showed for the first time that IGF2 is elevated in CRPC tissues treated with chemotherapy from

both public domain datasets and paraffin embedded tissues. Importantly, IGF2 efficiently

signaled through INSR in our three independent in vitro models, and this phenomenon would

also be expected in prostate cancer tissues. Therefore, our data predict that IGF1R blockade

would be ineffectual in the setting of docetaxel treatment for CRPC.

Overall, our results suggest that a detailed knowledge of the relevant molecular

pathophysiology may be instrumental in the design of future clinical trials, particularly those

88

involving targeted therapies. Indeed, it is possible that our characterization of IGF2 in CRPC

would have discouraged the design of a clinical trial with a combination of chemotherapy and

IGF1R blockade (de Bono et al., 2014). Fortunately, a second generation of IGF signaling

inhibitors has emerged which may circumvent IGF1R blockade resistance involving IGF2 and

INSR. Indeed, OSI-906 is a potent, selective and orally bioavailable small molecule inhibitor of

IGF1R and INSR that is in phase III development (Mulvihill et al., 2009), and OSI-906 improved

survival in combination with chemotherapy in our preclinical models. Moreover, several other

relevant therapeutics are currently in phase I and phase II development. For example, BMS-

754807 is a small molecule with similar properties to OSI-906 (Carboni et al., 2009; Wittman et

al., 2009), and MEDI-573 and BI-836845 are antibodies that neutralize IGF1 and IGF2

(Friedbichler et al., 2014; Gao et al., 2011). As we have shown, these therapeutics are better

suited to the molecular pathophysiology of CRPC, and they may succeed in improving survival

in a challenging clinical context in which previous strategies have failed.

89

MATERIALS AND METHODS

90

Taxanes, targeted pathway inhibitors and other reagents:

Docetaxel and was obtained from Sigma-Aldrich. Cabazitaxel as well as LY294002, MK2206,

SP00125, AS60125, U0126, PD98059, BMS-563924, GSK-1904529A and OSI-906 were

obtained from Selleck Chemicals. Anti-IGF2 (clone S1F2) was obtained from Millipore. IgG1

isotype control and human recombinant IGF2 were obtained from R&D.

Docetaxel resistant prostate cancer cell models:

The sublines DU145-DR and 22Rv1-DR were previously described (Domingo-Domenech et al.,

2012). ARCaPM cells were obtained and maintained in Prostate Epithelial Cell Medium (both

Novicure Biotechnology) supplemented with 10% FBS. Cells were grown at 37oC in a

humidified atmosphere with 5% CO2. Resistant clones were selected by culturing cells with

docetaxel dissolved in DMSO for 72 hours. A dose escalation strategy was implemented for 7

months until a concentration of 1µM was reached. In parallel, parental ARCaPM cells were

exposed to equal volumes of DMSO.

Colony formation and cell viability assays:

Clonogenic survival assays in response to drug treatment were performed by plating 103 cells in

35mm culture dishes. After 24 hours cells were treated with vehicle controls or with drugs for 72

hours. After 10-14 days cells were fixed with 4% paraformaldehyde in PBS, stained with crystal

violet solution and formed colonies of ≥50 cells were quantified microscopically. Cell viability

was analyzed using the Cell titer 96 Aquos Non-Reactive Cell Proliferation Assay (MTs) kit

(Promega). Cells were seeded at a density of 104 in 96 well culture dishes and 24 hours later

medium was removed and replaced with new medium containing vehicle control or drugs. After

72 hours, color absorbance was measured on a microplate spectrophotometer (Molecular

Dynamics) at 450 nm (test wavelength) and 620 nm (reference wavelength). The percentage of

91

surviving cells was estimated by dividing the A450nm-A620nm of treated cells by the A450nm-

A620nm of control cells.

Immunoblot and immunoprecipitation:

Whole cell extracts from cells grown with or without serum were prepared in sample buffer and

analyzed by immunoblot using standard procedures. Extracts from xenograft tumors were

obtained using a tissue homogenizer (Fisher Scientific). For immunoprecipitation, extracts were

incubated with the indicated antibodies overnight at 4°C. Following a 2 hour incubation with

protein A/G Dynabeads (Invitrogen), beads were washed four times, resuspended in 1x Laemmli

sample buffer and boiled for 5 minutes. SDS-PAGE resolved proteins were transferred to PVDF

membranes and incubated with primary antibodies. Secondary antibodies were used at 1:5000.

Antibodies:

Unless otherwise specified, the following antibodies were used for all reported applications.

Primary antibodies against GATA2, IGF2 and ß-Actin were obtained from Sigma-Aldrich.

Primary antibodies against cleaved PARP, pTyr, pIGF1R (immunoblot), IGF1R, pINSR

(immunoblot), INSR, pSHC1, SHC1, pAKT, AKT, pp38, p38, pERK1/2, ERK1/2, cleaved

Caspase 3 and AR were obtained from Cell Signaling Technology. Primary antibodies against

pIRS1, IRS1, pJNK1-3, JNK1-3 and pIGF1R (immunohistochemistry) were obtained from Santa

Cruz Biotechnology. Primary antibody against GATA2 was obtained from Novus Biologicals

(immunoprecipitation). Primary antibody against pINSR (immunohistochemistry) and PSMA

(immunohistochemistry) were obtained from Abcam.

92

Animal husbandry and surgical castration:

All mouse procedures were performed with NOD.Cg-Prkdcscid IL2rgtm1Wjl (NSG) mice obtained

from Jackson Laboratories. All protocols for mouse experiments were in accordance with

institutional guidelines and were approved by the Mount Sinai Medical Center Institutional

Animal Care and Use Committee (IACUC). Unless specified otherwise all mouse experiments

were performed using castrated male animals. For castration, anesthetized and surgically

prepared animals were placed in dorsal recumbency. Both testes were then pushed down into the

scrotal sacs by pressuring the abdomen. A 1cm incision was made in the scrotum to expose the

tunica. The tunica was pierced, the testes were pushed out one at a time and then raised to expose

the underlying blood vessels. The vas deferens with the prominent blood vessels running along

them were located using a forceps and the testis were dissected away from the fat and removed.

The vas deferens and ducts were then replaced back into the tunica, and skin incisions were

closed with stainless steel wound closures and removed after 10 days.

Tumorigenicity assay:

Limiting dilutions of cells were implanted subcutaneously in a 1:1 mixture of complete growth

medium and Matrigel (BD Biosciences). Tumor incidence (number of tumors per number of

injections) and tumor latency (time from injection to first tumor palpability) were evaluated

weekly. Tumors formed were confirmed histologically. When tumors became palpable at a

single injection site, they were surgically removed to allow continued evaluation of other sites.

Mice were monitored for up to 36 weeks, and animals with no sign of tumor formation were

examined at necropsy for confirmation.

93

Analysis of apoptosis by flow cytometry:

Adherent and detached cells were pooled, washed and labeled with Annexin-V-FITC and PI

using the Annexin-V-FLUOS Staining Kit (Roche) in accordance with manufacturer’s

instructions. Samples were acquired with a FACscan Flow Cytometer (BD Biosciences) and

analyzed with CellQuest Pro software (BD Biosciences) to determine the percentage of cells

displaying Annexin V and PI staining.

RNA extraction and qRT-PCR:

For cultured cells as well as explanted tumors total RNA was isolated using the RNeasy Mini kit

(Qiagen) in accordance with manufacturer’s instructions. Complementary DNA was synthesized

from equivalent concentrations of total RNA using the SuperScript III First-Strand Synthesis

SuperMix Kit (Invitrogen) in accordance with manufacturer’s instructions. Primer sequences

used for amplification experiments are shown below:

Gene Forward Sequence (5' to 3') Reverse Sequence (5' to 3') ACTB CATGTACGTTGCTATCCAGGC CTCCTTAATGTCACGCACGAT ARRDC3 ATGCAAGAGGACATGCGAAAG GTGGAAGCCTTCTTCGGAATTAT FOXM1 TGCAGCTAGGGATGTGAATCTTC GGAGCCCAGTCCATCAGAACT GALNT7 GGTTCATCTTACGCAGTTTGCT GTGGAAGCCTTCTTCGGAATTAT GATA2 CATCAAGCCCAAGCGAAGA TTTGACAGCTCCTCGAAGCA IGF2 CCGTGCTTCCGGACAACT GGACTGCTTCCAGGTGTCATATT PAK4 TAGGCCATTTGTCCTGGAGTTT GTTCATCCTGGTGTGGGTGAC

POM121 AGTGGCAGTGGACATTCAGC CGTAAGCGCCTGTCAAGGA NGRN ATGGCGGTTACCCTGAGTCT GGAATCGGATTGCTTGTTTCTGT CENPF CTCTCCCGTCAACAGCGTTC GTTGTGCATATTCTTGGCTTGC KLK3 CACAGGCCAGGTATTTCAGGT GAGGCTCATATCGTAGAGCGG TMPRSS2 GTCCCCACTGTCTACGAGGT CAGACGACGGGGTTGGAAG KRT 18 GAGACGTACAGTCCAGTCCTTGG CCACCTCCCTCAGGCTGTT KRT 19 CTGCGGGACAAGATTCTTGGT CCAGACGGGCATTGTCGAT HLA A AAAAGGAGGGAGTTACACTCAGG GCTGTGAGGGACACATCAGAG HLA B CAGTTCGTGAGGTTCGACAG CAGCCGTACATGCTCTGGA NOTCH2 GGCATTAATCGCTACAGTTGTGTCT GGAGGCACACTCATCAATGTCA HES1 CTGGAAATGACAGTGAAGCACCT ATTGATCTGGGTCATGCAGTTG

94

SMO ATGGATGGTGCCCGCCGAGAG ATGGTCTCGTTGATCTTGCTGG GLI1 CCCAACTCCACAGGCATAC ACACGAACTCCTTCCGCTCC GLI2 ACACCAACCAGAACAAGCAG AGTCTTCCCAGTGGCAGTTG

Coding sequences for genes of interest as well as a loading control (ACTB) were amplified from

500ng of complementary DNA using QuantiTect SYBR Green PCR Kit (Qiagen). Amplification

was carried out using an Mx3005P qPCR System (Agilent Technologies). Cycle threshold values

were determined and normalized to the loading control for each experiment. Fold changes for

experimental groups relative to respective controls were calculated using MX Pro software

(Agilent Technologies).

Clinical transcriptome dataset mining:

A gene was considered deregulated during the progression from primary to disseminated

chemotherapy resistant disease in GSE35988 (Grasso et al., 2012) and GSE21032 (Taylor et al.,

2010) if it exhibited multiple testing significance (FDR<0.05) using the significance analysis of

microarrays (SAM) algorithm (Tusher et al., 2001). For analyses of representation of the 28 gene

GATA2 signature and AR signatures in primary and disseminated chemotherapy resistant

disease, we calculated a Pearson correlation coefficient comparing the average expression of the

signatures and the expression data for each patient sample as previously described (Liu et al.,

2007). Briefly, the expression levels of genes in control shRNA expressing cells divided by the

expression levels of genes in their corresponding GATA2 shRNA expressing cells in the three

chemotherapy resistant cultures were averaged, log transformed and the resulting ratios were

used as expression levels for genes in the two signatures. Similarly, for the downloaded clinical

transcriptome datasets expression levels were calculated by log transformation of the signal

intensity for each probe divided by the average signal intensity of that probe in the entire dataset.

For generation of the seven gene core subset from the consensus 28 gene signature, genes

95

exhibiting significantly deregulated expression (FDR<0.1) during the progression to

disseminated chemotherapy resistant disease as well as the expression pattern predicted by

GATA2 knockdown in the chemotherapy resistant models were selected. For generation of the

CRPC AR signature from the Wang et al. study, datasets were downloaded from GEO

(GSE11428) and genes were considered significantly deregulated between ‘abl-siControl’ and

‘abl-siAR’ biological replicates if they exhibited multiple testing significance (FDR<0.05) as

well as fold change ≥1.5 using the SAM algorithm.

Human paraffin embedded prostate cancer tissues:

Human formalin fixed paraffin embedded primary (n=56), metastatic CRPC (n=35) and taxane

treated metastatic CRPC (n=33) tissue samples were collected from the Mount Sinai Medical

Center tumor biorepository under an Institutional Review Board (IRB) approved protocol. All

tissue sections were reviewed by a pathologist to confirm prostate cancer origin.

Immunohistochemistry:

Immunohistochemistry analyses were conducted on prostate cancer formalin fixed paraffin

embedded tissue sections from human samples and cell line (22Rv1-DR) or LPC xenografts.

Tissue sections (5µm) were deparaffinized and submitted to standard peroxidase based

immunohistochemistry procedures. Quantification of positive cells was determined by counting

the number of tumor cells in 10 contiguous high power fields in three different areas of each

section, and referred to the total number of counted cancer cells.

Generation of stably expressing shRNA sublines:

For shRNA mediated suppression of GATA2, two clones (GATA2_2668 and GATA2_2829)

and a non-targeting (Renilla) control were selected following a screen of a custom library.

Briefly, de novo predictions were obtained using the DSIR algorithm (Vert et al., 2006) and

96

guide strands were subsequently filtered using “sensor rules” to enrich for predictions harboring

sequence features associated with effective shRNAmir processing and potent knockdown

(Fellmann et al., 2011). The control and GATA2 shRNAs were cloned into the LMP (Dickins et

al., 2005) vector. Retroviral particles were generated by transient co-transfection of 2x106 Eco

Phoenix cells in 10cm dishes using FuGene (Roche) with 4µg of a VSVG packaging plasmid and

15µg of individual shRNA plasmids. For shRNA mediated suppression of IGF2, two clones

(V3LHS_400817 and V3LHS_382742) and a non-targeting (scrambled) control (RHS4346)

were selected following a screen of commercially available libraries (Thermo Fisher Scientific).

Lentiviral particles were generated by transient co-transfection of 2x106 HEK293T cells in 10cm

dishes using FuGene with 3µg each of pMD.G and CMVR8.9l packaging plasmids and 6µg of

individual shRNA plasmids. For both retrovirsus and lentiviruses, infectious supernatant was

collected 48 and 72 hours after transfection, concentrated by ultracentrifugation, resuspended in

full growth medium and added to cell cultures with 10µg/ml of polybrene (Sigma-Aldrich). 48

hours after infection at high MOI (> 95% infection efficiency), pools of shRNA expressing cells

were purified with a FACSAria II Cell Sorter (BD Biosciences) using GFP expression. First

passage stocks were validated by qRT-PCR and immunoblot before being stored in liquid

nitrogen. All studies with stably expressing shRNA sublines were performed with pools

passaged no more than five times.

RNA sequencing and expression analysis:

To characterize the transcriptional program regulated by GATA2, we performed RNA

sequencing of first passage chemotherapy resistant sublines stably expressing shRNAs. For each

chemotherapy resistant culture, we analyzed shControl-, shGATA2#1- and shGATA2#2-

expressing cells in biological triplicates. RNA was extracted and its integrity was checked by

97

either a 2100 Bioanalyzer using the RNA 6000 Nano assay or a 2200 TapeStation using the R6K

ScreenTape (both Agilent). All processed total RNA samples had RIN value of 7.7 or greater.

The sequencing library was prepared with the standard TruSeq RNA Sample Prep Kit v2

protocol (Illumina). Briefly, total RNA was poly-A-selected and then fragmented. The cDNA

was synthesized using random hexamers, end-repaired and ligated with appropriate adaptors for

sequencing. The library then underwent size selection and purification using AMPure XP beads

(Beckman Coulter). The recommended (Illumina) 6bp barcode bases were introduced at one end

of the adaptors during the PCR amplification step. The size and concentration of the RNAseq

libraries was measured by Bioanalyzer and Qubit fluorometry (Life Technologies) before

loading onto the sequencer. The mRNA libraries were sequenced on the HiSeq 2500 System

with 100 nucleotide single-end reads, according to the standard manufacturer's protocol

(Illumina). Obtained reads were then aligned to the UCSC hg19 human genome using Tophat

software (Trapnell et al., 2009) and gene level expression values obtained using Cufflinks

(Trapnell et al., 2010). For each chemotherapy resistant culture, the Cuffdiff utility was used to

identify differentially expressed genes. Genes were considered to be regulated by GATA2 if they

exhibited a fold change difference (FC≥1.5) as well as multiple testing significance (Padj<0.05)

in both shGATA2#1 and shGATA2#2 expressing cells. The GATA2 regulated genes from each

chemotherapy resistant culture were analyzed with the Ingenuity Pathways Analysis program

(http://www.ingenuity.com/index.html) to identify molecular and cellular functions enriched in

these gene lists. Categories with a significance of P<0.01 were considered.

siRNA knockdown:

Predesigned Silencer Select (Ambion) siRNAs were transfected into cells seeded overnight at a

density of 1x105 in six well plates. siRNAs and RNAiMAX lipofectamine (Life Technologies)

98

were diluted in OptiMEM (Life Technologies), combined and added dropwise to cells at a final

concentration of 10nM. Knockdown efficiency was assayed 24 hours later by qRT-PCR and for

each gene the two most independently efficacious siRNAs were selected for chemotherapy

resistance and soft agar colony formation assays.

Gene Clone Catalog # Control#1 - 4390843 FOXM1 S5248 4392420 FOXM1 S5250 4392420 PAK4 S20134 4390824 PAK4 S20135 4390824 POM121 S19145 4392420 POM121 S59623 4392420 IGF2 S7214 4392420 IGF2 S7115 4392420 SMO S13164 4392420 SMO S13165 4392420 NOTCH2 S9638 4392420 NOTCH2 S9639 4392420

Generation of stably expressing cDNA sublines:

GATA2 (PLOHS_ccsbBEn_10902), IGF2 (PLOHS_100010430), NGRN

(PLOHS_ccsbBEn_11979), GALNT7 (PLOHS_100074154) and ARRDC3

(PLOHS_100071659) clones from the ORFeome Collaboration and an RFP (OHS5832) control

were obtained in the Precision LentiORF lentiviral expression plasmid (Open Biosystems).

Lentiviral particles were generated by transient co-transfection of 2x106 HEK293T cells in 10cm

dishes using FuGene with 3µg each of pMD.G and CMVR8.9l packaging plasmids and 6µg of

cDNA plasmid. Supernatant was collected 48 and 72 hours after transfection, concentrated by

ultracentrifugation, resuspended in full growth medium and added to shRNA cultures with

10µg/ml of polybrene (Sigma-Aldrich). 48 hours after infection at high MOI (> 95% infection

99

efficiency), pools of cDNA expressing cells were selected with blasticidin for 5-7 days.

Retroviral particles driving the expression of myristoylated AKT (myrAKT) was a kind gift of

Dr. Adolfo Ferrando (Columbia University, New York). The vector containing a scramble

sequence that doesn’t target any human sequence was used as a control (Empty Vector), pools of

cells were selected with puromycin to generate stable cell lines. Low passage stocks were

validated by qRT-PCR and all studies with stably expressing shRNA-cDNA sublines were

performed with pools passaged no more than three times.

Soft agar assay:

Anchorage independent growth was determined by plating 1x103 cells in 35mm culture dishes in

0.33% Seaplaque agarose (Lonza) and cultured in RPMI supplemented with 10% fetal calf

serum. Colonies were counted under light microscopy after 21 days. For studies involving

continuous treatment with kinase inhibitors, antibodies and recombinant IGF2, growth medium

was replaced every 72 hours with fresh reagents.

ELISA:

Levels of secreted IGF2 were measured in cells cultured for 72 hours in serum free conditions

and using a human IGF2 ELISA kit (MyBiosource) in accordance with manufacturer’s

instructions. A microplate reader (Dynex Technologies) was used to detect the signal at 450nm.

Identification of predicted GBEs:

Evolutionarily conserved GBEs in the human IGF2 promoter were identified computationally

using multiple sequence alignment software (Ovcharenko et al., 2005). Multiple gene sequences

containing the IGF2 promoter regions were analyzed using the TRANSFAC database with

default settings.

100

Chromatin immunoprecipitation:

DU145-DR, 22Rv1-DR and ARCaPM-DR cells were crosslinked with paraformaldehyde and

nuclear extracts obtained using truChIP High Cell Chromatin Shearing Kit with SDS Shearing

Buffer (Covaris). Nuclear extracts were incubated with protein G magnetic beads (Millipore)

conjugated to anti-GATA2. The solution was then successively washed with low salt, high salt,

and LiCl buffers prior to elution. Following RNAse and proteinase K digestion and DNA

extraction, the immunoprecipitated and control (input) DNA was analyzed by qPCR using the

following primers:

Forward Sequence (5' to 3') Reverse Sequence (5' to 3') GBE1 GAGTTGGCAGGAGTCATGG GGGACTCCTGTTTGGAAATG Control 1 CAGAAGCCCACCCTGGTATGTT CATCCTGGCCCCGCCCACC GBE2 CACTGCCCTGTGCAGAGATG CAGCCAGTCTCTTCTCTCCTG Control 2 CACTGACCAGCCTGCAAACT GCTCAGCAGAAGGCTCGCTG

Luciferase reporter assay, constructs, and site directed mutagenesis:

The IGF2 P4 promoter sequence (-500 to +100 nucleotides) was amplified by PCR and cloned

into the pGL3 luciferase reporter vector (Promega). The resulting vector was used as a template

to mutate GATA2 binding elements by site directed mutagenesis using the following primers:

GBE Primer Sequence

1 Forward ACATTCGCGACAaaaaaaaaACACCCCCTCTGGCTCTGT Reverse ACAGAGCCAGAGGGGGTGTttttttttTGTCGCGAATGT

2 Forward ATGGCACTTCGCTAaaaaaaaaCAGGGCTCCCGGTTGGGGGTGCAG

GAGA

Reverse TCTCCTGCACCCCCAACCGGGAGCCCTGttttttttTAGCGAAGTGCCAT

After synthesizing the mutant strand by PCR, template sequence was digested with DpnI

restriction enzyme for 2 hours at 37°C. The mutant vector was transformed into competent cells

for nick repair, plasmid DNA was recovered and mutation of the binding site was confirmed by

101

Sanger sequencing. For the reporter assay, NIH3T3 cells were seeded into 6 well plates at a

density of 3x105 and allowed to attach overnight. Transfection mix was prepared by combining

3µg of pGL3, 1µg of pCMV-Renilla, and 3µg of a GATA2 expression vector

(PLOHS_ccsbBEn_10902) or RFP control (both Open Biosystems). Cells were lysed 48 hours

after transfection with 500µl of Luciferase Assay Buffer (Dual Glo, Promega) and 100µl of

lysate were analyzed in an automatic luminometer. 100µl of Stop & Glo reagent were then added

and Renilla luminescence measured after 10 minutes of incubation. Ratios of Firefly versus

Renilla luciferase were calculated to determine promoter activity.

Monitoring of subcutaneous xenograft growth:

For in vivo studies involving chemotherapy and/or OSI-906, subcutaneous xenografts were

generated by implantation of 106 indicated prostate cancer cells in a 1:1 mixture of growth

medium and Matrigel into the flanks of NSG mice. When subcutaneous tumors became palpable,

mice were randomly assigned to treatment groups containing four animals. The vehicles for

chemotherapy and OSI-906 were 10% DMSO in sterile 1xPBS and 25mM tartaric acid,

respectively. Tumor dimensions were monitored weekly using Vernier calipers. Tumor volume

was calculated according to the formula V=(a2xb)/2 where a and b are the minimal and maximal

diameter in millimeters, respectively. In accordance with institutional guidelines, mice bearing

subcutaneous xenografts greater than 500mm3 were sacrificed. Explanted tumors were weighed,

formalin fixed, and embedded in paraffin for pathological analysis.

IC injections:

IC injections were performed as previously described (Campbell et al., 2012). Briefly, the ventral

thorax of 3-4 week old castrated males was shaved prior to anesthesia with an isoflurane

vaporizer and nose cone. The thorax was sterilized with iodine and alcohol and a sterile marker

102

was used to mark a location half way between the sternal notch and the xyphoid process. 100µl

from a 1x106cell/ml suspension of 22Rv1-DR-luciferase, LPC1 or LPC2 cells in sterile 1xPBS

was drawn into a 30.5 guage needle. The upright syringe was gently inserted through the mark

and for each injection successful penetration into the left ventricle was confirmed visually by a

pulse of bright red blood into the syringe. Following each experiment a detailed necropsy was

performed to grossly and histologically confirm disseminated tumor burden.

In vivo bioluminescence imaging:

Imaging was performed using an IVIS Spectrum imager (Xenogen). Animals were anesthetized

using an isoflurane vaporizer and placed onto the warmed stage inside the camera box. Animals

next received intraperitoneal luciferin (200mg/kg) five minutes prior to imaging. For

quantification, rectangular regions of interest (ROIs) incorporating the entire animal were

measured. The signal was measured in photons per second using Living Image software.

General toxicity monitoring:

Body weights for every mouse were recorded every three days and fluctuations were computed

by the percentage of current body weight relative to baseline. When animals showed signs of

weight loss therapy was discontinued until resolution of the toxicity and reinitiated at a 50% of

the initial dose. In accordance with institutional guidelines all animals experiencing greater than

20% weight loss were sacrificed.

Statistical analyses:

Statistical analysis was carried out with SPSS software unless otherwise specified. Experimental

data expressed as mean ± SD were analyzed by Student’s t-test. All t-tests were conducted at the

two sided 0.05 level of significance. For genomic analyses, multiple testing significance was

calculated using Significance Analysis of Microarrays (SAM) or Tuxedo software. For

103

preclinical studies, survival analyses were performed using the Kaplan–Meier method and curves

were compared by the log rank test.

104

REFERENCES

105

Ahmed, T., Shea, K., Masters, J.R., Jones, G.E., and Wells, C.M. (2008). A PAK4-LIMK1 pathway drives prostate cancer cell migration downstream of HGF. Cellular signalling 20, 1320-1328.

Akavia, U.D., Litvin, O., Kim, J., Sanchez-Garcia, F., Kotliar, D., Causton, H.C., Pochanard, P., Mozes, E., Garraway, L.A., and Pe'er, D. (2010). An integrated approach to uncover drivers of cancer. Cell 143, 1005-1017.

Atzori, F., Tabernero, J., Cervantes, A., Prudkin, L., Andreu, J., Rodriguez-Braun, E., Domingo, A., Guijarro, J., Gamez, C., Rodon, J., et al. (2011). A phase I pharmacokinetic and pharmacodynamic study of dalotuzumab (MK-0646), an anti-insulin-like growth factor-1 receptor monoclonal antibody, in patients with advanced solid tumors. Clinical cancer research : an official journal of the American Association for Cancer Research 17, 6304-6312.

Awasthi, N., Zhang, C., Ruan, W., Schwarz, M.A., and Schwarz, R.E. (2012). BMS-754807, a small-molecule inhibitor of insulin-like growth factor-1 receptor/insulin receptor, enhances gemcitabine response in pancreatic cancer. Molecular cancer therapeutics 11, 2644-2653.

Aytes, A., Mitrofanova, A., Lefebvre, C., Alvarez, M.J., Castillo-Martin, M., Zheng, T., Eastham, J.A., Gopalan, A., Pienta, K.J., Shen, M.M., et al. (2014). Cross-species regulatory network analysis identifies a synergistic interaction between FOXM1 and CENPF that drives prostate cancer malignancy. Cancer cell 25, 638-651.

Bae, S.K., Bae, M.H., Ahn, M.Y., Son, M.J., Lee, Y.M., Bae, M.K., Lee, O.H., Park, B.C., and Kim, K.W. (1999). Egr-1 mediates transcriptional activation of IGF-II gene in response to hypoxia. Cancer research 59, 5989-5994.

Banerjee, A., Roach, M.C., Trcka, P., and Luduena, R.F. (1990). Increased microtubule assembly in bovine brain tubulin lacking the type III isotype of beta-tubulin. The Journal of biological chemistry 265, 1794-1799.

Barlow, D.P., Stoger, R., Herrmann, B.G., Saito, K., and Schweifer, N. (1991). The mouse insulin-like growth factor type-2 receptor is imprinted and closely linked to the Tme locus. Nature 349, 84-87.

Bartolomei, M.S., Zemel, S., and Tilghman, S.M. (1991). Parental imprinting of the mouse H19 gene. Nature 351, 153-155.

Bates, P., Fisher, R., Ward, A., Richardson, L., Hill, D.J., and Graham, C.F. (1995). Mammary cancer in transgenic mice expressing insulin-like growth factor II (IGF-II). British journal of cancer 72, 1189-1193.

Beer, T.M., Armstrong, A.J., Rathkopf, D.E., Loriot, Y., Sternberg, C.N., Higano, C.S., Iversen, P., Bhattacharya, S., Carles, J., Chowdhury, S., et al. (2014). Enzalutamide in Metastatic Prostate Cancer before Chemotherapy. The New England journal of medicine.

106

Beer, T.M., Pierce, W.C., Lowe, B.A., and Henner, W.D. (2001). Phase II study of weekly docetaxel in symptomatic androgen-independent prostate cancer. Annals of oncology : official journal of the European Society for Medical Oncology / ESMO 12, 1273-1279.

Belfiore, A., Frasca, F., Pandini, G., Sciacca, L., and Vigneri, R. (2009). Insulin receptor isoforms and insulin receptor/insulin-like growth factor receptor hybrids in physiology and disease. Endocrine reviews 30, 586-623.

Belfiore, A., and Malaguarnera, R. (2011). Insulin receptor and cancer. Endocrine-related cancer 18, R125-147.

Bergman, D., Halje, M., Nordin, M., and Engstrom, W. (2013). Insulin-like growth factor 2 in development and disease: a mini-review. Gerontology 59, 240-249.

Berry, W., Dakhil, S., Gregurich, M.A., and Asmar, L. (2001). Phase II trial of single-agent weekly docetaxel in hormone-refractory, symptomatic, metastatic carcinoma of the prostate. Seminars in oncology 28, 8-15.

Berry, W., Dakhil, S., Modiano, M., Gregurich, M., and Asmar, L. (2002). Phase III study of mitoxantrone plus low dose prednisone versus low dose prednisone alone in patients with asymptomatic hormone refractory prostate cancer. The Journal of urology 168, 2439-2443.

Berthold, D.R., Pond, G.R., Soban, F., de Wit, R., Eisenberger, M., and Tannock, I.F. (2008). Docetaxel plus prednisone or mitoxantrone plus prednisone for advanced prostate cancer: updated survival in the TAX 327 study. Journal of clinical oncology : official journal of the American Society of Clinical Oncology 26, 242-245.

Bill-Axelson, A., Holmberg, L., Ruutu, M., Garmo, H., Stark, J.R., Busch, C., Nordling, S., Haggman, M., Andersson, S.O., Bratell, S., et al. (2011). Radical prostatectomy versus watchful waiting in early prostate cancer. The New England journal of medicine 364, 1708-1717.

Bishr, M., and Saad, F. (2013). Overview of the latest treatments for castration-resistant prostate cancer. Nature reviews Urology 10, 522-528.

Blum, W.F., Ranke, M.B., and Bierich, J.R. (1988). A specific radioimmunoassay for insulin-like growth factor II: the interference of IGF binding proteins can be blocked by excess IGF-I. Acta endocrinologica 118, 374-380.

Bohl, C.E., Gao, W., Miller, D.D., Bell, C.E., and Dalton, J.T. (2005). Structural basis for antagonism and resistance of bicalutamide in prostate cancer. Proceedings of the National Academy of Sciences of the United States of America 102, 6201-6206.

Bohm, M., Locke, W.J., Sutherland, R.L., Kench, J.G., and Henshall, S.M. (2009). A role for GATA-2 in transition to an aggressive phenotype in prostate cancer through modulation of key androgen-regulated genes. Oncogene 28, 3847-3856.

Bolla, M., Collette, L., Blank, L., Warde, P., Dubois, J.B., Mirimanoff, R.O., Storme, G., Bernier, J., Kuten, A., Sternberg, C., et al. (2002). Long-term results with immediate androgen

107

suppression and external irradiation in patients with locally advanced prostate cancer (an EORTC study): a phase III randomised trial. Lancet 360, 103-106.

Bolla, M., van Poppel, H., Tombal, B., Vekemans, K., Da Pozzo, L., de Reijke, T.M., Verbaeys, A., Bosset, J.F., van Velthoven, R., Colombel, M., et al. (2012). Postoperative radiotherapy after radical prostatectomy for high-risk prostate cancer: long-term results of a randomised controlled trial (EORTC trial 22911). Lancet 380, 2018-2027.

Boorjian, S.A., Karnes, R.J., Viterbo, R., Rangel, L.J., Bergstralh, E.J., Horwitz, E.M., Blute, M.L., and Buyyounouski, M.K. (2011). Long-term survival after radical prostatectomy versus external-beam radiotherapy for patients with high-risk prostate cancer. Cancer 117, 2883-2891.

Brown, J., Jones, E.Y., and Forbes, B.E. (2009). Keeping IGF-II under control: lessons from the IGF-II-IGF2R crystal structure. Trends in biochemical sciences 34, 612-619.

Buck, E., Gokhale, P.C., Koujak, S., Brown, E., Eyzaguirre, A., Tao, N., Rosenfeld-Franklin, M., Lerner, L., Chiu, M.I., Wild, R., et al. (2010). Compensatory insulin receptor (IR) activation on inhibition of insulin-like growth factor-1 receptor (IGF-1R): rationale for cotargeting IGF-1R and IR in cancer. Molecular cancer therapeutics 9, 2652-2664.

Cai, Y., Balli, D., Ustiyan, V., Fulford, L., Hiller, A., Misetic, V., Zhang, Y., Paluch, A.M., Waltz, S.E., Kasper, S., et al. (2013). Foxm1 expression in prostate epithelial cells is essential for prostate carcinogenesis. The Journal of biological chemistry 288, 22527-22541.

Campbell, J.P., Merkel, A.R., Masood-Campbell, S.K., Elefteriou, F., and Sterling, J.A. (2012). Models of bone metastasis. Journal of visualized experiments : JoVE, e4260.

Cantor, A.B., Iwasaki, H., Arinobu, Y., Moran, T.B., Shigematsu, H., Sullivan, M.R., Akashi, K., and Orkin, S.H. (2008). Antagonism of FOG-1 and GATA factors in fate choice for the mast cell lineage. The Journal of experimental medicine 205, 611-624.

Carboni, J.M., Wittman, M., Yang, Z., Lee, F., Greer, A., Hurlburt, W., Hillerman, S., Cao, C., Cantor, G.H., Dell-John, J., et al. (2009). BMS-754807, a small molecule inhibitor of insulin-like growth factor-1R/IR. Molecular cancer therapeutics 8, 3341-3349.

Carver, B.S., Chapinski, C., Wongvipat, J., Hieronymus, H., Chen, Y., Chandarlapaty, S., Arora, V.K., Le, C., Koutcher, J., Scher, H., et al. (2011). Reciprocal feedback regulation of PI3K and androgen receptor signaling in PTEN-deficient prostate cancer. Cancer cell 19, 575-586.

Chabner, B.A., and Roberts, T.G., Jr. (2005). Timeline: Chemotherapy and the war on cancer. Nature reviews Cancer 5, 65-72.

Chang, A.J., Autio, K.A., Roach, M., 3rd, and Scher, H.I. (2014). High-risk prostate cancer-classification and therapy. Nature reviews Clinical oncology 11, 308-323.

Chang, A.N., Cantor, A.B., Fujiwara, Y., Lodish, M.B., Droho, S., Crispino, J.D., and Orkin, S.H. (2002). GATA-factor dependence of the multitype zinc-finger protein FOG-1 for its

108

essential role in megakaryopoiesis. Proceedings of the National Academy of Sciences of the United States of America 99, 9237-9242.

Chang, K.H., Li, R., Papari-Zareei, M., Watumull, L., Zhao, Y.D., Auchus, R.J., and Sharifi, N. (2011). Dihydrotestosterone synthesis bypasses testosterone to drive castration-resistant prostate cancer. Proceedings of the National Academy of Sciences of the United States of America 108, 13728-13733.

Chen, C.D., Welsbie, D.S., Tran, C., Baek, S.H., Chen, R., Vessella, R., Rosenfeld, M.G., and Sawyers, C.L. (2004). Molecular determinants of resistance to antiandrogen therapy. Nature medicine 10, 33-39.

Chen, Y., Chi, P., Rockowitz, S., Iaquinta, P.J., Shamu, T., Shukla, S., Gao, D., Sirota, I., Carver, B.S., Wongvipat, J., et al. (2013). ETS factors reprogram the androgen receptor cistrome and prime prostate tumorigenesis in response to PTEN loss. Nature medicine 19, 1023-1029.

Cheng, H., Snoek, R., Ghaidi, F., Cox, M.E., and Rennie, P.S. (2006). Short hairpin RNA knockdown of the androgen receptor attenuates ligand-independent activation and delays tumor progression. Cancer research 66, 10613-10620.

Chi, K.N., Hotte, S.J., Yu, E.Y., Tu, D., Eigl, B.J., Tannock, I., Saad, F., North, S., Powers, J., Gleave, M.E., et al. (2010a). Randomized phase II study of docetaxel and prednisone with or without OGX-011 in patients with metastatic castration-resistant prostate cancer. Journal of clinical oncology : official journal of the American Society of Clinical Oncology 28, 4247-4254.

Chi, K.N., Siu, L.L., Hirte, H., Hotte, S.J., Knox, J., Kollmansberger, C., Gleave, M., Guns, E., Powers, J., Walsh, W., et al. (2008). A phase I study of OGX-011, a 2'-methoxyethyl phosphorothioate antisense to clusterin, in combination with docetaxel in patients with advanced cancer. Clinical cancer research : an official journal of the American Association for Cancer Research 14, 833-839.

Chi, P., Chen, Y., Zhang, L., Guo, X., Wongvipat, J., Shamu, T., Fletcher, J.A., Dewell, S., Maki, R.G., Zheng, D., et al. (2010b). ETV1 is a lineage survival factor that cooperates with KIT in gastrointestinal stromal tumours. Nature 467, 849-853.

Chiang, Y.T., Wang, K., Fazli, L., Qi, R.Z., Gleave, M.E., Collins, C.C., Gout, P.W., and Wang, Y. (2014). GATA2 as a potential metastasis-driving gene in prostate cancer. Oncotarget 5, 451-461.

Chien, A.J., and Moasser, M.M. (2008). Cellular mechanisms of resistance to anthracyclines and taxanes in cancer: intrinsic and acquired. Seminars in oncology 35, S1-S14; quiz S39.

Christofori, G., Naik, P., and Hanahan, D. (1994). A second signal supplied by insulin-like growth factor II in oncogene-induced tumorigenesis. Nature 369, 414-418.

Chun, T.H., Itoh, H., Subramanian, L., Iniguez-Lluhi, J.A., and Nakao, K. (2003). Modification of GATA-2 transcriptional activity in endothelial cells by the SUMO E3 ligase PIASy. Circulation research 92, 1201-1208.

109

Church, D.N., Phillips, B.R., Stuckey, D.J., Barnes, D.J., Buffa, F.M., Manek, S., Clarke, K., Harris, A.L., Carter, E.J., and Hassan, A.B. (2012). Igf2 ligand dependency of Pten(+/-) developmental and tumour phenotypes in the mouse. Oncogene 31, 3635-3646.

Cisternino, S., Bourasset, F., Archimbaud, Y., Semiond, D., Sanderink, G., and Scherrmann, J.M. (2003). Nonlinear accumulation in the brain of the new taxoid TXD258 following saturation of P-glycoprotein at the blood-brain barrier in mice and rats. British journal of pharmacology 138, 1367-1375.

Cohn, A.L., Tabernero, J., Maurel, J., Nowara, E., Sastre, J., Chuah, B.Y., Kopp, M.V., Sakaeva, D.D., Mitchell, E.P., Dubey, S., et al. (2013). A randomized, placebo-controlled phase 2 study of ganitumab or conatumumab in combination with FOLFIRI for second-line treatment of mutant KRAS metastatic colorectal cancer. Annals of oncology : official journal of the European Society for Medical Oncology / ESMO 24, 1777-1785.

Coma, S., Allard-Ratick, M., Akino, T., van Meeteren, L.A., Mammoto, A., and Klagsbrun, M. (2013). GATA2 and Lmo2 control angiogenesis and lymphangiogenesis via direct transcriptional regulation of neuropilin-2. Angiogenesis 16, 939-952.

Constancia, M., Hemberger, M., Hughes, J., Dean, W., Ferguson-Smith, A., Fundele, R., Stewart, F., Kelsey, G., Fowden, A., Sibley, C., et al. (2002). Placental-specific IGF-II is a major modulator of placental and fetal growth. Nature 417, 945-948.

Cooper, C.R., Chay, C.H., and Pienta, K.J. (2002). The role of alpha(v)beta(3) in prostate cancer progression. Neoplasia 4, 191-194.

Cooper, W.N., Luharia, A., Evans, G.A., Raza, H., Haire, A.C., Grundy, R., Bowdin, S.C., Riccio, A., Sebastio, G., Bliek, J., et al. (2005). Molecular subtypes and phenotypic expression of Beckwith-Wiedemann syndrome. European journal of human genetics : EJHG 13, 1025-1032.

Cortes, J., Hochhaus, A., Hughes, T., and Kantarjian, H. (2011). Front-line and salvage therapies with tyrosine kinase inhibitors and other treatments in chronic myeloid leukemia. Journal of clinical oncology : official journal of the American Society of Clinical Oncology 29, 524-531.

Cui, H. (2007). Loss of imprinting of IGF2 as an epigenetic marker for the risk of human cancer. Disease markers 23, 105-112.

Da Costa, T.H., Williamson, D.H., Ward, A., Bates, P., Fisher, R., Richardson, L., Hill, D.J., Robinson, I.C., and Graham, C.F. (1994). High plasma insulin-like growth factor-II and low lipid content in transgenic mice: measurements of lipid metabolism. The Journal of endocrinology 143, 433-439.

Dart, A.E., and Wells, C.M. (2013). P21-activated kinase 4--not just one of the PAK. European journal of cell biology 92, 129-138.

Daughaday, W.H., Hall, K., Raben, M.S., Salmon, W.D., Jr., van den Brande, J.L., and van Wyk, J.J. (1972). Somatomedin: proposed designation for sulphation factor. Nature 235, 107.

110

de Bono, J.S., Logothetis, C.J., Molina, A., Fizazi, K., North, S., Chu, L., Chi, K.N., Jones, R.J., Goodman, O.B., Jr., Saad, F., et al. (2011). Abiraterone and increased survival in metastatic prostate cancer. The New England journal of medicine 364, 1995-2005.

de Bono, J.S., Oudard, S., Ozguroglu, M., Hansen, S., Machiels, J.P., Kocak, I., Gravis, G., Bodrogi, I., Mackenzie, M.J., Shen, L., et al. (2010). Prednisone plus cabazitaxel or mitoxantrone for metastatic castration-resistant prostate cancer progressing after docetaxel treatment: a randomised open-label trial. Lancet 376, 1147-1154.

de Bono, J.S., Piulats, J.M., Pandha, H.S., Petrylak, D.P., Saad, F., Aparicio, L.M., Sandhu, S.K., Fong, P., Gillessen, S., Hudes, G.R., et al. (2014). Phase II randomized study of figitumumab plus docetaxel and docetaxel alone with crossover for metastatic castration-resistant prostate cancer. Clinical cancer research : an official journal of the American Association for Cancer Research 20, 1925-1934.

de Pater, E., Kaimakis, P., Vink, C.S., Yokomizo, T., Yamada-Inagawa, T., van der Linden, R., Kartalaei, P.S., Camper, S.A., Speck, N., and Dzierzak, E. (2013). Gata2 is required for HSC generation and survival. The Journal of experimental medicine 210, 2843-2850.

DeChiara, T.M., Efstratiadis, A., and Robertson, E.J. (1990). A growth-deficiency phenotype in heterozygous mice carrying an insulin-like growth factor II gene disrupted by targeting. Nature 345, 78-80.

DeChiara, T.M., Robertson, E.J., and Efstratiadis, A. (1991). Parental imprinting of the mouse insulin-like growth factor II gene. Cell 64, 849-859.

Dehm, S.M., Schmidt, L.J., Heemers, H.V., Vessella, R.L., and Tindall, D.J. (2008). Splicing of a novel androgen receptor exon generates a constitutively active androgen receptor that mediates prostate cancer therapy resistance. Cancer research 68, 5469-5477.

Derry, W.B., Wilson, L., Khan, I.A., Luduena, R.F., and Jordan, M.A. (1997). Taxol differentially modulates the dynamics of microtubules assembled from unfractionated and purified beta-tubulin isotypes. Biochemistry 36, 3554-3562.

Diaz, L.E., Chuan, Y.C., Lewitt, M., Fernandez-Perez, L., Carrasco-Rodriguez, S., Sanchez-Gomez, M., and Flores-Morales, A. (2007). IGF-II regulates metastatic properties of choriocarcinoma cells through the activation of the insulin receptor. Molecular human reproduction 13, 567-576.

Dickins, R.A., Hemann, M.T., Zilfou, J.T., Simpson, D.R., Ibarra, I., Hannon, G.J., and Lowe, S.W. (2005). Probing tumor phenotypes using stable and regulated synthetic microRNA precursors. Nature genetics 37, 1289-1295.

Dickinson, R.E., Griffin, H., Bigley, V., Reynard, L.N., Hussain, R., Haniffa, M., Lakey, J.H., Rahman, T., Wang, X.N., McGovern, N., et al. (2011). Exome sequencing identifies GATA-2 mutation as the cause of dendritic cell, monocyte, B and NK lymphoid deficiency. Blood 118, 2656-2658.

111

Dickinson, R.E., Milne, P., Jardine, L., Zandi, S., Swierczek, S.I., McGovern, N., Cookson, S., Ferozepurwalla, Z., Langridge, A., Pagan, S., et al. (2014). The evolution of cellular deficiency in GATA2 mutation. Blood 123, 863-874.

Dinchuk, J.E., Cao, C., Huang, F., Reeves, K.A., Wang, J., Myers, F., Cantor, G.H., Zhou, X., Attar, R.M., Gottardis, M., et al. (2010). Insulin receptor (IR) pathway hyperactivity in IGF-IR null cells and suppression of downstream growth signaling using the dual IGF-IR/IR inhibitor, BMS-754807. Endocrinology 151, 4123-4132.

Domingo-Domenech, J., Oliva, C., Rovira, A., Codony-Servat, J., Bosch, M., Filella, X., Montagut, C., Tapia, M., Campas, C., Dang, L., et al. (2006). Interleukin 6, a nuclear factor-kappaB target, predicts resistance to docetaxel in hormone-independent prostate cancer and nuclear factor-kappaB inhibition by PS-1145 enhances docetaxel antitumor activity. Clinical cancer research : an official journal of the American Association for Cancer Research 12, 5578-5586.

Domingo-Domenech, J., Vidal, S.J., Rodriguez-Bravo, V., Castillo-Martin, M., Quinn, S.A., Rodriguez-Barrueco, R., Bonal, D.M., Charytonowicz, E., Gladoun, N., de la Iglesia-Vicente, J., et al. (2012). Suppression of acquired docetaxel resistance in prostate cancer through depletion of notch- and hedgehog-dependent tumor-initiating cells. Cancer cell 22, 373-388.

Draheim, K.M., Chen, H.B., Tao, Q., Moore, N., Roche, M., and Lyle, S. (2010). ARRDC3 suppresses breast cancer progression by negatively regulating integrin beta4. Oncogene 29, 5032-5047.

Drake, J.M., Gabriel, C.L., and Henry, M.D. (2005). Assessing tumor growth and distribution in a model of prostate cancer metastasis using bioluminescence imaging. Clinical & experimental metastasis 22, 674-684.

Drummond, I.A., Madden, S.L., Rohwer-Nutter, P., Bell, G.I., Sukhatme, V.P., and Rauscher, F.J., 3rd (1992). Repression of the insulin-like growth factor II gene by the Wilms tumor suppressor WT1. Science 257, 674-678.

Dumontet, C., and Sikic, B.I. (1999). Mechanisms of action of and resistance to antitubulin agents: microtubule dynamics, drug transport, and cell death. Journal of clinical oncology : official journal of the American Society of Clinical Oncology 17, 1061-1070.

Dydensborg, A.B., Rose, A.A., Wilson, B.J., Grote, D., Paquet, M., Giguere, V., Siegel, P.M., and Bouchard, M. (2009). GATA3 inhibits breast cancer growth and pulmonary breast cancer metastasis. Oncogene 28, 2634-2642.

Eggener, S.E., Scardino, P.T., Walsh, P.C., Han, M., Partin, A.W., Trock, B.J., Feng, Z., Wood, D.P., Eastham, J.A., Yossepowitch, O., et al. (2011). Predicting 15-year prostate cancer specific mortality after radical prostatectomy. The Journal of urology 185, 869-875.

Evans, T., and Felsenfeld, G. (1989). The erythroid-specific transcription factor Eryf1: a new finger protein. Cell 58, 877-885.

112

Fabbri, F., Carloni, S., Brigliadori, G., Zoli, W., Lapalombella, R., and Marini, M. (2006). Sequential events of apoptosis involving docetaxel, a microtubule-interfering agent: a cytometric study. BMC cell biology 7, 6.

Fadilah, S.A., Cheong, S.K., Roslan, H., Rozie-Hanisa, M., and Yen, G.K. (2002). GATA-1 and GATA-2 gene expression is related to the severity of dysplasia in myelodysplastic syndrome. Leukemia 16, 1563-1565.

Farber, S., and Diamond, L.K. (1948). Temporary remissions in acute leukemia in children produced by folic acid antagonist, 4-aminopteroyl-glutamic acid. The New England journal of medicine 238, 787-793.

Fellmann, C., Zuber, J., McJunkin, K., Chang, K., Malone, C.D., Dickins, R.A., Xu, Q., Hengartner, M.O., Elledge, S.J., Hannon, G.J., et al. (2011). Functional identification of optimized RNAi triggers using a massively parallel sensor assay. Molecular cell 41, 733-746.

Ferguson-Smith, A.C. (2011). Genomic imprinting: the emergence of an epigenetic paradigm. Nature reviews Genetics 12, 565-575.

Ferlay, J., Shin, H.R., Bray, F., Forman, D., Mathers, C., and Parkin, D.M. (2010). Estimates of worldwide burden of cancer in 2008: GLOBOCAN 2008. International journal of cancer Journal international du cancer 127, 2893-2917.

Fizazi, K., Yang, J., Peleg, S., Sikes, C.R., Kreimann, E.L., Daliani, D., Olive, M., Raymond, K.A., Janus, T.J., Logothetis, C.J., et al. (2003). Prostate cancer cells-osteoblast interaction shifts expression of growth/survival-related genes in prostate cancer and reduces expression of osteoprotegerin in osteoblasts. Clinical cancer research : an official journal of the American Association for Cancer Research 9, 2587-2597.

Flanigan, S.A., Pitts, T.M., Eckhardt, S.G., Tentler, J.J., Tan, A.C., Thorburn, A., and Leong, S. (2010). The insulin-like growth factor I receptor/insulin receptor tyrosine kinase inhibitor PQIP exhibits enhanced antitumor effects in combination with chemotherapy against colorectal cancer models. Clinical cancer research : an official journal of the American Association for Cancer Research 16, 5436-5446.

Flanigan, S.A., Pitts, T.M., Newton, T.P., Kulikowski, G.N., Tan, A.C., McManus, M.C., Spreafico, A., Kachaeva, M.I., Selby, H.M., Tentler, J.J., et al. (2013). Overcoming IGF1R/IR resistance through inhibition of MEK signaling in colorectal cancer models. Clinical cancer research : an official journal of the American Association for Cancer Research 19, 6219-6229.

Forbes, K., Westwood, M., Baker, P.N., and Aplin, J.D. (2008). Insulin-like growth factor I and II regulate the life cycle of trophoblast in the developing human placenta. American journal of physiology Cell physiology 294, C1313-1322.

Fox, E.M., Miller, T.W., Balko, J.M., Kuba, M.G., Sanchez, V., Smith, R.A., Liu, S., Gonzalez-Angulo, A.M., Mills, G.B., Ye, F., et al. (2011). A kinome-wide screen identifies the insulin/IGF-I receptor pathway as a mechanism of escape from hormone dependence in breast cancer. Cancer research 71, 6773-6784.

113

Frasca, F., Pandini, G., Scalia, P., Sciacca, L., Mineo, R., Costantino, A., Goldfine, I.D., Belfiore, A., and Vigneri, R. (1999). Insulin receptor isoform A, a newly recognized, high-affinity insulin-like growth factor II receptor in fetal and cancer cells. Molecular and cellular biology 19, 3278-3288.

Freedland, S.J., Humphreys, E.B., Mangold, L.A., Eisenberger, M., Dorey, F.J., Walsh, P.C., and Partin, A.W. (2005). Risk of prostate cancer-specific mortality following biochemical recurrence after radical prostatectomy. JAMA : the journal of the American Medical Association 294, 433-439.

Freedland, S.J., and Moul, J.W. (2007). Prostate specific antigen recurrence after definitive therapy. The Journal of urology 177, 1985-1991.

Friedbichler, K., Hofmann, M.H., Kroez, M., Ostermann, E., Lamche, H.R., Koessl, C., Borges, E., Pollak, M.N., Adolf, G., and Adam, P.J. (2014). Pharmacodynamic and antineoplastic activity of BI 836845, a fully human IGF ligand-neutralizing antibody, and mechanistic rationale for combination with rapamycin. Molecular cancer therapeutics 13, 399-409.

Froesch, E.R., Buergi, H., Ramseier, E.B., Bally, P., and Labhart, A. (1963). Antibody-Suppressible and Nonsuppressible Insulin-Like Activities in Human Serum and Their Physiologic Significance. An Insulin Assay with Adipose Tissue of Increased Precision and Specificity. The Journal of clinical investigation 42, 1816-1834.

Fu, V.X., Dobosy, J.R., Desotelle, J.A., Almassi, N., Ewald, J.A., Srinivasan, R., Berres, M., Svaren, J., Weindruch, R., and Jarrard, D.F. (2008). Aging and cancer-related loss of insulin-like growth factor 2 imprinting in the mouse and human prostate. Cancer research 68, 6797-6802.

Fujiwara, Y., Browne, C.P., Cunniff, K., Goff, S.C., and Orkin, S.H. (1996). Arrested development of embryonic red cell precursors in mouse embryos lacking transcription factor GATA-1. Proceedings of the National Academy of Sciences of the United States of America 93, 12355-12358.

Gao, H., Ouyang, X., Banach-Petrosky, W.A., Gerald, W.L., Shen, M.M., and Abate-Shen, C. (2006). Combinatorial activities of Akt and B-Raf/Erk signaling in a mouse model of androgen-independent prostate cancer. Proceedings of the National Academy of Sciences of the United States of America 103, 14477-14482.

Gao, J., Chesebrough, J.W., Cartlidge, S.A., Ricketts, S.A., Incognito, L., Veldman-Jones, M., Blakey, D.C., Tabrizi, M., Jallal, B., Trail, P.A., et al. (2011). Dual IGF-I/II-neutralizing antibody MEDI-573 potently inhibits IGF signaling and tumor growth. Cancer research 71, 1029-1040.

Gao, X., Johnson, K.D., Chang, Y.I., Boyer, M.E., Dewey, C.N., Zhang, J., and Bresnick, E.H. (2013). Gata2 cis-element is required for hematopoietic stem cell generation in the mammalian embryo. The Journal of experimental medicine 210, 2833-2842.

114

Gao, X., Pang, J., Li, L.Y., Liu, W.P., Di, J.M., Sun, Q.P., Fang, Y.Q., Liu, X.P., Pu, X.Y., He, D., et al. (2010). Expression profiling identifies new function of collapsin response mediator protein 4 as a metastasis-suppressor in prostate cancer. Oncogene 29, 4555-4566.

Garraway, L.A., Widlund, H.R., Rubin, M.A., Getz, G., Berger, A.J., Ramaswamy, S., Beroukhim, R., Milner, D.A., Granter, S.R., Du, J., et al. (2005). Integrative genomic analyses identify MITF as a lineage survival oncogene amplified in malignant melanoma. Nature 436, 117-122.

Gartrell, B.A., and Saad, F. (2014). Managing bone metastases and reducing skeletal related events in prostate cancer. Nature reviews Clinical oncology 11, 335-345.

Gaziel-Sovran, A., Segura, M.F., Di Micco, R., Collins, M.K., Hanniford, D., Vega-Saenz de Miera, E., Rakus, J.F., Dankert, J.F., Shang, S., Kerbel, R.S., et al. (2011). miR-30b/30d regulation of GalNAc transferases enhances invasion and immunosuppression during metastasis. Cancer cell 20, 104-118.

Gebeshuber, C.A., and Martinez, J. (2013). miR-100 suppresses IGF2 and inhibits breast tumorigenesis by interfering with proliferation and survival signaling. Oncogene 32, 3306-3310.

Giannakakou, P., Sackett, D.L., Kang, Y.K., Zhan, Z., Buters, J.T., Fojo, T., and Poruchynsky, M.S. (1997). Paclitaxel-resistant human ovarian cancer cells have mutant beta-tubulins that exhibit impaired paclitaxel-driven polymerization. The Journal of biological chemistry 272, 17118-17125.

Giannoukakis, N., Deal, C., Paquette, J., Goodyer, C.G., and Polychronakos, C. (1993). Parental genomic imprinting of the human IGF2 gene. Nature genetics 4, 98-101.

Giddings, S.J., and Carnaghi, L.R. (1992). Insulin receptor gene expression during development: developmental regulation of insulin receptor mRNA abundance in embryonic rat liver and yolk sac, developmental regulation of insulin receptor gene splicing, and comparison to abundance of insulin-like growth factor 1 receptor mRNA. Molecular endocrinology 6, 1665-1672.

Goldman, J.M., and Melo, J.V. (2003). Chronic myeloid leukemia--advances in biology and new approaches to treatment. The New England journal of medicine 349, 1451-1464.

Gomella, L.G. (2009). Effective testosterone suppression for prostate cancer: is there a best castration therapy? Reviews in urology 11, 52-60.

Gonzalez-Garay, M.L., Chang, L., Blade, K., Menick, D.R., and Cabral, F. (1999). A beta-tubulin leucine cluster involved in microtubule assembly and paclitaxel resistance. The Journal of biological chemistry 274, 23875-23882.

Gottesman, M.M., Fojo, T., and Bates, S.E. (2002). Multidrug resistance in cancer: role of ATP-dependent transporters. Nature reviews Cancer 2, 48-58.

Grass, J.A., Boyer, M.E., Pal, S., Wu, J., Weiss, M.J., and Bresnick, E.H. (2003). GATA-1-dependent transcriptional repression of GATA-2 via disruption of positive autoregulation and

115

domain-wide chromatin remodeling. Proceedings of the National Academy of Sciences of the United States of America 100, 8811-8816.

Grasso, C.S., Wu, Y.M., Robinson, D.R., Cao, X., Dhanasekaran, S.M., Khan, A.P., Quist, M.J., Jing, X., Lonigro, R.J., Brenner, J.C., et al. (2012). The mutational landscape of lethal castration-resistant prostate cancer. Nature 487, 239-243.

Gravis, G., Fizazi, K., Joly, F., Oudard, S., Priou, F., Esterni, B., Latorzeff, I., Delva, R., Krakowski, I., Laguerre, B., et al. (2013). Androgen-deprivation therapy alone or with docetaxel in non-castrate metastatic prostate cancer (GETUG-AFU 15): a randomised, open-label, phase 3 trial. The lancet oncology 14, 149-158.

Grote, D., Boualia, S.K., Souabni, A., Merkel, C., Chi, X., Costantini, F., Carroll, T., and Bouchard, M. (2008). Gata3 acts downstream of beta-catenin signaling to prevent ectopic metanephric kidney induction. PLoS genetics 4, e1000316.

Group, P.C.T.C. (2000). Maximum androgen blockade in advanced prostate cancer: an overview of the randomised trials. Prostate Cancer Trialists' Collaborative Group. Lancet 355, 1491-1498.

Group, T.V.A.C.-o.U.R. (1967). Treatment and survival of patients with cancer of the prostate. The Veterans Administration Co-operative Urological Research Group. Surgery, gynecology & obstetrics 124, 1011-1017.

Guo, Z., Yang, X., Sun, F., Jiang, R., Linn, D.E., Chen, H., Chen, H., Kong, X., Melamed, J., Tepper, C.G., et al. (2009). A novel androgen receptor splice variant is up-regulated during prostate cancer progression and promotes androgen depletion-resistant growth. Cancer research 69, 2305-2313.

Hahn, C.N., Chong, C.E., Carmichael, C.L., Wilkins, E.J., Brautigan, P.J., Li, X.C., Babic, M., Lin, M., Carmagnac, A., Lee, Y.K., et al. (2011). Heritable GATA2 mutations associated with familial myelodysplastic syndrome and acute myeloid leukemia. Nature genetics 43, 1012-1017.

Hahn, H., Wojnowski, L., Specht, K., Kappler, R., Calzada-Wack, J., Potter, D., Zimmer, A., Muller, U., Samson, E., Quintanilla-Martinez, L., et al. (2000). Patched target Igf2 is indispensable for the formation of medulloblastoma and rhabdomyosarcoma. The Journal of biological chemistry 275, 28341-28344.

Haldar, S., Chintapalli, J., and Croce, C.M. (1996). Taxol induces bcl-2 phosphorylation and death of prostate cancer cells. Cancer research 56, 1253-1255.

Haluska, P., Menefee, M., Plimack, E.R., Rosenberg, J., Northfelt, D., LaVallee, T., Shi, L., Yu, X.Q., Burke, P., Huang, J., et al. (2014). Phase I Dose-Escalation Study of MEDI-573, a Bispecific, Antiligand Monoclonal Antibody against IGFI and IGFII, in Patients with Advanced Solid Tumors. Clinical cancer research : an official journal of the American Association for Cancer Research 20, 4747-4757.

Haluska, P., Shaw, H.M., Batzel, G.N., Yin, D., Molina, J.R., Molife, L.R., Yap, T.A., Roberts, M.L., Sharma, A., Gualberto, A., et al. (2007). Phase I dose escalation study of the anti insulin-

116

like growth factor-I receptor monoclonal antibody CP-751,871 in patients with refractory solid tumors. Clinical cancer research : an official journal of the American Association for Cancer Research 13, 5834-5840.

Hamilton, G.S., Lysiak, J.J., Han, V.K., and Lala, P.K. (1998). Autocrine-paracrine regulation of human trophoblast invasiveness by insulin-like growth factor (IGF)-II and IGF-binding protein (IGFBP)-1. Experimental cell research 244, 147-156.

Han, V.K., Bassett, N., Walton, J., and Challis, J.R. (1996). The expression of insulin-like growth factor (IGF) and IGF-binding protein (IGFBP) genes in the human placenta and membranes: evidence for IGF-IGFBP interactions at the feto-maternal interface. The Journal of clinical endocrinology and metabolism 81, 2680-2693.

Han, V.K., D'Ercole, A.J., and Lund, P.K. (1987). Cellular localization of somatomedin (insulin-like growth factor) messenger RNA in the human fetus. Science 236, 193-197.

Harris, L.K., and Westwood, M. (2012). Biology and significance of signalling pathways activated by IGF-II. Growth factors 30, 1-12.

Hayakawa, F., Towatari, M., Ozawa, Y., Tomita, A., Privalsky, M.L., and Saito, H. (2004). Functional regulation of GATA-2 by acetylation. Journal of leukocyte biology 75, 529-540.

Heilmann, A.M., Perera, R.M., Ecker, V., Nicolay, B.N., Bardeesy, N., Benes, C.H., and Dyson, N.J. (2014). CDK4/6 and IGF1 receptor inhibitors synergize to suppress the growth of p16INK4A-deficient pancreatic cancers. Cancer research 74, 3947-3958.

Hellerstedt, B.A., and Pienta, K.J. (2002). The current state of hormonal therapy for prostate cancer. CA: a cancer journal for clinicians 52, 154-179.

Holohan, C., Van Schaeybroeck, S., Longley, D.B., and Johnston, P.G. (2013). Cancer drug resistance: an evolving paradigm. Nature reviews Cancer 13, 714-726.

Holthuizen, P., Van Dijk, M.A., Rodenburg, R.J., Koonen-Reemst, A.M., and Sussenbach, J.S. (1993). Transcriptional regulation of the major promoters of the human IGF-II gene. Molecular reproduction and development 35, 391-393.

Hou, X., Huang, F., Macedo, L.F., Harrington, S.C., Reeves, K.A., Greer, A., Finckenstein, F.G., Brodie, A., Gottardis, M.M., Carboni, J.M., et al. (2011). Dual IGF-1R/InsR inhibitor BMS-754807 synergizes with hormonal agents in treatment of estrogen-dependent breast cancer. Cancer research 71, 7597-7607.

Hsu, A.P., Sampaio, E.P., Khan, J., Calvo, K.R., Lemieux, J.E., Patel, S.Y., Frucht, D.M., Vinh, D.C., Auth, R.D., Freeman, A.F., et al. (2011). Mutations in GATA2 are associated with the autosomal dominant and sporadic monocytopenia and mycobacterial infection (MonoMAC) syndrome. Blood 118, 2653-2655.

Hu, R., Dunn, T.A., Wei, S., Isharwal, S., Veltri, R.W., Humphreys, E., Han, M., Partin, A.W., Vessella, R.L., Isaacs, W.B., et al. (2009). Ligand-independent androgen receptor variants

117

derived from splicing of cryptic exons signify hormone-refractory prostate cancer. Cancer research 69, 16-22.

Huang, Z., Dore, L.C., Li, Z., Orkin, S.H., Feng, G., Lin, S., and Crispino, J.D. (2009). GATA-2 reinforces megakaryocyte development in the absence of GATA-1. Molecular and cellular biology 29, 5168-5180.

Huggins, C., Stevens, R.E., Jr, and Hodges, C.V. (1941). Studies on prostatic cancer: Ii. the effects of castration on advanced carcinoma of the prostate gland. Archives of Surgery 43, 209-223.

Ikonomi, P., Noguchi, C.T., Miller, W., Kassahun, H., Hardison, R., and Schechter, A.N. (2000). Levels of GATA-1/GATA-2 transcription factors modulate expression of embryonic and fetal hemoglobins. Gene 261, 277-287.

Irving, J.A., and Lala, P.K. (1995). Functional role of cell surface integrins on human trophoblast cell migration: regulation by TGF-beta, IGF-II, and IGFBP-1. Experimental cell research 217, 419-427.

Jarrard, D.F., Bussemakers, M.J., Bova, G.S., and Isaacs, W.B. (1995). Regional loss of imprinting of the insulin-like growth factor II gene occurs in human prostate tissues. Clinical cancer research : an official journal of the American Association for Cancer Research 1, 1471-1478.

Jemal, A., Bray, F., Center, M.M., Ferlay, J., Ward, E., and Forman, D. (2011). Global cancer statistics. CA: a cancer journal for clinicians 61, 69-90.

Jennbacken, K., Tesan, T., Wang, W., Gustavsson, H., Damber, J.E., and Welen, K. (2010). N-cadherin increases after androgen deprivation and is associated with metastasis in prostate cancer. Endocrine-related cancer 17, 469-479.

Jiao, J., Wang, S., Qiao, R., Vivanco, I., Watson, P.A., Sawyers, C.L., and Wu, H. (2007). Murine cell lines derived from Pten null prostate cancer show the critical role of PTEN in hormone refractory prostate cancer development. Cancer research 67, 6083-6091.

Jin, J.K., Dayyani, F., and Gallick, G.E. (2011). Steps in prostate cancer progression that lead to bone metastasis. International journal of cancer Journal international du cancer 128, 2545-2561.

Johannessen, C.M., Johnson, L.A., Piccioni, F., Townes, A., Frederick, D.T., Donahue, M.K., Narayan, R., Flaherty, K.T., Wargo, J.A., Root, D.E., et al. (2013). A melanocyte lineage program confers resistance to MAP kinase pathway inhibition. Nature 504, 138-142.

Johansson, J.E., Andren, O., Andersson, S.O., Dickman, P.W., Holmberg, L., Magnuson, A., and Adami, H.O. (2004). Natural history of early, localized prostate cancer. JAMA : the journal of the American Medical Association 291, 2713-2719.

Johnson, K.D., Hsu, A.P., Ryu, M.J., Wang, J., Gao, X., Boyer, M.E., Liu, Y., Lee, Y., Calvo, K.R., Keles, S., et al. (2012). Cis-element mutated in GATA2-dependent immunodeficiency

118

governs hematopoiesis and vascular integrity. The Journal of clinical investigation 122, 3692-3704.

Jones, R.L., Kim, E.S., Nava-Parada, P., Alam, S., Johnson, F.M., Stephens, A.W., Simantov, R., Poondru, S., Gedrich, R., Lippman, S.M., et al. (2014). Phase I Study of Intermittent Oral Dosing of the Insulin-like Growth Factor-1 and Insulin Receptors Inhibitor OSI-906 in Patients With Advanced Solid Tumors. Clinical cancer research : an official journal of the American Association for Cancer Research.

Kalin, T.V., Wang, I.C., Ackerson, T.J., Major, M.L., Detrisac, C.J., Kalinichenko, V.V., Lyubimov, A., and Costa, R.H. (2006). Increased levels of the FoxM1 transcription factor accelerate development and progression of prostate carcinomas in both TRAMP and LADY transgenic mice. Cancer research 66, 1712-1720.

Kantoff, P.W., Halabi, S., Conaway, M., Picus, J., Kirshner, J., Hars, V., Trump, D., Winer, E.P., and Vogelzang, N.J. (1999). Hydrocortisone with or without mitoxantrone in men with hormone-refractory prostate cancer: results of the cancer and leukemia group B 9182 study. Journal of clinical oncology : official journal of the American Society of Clinical Oncology 17, 2506-2513.

Kantoff, P.W., Higano, C.S., Shore, N.D., Berger, E.R., Small, E.J., Penson, D.F., Redfern, C.H., Ferrari, A.C., Dreicer, R., Sims, R.B., et al. (2010). Sipuleucel-T immunotherapy for castration-resistant prostate cancer. The New England journal of medicine 363, 411-422.

Karl, P.I. (1995). Insulin-like growth factor-1 stimulates amino acid uptake by the cultured human placental trophoblast. Journal of cellular physiology 165, 83-88.

Kastner, P., and Chan, S. (2008). PU.1: a crucial and versatile player in hematopoiesis and leukemia. The international journal of biochemistry & cell biology 40, 22-27.

Kazenwadel, J., Secker, G.A., Liu, Y.J., Rosenfeld, J.A., Wildin, R.S., Cuellar-Rodriguez, J., Hsu, A.P., Dyack, S., Fernandez, C.V., Chong, C.E., et al. (2012). Loss-of-function germline GATA2 mutations in patients with MDS/AML or MonoMAC syndrome and primary lymphedema reveal a key role for GATA2 in the lymphatic vasculature. Blood 119, 1283-1291.

Kim, K.W., Bae, S.K., Lee, O.H., Bae, M.H., Lee, M.J., and Park, B.C. (1998). Insulin-like growth factor II induced by hypoxia may contribute to angiogenesis of human hepatocellular carcinoma. Cancer research 58, 348-351.

Klotz, L., and Emberton, M. (2014). Management of low risk prostate cancer-active surveillance and focal therapy. Nature reviews Clinical oncology 11, 324-334.

Kniss, D.A., Shubert, P.J., Zimmerman, P.D., Landon, M.B., and Gabbe, S.G. (1994). Insulinlike growth factors. Their regulation of glucose and amino acid transport in placental trophoblasts isolated from first-trimester chorionic villi. The Journal of reproductive medicine 39, 249-256.

Kobayashi-Osaki, M., Ohneda, O., Suzuki, N., Minegishi, N., Yokomizo, T., Takahashi, S., Lim, K.C., Engel, J.D., and Yamamoto, M. (2005). GATA motifs regulate early hematopoietic lineage-specific expression of the Gata2 gene. Molecular and cellular biology 25, 7005-7020.

119

Koga, S., Yamaguchi, N., Abe, T., Minegishi, M., Tsuchiya, S., Yamamoto, M., and Minegishi, N. (2007). Cell-cycle-dependent oscillation of GATA2 expression in hematopoietic cells. Blood 109, 4200-4208.

Koo, C.Y., Muir, K.W., and Lam, E.W. (2012). FOXM1: From cancer initiation to progression and treatment. Biochimica et biophysica acta 1819, 28-37.

Kooijman, R., van Buul-Offers, S.C., Scholtens, L.E., Schuurman, H.J., Van den Brande, L.J., and Zegers, B.J. (1995). T cell development in insulin-like growth factor-II transgenic mice. Journal of immunology 154, 5736-5745.

Kouros-Mehr, H., Bechis, S.K., Slorach, E.M., Littlepage, L.E., Egeblad, M., Ewald, A.J., Pai, S.Y., Ho, I.C., and Werb, Z. (2008). GATA-3 links tumor differentiation and dissemination in a luminal breast cancer model. Cancer cell 13, 141-152.

Kouros-Mehr, H., Slorach, E.M., Sternlicht, M.D., and Werb, Z. (2006). GATA-3 maintains the differentiation of the luminal cell fate in the mammary gland. Cell 127, 1041-1055.

Kuhn, D.J., Berkova, Z., Jones, R.J., Woessner, R., Bjorklund, C.C., Ma, W., Davis, R.E., Lin, P., Wang, H., Madden, T.L., et al. (2012). Targeting the insulin-like growth factor-1 receptor to overcome bortezomib resistance in preclinical models of multiple myeloma. Blood 120, 3260-3270.

Kumar, M.S., Hancock, D.C., Molina-Arcas, M., Steckel, M., East, P., Diefenbacher, M., Armenteros-Monterroso, E., Lassailly, F., Matthews, N., Nye, E., et al. (2012). The GATA2 transcriptional network is requisite for RAS oncogene-driven non-small cell lung cancer. Cell 149, 642-655.

Kyle, A.H., Huxham, L.A., Yeoman, D.M., and Minchinton, A.I. (2007). Limited tissue penetration of taxanes: a mechanism for resistance in solid tumors. Clinical cancer research : an official journal of the American Association for Cancer Research 13, 2804-2810.

Langer, C.J., Novello, S., Park, K., Krzakowski, M., Karp, D.D., Mok, T., Benner, R.J., Scranton, J.R., Olszanski, A.J., and Jassem, J. (2014). Randomized, phase III trial of first-line figitumumab in combination with paclitaxel and carboplatin versus paclitaxel and carboplatin alone in patients with advanced non-small-cell lung cancer. Journal of clinical oncology : official journal of the American Society of Clinical Oncology 32, 2059-2066.

Lau, M.M., Stewart, C.E., Liu, Z., Bhatt, H., Rotwein, P., and Stewart, C.L. (1994). Loss of the imprinted IGF2/cation-independent mannose 6-phosphate receptor results in fetal overgrowth and perinatal lethality. Genes & development 8, 2953-2963.

Lee, Y.I., Lee, S., Lee, Y., Bong, Y.S., Hyun, S.W., Yoo, Y.D., Kim, S.J., Kim, Y.W., and Poo, H.R. (1998). The human hepatitis B virus transactivator X gene product regulates Sp1 mediated transcription of an insulin-like growth factor II promoter 4. Oncogene 16, 2367-2380.

Lehr, J.E., and Pienta, K.J. (1998). Preferential adhesion of prostate cancer cells to a human bone marrow endothelial cell line. Journal of the National Cancer Institute 90, 118-123.

120

Li, B., Tsao, S.W., Chan, K.W., Ludwig, D.L., Novosyadlyy, R., Li, Y.Y., He, Q.Y., and Cheung, A.L. (2014). Id1-induced IGF-II and its autocrine/endocrine promotion of esophageal cancer progression and chemoresistance--implications for IGF-II and IGF-IR-targeted therapy. Clinical cancer research : an official journal of the American Association for Cancer Research 20, 2651-2662.

Li, P., Cavallero, S., Gu, Y., Chen, T.H., Hughes, J., Hassan, A.B., Bruning, J.C., Pashmforoush, M., and Sucov, H.M. (2011). IGF signaling directs ventricular cardiomyocyte proliferation during embryonic heart development. Development 138, 1795-1805.

Liao, S., Howell, D.K., and Chang, T.M. (1974). Action of a nonsteroidal antiandrogen, flutamide, on the receptor binding and nuclear retention of 5 alpha-dihydrotestosterone in rat ventral prostate. Endocrinology 94, 1205-1209.

Lim, K.C., Hosoya, T., Brandt, W., Ku, C.J., Hosoya-Ohmura, S., Camper, S.A., Yamamoto, M., and Engel, J.D. (2012). Conditional Gata2 inactivation results in HSC loss and lymphatic mispatterning. The Journal of clinical investigation 122, 3705-3717.

Ling, K.W., Ottersbach, K., van Hamburg, J.P., Oziemlak, A., Tsai, F.Y., Orkin, S.H., Ploemacher, R., Hendriks, R.W., and Dzierzak, E. (2004). GATA-2 plays two functionally distinct roles during the ontogeny of hematopoietic stem cells. The Journal of experimental medicine 200, 871-882.

Litzenburger, B.C., Creighton, C.J., Tsimelzon, A., Chan, B.T., Hilsenbeck, S.G., Wang, T., Carboni, J.M., Gottardis, M.M., Huang, F., Chang, J.C., et al. (2011). High IGF-IR activity in triple-negative breast cancer cell lines and tumorgrafts correlates with sensitivity to anti-IGF-IR therapy. Clinical cancer research : an official journal of the American Association for Cancer Research 17, 2314-2327.

Liu, R., Wang, X., Chen, G.Y., Dalerba, P., Gurney, A., Hoey, T., Sherlock, G., Lewicki, J., Shedden, K., and Clarke, M.F. (2007). The prognostic role of a gene signature from tumorigenic breast-cancer cells. The New England journal of medicine 356, 217-226.

Liu, Y., Gong, Z., Sun, L., and Li, X. (2014). FOXM1 and androgen receptor co-regulate CDC6 gene transcription and DNA replication in prostate cancer cells. Biochimica et biophysica acta 1839, 297-305.

Livingstone, C. (2013). IGF2 and cancer. Endocrine-related cancer 20, R321-339.

Locke, J.A., Guns, E.S., Lubik, A.A., Adomat, H.H., Hendy, S.C., Wood, C.A., Ettinger, S.L., Gleave, M.E., and Nelson, C.C. (2008). Androgen levels increase by intratumoral de novo steroidogenesis during progression of castration-resistant prostate cancer. Cancer research 68, 6407-6415.

Lubik, A.A., Gunter, J.H., Hollier, B.G., Ettinger, S., Fazli, L., Stylianou, N., Hendy, S.C., Adomat, H.H., Gleave, M.E., Pollak, M., et al. (2013). IGF2 increases de novo steroidogenesis in prostate cancer cells. Endocrine-related cancer 20, 173-186.

121

Luesink, M., Hollink, I.H., van der Velden, V.H., Knops, R.H., Boezeman, J.B., de Haas, V., Trka, J., Baruchel, A., Reinhardt, D., van der Reijden, B.A., et al. (2012). High GATA2 expression is a poor prognostic marker in pediatric acute myeloid leukemia. Blood 120, 2064-2075.

Lui, J.C., and Baron, J. (2013). Evidence that Igf2 down-regulation in postnatal tissues and up-regulation in malignancies is driven by transcription factor E2f3. Proceedings of the National Academy of Sciences of the United States of America 110, 6181-6186.

Manzo, F., Tambaro, F.P., Mai, A., and Altucci, L. (2009). Histone acetyltransferase inhibitors and preclinical studies. Expert opinion on therapeutic patents 19, 761-774.

May, G., Soneji, S., Tipping, A.J., Teles, J., McGowan, S.J., Wu, M., Guo, Y., Fugazza, C., Brown, J., Karlsson, G., et al. (2013). Dynamic analysis of gene expression and genome-wide transcription factor binding during lineage specification of multipotent progenitors. Cell stem cell 13, 754-768.

McDonnell, T.J., Troncoso, P., Brisbay, S.M., Logothetis, C., Chung, L.W., Hsieh, J.T., Tu, S.M., and Campbell, M.L. (1992). Expression of the protooncogene bcl-2 in the prostate and its association with emergence of androgen-independent prostate cancer. Cancer research 52, 6940-6944.

McGrath, K.E., Koniski, A.D., Malik, J., and Palis, J. (2003). Circulation is established in a stepwise pattern in the mammalian embryo. Blood 101, 1669-1676.

Mehra, R., Varambally, S., Ding, L., Shen, R., Sabel, M.S., Ghosh, D., Chinnaiyan, A.M., and Kleer, C.G. (2005). Identification of GATA3 as a breast cancer prognostic marker by global gene expression meta-analysis. Cancer research 65, 11259-11264.

Menghini, R., Marchetti, V., Cardellini, M., Hribal, M.L., Mauriello, A., Lauro, D., Sbraccia, P., Lauro, R., and Federici, M. (2005). Phosphorylation of GATA2 by Akt increases adipose tissue differentiation and reduces adipose tissue-related inflammation: a novel pathway linking obesity to atherosclerosis. Circulation 111, 1946-1953.

Merika, M., and Orkin, S.H. (1993). DNA-binding specificity of GATA family transcription factors. Molecular and cellular biology 13, 3999-4010.

Migliaccio, A.R., and Bieker, J.J. (2011). GATA2 finds its macrophage niche. Blood 118, 2647-2649.

Mimeault, M., Johansson, S.L., Vankatraman, G., Moore, E., Henichart, J.P., Depreux, P., Lin, M.F., and Batra, S.K. (2007). Combined targeting of epidermal growth factor receptor and hedgehog signaling by gefitinib and cyclopamine cooperatively improves the cytotoxic effects of docetaxel on metastatic prostate cancer cells. Molecular cancer therapeutics 6, 967-978.

Minegishi, N., Suzuki, N., Kawatani, Y., Shimizu, R., and Yamamoto, M. (2005). Rapid turnover of GATA-2 via ubiquitin-proteasome protein degradation pathway. Genes to cells : devoted to molecular & cellular mechanisms 10, 693-704.

122

Molkentin, J.D. (2000). The zinc finger-containing transcription factors GATA-4, -5, and -6. Ubiquitously expressed regulators of tissue-specific gene expression. The Journal of biological chemistry 275, 38949-38952.

Moller, N., and Jorgensen, J.O. (2009). Effects of growth hormone on glucose, lipid, and protein metabolism in human subjects. Endocrine reviews 30, 152-177.

Montgomery, R.B., Mostaghel, E.A., Vessella, R., Hess, D.L., Kalhorn, T.F., Higano, C.S., True, L.D., and Nelson, P.S. (2008). Maintenance of intratumoral androgens in metastatic prostate cancer: a mechanism for castration-resistant tumor growth. Cancer research 68, 4447-4454.

Moorehead, R.A., Sanchez, O.H., Baldwin, R.M., and Khokha, R. (2003). Transgenic overexpression of IGF-II induces spontaneous lung tumors: a model for human lung adenocarcinoma. Oncogene 22, 853-857.

Moulton, T., Crenshaw, T., Hao, Y., Moosikasuwan, J., Lin, N., Dembitzer, F., Hensle, T., Weiss, L., McMorrow, L., Loew, T., et al. (1994). Epigenetic lesions at the H19 locus in Wilms' tumour patients. Nature genetics 7, 440-447.

Mulholland, D.J., Tran, L.M., Li, Y., Cai, H., Morim, A., Wang, S., Plaisier, S., Garraway, I.P., Huang, J., Graeber, T.G., et al. (2011). Cell autonomous role of PTEN in regulating castration-resistant prostate cancer growth. Cancer cell 19, 792-804.

Mulvihill, M.J., Cooke, A., Rosenfeld-Franklin, M., Buck, E., Foreman, K., Landfair, D., O'Connor, M., Pirritt, C., Sun, Y., Yao, Y., et al. (2009). Discovery of OSI-906: a selective and orally efficacious dual inhibitor of the IGF-1 receptor and insulin receptor. Future medicinal chemistry 1, 1153-1171.

Murphy, L.J., Bell, G.I., and Friesen, H.G. (1987). Tissue distribution of insulin-like growth factor I and II messenger ribonucleic acid in the adult rat. Endocrinology 120, 1279-1282.

Murray, B.W., Guo, C., Piraino, J., Westwick, J.K., Zhang, C., Lamerdin, J., Dagostino, E., Knighton, D., Loi, C.M., Zager, M., et al. (2010). Small-molecule p21-activated kinase inhibitor PF-3758309 is a potent inhibitor of oncogenic signaling and tumor growth. Proceedings of the National Academy of Sciences of the United States of America 107, 9446-9451.

Murray, S., Briasoulis, E., Linardou, H., Bafaloukos, D., and Papadimitriou, C. (2012). Taxane resistance in breast cancer: mechanisms, predictive biomarkers and circumvention strategies. Cancer treatment reviews 38, 890-903.

Nordin, M., Bergman, D., Halje, M., Engstrom, W., and Ward, A. (2014). Epigenetic regulation of the Igf2/H19 gene cluster. Cell proliferation 47, 189-199.

O'Dell, S.D., and Day, I.N. (1998). Insulin-like growth factor II (IGF-II). The international journal of biochemistry & cell biology 30, 767-771.

123

Ogawa, O., Eccles, M.R., Szeto, J., McNoe, L.A., Yun, K., Maw, M.A., Smith, P.J., and Reeve, A.E. (1993). Relaxation of insulin-like growth factor II gene imprinting implicated in Wilms' tumour. Nature 362, 749-751.

Ogawa, T., Ogawa, K., Shiga, K., Furukawa, T., Nagase, H., Hashimoto, S., Kobayashi, T., and Horii, A. (2010). Upregulation of IGF2 is associated with an acquired resistance for cis-diamminedichloroplatinum in human head and neck squamous cell carcinoma. European archives of oto-rhino-laryngology : official journal of the European Federation of Oto-Rhino-Laryngological Societies 267, 1599-1606.

Ohlsson, R., Nystrom, A., Pfeifer-Ohlsson, S., Tohonen, V., Hedborg, F., Schofield, P., Flam, F., and Ekstrom, T.J. (1993). IGF2 is parentally imprinted during human embryogenesis and in the Beckwith-Wiedemann syndrome. Nature genetics 4, 94-97.

Orlic, D., Anderson, S., Biesecker, L.G., Sorrentino, B.P., and Bodine, D.M. (1995). Pluripotent hematopoietic stem cells contain high levels of mRNA for c-kit, GATA-2, p45 NF-E2, and c-myb and low levels or no mRNA for c-fms and the receptors for granulocyte colony-stimulating factor and interleukins 5 and 7. Proceedings of the National Academy of Sciences of the United States of America 92, 4601-4605.

Ostergaard, P., Simpson, M.A., Connell, F.C., Steward, C.G., Brice, G., Woollard, W.J., Dafou, D., Kilo, T., Smithson, S., Lunt, P., et al. (2011). Mutations in GATA2 cause primary lymphedema associated with a predisposition to acute myeloid leukemia (Emberger syndrome). Nature genetics 43, 929-931.

Ovcharenko, I., Loots, G.G., Giardine, B.M., Hou, M., Ma, J., Hardison, R.C., Stubbs, L., and Miller, W. (2005). Mulan: multiple-sequence local alignment and visualization for studying function and evolution. Genome research 15, 184-194.

Ozawa, Y., Towatari, M., Tsuzuki, S., Hayakawa, F., Maeda, T., Miyata, Y., Tanimoto, M., and Saito, H. (2001). Histone deacetylase 3 associates with and represses the transcription factor GATA-2. Blood 98, 2116-2123.

Panda, D., Miller, H.P., Banerjee, A., Luduena, R.F., and Wilson, L. (1994). Microtubule dynamics in vitro are regulated by the tubulin isotype composition. Proceedings of the National Academy of Sciences of the United States of America 91, 11358-11362.

Pandini, G., Frasca, F., Mineo, R., Sciacca, L., Vigneri, R., and Belfiore, A. (2002). Insulin/insulin-like growth factor I hybrid receptors have different biological characteristics depending on the insulin receptor isoform involved. The Journal of biological chemistry 277, 39684-39695.

Pandolfi, P.P., Roth, M.E., Karis, A., Leonard, M.W., Dzierzak, E., Grosveld, F.G., Engel, J.D., and Lindenbaum, M.H. (1995). Targeted disruption of the GATA3 gene causes severe abnormalities in the nervous system and in fetal liver haematopoiesis. Nature genetics 11, 40-44.

124

Park, M.H., Lee, H.S., Lee, C.S., You, S.T., Kim, D.J., Park, B.H., Kang, M.J., Heo, W.D., Shin, E.Y., Schwartz, M.A., et al. (2013). p21-Activated kinase 4 promotes prostate cancer progression through CREB. Oncogene 32, 2475-2482.

Parker, C., Nilsson, S., Heinrich, D., Helle, S.I., O'Sullivan, J.M., Fossa, S.D., Chodacki, A., Wiechno, P., Logue, J., Seke, M., et al. (2013). Alpha emitter radium-223 and survival in metastatic prostate cancer. The New England journal of medicine 369, 213-223.

Pchejetski, D., Doumerc, N., Golzio, M., Naymark, M., Teissie, J., Kohama, T., Waxman, J., Malavaud, B., and Cuvillier, O. (2008). Chemosensitizing effects of sphingosine kinase-1 inhibition in prostate cancer cell and animal models. Molecular cancer therapeutics 7, 1836-1845.

Perez-Stable, C.M., Pozas, A., and Roos, B.A. (2000). A role for GATA transcription factors in the androgen regulation of the prostate-specific antigen gene enhancer. Molecular and cellular endocrinology 167, 43-53.

Peters, J. (2014). The role of genomic imprinting in biology and disease: an expanding view. Nature reviews Genetics 15, 517-530.

Petrylak, D.P., Tangen, C.M., Hussain, M.H., Lara, P.N., Jr., Jones, J.A., Taplin, M.E., Burch, P.A., Berry, D., Moinpour, C., Kohli, M., et al. (2004). Docetaxel and estramustine compared with mitoxantrone and prednisone for advanced refractory prostate cancer. The New England journal of medicine 351, 1513-1520.

Pilepich, M.V., Winter, K., Lawton, C.A., Krisch, R.E., Wolkov, H.B., Movsas, B., Hug, E.B., Asbell, S.O., and Grignon, D. (2005). Androgen suppression adjuvant to definitive radiotherapy in prostate carcinoma--long-term results of phase III RTOG 85-31. International journal of radiation oncology, biology, physics 61, 1285-1290.

Pivot, X., Koralewski, P., Hidalgo, J.L., Chan, A., Goncalves, A., Schwartsmann, G., Assadourian, S., and Lotz, J.P. (2008). A multicenter phase II study of XRP6258 administered as a 1-h i.v. infusion every 3 weeks in taxane-resistant metastatic breast cancer patients. Annals of oncology : official journal of the European Society for Medical Oncology / ESMO 19, 1547-1552.

Ploussard, G., Terry, S., Maille, P., Allory, Y., Sirab, N., Kheuang, L., Soyeux, P., Nicolaiew, N., Coppolani, E., Paule, B., et al. (2010). Class III beta-tubulin expression predicts prostate tumor aggressiveness and patient response to docetaxel-based chemotherapy. Cancer research 70, 9253-9264.

Pollak, M. (2008). Insulin and insulin-like growth factor signalling in neoplasia. Nature reviews Cancer 8, 915-928.

Pollak, M. (2012). The insulin and insulin-like growth factor receptor family in neoplasia: an update. Nature reviews Cancer 12, 159-169.

125

Pommier, Y., Sordet, O., Antony, S., Hayward, R.L., and Kohn, K.W. (2004). Apoptosis defects and chemotherapy resistance: molecular interaction maps and networks. Oncogene 23, 2934-2949.

Pound, C.R., Partin, A.W., Eisenberger, M.A., Chan, D.W., Pearson, J.D., and Walsh, P.C. (1999). Natural history of progression after PSA elevation following radical prostatectomy. JAMA : the journal of the American Medical Association 281, 1591-1597.

Pravtcheva, D.D., and Wise, T.L. (1998). Metastasizing mammary carcinomas in H19 enhancers-Igf2 transgenic mice. The Journal of experimental zoology 281, 43-57.

Puzanov, I., Lindsay, C.R., Goff, L.W., Sosman, J.A., Gilbert, J., Berlin, J., Poondru, S., Simantov, R., Gedrich, R., Stephens, A., et al. (2014). A Phase I Study of Continuous Oral Dosing of OSI-906, a Dual Inhibitor of Insulin-Like Growth Factor-1 and Insulin Receptors in Patients with Advanced Solid Tumors. Clinical cancer research : an official journal of the American Association for Cancer Research.

Qian, D.Z., Rademacher, B.L., Pittsenbarger, J., Huang, C.Y., Myrthue, A., Higano, C.S., Garzotto, M., Nelson, P.S., and Beer, T.M. (2010). CCL2 is induced by chemotherapy and protects prostate cancer cells from docetaxel-induced cytotoxicity. The Prostate 70, 433-442.

Radu, M., Semenova, G., Kosoff, R., and Chernoff, J. (2014). PAK signalling during the development and progression of cancer. Nature reviews Cancer 14, 13-25.

Rainier, S., Johnson, L.A., Dobry, C.J., Ping, A.J., Grundy, P.E., and Feinberg, A.P. (1993). Relaxation of imprinted genes in human cancer. Nature 362, 747-749.

Ramalingam, S.S., Spigel, D.R., Chen, D., Steins, M.B., Engelman, J.A., Schneider, C.P., Novello, S., Eberhardt, W.E., Crino, L., Habben, K., et al. (2011). Randomized phase II study of erlotinib in combination with placebo or R1507, a monoclonal antibody to insulin-like growth factor-1 receptor, for advanced-stage non-small-cell lung cancer. Journal of clinical oncology : official journal of the American Society of Clinical Oncology 29, 4574-4580.

Ravenel, J.D., Broman, K.W., Perlman, E.J., Niemitz, E.L., Jayawardena, T.M., Bell, D.W., Haber, D.A., Uejima, H., and Feinberg, A.P. (2001). Loss of imprinting of insulin-like growth factor-II (IGF2) gene in distinguishing specific biologic subtypes of Wilms tumor. Journal of the National Cancer Institute 93, 1698-1703.

Rinderknecht, E., and Humbel, R.E. (1976a). Amino-terminal sequences of two polypeptides from human serum with nonsuppressible insulin-like and cell-growth-promoting activities: evidence for structural homology with insulin B chain. Proceedings of the National Academy of Sciences of the United States of America 73, 4379-4381.

Rinderknecht, E., and Humbel, R.E. (1976b). Polypeptides with nonsuppressible insulin-like and cell-growth promoting activities in human serum: isolation, chemical characterization, and some biological properties of forms I and II. Proceedings of the National Academy of Sciences of the United States of America 73, 2365-2369.

126

Ritchie, C.K., Andrews, L.R., Thomas, K.G., Tindall, D.J., and Fitzpatrick, L.A. (1997). The effects of growth factors associated with osteoblasts on prostate carcinoma proliferation and chemotaxis: implications for the development of metastatic disease. Endocrinology 138, 1145-1150.

Robert-Moreno, A., Espinosa, L., de la Pompa, J.L., and Bigas, A. (2005). RBPjkappa-dependent Notch function regulates Gata2 and is essential for the formation of intra-embryonic hematopoietic cells. Development 132, 1117-1126.

Robertson, J.F., Ferrero, J.M., Bourgeois, H., Kennecke, H., de Boer, R.H., Jacot, W., McGreivy, J., Suzuki, S., Zhu, M., McCaffery, I., et al. (2013). Ganitumab with either exemestane or fulvestrant for postmenopausal women with advanced, hormone-receptor-positive breast cancer: a randomised, controlled, double-blind, phase 2 trial. The lancet oncology 14, 228-235.

Rodrigues, N.P., Janzen, V., Forkert, R., Dombkowski, D.M., Boyd, A.S., Orkin, S.H., Enver, T., Vyas, P., and Scadden, D.T. (2005). Haploinsufficiency of GATA-2 perturbs adult hematopoietic stem-cell homeostasis. Blood 106, 477-484.

Rogler, C.E., Yang, D., Rossetti, L., Donohoe, J., Alt, E., Chang, C.J., Rosenfeld, R., Neely, K., and Hintz, R. (1994). Altered body composition and increased frequency of diverse malignancies in insulin-like growth factor-II transgenic mice. The Journal of biological chemistry 269, 13779-13784.

Rosa-Ribeiro, R., Nishan, U., Vidal, R.O., Barbosa, G.O., Reis, L.O., Cesar, C.L., and Carvalho, H.F. (2014). Transcription factors involved in prostate gland adaptation to androgen deprivation. PloS one 9, e97080.

Ross, H.M., Kryvenko, O.N., Cowan, J.E., Simko, J.P., Wheeler, T.M., and Epstein, J.I. (2012). Do adenocarcinomas of the prostate with Gleason score (GS) </=6 have the potential to metastasize to lymph nodes? The American journal of surgical pathology 36, 1346-1352.

Rossetti, L., Barzilai, N., Chen, W., Harris, T., Yang, D., and Rogler, C.E. (1996). Hepatic overexpression of insulin-like growth factor-II in adulthood increases basal and insulin-stimulated glucose disposal in conscious mice. The Journal of biological chemistry 271, 203-208.

Roth, B.J., Yeap, B.Y., Wilding, G., Kasimis, B., McLeod, D., and Loehrer, P.J. (1993). Taxol in advanced, hormone-refractory carcinoma of the prostate. A phase II trial of the Eastern Cooperative Oncology Group. Cancer 72, 2457-2460.

Ryan, C.J., Smith, M.R., de Bono, J.S., Molina, A., Logothetis, C.J., de Souza, P., Fizazi, K., Mainwaring, P., Piulats, J.M., Ng, S., et al. (2013). Abiraterone in metastatic prostate cancer without previous chemotherapy. The New England journal of medicine 368, 138-148.

Saad, F., Hotte, S., North, S., Eigl, B., Chi, K., Czaykowski, P., Wood, L., Pollak, M., Berry, S., Lattouf, J.B., et al. (2011). Randomized phase II trial of Custirsen (OGX-011) in combination with docetaxel or mitoxantrone as second-line therapy in patients with metastatic castrate-

127

resistant prostate cancer progressing after first-line docetaxel: CUOG trial P-06c. Clinical cancer research : an official journal of the American Association for Cancer Research 17, 5765-5773.

Sallman, D.A., Chen, X., Zhong, B., Gilvary, D.L., Zhou, J., Wei, S., and Djeu, J.Y. (2007). Clusterin mediates TRAIL resistance in prostate tumor cells. Molecular cancer therapeutics 6, 2938-2947.

Salmon, W.D., Jr., and Daughaday, W.H. (1957). A hormonally controlled serum factor which stimulates sulfate incorporation by cartilage in vitro. The Journal of laboratory and clinical medicine 49, 825-836.

Schally, A.V., Comaru-Schally, A.M., Plonowski, A., Nagy, A., Halmos, G., and Rekasi, Z. (2000). Peptide analogs in the therapy of prostate cancer. The Prostate 45, 158-166.

Scher, H.I., Fizazi, K., Saad, F., Taplin, M.E., Sternberg, C.N., Miller, K., de Wit, R., Mulders, P., Chi, K.N., Shore, N.D., et al. (2012). Increased survival with enzalutamide in prostate cancer after chemotherapy. The New England journal of medicine 367, 1187-1197.

Schutz, F.A., Buzaid, A.C., and Sartor, O. (2014). Taxanes in the management of metastatic castration-resistant prostate cancer: Efficacy and management of toxicity. Critical reviews in oncology/hematology.

Sciacca, L., Mineo, R., Pandini, G., Murabito, A., Vigneri, R., and Belfiore, A. (2002). In IGF-I receptor-deficient leiomyosarcoma cells autocrine IGF-II induces cell invasion and protection from apoptosis via the insulin receptor isoform A. Oncogene 21, 8240-8250.

Seidenfeld, J., Samson, D.J., Hasselblad, V., Aronson, N., Albertsen, P.C., Bennett, C.L., and Wilt, T.J. (2000). Single-therapy androgen suppression in men with advanced prostate cancer: a systematic review and meta-analysis. Annals of internal medicine 132, 566-577.

Seruga, B., Ocana, A., and Tannock, I.F. (2011). Drug resistance in metastatic castration-resistant prostate cancer. Nature reviews Clinical oncology 8, 12-23.

Sferruzzi-Perri, A.N., Owens, J.A., Pringle, K.G., Robinson, J.S., and Roberts, C.T. (2006). Maternal insulin-like growth factors-I and -II act via different pathways to promote fetal growth. Endocrinology 147, 3344-3355.

Shan, S.W., Fang, L., Shatseva, T., Rutnam, Z.J., Yang, X., Du, W., Lu, W.Y., Xuan, J.W., Deng, Z., and Yang, B.B. (2013). Mature miR-17-5p and passenger miR-17-3p induce hepatocellular carcinoma by targeting PTEN, GalNT7 and vimentin in different signal pathways. Journal of cell science 126, 1517-1530.

Sharma, N.L., Massie, C.E., Ramos-Montoya, A., Zecchini, V., Scott, H.E., Lamb, A.D., MacArthur, S., Stark, R., Warren, A.Y., Mills, I.G., et al. (2013). The androgen receptor induces a distinct transcriptional program in castration-resistant prostate cancer in man. Cancer cell 23, 35-47.

128

Shiau, A.K., Barstad, D., Loria, P.M., Cheng, L., Kushner, P.J., Agard, D.A., and Greene, G.L. (1998). The structural basis of estrogen receptor/coactivator recognition and the antagonism of this interaction by tamoxifen. Cell 95, 927-937.

Shin, S.I., Freedman, V.H., Risser, R., and Pollack, R. (1975). Tumorigenicity of virus-transformed cells in nude mice is correlated specifically with anchorage independent growth in vitro. Proceedings of the National Academy of Sciences of the United States of America 72, 4435-4439.

Shivdasani, R.A., Fujiwara, Y., McDevitt, M.A., and Orkin, S.H. (1997). A lineage-selective knockout establishes the critical role of transcription factor GATA-1 in megakaryocyte growth and platelet development. The EMBO journal 16, 3965-3973.

Siegel, R., Ma, J., Zou, Z., and Jemal, A. (2014). Cancer statistics, 2014. CA: a cancer journal for clinicians 64, 9-29.

Snoek, R., Cheng, H., Margiotti, K., Wafa, L.A., Wong, C.A., Wong, E.C., Fazli, L., Nelson, C.C., Gleave, M.E., and Rennie, P.S. (2009). In vivo knockdown of the androgen receptor results in growth inhibition and regression of well-established, castration-resistant prostate tumors. Clinical cancer research : an official journal of the American Association for Cancer Research 15, 39-47.

Soares, M.B., Ishii, D.N., and Efstratiadis, A. (1985). Developmental and tissue-specific expression of a family of transcripts related to rat insulin-like growth factor II mRNA. Nucleic acids research 13, 1119-1134.

Spinner, M.A., Sanchez, L.A., Hsu, A.P., Shaw, P.A., Zerbe, C.S., Calvo, K.R., Arthur, D.C., Gu, W., Gould, C.M., Brewer, C.C., et al. (2014). GATA2 deficiency: a protean disorder of hematopoiesis, lymphatics, and immunity. Blood 123, 809-821.

Springate, C.M., Jackson, J.K., Gleave, M.E., and Burt, H.M. (2005). Efficacy of an intratumoral controlled release formulation of clusterin antisense oligonucleotide complexed with chitosan containing paclitaxel or docetaxel in prostate cancer xenograft models. Cancer chemotherapy and pharmacology 56, 239-247.

Sramkoski, R.M., Pretlow, T.G., 2nd, Giaconia, J.M., Pretlow, T.P., Schwartz, S., Sy, M.S., Marengo, S.R., Rhim, J.S., Zhang, D., and Jacobberger, J.W. (1999). A new human prostate carcinoma cell line, 22Rv1. In vitro cellular & developmental biology Animal 35, 403-409.

Stanbrough, M., Bubley, G.J., Ross, K., Golub, T.R., Rubin, M.A., Penning, T.M., Febbo, P.G., and Balk, S.P. (2006). Increased expression of genes converting adrenal androgens to testosterone in androgen-independent prostate cancer. Cancer research 66, 2815-2825.

Steenman, M.J., Rainier, S., Dobry, C.J., Grundy, P., Horon, I.L., and Feinberg, A.P. (1994). Loss of imprinting of IGF2 is linked to reduced expression and abnormal methylation of H19 in Wilms' tumour. Nature genetics 7, 433-439.

129

Steingrimsson, E., Copeland, N.G., and Jenkins, N.A. (2004). Melanocytes and the microphthalmia transcription factor network. Annual review of genetics 38, 365-411.

Stylianopoulou, F., Efstratiadis, A., Herbert, J., and Pintar, J. (1988). Pattern of the insulin-like growth factor II gene expression during rat embryogenesis. Development 103, 497-506.

Sudo, T., Nitta, M., Saya, H., and Ueno, N.T. (2004). Dependence of paclitaxel sensitivity on a functional spindle assembly checkpoint. Cancer research 64, 2502-2508.

Sullivan, K.F., and Cleveland, D.W. (1986). Identification of conserved isotype-defining variable region sequences for four vertebrate beta tubulin polypeptide classes. Proceedings of the National Academy of Sciences of the United States of America 83, 4327-4331.

Sullivan, M.J., Taniguchi, T., Jhee, A., Kerr, N., and Reeve, A.E. (1999). Relaxation of IGF2 imprinting in Wilms tumours associated with specific changes in IGF2 methylation. Oncogene 18, 7527-7534.

Sun, S., Sprenger, C.C., Vessella, R.L., Haugk, K., Soriano, K., Mostaghel, E.A., Page, S.T., Coleman, I.M., Nguyen, H.M., Sun, H., et al. (2010). Castration resistance in human prostate cancer is conferred by a frequently occurring androgen receptor splice variant. The Journal of clinical investigation 120, 2715-2730.

Sun, Y., Wang, B.E., Leong, K.G., Yue, P., Li, L., Jhunjhunwala, S., Chen, D., Seo, K., Modrusan, Z., Gao, W.Q., et al. (2012). Androgen deprivation causes epithelial-mesenchymal transition in the prostate: implications for androgen-deprivation therapy. Cancer research 72, 527-536.

Sussenbach, J.S., Steenbergh, P.H., and Holthuizen, P. (1992). Structure and expression of the human insulin-like growth factor genes. Growth regulation 2, 1-9.

Suzuki, N., Ohneda, O., Minegishi, N., Nishikawa, M., Ohta, T., Takahashi, S., Engel, J.D., and Yamamoto, M. (2006). Combinatorial Gata2 and Sca1 expression defines hematopoietic stem cells in the bone marrow niche. Proceedings of the National Academy of Sciences of the United States of America 103, 2202-2207.

Sweeney, C.C., Y.H.; Carducci, M.A.; Liu, G.; Jarrard, D.F.; Eisenberger, M.A.; Wong, Y.N.; Hahn, N.M.; Kohli, M.; Vogelzang, N.J.; Cooney, M.M.; Dreicer, R.; Picus, J.; Shevrin, D.H.; Hussain, M.; Garcia, J.A.; DiPaola R.S. (2014). Impact on overall survival (OS) with chemohormonal therapy versus hormonal therapy for hormone-sensitive newly metastatic prostate cancer (mPrCa): An ECOG-led phase III randomized trial. (2014 ASCO Annual Meeting).

Tada, Y., Yamaguchi, Y., Kinjo, T., Song, X., Akagi, T., Takamura, H., Ohta, T., Yokota, T., and Koide, H. (2014). The stem cell transcription factor ZFP57 induces IGF2 expression to promote anchorage-independent growth in cancer cells. Oncogene.

130

Taichman, R.S., Cooper, C., Keller, E.T., Pienta, K.J., Taichman, N.S., and McCauley, L.K. (2002). Use of the stromal cell-derived factor-1/CXCR4 pathway in prostate cancer metastasis to bone. Cancer research 62, 1832-1837.

Tamaki, A., Ierano, C., Szakacs, G., Robey, R.W., and Bates, S.E. (2011). The controversial role of ABC transporters in clinical oncology. Essays in biochemistry 50, 209-232.

Tanaka, H., Kono, E., Tran, C.P., Miyazaki, H., Yamashiro, J., Shimomura, T., Fazli, L., Wada, R., Huang, J., Vessella, R.L., et al. (2010). Monoclonal antibody targeting of N-cadherin inhibits prostate cancer growth, metastasis and castration resistance. Nature medicine 16, 1414-1420.

Taniguchi, T., Sullivan, M.J., Ogawa, O., and Reeve, A.E. (1995). Epigenetic changes encompassing the IGF2/H19 locus associated with relaxation of IGF2 imprinting and silencing of H19 in Wilms tumor. Proceedings of the National Academy of Sciences of the United States of America 92, 2159-2163.

Tannock, I.F., de Wit, R., Berry, W.R., Horti, J., Pluzanska, A., Chi, K.N., Oudard, S., Theodore, C., James, N.D., Turesson, I., et al. (2004). Docetaxel plus prednisone or mitoxantrone plus prednisone for advanced prostate cancer. The New England journal of medicine 351, 1502-1512.

Tannock, I.F., Osoba, D., Stockler, M.R., Ernst, D.S., Neville, A.J., Moore, M.J., Armitage, G.R., Wilson, J.J., Venner, P.M., Coppin, C.M., et al. (1996). Chemotherapy with mitoxantrone plus prednisone or prednisone alone for symptomatic hormone-resistant prostate cancer: a Canadian randomized trial with palliative end points. Journal of clinical oncology : official journal of the American Society of Clinical Oncology 14, 1756-1764.

Taylor, B.S., Schultz, N., Hieronymus, H., Gopalan, A., Xiao, Y., Carver, B.S., Arora, V.K., Kaushik, P., Cerami, E., Reva, B., et al. (2010). Integrative genomic profiling of human prostate cancer. Cancer cell 18, 11-22.

Terui, K., Takahashi, Y., Kitazawa, J., Toki, T., Yokoyama, M., and Ito, E. (2000). Expression of transcription factors during megakaryocytic differentiation of CD34+ cells from human cord blood induced by thrombopoietin. The Tohoku journal of experimental medicine 192, 259-273.

Tessema, M., Yingling, C.M., Snider, A.M., Do, K., Juri, D.E., Picchi, M.A., Zhang, X., Liu, Y., Leng, S., Tellez, C.S., et al. (2014). GATA2 is epigenetically repressed in human and mouse lung tumors and is not requisite for survival of KRAS mutant lung cancer. Journal of thoracic oncology : official publication of the International Association for the Study of Lung Cancer 9, 784-793.

Thompson, I., Thrasher, J.B., Aus, G., Burnett, A.L., Canby-Hagino, E.D., Cookson, M.S., D'Amico, A.V., Dmochowski, R.R., Eton, D.T., Forman, J.D., et al. (2007). Guideline for the management of clinically localized prostate cancer: 2007 update. The Journal of urology 177, 2106-2131.

Tognon, C.E., Somasiri, A.M., Evdokimova, V.E., Trigo, G., Uy, E.E., Melnyk, N., Carboni, J.M., Gottardis, M.M., Roskelley, C.D., Pollak, M., et al. (2011). ETV6-NTRK3-mediated breast

131

epithelial cell transformation is blocked by targeting the IGF1R signaling pathway. Cancer research 71, 1060-1070.

Tolcher, A.W., Sarantopoulos, J., Patnaik, A., Papadopoulos, K., Lin, C.C., Rodon, J., Murphy, B., Roth, B., McCaffery, I., Gorski, K.S., et al. (2009). Phase I, pharmacokinetic, and pharmacodynamic study of AMG 479, a fully human monoclonal antibody to insulin-like growth factor receptor 1. Journal of clinical oncology : official journal of the American Society of Clinical Oncology 27, 5800-5807.

Trainor, C.D., Ghirlando, R., and Simpson, M.A. (2000). GATA zinc finger interactions modulate DNA binding and transactivation. The Journal of biological chemistry 275, 28157-28166.

Tran, C., Ouk, S., Clegg, N.J., Chen, Y., Watson, P.A., Arora, V., Wongvipat, J., Smith-Jones, P.M., Yoo, D., Kwon, A., et al. (2009). Development of a second-generation antiandrogen for treatment of advanced prostate cancer. Science 324, 787-790.

Trapnell, C., Pachter, L., and Salzberg, S.L. (2009). TopHat: discovering splice junctions with RNA-Seq. Bioinformatics 25, 1105-1111.

Trapnell, C., Williams, B.A., Pertea, G., Mortazavi, A., Kwan, G., van Baren, M.J., Salzberg, S.L., Wold, B.J., and Pachter, L. (2010). Transcript assembly and quantification by RNA-Seq reveals unannotated transcripts and isoform switching during cell differentiation. Nature biotechnology 28, 511-515.

Tsai, F.Y., Keller, G., Kuo, F.C., Weiss, M., Chen, J., Rosenblatt, M., Alt, F.W., and Orkin, S.H. (1994). An early haematopoietic defect in mice lacking the transcription factor GATA-2. Nature 371, 221-226.

Tsai, F.Y., and Orkin, S.H. (1997). Transcription factor GATA-2 is required for proliferation/survival of early hematopoietic cells and mast cell formation, but not for erythroid and myeloid terminal differentiation. Blood 89, 3636-3643.

Tsai, S.F., Martin, D.I., Zon, L.I., D'Andrea, A.D., Wong, G.G., and Orkin, S.H. (1989). Cloning of cDNA for the major DNA-binding protein of the erythroid lineage through expression in mammalian cells. Nature 339, 446-451.

Tusher, V.G., Tibshirani, R., and Chu, G. (2001). Significance analysis of microarrays applied to the ionizing radiation response. Proceedings of the National Academy of Sciences of the United States of America 98, 5116-5121.

Udan, R.S., Culver, J.C., and Dickinson, M.E. (2013). Understanding vascular development. Wiley interdisciplinary reviews Developmental biology 2, 327-346.

Ueno, H., and Weissman, I.L. (2010). The origin and fate of yolk sac hematopoiesis: application of chimera analyses to developmental studies. The International journal of developmental biology 54, 1019-1031.

132

Underwood, L.E., Hintz, R.L., Voina, S.J., and Van Wyk, J.J. (1972). Human somatomedin, the growth hormone dependent sulfation factor, is antilipolytic. The Journal of clinical endocrinology and metabolism 35, 194-198.

van Bokhoven, A., Varella-Garcia, M., Korch, C., Johannes, W.U., Smith, E.E., Miller, H.L., Nordeen, S.K., Miller, G.J., and Lucia, M.S. (2003). Molecular characterization of human prostate carcinoma cell lines. The Prostate 57, 205-225.

van Buul-Offers, S.C., de Haan, K., Reijnen-Gresnigt, M.G., Meinsma, D., Jansen, M., Oei, S.L., Bonte, E.J., Sussenbach, J.S., and Van den Brande, J.L. (1995). Overexpression of human insulin-like growth factor-II in transgenic mice causes increased growth of the thymus. The Journal of endocrinology 144, 491-502.

van Dijk, M.A., Rodenburg, R.J., Holthuizen, P., and Sussenbach, J.S. (1992). The liver-specific promoter of the human insulin-like growth factor II gene is activated by CCAAT/enhancer binding protein (C/EBP). Nucleic acids research 20, 3099-3104.

Vasan, N., Boyer, J.L., and Herbst, R.S. (2014). A RAS Renaissance: Emerging Targeted Therapies for KRAS-Mutated Non-Small Cell Lung Cancer. Clinical cancer research : an official journal of the American Association for Cancer Research.

Verrills, N.M., Po'uha, S.T., Liu, M.L., Liaw, T.Y., Larsen, M.R., Ivery, M.T., Marshall, G.M., Gunning, P.W., and Kavallaris, M. (2006). Alterations in gamma-actin and tubulin-targeted drug resistance in childhood leukemia. Journal of the National Cancer Institute 98, 1363-1374.

Vert, J.P., Foveau, N., Lajaunie, C., and Vandenbrouck, Y. (2006). An accurate and interpretable model for siRNA efficacy prediction. BMC bioinformatics 7, 520.

Vicente, C., Conchillo, A., Garcia-Sanchez, M.A., and Odero, M.D. (2012a). The role of the GATA2 transcription factor in normal and malignant hematopoiesis. Critical reviews in oncology/hematology 82, 1-17.

Vicente, C., Vazquez, I., Conchillo, A., Garcia-Sanchez, M.A., Marcotegui, N., Fuster, O., Gonzalez, M., Calasanz, M.J., Lahortiga, I., and Odero, M.D. (2012b). Overexpression of GATA2 predicts an adverse prognosis for patients with acute myeloid leukemia and it is associated with distinct molecular abnormalities. Leukemia 26, 550-554.

Visakorpi, T., Hyytinen, E., Koivisto, P., Tanner, M., Keinanen, R., Palmberg, C., Palotie, A., Tammela, T., Isola, J., and Kallioniemi, O.P. (1995). In vivo amplification of the androgen receptor gene and progression of human prostate cancer. Nature genetics 9, 401-406.

Wang, Q., Li, W., Liu, X.S., Carroll, J.S., Janne, O.A., Keeton, E.K., Chinnaiyan, A.M., Pienta, K.J., and Brown, M. (2007). A hierarchical network of transcription factors governs androgen receptor-dependent prostate cancer growth. Molecular cell 27, 380-392.

Wang, Q., Li, W., Zhang, Y., Yuan, X., Xu, K., Yu, J., Chen, Z., Beroukhim, R., Wang, H., Lupien, M., et al. (2009). Androgen receptor regulates a distinct transcription program in androgen-independent prostate cancer. Cell 138, 245-256.

133

Wang, S., Wang, Z., Boise, L., Dent, P., and Grant, S. (1999). Loss of the bcl-2 phosphorylation loop domain increases resistance of human leukemia cells (U937) to paclitaxel-mediated mitochondrial dysfunction and apoptosis. Biochemical and biophysical research communications 259, 67-72.

Wang, S.S., Esplin, E.D., Li, J.L., Huang, L., Gazdar, A., Minna, J., and Evans, G.A. (1998). Alterations of the PPP2R1B gene in human lung and colon cancer. Science 282, 284-287.

Wang, Z.Q., Fung, M.R., Barlow, D.P., and Wagner, E.F. (1994). Regulation of embryonic growth and lysosomal targeting by the imprinted Igf2/Mpr gene. Nature 372, 464-467.

Ward, A., Bates, P., Fisher, R., Richardson, L., and Graham, C.F. (1994). Disproportionate growth in mice with Igf-2 transgenes. Proceedings of the National Academy of Sciences of the United States of America 91, 10365-10369.

Wells, C.M., Whale, A.D., Parsons, M., Masters, J.R., and Jones, G.E. (2010). PAK4: a pluripotent kinase that regulates prostate cancer cell adhesion. Journal of cell science 123, 1663-1673.

Weroha, S.J., and Haluska, P. (2012). The insulin-like growth factor system in cancer. Endocrinology and metabolism clinics of North America 41, 335-350, vi.

Whale, A.D., Dart, A., Holt, M., Jones, G.E., and Wells, C.M. (2013). PAK4 kinase activity and somatic mutation promote carcinoma cell motility and influence inhibitor sensitivity. Oncogene 32, 2114-2120.

Wittman, M.D., Carboni, J.M., Yang, Z., Lee, F.Y., Antman, M., Attar, R., Balimane, P., Chang, C., Chen, C., Discenza, L., et al. (2009). Discovery of a 2,4-disubstituted pyrrolo[1,2-f][1,2,4]triazine inhibitor (BMS-754807) of insulin-like growth factor receptor (IGF-1R) kinase in clinical development. Journal of medicinal chemistry 52, 7360-7363.

Wolf, E., Hoeflich, A., and Lahm, H. (1998). What is the function of IGF-II in postnatal life? Answers from transgenic mouse models. Growth hormone & IGF research : official journal of the Growth Hormone Research Society and the International IGF Research Society 8, 185-193.

Wolf, E., Kramer, R., Blum, W.F., Foll, J., and Brem, G. (1994). Consequences of postnatally elevated insulin-like growth factor-II in transgenic mice: endocrine changes and effects on body and organ growth. Endocrinology 135, 1877-1886.

Wong, Y.N., Ferraldeschi, R., Attard, G., and de Bono, J. (2014). Evolution of androgen receptor targeted therapy for advanced prostate cancer. Nature reviews Clinical oncology 11, 365-376.

Wu, D., Sunkel, B., Chen, Z., Liu, X., Ye, Z., Li, Q., Grenade, C., Ke, J., Zhang, C., Chen, H., et al. (2014). Three-tiered role of the pioneer factor GATA2 in promoting androgen-dependent gene expression in prostate cancer. Nucleic acids research 42, 3607-3622.

134

Wu, L., Birle, D.C., and Tannock, I.F. (2005). Effects of the mammalian target of rapamycin inhibitor CCI-779 used alone or with chemotherapy on human prostate cancer cells and xenografts. Cancer research 65, 2825-2831.

Yagoda, A., and Petrylak, D. (1993). Cytotoxic chemotherapy for advanced hormone-resistant prostate cancer. Cancer 71, 1098-1109.

Yao, N., Yao, D., Wang, L., Dong, Z., Wu, W., Qiu, L., Yan, X., Yu, D., Chen, J., Sai, W., et al. (2012). Inhibition of autocrine IGF-II on effect of human HepG2 cell proliferation and angiogenesis factor expression. Tumour biology : the journal of the International Society for Oncodevelopmental Biology and Medicine 33, 1767-1776.

Yee, D. (2012). Insulin-like growth factor receptor inhibitors: baby or the bathwater? Journal of the National Cancer Institute 104, 975-981.

Yeh, J.E., Toniolo, P.A., and Frank, D.A. (2013). Targeting transcription factors: promising new strategies for cancer therapy. Current opinion in oncology 25, 652-658.

Yerbury, J.J., Poon, S., Meehan, S., Thompson, B., Kumita, J.R., Dobson, C.M., and Wilson, M.R. (2007). The extracellular chaperone clusterin influences amyloid formation and toxicity by interacting with prefibrillar structures. FASEB journal : official publication of the Federation of American Societies for Experimental Biology 21, 2312-2322.

Yuasa, H., Oike, Y., Iwama, A., Nishikata, I., Sugiyama, D., Perkins, A., Mucenski, M.L., Suda, T., and Morishita, K. (2005). Oncogenic transcription factor Evi1 regulates hematopoietic stem cell proliferation through GATA-2 expression. The EMBO journal 24, 1976-1987.

Zapf, J., Walter, H., and Froesch, E.R. (1981). Radioimmunological determination of insulinlike growth factors I and II in normal subjects and in patients with growth disorders and extrapancreatic tumor hypoglycemia. The Journal of clinical investigation 68, 1321-1330.

Zegarra-Moro, O.L., Schmidt, L.J., Huang, H., and Tindall, D.J. (2002). Disruption of androgen receptor function inhibits proliferation of androgen-refractory prostate cancer cells. Cancer research 62, 1008-1013.

Zeng, X., Zhang, H., Oh, A., Zhang, Y., and Yee, D. (2012). Enhancement of doxorubicin cytotoxicity of human cancer cells by tyrosine kinase inhibition of insulin receptor and type I IGF receptor. Breast cancer research and treatment 133, 117-126.

Zhang, L., Zhou, W., Velculescu, V.E., Kern, S.E., Hruban, R.H., Hamilton, S.R., Vogelstein, B., and Kinzler, K.W. (1997). Gene expression profiles in normal and cancer cells. Science 276, 1268-1272.

Zhang, P., Behre, G., Pan, J., Iwama, A., Wara-Aswapati, N., Radomska, H.S., Auron, P.E., Tenen, D.G., and Sun, Z. (1999). Negative cross-talk between hematopoietic regulators: GATA proteins repress PU.1. Proceedings of the National Academy of Sciences of the United States of America 96, 8705-8710.

135

Zhang, S.J., Ma, L.Y., Huang, Q.H., Li, G., Gu, B.W., Gao, X.D., Shi, J.Y., Wang, Y.Y., Gao, L., Cai, X., et al. (2008). Gain-of-function mutation of GATA-2 in acute myeloid transformation of chronic myeloid leukemia. Proceedings of the National Academy of Sciences of the United States of America 105, 2076-2081.

Zhao, H., Desai, V., Wang, J., Epstein, D.M., Miglarese, M., and Buck, E. (2012). Epithelial-mesenchymal transition predicts sensitivity to the dual IGF-1R/IR inhibitor OSI-906 in hepatocellular carcinoma cell lines. Molecular cancer therapeutics 11, 503-513.

Zhau, H.Y., Chang, S.M., Chen, B.Q., Wang, Y., Zhang, H., Kao, C., Sang, Q.A., Pathak, S.J., and Chung, L.W. (1996). Androgen-repressed phenotype in human prostate cancer. Proceedings of the National Academy of Sciences of the United States of America 93, 15152-15157.

Zhong, H., Fazenbaker, C., Breen, S., Chen, C., Huang, J., Morehouse, C., Yao, Y., and Hollingsworth, R.E. (2014). MEDI-573 Alone or in Combination with Mammalian Target of Rapamycin Inhibitors, Targets the Insulin-like Growth Factor Pathway in Sarcomas. Molecular cancer therapeutics.

Zhou, Y., Lim, K.C., Onodera, K., Takahashi, S., Ohta, J., Minegishi, N., Tsai, F.Y., Orkin, S.H., Yamamoto, M., and Engel, J.D. (1998). Rescue of the embryonic lethal hematopoietic defect reveals a critical role for GATA-2 in urogenital development. The EMBO journal 17, 6689-6700.

Zhu, Y., Liu, C., Nadiminty, N., Lou, W., Tummala, R., Evans, C.P., and Gao, A.C. (2013). Inhibition of ABCB1 expression overcomes acquired docetaxel resistance in prostate cancer. Molecular cancer therapeutics 12, 1829-1836.

Zingg, A.E., and Froesch, E.R. (1973). Effects of partially purified preparations with nonsuppressible insulin-like activity (NSILA-S) on sulfate incorporation into rat and chicken cartilage. Diabetologia 9, 472-476.

Zinn, R.L., Gardner, E.E., Marchionni, L., Murphy, S.C., Dobromilskaya, I., Hann, C.L., and Rudin, C.M. (2013). ERK phosphorylation is predictive of resistance to IGF-1R inhibition in small cell lung cancer. Molecular cancer therapeutics 12, 1131-1139.

Zong, Y., and Goldstein, A.S. (2013). Adaptation or selection--mechanisms of castration-resistant prostate cancer. Nature reviews Urology 10, 90-98.

Zoubeidi, A., Chi, K., and Gleave, M. (2010). Targeting the cytoprotective chaperone, clusterin, for treatment of advanced cancer. Clinical cancer research : an official journal of the American Association for Cancer Research 16, 1088-1093.