PI3K Regulatory Subunit P85alpha Plays a Tumor Suppressive Role in the Transformation of Mammary Epithelial Cells

PI3K regulatory subunit p85alpha plays a tumor suppressive role in the transformation of mammary epithelial cells

The Harvard community has made this article openly available. Please share how this access benefits you. Your story matters

Citation Thorpe, Lauren Marie. 2015. PI3K regulatory subunit p85alpha plays a tumor suppressive role in the transformation of mammary epithelial cells. Doctoral dissertation, Harvard University, Graduate School of Arts & Sciences.

Citable link http://nrs.harvard.edu/urn-3:HUL.InstRepos:14226045

Terms of Use This article was downloaded from Harvard University’s DASH repository, and is made available under the terms and conditions applicable to Other Posted Material, as set forth at http:// nrs.harvard.edu/urn-3:HUL.InstRepos:dash.current.terms-of- use#LAA

PI3K regulatory subunit p85alpha plays a tumor suppressive role

in the transformation of mammary epithelial cells

A dissertation presented

Lauren Marie Thorpe

The Division of Medical Sciences

in partial fulfillment of the requirements

for the degree of

Doctor of Philosophy

in the subject of

Virology

Harvard University

Cambridge, Massachusetts

December 2014

Dissertation Advisor: Dr. Jean J. Zhao Lauren Marie Thorpe

PI3K regulatory subunit p85alpha plays a tumor suppressive role

in the transformation of mammary epithelial cells

Abstract

Hyperactivation of the phosphatidylinositol 3-kinase (PI3K) pathway is one of the most common events in human cancers. Class IA PI3Ks are heterodimers of a p110 catalytic and a p85 regulatory subunit that coordinate the cellular response to extracellular stimuli.

Activating mutations in class IA PIγK catalytic isoform p110 are well established as causative in a number of cancer types. More recently, mutation or loss of the class IA regulatory isoform p85 (encoded by PIK3R1) has emerged as contributing to oncogenesis. In this dissertation, we use both in vitro and in vivo approaches to examine the role of p85 as a tumor suppressor in the transformation of mammary epithelial cells.

Using publically available online databases, we find heterozygous deletion of PIK3R1 occurs in 19-26% of breast tumors. Moreover, PIK3R1 expression is significantly decreased in breast tumors compared to normal breast tissue. In human mammary epithelial cells expressing dominant negative p53 (DDp53-HMECs), RNAi-mediated knockdown of PIK3R1 increases PI3K/AKT activation in response to growth factor stimulation and leads to transformation as assessed by anchorage-independent growth.

PIK3R1 knockdown also augments transformation of DDp53-HMECs by oncogenes, including activated HER2/neu. In a mouse model of HER2/neu-driven breast cancer, genetic ablation of Pik3r1 accelerates mammary tumor development. Transformation driven by p85 loss is largely mediated by signaling through catalytic isoform p110, as

iii

selective pharmacological inhibition of p110 but not p110 effectively blocks colony formation of PIK3R1 knockdown DDp53-HMECs and growth of Pik3r1 knockout tumors.

Mechanistically, we find that partial reduction of p85 increases the amount of p85- p110 bound to activated receptors, augmenting PI3K signaling and oncogenic transformation.

Together the work presented in this dissertation suggests that p85 depletion selectively targets a free negative regulator pool of this regulatory subunit that modulates PI3K activation under normal conditions, and transforms cells when lost. Furthermore, our work indicates that p85 plays a tumor suppressive role in the pathogenesis of breast tumors. Isoform-selective PI3K inhibitors are currently emerging in the clinic, and may offer improved specificity and reduced toxicity over first-generation pan-PI3K inhibitors.

Our findings suggest p110-selective therapies may be an effective treatment for breast cancers with reduced p85 expression.

Table of Contents

Abstract iii

Acknowledgements vi

Index of Figures vii

Index of Tables x

List of Abbreviations xi

Glossary of Terms xvi

Chapter 1: Introduction 1

Chapter 2: PI3K regulatory subunit p85alpha plays a tumor suppressive role in human mammary epithelial cells 46

Chapter 3: PI3K regulatory subunit p85alpha plays a tumor suppressive role in a genetically engineered mouse model of mammary tumorigenesis 94

Chapter 4: Summary, discussion, and future directions 132

Materials and Methods 150

References 165

Appendix A: Supplemental table of class I PI3K alterations in cancer, with complete references 192

Appendix B: Supplemental table of genetically engineered mouse models of PI3K isoforms in cancer, with complete references 200

Acknowledgements

I am deeply grateful to the many people who have supported me both professionally and personally throughout my years of study. Thank you to my undergraduate mentor, Dr.

Brooke McCartney, who gave me my first true research experience at Carnegie Mellon

University, and encouraged me to pursue graduate school. Thank you to my graduate mentor, Dr. Jean Zhao, an insightful scientist and caring mentor, who has afforded me many great opportunities during my time at Harvard University. Thank you to the past and current members of the Zhao lab for their guidance, in particular Dr. Hailing Cheng,

Thanh Von, Stephanie Santiago, Dr. Linda Clayton, Carolynn Ohlson, and Dr. Haluk

Yuzugullu. Thank you to my Dissertation Advisory Committee, Drs. Karl Münger, Lewis

Cantley, and Myles Brown, for their time and scientific expertise over the years; I am especially grateful to Karl for his mentoring and advice. Thank you to the Virology program, in particular Dr. David Knipe, and to my cohort of twelve, the Vironauts. Thank you to all of my friends and teammates, who have helped remind me that there is life outside the lab, and that it is good. Thank you to Hyun Kim, who has supported me every day and in every possible way. Thank you to my family: my parents Tom and Deb, my sister Jessica, and my brother David. You’ve been there for me through it all. This dissertation is dedicated to you.

Index of Figures

Chapter 1 Figure 1.1 The PI3K family comprises multiple classes and isoforms 7

Figure 1.2 Signaling by class I PI3K isoforms 8

Figure 1.3 Signaling by class II PI3K isoforms 10

Figure 1.4 Signaling by class III PI3K isoforms 11

Figure 1.5 Divergent roles of class I PI3K catalytic isoforms in the context of RTK, GPCR, and small GTPase inputs 19

Figure 1.6 Competition model for p110 and p110 regulation of RTK- mediated PI3K signaling 20

Figure 1.7 Molecular contexts dictating applications for isoform-selective PI3K inhibitors 21

Figure 1.8 Rational combination of PI3K inhibitors and other targeted therapeutics 32

Chapter 2 Figure 2.1 PIK3R1 expression is significantly reduced in breast cancers 54

Figure 2.2 Generation of DDp53-HMECs with stable RNAi-mediated PIK3R1 knockdown 57

Figure 2.3 PIK3R1 knockdown transforms DDp53-HMECs and increases growth factor-stimulated PI3K/AKT activation 59

Figure 2.4 Augmented PI3K/AKT activation in PIK3R1 knockdown DDp53-HMECs is rescued by ectopic expression of PIK3R1 61

Figure 2.5 Generation of DDp53-HMECs with activated HER2/neu and RNAi-mediated PIK3R1 knockdown 63

Figure 2.6 PIK3R1 knockdown increases transformation and PI3K/AKT signaling driven by activated HER2/neu in DDp53-HMECs 64

Figure 2.7 PIK3R1 knockdown increases transformation and PI3K/AKT signaling driven by p110-H1047R in DDp53-HMECs 66

Figure 2.8 Transformation of PIK3R1 knockdown DDp53-HMECs is blocked by p110-selective pharmacological inhibition 68

vii

Figure 2.9 Transformation of PIK3R1 knockdown DDp53-HMECs with activated HER2/neu is blocked by p110-selective inhibition 69

Figure 2.10 PI3K/AKT signaling in PIK3R1 knockdown DDp53-HMECs with activated HER2/neu is blocked by p110-selective inhibition 71

Figure 2.11 Endogenous p85 and PTEN do not appear to interact in DDp53-HMECs 72

Figure 2.12 PTEN and p85 do not appear to interact in a variety of cell types 74

Figure 2.13 PIK3R1 knockdown does not affect PTEN mRNA levels or lipid phosphatase activity in DDp53-HMECs 77

Figure 2.14 PIK3R1 knockdown does not increase growth factor-stimulated RTK phosphorylation in DDp53-HMECs 80

Figure 2.15 PIK3R1 knockdown does not affect growth factor-stimulated RTK trafficking in DDp53-HMECs 82

Figure 2.16 PIK3R1 knockdown increases transformation of DDp53-HMECs expressing activated ErbB3 84

Figure 2.17 PIK3R1 knockdown increases the amount of p85-p110 bound to activated RTKs in DDp53-HMECs 86

Figure 2.18 Model: partial p85 loss leads to increased PI3K/AKT signaling and transformation 87

Chapter 3 Figure 3.1 Schematic of mammary gland development in the mouse 97

Figure 3.2 Schematic of Pik3r1 conditional knockout allele and breeding scheme for mammary-specific Pik3r1 ablation 101

Figure 3.3 Transgenic MMTV-Cre ablates Pik3r1 expression in mouse mammary epithelial cells 103

Figure 3.4 Pik3r1 expression is not required for mouse mammary gland development 104

Figure 3.5 PI3K/AKT pathway activation in spontaneous mammary tumors from Pik3r1 knockout mice 109

Figure 3.6 Pathology of primary spontaneous mammary tumors and lung metastases from Pik3r1 knockout mice 110

Figure 3.7 Adjacent mammary glands from Pik3r1 knockout mice with spontaneous mammary tumors have a hypermorphic phenotype 111

viii

Figure 3.8 Schematic of the transgenic NIC allele and breeding scheme for mammary-specific HER2/neu expression and Pik3r1 ablation 113

Figure 3.9 Pik3r1 ablation reduces the latency of HER2/neu-driven mammary tumor development 114

Figure 3.10 Effect of Pik3r1 ablation on PI3K/AKT pathway activation in HER2/neu-driven mammary tumors 116

Figure 3.11 Quantification of the effect of Pik3r1 ablation on PI3K/AKT pathway activation in HER2/neu-driven mammary tumors 117

Figure 3.12 Effect of Pik3r1 ablation on tumor pathology and proliferation of HER2/neu-driven mammary tumors 119

Figure 3.13 Pan-PIγK or p110-selective inhibitors block the growth of transplanted HER2/neu tumors with Pik3r1 ablation 121

Figure 3.14 Pan-PIγK or p110-selective inhibitors suppress PI3K/AKT activation in transplanted HER2/neu tumors with Pik3r1 ablation 123

Figure 3.15 Pan-PIγK or p110-selective inhibitors suppress proliferation and induce apoptosis in transplanted HER2/neu tumors with Pik3r1 ablation 124

Figure 3.16 Model: heterozygous or homozygous Pik3r1 ablation has a similar effect on HER2/neu-driven tumorigenesis 130

Chapter 4 Figure 4.1 Model: modulation of p85 levels might produce a range of RTK-mediated PI3K activity 140

Index of Tables

Chapter 1 Table 1.1 Class I PI3K isoform alterations in cancer 4

Table 1.2 Pan-PI3K inhibitors and their clinical applications 29

Table 1.3 Dual pan-PI3K/mTOR inhibitors and their clinical applications 30

Table 1.4 Isoform-selective PI3K inhibitors and their clinical applications 31

Table 1.5 Combination of PI3K inhibitors with other targeted therapies in the clinic 33

Chapter 2 Table 2.1 PIK3R1 expression is significantly reduced in breast cancers across multiple microarray datasets 55

Chapter 3 Table 3.1 Nulliparous female mice with mammary-specific Pik3r1 ablation develop spontaneous mammary tumors 106

Table 3.2 Comparison of mammary tumor development in Pik3r1 knockout mice to other established GEMMs of breast cancer 107

List of Abbreviations

4G10 antibody clone specific to phosphotyrosines

ALL acute lymphoblastic leukemia

AML acute myeloid leukemia

ANOVA analysis of variance

BCL2 B cell lymphoma 2

BCL-XL B cell lymphoma-extra large

BCR B cell receptor

BET bromodomain and extraterminal domain

BH breakpoint cluster homology domain

BSA bovine serum albumin

C/EBP CCAAT/enhancer binding protein beta

Ca2+ calcium ion

CDK4 cyclin-dependent kinase 4

CDK6 cyclin-dependent kinase 6

CEF chicken embryo fibroblast

CLL chronic lymphocytic leukemia

CML chronic myelogenous leukemia

CRPC castration-resistant prostate cancer

CSF1 colony stimulating factor 1

DDp53 dominant negative p53 mutant

DM1 mertansine, a cytotoxic agent

DMEM Dulbecco’s modified eagle medium

DMEM/F12 Dulbecco’s modified eagle medium, nutrient mixture F1β

DTT dithiothreitol

EDTA ethylenediaminetetraacetic acid

EGF epidermal growth factor

EGFR epidermal growth factor receptor

ERK extracellular signal-regulated kinase

ESCC esophageal squamous cell carcinoma

FBS fetal bovine serum

FOXO forkhead box O transcription factors

GAP GTPase activating protein

GBM glioblastoma multiforme

GEF guanine nucleotide exchange factor

GEMM genetically engineered mouse model

GIST gastrointestinal stromal tumor

GPCR G-protein coupled receptor

H&E hematoxylin and eosin

HCC hepatocellular carcinoma

HMEC human mammary epithelial cell hTERT human telomerase reverse transcriptase

ID2 inhibitor of DNA binding 2

IGF1 insulin-like growth factor 1

IHC immunohistochemistry

INHL indolent non-Hodgkin lymphoma

INPP4B type II inositol 3,4-bisphosphate 4-phosphatase

IP immunoprecipitation

IR insulin receptor

IRES internal ribosomal entry site

IRS1 insulin receptor substrate 1 iSH2 inter-SH2 domain

xii

JAK2 Janus kinase 2

KI gene knock-in

KO gene knockout

LPA lysophosphatidic acid

LTR long terminal repeat

MAPK mitogen-activated protein kinase

MCL mantle cell lymphoma

MCL1 myeloid cell leukemia sequence 1

MEF mouse embryo fibroblast

MEK MAPK/ERK kinase miRNA micro RNA

MM multiple myeloma

MMEC mouse mammary epithelial cell

MMTV mouse mammary tumor virus mRNA messenger RNA

MTM myotubularin family phosphatases mTOR mammalian target of rapamycin mTORC1 mTOR complex 1

NcrNu Ncr nude athymic mouse strain

NRG1 neuregulin 1

NSCLC non-small cell lung carcinoma

P proline-rich domain p16INK4A cyclin-dependent kinase 4 inhibitor A p85 BD p85-binding domain p90RSK p90 ribosomal S6 kinase

PAGE polyacrylamide gel electrophoresis

xiii

PARP poly-(ADP-ribose) polymerase

PBS phosphate buffered saline

PCR polymerase chain reaction

PHTS PTEN hamartoma tumor syndrome

PI phosphatidylinositide; see also PtdIns

PI3K phosphatidylinositol 3-kinase

PIN prostatic intraepithelial neoplasia

PIP phosphatidylinositol 3-phosphate; see also PtdIns(3)P

PIP2 PtdIns 4,5-bisphosphate; see also PtdIns(4,5)P2

PIP3 PtdIns 3,4,5-trisphosphate; see also PtdIns(3,4,5)P3

PPAR peroxisome proliferator-activated receptor gamma

PPRE peroxisome proliferator response element

PR prolactin receptor

PtdIns phosphatidylinositide; see also PI

PtdIns(3)P phosphatidylinositol 3-phosphate; see also PIP

PtdIns(3,4)P2 PtdIns 3,4-bisphosphate

PtdIns(3,4,5)P3 PtdIns 3,4,5-trisphosphate; see also PIP3

PtdIns(4,5)P2 phosphatidylinositol 4,5-bisphosphate; see also PIP2

PTEN phosphatase and tensin homolog phosphatase

PyMT polyoma middle T qPCR quantitative polymerase chain reaction

RANKL receptor activator of nuclear factor kappa-B ligand

RBD RAS-binding domain

RNAi RNA interference

RTK receptor tyrosine kinase

S6K p70 ribosomal protein S6 kinase

xiv

SABCS San Antonio Breast Cancer Symposium

SCCHN squamous cell carcinoma of the head and neck

SD standard deviation

SDS sodium dodecyl sulfate

SEM standard error of the mean

SHH sonic hedgehog shRNA short hairpin RNA

SLL small lymphocytic leukemia

SMO smoothened sqNSCLC squamous non-small cell lung cancer

STAT3 signal transducer and activator of transcription 3

STAT5 signal transducer and activator of transcription 5

T-ALL T cell acute lymphoblastic leukemia

TBS tris-buffered saline

TBST tris-buffered saline with tween

TCC transitional cell carcinoma

TCGA The Cancer Genome Atlas

TCR T cell receptor

T-DM1 conjugate of DM1 to the monoclonal antibody Trastuzumab

TEB terminal end bud

TNBC triple-negative breast cancer

TUNEL terminal deoxynucleotidyl transferase dUTP nick end labeling

UTR untranslated region

WCL whole cell lysate

Glossary of Terms

Angiogenesis: the formation of new blood vessels from pre-existing vessels. Physiological angiogenesis is critical for normal growth and development, while pathophysiological angiogenesis is important for tumor growth.

Myristoylated: irreversible co-translational modification of proteins in which a myristoyl group is covalently attached to an N-terminal amino acid of a nascent polypeptide, promoting membrane localization of the modified protein.

Congenital mosaic overgrowth syndromes: a clinically heterogeneous group of genetic disorders characterized by abnormal progressive localized growth. They are caused by diverse somatic mutations and associated with increased cancer risk.

Inter-SH2 (iSH2) domain: the domain of p85 regulatory PI3K isoforms that is located between the C- and N-terminal SH2 domains and directly interacts with class IA p110 catalytic isoforms.

Megalencephaly syndromes: a collection of sporadic overgrowth disorders characterized by enlarged brain size and other distinct features.

SH2 domain: SRC homology 2 domain; a structurally conserved protein–protein interaction domain that facilitates interaction with phosphorylated tyrosine residues on other proteins.

RAS superfamily proteins: small monomeric membrane-associated GTPases, which are divided into the RAS, RHO, RAB, ARF, and RAN subfamilies based on structure and function.

RAS GTPases: subfamily of RAS superfamily GTPases that plays critical roles in signal transduction. In mammals, the three major RAS subfamily members are HRAS, KRAS, and NRAS.

RHO GTPases: subfamily of RAS superfamily proteins that shares similar roles in signal transduction to RAS GTPases and is best characterized for the regulation of cell shape, movement, and polarity.

Xenograft: transplantation of living cells, tissues, or organs from one species to another. Human cell lines are often xenografted into mice to study factors affecting tumor growth.

xvi

Chapter 1: Introduction

Acknowledgements

A manuscript based on the work in this chapter was accepted for publication in Nature

Reviews Cancer and is currently in press with the following authors: Lauren Thorpe,

Haluk Yuzugullu, and Jean Zhao. This content has been used in accordance with Nature

Publishing Group policy, which states that authors retain the right to reproduce their contribution in whole or in part in any printed volume of which they are an author. It has been reformatted to adhere to dissertation formatting guidelines.

Haluk Yuzugullu and Jean Zhao wrote the section on therapeutic targeting of PI3K isoforms in cancer, and Lauren Thorpe wrote all other sections. Lauren Thorpe, Haluk

Yuzugullu, and Jean Zhao edited the manuscript, and Lauren Thorpe and Haluk

Yuzugullu prepared the tables and figures.

We would additionally like to thank Tom Roberts for critical reading of the manuscript, and Tom Roberts and Lewis Cantley for helpful discussions.

Preface

Phosphatidylinositol 3-Kinases (PI3Ks) are critical coordinators of intracellular signaling in response to extracellular stimuli. Hyperactivation of PI3K signaling cascades is one of the most common events in human cancers. In this chapter, we discuss recent advances in our knowledge of the roles of distinct PI3K isoforms in normal and oncogenic signaling, the different ways in which PI3K can be upregulated, and the current state and future potential of targeting this pathway in the clinic.

Introduction

Phosphatidylinositol 3-Kinases (PI3Ks) are a family of lipid kinases that integrate signals from growth factors, cytokines, and other environmental cues, translating them into intracellular signals that regulate multiple signaling pathways. These pathways control many physiological functions and cellular processes, including cell proliferation, growth, survival, motility, and metabolism (Engelman et al., 2006; Liu et al., 2009;

Vanhaesebroeck et al., 2010). Activating alterations in PI3K are frequent in a variety of cancers (Table 1.1; for a fully referenced version see Appendix A), making this class of enzymes a prime drug target (Engelman, 2009; Liu et al., 2009). Tremendous efforts have been devoted to the development of effective PI3K inhibitors for cancer therapy.

Initial PI3K-directed drugs in clinical trials, consisting largely of non-isoform-selective pan-PI3K inhibitors, have not yielded exciting results. However, recent preclinical studies have demonstrated that different PI3K isoforms play divergent roles in cellular signaling and cancer, suggesting that inhibitors targeting individual isoforms may be able to achieve greater therapeutic efficacy. Isoform-selective inhibitors are now emerging in the clinic, and have had promising success. In this chapter, we provide an update on what has been learned in recent years about PI3K isoform-specific functions, differences in the modes of PI3K isoform activation, and the progress of isoform-selective inhibitors in

Table 1.1: Class I PI3K isoform alterations in cancer

Alteration Type Cancer Type Frequency of Sample Size Alteration Range Class IA PIK3CA (p110) Mutation Endometrial 10.3-53.0% 29-232 Breast 7.1-35.5% 65-507 Ovarian 33.0% 97 Colorectal 16.9†-30.6% 72-195 Bladder 5.0-20.0% 20-130 Lung 0.6-20.0% 5-183 Cervical 13.6% 22 Glioblastoma 4.3-11.0% 91-291 Head and neck 8.1-9.4% 32-74 Esophageal 5.5% 145 Melanoma 5.0% 121 Prostate 1.3-3.6% 55-156 Sarcoma 2.9% 207 Renal 1.0-2.9% 98-417 Liver 1.6% 125 Megalencephaly‡ 48.0% 50 Copy number Head and neck 9.1-100% 11-117 gain/amplification Cervical 9.1-76.4% 22-55 Lung 9.5-69.6% 3-92 Lymphoma 16.7-68.2% 22-60 Ovarian 13.3-39.8% 60-93 Gastric 36.4% 55 Thyroid 30.0% 110 Prostate 28.1% 32 Breast 8.7-13.4% 92-209 Glioblastoma 1.9-12.2% 139-206 Endometrial 10.3% 29 Thyroid 9.4% 128 Esophageal 5.7% 87 Leukemia 5.6% 161 Increased expression Prostate 40.0% 25 PIK3CB (p110) Mutation Breast 0.5% 183 Copy number Lung 56.5% 46 gain/amplification Thyroid 42.3% 97 Ovarian 5-26.9% NA-93 Lymphoma 20.0% 60 Glioblastoma 5.8% 103 Breast 4.9-5% NA-81 Increased expression Prostate 46.7% 30 Glioblastoma 3.9% 103 PIK3CD (p110) Copy number gain Glioblastoma 40.0% 10 Increased expression Neuroblastoma 52.6% 19 Glioblastoma 5.8% 103

Table 1.1: Class I PI3K isoform alterations in cancer (continued)

Alteration Type Cancer Type Frequency of Sample Size Alteration Range Class IA PIK3R1 (p85, p55, p50) Mutation Endometrial 19.8-32.8% 108-243 Pancreatic 16.7% 6 Glioblastoma 7.6-11.3% 91-291 Colorectal 4.6†-8.3% 108-195 Melanoma 4.4% 68 Ovarian 3.8% 80 Esophageal 3.4% 145 Breast 1.1-2.8% 62-507 Colon 1.7% 60 Decreased expression Breast 61.8% 458 Prostate 17-75%* NA Lung 19-46%* NA Ovarian 22%* NA Breast 18%* NA Bladder 18%* NA Copy number loss Ovarian 21.5% 93 PIK3R2 (p85) Mutation Endometrial 4.9% 243 Colorectal 0.9% 108 Megalencephaly‡ 22.0% 50 Amplification Lymphoma 23.3% 60 Increased expression Colon 55.0% 20 Breast 45.7% 35 PIK3R3 (p55) Copy number gain Ovarian 15.0% 93 Class IB PIK3CG (p110) Copy number gain Ovarian 19.3% 93 Increased expression Breast 77.5% 40 Prostate 72.4% 29 Medulloblastoma 52.9% 17 PIK3R5 (p101) Mutation Melanoma 38.2% 68 Gastric 2.7% 37

For further detail and references, see the expanded version of this table in Appendix A.

‡ Megalencephaly syndromes are a collection of sporadic overgrowth disorders characterized by enlarged brain size and other distinct features.

† Combined number of hypermutated and non-hypermutated colon and colorectal patient samples with mutations in the indicated gene.

* Represents the percent reduction in gene expression.

NA Sample size not available for this study.

preclinical and early clinical studies.

Multiple PI3K classes and isoforms

PIγKs phosphorylate the γ’-hydroxyl group of phosphatidylinositides (PtdIns). They are divided into three classes based on their structures and substrate specificities (Figure

1.1). In mammals, class I PI3Ks are further divided into subclasses IA and IB based on their modes of regulation. Class IA PI3Ks are heterodimers of a p110 catalytic subunit and a p85 regulatory subunit. The genes PIK3CA, PIK3CB, and PIK3CD respectively encode three highly homologous class IA catalytic isoforms, p110, p110, and p110.

These isoforms associate with any of five regulatory isoforms, p85 (and its splicing variants p55 and p50, encoded by PIK3R1), p85 (PIK3R2), and p55 (PIK3R3), collectively called p85 type regulatory subunits (reviewed in (Engelman et al., 2006;

Mellor et al., 2012)). Class IB PIγKs are heterodimers of a p110 catalytic subunit

(encoded by PIK3CG) coupled with regulatory isoforms p101 (PIK3R5) or p87 (p84 or p87PIKAP, encoded by PIK3R6). While p110 and p110 are ubiquitously expressed, p110 and p110 expression is largely restricted to leukocytes (Okkenhaug and

Vanhaesebroeck, 2003).

In the absence of activating signals, p85 interacts with p110, inhibiting p110 kinase activity. Upon receptor tyrosine kinase (RTK) or G-protein coupled receptor (GPCR) activation, class I PI3Ks are recruited to the plasma membrane, where p85 inhibition of p110 is relieved and p110 phosphorylates PtdIns 4,5-bisphosphate (PtdIns(4,5)P2) to generate PtdIns(3,4,5)P3 (Figure 1.2). This lipid product acts as a second messenger, activating AKT-dependent and –independent downstream signaling pathways (reviewed in (Engelman et al., 2006; Liu et al., 2009; Vanhaesebroeck et al., 2010)). The phosphatase and tensin homolog (PTEN) lipid phosphatase removes the γ’ phosphate

Figure 1.1: The PI3K family comprises multiple classes and isoforms. PI3Ks are classified based on their substrate specificities and structures. In vivo, class IA and IB PI3Ks phosphorylate PtdIns(4,5)P2, while class III PI3Ks phosphorylate PtdIns. Some evidence suggests that class II PI3Ks may also preferentially phosphorylate PtdIns in vivo (Falasca et al., 2007; Maffucci et al., 2005; Yoshioka et al., 2012). Class IA PI3Ks are heterodimers of a p110 catalytic subunit and a p85 regulatory subunit. Class IA catalytic isoforms (p110, p110, and p110) possess a p85-binding domain (p85 BD), RAS-binding domain (RBD), helical domain, and catalytic domain. Class IA p85 regulatory isoforms (p85, p85, p55, p55, and p50) possess an inter-SH2 (iSH2) domain that binds class IA catalytic subunits, flanked by SH2 domains that bind phosphorylated YXXM motifs. The longer isoforms, p85 and p85, additionally possess N-terminal SH3 and breakpoint cluster homology (BH) domains. Class IB PI3Ks are heterodimers of a p110 catalytic subunit and a p101 or p87 regulatory subunit. p110 possesses an RBD, helical domain, and catalytic domain. The domain structures of p101 and p87 are not fully known, but a C-terminal region of p101 has been identified as binding G subunits (Vadas et al., 2013). The monomeric class II isoforms (PI3K-Cβ, PI3K-Cβ, and PIγK-Cβ) possess an RBD, helical domain, and catalytic domain. VPS34, the only class III PI3K, possesses helical and catalytic domains. VPS34 forms a constitutive heterodimer with the myristoylated, membrane-associated VPS15 protein. Other indicated domains include proline-rich (P) domains and membrane-interacting C2 domains. Modified with permission from (Liu et al., 2009).

Figure 1.2: Signaling by class I PI3K isoforms. Upon receptor tyrosine kinase (RTK) or G-protein coupled receptor (GPCR) activation, class I PI3Ks are recruited to the plasma membrane by interaction with phosphorylated YXXM motifs on RTKs or their adaptors, or with GPCR-associated G subunits. There they phosphorylate PtdIns(4,5)P2 (PIP2) to generate PtdIns(3,4,5)P3 (PIP3), a second messenger which activates a number of AKT-dependent and –independent downstream signaling pathways regulating diverse cellular functions including growth, metabolism, motility, survival, and transformation. The phosphatase and tensin homolog (PTEN) lipid phosphatase removes the γ’ phosphate from PtdIns(γ,4,5)P3 to inactivate class I PI3K signaling. Modified with permission from (Liu et al., 2009).

from PtdIns(3,4,5)P3 to inactivate PI3K signaling.

Relatively little is known about class II PI3Ks. There are three class II isoforms, PI3K-

Cβ, PIγK-Cβ, and PIγK-Cβ, respectively encoded by PIK3C2A, PIK3C2B, and

PIK3C2G. These monomeric lipid kinases do not possess a regulatory subunit. PI3K-

Cβ and PIγK-Cβ are broadly expressed, while PIγK-Cβ expression is limited to the liver, prostate, and breast (Falasca and Maffucci, 2012). Although early experiments indicated that PI3K-Cβ and PIγK-Cβ could phosphorylate both PtdIns and PtdIns(4)P, in vivo PtdIns may be the preferred substrate, generating PtdIns(3)P (Falasca et al.,

2007; Maffucci et al., 2005; Yoshioka et al., 2012). The physiological roles of class II

PI3Ks are not fully understood, but recent studies suggest that PI3K-Cβ is important in angiogenesis (Yoshioka et al., 2012) and primary cilium function (Franco et al., 2014). In addition, PI3K-Cβ and PIγK-Cβ have been reported to regulate cellular functions including growth and survival (reviewed in (Falasca and Maffucci, 2012;

Vanhaesebroeck et al., 2010)) (Figure 1.3).

The single class III PI3K, VPS34, is encoded by PIK3C3. VPS34 forms a constitutive heterodimer with the myristoylated, membrane-associated VPS15 (encoded by PIK3R4), and phosphorylates PtdIns to produce PtdIns(3)P (Schu et al., 1993; Volinia et al.,

1995). In mammals, VPS34 is ubiquitously expressed (Volinia et al., 1995). The VPS34-

VPS15 dimer is found in distinct multiprotein complexes, which have critical roles in intracellular trafficking and autophagy (reviewed in (Backer, 2008; Vanhaesebroeck et al., 2010)) (Figure 1.4). The myotubularin (MTM) family phosphatases MTM1 and

MTMRβ remove the γ’ phosphate from PtdIns(γ)P, regulating the lipid products of class

II and III PI3Ks (Blondeau et al., 2000; Cao et al., 2008; Lu et al., 2012; Velichkova et al.,

2010).

Figure 1.3: Signaling by class II PI3K isoforms. Class II PI3Ks are not well understood, but may be activated by a number of different stimuli, including hormones, growth factors, chemokines, cytokines, phospholipids, and calcium (Ca2+). Although in vitro class II PI3Ks can phosphorylate both PtdIns and PtdIns(4)P, in vivo this class may preferentially phosphorylate PtdIns (PI) to generate PtdIns(3)P (PIP) (Falasca et al., 2007; Maffucci et al., 2005; Yoshioka et al., 2012). Class II PI3Ks regulate cellular functions including glucose transport, endocytosis, cell migration, and survival. Myotubularin (MTM) family phosphatases remove the γ’ phosphate from PtdIns(γ)P to inactivate class II PI3K signaling.

Figure 1.4: Signaling by class III PI3K isoforms. The class III VPS34-VPS15 heterodimer is found in distinct multiprotein complexes, which perform specific cellular functions. VPS34 may be activated by stimuli including amino acids, glucose, and other nutrients, and phosphorylates PtdIns (PI) to generate PtdIns(3)P (PIP). It plays critical roles in autophagy, endosomal trafficking, and phagocytosis. MTM family phosphatases remove the γ’ phosphate from PtdIns(γ)P to inactivate class III PIγK signaling.

Alterations of PI3K isoforms in cancer

Overactivation of the PI3K pathway is one of the most frequent events in human cancers. The most common mechanism leading to aberrant PI3K signaling is somatic loss of PTEN via genetic or epigenetic alterations (reviewed in (Parsons, 2004; Song et al., 2012)). The PI3K pathway can also be upregulated by activation of RTKs, or alterations in isoforms of PI3K itself (Table 1.1).

Class I PI3K catalytic isoform alterations

The transforming potential of class I PI3K catalytic isoforms was first demonstrated by studies in the late 1990s and early β000s, which showed that fusion of p110 to viral sequences (Chang et al., 1997) or the SRC myristoylation sequence (Klippel et al.,

1996; Zhao et al., 2003; Zhao et al., 2005) was activating and highly oncogenic. The

2004 discovery of frequent PIK3CA mutations in human cancers (Samuels et al., 2004) brought PI3K to the forefront as a major cancer driver and potential drug target. PIK3CA mutation has since been firmly established as causative in many cancer types (Table

1.1). Missense mutations occur in all domains of p110, but the majority cluster in two hotspots, the most common being E542K and E545K in the helical domain and H1047R in the kinase domain. Cell-based analyses confirmed that these hotspot mutations confer transformation via constitutive activation of p110 (Isakoff et al., 2005; Kang et al., 2005; Zhao et al., 2005). Subsequently, several studies using genetically engineered mouse models (GEMMs) demonstrated roles for mutant PIK3CA in tumor initiation, progression, and maintenance (Engelman et al., 2008; Kinross et al., 2012; Liu et al.,

2011; Wu et al., 2013; Yuan et al., 2013) (Appendix B). Helical domain mutations reduce inhibition of p110 by p85 (Burke et al., 2012; Huang et al., 2007; Miled et al.,

2007; Zhao and Vogt, 2010) or facilitate direct interaction of p110 with insulin receptor substrate 1 (IRS1) (Hao et al., 2013), while kinase domain mutations increase interaction

of p110 with lipid membranes (Burke et al., 2012; Huang et al., 2007; Mandelker et al.,

2009). Other PIK3CA mutations mimic distinct structural conformation changes that occur during activation of PI3K (Burke et al., 2012). Interestingly, some of these mutations in PIK3CA have also been reported in congenital mosaic overgrowth syndromes (Kurek et al., 2012; Orloff et al., 2013; Rios et al., 2013; Riviere et al., 2012).

In contrast, mutations in other class I catalytic isoforms are rare. While activating

PIK3CD mutations have been described in immune deficiencies (Angulo et al., 2013;

Lucas et al., 2014), they have not been linked to cancer. One PIK3CB mutation was detected in a single case of breast cancer (Kan et al., 2010); this helical domain substitution enhances basal PIγK activation, potentially by increasing p110 association with membranes (Dbouk et al., 2013). Recent structural studies have indicated that p110 may be less inhibited by p85 (Dbouk et al., 2010; Vogt, 2011; Zhang et al., 2011) and thus has higher basal transforming potential. Interestingly, p110 expression has been detected in some human solid cancer cell lines (Sawyer et al., 2003), and overexpression of wildtype p110, p110, or p110, but not p110, transforms cells in vitro (Kang et al., 2006). This is consistent with the fact that PIK3CB, PIK3CD, and

PIK3CG are generally amplified or overexpressed, but not mutated, in cancers (Table

1.1).

Class I PI3K regulatory isoform alterations

Recent studies have converged to implicate the p85 regulatory isoforms in tumorigenesis. Since the initial discovery of PIK3R1 mutations in human cancer cell lines and primary tumors (Philp et al., 2001), somatic mutations in PIK3R1 have been identified in a number of different cancers (Cancer Genome Atlas Research, 2008;

Cheung et al., 2011; Cizkova et al., 2013; Jaiswal et al., 2009; Urick et al., 2011) (Table

1.1). The majority are substitutions or in-frame insertions or deletions in the inter-SH2

(iSHβ) domain of p85 (Cancer Genome Atlas Research, 2008; Cheung et al., 2011;

Jaiswal et al., 2009; Urick et al., 2011), the region of the protein that makes contact with p110 (Huang et al., 2007), indicating this domain as a mutation hotspot (Cheung et al.,

2011). A number of these iSH2 domain mutants retain the ability to bind and stabilize p110 isoforms, but promote enhanced PI3K activity and transformation due to reduced ability to inhibit p110 (Cheung et al., 2011; Jaiswal et al., 2009; Sun et al., 2010; Urick et al., 2011; Wu et al., 2009).

In addition, reduced expression of PIK3R1 has been reported in some cancers (Cizkova et al., 2013; Taniguchi et al., 2010) (Table 1.1). PIK3R1 mRNA levels inversely correlated with malignancy grade and incidence of metastasis in both breast and liver cancers (Cizkova et al., 2013; Taniguchi et al., 2010). In mice, Pik3r1 ablation increased epithelial neoplasia driven by Pten loss (Luo et al., 2005c) and led to spontaneous development of aggressive liver tumors (Taniguchi et al., 2010). This work indicates that p85 can negatively regulate PIγK signaling in cancer, and suggests that p85 has tumor suppressive functions in certain tissues (Luo and Cantley, 2005).

Alterations in genes encoding other regulatory isoforms have also been detected, albeit at a lower frequency. Increased PIK3R2 expression has been reported in breast and colon cancers (Cortes et al., 2012) (Table 1.1). Consistent with this, overexpression of wildtype p85 increased PIγK pathway activation in cells and tumor formation in mice

(Cortes et al., 2012). Somatic PIK3R2 mutations have been found in endometrial and colorectal cancers (Cheung et al., 2011; Jaiswal et al., 2009), and causative germline

PIK3R2 mutations have been reported in megalencephaly syndromes (Riviere et al.,

2012). All PIK3R2 mutations described to date are substitutions with no apparent

hotspot region, and similar to some p85 mutants, mutations in p85 increase PIγK activation without affecting p110 binding (Cheung et al., 2011). Together these studies indicate that PI3K regulatory isoforms may contribute to tumorigenesis by multiple mechanisms.

Class II PI3K isoform alterations

Although class II PI3Ks are not well understood, PIK3C2A or PIK3C2B expression has been implicated in physiological functions important to tumorigenesis (Biswas et al.,

2013; Diouf et al., 2011; Elis et al., 2008; Katso et al., 2006; Maffucci et al., 2005).

PIK3C2B amplification has been reported in glioblastoma (Knobbe and Reifenberger,

2003; Nobusawa et al., 2010; Rao et al., 2010), and somatic PIK3C2B mutations were detected in non-small cell lung cancer (Liu et al., 2012), but the functional consequence of these mutations is unknown. Perhaps the most convincing evidence towards a role for class II PI3Ks in tumorigenesis comes from a recent study demonstrating that mice with

Pik3c2a ablation had compromised angiogenesis and vascular barrier integrity, and significant reduction in the size and microvessel density of implanted tumors (Yoshioka et al., 2012). Since mice with embryonic Pik3c2a or Pik3c2b knockout (KO) are viable

(Harada et al., 2005; Harris et al., 2011), a class II-selective PI3K inhibitor might target tumor angiogenesis with tolerable side effects, although toxicity due to the critical role of

PI3K-Cβ in maintaining normal renal homeostasis (Harris et al., 2011) would need to be considered.

Type II inositol 3,4-bisphosphate 4-phosphatase (INPP4B), the phosphatase responsible for dephosphorylation of PtdIns(3,4)P2 to PtdIns(3)P (Gewinner et al., 2009; Norris et al.,

1997), has also been implicated in cancer. In human mammary cell lines, INPP4B knockdown increased AKT activation and transformation (Fedele et al., 2010; Gewinner

et al., 2009). INPP4B loss-of-heterozygosity has been detected in cancers (Gewinner et al., 2009; Stjernstrom et al., 2014), and reduced INPP4B expression has been correlated with high tumor grade, earlier recurrence, and decreased survival (Fedele et al., 2010;

Gewinner et al., 2009; Hodgson et al., 2011). Identification of INPP4B as a tumor suppressor suggests that deregulation of the class II PI3K lipid products may contribute to tumorigenesis.

Class III PI3K isoform alterations

There is currently little evidence indicating an oncogenic role for VPS34. One recent study suggested that VPS34 is tyrosine-phosphorylated and activated downstream of

SRC, and its lipid kinase activity is required for SRC-mediated transformation (Hirsch et al., 2010). However, overexpression of wildtype or myristoylated VPS34 was not sufficient to induce cellular transformation (Denley et al., 2009). Another study indicated that VPS34 activity might be decreased in the context of activated epidermal growth factor receptor (EGFR) (Wei et al., 2013). Further investigation is needed to determine whether VPS34 plays a role in transformation.

Divergent roles of class I PI3K catalytic isoforms

Class I PI3K catalytic isoforms share a conserved domain structure. They utilize the same lipid substrates and generate the same lipid products. Despite their similarities, accumulating evidence indicates these isoforms have distinct roles in mediating PI3K signaling in physiological and oncogenic contexts.

GEMMs have been used to elucidate the roles of individual class I PI3K isoforms. Mice with germline KO of Pik3ca or knock-in (KI) of a kinase-dead Pik3ca allele die at day

E10.5 (Bi et al., 2002; Bi et al., 1999). Interestingly, Pik3cb KO mice die much earlier at

day E3.5 (Bi et al., 2002), while kinase-dead Pik3cb KI mice develop to maturity with minor defects in size and glucose metabolism, and major defects in male fertility (Ciraolo et al., 2008; Ciraolo et al., 2010). These differences suggest an important kinase- independent scaffolding role for p110 (Ciraolo et al., 2008). Germline inactivation of

Pik3cd or Pik3cg by KO or KI of kinase-dead alleles yields viable mice that grow to adulthood; however, loss of p110 results in functional defects in lymphocytes, neutrophils, and mast cells (Ali et al., 2004; Clayton et al., 2002; Jou et al., 2002;

Okkenhaug et al., 2002), while loss of p110 impairs thymocyte development, T cell activation, and neutrophil migration (Martin et al., 2008; Sasaki et al., 2000; Yum et al.,

2001). These studies indicate non-redundant roles in mouse embryonic development for p110 and p110, the two ubiquitously expressed class I PIγK isoforms, and distinct roles in the immune system and inflammatory response for p110 and p110, the two leukocyte-restricted isoforms.

Technological developments have facilitated further insight into the individual roles of

PI3K enzymes. The generation of conditional KO animals using the Cre/loxP recombination system has allowed the functions of each isoform to be studied in different tissues, stages of development, and pathological settings (Appendix B).

Additional progress has come from studies using RNA interference (RNAi) and a new generation of isoform-selective PI3K inhibitors. These have advanced our understanding of the roles of class I catalytic isoforms in mediating signaling downstream of RTKs,

GPCRs, and small GTPases (Figure 1.5 and Figure 1.6), and in the context of PTEN deficiency (Figure 1.7).

In mediating RTK signaling

Binding of growth factor ligands induces RTK dimerization, activation, and auto-

phosphorylation of tyrosine-containing YXXM motifs on the receptors or their associated adaptor proteins. Class IA p110-p85 heterodimers are then recruited to activated RTKs through direct interaction of p85 SH2 domains with these phosphorylated YXXM motifs

(Rameh et al., 1995; Yu et al., 1998a; Yu et al., 1998b) (Figure 1.2). Accordingly p110, p110, and p110 can complex with activated RTKs (Figure 1.5), and might be expected to mediate growth factor signaling.

Studies using isoform-selective pharmacological inhibitors and genetic inactivation or ablation indicated that loss of p110 activity was sufficient to largely block PI3K signaling in response to a number of growth factors (Foukas et al., 2006; Graupera et al., 2008;

Knight et al., 2006; Sopasakis et al., 2010; Utermark et al., 2012; Zhao et al., 2006).

Notably, genetic ablation or inactivation of p110 had only a modest effect on PIγK signaling following acute RTK activation (Ciraolo et al., 2008; Guillermet-Guibert et al.,

2008; Jia et al., 2008). It was suggested that the relative abundance of catalytic isoforms in a particular tissue might dictate which isoforms are dominant in mediating RTK signaling (Chaussade et al., 2007). This may explain the role of p110, which is mainly expressed in leukocytes and is the primary isoform regulating PI3K signaling downstream of certain RTKs in mast cells and macrophages (Ali et al., 2004;

Papakonstanti et al., 2008; Vanhaesebroeck et al., 1999). However, differential expression does not completely explain isoform dependence, as in many tissues p110 levels are comparable to or even higher than levels of p110 (Geering et al., 2007).

The involvement of p110 in RTK signaling remained puzzling, until a recent study from our group suggested a new model. In mice, while p110 ablation blocked normal mammary development and mammary tumorigenesis driven by polyoma middle T

(PyMT) or HERβ (also known as ERBBβ), p110 ablation increased mammary gland

Figure 1.5: Divergent roles of class I PI3K catalytic isoforms in the context of RTK, GPCR, and small GTPase inputs. Class I PI3Ks mediate signaling downstream of RTKs, GPCRs, and small GTPases. Left: p85 regulatory subunits bind phosphorylated YXXM motifs on activated RTKs. Because p110, p110, and p110 bind p85, these isoforms mediate signaling downstream of RTKs. Recent evidence also suggests that p87-p110 may be activated by certain RTKs (Schmid et al., 2011). Middle: Small GTPases synergize with RTK and GPCR signals to directly activate PI3Ks by interacting with their RAS-binding domains (RBDs). Isoforms p110, p110, and p110 bind RAS family GTPases, while p110 binds the RHO family GTPases RAC1 and CDC42 (Fritsch et al., 2013). Right: G and G proteins dissociate from activated GPCRs. Catalytic isoforms p110 and p110, and regulatory isoform p101, directly bind and are activated by G. p110 may be activated downstream of GPCRs, but the mechanism is unknown (Durand et al., 2009; Reif et al., 2004; Saudemont et al., 2009). G proteins have been reported to directly bind and inhibit p110 (Ballou et al., 2006; Ballou et al., 2003; Yeung and Wong, 2010). Modified with permission from (Vanhaesebroeck et al., 2010).

Figure 1.6: Competition model for p110 and p110 regulation of RTK-mediated PI3K signaling. Based on work presented in (Utermark et al., 2012). Both p85-p110 and p85-p110 compete for phosphorylated YXXM sites on activated RTKs. However, the maximal specific activity and enzymatic rate of p110 are higher than that of p110 (Beeton et al., 2000; Meier et al., 2004), and RTK-associated p110 may have higher lipid kinase activity than p110 (Utermark et al., 2012). By this model, loss or inactivation of p110 or p110 differentially modulates RTK signaling. Knockout of p110 allows all sites to be occupied by the less active p110, decreasing RTK output. Conversely, knockout of p110 allows all sites to be bound by the more active p110, increasing RTK output. Genetically or pharmacologically inactivated p110 or p110 can still bind RTKs but cannot signal, reducing RTK output.

Figure 1.7: Molecular contexts dictating applications for isoform-selective PI3K inhibitors. Red box: Upregulation or mutation of receptor tyrosine kinases (RTKs), oncogenic RAS mutations, or activating p110 mutations all increase PtdIns(γ,4,5)P3 production through p110, which can be amplified by mutation or loss of PTEN. In these contexts use of p110-selective inhibitors is effective. Blue box: In the absence of other oncogenic alterations, PTEN loss or mutation increases PtdIns(3,4,5)P3 production through p110, perhaps due to RAC1- or CDC42-mediated p110 activation, or the basal activity of this isoform. In this context use of p110-selective inhibitors is effective. Green box: Upregulation or mutation of B cell receptors (BCRs), cytokine receptors, or other immune cell surface markers increases PtdIns(3,4,5)P3 production through p110. In this context use of p110-selective inhibitors is effective.

outgrowth and accelerated tumor formation driven by these oncogenic RTKs (Utermark et al., 2012). To explain this negative role of p110, a competition model was proposed: if p110 has higher RTK-associated lipid kinase activity than p110, the less-active p110 could compete with p110 for phosphorylated YXXM sites on receptors to modulate PI3K signal strength downstream of RTKs (Utermark et al., 2012) (Figure 1.6).

Although direct comparison of RTK-associated p110 and p110 lipid kinase activity has not been shown, the maximal specific activity and enzymatic rate of p110 are higher than that of p110 (Beeton et al., 2000; Meier et al., 2004). Biochemical data were consistent with this proposed model, demonstrating that in p110 KO cells, activated

RTKs had more bound p110 and higher associated lipid kinase activity (Utermark et al.,

2012). Furthermore, pharmacologically inactivated p110 could still compete with p110 for binding sites on activated receptors, modestly reducing signaling and tumor growth driven by PyMT or HER2 (Utermark et al., 2012). This model also explains moderately decreased AKT activation, mild hyperglycemia, and delayed HER2-driven tumor formation observed in mice with KI of kinase-dead p110 (Ciraolo et al., 2008), a scenario mimicking p110-selective kinase inhibition. These studies not only reveal a novel p110-based regulatory mechanism in RTK-mediated PI3K signaling, but also identify p110 as an important target in cancers driven by oncogenic RTKs.

Initial studies suggested that class IA isoforms mediated signaling downstream of RTKs, while the class IB isoform signaled downstream of GPCRs. Although p110 activation by

GPCRs is well established, a recent report suggested that this class IB isoform might also function downstream of RTKs through regulatory isoform p87 in mouse myeloid cells (Schmid et al., 2011) (Figure 1.5). Given that p87 and p101 may have distinct tissue distribution (Bohnacker et al., 2009; Shymanets et al., 2013; Voigt et al., 2006) and non-redundant functions (Bohnacker et al., 2009; Kurig et al., 2009; Schmid et al.,

2011; Shymanets et al., 2013), this suggests that the two class IB regulatory isoforms may mediate p110 activation in response to specific upstream signals.

In mediating GPCR signaling

GPCRs are a family of seven-transmembrane domain receptors that associate with heterotrimeric G proteins composed of the G and G subunits. Ligand binding to

GPCRs results in allosteric activation and disassociation of bound G proteins into their separate subunits, which can then act on intracellular targets.

The single class IB PIγK isoform, p110, is activated by G proteins (Brock et al., 2003;

Maier et al., 1999; Stoyanov et al., 1995) (Figure 1.2). Although association of p110 with either its p101 or p87 regulatory isoforms increased its activation in response to G

(Brock et al., 2003; Stephens et al., 1997; Suire et al., 2005), recent evidence indicated that p101 is the main regulatory isoform involved in GPCR-mediated p110 signaling

(Kurig et al., 2009; Schmid et al., 2011) (Figure 1.5). Both p110 and p101 interact directly with G heterodimers, and these contacts are critical for signaling and transformation mediated by p110 (Brock et al., 2003; Vadas et al., 2013). Recent studies have shown that in myeloid cells, p110 can be activated by GPCR and RTK signals in a RAS- or RAP1A-dependent manner to mediate integrin 41 activity, leading to tumor inflammation and progression (Schmid et al., 2011; Schmid et al.,

2013). Thus p110-mediated signaling may contribute to tumorigenesis by controlling both tumor cell characteristics and the tumor microenvironment.

Interestingly, in vitro experiments (Kubo et al., 2005; Kurosu et al., 1997; Maier et al.,

1999; Murga et al., 2000) and subsequent GEMM studies (Ciraolo et al., 2008;

Guillermet-Guibert et al., 2008; Jia et al., 2008) demonstrated a role for p110 in G

protein-mediated PI3K signaling (Figure 1.2). Recently a region in the C2-helical domain linker of p110 was shown to bind G subunits (Figure 1.5); this region is not present in other class IA isoforms (Dbouk et al., 2012), and is similar to the region of p110 that binds G (Vadas et al., 2013). Abrogation of p110-G interaction blocked p110- mediated signaling and transformation downstream of GPCRs, and inhibited the proliferation and invasiveness of cancer cells (Dbouk et al., 2012). Although p110 does not directly interact with G proteins, a non-redundant role for this isoform in GPCR- mediated leukocyte migration has been demonstrated in certain contexts (Durand et al.,

2009; Reif et al., 2004; Saudemont et al., 2009); however, the mechanism of p110 activation downstream of GPCRs is unknown. It has also been reported that some G proteins directly bind and inhibit p110 (Ballou et al., 2006; Ballou et al., 2003; Yeung and Wong, 2010). Clearly, class I PI3K isoforms cooperate with GPCRs in a number of different ways to regulate signaling and transformation.

Downstream of RAS and other small GTPases

RAS superfamily proteins are direct activators of the PI3K pathway. All class I PI3K catalytic isoforms possess an N-terminal RAS-binding domain (RBD) (Figure 1.1) allowing them to interact with RAS GTPases or other RAS superfamily members (Figure

1.5).

Activated or oncogenic mutant RAS proteins directly bind and increase the enzymatic activity of both p110 (Rodriguez-Viciana et al., 1994; Rodriguez-Viciana et al., 1996) and p110 (Pacold et al., 2000; Rubio et al., 1997; Suire et al., 2002). Cellular and structural studies suggest that p110 association with RAS might both increase its membrane translocation (Kurig et al., 2009; Pacold et al., 2000) and allosterically increase p110 kinase activity (Pacold et al., 2000). Interestingly, RAS is required for

activation of p110 bound to regulatory isoform p87, but not p101 (Kurig et al., 2009). In vitro, the transforming capability of both helical domain p110 mutants (Zhao and Vogt,

2008, 2010) and of overexpressed wildtype p110 (Denley et al., 2008; Kang et al.,

2006) are dependent on their association with RAS. GEMM studies using KI of Pik3ca with an RBD mutation or KO of endogenous Pik3ca revealed that the p110-RAS interaction is critical for both the initiation and maintenance of lung tumors (Castellano et al., 2013; Gupta et al., 2007) and the development of myeloid leukemia (Gritsman et al.,

2014) driven by oncogenic KRAS. In mice, p110-RAS binding is required for inflammation-induced PtdIns(3,4,5)P3 accumulation (Suire et al., 2006) and inflammation-associated tumor progression (Schmid et al., 2011; Schmid et al., 2013).

These studies highlight the importance of p110 or p110 interaction with RAS in both normal PI3K signaling and transformation.

Although p110 was shown to bind RAS in vitro (Fritsch et al., 2013; Vanhaesebroeck et al., 1997), some studies indicated that p110 kinase activity was not stimulated by

HRAS, NRAS, or KRAS, but instead by RRAS and TC21 (also known as RRAS2)

(Murphy et al., 2002; Rodriguez-Viciana et al., 2004). Furthermore, B and T cells derived from Tc21 KO mice displayed diminished PIγK activity and recruitment of p110 to T cell receptors (TCRs) and B cell receptors (BCRs), suggesting that TC21 might function upstream of p110 (Delgado et al., 2009). Thus PI3K signaling through p110 may be regulated by additional RAS subfamily members.

It was initially anticipated that all p110 isoforms bearing a RBD might interact with RAS

GTPases. Surprisingly, in vitro studies determined that p110 kinase activity was not stimulated by any RAS subfamily members (Rodriguez-Viciana et al., 2004). A recent extensive biochemical study demonstrated that p110 is instead regulated by RAC1 and

CDC42 of the RHO GTPase subfamily (Fritsch et al., 2013) (Figure 1.5). Direct interaction between the p110 RBD and RAC1 is important for GPCR-mediated activation of p110 (Fritsch et al., 2013), indicating cooperative G and RHO GTPase signaling through p110. Previous studies reported that an intact RBD was required for signaling and oncogenic transformation by wildtype p110 in cultured cells (Denley et al., 2008; Kang et al., 2006), suggesting a potential role for RHO GTPase interaction with p110 in transformation. Notably, RAC1 and CDC4β can also be activated downstream of PI3K by PtdIns(3,4,5)P3-dependent guanine nucleotide exchange factors

(GEFs) and GTPase activating proteins (GAPs) (Klarlund et al., 1997; Krugmann et al.,

2002; Welch et al., 2002). The finding of distinct p110 regulation by RAC1 and CDC4β expands PI3K signaling input by GTPases beyond the RAS subfamily, and also supports the notion that PI3K can act both upstream and downstream of GTPases, potentially allowing for positive feedback loops in cancer settings.

In PTEN deficiency

The PTEN lipid phosphatase counteracts class I PI3K activity, making it an important tumor suppressor. Somatic loss of PTEN in human cancers is common. Germline PTEN mutations are also found in several genetic disorders characterized by multiple hamartomas with overgrowth phenotypes, collectively termed PTEN hamartoma tumor syndromes (PHTS) (Liaw et al., 1997).

Pten KO mouse models provided a tool to explore the molecular mechanisms underlying diseases caused by PTEN loss. While embryonic Pten KO is lethal (Di Cristofano et al.,

1998; Stambolic et al., 1998), heterozygous or conditional Pten KO animals recapitulated human disease phenotypes, including development of prostate cancer

(Alimonti et al., 2010; Kwabi-Addo et al., 2001; Wang et al., 2003). Surprisingly, ablation

of p110, but not p110, blocked prostatic intraepithelial neoplasia (PIN) induced by

PTEN loss (Jia et al., 2008). Subsequent studies demonstrated a correlation between

PTEN deficiency and sensitivity to p110 knockdown or inhibition in human cancer cell lines both in vitro and in mouse xenografts (Ni et al., 2012; Torbett et al., 2008; Wee et al., 2008). However, the mechanism governing the specific importance of p110 in the context of PTEN loss remains elusive. Perhaps the unique role for p110 as a convergence point for GPCR and RAC1 or CDC42 signals (Figure 1.7) contributes to transformation induced by PTEN deficiency. Structural studies have also suggested that compared to p110, p110 is less inhibited by p85, and may supply a basal level of

PtdIns(3,4,5)P3 (Dbouk et al., 2010; Vogt, 2011; Zhang et al., 2011). This may explain why wildtype p110 can be oncogenic when it is overexpressed (Denley et al., 2008;

Kang et al., 2006) or when PTEN is lost.

Although p110 is the primary PIγK isoform involved in many cases of tumorigenesis driven by PTEN loss, studies have shown that depending on the tissue type and pathology both p110 and p110 may be involved (Berenjeno et al., 2012; Jia et al.,

2013; Schmit et al., 2014). Mice with Pten ablation in the basal epidermal compartment require both p110 and p110 for the development of hyperproliferative epidermal lesions closely resembling PHTS (Wang et al., 2013a; Wang et al., 2013b). In this model, spatially distinct roles for these isoforms in epidermal compartments were identified: p110 is responsible for RTK signaling in and survival of suprabasal cells, whereas p110 is important for GPCR signaling in and proliferation of basal cells (Wang et al., 2013a). In mice with thymocyte-specific Pten KO, not surprisingly both p110 and p110 were required for the development of T cell acute lymphoblastic leukemia (T-ALL)

(Subramaniam et al., 2012). This suggests that in certain contexts, transformation driven by PTEN loss may be governed by the PI3K isoforms that are dominant in that tissue or

compartment.

Since PTEN loss removes one mechanism of PI3K pathway negative regulation, the specific roles of p110 isoforms in this pathogenic context can be influenced by other activating inputs. These can be cues from the tissue microenvironment, or other co- existing genetic events. A recent GEMM study demonstrated that concomitant activation of oncogenic KRAS in ovarian endometrioid adenocarcinoma driven by Pten ablation shifted the PIγK isoform reliance from p110 to p110 (Schmit et al., 2014) (Figure 1.7).

Consistent with this, a subset of PTEN-mutant human endometrioid endometrial cancer cell lines harboring other PI3K-activating mutations were found to be resistant to p110 inhibition (Weigelt et al., 2013). It is also possible that other genetic events downstream of PI3K or in PI3K-independent pathways may render PTEN-null tumors less reliant on

PI3K. Thus determination of isoform dependency in PTEN-deficient tumors remains a challenge.

Therapeutic targeting of PI3K isoforms in cancer

The central role of PI3K in cancer makes it an attractive therapeutic target. Enormous efforts have focused on the development of drugs targeting PI3K, many of which are undergoing clinical evaluation (Table 1.2, Table 1.3, and Table 1.4). Unlike drugs targeting other oncogenic kinases, such as EGFR, BRAF, and ALK, PI3K inhibitors have shown limited efficacy as mono-therapies in early trials on patients with tumors harboring

PI3K pathway activation (Rodon et al., 2013). The effectiveness of these early PI3K inhibitors may have been limited by their lack of specificity, and by compensatory signaling feedback loops and co-existing genetic and epigenetic alterations. The development of novel isoform-selective PI3K inhibitors (Figure 1.7) and their rational combination with other therapeutics (Figure 1.8 and Table 1.5) may substantially

Table 1.2: Pan-PI3K inhibitors and their clinical applications

Agent Company Target Trial stage* Tumor types* BKM120 Novartis Class I PI3Ks I, II, and III . NSCLC . Endometrial . Thyroid . CRPC . Breast . Colorectal . Head and neck . GBM . Renal cell . B cell lymphoma . GIST . Melanoma . Ovarian . Prostate . Pancreatic . Leukemia . Esophageal . Cervical . Non-Hodgkin lymphoma . Squamous NSCLC . Adv. solid tumors GDC0941 Genentech Class I PI3Ks I and II . Breast . NSCLC . Non-Hodgkin lymphoma . Adv. solid tumors BAY80-6946 Bayer Class I PI3Ks I and II . Non-Hodgkin lymphoma . Adv. solid tumors ZSTK474 Zenyaku Kogyo Co. Class I PI3Ks I and II . Adv. solid tumors PX866 Oncothyreon Class I PI3Ks I and II . Colorectal . SCCHN . Melanoma . NSCLC . Prostate . GBM . Adv. solid tumors XL147 Exelixis/Sanofi- Class I PI3Ks I and II . Breast Aventis . Endometrial . Ovarian . Lymphoma . GBM . NSCLC . Adv. solid tumors CH5132799 Chugai Pharma Class I PI3Ks I . Adv. solid tumors Europe

NSCLC, non-small cell lung carcinoma; CRPC, castration-resistant prostate cancer; GIST, gastrointestinal stromal tumor; SCCHN, squamous cell carcinoma of the head and neck; GBM, glioblastoma multiforme.

* Data taken from an April 2014 search of http://www.clinicaltrials.gov.

Table 1.3: Dual pan-PI3K/mTOR inhibitors and their clinical applications

Agent Company Target Trial stage* Tumor types* GDC0980 Genentech PI3K/mTOR I and II . Prostate . Breast . Endometrial . Renal cell . Non-Hodgkin lymphoma . Adv. solid tumors PF04691502 Pfizer PI3K/mTOR I . Adv. solid tumors BGT226 Novartis PI3K/mTOR I and II . Adv. solid tumors BEZ235 Novartis PI3K/mTOR I and II . Breast . Renal cell . Prostate . TCC . ALL . AML . CML . Pancreatic neuroendocrine . Adv. solid tumors XL765 Sanofi PI3K/mTOR I . GBM . Adv. solid tumors GSK2126458 GlaxoSmithKline PI3K/mTOR I . Solid tumors . Lymphoma DS7423 Daiichi Sankyo PI3K/mTOR I . Solid tumors PWT33597 Pathway Therapeutics PI3K/mTOR I . Adv. solid tumors SF1126 Semafore PI3K/mTOR I . Adv. solid tumors Pharmaceuticals PF05212384 Pfizer PI3K/mTOR I and II . Adv. solid tumors

TCC, transitional cell carcinoma; ALL, acute lymphoblastic leukemia; AML, acute myeloid leukemia; CML, chronic myelogenous leukemia; GBM, glioblastoma multiforme.

* Data taken from an April 2014 search of http://www.clinicaltrials.gov.

Table 1.4: Isoform-selective PI3K inhibitors and their clinical applications

Agent Company Target Trial stage* Tumor types* BYL719 Novartis p110 I and II . SCCHN . ESCC . Colorectal . Breast . GIST . Kidney . Pancreas . Gastric . Adv. solid tumors GDC0032 Genentech p110 I . Breast . Adv. solid tumors INK1117 Intellikine/Millenium p110 I . Adv. solid tumors AZD8186 Astra-Zeneca p110 I . CRPC . sqNSCLC . TNBC . Adv. solid tumors with PTEN deficiency GSK2636771 GlaxoSmithKline p110 I and II . Adv. solid tumors with PTEN deficiency SAR260301 Sanofi p110 I . Adv. solid tumors . Lymphoma IPI145 Infinity p110 and I, II, and III . CLL p110 . SLL . ALL . INHL . Hematologic malignancies AMG319 Amgen p110 I . Lymphoid malignancies CAL101 Gilead Sciences p110 I, II, and III . INHL (GS101) . CLL . MCL . SLL . Hodgkin lymphoma . Non-Hodgkin lymphoma . Other lymphomas . AML . MM . Hematologic malignancies GS9820 Gilead Sciences p110 and I . Lymphoid malignancies p110

SCCHN, squamous cell carcinoma of the head and neck; ESCC, esophageal squamous cell carcinoma; GIST, gastrointestinal stromal tumor; CRPC, castration-resistant prostate cancer; sqNSCLC, squamous non-small cell lung cancer; TNBC, triple-negative breast cancer; CLL, chronic lymphocytic leukemia; SLL, small lymphocytic leukemia; ALL, acute lymphoblastic leukemia; INHL, indolent non-Hodgkin lymphoma; MCL, mantle cell lymphoma; AML, acute myeloid leukemia; MM, multiple myeloma.

* Data taken from an April 2014 search of http://www.clinicaltrials.gov.

Figure 1.8: Rational combination of PI3K inhibitors and other targeted therapeutics. Pan-PI3K and dual pan-PI3K/mTOR inhibitors are currently being tested in clinical trials (grey box). These agents are being combined with mTOR-selective inhibitors (also in grey box), RAS-RAF-MEK-ERK pathway inhibitors (yellow box), RTK or other membrane-associated protein inhibitors (blue box), hormone signaling inhibitors (purple box), and other agents inhibiting the cell cycle, apoptosis machinery, or other signaling pathways (red box). Colored number symbols indicate targeted therapeutics currently in clinical trials for combination with the designated PI3K inhibitor. Asterisks denote targeted therapeutics expected to cooperate with PI3K therapies based on preclinical studies. For further detail, see Table 1.5.

Table 1.5: Combination of PI3K inhibitors with other targeted therapies in the clinic

Combination therapy trials Agent Company Target Agent Target Tumor types* Clinical trial* Class I pan-PI3K inhibitors BKM120 Novartis Class I PI3Ks . Lapatinib . EGFR/HER2 . Breast NCT01589861 . Fulvestrant . ER NCT01339442 . Trastuzumab . HER2 NCT01132664 . Letrozole . Aromatase NCT01248494 . Gefitinib . EGFR . NSCLC NCT01570296 . Erlotinib . EGFR NCT01487265 . Panitumumab . EGFR . Colorectal NCT01591421 . Cetuximab . EGFR . Head and neck NCT01816984 . Bevacizumab . VEGFR† . GBM NCT01349660 . Renal cell NCT01283048 . INC280 . c-MET . GBM NCT01870726

33 . Rituximab . CD20 . B cell lymphoma NCT02049541

. Imatinib . BCR-ABL . GIST NCT01468688 . Vemurafenib . BRAF . Melanoma NCT01512251 . Encorafenib . BRAF NCT01820364 . Olaparib . PARP . TNBC NCT01623349 . Ovarian . Abiraterone . CYP17 . Prostate NCT01634061 acetate . Erismodegib . Smoothened . Adv. solid tumors NCT01576666 . Trametinib . MEK1/2 NCT01155453 . MEK162 . MEK1/2 NCT01363232 . Everolimus . mTOR NCT01470209 GDC0941 Genentech Class I PI3Ks . Fulvestrant . ER . Breast NCT01437566 . Trastuzumab . HER2 NCT00928330 . Bevacizumab . VEGFR† . Breast NCT00960960 . NSCLC NCT00974584 . Erlotinib . EGFR . Adv. solid tumors NCT00975182 . Cobimetinib . MEK1 NCT00996892

Table 1.5: Combination of PI3K inhibitors with other targeted therapies in the clinic (continued)

Combination therapy trials Agent Company Target Agent Target Tumor types* Clinical trial* Class I pan-PI3K inhibitors PX866 Oncothyreon Class I PI3Ks . Cetuximab . EGFR . Colorectal NCT01252628 . SCCHN NCT01252628 . Vemurafenib . BRAF . Melanoma NCT01616199 XL147 Exelixis/ Class I PI3Ks . Trastuzumab . HER2 . Breast NCT01042925 Sanofi-Aventis . Letrozole . Aromatase NCT01082068 . Erlotinib . EGFR . Adv. solid tumors NCT00692640 . MM121 . HER3 NCT00704392 . XL647 . RTKs NCT01436565 Isoform-selective PI3K inhibitors BYL719 Novartis p110 . Cetuximab . EGFR . SCCHN NCT01602315 . LJM716 . HER3 . ESCC NCT01822613

34 . Encorafenib . BRAF . Colorectal NCT01719380

. Cetuximab . EGFR . Fulvestrant . ER . Breast NCT02088684 . Adv. solid tumors NCT01219699 . Imatinib . BCR-ABL . GIST NCT01735968 . Letrozole . Aromatase . Breast NCT01870505 . Exemestane . Aromatase NCT01870505 . TDM-1 . HER2‡ NCT02038010 . LEE011 . CDK4/6 NCT02088684 . Everolimus . mTOR . Breast NCT02077933 . Exemestane . Aromatase . Kidney . Pancreas . AUY922 . HSP90 . Gastric NCT01613950 . Ganitumab . IGF1R . Adv. solid tumors NCT01708161 . BGJ398 . FGFR NCT01928459 . MEK162 . MEK1/2 NCT01449058 GDC0032 Genentech p110 . Letrozole . Aromatase . Breast NCT01296555 . Fulvestrant . ER

Table 1.5: Combination of PI3K inhibitors with other targeted therapies in the clinic (continued)

Combination therapy trials Agent Company Target Agent Target Tumor types* Clinical trial* Isoform-selective PI3K inhibitors INK1117 Intellikine/ p110 . MLN0128 . mTORC1/2 . Adv. non- NCT01899053 Millenium hematological malignancies SAR260301 Sanofi p110 . Vemurafenib . BRAF . Melanoma NCT01673737 IPI145 Infinity p110 and . Ofatumumab . CD20 . CLL NCT02049515 p110 . SLL . Rituximab . CD20 . Hematologic NCT01871675 malignancies AMG319 Amgen p110 NCT01300026 CAL101 (GS101) Gilead Sciences p110 . Rituximab . CD20 . INHL NCT01088048 . Ofatumumab . CD20 . CLL

35 . Everolimus . mTOR . MCL

. Bortezomib . NFB . GS9973 . SYK . Hematologic NCT01796470 malignancies . Everolimus . mTOR . MCL NCT01088048 Dual pan-PI3K/mTOR inhibitors GDC0980 Genentech PI3K/mTOR . Fulvestrant . ER . Breast NCT01437566 . Abiraterone . CYP17 . Prostate NCT01485861 acetate . Bevacizumab . VEGFR† . Breast NCT01254526 . Adv. solid tumors NCT01332604 PF04691502 Pfizer PI3K/mTOR . PD0325901 . MEK . Adv. solid tumors NCT01347866

Table 1.5: Combination of PI3K inhibitors with other targeted therapies in the clinic (continued)

Combination therapy trials Agent Company Target Agent Target Tumor types* Clinical trial* Dual pan-PI3K/mTOR inhibitors BEZ235 Novartis PI3K/mTOR . Trastuzumab . HER2 . Breast NCT01471847 . Adv. solid tumors NCT01285466 . Everolimus . mTOR . Breast NCT01482156 . Renal cell . Adv. solid tumors NCT01508104 . Abiraterone . CYP17 . Prostate NCT01717898 acetate . Letrozole . Aromatase . Breast NCT01248494 . Everolimus . mTOR . Adv. solid tumors NCT01482156 . MEK162 . MEK NCT01337765 XL765 Sanofi PI3K/mTOR . Letrozole . Aromatase . Breast NCT01082068

36 . Erlotinib . EGFR . Adv. solid tumors NCT00777699

PF05212384 Pfizer PI3K/mTOR . PD0325901 . MEK . Adv. solid tumors NCT01347866 . Cetuximab . EGFR . Colorectal cancer NCT01925274 . Bevacizumab . VEFGR . Colorectal cancer NCT01937715

TNBC, triple-negative breast cancer; GIST, gastrointestinal stromal tumor; NSCLC, non-small cell lung carcinoma; ESCC, esophageal squamous cell carcinoma; SCCHN, squamous cell carcinoma of the head and neck; sqNSCLC, squamous non-small cell lung cancer; TCC, transitional cell carcinoma; INHL, indolent non-Hodgkin lymphoma; MCL, mantle cell lymphoma; CLL, chronic lymphocytic leukemia; SLL, small lymphocytic leukemia; ALL, acute lymphoblastic leukemia; AML, acute myeloid leukemia; CML, chronic myelogenous leukemia; MM, multiple myeloma; CRPC, castration-resistant prostate cancer; GBM, glioblastoma multiforme.

* Data taken from an April 2014 search of http://www.clinicaltrials.gov.

† Bevacizumab is a monoclonal antibody targeting VEGF that prevents signaling through VEFGR.

‡ T-DM1 is a conjugate of the cytotoxic agent mertansine (DM1) to the monoclonal antibody Trastuzumab targeting HER2.

improve therapeutic outcomes.

Emerging isoform-selective PI3K inhibitors

Most PI3K inhibitors in early clinical trials are ATP-competitive agents that target all class I isoforms with similar potencies. These include pan-PI3K inhibitors (Table 1.2) such as GDC0941 (Raynaud et al., 2009) and dual pan-PI3K/mTOR inhibitors (Table

1.3) such as BEZ235 (Maira et al., 2008). Though these drugs display potent preclinical anti-tumor activity, their success in clinical trials as single agents has been modest

(Rodon et al., 2013). The therapeutic window and efficacy of pan-PI3K inhibitors are limited in some cases by adverse effects arising from a broader spectrum of off-target effects (Fruman and Rommel, 2014). Furthermore, while both pan-PI3K and isoform- selective inhibitors have on-target effects from suppression of essential PI3K functions, for example glucose homeostasis, pan-PI3K inhibitors likely have additional on-target effects from inhibiting isoforms that are not contributing to tumorigenesis. Isoform- selective inhibitors may achieve greater efficacy with fewer toxic effects, and are emerging in the clinic (Table 1.4).

The most effective single agent PI3K-based therapy to date is idelalisib (CAL101 or

GS1101), a p110-selective inhibitor. Idelalisib has achieved notable success in early trials for patients with chronic lymphocytic leukemia or indolent lymphoma, and is currently in phase III clinical trials (Furman et al., 2014; Gopal et al., 2014). Interestingly, this dramatic response is not due to genetic activation of the PI3K pathway, as neither

PI3K mutation nor PTEN loss is common in these malignancies. Given the important role of p110 in signaling downstream of BCRs (Clayton et al., 2002; Jou et al., 2002;

Okkenhaug et al., 2002) and the fact that leukemic B cells have been shown to be dependent on BCR signaling, it is likely that idelalisib functions by blocking essential

BCR signals. Two recent articles provide great insight into the success of idelalisib trials

(see (Fruman and Cantley, 2014) and (Vanhaesebroeck and Khwaja, 2014)).

In addition to the role of p110 in B cell malignancies, a recent preclinical study showed that this isoform also contributes to PTEN-null T-ALL (Subramaniam et al., 2012).

However, p110-selective inhibition in this study was insufficient to suppress tumorigenesis; combined inhibition of both p110 and p110 was required for effective anti-PI3K therapy (Subramaniam et al., 2012). The involvement of p110 and p110 in leukocyte signaling and hematological malignancies has drawn great attention, and new inhibitors that target both isoforms simultaneously are in clinical trials for B and T cell lymphomas (Table 1.4). These isoforms may also mediate immune responses that support the growth of solid tumors. In a mouse model, p110 inhibition blocked myeloid cell recruitment to tumors, thus suppressing malignancy by targeting the tumor microenvironment (Schmid et al., 2011). Another study indicated that p110 inhibition impaired tumor growth by disrupting regulatory T cell-mediated immune tolerance (Ali et al., 2014). These findings indicate potential new applications for p110- or p110- selective therapies in cancer.

The frequency of PIK3CA mutations in solid tumors has generated great interest in the potential for p110-selective inhibitors in targeting these cancers. Data presented at the

2013 San Antonio Breast Cancer Symposium (SABCS) indicated promising early clinical activity of p110-selective inhibitors BYL719 or GDC0032 as single agents in patients with PIK3CA-mutant advanced breast tumors (Juric et al., 2013b). Recent preclinical findings that HER2- or KRAS-driven tumors rely on p110 (Castellano et al., 2013;

Gritsman et al., 2014; Gupta et al., 2007; Schmit et al., 2014; Utermark et al., 2012) underscore the need for clinical evaluation of p110-selective drugs in these disease

settings. In these studies, growth of HER2- or KRAS-driven solid tumors is inhibited similarly by pan- and p110-selective inhibitors (Castellano et al., 2013; Utermark et al.,

2012), but only modestly by p110-selective inhibition (Schmit et al., 2014; Utermark et al., 2012). However, further study is needed to determine the contexts in which simultaneous inhibition of p110 and p110 can improve outcomes of KRAS- or HER2- driven disease.

One drawback of p110-selective inhibitors is their inevitable on-target adverse effects on insulin signaling and glucose metabolism, since p110 is the major isoform mediating these functions (Knight et al., 2006; Sopasakis et al., 2010). In the clinic, the effect of p110-selective inhibitors on glucose homeostasis must be carefully managed (Busaidy et al., 2012), and is in some cases limiting (Rodon et al., 2013). To circumvent this, inhibitors are being developed that specifically target p110 harboring hotspot mutations.

Such agents might be used at high doses with low toxicity, similar to mutant-selective

BRAF inhibitors that have had great clinical success (Bollag et al., 2010; Chapman et al.,

2011). A major obstacle to this approach is the heterogeneity of oncogenic PIK3CA mutations. Some progress has been made with the discovery of GDC0032, which was reported at the 2013 SABCS to have enhanced potency in PIK3CA mutant breast cancer models (Sampath et al., 2013); one preclinical study also reported success using stapled peptides to specifically disrupt the interaction of p110-E545K with IRS1 (Hao et al.,

2013). However, devising strategies to selectively interrupt mutant-specific function remains challenging. If developed, this class of inhibitor will likely be most effective in early stage tumors with PIK3CA mutations, as advanced PIK3CA-mutant tumors may have escaped their dependency on oncogenic p110 (Liu et al., 2011). Such drugs would also be ideal for treating congenital overgrowth syndromes caused by PIK3CA mutations occurring during early embryonic development (Kurek et al., 2012; Orloff et

al., 2013; Rios et al., 2013; Riviere et al., 2012). In these contexts, p110 mutant- selective inhibitors may yield improved therapeutic index.

Several preclinical studies have documented that certain PTEN-deficient tumors depend on p110 (Jia et al., 2008; Ni et al., 2012; Wee et al., 2008), prompting a new clinical trial with the p110-selective inhibitor GSK2636771 in patients with PTEN-deficient advanced solid tumors (NCT01458067). However, since PTEN is a negative regulator of

PI3K, isoform-dependency of PTEN-deficient tumors can be complicated as it can be affected by tissue type, co-existing genetic events, and microenvironmental cues that fuel cancer cells. In model systems where PTEN-deficient tumors are found to be dependent on p110, addition of oncogenic RTKs, RAS, or mutant PIK3CA can shift dependency partially or totally to p110 (Figure 1.7). Recent studies also show that prolonged treatment of PTEN-deficient tumor cells with p110-selective inhibitors can shift isoform dependency from p110 to p110 (N. Rosen, unpublished observations).

Therefore in most PTEN-deficient solid tumors, both p110 and p110 should be targeted.

Although development of dual p110/p110-selective inhibitors has proven difficult

(Knight et al., 2006), combination of individual p110- and p110-selective inhibitors might offer flexibility in the dosing of each isoform-selective inhibitor to further reduce toxicity and increase the therapeutic window. One approach could involve continuous inhibition of p110 to suppress elevated basal PIγK activity due to PTEN loss, combined with pulsatile inhibition of p110 to avoid toxicity due to glucose elevation. Such a strategy might also avoid the reported shift in isoform dependency of tumors from p110 to p110 after prolonged treatment with the p110-selective inhibitor BYL719 (J.A.

Engelman, unpublished observations). Ultimately, the success of targeting PI3K in

cancer will likely require better understanding of which PI3K isoforms to target in a given disease setting, improved inhibitors, and more careful dosing strategies.

Resistance mechanisms and combination therapeutic strategies

PI3K-based therapeutic approaches have encountered a number of roadblocks in the form of intrinsic and acquired resistance mechanisms. A large body of work has identified multiple signaling feedback loops, compensatory parallel signaling pathways, and modes of downstream pathway activation that may result in clinical resistance to

PI3K inhibitors. Consequently, combination therapies are being developed and evaluated in both preclinical and clinical settings (Figure 1.8 and Table 1.5), and will be necessary to maximize clinical efficacy of PI3K inhibitors.

The first indication of feedback loops in the PI3K pathway came from experiments with mTOR inhibitors. In early studies mTOR inhibition led to p70 ribosomal protein S6 kinase

(S6K) suppression, IRS1 upregulation, and PI3K-AKT activation (O'Reilly et al., 2006).

This prompted the development of dual pan-PI3K/mTOR inhibitors that are currently in clinical trials (Table 1.3). Interestingly, feedback loops can also arise from dual pan-

PI3K/mTOR inhibition. A recent preclinical report suggested that PI3K and mTOR blockade activated the Janus kinase 2 (JAK2)-signal transducer and activator of transcription 5 (STAT5) signaling axis via IRS1, generating resistance to dual

PI3K/mTOR inhibition, which could be overcome by targeting JAK2 (Britschgi et al.,

2012). Similarly, in another preclinical study treatment with BEZ235 increased phosphorylation of multiple signaling molecules, including STAT3, STAT5, JUN, and p90 ribosomal S6 kinase (p90RSK) (Muranen et al., 2012). Isoform-selective PI3K inhibitors can also generate feedback loops: in a recent study of PIK3CA mutant breast tumors, mTOR complex 1 (mTORC1) reactivation by insulin-like growth factor 1 (IGF1) and

neuregulin 1 (NRG1) was associated with tumor resistance to the p110-selective agent

BYL719, necessitating concurrent mTORC1 inhibition using RAD001 (Elkabets et al.,

2013). Inhibiting both PI3K and mTOR, possibly in conjunction with additional signaling pathways, may be required to achieve effective anti-tumor activity.

Another important resistance mechanism to PI3K pathway inhibition is increased expression of RTKs, such as HER3, IGF1R, insulin receptor (IR), and EGFR, via forkhead box O (FOXO)-mediated transcriptional upregulation (Chandarlapaty et al.,

2011). Robust HER3 induction in response to PI3K inhibition has been reported in several tumor types (Garrett et al., 2011; Muranen et al., 2012; Sergina et al., 2007).

While HER3 itself does not possess strong tyrosine kinase activity, it dimerizes with

EGFR, HER2, or HER4, hyperactivating the PI3K pathway and dampening the efficacy of PI3K drugs. A preclinical study demonstrated that combination of the HER3- neutralizing antibody LJM716 and the p110-selective inhibitor BYL719 potently blocked

PI3K signaling and growth of HER2-positive breast tumor xenografts, even without a direct HER2 antagonist (Garrett et al., 2013). Similarly, combination of the dual

EGFR/HER3 inhibitor MEHD7945A with a PI3K inhibitor (GDC0941) or AKT inhibitor

(GDC0068) effectively blocked the growth of triple-negative breast cancer cells in vitro and in xenografts in a preclinical study (Tao et al., 2014). Blockade of PI3K along with upstream RTKs may therefore circumvent certain PI3K therapy resistance mechanisms

(Figure 1.8).

Activation of convergent signaling pathways, for example the RAS-RAF-MEK-ERK pathway, can also lead to PI3K pathway inhibition resistance. Mutant RAS can activate both the RAF-ERK and PI3K-AKT-mTOR pathways in cancer cells; blocking the PI3K pathway in such cells leads to upregulation of the ERK pathway (Serra et al., 2011).

Inhibition of both PI3K and ERK pathways successfully suppressed the growth of cancer cells in mouse models (Castellano et al., 2013; Engelman et al., 2008; Will et al., 2014), and combinations of MEK inhibitors and pan- or isoform-selective PI3K agents are being evaluated in clinical trials. However, there is preclinical evidence that some of these combinations may be limited due to synergistic toxicity (Castellano et al., 2013).

Preclinical studies indicate that pulsatile inhibition of both PI3K and ERK pathways may provide more effective anti-tumor activity while limiting toxic effects (Will et al., 2014), suggesting that optimization of such combinations in the clinic will require careful dosing strategies.

Another mode of resistance to PI3K-directed therapies arises from the activation of transcription downstream or outside of the PI3K pathway. Several reports have indicated

MYC amplification or overexpression (Ilic et al., 2011; Liu et al., 2011) or activation of the

Notch and WNT/-catenin pathways (Muellner et al., 2011; Tenbaum et al., 2012) as mechanisms of resistance to PI3K inhibition. Recently, the bromodomain and extraterminal (BET) inhibitor JQ1 has been shown to downregulate transcription of MYC, among other targets (Delmore et al., 2011). XAV939 has also been identified as an inhibitor of WNT/-catenin-mediated transcription (Huang et al., 2009). Combination of

PI3K inhibition with these agents is being actively pursued in preclinical settings.

Other combination therapies have been suggested by assessing pathways that may synergize with PI3K (Figure 1.8). As presented at the 2012 and 2013 SABCS, anti- estrogen therapies are being tested in combination with PI3K inhibitors in clinical trials for breast cancer patients (Juric et al., 2012; Juric et al., 2013a; Juric et al., 2013b). In a brain tumor study, coordinate activation of sonic hedgehog (SHH) and PI3K signaling was found in PTEN-deficient glioblastoma; combination of BKM120, a pan-PI3K

inhibitor, and LED225, a smoothened (SMO) inhibitor that blocks SHH signaling, resulted in synergistic anti-tumor effects (Gruber Filbin et al., 2013). Poly-(ADP-ribose) polymerase (PARP) and PI3K inhibitors have been found to cooperate in prostate and triple-negative breast cancers (Gonzalez-Billalabeitia et al., 2014; Ibrahim et al., 2012;

Juvekar et al., 2012). It appears that PI3K inhibition downregulates BRCA1 and BRCA2, impairing homologous recombination and sensitizing BRCA-wildtype cancer cells to

PARP inhibition. Another attractive approach is combination of PI3K-targeted agents with drugs that suppress anti-apoptotic factors. B cell lymphoma 2 (BCL2), myeloid cell leukemia sequence 1 (MCL1), and other pro-survival proteins are frequently upregulated in cancer, and may explain why PI3K inhibition is often cytostatic in tumor cells. BCL2 or

MCL1 suppression may induce cytotoxicity in response to PI3K inhibition (Rahmani et al., 2013). Finally, an emerging approach is to combine PI3K inhibitors with agents that disrupt cell cycle machinery (Vora et al., 2014). The p16-Cyclin D-cyclin-dependent kinase 4 (CDK4)-CDK6 pathway is frequently dysregulated in cancer. A number of

CDK4/CDK6 inhibitors, including LEE011 and palbociclib (PD0332991), are entering clinical trials for combination with pan- or p110-selective inhibitors. Such rational combination therapies will be required to increase the success of PI3K inhibitors.

Conclusions and perspective

Targeting the PI3K pathway remains both an opportunity and a challenge for cancer therapy. Recent advances have provided the framework and rationale for inhibiting select class I PI3K catalytic isoforms. We have learned a great deal about the divergent roles of these isoforms in different signaling contexts, and are beginning to understand the importance of each isoform in various tissues, compartments, and cancer types.

These findings have informed preclinical and clinical studies with isoform-selective PI3K agents, which offer improved specificity and reduced toxicity over first-generation pan-

PI3K drugs. Isoform-selective PI3K inhibitors have seen promising success in early- and late-stage clinical trials for solid and hematological malignancies, highlighting the potential for isoform-selective PI3K therapeutics.

Although we have made substantial progress, further efforts are needed. We have only recently begun to appreciate the importance of class I regulatory isoforms in tumorigenesis. The different ways in which p85 subunits contribute to cancer, and the effective means to pharmacologically inhibit these mechanisms, are still not fully understood. Similarly, while a recent study indicates that class II isoform PI3K-Cβ is important for pathophysiological angiogenesis, the roles of class II and III PI3Ks in cancer remain unclear.

For the class I catalytic isoforms, we must continue to precisely define the disease settings in which different PI3K isoforms will need to be targeted. To better inform isoform-selective therapeutic strategies, a set of biomarkers to predict the active p110 isoforms in a given tumor would be ideal, but development of this will require systematic studies. Continued work to understand the underlying cellular programs that protect tumors with aberrant PI3K activation from PI3K-targeted therapy will also be important.

This will allow for better rational design of combination therapies, which will be necessary to overcome compensatory pathway activation and acquired resistance mechanisms and maximize the anti-tumor activity of PI3K inhibitors. Dosing strategies will also need to be carefully considered, as recent studies suggest that in some cases pulsatile inhibition may reduce toxicity without sacrificing efficacy. Progress in these areas should increase the effectiveness of PI3K-directed therapies in the clinic.

Chapter 2: PI3K regulatory subunit p85alpha plays a tumor

suppressive role in human mammary epithelial cells

Acknowledgements

Hailing Cheng and Jean Zhao conceived the initial project idea. Lauren Thorpe performed all experiments and data analysis.

I would additionally like to thank Hailing Cheng for providing HMECs and plasmids, for teaching initial techniques, and for assistance in troubleshooting experiments. Thank you to Haluk Yuzugullu for assistance in subcloning the neuT construct. Thank you to

Jennifer Spangle for providing detailed protocols for receptor internalization and degradation assays. I would also like to thank Lewis Cantley for helpful discussions conceptualizing the mechanism portion of this work, and Lewis Cantley, Karl Münger, and Myles Brown for helpful discussions on experimental confirmation of this mechanism.

Preface

In this chapter, we use human mammary epithelial cells (HMECs), a normal, non- transformed cell culture system, to examine the in vitro effects of partial p85 loss on

PI3K signaling and cell transformation. We find that RNAi-mediated PIK3R1 knockdown increases growth factor-dependent PIγK signaling through catalytic isoform p110, facilitating cellular transformation. We furthermore demonstrate that knockdown of

PIK3R1 augments PI3K signaling and transformation mediated by oncogenes, including activated HERβ/neu. Finally, we show that partial reduction of p85 leads to an increased amount of p110 bound to activated receptor tyrosine kinases (RTKs).

Together these findings suggest that p85 depletion selectively targets a free negative regulator pool of the PI3K regulatory subunit that fine tunes RTK-mediated PI3K activation under normal conditions, and transforms cells when lost.

Introduction

Class IA phosphatidylinositol 3-kinases (PI3Ks) are critical coordinators of the cellular response to extracellular signals. They are heterodimers comprised of a p110 catalytic subunit (p110, p110, or p110) and a p85 regulatory subunit (p85, p55, p50, p85, or p55, collectively referred to as p85) (Figure 1.1). In quiescent cells, the iSH2 domain of the p85 regulatory subunit binds and stabilizes the p110 catalytic subunit while maintaining p110 in a low-activity state (Yu et al., 1998b). Upon growth factor stimulation, the SH2 domains of p85 bind tyrosine-phosphorylated YXXM motifs on activated RTKs or their adaptors, recruiting p110 to the plasma membrane and simultaneously relieving its inhibition (Yu et al., 1998a) (Figure 1.2). The activated p110 catalytic subunit phosphorylates the γ’-hydroxyl group of phosphatidylinositol 4,5- bisphosphate (PtdIns(4,5)P2) to produce phosphatidylinositol 3,4,5-trisphosphate

(PtdIns(3,4,5)P3), a cellular second messenger which goes on to activate multiple AKT-

dependent and –independent downstream signaling pathways. PI3K signaling controls diverse cellular activities, including cell growth, proliferation, survival, and transformation

(Liu et al., 2009; Wong et al., 2010). The phosphatase and tensin homologue (PTEN) lipid phosphatase removes the γ’ phosphate from PtdIns(γ,4,5)P3 to inactivate PI3K signals.

Several reports have linked p85 to activity or stability of the PTEN lipid phosphatase. In liver lysates from mice with liver-specific Pik3r1 knockout, PtdIns(3,4,5)P3 production was more sustained and AKT activation was increased, while the lipid phosphatase activity of PTEN was found to be reduced, suggesting that p85 may enhance PTEN activity (Taniguchi et al., 2006). This finding was corroborated by in vitro assays, where addition of increasing amounts of purified p85 augmented dephosphorylation of

PtdIns(3,4,5)P3 by PTEN (Chagpar et al., 2010). Furthermore, co-immunoprecipitation experiments have suggested an interaction between p85 and PTEN (Chagpar et al.,

2010; Cheung et al., 2011; Rabinovsky et al., 2009); this interaction was shown to be direct and to require the N-terminal SHγ and BH domains of p85 (Chagpar et al., 2010).

In addition to effects on PTEN lipid phosphatase activity, p85 has been linked to PTEN expression and protein stability. Although liver-specific ablation of Pik3r1 had no effect on the levels of PTEN protein in mouse liver lysates (Taniguchi et al., 2006; Taniguchi et al., 2010), these mice developed liver tumors within 14-20 months, and lysates from these tumors had significantly reduced PTEN protein and mRNA levels (Taniguchi et al.,

2010). In a separate study, ectopic expression of wildtype p85 increased PTEN protein levels, while expression of the endometrial cancer-associated p85-E160* mutant enhanced ubiquitination and proteasomal degradation of PTEN; in these contexts, PTEN mRNA levels were unchanged (Cheung et al., 2011). Together these findings suggest that p85 may affect PTEN expression, stability, or activity, potentially contributing to

negative regulation of the PI3K pathway.

In addition to its reported effects on PTEN, p85 has also been shown to have a

GTPase activating protein (GAP) function towards select Rab proteins. Rabs are a large family of small GTPases critical for the regulation of intracellular trafficking (Jean and

Kiger, 2012; Stenmark, 2009). One mechanism controlling receptor-mediated signaling is the endocytosis of receptors following their activation; receptors are targeted to the early endosome, then either transported to the lysosome for degradation, terminating signaling, or recycled back to the plasma membrane, allowing for potentially sustained signaling (Miaczynska, 2013). Of the many Rab proteins, in particular Rab4 and Rab5 are important for transport from the plasma membrane to early endosomes, and Rab11 is important for recycling of endosomes back to the plasma membrane (Jean and Kiger,

2012; Stenmark, 2009). The N-terminal breakpoint cluster homology (BH) domain of p85 has been reported to directly bind and have GAP activity towards Rab5; in addition, p85 showed GAP activity toward Rab4, but not Rab11 (Chamberlain et al.,

2004). Ectopic expression of a p85 mutant with disrupted Rab-GAP activity led to increased growth factor-stimulated PI3K/AKT activation and cellular transformation

(Chamberlain et al., 2004; Chamberlain et al., 2008), apparently due to more rapid and sustained internalization and reduced degradation of RTKs (Chamberlain et al., 2008;

Chamberlain et al., 2010). Similarly, a recent report demonstrated that RNAi-mediated p85 downregulation lead to an increase in active GTP-bound Rab5 and PI3K/AKT pathway activation (Dou et al., 2013). Together this suggests that p85 might also contribute to regulation of PI3K signaling by effects on activation of Rab proteins controlling receptor trafficking.

Hyperactivation of the PI3K pathway is one of the most common events in human

cancers (Table 1.1 and Appendix A). This can be a result of lesions along the pathway, such as loss of the PTEN phosphatase opposing PI3K (Cully et al., 2006), or of mutations in PI3K itself. Oncogenic mutations frequently occur in the PIK3CA gene encoding catalytic isoform p110 (Cancer Genome Atlas Research, 2008; Parsons et al., 2008; Samuels et al., 2004; Thomas et al., 2007), and confer constitutive p110 activation leading to transformation (Gymnopoulos et al., 2007; Hon et al., 2012; Isakoff et al., 2005; Kang et al., 2005; Mandelker et al., 2009; Samuels et al., 2005; Zhao et al.,

2005). More recently, somatic mutations in PIK3R1 (encoding p85 and its splicing variants p55 and p50) and PIK3R2 (encoding p85) have been reported in glioblastoma (Cancer Genome Atlas Research, 2008; Parsons et al., 2008) and in endometrial cancer (Cheung et al., 2011; Urick et al., 2011). A majority of PIK3R1 mutations occur in iSH2 domain hotspots (Cancer Genome Atlas Research, 2008;

Cheung et al., 2011; Jaiswal et al., 2009; Urick et al., 2011); subsequent studies demonstrated that many of these iSH2 mutants retain the ability to bind and stabilize p110, but are less able to inhibit p110 catalytic activity, resulting in increased PI3K activation and transformation (Cheung et al., 2011; Jaiswal et al., 2009; Urick et al.,

2011; Wu et al., 2009). A smaller fraction of PIK3R1 mutations found in cancers occur in other domains of the protein; interestingly, some of these mutants lack the iSH2 domain and are unable to bind p110 (Cheung et al., 2011; Jaiswal et al., 2009; Urick et al.,

2011). Ectopic expression of the truncation mutant p85-E160* has been reported to contribute to destabilization of PTEN protein (Cheung et al., 2011), but the effects of other truncation mutations has not yet been determined. The discovery and characterization of mutations in PIK3R1 has established a previously unreported role for the p85 regulatory subunit in cancer, potentially by multiple distinct mechanisms.

In addition to oncogenic mutations in PIK3R1, accumulating evidence suggests that

changes in the levels of p85 can have an important effect on PI3K activation. In mice, although complete Pik3r1 ablation is perinatally lethal (Fruman et al., 2000), partial loss of different p85 isoforms improves insulin sensitivity and insulin-stimulated AKT activation (Chen et al., 2004; Mauvais-Jarvis et al., 2002; Terauchi et al., 1999; Ueki et al., 2002a; Ueki et al., 2002b). It has been suggested that in some tissues, the p85 regulatory subunit is present in excess of p110, and that monomeric free p85 is capable of acting as a negative regulator of PI3K signaling (Luo and Cantley, 2005; Mauvais-

Jarvis et al., 2002; Ueki et al., 2002a; Ueki et al., 2003). In support of this model, p85 overexpression in L6 myotubes significantly reduced insulin-stimulated PI3K/AKT activation (Ueki et al., 2000), and mice with p85 overexpression had decreased skeletal muscle insulin signaling (Barbour et al., 2005). Furthermore, monomeric p85 has been shown to downregulate insulin-stimulated PI3K activity by forming a sequestration complex with non-signaling IRS1 adaptors (Luo et al., 2005a). Loss of p85 has also recently been proposed to play a role in cancer. Significant PIK3R1 underexpression was recently detected in breast cancers, and correlated with poorer metastasis-free survival (Cizkova et al., 2013). In mice, liver-specific Pik3r1 deletion eventually led to development of aggressive high-grade hepatocellular carcinoma (HCC)-like tumors with upregulated AKT activation (Taniguchi et al., 2010). Together these studies indicate that partial reduction of p85 can upregulate PI3K in response to insulin, and that furthermore loss of p85 may contribute to tumorigenesis in the breast and liver.

In this chapter, we examine the importance of p85 levels in regulating PIγK/AKT activation in mammary epithelial cells. We use RNAi techniques to downregulate

PIK3R1 expression in human mammary epithelial cells, and explore the effects on

PI3K/AKT signaling and cellular transformation. We then use isoform-selective pharmacological inhibitors to identify the PI3K catalytic isoforms that contribute to

signaling in the context of reduced p85. Finally, we use immunoprecipitation experiments to identify a potential mechanism linking the levels of p85 to the magnitude of RTK-mediated PI3K/AKT signaling activation. Our findings support previous reports that p85 levels modulate PIγK output, and importantly demonstrate that a reduction in p85 is sufficient to transform mammary epithelial cells in vitro.

Results

PIK3R1 expression is significantly reduced in breast cancer

To determine whether p85 may play a tumor suppressive role in breast cancer, we analyzed expression levels of the PIK3R1 gene encoding p85 in different publically available datasets from breast cancer patients. We used the cBioPortal for Cancer

Genomics (http://www.cbioportal.org, (Cerami et al., 2012; Gao et al., 2013)) to query data from a 2012 comprehensive study of human breast tumors by The Cancer Genome

Atlas Network (TCGA) (Cancer Genome Atlas, 2012) and an additional provisional

TCGA data set for copy number loss (as determined by GISTIC) or mutation of PIK3R1.

Such alterations in PIK3R1 occurred in 23% and 28% of breast cancer cases in these studies respectively, with the vast majority of these being heterozygous loss of the gene

(Figure 2.1 A). We also used Oncomine, an online database compiling expression data from thousands of microarray experiments (https://www.oncomine.org, (Rhodes et al.,

2007; Rhodes et al., 2004)), to analyze PIK3R1 expression in other breast cancer datasets. Across multiple studies, PIK3R1 mRNA expression was significantly reduced by 50-77% in breast cancer samples when compared to normal breast tissue (Figure

2.1 B and Table 2.1). Together these results indicate that in breast cancers, PIK3R1 is consistently decreased at both the genomic and mRNA expression levels, suggesting that reduced levels of p85 may play a functional role in tumorigenesis of this tissue.

Figure 2.1: PIK3R1 expression is significantly reduced in breast cancers. Publically available datasets were queried for changes in PIK3R1 associated with breast cancer. (A) The cBio Portal for Cancer Genomics (http://www.cbioportal.org) was used to determine the incidence of PIK3R1 copy number loss or mutation in both a comprehensive 2012 study and an additional provisional study of breast cancer from The Cancer Genome Atlas Network (TCGA). In the 2012 study, such alterations occurred in 23% (111/482) of cases; in the provisional study, these alterations occurred in 28% (271/962) of cases. (B) The Oncomine database (https://www.oncomine.org/) was used to determine changes in PIK3R1 expression in breast cancer as compared to normal breast tissue. Microarray data from representative studies was converted to raw expression levels by taking the inverse log2, and then normalized to the mean raw expression level of normal breast tissue in that specific study. Means ± SEM are shown. Statistical significance was determined using unpaired t-test. For more detail on individual studies, see Table 2.1. *, P < 0.05; **, P < 0.01; ****, P < 0.0001.

Table 2.1: PIK3R1 expression is significantly reduced in breast cancers across multiple microarray datasets

Study name Publication Control tissue Control N Cancer tissue Cancer N % reduction p value in PIK3R1

Sorlie Breast (Sorlie et al., 2001) Normal breast 4 Ductal breast 65 50.80% 0.0160 carcinoma

Sorlie Breast 2 (Sorlie et al., 2003) Normal breast 4 Ductal breast 81 50.06% 0.0079 carcinoma

Richardson Breast 2 (Richardson et al., Normal breast 7 Ductal breast 40 51.77% 0.0186 2006) carcinoma

Ma Breast 4 (Ma et al., 2009) Normal breast 28 Invasive ductal 18 77.34% 0.0034 breast carcinoma 55

TCGA Breast TCGA Network Normal breast 61 Invasive ductal 389 66.71% < 0.0001 2011, no associated breast carcinoma publication

The Oncomine database version 4.4.4.3 (https://www.oncomine.org) was queried for expression of PIK3R1 in breast cancer microarray studies. The data from each study was converted to raw expression levels of PIK3R1 by taking the inverse log2, and then normalized to the mean raw PIK3R1 expression level of the normal breast tissue from that specific study. Unpaired t-test was performed using GraphPad Prism 6.0 to determine p values for expression in cancer tissue as compared to normal breast.

RNAi-mediated PIK3R1 knockdown transforms human mammary epithelial cells

To study whether partial loss of p85 could contribute to transformation, we chose to use human mammary epithelial cells (HMECs), a well-established system for the study of

PI3K-mediated transformation in vitro. In this system, HMECs are immortalized by expression of the human telomerase reverse transcriptase (hTERT) catalytic subunit; these cells have also been characterized to lack functional cyclin-dependent kinase 4 inhibitor A (p16INK4A) and have increased MYC expression (Romanov et al., 2001;

Utermark et al., 2007; Wang et al., 2000). HMECs expressing a dominant negative p53 mutant (DDp53) are unable to form colonies in agar (Zhao et al., 2003; Zhao et al.,

2005), an experimental measure of the transformation hallmark of anchorage- independent growth (Hanahan and Weinberg, 2011). Activation of PI3K by expression of myristoylated p110 or p110, or mutant alleles of p110 found in cancers, transforms these cells (Zhao et al., 2003; Zhao et al., 2005).

To address whether decreased PIK3R1 expression might play a role in the transformation of mammary tissue, we generated polyclonal DDp53-HMEC lines with stable RNAi-mediated knockdown of PIK3R1 (Figure 2.2 A). Two different shRNAs targeting PIK3R1 reduced p85 protein levels by 76.6 ± 0.7% and 77.3 ± 2.9% in comparison to the shControl (N = 3 for all) (Figure 2.2 B). While the positive control cells expressing oncogenic p110-H1047R (Zhao et al., 2003) showed dramatically increased activation of AKT even after 4 hours of serum and growth factor starvation, with a fold increase of 35.1 ± 9.0 in phosphorylation of AKT at S473 as compared to the shControl line, PIK3R1 knockdown only slightly increased AKT S473 phosphorylation, by a fold increase of 2.7 ± 0.4 and 4.0 ± 0.9 (N = 3 for all) (Figure 2.2 D). Although this increase was statistically significant, it is clear that partial loss of p85 does not increase constitutive PI3K pathway activation under un-stimulated conditions to a similar extent

Figure 2.2: Generation of DDp53-HMECs with stable RNAi-mediated PIK3R1 knockdown. Human mammary epithelial cells (HMECs) expressing a dominant negative p53 mutant (DDp53) were infected with lentivirus containing a negative control scrambled shRNA or one of two distinct shRNAs targeting PIK3R1, or with retrovirus containing the cancer-associated p110-H1047R mutant as a positive control. Following selection for stable polyclonal lines, cells were starved of serum and growth factors for 4 hours; protein lysates were collected and subjected to immunoblotting for PI3K pathway components. Membranes were stripped and re-probed for total proteins and vinculin as a loading control. (A) Representative immunoblot. (B-D) Bands from immunoblots of three independent sets of protein lysates were quantified by densitometry; protein levels were normalized first to the vinculin loading control (for p85 and PTEN) or the total unphosphorylated protein (for phospho-AKTS473), then to the corresponding mean value for the shControl. Means ± SEM are shown; N = 3 for each. Statistical significance was determined by unpaired t-test. Significance for comparison to the shControl is shown. *, P < 0.05; **, P < 0.01; ****, P < 0.0001; ns, not significant.

as oncogenic mutant p110.

Next we investigated whether PIK3R1 knockdown was sufficient to induce transformation of DDp53-HMECs. It has previously been shown that DDp53-HMECs are unable to form colonies in agar, but become transformed upon activation of PI3K (Zhao et al., 2006; Zhao et al., 2003). We plated single-cell suspensions of our HMEC lines in fully supplemented growth medium containing 0.3% agar and allowed them to grow for 4 weeks before scoring colonies. Consistent with previous reports, the shControl line was unable to form colonies in agar, with only 16 ± 0.6 colonies per 50,000 cells plated, while the positive control line expressing p110-H1047R formed abundant colonies, with 909.7

± 35.0 colonies per 50,000 cells plated (N = 3 for both) (Figure 2.3 A). In comparison, both shPIK3R1 lines readily formed colonies when grown in agar, with 633.7 ± 7.9 and

1,229 ± 39.0 colonies per 50,000 cells plated, respectively (N = 3 for both) (Figure 2.3

A). These results demonstrate that PIK3R1 knockdown is sufficient to transform DDp53-

HMECs in vitro.

To further assess the effects of PIK3R1 knockdown on PI3K signaling, we evaluated activation of downstream effectors of this pathway in response to acute growth factor stimulation. Cells were synchronized via a 4 hour serum and growth factor starvation, and then stimulated with 20ng/ml EGF for timepoints up to 60 minutes; protein lysates were collected and subjected to immunoblotting for phosphorylation at activation sites of

AKT and S6 ribosomal protein. Compared to the control line, PIK3R1 knockdown lines exhibited both increased and sustained phosphorylation of AKT at activating residues

T308 and S473, and of S6 ribosomal protein at S235/236 (Figure 2.3 B). Comparable increases in EGF-stimulated PI3K/AKT pathway activation were seen in DDp53-HMECs expressing other shRNAs targeting PIK3R1, and a similar effect was seen upon

Figure 2.3: PIK3R1 knockdown transforms DDp53-HMECs and increases growth factor-stimulated PI3K/AKT activation. (A) DDp53-HMECs with stable expression of shControl, shPIK3R1, or p110-H1047R were tested for anchorage-independent growth. Single-cell suspensions were plated in 0.3% agar and grown for 4 weeks. 50,000 cells were plated per 6cm plate. The number of colonies was counted for three independent experiments of at least 3 plates each. Means ± SEM are shown; N = 3 for each group. Statistical significance was determined by unpaired t-test. Significance for comparison to the shControl is shown. ****, P < 0.0001. (B) The indicated cell lines were starved of serum and growth factors for 4 hours and then stimulated with 20ng/ml EGF for the indicated amounts of time. Protein lysates were collected and subjected to immunoblotting. Membranes were stripped and re-probed for total proteins and vinculin as a loading control. Bands were quantified by densitometry and normalized first to the corresponding total protein, then to the most intense band for the shControl. One representative experiment is shown.

stimulation with insulin (data not shown). These results indicate that partial reduction of p85 increases PIγK/AKT signaling in DDp5γ-HMECs, and that this increase is most pronounced upon stimulation with growth factors.

To establish that the increased PI3K signaling seen in our PIK3R1 knockdown DDp53-

HMEC lines was specifically from decreased p85 expression and not an off-target effect of stable shRNA introduction, we sought to rescue p85 protein levels in our knockdown lines. For shRNAs targeting the γ’ UTR of PIK3R1, including shPIK3R1 #1, a wildtype flag-tagged PIK3R1 construct was used; for shRNAs targeting the coding region of PIK3R1, including shPIK3R1 #2, we introduced silent wobble mutations rendering resistance to specific shRNAs into this wildtype construct. Stable expression of these constructs in our PIK3R1 knockdown DDp53-HMECs restored p85 protein expression to levels comparable to the shControl line (Figure 2.4 and data not shown). We then evaluated PI3K/AKT pathway activation in shControl, shPIK3R1, and rescue cell lines in response to acute growth factor stimulation. Cells were starved for 4 hours and then stimulated with 20ng/ml EGF for timepoints up to one hour; protein lysates were collected and subjected to immunoblotting for PI3K/AKT pathway components. As expected, compared to the shControl line, shPIK3R1 cells exhibited increased and sustained phosphorylation of AKT and S6 ribosomal protein; rescue of p85 protein levels in our shPIK3R1 cells reduced AKT and S6 phosphorylation to levels comparable to the shControl line (Figure 2.4). Similar results were seen with other shPIK3R1 rescue

DDp53-HMECs (data not shown). This data indicates that the augmented PI3K/AKT signaling seen in our knockdown lines is likely due to decreased expression of p85.

PIK3R1 knockdown augments HMEC transformation mediated by oncogenes

Because the increased PI3K/AKT signaling in PIK3R1 knockdown DDp53-HMECs was

Figure 2.4: Augmented PI3K/AKT activation in PIK3R1 knockdown DDp53-HMECs is rescued by ectopic expression of PIK3R1. Silent wobble mutations were introduced into a wildtype, flag-tagged PIK3R1 construct to specifically render resistant to shPIK3R1 #2. This construct was stably introduced into DDp53-HMECs expressing shPIK3R1 #2 to rescue expression levels of p85. The cell lines indicated were starved of serum and growth factors for 4 hours and then stimulated with 20ng/ml EGF for timepoints up to one hour. Protein lysates were collected and subjected to immunoblotting. Membranes were stripped and re-probed for total proteins and vinculin as a loading control. Bands were quantified by densitometry and normalized first to the corresponding total protein, then to the most intense band for the shControl. One representative experiment is shown.

dependent on growth factors to initiate PI3K signaling, we were interested to know whether reduced p85 could also cooperate with PIγK-activating oncogenes to augment

PI3K output and transformation. To address this, we generated DDp53-HMECs stably expressing neuT, an activated rat form of HER2/neu, in addition to shRNA targeting

PIK3R1 (Figure 2.5 A). Similar to the DDp53-HMEC lines, shRNAs targeting PIK3R1 reduced p85 protein levels in neuT-expressing lines by 80.3 ± 0.4% and 70.6 ± 3.9% compared to the shControl (N = 3 for all) (Figure 2.5 B). Expression of neuT increased

PI3K/AKT pathway activation even following overnight starvation, and as expected, expression of p110-H1047R further augmented AKT activation as assessed by S473 phosphorylation by 3.8 ± 0.5 fold, as compared to the shControl; both PIK3R1 knockdown lines also exhibited increased AKT S473 phosphorylation over the shControl line, by 1.6 ± 0.1 and 1.6 ± 0.2 fold respectively (N = 3 for all) (Figure 2.5 D). These results suggest that partial p85 loss may be able to cooperate with activated HER2/neu to increase PI3K/AKT activation.

To study the potential synergy between activated HERβ/neu and reduced p85, we examined whether PIK3R1 knockdown could increase the in vitro transformation of

DDp53-HMECs by neuT. We plated single-cell suspensions of these HMEC lines in fully supplemented growth medium containing 0.3% agar and allowed them to grow for 3 weeks before scoring colonies. While the neuT shControl line was able to form colonies in agar, with 821.1 ± 69.2 colonies per 25,000 cells plated, additional expression of p110-H1047R increased colony formation, with 1,330 ± 190.3 colonies per 25,000 cells plated (N = 4 for both) (Figure 2.6 A). Similarly, both neuT shPIK3R1 lines exhibited significantly augmented colony formation over the neuT shControl line, with 1,227 ± 79.3 and 1,321 ± 125.6 colonies per 25,000 cells plated, respectively (N = 4 for both) (Figure

2.6 A). These results indicate that partial p85 loss can cooperate with oncogenic

Figure 2.5: Generation of DDp53-HMECs with activated HER2/neu and RNAi- mediated PIK3R1 knockdown. DDp53-HMECs stably expressing neuT were infected with lentivirus containing a negative control scrambled shRNA or one of two distinct shRNAs targeting PIK3R1, or with retrovirus containing the cancer-associated p110- H1047R mutant as a positive control. Following selection for stable polyclonal lines, cells were starved of serum and growth factors overnight; protein lysates were collected and subjected to immunoblotting for PI3K pathway components. Membranes were stripped and re-probed for total proteins and vinculin as a loading control. (A) Representative immunoblot. (B-D) Bands from immunoblots of three independent sets of protein lysates were quantified by densitometry; protein levels were normalized first to the vinculin loading control (for p85 and PTEN) or the total unphosphorylated protein (for phospho- AKTS473), then to the corresponding mean value for the shControl. Means ± SEM are shown; N = 3 for each. Statistical significance was determined by unpaired t-test. Significance for comparison to the shControl line is shown. *, P < 0.05; **, P < 0.01; ***, P < 0.001; ****, P < 0.0001; ns, not significant.

Figure 2.6: PIK3R1 knockdown increases transformation and PI3K/AKT signaling driven by activated HER2/neu in DDp53-HMECs. (A) DDp53-HMECs with stable expression of neuT and shControl, shPIK3R1, or p110-H1047R were tested for anchorage-independent growth. Single-cell suspensions were plated in 0.3% agar and grown for 3 weeks. 25,000 cells were plated per 6cm plate. The number of colonies was counted for four independent experiments of at least 3 plates each. Means ± SEM are shown; N = 4 for each group. Statistical significance was determined by unpaired t-test. Significance for comparison to the shControl is shown. *, P < 0.05; **, P < 0.01. (B) The indicated cell lines were starved of serum and growth factors overnight and then stimulated with 20ng/ml EGF for the indicated amounts of time. Protein lysates were collected and subjected to immunoblotting. Membranes were stripped and re-probed for total proteins and vinculin as a loading control. Bands were quantified by densitometry and normalized first to the corresponding total protein, then to the most intense band for the shControl line. One representative experiment is shown.

HER2/neu to increase transformation of DDp53-HMECs.

We also assessed the ability of reduced p85 expression to increase PIγK/AKT signaling in DDp53-HMECs with activated HER2/neu. DDp53-HMECs expressing neuT and control or PIK3R1-targeting shRNAs were synchronized via overnight serum and growth factor starvation, then stimulated with 20ng/ml EGF for timepoints up to 24 hours; protein lysates were collected and subjected to immunoblotting for phosphorylation of activation sites on AKT and S6 ribosomal protein. Compared to the shControl cells,

DDp53-HMECs expressing neuT and shPIK3R1 had increased phosphorylation of AKT at both T308 and S473, and of S6 at S235/236, throughout the timecourse of EGF stimulation (Figure 2.6 B). Similar results were seen with DDp53-HMECs expressing neuT and other shRNAs targeting PIK3R1 (data not shown).Together this data indicates that reduced p85 synergistically augments PIγK signaling driven by the pathway- activating oncogenic HER2/neu in starved or growth factor-stimulated conditions.

To further explore the ability of partial p85 loss to synergize with oncogenes common to breast cancer, we generated DDp53-HMECs stably expressing oncogenic p110-

H1047R in addition to stable knockdown of PIK3R1. Compared to DDp53-HMECs expressing p110-H1047R and shControl, both PIK3R1 knockdown cell lines formed approximately 4 fold more colonies when grown in 0.3% agar for 3 weeks (Figure 2.7

A). We also examined PI3K/AKT activation in these cells in response to EGF stimulation following overnight starvation. In DDp53-HMECs expressing p110-H1047R, PIK3R1 increased phosphorylation of AKT at both T308 and S473 in comparison to the shControl (Figure 2.7 B). Together this data demonstrates that partial loss of p85 can further augment transformation mediated by PI3K-activating oncogenes clinically relevant to breast cancer, including HERβ/neu and p110-H1047R.

Figure 2.7: PIK3R1 knockdown increases transformation and PI3K/AKT signaling driven by p110-H1047R in DDp53-HMECs. (A) DDp53-HMECs with stable expression of p110-H1047R and shControl or shPIK3R1 were tested for anchorage-independent growth. Single-cell suspensions were plated in 0.3% agar and grown for 3 weeks. 25,000 cells were plated per 6cm plate. The number of colonies was counted for one independent experiment of at least 3 plates per cell line. Means ± SD are shown; N ≥ γ for each. (B) The indicated cell lines were starved of serum and growth factors overnight and then stimulated with 20ng/ml EGF for the indicated amounts of time. Protein lysates were collected and subjected to immunoblotting. Membranes were stripped and re- probed for vinculin as a loading control.

Transformation driven by PIK3R1 knockdown is mediated by signaling through p110

The PI3K catalytic isoforms play divergent roles in physiological and pathophysiological signaling. Knowing which isoforms are functionally critical in different contexts can inform future clinical use of isoform-selective inhibitors. Accordingly, we used pan-PI3K and isoform-selective inhibitors to determine the contribution of different p110 isoforms to the transformation of HMECs mediated by PIK3R1 knockdown. We plated DDp53-HMECs expressing control or PIK3R1-targeting shRNAs in 0.3% agar, and grew cells for 4 weeks in the presence of either pan-PI3K or isoform-selective PI3K inhibitors. Either the pan-PI3K inhibitor GDC0941 (Folkes et al., 2008) or the p110-selective inhibitor

BYL719 (Furet et al., 2013) effectively inhibited colony formation of PIK3R1 knockdown lines in a dose-dependent manner (Figure 2.8 A-B). Treatment with the p110-selective inhibitor TGX221 did not have a substantial effect on colony growth, even at doses well above the IC50 of this compound (Jackson et al., 2005) (Figure 2.8 C). These results suggest that transformation of DDp53-HMECs with partial p85 is mediated by PIγK signaling through p110 and not p110.

While it has been demonstrated that similar to our findings with PIK3R1 knockdown

DDp53-HMECs, HER2/neu-driven transformation is dependent on p110 (Utermark et al., 2012), recent work has also demonstrated that the presence of co-existing oncogenic events can shift PI3K isoform dependency (Schmit et al., 2014). Therefore, we were interested to determine the isoform dependency of HER2/neu-driven transformation in the context of partial p85 loss. We plated single-cell suspensions of

DDp53-HMECs with stable expression of neuT and either shControl, shPIK3R1, or p110-H1047R in 0.3% agar and allowed cells to grow for 3 weeks with different PI3K inhibitors. Treatment with the pan-PIγK inhibitor GDC0941 or the p110-selective inhibitor BYL719 effectively inhibited colony formation of these cells (Figure 2.9 A-B),

Figure 2.8: Transformation of PIK3R1 knockdown DDp53-HMECs is blocked by p110-selective pharmacological inhibition. DDp53-HMECs with stable expression of shControl or shPIK3R1 were plated as single-cell suspensions in 0.3% agar and grown for 4 weeks. 9,000 cells were plated per well of a 12-well plate. Fresh growth medium containing either the pan-PI3K inhibitor GDC0941 (A), the p110-selective inhibitor BYL719 (B), or the p110-selective inhibitor TGX221 (C) was given every three days. Means ± SD are shown for triplicate wells from representative experiments.

Figure 2.9: Transformation of PIK3R1 knockdown DDp53-HMECs with activated HER2/neu is blocked by p110-selective inhibition. DDp53-HMECs with stable expression of neuT and shControl, shPIK3R1, or p110-H1047R were plated as single- cell suspensions in 0.3% agar and grown for 3 weeks. 4,500 cells were plated per well of a 12-well plate. Fresh growth medium containing either the pan-PI3K inhibitor GDC0941 (A), the p110-selective inhibitor BYL719 (B), or the p110-selective inhibitor KIN193 (C) was given every three days. Means ± SD are shown for triplicate wells from representative experiments.

while treatment with the p110-selective inhibitor KIN193 did not substantially effect colony growth (Figure 2.9 C), even at concentrations that completely block AKT activation in PTEN-null cancer cell lines (Ni et al., 2012). Consistent with these findings, either GDC0941 or BYL719 blocked PI3K/AKT pathway activation in DDp53-HMECs expressing neuT and shPIK3R1 in a dose-dependent manner, while KIN193 only slightly reduced AKT and S6 phosphorylation at the highest doses (Figure 2.10). We would note that these inhibitor experiments are missing an important positive control for the p110 inhibitors: to demonstrate that TGX221 and KIN193 are acting as expected, we plan to carry out similar experiments using the PTEN null breast cancer cells HCC70, MDA-MB-

468, and BT-549; signaling in and transformation of these cells has been shown to be largely reliant on p110 (Jia et al., 2008; Ni et al., 2012; Torbett et al., 2008; Wee et al.,

2008). Nonetheless, the results presented here indicate that the augmented PI3K/AKT signaling in and transformation of PIK3R1 knockdown DDp53-HMECs with or without expression of neuT is primarily mediated by p110.

PIK3R1 knockdown does not affect PTEN levels or lipid phosphatase activity

Next we were interested in determining the molecular mechanism by which decreased p85 levels could increase PIγK/AKT output and mediate transformation. Recent studies have reported that p85 may directly bind PTEN (Chagpar et al., 2010; Rabinovsky et al., 2009) and increase its phosphatase activity (Chagpar et al., 2010); this binding was found to be dependent on EGF stimulation in one study (Chagpar et al., 2010). Other publications have suggested that p85 may be important for PTEN stability at the mRNA or protein level (Cheung et al., 2011; Taniguchi et al., 2010). We hypothesized that if p85 had a stabilizing or activating function on this negative regulator of the PIγK/AKT pathway, then partial p85 loss may lead to reduced levels or activity of PTEN, thereby increasing PI3K/AKT signaling.

Figure 2.10: PI3K/AKT signaling in PIK3R1 knockdown DDp53-HMECs with activated HER2/neu is blocked by p110-selective inhibition. DDp53-HMECs stably expressing neuT and shPIK3R1 #2 were starved of serum and growth factors overnight, then stimulated with starvation medium containing 20ng/ml EGF and the indicated concentrations of PI3K inhibitors for 15 minutes. Protein lysates were collected and subjected to immunoblotting. Membranes were stripped and re-probed for total proteins and vinculin as a loading control. Bands were quantified by densitometry and normalized first to the corresponding total protein, then to the corresponding band for the DMSO control. One representative experiment is shown.

Figure 2.11: Endogenous p85 and PTEN do not appear to interact in DDp53- HMECs. DDp53-HMECs with stable expression of scrambled control or PIK3R1- targeting shRNAs were starved of serum and growth factors for 4 hours and then given either fresh starvation medium or 20ng/ml EGF for the indicated amounts of time. Protein lysates were collected and subjected to immunoprecipitation reactions for either endogenous PTEN (A) or endogenous p85 (B). Samples were then analyzed by immunoblotting. Membranes were stripped and re-probed for vinculin as a loading control.

We first sought to determine whether an interaction between endogenous p85 and

PTEN proteins occurred in our DDp53-HMECs. We synchronized shControl and shPIK3R1 DDp53-HMECs by a 4 hour serum and growth factor starvation, stimulated half of the cells with 20ng/ml EGF, collected protein lysates, and carried out immunoprecipitation reactions for PTEN. Although we were able to achieve reliable immunoprecipitation of endogenous PTEN from both cell lines, we could not detect any evidence of p85 co-immunoprecipitation either under starved or stimulated conditions

(Figure 2.11 A). We also carried out similar experiments with immunoprecipitation of p85 from synchronized or EGF-stimulated DDp53-HMECs. We consistently immunoprecipitated endogenous p85 from shControl and shPIK3R1 cell lines, but while we observed co-immunoprecipitation of endogenous p110 in these experiments, we did not detect endogenous PTEN in immunoprecipitations from either starved or stimulated cells (Figure 2.11 B). Despite altering a number of conditions for these immunoprecipitation reactions, including the antibodies and buffers used, similar results were seen across multiple experiments (data not shown). Thus we could not positively demonstrate an interaction between endogenous p85 and PTEN in our HMECs.

We considered that a p85-PTEN interaction might be easier to demonstrate in certain cell types over others. Neither of the publications showing this interaction used HMECs; however, both studies used HeLa cells among other cell types in their immunoprecipitation experiments (Chagpar et al., 2010; Rabinovsky et al., 2009). We therefore used HeLa cells in addition to wildtype HMECs without p53 inactivation. These cell lines were starved of serum and growth factors for 4 hours, and then half of the cells were stimulated with 20ng/ml EGF. Protein lysates were collected and subjected to immunoprecipitation for endogenous p85. Although p85 and p110 were reliably co- immunoprecipitated from both cell lines, PTEN was not detected in these samples under

Figure 2.12: PTEN and p85 do not appear to interact in a variety of cell types. (A) HeLa cells and HMECs without p53 inactivation were starved of serum and growth factors for 4 hours before being stimulated with 20ng/ml EGF for the indicated amounts of time. Protein lysates were collected and used in immunoprecipitation reactions for endogenous p85 or PTEN. (B) 293T cells were either mock transfected or transiently transfected with wildtype PIK3R1. Protein lysates were collected and subjected to immunoprecipitation for PTEN or p85. Membranes were stripped and re-probed for vinculin as a loading control.

either starved or stimulated conditions (Figure 2.12 A). Based on these results we were unable to conclude that p85 and PTEN interact in HeLa cells.

Finally, we speculated that the amount of interacting p85 and PTEN might be a small fraction of the total amount of these proteins, potentially making it difficult to detect endogenous co-immunoprecipitation. We therefore transiently transfected 293T cells with a wildtype PIK3R1 construct, collected protein lysates from these cells, and subjected them to immunoprecipitation for either PTEN or p85. Compared to the mock transfected cells, transient transfection with PIK3R1 led to substantial overexpression of p85 (Figure 2.12 B). Although immunoblots revealed a small amount of p85 present in PTEN immunoprecipitates, a comparable amount of p85 was detected in the beads only control, suggesting that when p85 was highly overexpressed it bound at a low, non-specific level to the beads used for immunoprecipitation (Figure 2.12 B). Similar results were seen in multiple independent experiments (data not shown). Consistent with this conclusion, we did not observe any co-immunoprecipitated PTEN along with p85 in

PIK3R1-transfected 293T cells, despite achieving co-immunoprecipitation of both p110 and p110 (Figure 2.12 B). Taken together, we were unable to satisfactorily demonstrate an interaction between p85 and PTEN in a variety of cell types and conditions.

While we were unable to positively shown an interaction between p85 and PTEN in our experiments, it has been also been reported that mutation or loss of PIK3R1 might destabilize PTEN, either at the level of transcription or translation (Cheung et al., 2011;

Taniguchi et al., 2010). To explore this possibility, we first examined PTEN protein levels in our HMECs with PIK3R1 knockdown. We collected three independent sets of protein lysates from cells synchronized by serum and growth factor starvation, subjected them

to immunoblotting for PTEN, and used densitometry to quantify the intensity of each band relative to the vinculin loading control. In DDp53-HMECs, stable PIK3R1 knockdown did not have a consistent effect on the steady-state PTEN levels; only one of the shPIK3R1 lines had significantly reduced PTEN protein levels, of 29.8 ± 4.1% compared to the shControl (N = 3 for both) (Figure 2.2 C). We also assessed PTEN protein levels in our neuT-expressing DDp53-HMECs: compared to the shControl, only one shPIK3R1 line showed a significant reduction in PTEN protein levels of 32.9 ± 4.0%; the p110-H1047R line also had a significant reduction of 17.1 ± 1.7% in PTEN levels (N

= 3 for all) (Figure 2.5 C). These results suggest that in HMECs, partial reduction of p85 does not necessarily lead to destabilization of PTEN protein.

We also sought to confirm that PTEN transcripts were not affected by PIK3R1 knockdown in our HMEC lines. We used quantitative PCR (qPCR) to assess the levels of PTEN mRNA in extracts from DDp53-HMECs stably expressing shControl or shPIK3R1. As an additional control, we also examined mRNA levels of the INPP4B phosphatase; INPP4B is a lipid phosphatase that removes phosphorylation of the 4’ hydroxyl group of PtdIns(3,4)P2, and has been identified as a tumor suppressor in breast cancer in a screen using HMECs (Gewinner et al., 2009). Compared to the shControl line, both shPIK3R1 lines had somewhat increased levels of INPP4B transcripts (Figure

2.13 A). In addition, while one shPIK3R1 line had somewhat decreased PTEN transcripts compared to the shControl line, the other shPIK3R1 line showed slightly elevated PTEN mRNA levels (Figure 2.13 A). Together these qPCR results suggest that

PIK3R1 expression is not a reliable predictor of the mRNA levels for either the INPP4B or PTEN tumor suppressor genes.

Although we were unable to reproduce a p85-PTEN interaction in our experiments, and

Figure 2.13: PIK3R1 knockdown does not affect PTEN mRNA levels or lipid phosphatase activity in DDp53-HMECs. (A) Quantitative PCR (qPCR) was used to determine the levels of INPP4B and PTEN mRNA in extracts from DDp53-HMECs stably expressing scrambled control or PIK3R1-targeting shRNAs. Means ± SD are shown for one experiment performed in triplicate. Levels were normalized to the corresponding mean for shControl. (B-C) In vitro lipid phosphatase assays for PTEN immunoprecipitated from shControl or shPIK3R1 DDp53-HMECs. A malachite green reagent was used to detect phosphate in prepared standards (B) or released upon incubation of PtdIns(3,4,5)P3 substrate with PTEN immunoprecipitates. Three negative controls were used: PIP3 Only, assay performed using substrate but no source of PTEN; PTEN Only, assay performed using PTEN immunoprecipitated from shControl cells but with no substrate added; and Beads Only, assay performed using precipitates from shControl cells in which beads but no PTEN antibody was used. The amount of phosphate released in in vitro assays was interpolated from comparison to the phosphate standards. For (B), means ± SD are shown for standards in triplicate. For (C), means ± SEM are shown for assays performed using three independent immunoprecipitation reactions per group. Statistical analysis was performed using unpaired t-test as compared to the shControl. *, P < 0.05; ns, not significant.

we did not find a consistent effect of PIK3R1 knockdown on the levels of PTEN mRNA or protein, the possibility remained that partial p85 loss might affect PTEN lipid phosphatase activity. To assess PTEN activity in our HMEC lines, we carried out PTEN immunoprecipitations and performed in vitro lipid phosphatase assays using

PtdIns(3,4,5)P3 as a substrate. In these assays, a malachite green reagent was used to detect phosphate liberated by lipid phosphatase activity towards the provided substrate.

We used three negative controls, consisting of assays performed either without PTEN added (“PIPγ Only,” using substrate but no source of PTEN, and “Beads Only,” using precipitate from shControl cells with beads but no antibody), or without substrate added

(“PTEN Only,” using PTEN immunoprecipitated from shControl cells). These controls produced negligible phosphate release (Figure 2.13 C). Compared to these controls,

PTEN immunoprecipitated from shControl DDp53-HMECs was able to convert 25.0 ±

0.6% of the PtdIns(3,4,5)P3 substrate provided to PtdIns(4,5)P2 (N = 3) (Figure 2.13 C).

PTEN immunoprecipitated from one of the shPIK3R1 lines converted significantly more substrate, with 30.5 ± 1.8% conversion, while PTEN immunoprecipitated from the other shPIK3R1 line was similar to the control, with 25.2 ± 1.0% substrate conversion (N = 3 for both) (Figure 2.13 C). Admittedly, the conclusions from this experiment are limited by the inability to control for the exact amount of immunoprecipitated PTEN in each in vitro lipid phosphatase assay; however, previous experiments from these same cell lines produced a comparable amount of PTEN protein in immunoprecipitation reactions

(Figure 2.11 A). These results suggest that in DDp53-HMECs, partial p85 loss does not lead to reduced PTEN lipid phosphatase activity.

PIK3R1 knockdown does not increase RTK activation or alter RTK trafficking

Because we were unable to demonstrate an interaction between p85 and PTEN, and did not find that PIK3R1 knockdown significantly impacted PTEN levels or lipid

phosphatase activity, we looked for a mechanism of transformation mediated by partial p85 loss that did not involve PTEN. To determine whether the augmented growth factor-mediated PI3K signaling in PIK3R1 knockdown cells was due to increased receptor activation, we assessed EGFR phosphorylation in response to EGF stimulation using two different methods. First, we synchronized shControl and shPIK3R1 DDp53-

HMECs by a 4 hour serum and growth factor starvation, stimulated the cells with

20ng/ml EGF for timepoints up to 1 hour, and collected protein lysates. We then performed immunoblotting using an antibody specific to EGFR phosphorylated on

Y1068, one of the three major activating autophosphorylation sites (Downward et al.,

1984). Although PIK3R1 knockdown cells exhibited significantly increased and sustained phosphorylation of AKT and S6 ribosomal protein, phosphorylation of EGFR at Y1068 was largely unchanged (Figure 2.14 A). While Y1068 is one of the main residues phosphorylated during EGFR activation, other phosphorylated tyrosines also contribute to activation of this receptor (Downward et al., 1984). To completely assess EGFR phosphorylation status, we next carried out the same EGF stimulation timecourse, performed immunoprecipitations for tyrosine-phosphorylated proteins, and subjected these samples to immunoblotting for total EGFR. Following EGF stimulation, PIK3R1 knockdown cells showed a similar or even slightly reduced amount of total tyrosine- phosphorylated EGFR when compared to control cells (Figure 2.14 B). Together, these results indicate that the increase in growth factor-stimulated PI3K signaling in PIK3R1 knockdown cells likely occurs at a step in pathway activation that is downstream of receptor autophosphorylation and activation.

Following activation, RTKs are internalized in endosomes, and are then either recycled back to the cell surface or targeted to lysosomes for degradation (Miaczynska, 2013).

Intracellular trafficking of receptors can control the magnitude and duration of signaling

Figure 2.14: PIK3R1 knockdown does not increase growth factor-stimulated RTK phosphorylation in DDp53-HMECs. (A) The indicated cell lines were starved of serum and growth factors for 4 hours and then stimulated with 20ng/ml EGF for the indicated amounts of time. Protein lysates were collected and subjected to immunoblotting. Membranes were stripped and re-probed for vinculin as a loading control. One representative experiment is shown. (B) Growth factor stimulation was performed as in (A); protein lysates were collected and subjected to immunoprecipitation using beads conjugated to the 4G10 antibody recognizing tyrosine-phosphorylated residues. Immunoprecipitates were subjected to immunoblotting for phosphorylated tyrosines using the 4G10 antibody; membranes were stripped and re-probed for total EGFR and vinculin as a loading control. One representative experiment is shown.

(Miaczynska, 2013), and is regulated by a number of Rab small GTPases (Jean and

Kiger, 2012; Stenmark, 2009). The BH domain of p85 reportedly has GAP activity towards select small GTPases, including Rab4 and Rab5 (Chamberlain et al., 2004), and disruption of this activity leads to increased PI3K/AKT activation and cellular transformation (Chamberlain et al., 2004; Chamberlain et al., 2008), apparently due to more rapid and sustained internalization and reduced degradation of RTKs

(Chamberlain et al., 2008; Chamberlain et al., 2010). Therefore, we were interested to know whether the increased PI3K signaling and transformation in our PIK3R1 knockdown cells could be due to altered intracellular trafficking of receptors, as a consequence of reduced p85 Rab-GAP function.

To assess effects of PIK3R1 knockdown on RTK trafficking, we used biotin labeling to track both internalization and degradation of surface proteins. First, to analyze receptor internalization, shControl and shPIK3R1 cells were synchronized by overnight serum and growth factor starvation, and then all surface proteins were labeled with biotin. Cells were stimulated with 20ng/ml EGF to initiate receptor internalization, remaining surface biotin was cleaved, and cells were lysed and subjected to immunoprecipitation with streptavidin beads to capture internalized biotinylated proteins. Immunoblotting was then used to assess the amount of total EGFR in the samples. Both shControl and shPIK3R1 cells had similar rates of EGFR internalization (Figure 2.15 A). Next, to analyze receptor degradation, all surface proteins of synchronized shControl and shPIK3R1 cells were labeled with biotin, cells were stimulated with EGF, and cells were lysed and subjected to immunoprecipitation with streptavidin beads to capture all remaining biotinylated proteins. Immunoblotting for total EGFR revealed that PIK3R1 knockdown cells had a rate of EGFR degradation comparable to that of the control cells (Figure 2.15 B). These experiments indicate that in DDp53-HMECs, partial loss of p85 does not have a

Figure 2.15: PIK3R1 knockdown does not affect growth factor-stimulated RTK trafficking in DDp53-HMECs. (A) EGFR internalization in response to EGF. The indicated cell lines were starved of serum and growth factors overnight. Surface proteins were labeled with biotin, cells were stimulated with 20ng/ml EGF for the indicated amounts of time to initiate receptor internalization, and remaining surface biotin was cleaved. Protein lysates were generated and used in streptavidin immunoprecipitation to capture internalized biotinylated proteins. Immunoblotting was used to determine the amount of tyrosine-phosphorylated EGFR in the samples. Membranes were stripped and re-probed for total EGFR and vinculin as a loading control. T.S., total surface protein prior to stimulation. (B) EGFR degradation in response to EGF. The indicated cell lines were starved as in (A). Surface proteins were biotinylated, and then cells were stimulated with 20ng/ml EGF for the indicated amounts of time. Protein lysates were generated and then subjected to streptavidin immunoprecipitation to capture all remaining biotinylated proteins. Immunoblotting was used to determine the amount of total EGFR in the samples. Membranes were stripped and re-probed for vinculin as a loading control. T.S., total surface protein prior to stimulation. Representative experiments for (A) and (B) are shown.

substantial effect on EGFR trafficking.

PIK3R1 knockdown increases the amount of p85-p110 bound to activated RTKs

Binding to the p85 regulatory subunit is required for p110 catalytic subunit stability (Yu et al., 1998b). Furthermore, p85 serves as a necessary adaptor for recruiting p110 to activated RTKs, where the p85-p110 heterodimer becomes activated (Rameh et al.,

1995; Yu et al., 1998a; Yu et al., 1998b). Research from the lab of Dr. Cantley has suggested that in some tissues, p85 may be present in excess of p110, and that free p85 might function as a negative regulator of PI3K signaling by competing with p85-p110 heterodimers for binding to phosphorylated RTKs (Luo and Cantley, 2005; Ueki et al.,

2003; Ueki et al., 2002b). Therefore, we were interested to know whether PIK3R1 knockdown in DDp53-HMECs might selectively reduce free p85, allowing more p85- p110 heterodimers to bind activated RTKs, thereby increasing PI3K/AKT signaling.

To test whether partial loss of p85 in our cells augments PI3K activation by selectively depleting free p85, we generated polyclonal DDp53-HMEC lines with stable expression of flag-tagged, activated ErbB3 (Flag-TEL-ErbB3) along with shRNA targeting PIK3R1.

Expression of Flag-TEL-ErbB3 lead to low-level colony formation of DDp53-HMECs, with

150.7 ± 18.8 colonies per 50,000 cells, while additional expression of p110-H1047R robustly increased transformation of these cells, to 891.5 ± 16.6 colonies per 50,000 cells (N = 3 for both) (Figure 2.16). Consistent with our previous data, knockdown of

PIK3R1 in these cells increased colony formation to a similar extent as p110-H1047R, with 685.8 ± 20.4 and 725.2 ± 53.9 colonies per 50,000 cells (N = 3 for both) (Figure

2.16). These findings suggest that in DDp53-HMECs, the expressed Flag-TEL-ErbB3 is weakly activated and transforming, and can cooperate with partial p85 loss.

Figure 2.16: PIK3R1 knockdown increases transformation of DDp53-HMECs expressing activated ErbB3. DDp53-HMECs stably expressing flag-tagged, activated ErbB3 (Flag-TEL-ErbB3) were infected with lentivirus containing a negative control scrambled shRNA or one of two distinct shRNAs targeting PIK3R1, or with retrovirus containing the cancer-associated p110-H1047R mutant as a positive control. Following selection for stable polyclonal lines, these cells were tested for anchorage-independent growth. Single-cell suspensions were plated in 0.3% agar and grown for 4 weeks. 50,000 cells were plated per 6cm plate. The number of colonies was counted for three independent experiments of at least 3 plates each. Means ± SEM are shown; N = 3 for each group. Statistical significance was determined by unpaired t-test. Significance for comparison to the shControl is shown. ***, P < 0.001; ****, P < 0.0001.

To assess the effect of PIK3R1 knockdown on binding of p85 and p110 to activated

RTKs in DDp53-HMECs, we starved the cells of serum and growth factors to limit RTK activation to just the Flag-TEL-ErbB3, then used the Flag tag to immunoprecipitate this activated receptor. Immunoblotting and densitometric quantification was then used to determine the amounts of pan-p85 and p110 bound to Flag-TEL-ErbB3 (Figure 2.17

A). Compared to the shControl, both shPIK3R1 lines had significantly reduced pan-p85 levels, by 68.6 ± 3.7% and 69.3 ± 1.7% (N = 4 for all) (Figure 2.17 E). PIK3R1 knockdown reduced the amount of pan-p85 bound to Flag-TEL-ErbB3 by 43.6 ± 14.7% and 53.4 ± 4.2% compared to the shControl (N = 4 for all) (Figure 2.17 B). Both shPIKγR1 lines also exhibited a slight but significant reduction in p110 protein in whole cell lysates as compared to the shControl line, of 20.9 ± 7.1% and 32.2 ± 6.4% (N = 4 for all) (Figure 2.17 F); despite this, the amount of p110 associated with Flag-TEL-ErbB3 was actually increased in shPIK3R1 lines, with fold increases of 1.5 ± 0.3 and 1.4 ± 0.1 as compared to the shControl (N = 4 for all) (Figure 2.17 C). The net result was that knockdown of PIK3R1 in DDp53-HMECs increased the ratio of p110:pan-p85 bound to activated ErbB3 by a fold of 3.7 ± 0.4 and 2.5 ± 0.1 as compared to the shControl (N = 4 for all) (Figure 2.17 D). These ratios correlated well with the fold increase in AKT activation in these cells, as determined by phosphorylation at S473, of 2.4 ± 0.4 and 1.7

± 0.2 as compared to the shControl line (N = 4 for all) (Figure 2.17 G). These findings were recapitulated with two additional PIK3R1 knockdown DDp53-HMEC lines, and also confirmed using a different pan-p85 antibody (data not shown).

Together, the data presented in this chapter is consistent with the model of increased

PI3K/AKT signaling and transformation mediated by partial p85 loss depicted in Figure

2.18: in mammary epithelial cells with normal p85 levels, p85 is present in excess of p110, and activated sites on receptors are able to be bound by both non-signaling free

Figure 2.17: PIK3R1 knockdown increases the amount of p85-p110 bound to activated RTKs in DDp53-HMECs. DDp53-HMECs stably expressing Flag-TEL-ErbB3 and either shControl, shPIK3R1, or p110-H1047R were starved of serum and growth factors for 4 hours; protein lysates were collected and subjected to immunoprecipitation for ErbB3 using beads conjugated to anti-flag antibody. (A) Representative immunoblot. Membranes were stripped and re-probed for total proteins and vinculin as a loading control. (B-D) Bands for proteins in ErbB3 immunoprecipitates from immunoblots were quantified by densitometry. For (B-C), protein levels were normalized first to the amount of ErbB3 in the immunoprecipitates, then to the corresponding mean for the shControl. For (D), protein levels of p110 were normalized to the levels of pan-p85 in immunoprecipitates, then to the corresponding mean for the shControl. (E-G) Bands for proteins in whole cell lysates from immunoblots were quantified by densitometry. Protein levels were normalized first to the vinculin loading control (for pan-85 and p110) or the total unphosphorylated protein (for phospho-AKTS473), then to the corresponding mean value for the shControl. For (B-G), means ± SEM are shown for four independent experiments, N = 4 for each. Statistical significance was determined by unpaired t-test. Significance for comparison to the shControl is shown. *, P < 0.05; **, P < 0.01; ***, P < 0.001; ****, P < 0.0001; ns, not significant.

Figure 2.18: Model: partial p85 loss leads to increased PI3K/AKT signaling and transformation. The data presented in Chapter 2 is consistent with a model in which p85 monomers compete with p85-p110 heterodimers to negatively regulate RTK- mediated PI3K/AKT signaling. Top: In mammary epithelial cells with normal p85 levels, p85 is present in excess of p110. Both monomeric p85 and heterodimeric p85-p110 can compete for binding to sites on activated RTKs, but only p85-p110 is capable of signaling. Bottom: In mammary epithelial cells with partial p85 reduction, the pool of monomeric p85 is selectively depleted, since p85 is required for p110 stability; more binding sites on activated RTKs are available for p85-p110 heterodimers, allowing for upregulated RTK-mediated PI3K signaling. While it is likely that reduced p85 levels similarly affect p110 and p110, the p110 isoform has a minimal role in RTK-mediated signaling, possibly due to lower RTK-associated lipid kinase activity of p110 in comparison to p110 (Utermark et al., 2012).

p85 and signaling-capable p85-p110 heterodimers; under conditions of partial p85 loss, the pool of free p85 is selectively reduced, allowing for more p85-p110 heterodimers to bind activated receptors, resulting in increased PI3K signal output. Although we expect p85 reduction to similarly affect RTK signaling through p110 and p110, since the p110 isoform contributes little to PI3K signaling downstream of RTKs (Utermark et al.,

2012), it likely plays a minor role in this context. By this model, a balance between p85 and p110 subunits is expected to be critical for regulation of PI3K signaling in response to inputs from activated RTKs.

Summary and discussion

In this chapter, we find that heterozygous deletion of PIK3R1 is frequent in human breast cancers, and that furthermore PIK3R1 expression is significantly reduced in breast tumors when compared to normal breast tissue. We show that RNAi-mediated downregulation of p85 in DDp5γ-HMECs augments PI3K/AKT signaling in response to growth factor stimulation, and increases colony formation in agar. PIK3R1 knockdown also augments PI3K/AKT signaling and transformation driven by oncogenes common in breast cancer, including p110-H1047R and activated HER2/neu. Studies using pan-

PI3K and isoform-selective inhibitors suggest that HMEC transformation driven by partial p85 loss is primarily mediated by signaling through p110. Contrary to reports linking p85 to PTEN stability or activity, we find that PIK3R1 downregulation in HMECs does not substantially change steady-state PTEN mRNA or protein levels, or in vitro PTEN lipid phosphatase activity. In addition, although p85 has been reported to have GAP activity towards the Rab5 GTPase important for intracellular trafficking, we find that RTK internalization and degradation is largely unchanged in PIK3R1 HMECs. Instead, we find that partial reduction of p85 increases the amount of p85-p110 associated with activated RTKs, suggesting a model where excess p85 monomers compete with p85-

p110 heterodimers to negatively regulate RTK-mediated PI3K signaling.

The data presented in Chapter 2 strongly suggest that partial loss of p85 leads to an increase in growth factor-stimulated PI3K/AKT pathway activation. However, a number of additional experiments will help to confirm these findings. We have shown that

PIK3R1 knockdown increases phosphorylation of AKT at both activation sites in response to growth factors, and while AKT phosphorylation is in general a good readout for PI3K pathway activation, a more direct measure will be to assess production of

PtdIns(3,4,5)P3. We expect that in our HMECs, PtdIns(3,4,5)P3 levels will correlate well with AKT phosphorylation, but in certain contexts it seems that this is not always the case (Vora et al., 2014) and (J.A. Engelman, unpublished observations), so this will be important to confirm. We also plan to extend our immunoblot experiments (Figure 2.2,

Figure 2.3, Figure 2.4, Figure 2.5, and Figure 2.6) to include other activated downstream components of this pathway, in particular 4EBP1 phosphorylated at S65,

S6K phosphorylated at T389, PRAS40 phosphorylated at Tβ46, and GSKγ phosphorylated at S9, to further confirm PI3K/AKT pathway upregulation in our PIK3R1 knockdown cells. Finally, we will expand on our pharmacological inhibition experiments

(Figure 2.8, Figure 2.9, and Figure 2.10) by using AKT-selective inhibitors such as the

ATP-competitive small molecule inhibitor GSK690693 (Rhodes et al., 2008) or the allosteric inhibitor MK2206 (Hirai et al., 2010) to treat our PIK3R1 knockdown HMECs. If transformation of these cells is indeed driven by PI3K/AKT pathway upregulation, we expect that these agents will effectively block signaling in and transformation of these cells in a manner similar to PI3K-selective inhibitors. Together, these proposed experiments will further confirm that partial p85 loss transforms HMECs by increasing

PI3K/AKT pathway activation.

Although the data presented in Chapter 2 are consistent with the competition model we propose (Figure 2.18), additional experiments may substantially strengthen the argument for this model. Using immunoblotting techniques, we have shown that PIK3R1 knockdown in HMECs increases the amount of p85-p110 bound to activated RTKs

(Figure 2.17 D). We plan to verify this result using mass spectrometry to determine the number of p85 and p110 molecules bound to these activated RTKs. It will also be critical to use in vitro lipid kinase assays to demonstrate that in these same cells, the PI3K activity associated with activated RTKs is increased upon PIK3R1 knockdown. Our model also relies on p85 being present in excess of p110 in HMECs. We performed immunodepletion experiments similar to those published by other groups (Geering et al.,

2007; Mauvais-Jarvis et al., 2002; Ueki et al., 2002a) which indicated that this may in fact be the case (data not shown), but technical difficulties precluded a definitive conclusion from these results. An alternative approach will be to perform size-exclusion chromatography on whole cell lysates from control and PIK3R1 knockdown HMECs, followed by immunoblotting of fractions for p85 and p110 isoforms. We expect that if p85 is present in excess of p110, p85 will be detected in two distinct sets of fractions: monomeric p85 without p110 will be found in earlier fractions of smaller molecular weight, while heterodimeric p85 with p110 will be found in later fractions of higher molecular weight. We additionally expect that compared to the control cells, PIK3R1 knockdown HMECs will have a reduced amount of monomeric p85 detected in earlier fractions of lower molecular weight. Finally, to determine the absolute amounts of PI3K subunits in our cells, we plan to perform mass spectrometry or immunoblotting of total cell lysates in comparison to known amounts of recombinant p110 and p85 proteins.

Together, these experiments may strengthen the argument for the competition model we have proposed here.

One caveat to the immunoprecipitation experiments presented in this chapter to support this competition model is the use of Flag-TEL-ErbB3 as an activated RTK probe for bound PI3K isoforms. HER3/ErbB3 is generally considered to be a pseudokinase because it lacks residues conserved among other HER/ErbB family RTKs thought to be required for autophosphorylation and catalytic function (Guy et al., 1994; Jura et al.,

2009; Shi et al., 2010; Sierke et al., 1997); it instead propagates RTK signals by forming heterodimers with other HER/ErbB family members, in particular EGFR/ErbB1 and

HER2/ErbB2 (Pinkas-Kramarski et al., 1996; Tzahar et al., 1996), and does not naturally form homodimers (Berger et al., 2004). Here we use a fusion of ErbB3 to a TEL domain that facilitates homodimerization and activation of RTKs (Carroll et al., 1996; Golub et al., 1994; Jousset et al., 1997). Although it might be expected that ErbB3 homodimerization would therefore not activate RTK signaling, a recent report indicated that ErbB3 might in fact retain weak catalytic activity (Shi et al., 2010). Our data demonstrating that ectopic expression of Flag-TEL-ErbB3 in HMECs modestly increases

AKT phosphorylation under starvation conditions (Figure 2.17 A compared to Figure 2.2

A) and augments colony formation (Figure 2.16 compared to Figure 2.3 A) are consistent with this report. Because ErbB3 possess direct p85-binding YXXM motifs

(Hellyer et al., 1998; Prigent and Gullick, 1994; Soltoff et al., 1994) while EGFR/ErbB1 and HER2/ErbB2 require adaptors to interact with p85 (Baselga and Swain, 2009), it was an ideal probe for bound p85 and p110 in our system.

In addition to the competition model favored here, we explored other possible explanations for the augmented PI3K/AKT signaling and transformation seen in HMECs with reduced PIK3R1 expression. The BH domain of p85 has been reported to have

GAP activity towards Rab4 and Rab5, small GTPases critical for endosomal trafficking

(Chamberlain et al., 2004). Disruption of the Rab-GAP function of p85 leads to

increased growth factor-stimulated PI3K/AKT activation and cellular transformation as a result of more rapid and sustained Rab-mediated RTK internalization (Chamberlain et al., 2004; Chamberlain et al., 2008; Chamberlain et al., 2010). Furthermore, a recent report demonstrated that in MEFs, RNAi-mediated PIK3R1 knockdown increased the amount of active GTP-bound Rab5, augmented PI3K/AKT pathway activation, and induced autophagy (Dou et al., 2013). Accordingly, we examined whether p85 downregulation affected growth factor-stimulated trafficking of RTKs in our HMEC lines.

We found that PIK3R1 knockdown did not have a substantial effect on EGFR internalization or degradation in response to EGF stimulation (Figure 2.15). However, we did not examine intracellular trafficking of EGFR by additional means, for example immunofluorescence using fluorophore-conjugated EGF and known markers for different intracellular compartments. We also did not explore whether PIK3R1 knockdown in

HMECs affected activation of Rab GTPases. Therefore, while our work suggests that the increased growth factor-stimulated PI3K/AKT activation in these cells is likely not due to altered RTK trafficking, we cannot completely rule out an effect on Rab GTPase activation or function.

Based on published reports implicating p85 in PTEN binding (Chagpar et al., 2010;

Rabinovsky et al., 2009), stability (Cheung et al., 2011; Taniguchi et al., 2010), or lipid phosphatase activity (Chagpar et al., 2010; Taniguchi et al., 2006), we also explored the possibility that PIK3R1 knockdown had an effect on PTEN. Despite repeated attempts, we were unable to confirm interaction of either endogenous or ectopically expressed p85 with PTEN in HMECs or in a number of other cell types (Figure 2.11, Figure 2.12, and data not shown). We also did not find that PIK3R1 knockdown substantially affected

PTEN mRNA or protein levels at the steady state (Figure 2.2 C, Figure 2.5 C, and

Figure 2.13 A). In addition, based on reports that in some contexts p85 is translocated

to the nucleus (Chiu et al., 2014; Park et al., 2010; Winnay et al., 2010) and a multitude of publications demonstrating a poorly-understood role for nuclear PTEN (reviewed in

(Planchon et al., 2008)), we examined whether PIK3R1 knockdown in HMECs affected the subcellular localization of PTEN. Cellular fractionation experiments indicated that p85 downregulation did not affect the proportion of PTEN associated with lipid membrane, cytosolic, or nuclear fractions in either quiescent or EGF-stimulated HMECs

(data not shown). This is consistent with a recent study suggesting that p85 and PTEN nuclear localization are not related (Chiu et al., 2014). Together our findings suggest that in HMECs with PIK3R1 knockdown, upregulated PI3K/AKT signaling and transformation may not be mediated by an effect of p85 directly on PTEN.

In summary, the data presented in this chapter demonstrates that the levels of p85 modulate PI3K/AKT activation in mammary epithelial cells. RNAi-mediated PIK3R1 downregulation increases growth factor-stimulated PI3K signaling in and transformation of HMECs. Partial reduction of p85 also synergizes with oncogenes common in breast cancer, including p110-H1047R and oncogenic HER2/neu. Both pan-PIγK and p110- selective pharmacological inhibitors are equally effective at blocking PI3K/AKT signaling and colony formation mediated by PIK3R1 knockdown. Data from immunoprecipitation of activated RTKs are consistent with a model where p85 is in excess of p110, and monomeric p85 can compete with p85-p110 heterodimers for binding to RTKs to fine- tune PIγK output. While others have reported a role for p85 in regulation of Rab- mediated receptor trafficking, we find that PIK3R1 knockdown does not affect EGFR internalization or degradation in HMECs. In contrast to published reports, we also do not find that partial p85 reduction directly affects PTEN levels or lipid phosphatase activity.

The current literature on this topic, and the implications our work in this context, are discussed in further detail in Chapter 4 of this dissertation.

Chapter 3: PI3K regulatory subunit p85alpha plays a tumor suppressive role in

a genetically engineered mouse model of mammary tumorigenesis

Acknowledgements

The HMS Rodent Histopathology Core prepared slides of fixed tissue for histological analysis and performed hematoxylin and eosin (H&E) staining of these slides. Carolynn

Ohlson performed mammary tumor transplants, daily drug administrations, and immunohistochemical staining of fixed tissue samples. Lauren Thorpe performed all other experiments and data analysis.

I would additionally like to thank Lewis Cantley for the generous contribution of floxed

Pik3r1 mice, and William Muller for the generous contribution of MMTV-NIC mice. Thank you to Thanh Von for sharing his extensive knowledge of in vivo techniques. Thank you to Roderick Bronson of the HMS Rodent Histopathology Core for help in pathological analysis. Thank you to Qi Wang for sharing protocols for mouse mammary epithelial cell

(MMEC) isolation, and Hailing Cheng for sharing protocols and reagents for the isolation and culture of mouse tumor cells. Finally, I would like to thank Stephanie Santiago for sharing tips and tricks for mouse mammary gland whole mount preparation.

Preface

In this chapter, we use genetically engineered mouse models (GEMMs) to explore the consequences of Cre/loxP-mediated conditional Pik3r1 ablation in the mouse mammary epithelium. We find that p85 expression is not required for normal mouse mammary gland development during puberty, pregnancy, or lactation. We then use a GEMM of

HER2/neu-driven breast cancer to demonstrate that Pik3r1 ablation significantly reduces the latency of mammary tumor onset. When transplanted into recipient mice, the growth of these tumors is blocked by treatment with pan-PIγK or p110-selective inhibitors.

Together, these findings demonstrate the in vivo importance of p85 as a tumor suppressor in the mammary epithelium, and suggest that isoform-selective PI3K therapies may be effective in breast cancers characterized by a decrease in p85.

Introduction

Genetically engineered mouse models are a convenient system in which to study contributions of specific genes to both normal mammary gland development and mammary tumorigenesis. Both mice and humans have three main stages of mammary development: embryonic, pubertal, and adult (Watson and Khaled, 2008) (Figure 3.1).

During mouse embryonic development, five pairs of mammary buds form and undergo limited initial branching, forming a rudimentary ductal tree. Following birth, terminal end buds (TEBs) appear at the tips of the ducts and begin to invade the mammary fat pad; during puberty, estrogen and other signals stimulate TEB proliferation and clefting, resulting in ductal elongation and branching. By approximately 12 weeks after birth, the mammary fat pad is filled, TEBs disappear, and ductal growth ceases. Adult mammary glands undergo further change during pregnancy and lactation: progesterone and prolactin induce side-branching and alveolar bud formation, differentiation into alveoli, and milk secretion. After weaning, cell death, alveolar collapse, and remodeling occur

Figure 3.1: Schematic of mammary gland development in the mouse. Mouse mammary gland development progresses through several distinct stages and is tightly regulated by a number of hormones and signaling pathways. At birth, a few rudimentary ducts surround the nipple. During puberty, these ducts undergo pronounced outgrowth and branching; at the end of puberty, the mammary fat pad is filled with a ductal network. During pregnancy, additional ductal branching occurs, along with extensive lobular-alveolar development; in preparation for lactation, the secretory epithelium undergoes functional differentiation, allowing milk to be produced and secreted. Following weaning, the alveolar compartment undergoes remodeling, termed involution, and the mammary gland returns to a pre-pregnancy-like state. Hormones and growth factors known to be important for transition from one state to another are indicated above blue arrows. Signaling pathways known to be important during certain stages are indicated above that stage. Adapted from (Hennighausen and Robinson, 2001) and (Hennighausen and Robinson, 1998).

during a process called involution, returning the mammary gland to a pre-pregnancy-like state (Hennighausen and Robinson, 1998). In addition to hormone signals, the different stages of mammary gland development are tightly regulated by a number of other signaling pathways (Hennighausen and Robinson, 2001; Hynes and Watson, 2010;

Watson and Khaled, 2008).

Several studies have established the important role of the PI3K pathway in mouse mammary gland development and tumorigenesis. Conditional ablation of Pten in the mammary epithelium led to accelerated ductal outgrowth and precocious lobulo-alveolar development during puberty and pregnancy; these mice also frequently developed mammary tumors with heterogeneous pathology as early as 2 months (Li et al., 2002).

Mice with mammary-specific expression of an inducible oncogenic p110-H1047R transgene developed mammary tumors with a mean latency of 7 months (Liu et al.,

2011), whereas Pik3ca ablation in the mammary epithelium significantly impaired pubertal mammary gland branching and outgrowth and post-partum lactation, and also blocked mammary tumor development driven by polyoma middle T antigen (pyMT) or oncogenic HER2/neu (Utermark et al., 2012). Surprisingly, mammary-specific deletion of

Pik3cb resulted in modestly hypermorphic mammary gland development, with precocious lobulo-alveolar growth and increased ductal branching, and moderately accelerated pyMT- or HER2/neu-driven mammary tumor development (Utermark et al.,

2012). The distinct roles of p110 and p110 in the mammary gland were explained by a proposed model in which p110 may have higher RTK-associated kinase activity than p110, allowing p110 to compete with p110 for binding sites on RTKs to regulate

PI3K output (Utermark et al., 2012) (Figure 1.6). This model was supported by extensive biochemical experiments demonstrating increased RTK-bound p110 and RTK- associated PIγK activity in mouse mammary epithelial cells with p110 knockout

(Utermark et al., 2012). Together, these studies demonstrate that in mouse models, genetic alterations that increase PI3K/AKT signaling lead to accelerated mammary gland development and tumorigenesis, while alterations that reduce PI3K/AKT signaling impede mammary gland development and block tumorigenesis in certain contexts.

A number of publications have suggested that in vivo, p85 expression levels modulate

PI3K/AKT activation. Mice with genetic ablation of the p85 isoform only (Terauchi et al.,

1999), of all three regulatory isoforms arising from Pik3r1 (p85, p55, and p50)

(Fruman et al., 2000), or of p55 and p50 only (Chen et al., 2004) exhibited hypoglycemia, increased insulin sensitivity, and increased PI3K/AKT pathway activation upon insulin stimulation. Conversely, mice with increased p85 expression displayed increased insulin resistance and reduced PtdIns(3,4,5)P3 production (Barbour et al.,

2005). Expression levels of p85 have also been shown to modulate pathophysiological signals in mice. Heterozygous deletion of Pik3r1 increased the incidence of prostatic intraepithelial neoplasia induced by heterozygous Pten knockout (Luo et al., 2005c) and the number of lung tumors in a GEMM of lung cancer driven by oncogenic KRAS

(Engelman et al., 2008). Mice with liver-specific Pik3r1 ablation developed hepatitis and dysplastic liver nodules by 6 months of age, and liver tumors resembling hepatocellular carcinoma by 14 to 20 months (Taniguchi et al., 2010). In this study, liver lysates from

Pik3r1 knockout mice demonstrated upregulated PI3K/AKT activation and

PtdIns(3,4,5)P3 accumulation; liver tumors were found to have significantly reduced Pten mRNA and protein levels (Taniguchi et al., 2010). This work indicates that reduced p85 expression can increase PI3K/AKT signaling in vivo, and that additionally in some tissues p85 downregulation augments tumorigenesis.

In this chapter, we examine the specific role of p85 in the mammary gland. We use

conditional knockout techniques to determine the consequences of Pik3r1 ablation on normal mouse mammary gland development. We then combine mammary-specific

Pik3r1 ablation with an established GEMM of HER2/neu-driven breast cancer to examine the role of p85 in mammary tumor development. Finally, we explore the ability of pan- and isoform-selective PI3K inhibitors to block in vivo mammary tumorigenesis in the context of reduced p85. As some of these agents are currently in early clinical trials

(Table 1.2 and Table 1.4), this work has important implications for therapeutic targeting of breast cancers with reduced p85 expression.

Results

Pik3r1 expression is not required for mouse mammary gland development

The data presented in Chapter β indicates a tumor suppressive role for p85 in the in vitro transformation of mammary epithelial cells. To determine whether these observations hold true in vivo, we used GEMMs to evaluate the physiological and pathophysiological consequences of p85 loss in this tissue. In mice, embryonic Pik3r1 knockout is lethal (Fruman et al., 2000). Therefore, we took advantage of the Cre/loxP recombination system to conditionally ablate Pik3r1 in the mammary epithelium (Figure

3.2). Mice bearing a floxed Pik3r1 allele (Luo et al., 2005b) were backcrossed more than ten generations to the FVB/N wildtype background to eliminate variations in mammary development (MacLennan et al., 2011) and mammary tumor latency (Davie et al., 2007) arising from different genetic backgrounds. To study the role of p85 in mouse mammary gland development, these Pik3r1 floxed mice were then crossed with MMTV-

Cre transgenic mice, in which expression of the Cre recombinase is under control of the

MMTV LTR promoter and occurs in the secretory epithelium, including the mammary gland, beginning at early stages of development (Wagner et al., 1997). The resulting

MMTV-Cre; Pik3r1+/loxP and MMTV-Cre; Pik3r1loxP/loxP mice have mosaic ablation of one

100

Figure 3.2: Schematic of Pik3r1 conditional knockout allele and breeding scheme for mammary-specific Pik3r1 ablation. In mice, Pik3r1 generates three regulatory isoforms, p85, p55, and p50, by alternative transcription initiation and splicing. The full-length isoform p85 has N-terminal SH3 and BH domains encoded by exons 1A through 6, while p55 and p50 have the unique first exons 1C and 1B, respectively. The first exon common to all three isoforms arising from Pik3r1 is exon 7. The conditional Pik3r1 knockout mouse used in these studies have knock-in of an engineered Pik3r1 allele in which exon 7 is flanked by loxP sites (Luo et al., 2005b); Cre recombinase mediates recombination of the loxP sites, resulting in deletion of exon 7. Transcription of the recombined gene produces a truncated p85 consisting of the SHγ and BH domains, but lacking the domains necessary for binding to p110 or activated RTKs, and essentially does not produce any p55 or p50 (Luo et al., 2005b). These floxed Pik3r1 mice were interbred with MMTV-Cre mice (Wagner et al., 1997), which express the Cre transgene under control of the MMTV LTR promoter, to achieve mammary-specific Pik3r1 ablation.

101

or both Pik3r1 alleles mainly in luminal mammary epithelial cells, and will hereafter be referred to as MMTV-Cre; Pik3r1+/- and MMTV-Cre; Pik3r1-/-.

To confirm the successful ablation of Pik3r1 in these mice, we isolated mouse mammary epithelial cells (MMECs) from resected mammary glands of adult nulliparous female

MMTV-Cre; Pik3r1+/- and MMTV-Cre; Pik3r1-/- mice, generated protein lysates from these cells, and used immunoblotting to determine expression of PI3K catalytic and regulatory isoforms. Lysates of MMECs isolated from adult nulliparous female MMTV-

Cre mice were used as a wildtype control. MMECs derived from MMTV-Cre; Pik3r1+/- and MMTV-Cre; Pik3r1-/- mice exhibited dramatically reduced p85 protein levels in comparison to the MMTV-Cre control MMECs (Figure 3.3). In addition, as p85 is known to be required for the stability of p110 (Fruman et al., 2000; Luo et al., 2005b; Yu et al.,

1998b), we were not surprised to find a reduction in p110 protein levels in MMTV-Cre;

Pik3r1+/- and MMTV-Cre; Pik3r1-/- MMECs. We also found that the level of p85 protein varied slightly in each individual mouse, likely due to the mosaic nature of MMTV-Cre expression (Wagner et al., 1997). Nonetheless, this protein analysis confirmed that this

GEMM could be used to determine the consequences of p85 loss in the mouse mammary epithelium.

We then used whole mount techniques to analyze the effects of p85 loss in these mice on the different stages of mammary gland development. The fourth inguinal mammary glands from control and Pik3r1 floxed female mice were excised, fixed on slides, and prepared with Carmine staining. Mammary gland development was assessed during puberty in nulliparous female mice at 6 weeks of age, at the completion of puberty in 12- week-old nulliparous females, during pregnancy at 14 days, and during lactation at 2 days postpartum (Figure 3.4). We found that there was no appreciable difference in the

102

Figure 3.3: Transgenic MMTV-Cre ablates Pik3r1 expression in mouse mammary epithelial cells. MMECs were derived from the mammary glands of individual nulliparous MMTV-Cre, MMTV-Cre; Pik3r1+/-, and MMTV-Cre; Pik3r1-/- females. Protein lysates were prepared from these MMECs and subjected to immunoblotting for PI3K isoforms. The membrane was stripped and re-probed for vinculin as a loading control.

103

Figure 3.4: Pik3r1 expression is not required for mouse mammary gland development. Whole mounts were prepared from the fourth inguinal mammary glands of MMTV-Cre; Pik3r1+/-, and MMTV-Cre; Pik3r1-/- female mice during puberty (A-C), at the end of puberty (D-F), during pregnancy (G-I), and during lactation (J-L). Mammary glands were stained with Carmine red to visualize ducts. Representative images from each genotype and stage are shown.

104

mammary gland development of MMTV-Cre; Pik3r1+/- and MMTV-Cre; Pik3r1-/- mice as compared to the MMTV-Cre control at all developmental stages examined. Halfway through puberty, at 6 weeks of age, the duct outgrowth of all three genotypes was similar relative to the lymph node (Figure 3.4 A-C). At the end of puberty at 12 weeks, the ducts had completely filled the mammary fat pads in all three genotypes, and ductal branching was comparable (Figure 3.4 D-F). Heterozygous or homozygous Pik3r1 ablation had no substantial effect on lobulo-alveolar development during pregnancy (Figure 3.4 G-I) or lactation (Figure 3.4 J-L). Together, these findings demonstrate that p85 is not required for these stages of mammary gland development in the mouse.

Mammary-specific Pik3r1 ablation leads to spontaneous mammary tumor development

Although deletion of Pik3r1 did not have a substantial effect on mouse mammary gland development, over a longer time frame nulliparous MMTV-Cre; Pik3r1+/- and MMTV-Cre;

Pik3r1-/- females eventually developed focal or multifocal spontaneous mammary tumors. These tumors formed with an average latency of 14.1 months, an average survival of 14.4 months, and a penetrance of 90% (9/10) (Table 3.1). In addition, we found metastasis of the primary mammary tumor to the lungs in 11% (1/9) of the mice with spontaneous mammary tumors. Many of the characteristics of mammary tumor development in Pik3r1 knockout mice are comparable to those of other established mouse models of breast cancer (Table 3.2).

To determine the expression of PI3K isoforms and activation of the PI3K/AKT pathway in

Pik3r1 knockout spontaneous mammary tumors, we generated protein lysates from tumor tissue and subjected them to immunoblotting. Compared to whole mammary gland lysates from MMTV-Cre nulliparous females, mammary tumors from MMTV-Cre;

Pik3r1+/- and MMTV-Cre; Pik3r1-/- mice had reduced p85 protein levels; tumors from

105

Table 3.1: Nulliparous female mice with mammary-specific Pik3r1 ablation develop spontaneous mammary tumors

Tumor Survival Lung Genotype Mouse ID Latency Primary Tumor Pathology Comments (months) Metastasis? (months)

Undifferentiated MMTV-Cre; Pik3r1+/- A659 10.7 11.6 No Focal mammary tumor (1) adenocarcinoma, sarcoma Undifferentiated MMTV-Cre; Pik3r1+/- A609 12.7 15.9 No Multifocal mammary tumors (3) adenocarcinoma, sarcoma Undifferentiated MMTV-Cre; Pik3r1+/- A603 13.5 14.9 No Multifocal mammary tumors (2) adenocarcinoma, sarcoma

MMTV-Cre; Pik3r1+/- A605 17.1 17.8 ND No Multifocal mammary tumors (2) 106 +/-

MMTV-Cre; Pik3r1 A607 17.9 ND ND No

Undifferentiated MMTV-Cre; Pik3r1-/- A660 ND 10.7 Yes Focal mammary tumor (1) adenocarcinoma, sarcoma Undifferentiated MMTV-Cre; Pik3r1-/- A606 12.2 14.1 No Multifocal mammary tumors (3) adenocarcinoma, sarcoma Mammary adenocarcinoma, MMTV-Cre; Pik3r1-/- A602 13.9 15.2 No Focal mammary tumor (1) mammary hyperplasia Undifferentiated MMTV-Cre; Pik3r1-/- A597 14.7 14.9 No Multifocal mammary tumors (2) adenocarcinoma, sarcoma Sacrificed mouse at 22 months MMTV-Cre; Pik3r1-/- A618 NA NA NA NA due to severe dermatitis; did not have tumors

ND, not determined; NA, not applicable.

Table 3.2: Comparison of mammary tumor development in Pik3r1 knockout mice to other established GEMMs of breast cancer

Tumor Tumor Lung Genotype Primary Tumor Pathology Reference Latency Penetrance Metastasis? MMTV-Cre; Pik3r1+/- and Mixed (sarcoma, mammary L.M. Thorpe and J.J. Zhao, 14.1 months 90% (9/10) 11% (1/9) MMTV-Cre; Pik3r1-/- adenocarcinoma) unpublished observations Mixed mammary MMTV-rtTA; tetO-Pik3caH1047R 6.8 months 95% (adenocarcinoma, NR (Liu et al., 2011) adenosquamous carcinoma)

Mammary and skin (IDC-nos, K14-Cre; p53-/-; Brca1-/- 7.0 months 100% carcinosarcoma, NR (Liu et al., 2007) adenomyoepithelioma)

KI Focal mammary comedo- 107 MMTV-Cre; ErbB2 13.8 months 83% 6% (Andrechek et al., 2000) adenocarcinomas

Multifocal, solid nodular MMTV-NIC 6.4 months 100% 56% (Schade et al., 2009) mammary carcinoma Solid nodular mammary MMTV-Cre; Pten+/- 15.5 months 75% 18% (2/11) (Dourdin et al., 2008) carcinoma Mixed (fibroadenoma, MMTV-Cre; Pten-/- 11 months 75% NR (Li et al., 2002) adenocarcinoma)

Mixed mammary (solid nodular mammary carcinoma, MMTV-Cre; ErbB2KI; Pten+/- 6.5 months 100% 35% (6/17) (Dourdin et al., 2008) adenomyoepithelioma, adenosquamous carcinoma)

NR, not reported.

MMTV-Cre; Pik3r1-/- mice also generally had a greater reduction in total p85 levels when compared to tumors from MMTV-Cre; Pik3r1+/- mice (Figure 3.5). Interestingly, though

PTEN loss has been reported in spontaneous liver tumors of mice with Pik3r1 knockout

(Taniguchi et al., 2010), we did not observe a pattern of PTEN protein reduction in

Pik3r1 knockout mammary tumors. In addition, we found that PI3K/AKT pathway activation in these mammary tumors as assessed by phosphorylation of AKT and S6 ribosomal protein was highly variable (Figure 3.5). Thus while we find that mice with mammary-specific Pik3r1 ablation develop spontaneous mammary tumors, our data is inconclusive as to whether these tumors arise due to upregulated PI3K/AKT signaling.

We were further interested to know the pathology of spontaneous mammary tumors from

Pik3r1 knockout mice. Analysis of formalin-fixed tumor tissue by hematoxylin and eosin

(H&E) staining revealed that these tumors ranged in pathology, from sarcoma to mammary adenocarcinoma (Figure 3.6 A-F). In addition, metastasis of the primary mammary tumor to the lung was observed in 11% (1/9) of the mice with spontaneous tumors (Figure 3.6 G-I). We also used whole mount techniques to examine the adjacent mammary glands from Pik3r1 knockout mice with spontaneous mammary tumors.

Interestingly, non-tumor-bearing mammary glands from these mice displayed a slightly hypermorphic phenotype with apparent lobulo-alveolar development and excessive ductal branching (Figure 3.7) resembling mammary glands from pregnant females

(Figure 3.4 G-I). Together these findings suggest that Pik3r1 ablation alone is sufficient for mammary tumor development in mice.

Pik3r1 ablation reduces the latency of HER2/neu-driven mouse mammary tumors

To further study the contribution of p85 loss to in vivo mammary tumorigenesis, we combined a well-defined GEMM of HER2/neu-driven breast cancer with our conditional

108

Figure 3.5: PI3K/AKT pathway activation in spontaneous mammary tumors from Pik3r1 knockout mice. Mammary tumor tissue was collected from random-fed MMTV- Cre; Pik3r1+/- and MMTV-Cre; Pik3r1-/- female mice. As a control, normal mammary gland tissue was collected from MMTV-Cre mice (“MG”). Protein lysates were generated and subjected to immunoblotting for PI3K/AKT pathway components and activation of PI3K/AKT pathway effectors. Membranes were stripped and re-probed for total proteins and vinculin as a loading control.

109

Figure 3.6: Pathology of primary spontaneous mammary tumors and lung metastases from Pik3r1 knockout mice. (A-F) Representative images of formalin-fixed primary mammary tumor tissue from MMTV-Cre; Pik3r1+/- and MMTV-Cre; Pik3r1-/- females stained with hematoxylin and eosin (H&E). These spontaneous tumors display a range of pathologies, from sarcoma and undifferentiated adenocarcinoma (A-C) to mammary hyperplasia and mammary adenocarcinoma (D-F). White arrowheads designate areas of the tumor where duct-like formations occur. Scale bars = 50m. (G-I) Representative images of formalin-fixed lung tissue from one mouse in (A-F) stained with H&E. Black arrows designate metastases. Scale bars = 100m.

110

Figure 3.7: Adjacent mammary glands from Pik3r1 knockout mice with spontaneous mammary tumors have a hypermorphic phenotype. Mammary gland tissue from an approximately 1 year old nulliparous MMTV-Cre; Pik3r1-/- female with palpable mammary tumors was excised, fixed, and subjected to Carmine staining. The hypermorphic phenotype of these mammary glands, with what appears to be lobular- alveolar development and ductal branching similar to that occurring during pregnancy, can be seen clearly in (C) and (D), magnified images taken from the white boxes shown in (A) and (B) respectively.

111

Pik3r1 knockout mice. We bred Pik3r1 floxed mice with MMTV-NIC mice, which express a bicistronic transgene consisting of an activated HER2/neu allele and Cre recombinase under control of the MMTV promoter (Schade et al., 2009; Ursini-Siegel et al., 2008)

(Figure 3.8). The resulting MMTV-NIC; Pik3r1+/loxP and MMTV-NIC; Pik3r1loxP/loxP mice have ablation of one or both Pik3r1 alleles in the same luminal mammary epithelial cells that express oncogenic HER2/neu. MMTV-NIC mice were used as a control. These mice will hereafter be referred to as NIC, NIC; Pik3r1+/-, and NIC; Pik3r1-/-.

We monitored female cohorts of these mice for progression of mammary tumors. All three genotypes developed multifocal mammary tumors with 100% penetrance. NIC mice developed palpable tumors with a mean latency of 140 days (range 97-191, N =

25), consistent with other published information for this strain (Schade et al., 2009;

Ursini-Siegel et al., 2008; Utermark et al., 2012). Either heterozygous or homozygous

Pik3r1 ablation significantly reduced the time to tumor onset: NIC; Pik3r1+/- and NIC;

Pik3r1-/- mice developed tumors with mean latencies of 125 days (range 101-162, N =

38) and 126 days (range 86-155, N = 43), respectively (Figure 3.5 A). We additionally determined the average number of mammary tumors, total tumor mass, and number of lung metastases per mouse for all three genotypes five weeks after the onset of the first palpable tumor. NIC mice developed on average 9.4 ± 1.5 tumors (N = 16), while NIC;

Pik3r1+/- mice had a significantly higher number of tumors, with on average 16.0 ± 1.2 tumors per mouse (N = 20); NIC; Pik3r1-/- mice developed a comparable number of tumors to the NIC control, with on average 10.4 ± 4.3 tumors per mouse (N = 20)

(Figure 3.9 B). All three genotypes had comparable total tumor weight; NIC mice had an average total tumor mass of 4.6 ± 0.8 grams (N = 16), while NIC; Pik3r1+/- and NIC;

Pik3r1-/- mice had an average total tumor mass of 5.4 ± 0.5 grams and 4.7 ± 0.4 grams, respectively (N = 20 for both) (Figure 3.9 C). To determine the number of lung

112

Figure 3.8: Schematic of the transgenic NIC allele and breeding scheme for mammary-specific HER2/neu expression and Pik3r1 ablation. MMTV-NIC mice are a well established genetically engineered model of HER2/neu-driven breast cancer (Schade et al., 2009; Ursini-Siegel et al., 2008). These mice express a bicistronic transgene consisting of a HER2/neu allele containing an activating in-frame deletion (Schade et al., 2009; Siegel et al., 1999; Ursini-Siegel et al., 2008) and the Cre recombinase gene, linked by an internal ribosomal entry site (IRES). This transgene is expressed under the control of the MMTV LTR promoter. MMTV-NIC mice develop multifocal mammary tumors with 100% penetrance and an average latency of 198 ± 43 days. Approximately 60% of MMTV-NIC mice also develop lung metastases. In this study, we interbred MMTV-NIC and conditional Pik3r1 knockout mice to achieve expression of activated HER2/neu and deletion of Pik3r1 in the same luminal mammary epithelial cells.

113

Figure 3.9: Pik3r1 ablation reduces the latency of HER2/neu-driven mammary tumor development. (A) Cohorts of NIC, NIC; Pik3r1+/-, and NIC; Pik3r1-/- female mice were observed every three days for mammary tumor onset as determined by first palpation. Median tumor-free survival: NIC, 140 days (N = 25); NIC; Pik3r1+/-, 125 days (N = 38); NIC; Pik3r1-/-, 126 days (N = 43). Statistical significance was determined by Log-rank (Mantel-Cox) test. (B) The total number of mammary tumors per mouse was determined for NIC (N = 16), NIC; Pik3r1+/- (N = 20), and NIC; Pik3r1-/- (N = 20) females. (C) The total wet weight of mammary tumors plus associated mammary gland tissue per mouse was determined for NIC (N = 16), NIC; Pik3r1+/- (N = 20), and NIC; Pik3r1-/- (N = 20) females. (D) The number of lung metastases per mouse was determined for NIC, NIC; Pik3r1+/-, and NIC; Pik3r1-/- females (all groups N = 8). For (B-D), numbers were determined exactly 5 weeks after tumor onset. Means ± SEM are shown. Statistical significance was determined by unpaired t-test. Significance for NIC; Pik3r1+/- and NIC; Pik3r1-/- in comparison to the NIC control is shown. **, P < 0.01; ***, P < 0.001; ****, P < 0.0001; ns, not significant.

114

metastases per mouse, we examined H&E-stained lung sections by microscopy.

Although the difference was not statistically significant, there was a trend towards higher incidence of lung metastasis in mice with Pik3r1 ablation, with NIC mice having an average of 2.7 ± 0.5 mets, and NIC; Pik3r1+/- and NIC; Pik3r1-/- having an average of 5.6

± 2.2 mets and 6.1 ± 4.9 mets, respectively (N = 8 for all groups) (Figure 3.9 D).

Together these results indicate that partial p85 loss reduces the latency of HER2/neu- driven mammary tumorigenesis in mice, and may contribute to the severity of the disease.

To analyze PI3K isoform expression and PI3K/AKT pathway activation in these tumors, we isolated primary tumor tissue from random-fed mice, generated protein lysates, and subjected them to immunoblotting (Figure 3.10). We also quantified the bands from these blots by densitometry (Figure 3.11). Although there was some mouse-to-mouse variation, likely owing to the mosaic expression of the MMTV LTR promoter (Wagner et al., 1997), NIC; Pik3r1+/- mammary tumors had a 25.8 ± 15.1% reduction in p85 protein levels, while NIC; Pik3r1-/- mammary tumors had a 83.9 ± 2.2% reduction in p85 protein levels (N = 4 for both) (Figure 3.11 A). Expression of PI3K catalytic isoforms correlated with the protein levels of p85: NIC; Pik3r1+/- mammary tumors had a 46.5 ± 14.0% reduction in p110 protein levels and a 9.8 ± 10.8% reduction in p110 levels, while

NIC; Pik3r1-/- mammary tumor cells had a 69.8 ± 4.8% reduction in p110 protein levels and a 55.8 ± 5.4% reduction in p110 levels (N = 4 for all) (Figure 3.11 D-E). Notably, although others have reported that p85 may important for the stability of PTEN

(Cheung et al., 2011; Taniguchi et al., 2010), we did not observe a significant change in

PTEN protein levels in NIC; Pik3r1+/- or NIC; Pik3r1-/- mammary tumors as compared to the NIC control (Figure 3.11 F). Finally, we did not observe a significant difference in activation of the PI3K/AKT pathway, as determined by phosphorylation of AKT and S6

115

Figure 3.10: Effect of Pik3r1 ablation on PI3K/AKT pathway activation in HER2/neu-driven mammary tumors. Mammary tumor tissue was collected from random-fed NIC, NIC; Pik3r1+/-, and NIC; Pik3r1-/- female mice. Protein lysates were generated and subjected to immunoblotting for PI3K/AKT pathway components and activation of PI3K/AKT pathway effectors. Membranes were stripped and re-probed for total proteins and vinculin as a loading control.

116

Figure 3.11: Quantification of the effect of Pik3r1 ablation on PI3K/AKT pathway activation in HER2/neu-driven mammary tumors. Bands in the immunoblots of NIC, NIC; Pik3r1+/-, and NIC; Pik3r1-/- mammary tumor lysates from Figure 3.10 were quantified by densitometry. Levels of p85, pan-p85, p110, p110, and PTEN were normalized to the corresponding vinculin loading control on the re-probed membrane, then to the corresponding NIC control mean. Levels of phosphorylated HER2/neu, AKT, and S6 were normalized to the corresponding total proteins on the re-probed membrane, then to the corresponding NIC control mean. Means ± SEM are shown; N = 4 for each group. Statistical significance was determined by unpaired t-test. Significance is shown for NIC; Pik3r1+/- and NIC; Pik3r1-/- compared to the NIC control. *, P < 0.05; **, P < 0.01; ***, P < 0.001; ns, not significant.

117

ribosomal protein (Figure 3.11 G-I). Thus although we have shown that partial p85 loss reduces the latency of HER2/neu-driven mammary tumors, it is not clear from this data whether PI3K signaling is upregulated in Pik3r1 knockout tumors.

We additionally examined the histology of formalin-fixed tumor tissue of each genotype.

Immunohistochemical (IHC) staining for p85 revealed that in NIC tumors, p85 was mainly cytoplasmic, with strong and uniform signal throughout the tumor and perhaps slightly elevated levels at the tumor edge (Figure 3.12 A); NIC; Pik3r1+/- tumors showed a reduction in p85, while NIC; Pik3r1-/- tumors had very little signal for p85 (Figure

3.12 B-C), correlating well with the immunoblot results (Figure 3.11 A). H&E staining of tumor tissue showed that tumors from mice of all three genotypes had similar solid nodular carcinoma histology (Figure 3.12 D-F); it is unsurprising that Pik3r1 ablation had no effect on tumor pathology, since it has been shown previously that Pten ablation also does not change the pathology of MMTV-NIC tumors (Schade et al., 2009). Finally, IHC was used to stain for Ki67, a nuclear protein associated with cellular proliferation. Ki67

IHC revealed that compared to NIC tumors, which had 9.3 ± 0.9% proliferating cells,

NIC; Pik3r1+/- and NIC; Pik3r1-/- tumors had nearly double the proliferation indices, with

16.5 ± 2.1% and 17.5 ± 1.5% proliferating cells respectively (N = 12 for all groups)

(Figure 3.12 G-J). Together these results demonstrate that in mice, reduced p85 expression significantly increases the proliferation of HER2/neu-driven mammary tumor cells, correlating with a significant reduction in the latency of tumor onset.

Growth of Pik3r1 knockout tumors is blocked by p110-selective inhibitors

Since mammary epithelial cell transformation driven by p85 loss is blocked in vitro by pan-PI3K or p110-selective inhibitors, we were interested to know whether these agents could block growth of HER2/neu-driven mammary tumors with p85 loss in vivo.

118

Figure 3.12: Effect of Pik3r1 ablation on tumor pathology and proliferation of HER2/neu-driven mammary tumors. Formalin-fixed tissue from NIC, NIC; Pik3r1+/-, and NIC; Pik3r1-/- mammary tumors was subjected to immunohistochemical staining for p85 (A-C), staining with hematoxylin and eosin (H&E) (D-F), or subjected to immunohistochemical staining for the proliferation marker Ki67 (G-I). The percentage of Ki67-positive nuclei (J) was calculated by dividing the number of positively-stained nuclei by the total number of nuclei in the field of view. Means ± SEM are shown; all groups N = 12. Statistical significance was determined by unpaired t-test. Significance is shown for NIC; Pik3r1+/- and NIC; Pik3r1-/- compared to the NIC control. **, P < 0.01; ***, P < 0.001. Scale bars = 50m.

119

Primary tumors were excised from a NIC; Pik3r1+/- donor female and orthotopically transplanted into eight week old NcrNu female recipients. Recipient mice were randomly assigned to four cohorts, and treated daily with methylcellulose vehicle, the pan-PI3K inhibitor GDC0941 (125mg/kg), the p110-selective inhibitor BYL719 (45mg/kg), or the p110-selective inhibitor KIN193 (20mg/kg); tumor size was measured every three days using calipers. All agents were administered by oral gavage with the exception of

KIN193, which was administered by intraperitoneal injection. Compared to the vehicle control, which had an average final tumor volume of 942.6 ± 222.9 mm3, KIN193 did not have a substantial effect on tumor growth, with an average final tumor volume of 801.6 ±

111.1 mm3; treatment with either GDC0941 or BYL719 significantly blocked the growth of transplanted tumors, with average final volumes of 299.5 ± 66.8 mm3 and 293.2 ±

85.0 mm3 respectively (N ≥ 10 for all cohorts) (Figure 3.13 A). Similar results were obtained upon treatment of transplanted NIC; Pik3r1-/- tumors (data not shown).

Although all treatment groups maintained good body weight, signifying that none of the drugs were excessively toxic over the course of treatment (Figure 3.13 B), it should be noted that compared to the vehicle control, treatment with GDC0941 did lead to a slight but significant percent decrease in body weight at later time points (Figure 3.13 C).

Together, these results indicate that pan-PIγK and p110-selective inhibitors are similarly able block in vivo tumor growth in the context of p85 loss.

To further study the effects of pan- and isoform-selective PI3K inhibitors on Pik3r1 ablated tumors in vivo, we analyzed activation of the PI3K pathway in tumors under each treatment condition. Recipient mice were treated for four days, and tumors were excised from recipient mice one hour after the last administration. A portion of each tumor was used to generate protein lysates, which were then subjected to immunoblotting to assess activation of components downstream of PI3K. Acute treatment with either

120

Figure 3.13: Pan-PI3K or p110-selective inhibitors block the growth of transplanted HER2/neu tumors with Pik3r1 ablation. NIC; Pik3r1+/- mammary tumors were orthotopically transplanted into female NcrNu mice. Recipients were randomly assigned to cohorts for once daily treatment with the pan-PI3K inhibitor GDC0941 (1β5mg/kg, oral gavage), the p110-selective inhibitor BYL719 (45mg/kg, oral gavage), the p110-selective inhibitor KIN193 (20mg/kg, intraperitoneal injection), or the vehicle control (methylcellulose, oral gavage). Tumor size was measured with calipers (A) and body weight was determined (B) every three days. Percent change in body weight (C) was calculated for each mouse relative to its weight at the start of treatment (day 0). Means ± SEM are shown; all treatment groups N ≥ 10. Statistical significance was determined by two-way ANOVA with Tukey’s multiple comparisons test. In (A), only significance for comparison of the final tumor volumes (day 18) of each treatment group to the vehicle control is shown. In (C), significance is shown for comparison of GDC0941 treatment to the vehicle control for days 12, 15, and 18; no other comparisons were statistically significant. *, P < 0.05; **, P < 0.01; ****, P < 0.0001; ns, not significant.

121

GDC0941 or BYL719, but not KIN193, substantially reduced phosphorylation of AKT at both activation sites and phosphorylation of S6 ribosomal protein at activation residues

S235/236 (Figure 3.14 A). The other portion of the excised treated tumor tissue was fixed in formalin and used to assess histopathology. H&E staining revealed that the transplanted tumors retained similar histology to the primary tumors (Figure 3.14 B-E).

IHC staining for AKT phosphorylated at S473 showed that tumors treated with vehicle or

KIN193 had moderate to strong nuclear and cytoplasmic AKT activation in most cells, while tumors treated with GDC0941 or BYL719 had only slight AKT activation in a limited number of cells (Figure 3.14 F-I). Similarly, IHC staining for S235/236-phosphorylated

S6 ribosomal protein showed strong cytoplasmic activated S6 signal in most cells of tumors treated with vehicle or KIN193, while tumors treated with GDC0941 or BYL719 showed strong cytoplasmic S6 activation in only a few select cells (Figure 3.14 J-M).

These results demonstrate that pan-PI3K or p110-selective inhibition, but not p110- selective inhibition, effectively decreases PI3K/AKT pathway signals in transplanted

HER2/neu-driven tumors with reduced p85.

Finally, IHC staining was used to assess cellular proliferation and apoptosis in transplanted tumors treated with pan-PI3K or isoform-selective inhibitors. Ki67 IHC was used to visualize the nuclei of proliferating tumor cells. Compared to the vehicle-treated sample with 34.2 ± 2.7% Ki67-positive nuclei, either GDC0941 or BYL719 treatment significantly reduced the percentage of Ki67-positive nuclei to 9.7 ± 1.1% and 14.3 ±

1.4%, respectively, while KIN193 treatment had no effect on the proliferation index of tumor cells, with 33.6 ± 2.9% Ki67-positive nuclei (N = 8 for all groups) (Figure 3.15 A-D and Figure 3.15 I). IHC using the TUNEL method was performed to visualize the fragmented DNA of cells undergoing apoptosis. While vehicle treatment did not induce apoptosis of tumor cells, with only 1.2 ± 0.3% TUNEL-positive nuclei, either GDC0941

122

Figure 3.14: Pan-PI3K or p110-selective inhibitors suppress PI3K/AKT activation in transplanted HER2/neu tumors with Pik3r1 ablation. NIC; Pik3r1+/- mammary tumors orthotopically transplanted into NcrNu females and treated with the pan-PI3K inhibitor GDC0941, the p110-selective inhibitor BYL719, the p110-selective inhibitor KIN193, or the vehicle control as described in Figure 3.13. One hour after treatment on day 4, recipients were sacrificed and tumor tissue was collected. (A) Protein lysates from tumors in each treatment group were subjected to immunoblotting for PI3K/AKT pathway activation. Membranes were stripped and re-probed for total proteins and vinculin as a loading control. (B-M) Formalin-fixed tumor tissue was stained with hematoxylin and eosin (H&E) (B-E), or subjected to immunohistochemical staining for AKT phosphorylated at S473 (F-I) or S6 ribosomal protein phosphorylated at S235/236. Scale bars = 50m.

123

Figure 3.15: Pan-PI3K or p110-selective inhibitors suppress proliferation and induce apoptosis in transplanted HER2/neu tumors with Pik3r1 ablation. NIC; Pik3r1+/- mammary tumors were orthotopically transplanted into NcrNu females. Recipients were treated with the pan-PI3K inhibitor GDC0941, the p110-selective inhibitor BYL719, the p110-selective inhibitor KIN193, or the vehicle control as described in Figure 3.13. One hour after treatment on day 4, recipients were sacrificed and tumor tissue was collected. Formalin-fixed tissue was subjected to immunohistochemical staining for the proliferation marker Ki67 (A-D) or for fragmented DNA using the TUNEL method (E-H). The percentage of Ki67-positive (I) or TUNEL- positive (J) nuclei was calculated by dividing the number of positively-stained nuclei by the total number of nuclei in the field of view. Means ± SEM are shown; all groups N = 8. Statistical significance was determined by unpaired t-test. Significance is shown for each treatment group compared to the vehicle control. **, P < 0.01, ****, P < 0.0001; ns, not significant. Scale bars = 50m.

124

or BYL719 treatment significantly increased the percentage of TUNEL-positive nuclei to

5.8 ± 1.0% and 4.8 ± 0.5%, respectively, while KIN193 treatment had no effect on tumor cell apoptosis with 0.9 ± 0.3% TUNEL-positive nuclei (N = 8 for all groups) (Figure 3.15

E-H and Figure 3.15 J). Together this data demonstrates that pan-PI3K or p110- selective inhibitors, but not p110-selective inhibitors, effectively block PI3K pathway activation in mammary tumors in vivo in the context of p85 loss, and furthermore reduce proliferation and induce apoptosis of mammary tumor cells.

Summary and discussion

In this chapter, we use Cre/loxP-mediated conditional deletion to study the effects of

Pik3r1 ablation on the mouse mammary gland. We find that Pik3r1 expression is not required for mammary development during puberty, pregnancy, or lactation. However, mice with mammary-specific Pik3r1 knockout develop spontaneous mammary tumors with a mean latency of 14.1 months. We furthermore show that when combined with an established GEMM of HER2/neu-driven breast cancer, mammary-specific Pik3r1 ablation significantly reduces the latency and increases the cellular proliferation of mammary tumors. Growth of these tumors is equally and effectively blocked by pan-

PIγK or p110-selective pharmacological inhibitors. Pan-PIγK or p110-selective therapeutics also block PI3K/AKT signaling, reduce proliferation, and induce apoptosis in mammary tumors with Pik3r1 knockout. These findings help elucidate the previously unstudied role of p85 in the mammary epithelium, and also have important implications for therapeutic targeting of breast cancers with decreased p85.

Because PI3K signaling is known to be important for mouse mammary gland development, we were surprised to find that Pik3r1 ablation did not have an appreciable effect on the morphology of this tissue. Previous studies have shown that increased

125

PI3K activation via Pten ablation leads to hypermorphic mammary gland development during puberty and pregnancy (Li et al., 2002), while reduced PI3K signaling due to

Pik3ca ablation significantly impairs pubertal mammary development and post-partum lactation (Utermark et al., 2012). Based on our data in Chapter 2 demonstrating that partial p85 loss leads to upregulated PIγK signaling, we expected to find mildly accelerated mammary gland development in mice with conditional Pik3r1 ablation.

Although we do not have a definitive explanation for our observations, we propose that because partial p85 loss has a subtle effect on PIγK activation in comparison to strong oncogenes such as p110-H1047R (Figure 2.2 D), it may not upregulate PI3K signaling strongly enough in this tissue during development to substantially affect its morphology.

This idea is supported by the published result that Pten ablation leads to tumor development in mice as early as 2 months (Li et al., 2002), while we find that the earliest onset of spontaneous tumors in mice with mammary-specific Pik3r1 ablation is about

10.7 months (Table 3.1). Analysis of adjacent mammary glands from these mice using whole mount techniques suggested that by the time spontaneous mammary tumors developed, the remaining mammary glands without tumors showed signs of alveolar differentiation (Figure 3.7) similar to normal mammary gland development during pregnancy (Figure 3.4 G-I). It is likely that in mice with mammary-specific Pik3r1 ablation, PI3K/AKT signaling upregulation is modest and does not substantially affect mammary gland development during puberty or pregnancy, but over a longer time period leads to a slightly hypermorphic phenotype and spontaneous tumor development.

An additional confounding factor to our examination of the effect of conditional Pik3r1 ablation on mammary gland development is the levels of other PI3K regulatory and catalytic isoforms. Although we were unable to identify an antibody capable of selectively recognizing murine p85 by immunoblot (data not shown), we expect that MMTV-Cre;

126

Pik3r1+/- and MMTV-Cre; Pik3r1-/- mice should have comparable levels of p85 to the

MMTV-Cre control. Retained Pik3r2 expression is expected to be critical for the normal mammary gland morphology in our conditional Pik3r1 knockout mice, as p85 is necessary to sustain PIγK signaling in conditions of p85 depletion (Ueki et al., 2002a); ablation of both Pik3r1 and Pik3r2 would likely lead to substantially impaired mammary gland development. This hypothesis is supported by the finding that conditional Pik3r1 ablation increases insulin-stimulated PI3K/AKT activity in the heart, while double Pik3r1 and Pik3r2 knockout largely blocks PI3K signaling (Luo et al., 2005a). Although Pik3r1 ablation should not affect p85 levels, we do find that it leads to significant reductions in protein levels of both the p110 and p110 catalytic isoforms (Figure 3.3, Figure 3.5,

Figure 3.10, and Figure 3.11 D-E). Since it has previously been demonstrated that

Pik3ca ablation impairs mammary gland development during puberty and pregnancy, while Pik3cb ablation leads to precocious lobulo-alveolar development (Utermark et al.,

2012), p110 and p110 downregulation likely contribute to the mammary phenotype of mice with Pik3r1 ablation. The overall effect of mammary-specific Pik3r1 ablation on

PI3K/AKT activation and mammary gland development in these mice is probably a combination of the changes to all class IA PI3K isoforms in this tissue.

Although Pik3r1 ablation does not have a substantial impact on mouse mammary gland development, we find that 90% of MMTV-Cre; Pik3r1+/- and MMTV-Cre; Pik3r1-/- females develop spontaneous mammary tumors by an average of 14.1 months (Table 3.1). This timeline to spontaneous tumor development is comparable to the 14 to 20 month latency for development of spontaneous liver tumors resembling hepatocellular carcinoma in mice with liver-specific Pik3r1 ablation (Taniguchi et al., 2010), and to other GEMMs of breast cancer (Table 3.2). While tumor lysates prepared from random-fed mice were confirmed to have reduced p85 protein levels by immunoblotting, they exhibited

127

variable PI3K/AKT activation (Figure 3.5). While we believe that these spontaneous tumors likely arose due to persistent low-level upregulation of PI3K, which may be difficult to demonstrate in tumor tissue from random-fed mice, it will also be interesting to determine whether these tumors are prone to the accumulation of other genetic alterations, as has been observed with other GEMM models of breast cancer (Liu et al.,

2011). Further study of these spontaneous Pik3r1 knockout mammary tumors using

RNA sequencing techniques will provide insight into this question.

Another puzzling result from our in vivo studies is the finding that in mice, heterozygous or homozygous Pik3r1 ablation has a nearly identical effect on the latency of HER2/neu- driven mammary tumors (Figure 3.9). Because heterozygous PIK3R1 loss is much more frequent than homozygous deletion in human breast cancers (Figure 2.1 A), we expected that in our mouse model heterozygous Pik3r1 ablation would be more tumorigenic. While a number of studies have shown that either heterozygous (Mauvais-

Jarvis et al., 2002) or homozygous (Fruman et al., 2000; Terauchi et al., 1999) Pik3r1 ablation leads to hypoglycemia and enhanced insulin sensitivity as compared to wildtype controls, few have made direct comparisons between the consequences of heterozygous or homozygous loss of this gene on PI3K/AKT signaling. One publication compared wildtype, heterozygous Pik3r1 knockout, and homozygous Pik3r1 knockout

MEFs, and found that while heterozygous knockout increased IGF1-stimulated PI3K activity associated with p110 or tyrosine-phosphorylated proteins and robustly enhanced IGF1-stimulated PtdIns(3,4,5)P3 production in comparison to wildtype MEFs, homozygous Pik3r1 knockout had a less pronounced effect (Ueki et al., 2002a). Thus it was unclear from the literature why heterozygous or homozygous Pik3r1 ablation would similarly increase tumorigenesis in the mouse mammary gland.

128

To explain our observations, we considered that based on our model from Chapter 2

(Figure 2.18) heterozygous Pik3r1 ablation in the mouse mammary epithelium might partially reduce free p85, while homozygous Pik3r1 ablation may completely reduce free p85 and also partially reduce p85-p110 heterodimers. Conceivably these two scenarios could result in similar levels of PI3K/AKT signaling upregulation (Figure 3.16).

We sought to address this by immunoprecipitating endogenous HER3/ErbB3 from NIC,

NIC; Pik3r1+/-, and NIC; Pik3r1-/- mammary tumor lysates, and using immunoblotting to assess the amount of p85 and p110 bound to these RTKs. Unfortunately, we found that although we could achieve robust HER3/ErbB3 immunoprecipitation from tumor lysates, the tissue homogenization process apparently disrupted all protein-protein interactions, as no PI3K isoforms were detectable in the immunoprecipitates by immunoblot (data not shown). We also cultured cells from these mammary tumors, generated protein lysates, and again performed HER3/ErbB3 immunoprecipitations; this protocol resulted in successful p85 and p110 pulldown, but the background on these immunoblots was too high to reliably quantify the protein bands (data not shown). Verification of this proposed explanation for our findings will likely require the use of other techniques such as mass spectrometry.

In summary, the data presented in this chapter demonstrates the important role of p85 in the mammary epithelium. While p85 expression is not required for normal mouse mammary gland development during puberty, pregnancy, or lactation, mammary-specific

Pik3r1 ablation leads to spontaneous mammary tumor development accompanied by a hypermorphic mammary gland phenotype within about one year. Pik3r1 ablation also significantly reduces the latency of mammary tumor development driven by HER2/neu in an established GEMM. Either pan-PIγK or p110-selective inhibitors effectively block growth of transplanted HER2/neu-driven mammary tumors with Pik3r1 ablation and

129

Figure 3.16: Model: heterozygous or homozygous Pik3r1 ablation has a similar effect on HER2/neu-driven tumorigenesis. Pik3r1 ablation reduces the latency of HER2/neu-driven mammary tumor development in the NIC mouse model. Surprisingly, NIC; Pik3r1+/- and NIC; Pik3r1-/- female mice have nearly identical average time to tumor development. Here we show one possible explanation for these observations. Top: In NIC mice, p85 is present in excess of p110; p85 monomers and p85-p110 heterodimers compete for binding to activated RTKs, but only p85-p110 heterodimers can signal. Middle: Heterozygous Pik3r1 ablation partially depletes the pool of free p85, but some monomers remain. Bottom: Homozygous Pik3r1 ablation completely depletes the pool of free p85, but also reduces the number of p85-p110 heterodimers. By this model, heterozygous or homozygous Pik3r1 ablation could result in similar low-level upregulation of RTK-mediated PI3K signaling. While we expect that partial p85 loss similarly affects signaling by p110 and p110, the p110 isoform is believed to have minimal contribution to RTK-mediated PI3K signaling due to substantially lower RTK- associated lipid kinase activity (Utermark et al., 2012).

130

PI3K/AKT signaling in these tumors. The implications of these findings, along with the data presented in Chapter 2, for the clinical application of PI3K-targeted therapies in breast cancers with reduced p85 are discussed in Chapter 4 of this dissertation.

131

Chapter 4: Summary, discussion, and future directions

Summary

Research in the past decade has established that PI3K isoforms play critical but divergent roles in both normal cellular signaling and in cancer. Chapter 1 discusses the current state of this field: the distinct roles of PI3K isoforms in different signaling contexts, the ways in which different PI3K isoforms are altered in cancer, and the promise of emerging isoform-selective therapeutics in the clinic. The roles of class I PI3K catalytic isoforms in cancer are relatively well studied; oncogenic mutation of p110 is frequent in human cancers, while p110, p110, and p110 are rarely mutated but can be overexpressed (Table 1.1 and Appendix A). Although PI3K regulatory isoforms had not previously been considered to contribute to oncogenesis, recent studies have converged to implicate p85 mutation or loss of expression in certain cancers. In this dissertation we find that heterozygous deletion of PIK3R1 is a frequent event in human breast cancers, and furthermore that PIK3R1 expression is significantly reduced in breast tumors when compared to normal breast tissue. Based on these findings, we use both in vitro and in vivo approaches to explore the role of p85 as a tumor suppressor in the transformation of mammary epithelial cells.

In Chapter 2, we use RNAi-mediated knockdown of PIK3R1 to assess the effects of p85 loss on human mammary epithelial cells (HMECs) in vitro. We find that partial p85 reduction leads to increased growth factor-stimulated PI3K signaling in and transformation of these cells. We further show that PIK3R1 knockdown augments HMEC transformation by oncogenes, including activated HER2/neu. Using pharmacological inhibitors, we demonstrate that the increased PI3K signaling and transformation driven by partial p85 loss is largely mediated by p110, and can be equally blocked by either the pan-PI3K inhibitor GDC0941 or the p110-selective inhibitor BYL719. Although others have reported a role for p85 in the stabilization or activation of the PTEN

133

phosphatase opposing PI3K, we were unable to demonstrate an effect of PIK3R1 knockdown on PTEN mRNA or protein levels, or on in vitro PTEN lipid phosphatase activity. We also did not detect a change in RTK endocytosis or degradation, counter to reports that p85 plays an important role in activation of Rabs critical for intracellular trafficking. Instead, we find that partial loss of p85 increases the amount of p85-p110 bound to activated RTKs. This result is consistent with a model in which p85 is in excess of p110 in wildtype HMECs, allowing monomeric p85 to compete with p85-p110 for binding to activated RTKs and negatively regulate PI3K signaling. PIK3R1 knockdown might selectively reduce the pool of monomeric p85, allowing more p85-p110 heterodimers to bind RTKs, increasing PI3K signaling and transformation (Figure 2.18).

In Chapter 3, we use the Cre/loxP system to study the effects of mammary-specific

Pik3r1 deletion on both normal mouse mammary gland development and mammary oncogenesis. Surprisingly, we find that Pik3r1 expression is not required for mouse mammary gland development during puberty, pregnancy, or lactation. However, mice with mammary-specific Pik3r1 ablation develop spontaneous mammary tumors of heterogeneous pathology with a mean latency of 14.1 months. Furthermore, Pik3r1 ablation significantly reduces the latency and increases the cellular proliferation of mammary tumors in a mouse model of HER2/neu-driven breast cancer. Either the pan-

PI3K inhibitor GDC0941 or the p110-selective inhibitor BYL719 effectively blocked growth of transplanted Pik3r1 knockout tumors. Treatment of these tumors with

GDC0941 or BYL719 also considerably reduced PI3K/AKT activation, and significantly reduced cellular proliferation and increased apoptosis. These findings indicate an important role for p85 as a tumor suppressor in mammary tumor formation in vivo, and furthermore suggest that pan-PI3K or p110-selective inhibitors might be effective therapeutics in breast cancers characterized by p85 loss.

134

Discussion and future directions

In this dissertation, we present in vitro data using HMECs and in vivo data using GEMMs demonstrating a tumor suppressive role for PIγK regulatory subunit p85 in mammary epithelial cells. Our data is consistent with a model where p85 is present in excess of p110 in normal mammary epithelial cells, allowing p85 monomers to compete with p85- p110 heterodimers for binding sites on activated RTKs to modulate PI3K signaling. We additionally show that in the context of transformation mediated by partial p85 loss, the efficacy of p110-selective inhibitors is comparable to that of pan-PI3K inhibitors. Our findings have several important implications for the current literature regarding p85- mediated transformation, the regulation of RTK-mediated PIγK signals by p85, and the therapeutic targeting of breast cancers characterized by partial p85 loss. They also raise interesting questions about the role of differential p85 expression in various tissues and metabolic contexts. These concepts will be important to address in continuing work on this project.

Transformation mediated by partial p85 loss does not appear to be through PTEN

A number of recent reports link p85 to PTEN stability or lipid phosphatase activity, and some studies have demonstrated an interaction between these two proteins. This proposed connection between p85 and PTEN has been invoked to explain increased

PI3K/AKT signaling and transformation in the context of p85 mutation or downregulation (Cheung et al., 2011; Taniguchi et al., 2006; Taniguchi et al., 2010). The results presented in this dissertation suggest that further study is needed to conclude whether PTEN contributes to transformation mediated by partial reduction of p85. At a minimum, our data suggest that these reported effects of p85 on PTEN may not be operative in all tissues.

135

An interaction between p85 and PTEN has been reported by two separate groups. In one study p85, p85, and p110 were found to participate in a complex with unphosphorylated PTEN (Rabinovsky et al., 2009); a second study demonstrated direct binding between the C-terminal SHγ and BH domains of p85 and PTEN that was dependent on EGF stimulation only in certain cell types (Chagpar et al., 2010). Thus the exact nature of the proposed p85-PTEN interaction and the contexts in which these proteins bind are still not clear. Although we were unable to positively show co- immunoprecipitation of p85 and PTEN in a number of different cell types and conditions

(Figure 2.11, Figure 2.12, and data not shown), we cannot absolutely rule out this interaction. It is possible that the antibodies we used for our co-immunoprecipitation assays obscured protein domains necessary for binding; although we tried many of the antibodies used in these two publications, at least one was not a commercial reagent

(D.H. Anderson, personal communication). However, we would note that at least one other study reported being unable to demonstrate binding of p85 and PTEN (Taniguchi et al., 2006).

Other publications have indicated that p85 is important for PTEN mRNA or protein stability. Ectopic expression of the cancer-associated truncation mutant p85-E160* reportedly led to PTEN protein destabilization via ubiquitin-mediated proteasomal degradation, while overexpression of wildtype p85 was shown to stabilize PTEN protein levels (Cheung et al., 2011). In another study, mice with liver-specific Pik3r1 ablation developed liver tumors resembling hepatocellular carcinoma within 14 to 20 months; compared to liver lysates from these mice at 6 months of age, livers from mice aged 16 to 18 months exhibited upregulated AKT phosphorylation and PtdIns(3,4,5)P3 production, and a corresponding reduction in both PTEN protein and mRNA levels

(Taniguchi et al., 2010). In HMECs, we found that RNAi-mediated PIK3R1 knockdown

136

did not have a substantial effect on steady-state PTEN mRNA levels (Figure 2.13 A) or protein levels (Figure 2.2 C and Figure 2.5 C). We also did not find that ablation of one or both Pik3r1 alleles had a consistent effect on PTEN protein levels either in spontaneous mammary tumors as compared to wildtype mammary glands (Figure 3.5) or in mammary tumors driven by HER2/neu (Figure 3.10 and Figure 3.11 F). These findings corroborate a report that liver-specific Pik3r1 ablation had no effect on PTEN protein levels in this tissue (Taniguchi et al., 2006). Together our results indicate that p85 may not be important in the regulation of PTEN mRNA or protein levels in mammary epithelial cells. Our data furthermore suggests that PTEN mRNA and protein reduction found in spontaneous Pik3r1 knockout liver tumors may be a secondary event that is not necessarily linked to p85 downregulation.

Finally, p85 has been reported to regulate PTEN lipid phosphatase activity. In liver extracts from mice with liver-specific Pik3r1 ablation, PtdIns(3,4,5)P3 production and

AKT activation were found to be upregulated; although Pik3r1 knockout had no effect on

PTEN protein levels, it significantly reduced both basal and insulin-stimulated PTEN lipid phosphatase activity (Taniguchi et al., 2006). Another study reported that in in vitro assays, addition of purified p85 increased PTEN lipid phosphatase activity in a concentration-dependent manner (Chagpar et al., 2010). However, we did not find a difference in the in vitro lipid phosphatase activity associated with PTEN immunoprecipitates from shControl and shPIK3R1 HMEC lines (Figure 2.13 C). This data suggests that at least in this cell type, p85 may not modulate PTEN activity.

Together our results indicate that in both mouse and human mammary epithelial cells, partial p85 loss does not substantially alter PTEN levels. They also suggest that in

HMECs, p85 may not have a significant effect on PTEN activity. We conclude it is

137

unlikely that the observed transformation mediated by p85 downregulation in our model systems is due to reduced PTEN function. It would be especially interesting to carry out experiments examining the ability of partial p85 loss to increase transformation of

PTEN-null mammary epithelial cells. One approach would be to use RNAi techniques to silence PIK3R1 in PTEN-null breast cancer cell lines or in HMECs with CRISPR- mediated PTEN deletion. A parallel approach would be to use Cre/loxP-mediated conditional ablation to knock out both Pik3r1 and Pten in the mouse mammary epithelium. The ability of reduced p85 to augment transformation and tumorigenesis in the complete absence of PTEN would further support the idea that this mechanism is independent of proposed effects of p85 on PTEN stability or activity.

Implications of the competition model for transformation mediated by p85 loss

We present data demonstrating that in HMECs, PIK3R1 knockdown increases PI3K/AKT pathway signaling in response to growth factors, cellular transformation, and the amount of p85-p110 associated with activated RTKs. Our results are consistent with a model where partial p85 loss selectively reduces a pool of monomeric p85 which competes with p85-p110 heterodimers for binding activated RTKs to negatively regulate PI3K/AKT signaling downstream of RTKs (Figure 2.18). While there are a number of remaining experiments that would more conclusively support this model, discussed in Chapter 2 of this dissertation, it has several important implications for p85-mediated regulation of

PI3K signaling, and the general role of p85 regulatory isoforms in transformation.

Although our work here has focused on PIK3R1, there are three genes encoding class

IA regulatory isoforms: PIK3R1, encoding p85 and its splicing variants p55 and p50,

PIK3R2, encoding p85, and PIK3R3, encoding p55. These five isoforms are collectively called p85 type regulatory subunits. All p85 isoforms possess the iSH2

138

domain responsible for binding class IA catalytic isoforms, and the SH2 domains responsible for binding to phosphorylated YXXM motifs on activated RTKs or their adaptors (Figure 1.1). The experiments presented in this dissertation reduced or ablated expression of all three p85 isoforms arising from PIK3R1, leaving expression of PIK3R2 and PIK3R3 intact. We expect that the remaining p85 isoforms are necessary to sustain

PI3K signaling in the context of reduced PIK3R1 expression.

Our model predicts that modulation of the ratio of p85 monomers to p85-p110 heterodimers might produce a range of RTK-mediated PI3K activity (Figure 4.1). When p85 is present in excess of p110, p85 monomers might negatively regulate RTK- mediated PI3K output. According to this model, PI3K signaling downstream of RTKs would be maximal when p85 isoform expression is reduced exactly to the level where the number of p85 subunits is equal to the number of p110 subunits; further reduction in p85 expression would decrease the number of signaling-capable p85-p110 heterodimers, reducing RTK-mediated PI3K signaling. Consistent with this idea, others have shown that ablation of both Pik3r1 and Pik3r2 significantly impairs PI3K lipid kinase activity and AKT phosphorylation in mouse cardiac tissue (Luo et al., 2005b). In a separate study using a GEMM of KRAS-driven lung cancer, Pik3r1+/-; Pik3r2-/- mice developed a greater number of lung tumors in comparison to Pik3r2-/- mice, while tumor development was nearly completely blocked in Pik3r1-/-; Pik3r2-/- mice (Engelman et al.,

2008). These results support the idea that a balance between p85 and p110 isoforms is critical for regulation of PI3K signaling.

If our model is correct, it might also be expected that partial loss of any p85 isoforms could increase RTK-mediated PI3K signaling. We have performed preliminary experiments indicating that in HMECs, RNAi-mediated PIK3R2 knockdown increases

139

140

Figure 4.1: Model: modulation of p85 levels might produce a range of RTK-mediated PI3K activity. Left: When the amount of p85 is very low, p110 levels are concomitantly reduced, since p85 is required for p110 stability; few p85-p110 heterodimers are available for binding to activated RTKs. In this context, RTK-mediated PI3K signaling might be minimal. Middle: When p85 and p110 are present in equal amounts, there is no monomeric p85 to compete with heterodimeric p85-p110 for binding to activated RTKs; all binding sites can be occupied by signaling p85-p110 heterodimers. In this context, RTK-mediated PI3K signaling might be maximal. Right: When p85 levels are very high, p85 regulatory isoforms may be in excess of p110 catalytic isoforms; both monomeric p85 and heterodimeric p85-p110 compete for binding to activated RTKs, but only p85-p110 heterodimers can signal. In this context, RTK- mediated PI3K signaling might be low. In all cases, while p85 levels are expected to similarly modulate RTK binding of p110- and p110-containing PI3K heterodimers, p110 is the main catalytic isoform mediating PI3K activation downstream of RTK inputs (Utermark et al., 2012).

growth factor-stimulated PI3K/AKT activation and anchorage-independent growth (data not shown). Others have shown that Pik3r2-/- mice display increased insulin sensitivity, and cells derived from these mice showed heightened insulin-stimulated AKT phosphorylation (Ueki et al., 2003; Ueki et al., 2002b). However, two recent publications demonstrated that PIK3R2 overexpression in chicken embryo fibroblasts (CEFs) or immortalized murine fibroblasts (NIH 3T3) increases PI3K/AKT activation and cellular transformation (Cortes et al., 2012; Ito et al., 2014). Furthermore, these studies suggested that PIK3R2 is overexpressed in human breast, colon, and ovarian cancers

(Cortes et al., 2012; Ito et al., 2014). It has been proposed that p85-p110 has greater kinase activity towards PtdIns(3,4,5)P3 than p85-p110 (Cortes et al., 2012) or that p85 is a less effective inhibitor of p110 catalytic activity (Ito et al., 2014). We queried https://www.oncomine.org and http://www.cbioportal.org, the same online databases we used to detect significant PIK3R1 underexpression in breast cancers (Figure 2.1 and

Table 2.1), to determine whether PIK3R2 alterations were a common event in human cancers. However, we did not identify a significant trend towards either overexpression or underexpression of PIK3R2 (data not shown). We conclude that further study is needed to clarify the role of PIK3R2 expression in transformation.

Our model also relies on an imbalance in the number of class IA p85 regulatory and p110 catalytic subunits. The presence of free p85 is currently a source of controversy in the field, with publications arguing both for and against the existence of p85 monomers.

One study used quantitative mass spectrometry to determine the absolute number of p85 and p110 molecules in murine B lymphocytes (WEHI-231), murine fibroblasts (NIH

3T3), and various homogenized mouse tissues (Geering et al., 2007). This work demonstrated that in WEHI-231 and NIH 3T3 cells, and in mouse muscle, liver, fat, and spleen tissue, p85 and p110 are present in approximately equal amounts; only in the

141

mouse brain was p85 found to be present in excess (Geering et al., 2007). In this study, it was concluded that class IA PI3Ks are obligate p85-p110 heterodimers.

In contrast, two studies from the Cantley and Kahn groups used immunodepletion techniques on wildtype and Pik3r1 ablated mouse cells and tissues to demonstrate the presence of p85 monomers (Mauvais-Jarvis et al., 2002; Ueki et al., 2002a). From these results it was estimated that the ratio of p85-p110 dimers to p85 monomers was 2:1 in wildtype MEFs, 3:1 in heterozygous Pik3r1 knockout MEFs, and 7:1 in homozygous

Pik3r1 knockout MEFs (Ueki et al., 2002a); in mouse livers, the ratio of p85-p110 dimers to p85 monomers was estimated to be 2:1 in wildtype mice and 4:1 in heterozygous

Pik3r1 knockout mice (Mauvais-Jarvis et al., 2002). However, none of these studies examined mammary cells or tissues. We used similar immunodepletion techniques in our HMECs, but while these results suggested that p85 might be in excess in these cells, our findings were inconclusive (data not shown). It is possible that the relative levels of p85 and p110 may vary depending on the cell type or tissue. In fact, in preliminary studies we found that RNAi-mediated downregulation of p85 in MCF10A human mammary epithelial cells or in HC11 murine mammary epithelial cells increased

PI3K/AKT signaling in response to EGF or insulin, while Pik3r1 knockdown in wildtype

MEFs had no effect on growth factor-stimulated PI3K/AKT activation (data not shown). It will be critical to assess the relative numbers of p85 and p110 isoforms in different cell and tissue types using sensitive techniques such as quantitative mass spectrometry.

If correct, our model also has several implications for RTKs relative to PI3K subunits.

First, the ratio of available RTKs to PI3K molecules will be important. It is likely that in mammary epithelial cells, the number of p85 monomers plus p85-p110 heterodimers is in excess in comparison to the number of RTKs, even when Flag-TEL-ErbB3 is

142

ectopically expressed. If instead RTKs were in excess, we would expect that partial reduction of p85 would have no effect on PIγK/AKT signaling, since all p85-p110 heterodimers would already be able to complex with activated RTKs in this instance.

Second, our model makes the assumption that p85 monomers and p85-p110 heterodimers bind to RTKs with approximately the same affinity, allowing monomers and heterodimers to compete for the same active sites. We could begin to address this by performing immunoprecipitations of Flag-TEL-ErbB3 similar to those shown in Chapter

2, and comparing the ratio of p85 to p110 in immunoprecipitates in comparison to the non-immunoprecipitated supernatant; in this experiment, we would expect the ratio of p85 to p110 to be similar in both the immunoprecipitate and supernatant fractions.

However, a more definitive demonstration will require careful biochemical experiments with purified proteins.

Finally, our data indicates that increased PIγK/AKT signaling in contexts of partial p85 reduction may be specific to RTK inputs. Our model predicts that p85 downregulation might only affect PI3K signaling downstream of receptors for which p85 serves as a necessary binding adaptor for the PI3K heterodimer. Growth factor-stimulated PI3K signaling relies on binding of the SH2 domains of p85 subunits to phosphorylated motifs on activated RTKs, recruiting class IA p110 catalytic subunits and activating their lipid kinase activity (Figure 1.2). A number of studies have demonstrated that p110 is the primary catalytic subunit involved in growth factor-stimulated PI3K activation (Foukas et al., 2006; Graupera et al., 2008; Knight et al., 2006; Utermark et al., 2012; Zhao et al.,

2006). Consistent with this, we find that increased signaling and transformation mediated by partial p85 loss is blocked by p110-selective pharmacological inhibition. It is likely that p85 depletion similarly affects both p110 and p110, but since p110 plays a minor role in RTK-mediated PIγK signaling, effects on p110 are difficult to detect.

143

Interestingly, of the class IA catalytic isoforms only p110 has been shown to also mediate GPCR-activated PI3K signaling (Ciraolo et al., 2008; Guillermet-Guibert et al.,

2008; Jia et al., 2008). A recent study demonstrated that a unique region of p110 directly binds GPCR-associated G subunits, facilitating p110-mediated PI3K signaling downstream of GPCRs (Figure 1.5) (Dbouk et al., 2012). This suggests that p85 subunits may not function as adaptors for PI3K signaling downstream of GPCRs, consistent with our data demonstrating that p110-selective inhibition does not suppress

PI3K/AKT signaling and transformation brought about by a partial reduction in p85. It may be that p85 downregulation uniquely synergizes with RTK-activating signals. It will be interesting to test this idea using GPCR agonists such as LPA to stimulate our control and PIK3R1 knockdown HMECs. If correct, we would expect to see no difference in

LPA-stimulated PI3K/AKT activation between these cell lines.

Implications for therapeutic targeting of PI3K in cancers with p85 loss

The discovery of frequent PI3K activation in cancers has made this pathway an attractive target for small-molecule therapeutics. The first PI3K inhibitors to enter clinical trials were pan-PI3K inhibitors (Table 1.2) and dual pan-PI3K/mTOR inhibitors (Table

1.3). Unfortunately, many of these drugs have had only modest single-agent success in the clinic (Rodon et al., 2013). This is in part due to compensatory feedback signaling networks which facilitate the development of resistance to targeted therapies; one major focus in the field is to identify these resistance pathways and develop rational combination therapy strategies to circumvent them (Figure 1.8). Another limiting aspect of many pan-PI3K inhibitors is their broad spectrum of off-target effects on PI3K-related kinases and other cellular components (Fruman and Rommel, 2014). This is likely because inhibitors targeting the active site of all class I PI3Ks will necessarily be promiscuous (Knight and Shokat, 2005). Isoform-selective PI3K inhibitors are now

144

emerging in clinical trials (Table 1.4), and should theoretically have both fewer off-target effects on molecules outside the PI3K family, and fewer on-target toxicities by sparing

PI3K isoforms that are not contributing to tumorigenesis. Some p110-selective inhibitors have shown promising success in early clinical trials (Juric et al., 2013b), and the p110-selective inhibitor idelalisib has recently become the first FDA-approved PI3K therapy due to its significant success in patients with B cell malignancies. Thus another major focus of the field is to identify the contexts in which isoform-selective PI3K inhibitors will be successful (Figure 1.7).

Using isoform-selective PI3K inhibitors, we demonstrate that the transformation of mammary epithelial cells with partial p85 loss is primarily mediated by catalytic isoform p110. Either the pan-PI3K inhibitor GDC0941 or the p110-selective inhibitor BYL719 effectively inhibited colony formation of PIK3R1 knockdown DDp53-HMECs, while p110-selective inhibition had no effect (Figure 2.8). GDC0941 and BYL719 also substantially reduced PI3K/AKT activation in and colony formation of DDp53-HMECs expressing oncogenic HER2/neu in conjunction with knockdown of PIK3R1 (Figure 2.9 and Figure 2.10). Pan-PIγK or p110-selective pharmacological inhibition of p110 blocked the in vivo growth of transplanted tumors with Pik3r1 ablation, while the p110- selective inhibitor KIN193 did not affect tumor growth (Figure 3.13). IHC and immunoblot analysis of tumor tissue revealed that GDC0941 or BYL719 treatment suppressed PI3K/AKT pathway activation in Pik3r1 knockout tumors, while the effect of

KIN193 treatment on PI3K/AKT signaling was indistinguishable from that of the vehicle control (Figure 3.14). Pan-PIγK and p110-selective inhibition also significantly reduced cellular proliferation and increased the percentage of cells undergoing apoptosis in transplanted tumors (Figure 3.15). These results demonstrate that in mammary epithelial cells with partial p85 loss, the efficacy of p110-selective pharmacological

145

inhibition at blocking transformation and tumorigenesis is comparable to that of a pan-

PI3K agent.

It is perhaps unsurprising that transformation in this context is governed by PI3K signaling through p110. If our model for increased PIγK signaling mediated by partial p85 loss (Figure 2.18) is correct, downregulation of p85 should only augment PIγK signaling through RTKs, and not through other inputs such as GPCRs. A number of studies using pharmacological inhibition and genetic inactivation or ablation have demonstrated that PI3K signaling in response to growth factors is principally through p110 (Foukas et al., 2006; Graupera et al., 2008; Knight et al., 2006; Sopasakis et al.,

2010; Utermark et al., 2012; Zhao et al., 2006) and not p110 (Ciraolo et al., 2008;

Guillermet-Guibert et al., 2008; Jia et al., 2008). Furthermore, Pik3ca ablation or p110- selective inhibition was sufficient to block mouse mammary tumorigenesis driven by

HER2/neu in a recent study from our group (Utermark et al., 2012). These findings were explained by a proposed model where p110 may have higher RTK-associated lipid kinase activity than p110, making p110 the primary catalytic isoform mediating PI3K signaling downstream of RTKs (Utermark et al., 2012) (Figure 1.6). Based on this model, oncogenic lesions activating RTK signaling will be effectively targeted by p110- selective inhibitors (Figure 1.7).

Early clinical data has reported promising activity of p110-selective agents BYL719 or

GDC00γβ in advanced breast tumors with p110 activation (Juric et al., 2013b). The data presented in this dissertation suggests that such inhibitors may also be successful in treatment of breast tumors with reduced expression of PIK3R1. Furthermore, another preclinical study has demonstrated that p110- but not p110-selective inhibitors block both PI3K/AKT signaling in and cellular transformation of cells with ectopic expression of

146

cancer-associated p85 mutants (Sun et al., 2010). Together this work highlights the potential for use of p110-selective agents in treatment of cancers with p85 loss or mutation. Since loss of PIK3R1 expression is common in breast cancers (Figure 2.1 and

Table 2.1) (Cizkova et al., 2013) and certain other cancer types (Taniguchi et al., 2010), and PIK3R1 mutation is a frequent event particularly in endometrial and pancreatic cancers (Table 1.1 and Appendix A), our findings along with other recent publications emphasize the need to consider alterations in PIK3R1 as a diagnostic marker in cancers, and underline the importance of evaluating p110-selective agents in the treatment of cancers with PIK3R1 alterations in a clinical setting.

Potential role of differential p85 expression in various metabolic contexts

The work presented in this dissertation supports the concept that modulation of p85 levels could provide a mechanism to fine-tune activation of the PI3K/AKT pathway

(Figure 4.1). We and others have shown that reduced p85 expression leads to enhanced growth factor-stimulated PI3K/AKT activation (Fruman et al., 2000; Mauvais-

Jarvis et al., 2002; Taniguchi et al., 2010; Terauchi et al., 1999; Ueki et al., 2002a), while several publications have demonstrated that increased p85 expression reduces

PI3K/AKT output (Barbour et al., 2005; Luo et al., 2005a; Ueki et al., 2000). Together these findings suggest that differential p85 expression could serve as a physiological mechanism to control the extent of PI3K/AKT signaling.

One potential application for PIγK/AKT pathway regulation by p85 levels might be in different metabolic contexts. Under conditions of glucose elevation, for example as a result of ingested nutrients, reduced p85 expression could increase sensitivity to insulin and reduce blood glucose levels. Conversely, in conditions of reduced glucose, such as during nutrient deprivation, elevated p85 levels could reduce insulin sensitivity and

147

favor fatty acid metabolism. Consistent with these ideas, one publication found that in

GEMMs of genetic insulin resistance via heterozygous deletion of IR or IRS1, heterozygous Pik3r1 ablation improved insulin sensitivity and glucose homeostasis, and protected mice from the development of diabetes (Mauvais-Jarvis et al., 2002). A second study demonstrated that Pik3r1 expression was highly induced in adipose tissue from mice with obesity induced by a high fat diet; heterozygous Pik3r1 ablation preserved insulin sensitivity in these mice (McCurdy et al., 2012). Thus it will be especially interesting to explore whether p85 expression levels are dynamic, particularly in those tissues critical for metabolism, depending on nutrient conditions and disease states.

It will also be important to identify the mechanisms governing p85 expression. Several recent publications have identified microRNAs (miRNAs) which functionally regulate

Pik3r1 expression (Huang et al., 2014; Huang et al., 2012; Nicoli et al., 2012; Peng et al., 2013; Tian et al., 2013; Zheng et al., 2012). A separate study reported PIK3R1 transcriptional upregulation in adipocytes by direct binding of peroxisome proliferator- activated receptor gamma (PPAR) to two peroxisome proliferator response elements

(PPREs) in the PIK3R1 promoter (Kim et al., 2014). However, the regulation of PIK3R1 expression remains an underexplored area of study. If dynamic changes in p85 levels in fact contribute to physiological processes such as regulation of metabolism, it will be critical to identify the different mechanisms which control expression of PIK3R1.

Conclusions and perspective

A tremendous amount of work using RNAi-mediated gene silencing, genetically engineered mouse models, and emerging isoform-selective pharmacological inhibitors has begun to elucidate the distinct roles of PI3K catalytic isoforms in different signaling contexts. Large-scale sequencing efforts in the past decade have also identified

148

frequently occurring oncogenic mutations in catalytic isoform p110 in a range of human cancers; catalytic isoforms p110, p110, and p110 are rarely mutated but can be overexpressed. While the roles of PI3K catalytic isoforms in signaling and cancer are beginning to be understood, the roles of the regulatory isoforms are less well studied.

The work presented in this dissertation adds to a growing number of recent publications indicating that PIγK regulatory isoform p85 may also contribute to tumorigenesis.

Frequent mutations in PIK3R1 have been identified in certain cancer types; studies have shown that some of these cancer-associated p85 mutants can still bind but not inhibit p110, leading to enhanced PI3K/AKT activation. One preclinical study has also indicated a potential role for p85 as a tumor suppressor in the liver. Here we show that partial reduction of p85 increases PIγK/AKT signaling in and transformation of human mammary epithelial cells in vitro and contributes to mammary tumor formation in vivo.

Our findings are consistent with a model in which excess monomeric p85 competes with p85-p110 heterodimers to negatively regulate PI3K signaling downstream of RTKs. This work begins to address the role of p85 in different physiological and pathophysiological

PI3K signaling contexts, and highlights the potential for PIK3R1 mutation or underexpression as a diagnostic and therapeutic marker for breast cancer.

149

Materials and Methods

Publically available clinical data

The Oncomine database version 4.4.4.3 (https://www.oncomine.org/) was queried for expression of PIK3R1 in breast cancer microarray studies. The data from each study was converted to raw expression levels of PIK3R1 by taking the inverse log2, and then normalized to the mean raw PIK3R1 expression level of the normal breast tissue from that specific study.

Vectors

Lentiviral pLKO.1-puromycin shRNA vectors were obtained from the RNAi Consortium at the Broad Institute, Cambridge, MA. The sequences for these vectors were as follows: shControl scrambled vector 5’-TCC TAA GGT TAA GTC GCC CTC G-γ’, shPIKγR1 #1

(TRCN0000039903, targeting the γ’ UTR of PIK3R1) 5’-GCG CTA TGC AAT TCT TAA

TTT-γ’, and shPIK3R1 #2 (TRCN0000033284, targeting the coding sequence of

PIK3R1) 5’-CCT TCA GTT CTG TGG TTG AAT-γ’. Retroviral vectors were as follows: pWZL-blasticidin-p53DD, pBabe-neomycin-p53DD, pBabe-blasticidin-neuT, pWZL- neomycin-HA-PIK3CAH1047R, pBabe-puromycin-HA-PIK3CAH1047R, and pWZL-neomycin-

Flag-tel-ErbB3. To generate constructs for rescue of PIK3R1 knockdown, a pCMV6 entry vector containing the cDNA ORF of wildtype human PIK3R1 with C-terminal Myc and Flag tags was obtained from Origene (RC210544) and cloned into the pWZL- blasticidin retroviral vector. This vector was used for rescue of shPIK3R1 #1. For rescue of shPIK3R1 #2, the QuikChange II XL site-directed mutagenesis kit (Agilent) was used to make wobble point mutations T1197C and G1203A with primers 5'-CTC TGA CCC

ATT AAC CTT CAG CTC TGT AGT TGA ATT AAT AAA CCA CTA CC-3' and 5’-GGT

AGT GGT TTA TTA ATT CAA CTA CAG AGC TGA AGG TTA ATG GGT CAG AG-γ’.

151

Cell culture and transduction hTERT-immortalized HMECs were cultured in HMEC Growth Medium (DMEM/F12

GlutaMAX [Gibco] supplemented with 0.01g/ml EGF [Sigma], 10g/ml insulin [Gibco],

0.0β5g/ml hydrocortisone [Sigma], 1ng/ml cholera toxin [Sigma], 0.6% FBS [Gibco], penicillin/streptomycin [Gibco], and antimycotic [Gibco]). 239T and phoenix cells used for virus production were cultured in DMEM 10% FBS (DMEM supplemented with 5% FBS, penicillin/streptomycin, and antimycotic, all from Gibco). All cells were maintained at

37°C and 5% CO2.

VSV lentivirus was produced by transfecting 293T cells with the pMD2-VSV-G and pCMV-R8.91 packaging vectors along with a pLKO.1 vector encoding the shRNA of interest. Retrovirus was produced by transfecting phoenix cells with the retroviral vector encoding the gene of interest. All transfections were carried out for 20 minutes at room temperature using the FuGene6 transfection reagent (Promega) in Opti-MEM reduced serum media (Gibco). Virus was harvested by passing culture supernatants through a

0.45m filter β-3 days post-transfection.

Stable HMEC lines were generated by infecting cells with lentiviral or retroviral supernatants in the presence of 4g/ml polybrene (Milipore) overnight. After infection, successfully transduced stable polyclonal lines selected for several days in HMEC

Growth Medium containing the appropriate antibiotic (β50g/ml neomycin [Gibco],

β.5g/ml blasticidin [Invitrogen], or 1.5g/ml puromycin [Calbiochem]) until a control plate of non-transduced cells were completely killed.

Growth factor stimulation timecourse assays

Cells were rinsed twice with PBS (Gibco) and starved in HMEC Starvation Medium

152

(DMEM/F12 GlutaMAX with penicillin/streptomycin and antimycotic [all from Gibco]) for either 4 hours or overnight as indicated. For EGF timecourses, human recombinant EGF

(Sigma) was prepared in HMEC Starvation Medium to a final concentration of 20mg/ml, and used to stimulate cells at 37°C and 5% CO2 for the indicated amounts of time.

Protein lysate preparation and immunoblotting

All protein lysates were prepared by scraping plates of cells on ice using NP40 lysis buffer (137mM NaCl, 20mM Tris-HCl [pH 8.0], 0.2mM EDTA, 10% glycerol, 1% NP40) with protease inhibitors (Roche) and phosphatase inhibitors (Thermo Scientific). Whole cell lysates were prepared by taking the supernatant following centrifugation at

14000rpm and 4°C. Protein concentrations of total cell lysates were determined by

Bradford assay (Bio-Rad) and then 4X SDS/DTT sample buffer (40% glycerol, 250mM

Tris-HCl [pH 6.8], 8% SDS, 0.04% bromophenol blue) was added to a final concentration of 1X. Samples were boiled at 100°C for 10 minutes and stored at -80°C.

For immunoblotting, proteins in the samples were separated by SDS-PAGE on 8%,

10%, or 1β% polyacrylamide gels and electrotransferred onto 0.45m NitroBind nitrocellulose membranes (Maine Manufacturing). Membranes were blocked for 1 hour at room temperature in a solution of 5% nonfat dry milk (Bio-Rad) in TBS (50mM Tris-

HCl [pH 7.5], 150mM NaCl). Primary antibodies were diluted in a solution of 5% BSA

(Research Products International) in TBST (50mM Tris-HCl [pH 7.5], 150mM NaCl,

0.15% Tween-20) and incubated with membranes at 4°C overnight. The following primary antibodies were used for immunoblotting at the dilutions specified: vinculin

(Sigma V9131 1:10000), pan-p85 (Millipore Absβγ4 1:600), p85 (Millipore 05β1β

1:1000), p85 (Millipore 0440γ 1:600), p85 (Santa Cruz sc569γ4 1:100), p110 (Cell

Signaling 4β49 1:1000), p110 (Cell Signaling γ011 1:1000), p110 (Santa Cruz sc60β

153

1:100), Flag (Origene TA50011 1:1000), Flag (Sigma F1804 1:1000), HA (Cell Signaling

2367 1:1000), ErbB2 (Calbiochem OP15 1:100), ErbB3 (Cell Signaling 4754 1:1000),

PTEN (Cell Signaling 9552 1:1000), phospho-AKTT308 (Cell Signaling 4056 1:1000), phospho-AKTS473 (Cell Signaling 9271 1:1000), total AKT (Cell Signaling 9271 1:1000), phospho-ERKT202/Y204 (Cell Signaling 9101 1:1000), total ERK (Cell Signaling 9102

1:1000), phospho-S6S235/236 (Cell Signaling 2211 1:1000), total S6 (Cell Signaling 2217

1:1000), phospho-EGFRY1068 (Cell Signaling 3777 1:1000), and EGFR (Cell Signaling

4267 1:1000). Anti-mouse IgG IRDye800-conjugated (Rockland) and anti-rabbit IgG

AlexaFluor680-conjugated (Invitrogen) secondary antibodies were used at a dilution of

1:5000 in a solution of 1.36% nonfat dry milk (Bio-Rad) and 0.01% SDS (Invitrogen) in

TBST, and were incubated with membranes for 1 hour at room temperature. Fluorescent protein signals were detected and quantified using a LI-COR Odyssey CLx imaging system and the accompanying ImageStudio software, version 3.1.4.

Anchorage-independent growth assays

Single-cell suspensions were plated in a solution of 0.3% agar in HMEC Growth Medium on top of a base layer of 0.6% agar in DMEM (Gibco). To prevent agar from drying out,

HMEC Growth Medium with or without PI3K inhibitors was added fresh every 3 days.

For HMECs stably expressing neuT, 2.5x104 viable cells were plated per 60mm dish or

4.5x103 viable cells per well of a 12-well plate, and grown for 3 weeks at 37°C and 5%

4 3 CO2; for all other HMEC lines, 5.0x10 viable cells were plated per 60mm dish or 9.0x10 viable cells per well of a 12-well plate, and grown for 4 weeks at 37°C and 5% CO2.

Pictures of unstained colonies were taken at 2X magnification using a Nikon SMZ-U dissecting microscope with a SPOT Flex 15.2 64Mp Shifting Pixel camera (Diagnostic

Instruments) and the accompanying SPOT advanced software, version 4.5.9.1. Plates were then stained overnight at 37°C and 5% CO2 with a solution of 0.5mg/ml

154

iodonitrotetrazolium chloride (Sigma) in HMEC Growth Medium; pictures of stained plates were taken using visualized using an AlphaImager EP transilluminator (Alpha

Innotech) and the accompanying AlphaView software, version 1.3.0.7.

Proliferation assays

Single-cell suspensions in HMEC Growth Medium were generated, and 1.25x104 viable cells were plated per well of 24-well plates and allowed to attach overnight. For cells grown under minimal growth conditions, wells were washed once with PBS (Gibco) and then given Minimal Growth Medium (HMEC Growth Medium with 0.5% of the normal growth factor and serum supplements). Fresh HMEC Growth Medium or Minimal Growth

Medium was given every other day. At each time point, wells were washed once with

PBS, cells were fixed for 15 minutes at room temperature with 10% phosphate-buffered formalin (Fisher), washed twice with ddH2O, stained for 30 minutes with 0.1% crystal violet (Sigma), washed three times with ddH2O, and allowed to dry completely. Cell- associated dye was extracted with 1ml of 10% acetic acid per well, and optical density at

590nm was read using a Benchmark Plus microplate spectrophotometer (Bio-Rad) and the accompanying Microplate Manager III software, version 1.133 for Mac OSX.

Immunoprecipitation assays

For immunoprecipitation of Flag-ErbB3 from HMECs, cells were rinsed twice with PBS and starved for 4 hours in HMEC Starvation Medium before being lysed in NP40 lysis buffer. For each reaction, 40l of anti-Flag beads (Sigma) were washed 3 times with

NP40 lysis buffer, and then mixed with 1mg total protein from the appropriate whole cell lysate in a total volume of 500l NP40 lysis buffer. Reactions were carried out on a rotator at 4°C for 1 hour. Beads were washed γ times with β50l NP40 lysis buffer and then resuspended in 2X SDS/DTT sample buffer and boiled at 100°C for 10 minutes.

155

For immunoprecipitation of endogenous ErbB3 from cultured mouse mammary tumor cells, cells were rinsed twice with PBS and starved overnight in DMEM containing penicillin/streptomycin and antimycotic before being lysed in NP40 lysis buffer. For each reaction, 500g total protein from the appropriate whole cell lysate was mixed with 4g anti-ErbB3 antibody (Millipore 05-390) and incubated on a rotator at 4°C for 1 hour.

Then, 40l of Protein A/G beads (Santa Cruz) were added and reactions were carried out on a rotator at 4°C for an additional 1 hour. The beads were washed 3 times with

β50l NP40 lysis buffer and then resuspended in βX SDS/DTT sample buffer and boiled at 100°C for 10 minutes.

For immunoprecipitation of phospho-tyrosine-containing proteins, HMECs were rinsed twice with PBS and given HMEC Starvation Medium for 4 hours before being lysed in

NP40 lysis buffer. Each reaction consisted of 10l 4G10 anti-phospho-tyrosine sepharose bead slurry (Millipore) and 100g total protein from the appropriate whole cell lysate in a final volume of β50l. Reactions were carried out on a rotator overnight at

4°C, and then the beads were washed γ times with β50l NP40 lysis buffer and then resuspended in 2X SDS/DTT sample buffer and boiled at 100°C for 10 minutes.

PTEN lipid phosphatase activity assay

HMECs were grown to confluency in 15cm tissue culture dishes. Asynchronous cells were lysed in NP40 lysis buffer. Endogenous PTEN was immunoprecipitated from cell lysates in triplicate reactions. For “Beads Only” reaction, 2mg total protein from shControl whole cell lysate was diluted to a final volume of 400l in NP40 lysis buffer.

For “PTEN Only” reaction, 2mg total protein from shControl whole cell lysate was mixed with 8l anti-PTEN antibody (Cell Signaling 9559) and diluted to a final volume of 400l in NP40 lysis buffer. For all other reactions, 2mg total protein from the appropriate whole

156

cell lysate was mixed with 8l anti-PTEN antibody (Cell Signaling 9559) and diluted to a final volume of 400l in NP40 lysis buffer. All tubes were incubated on a rotator overnight at 4°C. Then, 40l of Protein A/G beads (Santa Cruz) were added to each tube, and reactions were incubated on a rotator for 3 hours at 4°C. The beads were washed 3 times with 400l NP40 lysis buffer and then washed once with PTEN Reaction

Buffer (TBS [25 mM Tris-HCl (pH 7.5), 140mM NaCl, 2.7mM KCl] with 10mM DTT added fresh). The beads were resuspended in γ0l PTEN Reaction Buffer and used in in vitro

PTEN lipid phosphatase assays (Echelon Biosciences, method adapted from provided documents and (Song et al., 2011)). Briefly, phosphate standards from 0-2,000pmol were prepared using provided reagents. Then βl of 1mM diC8PtdIns(3,4,5)P3 was added to each of the standards, a “PIPγ Substrate Only” control (consisting of PTEN

Reaction Buffer only), and the immunoprecipitation reactions and accompanying controls, and brought to a final volume of 50l with PTEN Reaction Buffer. Reactions were incubated at γ7°C for γ0 minutes, and then 100l Malachite Green Reagent was added to each reaction and incubated for 10 minutes at room temperature. The optical density at 620nm, corresponding to phosphate released by PTEN enzymatic activity, was read using a Benchmark Plus microplate spectrophotometer (Bio-Rad) and the accompanying Microplate Manager III software, version 1.133 for Mac OSX.

Receptor internalization and degradation assays

To track internalization of EGFR, HMECs were rinsed with PBS and starved overnight in

HMEC Starvation Medium. Plates were washed with PBS-CM (PBS with 1mM MgCl2 and 0.1mM CaCl2 added), then incubated on a rocker for 40 minutes at 4°C with 5ml

PBS-CM containing 0.5mg/ml Sulfo-NHS-SS-Biotin (Pierce) to label surface proteins.

Plates were washed with PBS-CM, then incubated on a rocker for 10 minutes at 4°C with 10ml PBS-CM containing 50mM NH4Cl (Sigma) to inactivate excess unbound sulfo-

157

NHS-SS-biotin. Plates were washed with PBS-CM. One plate per cell line was then lysed with TritonX100 Lysis Buffer (200mM NaCl, 75mM Tris-HCl [pH 7.5], 15mM NaF,

2.5mM EDTA, 2.5mM EGTA, 1.5% TritonX100, 0.75% NP40, 0.1% SDS) for determination of total surface EGFR. The remaining plates were stimulated with

20mg/ml human recombinant EGF (Sigma) in HMEC Starvation Medium at 37°C and 5%

CO2 for the indicated amounts of time to induce receptor internalization. At the appropriate time, plates were rinsed with PBS-CM, washed twice with 10ml Glutathione

Buffer (90mM NaCl, 1mM MgCl2, 0.1mM CaCl2, 50mM reduced glutathione, 60mM

NaOH) to cleave any sulfo-NHS-SS-biotin remaining on the cell surface, and rinsed with

10ml PBS-IAA (50mM iodoacetamide in PBS-CM) on a rocker at 4°C for 15 minutes, to quench excess glutathione. After rinsing with PBS-CM, plates were lysed with

TritonX100 lysis buffer and protein samples were prepared as described above. A portion of these samples was reserved for whole cell lysates, and the rest was used in

Streptavidin immunoprecipitation. First, 1mg total protein per sample was pre-cleared by incubating with pansorbin (Calbiochem) on a rotator at 4°C for 1 hour. The pansorbin was removed by centrifugation, and 40ul Streptavidin beads (Pierce) were added for overnight immunoprecipitation reactions. Beads were rinsed 4 times with TritonX100 lysis buffer before being boiled in 2X SDS/DTT sample buffer for 10 minutes. Samples were then subjected to electrophoresis and immunoblotting as described above.

To track degradation of EGFR, HMECs were rinsed with PBS and starved overnight in

HMEC Starvation Medium. Plates were washed with PBS-CM, then incubated on a rocker for 40 minutes at 4°C with 5ml PBS-CM containing 0.5mg/ml Sulfo-NHS-LC-LC-

Biotin (Pierce) to label surface proteins. Plates were washed with PBS-CM, then incubated on a rocker for 10 minutes at 4°C with 10ml PBS-CM containing 50mM NH4Cl

(Sigma) to inactivate excess unbound sulfo-NHS-LC-LC-biotin. Plates were washed with

158

PBS-CM. One plate per cell line was then lysed with TritonX100 Lysis Buffer for determination of total surface EGFR. The remaining plates were stimulated with

20mg/ml human recombinant EGF (Sigma) in HMEC Starvation Medium at 37°C and 5%

CO2 for the indicated amounts of time to induce receptor internalization. At the appropriate time, plates were rinsed with PBS-CM and lysed with TritonX100 lysis buffer, and protein samples were prepared as described above. A portion of these samples was reserved for whole cell lysates, and the rest was used in Streptavidin immunoprecipitation. First, 1mg total protein per sample was pre-cleared by incubating with pansorbin (Calbiochem) on a rotator at 4°C for 1 hour. The pansorbin was removed by centrifugation, and 40ul Streptavidin beads (Pierce) were added for overnight immunoprecipitation reactions. Beads were rinsed 4 times with TritonX100 lysis buffer before being boiled in 2X SDS/DTT sample buffer for 10 minutes. Samples were then subjected to electrophoresis and immunoblotting as described above.

Animal husbandry and breeding strategy

MMTV-Cre (Wagner et al., 1997), NIC (Schade et al., 2009; Ursini-Siegel et al., 2008), and Pik3r1 floxed (Luo et al., 2005b) mice were backcrossed to the FVB/N wildtype background at least 10 generations. To generate the female mice used in these studies, male MMTV-Cre or NIC mice heterozygous for the floxed Pik3r1 allele were crossed with female mice heterozygous for the floxed Pik3r1 allele. For orthotopic tumor transplantations, eight-week-old NCrNu female mice (Harlan) were used. Transplant recipient mice were treated daily with vehicle control (0.5% [w/V] methylcellulose

[Sigma], administered by oral gavage at 1ml/kg body weight), BYL719 (in 0.5% methylcellulose, administered by oral gavage at 45mg/kg), GDC0941 (in 0.5% methylcellulose, administered by oral gavage at 125mg/kg), or KIN193 (in a solution of

7.5% NMP [1-methyl-2-pyrrolidinone, Sigma] and 40% PEG400 [Sigma], administered

159

by intraperitoneal injection at 20mg/kg). All animals were housed and treated in accordance with protocols approved by the Institutional Animal Care and Use

Committees of Dana-Farber Cancer Institute and Harvard Medical School.

Genotyping

Genomic DNA was extracted from 2-3mm of tail tissue by boiling in 150ul of an alkaline buffer containing 2mM EDTA and 25mM NaOH at 100C for 60 minutes, followed by neutralization with 150ul of a buffer containing 40mM Tris Base. PCR was then performed on the DNA extracts using GoTaq DNA polymerase (Promega) as follows for each gene: Pik3r1, primers 5'-CAC CGA GCA CTG GAG CAC TG-3' and 5'-CCA GTT

ACT TTC AAA TCA GCA CAG-3', wildtype Pik3r1 allele generates a fragment of 252bp, while floxed Pik3r1 allele generates a 301bp fragment; NIC transgene, primers 5'-TTC

CGG AAC CCA CAT CAG GCC-3' and 5'-GTT TCC TGC AGC AGC CTA CGC-3', transgene generates a 630bp fragment; MMTV-Cre transgene, primers 5’-CTG ATC

TGA GCT CTG AGT G-γ’, 5’-CAT CAC TCG TTG CAT CGA CC-γ’, transgene generates a 250bp fragment. PCR samples were then resolved by electrophoresis through a 2% agarose gel with SYBR Safe (Invitrogen) and visualized under UV light using an AlphaImager EP transilluminator (Alpha Innotech) and the accompanying

AlphaView software, version 1.3.0.7.

Mammary whole mount preparation

The fourth mammary gland tissue was excised, spread on glass slides, and fixed overnight in Carnoy’s Fixative (a 1:γ [V/V] solution of glacial acetic acid and 100% ethanol). Slides were then washed for 30 minutes in 70% ethanol and 30 minutes in ddH2O, and then stained overnight in Carmine Alum (1g carmine [Sigma] dissolved in

500ml ddH2O). Slides were then washed for 30 minutes in 70% ethanol, 30 minutes in

160

95% ethanol, and 30 minutes 100% ethanol before being immersed in toluene. All steps were carried out at room temperature. Images of fixed mammary tissue were taken with a Nikon SMZ-U dissecting microscope with a SPOT Flex 15.2 64Mp Shifting Pixel camera (Diagnostic Instruments) and the accompanying SPOT advanced software, version 4.5.9.1.

Mouse mammary epithelial cell isolation

For each mouse, the third and fourth mammary glands were excised, washed in PBS

(Gibco), finely chopped, combined, and digested overnight at 37°C and 5% CO2 in 5ml

Digestion Medium (DMEM/F12 GlutaMAX [Gibco] with penicillin/streptomycin [Gibco] and collagenase/hyaluronidase [Stem Cell Technologies] added to a final concentration of 1X). Following digestion, the tissue was pelleted by centrifugation, the supernatant removed, and the tissue resuspended in Red Blood Cell Lysis Buffer (a 1:4 [V/V] solution of cold HF Solution [Hank’s Balanced Salt Solution (Stem Cell Technologies) with β%

FBS (Gibco) and penicillin/streptomycin (Gibco)] and ammonium chloride [Stem Cell

Technologies]). The tissue was then pelleted by centrifugation, the supernatant removed, and the pellet resuspended in 2ml pre-warmed 0.25% Trypsin-EDTA (Gibco) for 1-3 minutes, followed by addition of 3ml FBS and 10ml cold HF Solution. The tissue was then pelleted by centrifugation, the supernatant removed, the pellet resuspended in

1ml Dispase and 100l DNAseI (both Stem Cell Technologies) for 1 minute followed by addition of 5-10ml cold HF Solution, and then passed through a 40m cell strainer.

Analysis of mammary tumors and lung metastases

Cohorts of female mice were examined every 3 days for the onset of tumors, defined by the first palpation. Five weeks after tumor onset, the mice were sacrificed by CO2 in accordance with protocols approved by the Institutional Animal Care and Use

161

Committees of Dana-Farber Cancer Institute and Harvard Medical School. The total tumor burden per mouse was determined by taking the wet weight of excised mammary tissue and associated tumors. The total number of tumors per mouse was determined by counting the number of distinct excised tumors with a diameter of at least 3mm. To determine the total number of lung metastases per mouse, lungs were excised and fixed overnight in 10% phosphate-buffered formalin (Fisher). All lung lobes were embedded in paraffin, and three sections 50m apart were mounted on slides and stained with hematoxylin and eosin. Metastases were visualized and quantified using a Nikon Eclipse

E600 microscope. For each mouse, the total number of lung metastases was taken to be that of the section with the highest count.

Histology and immunohistochemistry

Tissue was fixed overnight in 10% phosphate-buffered formalin (Fisher) and then transferred to 70% ethanol. Formalin-fixed tissue was embedded in paraffin blocks and mounted on slides by the HMS Rodent Histopathology Core. Hematoxylin and eosin staining was performed by the HMS Rodent Histopathology Core. Immunohistochemical staining of tissue sections was performed using a sodium citrate antigen retrieval method. Slides were deparaffinized in xylene and hydrated in ethanol washes of decreasing concentration. Antigen retrieval was then performed by boiling slides in

10mM sodium citrate (pH 6.0) (Boston BioProducts) for 20 minutes, followed by 30 minutes of cooling. Endogenous peroxidases were blocked by soaking the slides in 3% hydrogen peroxide (Sigma) for 10 minutes at room temperature, and then slides were further blocked by a 1 hour incubation with 5% serum in IHC-TBST (50mM Tris-HCl [pH

7.5], 150mM NaCl, 0.1% Tween-20) at room temperature. Slides were then incubated with primary antibodies overnight. Antibodies from Cell Signaling were diluted in the buffer provided by Cell Signaling, and all other antibodies were diluted in 5% serum in

162

IHC-TBST. The following primary antibodies were used: p85 (EMD Millipore 04403

1:500), phospho-AKTS473 (Cell Signaling 3787 1:50), phospho-ERKT202/Y204 (Cell

Signaling 4376 1:400), phospho-S6S235/236 (Cell Signaling 4858 1:400), and Ki67 (Vector

Laboratories VPK451 1:1000). Slides were then incubated for 30 minutes at room temperature with biotinylated goat anti-rabbit IgG secondary antibody (Vector

Laboratories BA1000 1:250-1:2000) followed by a 30-minute incubation at room temperature with ABC reagent (Vector Laboratories). Antibody signal was then developed by brief incubation with DAB horseradish peroxidase substrate (Vector

Laboratories) and slides were counterstained with hematoxylin (Vector Laboratories).

For TUNEL, the ApopTag Plus Peroxidase In Situ Apoptosis Detection Kit was used

(EMD Millipore) followed by a counterstain with methyl green (Vector Laboratories).

Tissue sections were then dehydrated by washes in increasing concentrations of ethanol followed by washes in xylene. Coverslips were mounted using Cytoseal XYL (Thermo

Scientific) and slides were visualized using a Nikon Eclipse E600 microscope, and images were obtained with a SPOT Flex 15.2 64Mp Shifting Pixel camera (Diagnostic

Instruments) and the accompanying SPOT advanced software, version 4.5.9.1.

Preparation of mouse tissue protein lysates

Tissue pieces were snap frozen in liquid nitrogen. To generate protein lysates, tissue was homogenized in NP40 lysis buffer using 0.5mm zirconium oxide beads (Next

Advance Inc.) in an air-cooling bullet blender (Next Advance). Once homogenized, lysates were cleared and analyzed as described above.

Isolation and culture of mouse mammary tumor cells

Mouse mammary tumor cells were cultured in DMEM 5% FBS (DMEM supplemented with 5% fetal bovine serum, penicillin/streptomycin, and antimycotic, all from Gibco) at

163

3 37°C and 5% CO2. To isolate cells for culture, tumor pieces of approximately 0.5cm were finely chopped, resuspended in 6-8ml DMEM 5% FBS containing 1mg/ml collagenase (Roche), and incubated on a rotator for 20 minutes at 37°C. Cells were then pelleted by a 5 minute centrifugation at 1200rpm, washed with PBS, pelleted again, resuspended in 10ml DMEM 5% FBS, and plated in 10cm tissue culture dishes.

Graphing software and statistical analysis

All graphs and statistical analysis were done using GraphPad Prism 6.0 for Mac OS X.

164

References

Ali, K., Bilancio, A., Thomas, M., Pearce, W., Gilfillan, A.M., Tkaczyk, C., Kuehn, N., Gray, A., Giddings, J., Peskett, E., et al. (2004). Essential role for the p110delta phosphoinositide 3-kinase in the allergic response. Nature 431, 1007-1011.

Ali, K., Soond, D.R., Pineiro, R., Hagemann, T., Pearce, W., Lim, E.L., Bouabe, H., Scudamore, C.L., Hancox, T., Maecker, H., et al. (2014). Inactivation of PI(3)K p110delta breaks regulatory T-cell-mediated immune tolerance to cancer. Nature.

Alimonti, A., Carracedo, A., Clohessy, J.G., Trotman, L.C., Nardella, C., Egia, A., Salmena, L., Sampieri, K., Haveman, W.J., Brogi, E., et al. (2010). Subtle variations in Pten dose determine cancer susceptibility. Nature genetics 42, 454-458.

Andrechek, E.R., Hardy, W.R., Siegel, P.M., Rudnicki, M.A., Cardiff, R.D., and Muller, W.J. (2000). Amplification of the neu/erbB-2 oncogene in a mouse model of mammary tumorigenesis. Proceedings of the National Academy of Sciences of the United States of America 97, 3444-3449.

Angulo, I., Vadas, O., Garcon, F., Banham-Hall, E., Plagnol, V., Leahy, T.R., Baxendale, H., Coulter, T., Curtis, J., Wu, C., et al. (2013). Phosphoinositide 3-kinase delta gene mutation predisposes to respiratory infection and airway damage. Science 342, 866-871.

Backer, J.M. (2008). The regulation and function of Class III PI3Ks: novel roles for Vps34. The Biochemical journal 410, 1-17.

Ballou, L.M., Chattopadhyay, M., Li, Y., Scarlata, S., and Lin, R.Z. (2006). Galphaq binds to p110alpha/p85alpha phosphoinositide 3-kinase and displaces Ras. The Biochemical journal 394, 557-562.

Ballou, L.M., Lin, H.Y., Fan, G., Jiang, Y.P., and Lin, R.Z. (2003). Activated G alpha q inhibits p110 alpha phosphatidylinositol 3-kinase and Akt. The Journal of biological chemistry 278, 23472-23479.

Barbour, L.A., Mizanoor Rahman, S., Gurevich, I., Leitner, J.W., Fischer, S.J., Roper, M.D., Knotts, T.A., Vo, Y., McCurdy, C.E., Yakar, S., et al. (2005). Increased P85alpha is a potent negative regulator of skeletal muscle insulin signaling and induces in vivo insulin resistance associated with growth hormone excess. The Journal of biological chemistry 280, 37489-37494.

Baselga, J., and Swain, S.M. (2009). Novel anticancer targets: revisiting ERBB2 and discovering ERBB3. Nature reviews. Cancer 9, 463-475.

Beeton, C.A., Chance, E.M., Foukas, L.C., and Shepherd, P.R. (2000). Comparison of the kinetic properties of the lipid- and protein-kinase activities of the p110alpha and p110beta catalytic subunits of class-Ia phosphoinositide 3-kinases. The Biochemical journal 350 Pt 2, 353-359.

Berenjeno, I.M., Guillermet-Guibert, J., Pearce, W., Gray, A., Fleming, S., and Vanhaesebroeck, B. (β01β). Both p110 and p110 isoforms of PIγK can modulate the impact of loss-of-function of the PTEN tumour suppressor. The Biochemical journal 442, 151-159.

166

Berger, M.B., Mendrola, J.M., and Lemmon, M.A. (2004). ErbB3/HER3 does not homodimerize upon neuregulin binding at the cell surface. FEBS letters 569, 332-336.

Bi, L., Okabe, I., Bernard, D.J., and Nussbaum, R.L. (2002). Early embryonic lethality in mice deficient in the p110beta catalytic subunit of PI 3-kinase. Mammalian genome : official journal of the International Mammalian Genome Society 13, 169-172.

Bi, L., Okabe, I., Bernard, D.J., Wynshaw-Boris, A., and Nussbaum, R.L. (1999). Proliferative defect and embryonic lethality in mice homozygous for a deletion in the p110alpha subunit of phosphoinositide 3-kinase. The Journal of biological chemistry 274, 10963-10968.

Biswas, K., Yoshioka, K., Asanuma, K., Okamoto, Y., Takuwa, N., Sasaki, T., and Takuwa, Y. (2013). Essential role of class II phosphatidylinositol-3-kinase-C2alpha in sphingosine 1-phosphate receptor-1-mediated signaling and migration in endothelial cells. The Journal of biological chemistry 288, 2325-2339.

Blondeau, F., Laporte, J., Bodin, S., Superti-Furga, G., Payrastre, B., and Mandel, J.L. (2000). Myotubularin, a phosphatase deficient in myotubular myopathy, acts on phosphatidylinositol 3-kinase and phosphatidylinositol 3-phosphate pathway. Human molecular genetics 9, 2223-2229.

Bohnacker, T., Marone, R., Collmann, E., Calvez, R., Hirsch, E., and Wymann, M.P. (2009). PI3Kgamma adaptor subunits define coupling to degranulation and cell motility by distinct PtdIns(3,4,5)P3 pools in mast cells. Science signaling 2, ra27.

Bollag, G., Hirth, P., Tsai, J., Zhang, J., Ibrahim, P.N., Cho, H., Spevak, W., Zhang, C., Zhang, Y., Habets, G., et al. (2010). Clinical efficacy of a RAF inhibitor needs broad target blockade in BRAF-mutant melanoma. Nature 467, 596-599.

Britschgi, A., Andraos, R., Brinkhaus, H., Klebba, I., Romanet, V., Muller, U., Murakami, M., Radimerski, T., and Bentires-Alj, M. (2012). JAK2/STAT5 inhibition circumvents resistance to PI3K/mTOR blockade: a rationale for cotargeting these pathways in metastatic breast cancer. Cancer cell 22, 796-811.

Brock, C., Schaefer, M., Reusch, H.P., Czupalla, C., Michalke, M., Spicher, K., Schultz, G., and Nurnberg, B. (2003). Roles of G beta gamma in membrane recruitment and activation of p110 gamma/p101 phosphoinositide 3-kinase gamma. The Journal of cell biology 160, 89-99.

Burke, J.E., Perisic, O., Masson, G.R., Vadas, O., and Williams, R.L. (2012). Oncogenic mutations mimic and enhance dynamic events in the natural activation of phosphoinositide 3-kinase p110alpha (PIK3CA). Proceedings of the National Academy of Sciences of the United States of America 109, 15259-15264.

Busaidy, N.L., Farooki, A., Dowlati, A., Perentesis, J.P., Dancey, J.E., Doyle, L.A., Brell, J.M., and Siu, L.L. (2012). Management of metabolic effects associated with anticancer agents targeting the PI3K-Akt-mTOR pathway. Journal of clinical oncology : official journal of the American Society of Clinical Oncology 30, 2919-2928.

167

Cancer Genome Atlas, N. (2012). Comprehensive molecular portraits of human breast tumours. Nature 490, 61-70.

Cancer Genome Atlas Research, N. (2008). Comprehensive genomic characterization defines human glioblastoma genes and core pathways. Nature 455, 1061-1068.

Cao, C., Backer, J.M., Laporte, J., Bedrick, E.J., and Wandinger-Ness, A. (2008). Sequential actions of myotubularin lipid phosphatases regulate endosomal PI(3)P and growth factor receptor trafficking. Molecular biology of the cell 19, 3334-3346.

Carroll, M., Tomasson, M.H., Barker, G.F., Golub, T.R., and Gilliland, D.G. (1996). The TEL/platelet-derived growth factor beta receptor (PDGF beta R) fusion in chronic myelomonocytic leukemia is a transforming protein that self-associates and activates PDGF beta R kinase-dependent signaling pathways. Proceedings of the National Academy of Sciences of the United States of America 93, 14845-14850.

Castellano, E., Sheridan, C., Thin, M.Z., Nye, E., Spencer-Dene, B., Diefenbacher, M.E., Moore, C., Kumar, M.S., Murillo, M.M., Gronroos, E., et al. (2013). Requirement for interaction of PI3-kinase p110alpha with RAS in lung tumor maintenance. Cancer cell 24, 617-630.

Cerami, E., Gao, J., Dogrusoz, U., Gross, B.E., Sumer, S.O., Aksoy, B.A., Jacobsen, A., Byrne, C.J., Heuer, M.L., Larsson, E., et al. (2012). The cBio cancer genomics portal: an open platform for exploring multidimensional cancer genomics data. Cancer discovery 2, 401-404.

Chagpar, R.B., Links, P.H., Pastor, M.C., Furber, L.A., Hawrysh, A.D., Chamberlain, M.D., and Anderson, D.H. (2010). Direct positive regulation of PTEN by the p85 subunit of phosphatidylinositol 3-kinase. Proceedings of the National Academy of Sciences of the United States of America 107, 5471-5476.

Chamberlain, M.D., Berry, T.R., Pastor, M.C., and Anderson, D.H. (2004). The p85alpha subunit of phosphatidylinositol 3'-kinase binds to and stimulates the GTPase activity of Rab proteins. The Journal of biological chemistry 279, 48607-48614.

Chamberlain, M.D., Chan, T., Oberg, J.C., Hawrysh, A.D., James, K.M., Saxena, A., Xiang, J., and Anderson, D.H. (2008). Disrupted RabGAP function of the p85 subunit of phosphatidylinositol 3-kinase results in cell transformation. The Journal of biological chemistry 283, 15861-15868.

Chamberlain, M.D., Oberg, J.C., Furber, L.A., Poland, S.F., Hawrysh, A.D., Knafelc, S.M., McBride, H.M., and Anderson, D.H. (2010). Deregulation of Rab5 and Rab4 proteins in p85R274A-expressing cells alters PDGFR trafficking. Cellular signalling 22, 1562-1575.

Chandarlapaty, S., Sawai, A., Scaltriti, M., Rodrik-Outmezguine, V., Grbovic-Huezo, O., Serra, V., Majumder, P.K., Baselga, J., and Rosen, N. (2011). AKT inhibition relieves feedback suppression of receptor tyrosine kinase expression and activity. Cancer cell 19, 58-71.

168

Chang, H.W., Aoki, M., Fruman, D., Auger, K.R., Bellacosa, A., Tsichlis, P.N., Cantley, L.C., Roberts, T.M., and Vogt, P.K. (1997). Transformation of chicken cells by the gene encoding the catalytic subunit of PI 3-kinase. Science 276, 1848-1850.

Chapman, P.B., Hauschild, A., Robert, C., Haanen, J.B., Ascierto, P., Larkin, J., Dummer, R., Garbe, C., Testori, A., Maio, M., et al. (2011). Improved survival with vemurafenib in melanoma with BRAF V600E mutation. The New England journal of medicine 364, 2507-2516.

Chaussade, C., Rewcastle, G.W., Kendall, J.D., Denny, W.A., Cho, K., Gronning, L.M., Chong, M.L., Anagnostou, S.H., Jackson, S.P., Daniele, N., et al. (2007). Evidence for functional redundancy of class IA PI3K isoforms in insulin signalling. The Biochemical journal 404, 449-458.

Chen, D., Mauvais-Jarvis, F., Bluher, M., Fisher, S.J., Jozsi, A., Goodyear, L.J., Ueki, K., and Kahn, C.R. (2004). p50alpha/p55alpha phosphoinositide 3-kinase knockout mice exhibit enhanced insulin sensitivity. Molecular and cellular biology 24, 320-329.

Cheung, L.W., Hennessy, B.T., Li, J., Yu, S., Myers, A.P., Djordjevic, B., Lu, Y., Stemke- Hale, K., Dyer, M.D., Zhang, F., et al. (2011). High frequency of PIK3R1 and PIK3R2 mutations in endometrial cancer elucidates a novel mechanism for regulation of PTEN protein stability. Cancer discovery 1, 170-185.

Chiu, Y.H., Lee, J.Y., and Cantley, L.C. (2014). BRD7, a tumor suppressor, interacts with p85alpha and regulates PI3K activity. Molecular cell 54, 193-202.

Ciraolo, E., Iezzi, M., Marone, R., Marengo, S., Curcio, C., Costa, C., Azzolino, O., Gonella, C., Rubinetto, C., Wu, H., et al. (2008). Phosphoinositide 3-kinase p110beta activity: key role in metabolism and mammary gland cancer but not development. Science signaling 1, ra3.

Ciraolo, E., Morello, F., Hobbs, R.M., Wolf, F., Marone, R., Iezzi, M., Lu, X., Mengozzi, G., Altruda, F., Sorba, G., et al. (2010). Essential role of the p110beta subunit of phosphoinositide 3-OH kinase in male fertility. Molecular biology of the cell 21, 704-711.

Cizkova, M., Vacher, S., Meseure, D., Trassard, M., Susini, A., Mlcuchova, D., Callens, C., Rouleau, E., Spyratos, F., Lidereau, R., et al. (2013). PIK3R1 underexpression is an independent prognostic marker in breast cancer. BMC cancer 13, 545.

Clayton, E., Bardi, G., Bell, S.E., Chantry, D., Downes, C.P., Gray, A., Humphries, L.A., Rawlings, D., Reynolds, H., Vigorito, E., et al. (2002). A crucial role for the p110delta subunit of phosphatidylinositol 3-kinase in B cell development and activation. The Journal of experimental medicine 196, 753-763.

Cortes, I., Sanchez-Ruiz, J., Zuluaga, S., Calvanese, V., Marques, M., Hernandez, C., Rivera, T., Kremer, L., Gonzalez-Garcia, A., and Carrera, A.C. (2012). p85beta phosphoinositide 3-kinase subunit regulates tumor progression. Proceedings of the National Academy of Sciences of the United States of America 109, 11318-11323.

169

Cully, M., You, H., Levine, A.J., and Mak, T.W. (2006). Beyond PTEN mutations: the PI3K pathway as an integrator of multiple inputs during tumorigenesis. Nature reviews. Cancer 6, 184-192.

Davie, S.A., Maglione, J.E., Manner, C.K., Young, D., Cardiff, R.D., MacLeod, C.L., and Ellies, L.G. (2007). Effects of FVB/NJ and C57Bl/6J strain backgrounds on mammary tumor phenotype in inducible nitric oxide synthase deficient mice. Transgenic research 16, 193-201.

Dbouk, H.A., Khalil, B.D., Wu, H., Shymanets, A., Nurnberg, B., and Backer, J.M. (2013). Characterization of a tumor-associated activating mutation of the p110beta PI 3- kinase. PloS one 8, e63833.

Dbouk, H.A., Pang, H., Fiser, A., and Backer, J.M. (2010). A biochemical mechanism for the oncogenic potential of the p110beta catalytic subunit of phosphoinositide 3-kinase. Proceedings of the National Academy of Sciences of the United States of America 107, 19897-19902.

Dbouk, H.A., Vadas, O., Shymanets, A., Burke, J.E., Salamon, R.S., Khalil, B.D., Barrett, M.O., Waldo, G.L., Surve, C., Hsueh, C., et al. (2012). G protein-coupled receptor- mediated activation of p110beta by Gbetagamma is required for cellular transformation and invasiveness. Science signaling 5, ra89.

Delgado, P., Cubelos, B., Calleja, E., Martinez-Martin, N., Cipres, A., Merida, I., Bellas, C., Bustelo, X.R., and Alarcon, B. (2009). Essential function for the GTPase TC21 in homeostatic antigen receptor signaling. Nature immunology 10, 880-888.

Delmore, J.E., Issa, G.C., Lemieux, M.E., Rahl, P.B., Shi, J., Jacobs, H.M., Kastritis, E., Gilpatrick, T., Paranal, R.M., Qi, J., et al. (2011). BET bromodomain inhibition as a therapeutic strategy to target c-Myc. Cell 146, 904-917.

Denley, A., Gymnopoulos, M., Kang, S., Mitchell, C., and Vogt, P.K. (2009). Requirement of phosphatidylinositol(3,4,5)trisphosphate in phosphatidylinositol 3-kinase- induced oncogenic transformation. Molecular cancer research : MCR 7, 1132-1138.

Denley, A., Kang, S., Karst, U., and Vogt, P.K. (2008). Oncogenic signaling of class I PI3K isoforms. Oncogene 27, 2561-2574.

Di Cristofano, A., Pesce, B., Cordon-Cardo, C., and Pandolfi, P.P. (1998). Pten is essential for embryonic development and tumour suppression. Nature genetics 19, 348- 355.

Diouf, B., Cheng, Q., Krynetskaia, N.F., Yang, W., Cheok, M., Pei, D., Fan, Y., Cheng, C., Krynetskiy, E.Y., Geng, H., et al. (2011). Somatic deletions of genes regulating MSH2 protein stability cause DNA mismatch repair deficiency and drug resistance in human leukemia cells. Nature medicine 17, 1298-1303.

Dou, Z., Pan, J.A., Dbouk, H.A., Ballou, L.M., DeLeon, J.L., Fan, Y., Chen, J.S., Liang, Z., Li, G., Backer, J.M., et al. (2013). Class IA PI3K p110beta subunit promotes autophagy through Rab5 small GTPase in response to growth factor limitation. Molecular cell 50, 29-42.

170

Dourdin, N., Schade, B., Lesurf, R., Hallett, M., Munn, R.J., Cardiff, R.D., and Muller, W.J. (2008). Phosphatase and tensin homologue deleted on chromosome 10 deficiency accelerates tumor induction in a mouse model of ErbB-2 mammary tumorigenesis. Cancer research 68, 2122-2131.

Downward, J., Parker, P., and Waterfield, M.D. (1984). Autophosphorylation sites on the epidermal growth factor receptor. Nature 311, 483-485.

Durand, C.A., Hartvigsen, K., Fogelstrand, L., Kim, S., Iritani, S., Vanhaesebroeck, B., Witztum, J.L., Puri, K.D., and Gold, M.R. (2009). Phosphoinositide 3-kinase p110 delta regulates natural antibody production, marginal zone and B-1 B cell function, and autoantibody responses. Journal of immunology 183, 5673-5684.

Elis, W., Triantafellow, E., Wolters, N.M., Sian, K.R., Caponigro, G., Borawski, J., Gaither, L.A., Murphy, L.O., Finan, P.M., and Mackeigan, J.P. (2008). Down-regulation of class II phosphoinositide 3-kinase alpha expression below a critical threshold induces apoptotic cell death. Molecular cancer research : MCR 6, 614-623.

Elkabets, M., Vora, S., Juric, D., Morse, N., Mino-Kenudson, M., Muranen, T., Tao, J., Campos, A.B., Rodon, J., Ibrahim, Y.H., et al. (2013). mTORC1 inhibition is required for sensitivity to PI3K p110alpha inhibitors in PIK3CA-mutant breast cancer. Science translational medicine 5, 196ra199.

Engelman, J.A. (2009). Targeting PI3K signalling in cancer: opportunities, challenges and limitations. Nature reviews. Cancer 9, 550-562.

Engelman, J.A., Chen, L., Tan, X., Crosby, K., Guimaraes, A.R., Upadhyay, R., Maira, M., McNamara, K., Perera, S.A., Song, Y., et al. (2008). Effective use of PI3K and MEK inhibitors to treat mutant Kras G12D and PIK3CA H1047R murine lung cancers. Nature medicine 14, 1351-1356.

Engelman, J.A., Luo, J., and Cantley, L.C. (2006). The evolution of phosphatidylinositol 3-kinases as regulators of growth and metabolism. Nature reviews. Genetics 7, 606-619.

Falasca, M., Hughes, W.E., Dominguez, V., Sala, G., Fostira, F., Fang, M.Q., Cazzolli, R., Shepherd, P.R., James, D.E., and Maffucci, T. (2007). The role of phosphoinositide 3-kinase C2alpha in insulin signaling. The Journal of biological chemistry 282, 28226- 28236.

Falasca, M., and Maffucci, T. (2012). Regulation and cellular functions of class II phosphoinositide 3-kinases. The Biochemical journal 443, 587-601.

Fedele, C.G., Ooms, L.M., Ho, M., Vieusseux, J., O'Toole, S.A., Millar, E.K., Lopez- Knowles, E., Sriratana, A., Gurung, R., Baglietto, L., et al. (2010). Inositol polyphosphate 4-phosphatase II regulates PI3K/Akt signaling and is lost in human basal-like breast cancers. Proceedings of the National Academy of Sciences of the United States of America 107, 22231-22236.

Folkes, A.J., Ahmadi, K., Alderton, W.K., Alix, S., Baker, S.J., Box, G., Chuckowree, I.S., Clarke, P.A., Depledge, P., Eccles, S.A., et al. (2008). The identification of 2-(1H- indazol-4-yl)-6-(4-methanesulfonyl-piperazin-1-ylmethyl)-4-morpholin-4-yl-t hieno[3,2-

171

d]pyrimidine (GDC-0941) as a potent, selective, orally bioavailable inhibitor of class I PI3 kinase for the treatment of cancer. Journal of medicinal chemistry 51, 5522-5532.

Foukas, L.C., Claret, M., Pearce, W., Okkenhaug, K., Meek, S., Peskett, E., Sancho, S., Smith, A.J., Withers, D.J., and Vanhaesebroeck, B. (2006). Critical role for the p110alpha phosphoinositide-3-OH kinase in growth and metabolic regulation. Nature 441, 366-370.

Franco, I., Gulluni, F., Campa, C.C., Costa, C., Margaria, J.P., Ciraolo, E., Martini, M., Monteyne, D., De Luca, E., Germena, G., et al. (2014). PI3K class II alpha controls spatially restricted endosomal PtdIns3P and Rab11 activation to promote primary cilium function. Developmental cell 28, 647-658.

Fritsch, R., de Krijger, I., Fritsch, K., George, R., Reason, B., Kumar, M.S., Diefenbacher, M., Stamp, G., and Downward, J. (2013). RAS and RHO families of GTPases directly regulate distinct phosphoinositide 3-kinase isoforms. Cell 153, 1050- 1063.

Fruman, D.A., and Cantley, L.C. (2014). Idelalisib--a PI3Kdelta inhibitor for B-cell cancers. The New England journal of medicine 370, 1061-1062.

Fruman, D.A., Mauvais-Jarvis, F., Pollard, D.A., Yballe, C.M., Brazil, D., Bronson, R.T., Kahn, C.R., and Cantley, L.C. (2000). Hypoglycaemia, liver necrosis and perinatal death in mice lacking all isoforms of phosphoinositide 3-kinase p85 alpha. Nature genetics 26, 379-382.

Fruman, D.A., and Rommel, C. (2014). PI3K and cancer: lessons, challenges and opportunities. Nature reviews. Drug discovery 13, 140-156.

Furet, P., Guagnano, V., Fairhurst, R.A., Imbach-Weese, P., Bruce, I., Knapp, M., Fritsch, C., Blasco, F., Blanz, J., Aichholz, R., et al. (2013). Discovery of NVP-BYL719 a potent and selective phosphatidylinositol-3 kinase alpha inhibitor selected for clinical evaluation. Bioorganic & medicinal chemistry letters 23, 3741-3748.

Furman, R.R., Sharman, J.P., Coutre, S.E., Cheson, B.D., Pagel, J.M., Hillmen, P., Barrientos, J.C., Zelenetz, A.D., Kipps, T.J., Flinn, I., et al. (2014). Idelalisib and rituximab in relapsed chronic lymphocytic leukemia. The New England journal of medicine 370, 997-1007.

Gao, J., Aksoy, B.A., Dogrusoz, U., Dresdner, G., Gross, B., Sumer, S.O., Sun, Y., Jacobsen, A., Sinha, R., Larsson, E., et al. (2013). Integrative analysis of complex cancer genomics and clinical profiles using the cBioPortal. Science signaling 6, pl1.

Garrett, J.T., Olivares, M.G., Rinehart, C., Granja-Ingram, N.D., Sanchez, V., Chakrabarty, A., Dave, B., Cook, R.S., Pao, W., McKinely, E., et al. (2011). Transcriptional and posttranslational up-regulation of HER3 (ErbB3) compensates for inhibition of the HER2 tyrosine kinase. Proceedings of the National Academy of Sciences of the United States of America 108, 5021-5026.

Garrett, J.T., Sutton, C.R., Kurupi, R., Bialucha, C.U., Ettenberg, S.A., Collins, S.D., Sheng, Q., Wallweber, J., Defazio-Eli, L., and Arteaga, C.L. (2013). Combination of

172

antibody that inhibits ligand-independent HER3 dimerization and a p110alpha inhibitor potently blocks PI3K signaling and growth of HER2+ breast cancers. Cancer research 73, 6013-6023.

Geering, B., Cutillas, P.R., Nock, G., Gharbi, S.I., and Vanhaesebroeck, B. (2007). Class IA phosphoinositide 3-kinases are obligate p85-p110 heterodimers. Proceedings of the National Academy of Sciences of the United States of America 104, 7809-7814.

Gewinner, C., Wang, Z.C., Richardson, A., Teruya-Feldstein, J., Etemadmoghadam, D., Bowtell, D., Barretina, J., Lin, W.M., Rameh, L., Salmena, L., et al. (2009). Evidence that inositol polyphosphate 4-phosphatase type II is a tumor suppressor that inhibits PI3K signaling. Cancer cell 16, 115-125.

Golub, T.R., Barker, G.F., Lovett, M., and Gilliland, D.G. (1994). Fusion of PDGF receptor beta to a novel ets-like gene, tel, in chronic myelomonocytic leukemia with t(5;12) chromosomal translocation. Cell 77, 307-316.

Gonzalez-Billalabeitia, E., Seitzer, N., Song, S.J., Song, M.S., Patnaik, A., Liu, X.S., Epping, M.T., Papa, A., Hobbs, R.M., Chen, M., et al. (2014). Vulnerabilities of PTEN- p53-deficient prostate cancers to compound PARP/PI3K inhibition. Cancer discovery.

Gopal, A.K., Kahl, B.S., de Vos, S., Wagner-Johnston, N.D., Schuster, S.J., Jurczak, W.J., Flinn, I.W., Flowers, C.R., Martin, P., Viardot, A., et al. (2014). PI3Kdelta inhibition by idelalisib in patients with relapsed indolent lymphoma. The New England journal of medicine 370, 1008-1018.

Graupera, M., Guillermet-Guibert, J., Foukas, L.C., Phng, L.K., Cain, R.J., Salpekar, A., Pearce, W., Meek, S., Millan, J., Cutillas, P.R., et al. (2008). Angiogenesis selectively requires the p110alpha isoform of PI3K to control endothelial cell migration. Nature 453, 662-666.

Gritsman, K., Yuzugullu, H., Von, T., Yan, H., Clayton, L., Fritsch, C., Maira, S.M., Hollingworth, G., Choi, C., Khandan, T., et al. (2014). Hematopoiesis and RAS-driven myeloid leukemia differentially require PI3K isoform p110alpha. The Journal of clinical investigation 124, 1794-1809.

Gruber Filbin, M., Dabral, S.K., Pazyra-Murphy, M.F., Ramkissoon, S., Kung, A.L., Pak, E., Chung, J., Theisen, M.A., Sun, Y., Franchetti, Y., et al. (2013). Coordinate activation of Shh and PI3K signaling in PTEN-deficient glioblastoma: new therapeutic opportunities. Nature medicine 19, 1518-1523.

Guillermet-Guibert, J., Bjorklof, K., Salpekar, A., Gonella, C., Ramadani, F., Bilancio, A., Meek, S., Smith, A.J., Okkenhaug, K., and Vanhaesebroeck, B. (2008). The p110beta isoform of phosphoinositide 3-kinase signals downstream of G protein-coupled receptors and is functionally redundant with p110gamma. Proceedings of the National Academy of Sciences of the United States of America 105, 8292-8297.

Gupta, S., Ramjaun, A.R., Haiko, P., Wang, Y., Warne, P.H., Nicke, B., Nye, E., Stamp, G., Alitalo, K., and Downward, J. (2007). Binding of ras to phosphoinositide 3-kinase p110alpha is required for ras-driven tumorigenesis in mice. Cell 129, 957-968.

173

Guy, P.M., Platko, J.V., Cantley, L.C., Cerione, R.A., and Carraway, K.L., 3rd (1994). Insect cell-expressed p180erbB3 possesses an impaired tyrosine kinase activity. Proceedings of the National Academy of Sciences of the United States of America 91, 8132-8136.

Gymnopoulos, M., Elsliger, M.A., and Vogt, P.K. (2007). Rare cancer-specific mutations in PIK3CA show gain of function. Proceedings of the National Academy of Sciences of the United States of America 104, 5569-5574.

Hanahan, D., and Weinberg, R.A. (2011). Hallmarks of cancer: the next generation. Cell 144, 646-674.

Hao, Y., Wang, C., Cao, B., Hirsch, B.M., Song, J., Markowitz, S.D., Ewing, R.M., Sedwick, D., Liu, L., Zheng, W., et al. (2013). Gain of interaction with IRS1 by p110alpha-helical domain mutants is crucial for their oncogenic functions. Cancer cell 23, 583-593.

Harada, K., Truong, A.B., Cai, T., and Khavari, P.A. (2005). The class II phosphoinositide 3-kinase C2beta is not essential for epidermal differentiation. Molecular and cellular biology 25, 11122-11130.

Harris, D.P., Vogel, P., Wims, M., Moberg, K., Humphries, J., Jhaver, K.G., DaCosta, C.M., Shadoan, M.K., Xu, N., Hansen, G.M., et al. (2011). Requirement for class II phosphoinositide 3-kinase C2alpha in maintenance of glomerular structure and function. Molecular and cellular biology 31, 63-80.

Hellyer, N.J., Cheng, K., and Koland, J.G. (1998). ErbB3 (HER3) interaction with the p85 regulatory subunit of phosphoinositide 3-kinase. The Biochemical journal 333 ( Pt 3), 757-763.

Hennighausen, L., and Robinson, G.W. (1998). Think globally, act locally: the making of a mouse mammary gland. Genes & development 12, 449-455.

Hennighausen, L., and Robinson, G.W. (2001). Signaling pathways in mammary gland development. Developmental cell 1, 467-475.

Hirai, H., Sootome, H., Nakatsuru, Y., Miyama, K., Taguchi, S., Tsujioka, K., Ueno, Y., Hatch, H., Majumder, P.K., Pan, B.S., et al. (2010). MK-2206, an allosteric Akt inhibitor, enhances antitumor efficacy by standard chemotherapeutic agents or molecular targeted drugs in vitro and in vivo. Molecular cancer therapeutics 9, 1956-1967.

Hirsch, D.S., Shen, Y., Dokmanovic, M., and Wu, W.J. (2010). pp60c-Src phosphorylates and activates vacuolar protein sorting 34 to mediate cellular transformation. Cancer research 70, 5974-5983.

Hodgson, M.C., Shao, L.J., Frolov, A., Li, R., Peterson, L.E., Ayala, G., Ittmann, M.M., Weigel, N.L., and Agoulnik, I.U. (2011). Decreased expression and androgen regulation of the tumor suppressor gene INPP4B in prostate cancer. Cancer research 71, 572-582.

Hon, W.C., Berndt, A., and Williams, R.L. (2012). Regulation of lipid binding underlies the activation mechanism of class IA PI3-kinases. Oncogene 31, 3655-3666.

174

Huang, C.H., Mandelker, D., Schmidt-Kittler, O., Samuels, Y., Velculescu, V.E., Kinzler, K.W., Vogelstein, B., Gabelli, S.B., and Amzel, L.M. (2007). The structure of a human p110alpha/p85alpha complex elucidates the effects of oncogenic PI3Kalpha mutations. Science 318, 1744-1748.

Huang, S.M., Mishina, Y.M., Liu, S., Cheung, A., Stegmeier, F., Michaud, G.A., Charlat, O., Wiellette, E., Zhang, Y., Wiessner, S., et al. (2009). Tankyrase inhibition stabilizes axin and antagonizes Wnt signalling. Nature 461, 614-620.

Huang, X., Hou, J., Shen, X., Huang, C., Zhang, X., Xie, Y., and Luo, X. (2014). MicroRNA-486-5p, which is downregulated in hepatocellular carcinoma, suppresses tumor growth by targeting PIK3R1. The FEBS journal.

Huang, X., Shen, Y., Liu, M., Bi, C., Jiang, C., Iqbal, J., McKeithan, T.W., Chan, W.C., Ding, S.J., and Fu, K. (2012). Quantitative proteomics reveals that miR-155 regulates the PI3K-AKT pathway in diffuse large B-cell lymphoma. The American journal of pathology 181, 26-33.

Hynes, N.E., and Watson, C.J. (2010). Mammary gland growth factors: roles in normal development and in cancer. Cold Spring Harbor perspectives in biology 2, a003186.

Ibrahim, Y.H., Garcia-Garcia, C., Serra, V., He, L., Torres-Lockhart, K., Prat, A., Anton, P., Cozar, P., Guzman, M., Grueso, J., et al. (2012). PI3K inhibition impairs BRCA1/2 expression and sensitizes BRCA-proficient triple-negative breast cancer to PARP inhibition. Cancer discovery 2, 1036-1047.

Ilic, N., Utermark, T., Widlund, H.R., and Roberts, T.M. (2011). PI3K-targeted therapy can be evaded by gene amplification along the MYC-eukaryotic translation initiation factor 4E (eIF4E) axis. Proceedings of the National Academy of Sciences of the United States of America 108, E699-708.

Isakoff, S.J., Engelman, J.A., Irie, H.Y., Luo, J., Brachmann, S.M., Pearline, R.V., Cantley, L.C., and Brugge, J.S. (2005). Breast cancer-associated PIK3CA mutations are oncogenic in mammary epithelial cells. Cancer research 65, 10992-11000.

Ito, Y., Hart, J.R., Ueno, L., and Vogt, P.K. (2014). Oncogenic activity of the regulatory subunit p85beta of phosphatidylinositol 3-kinase (PI3K). Proceedings of the National Academy of Sciences of the United States of America.

Jackson, S.P., Schoenwaelder, S.M., Goncalves, I., Nesbitt, W.S., Yap, C.L., Wright, C.E., Kenche, V., Anderson, K.E., Dopheide, S.M., Yuan, Y., et al. (2005). PI 3-kinase p110beta: a new target for antithrombotic therapy. Nature medicine 11, 507-514.

Jaiswal, B.S., Janakiraman, V., Kljavin, N.M., Chaudhuri, S., Stern, H.M., Wang, W., Kan, Z., Dbouk, H.A., Peters, B.A., Waring, P., et al. (2009). Somatic mutations in p85alpha promote tumorigenesis through class IA PI3K activation. Cancer cell 16, 463- 474.

Jean, S., and Kiger, A.A. (2012). Coordination between RAB GTPase and phosphoinositide regulation and functions. Nature reviews. Molecular cell biology 13, 463-470.

175

Jia, S., Gao, X., Lee, S.H., Maira, S.M., Wu, X., Stack, E.C., Signoretti, S., Loda, M., Zhao, J.J., and Roberts, T.M. (2013). Opposing effects of androgen deprivation and targeted therapy on prostate cancer prevention. Cancer discovery 3, 44-51.

Jia, S., Liu, Z., Zhang, S., Liu, P., Zhang, L., Lee, S.H., Zhang, J., Signoretti, S., Loda, M., Roberts, T.M., et al. (2008). Essential roles of PI(3)K-p110beta in cell growth, metabolism and tumorigenesis. Nature 454, 776-779.

Jou, S.T., Carpino, N., Takahashi, Y., Piekorz, R., Chao, J.R., Carpino, N., Wang, D., and Ihle, J.N. (2002). Essential, nonredundant role for the phosphoinositide 3-kinase p110delta in signaling by the B-cell receptor complex. Molecular and cellular biology 22, 8580-8591.

Jousset, C., Carron, C., Boureux, A., Quang, C.T., Oury, C., Dusanter-Fourt, I., Charon, M., Levin, J., Bernard, O., and Ghysdael, J. (1997). A domain of TEL conserved in a subset of ETS proteins defines a specific oligomerization interface essential to the mitogenic properties of the TEL-PDGFR beta oncoprotein. The EMBO journal 16, 69-82.

Jura, N., Shan, Y., Cao, X., Shaw, D.E., and Kuriyan, J. (2009). Structural analysis of the catalytically inactive kinase domain of the human EGF receptor 3. Proceedings of the National Academy of Sciences of the United States of America 106, 21608-21613.

Juric, D., Argiles, G., Burris, H., Gonzalez-Angulo, A., Saura, C., Quadt, C., Douglas, M., Demanse, D., De Buck, S., and Baselga, J. (2012). Phase I study of BYL719, an alpha- specific PI3K inhibitor, in patients with PIK3CA mutant advanced solid tumors: preliminary efficacy and safety in patients with PIK3CA mutant ER-positive (ER+) metastatic breast cancer (MBC). Cancer research 72, P6-10-07.

Juric, D., Gonzalez-Angulo, A., Burris, H., Schuler, M., Schellens, J., Berlin, J., Gupta, A., Seggewiss-Bernhardt, R., Adamo, B., Gil-Martin, M., et al. (2013a). Abstract P2-16- 14: Preliminary safety, pharmacokinetics and anti-tumor activity of BYL719, an alpha- specific PI3K inhibitor in combination with fulvestrant: Results from a phase I study. Cancer research 73, P2-16-14.

Juric, D., Saura, C., Cervantes, A., Kurkjian, C., Patel, M., Sachdev, J., Mayer, I., Krop, I., Oliveira, M., Sanabria, S., et al. (2013b). Abstract PD1-3: Ph1b study of the PI3K inhibitor GDC-0032 in combination with fulvestrant in patients with hormone receptor- positive advanced breast cancer. Cancer research 73, PD1-3.

Juvekar, A., Burga, L.N., Hu, H., Lunsford, E.P., Ibrahim, Y.H., Balmana, J., Rajendran, A., Papa, A., Spencer, K., Lyssiotis, C.A., et al. (2012). Combining a PI3K inhibitor with a PARP inhibitor provides an effective therapy for BRCA1-related breast cancer. Cancer discovery 2, 1048-1063.

Kan, Z., Jaiswal, B.S., Stinson, J., Janakiraman, V., Bhatt, D., Stern, H.M., Yue, P., Haverty, P.M., Bourgon, R., Zheng, J., et al. (2010). Diverse somatic mutation patterns and pathway alterations in human cancers. Nature 466, 869-873.

Kang, S., Bader, A.G., and Vogt, P.K. (2005). Phosphatidylinositol 3-kinase mutations identified in human cancer are oncogenic. Proceedings of the National Academy of Sciences of the United States of America 102, 802-807.

176

Kang, S., Denley, A., Vanhaesebroeck, B., and Vogt, P.K. (2006). Oncogenic transformation induced by the p110beta, -gamma, and -delta isoforms of class I phosphoinositide 3-kinase. Proceedings of the National Academy of Sciences of the United States of America 103, 1289-1294.

Katso, R.M., Pardo, O.E., Palamidessi, A., Franz, C.M., Marinov, M., De Laurentiis, A., Downward, J., Scita, G., Ridley, A.J., Waterfield, M.D., et al. (2006). Phosphoinositide 3- Kinase C2beta regulates cytoskeletal organization and cell migration via Rac-dependent mechanisms. Molecular biology of the cell 17, 3729-3744.

Kim, Y.J., Kim, H.J., Chung, K.Y., Choi, I., and Kim, S.H. (2014). Transcriptional activation of PIK3R1 by PPARgamma in adipocytes. Molecular biology reports 41, 5267- 5272.

Kinross, K.M., Montgomery, K.G., Kleinschmidt, M., Waring, P., Ivetac, I., Tikoo, A., Saad, M., Hare, L., Roh, V., Mantamadiotis, T., et al. (2012). An activating Pik3ca mutation coupled with Pten loss is sufficient to initiate ovarian tumorigenesis in mice. The Journal of clinical investigation 122, 553-557.

Klarlund, J.K., Guilherme, A., Holik, J.J., Virbasius, J.V., Chawla, A., and Czech, M.P. (1997). Signaling by phosphoinositide-3,4,5-trisphosphate through proteins containing pleckstrin and Sec7 homology domains. Science 275, 1927-1930.

Klippel, A., Reinhard, C., Kavanaugh, W.M., Apell, G., Escobedo, M.A., and Williams, L.T. (1996). Membrane localization of phosphatidylinositol 3-kinase is sufficient to activate multiple signal-transducing kinase pathways. Molecular and cellular biology 16, 4117-4127.

Knight, Z.A., Gonzalez, B., Feldman, M.E., Zunder, E.R., Goldenberg, D.D., Williams, O., Loewith, R., Stokoe, D., Balla, A., Toth, B., et al. (2006). A pharmacological map of the PI3-K family defines a role for p110alpha in insulin signaling. Cell 125, 733-747.

Knight, Z.A., and Shokat, K.M. (2005). Features of selective kinase inhibitors. Chemistry & biology 12, 621-637.

Knobbe, C.B., and Reifenberger, G. (2003). Genetic alterations and aberrant expression of genes related to the phosphatidyl-inositol-3'-kinase/protein kinase B (Akt) signal transduction pathway in glioblastomas. Brain pathology 13, 507-518.

Krugmann, S., Anderson, K.E., Ridley, S.H., Risso, N., McGregor, A., Coadwell, J., Davidson, K., Eguinoa, A., Ellson, C.D., Lipp, P., et al. (2002). Identification of ARAP3, a novel PI3K effector regulating both Arf and Rho GTPases, by selective capture on phosphoinositide affinity matrices. Molecular cell 9, 95-108.

Kubo, H., Hazeki, K., Takasuga, S., and Hazeki, O. (2005). Specific role for p85/p110beta in GTP-binding-protein-mediated activation of Akt. The Biochemical journal 392, 607-614.

Kurek, K.C., Luks, V.L., Ayturk, U.M., Alomari, A.I., Fishman, S.J., Spencer, S.A., Mulliken, J.B., Bowen, M.E., Yamamoto, G.L., Kozakewich, H.P., et al. (2012). Somatic

177

mosaic activating mutations in PIK3CA cause CLOVES syndrome. American journal of human genetics 90, 1108-1115.

Kurig, B., Shymanets, A., Bohnacker, T., Prajwal, Brock, C., Ahmadian, M.R., Schaefer, M., Gohla, A., Harteneck, C., Wymann, M.P., et al. (2009). Ras is an indispensable coregulator of the class IB phosphoinositide 3-kinase p87/p110gamma. Proceedings of the National Academy of Sciences of the United States of America 106, 20312-20317.

Kurosu, H., Maehama, T., Okada, T., Yamamoto, T., Hoshino, S., Fukui, Y., Ui, M., Hazeki, O., and Katada, T. (1997). Heterodimeric phosphoinositide 3-kinase consisting of p85 and p110beta is synergistically activated by the betagamma subunits of G proteins and phosphotyrosyl peptide. The Journal of biological chemistry 272, 24252- 24256.

Kwabi-Addo, B., Giri, D., Schmidt, K., Podsypanina, K., Parsons, R., Greenberg, N., and Ittmann, M. (2001). Haploinsufficiency of the Pten tumor suppressor gene promotes prostate cancer progression. Proceedings of the National Academy of Sciences of the United States of America 98, 11563-11568.

Li, G., Robinson, G.W., Lesche, R., Martinez-Diaz, H., Jiang, Z., Rozengurt, N., Wagner, K.U., Wu, D.C., Lane, T.F., Liu, X., et al. (2002). Conditional loss of PTEN leads to precocious development and neoplasia in the mammary gland. Development 129, 4159- 4170.

Liaw, D., Marsh, D.J., Li, J., Dahia, P.L., Wang, S.I., Zheng, Z., Bose, S., Call, K.M., Tsou, H.C., Peacocke, M., et al. (1997). Germline mutations of the PTEN gene in Cowden disease, an inherited breast and thyroid cancer syndrome. Nature genetics 16, 64-67.

Liu, P., Cheng, H., Roberts, T.M., and Zhao, J.J. (2009). Targeting the phosphoinositide 3-kinase pathway in cancer. Nature reviews. Drug discovery 8, 627-644.

Liu, P., Cheng, H., Santiago, S., Raeder, M., Zhang, F., Isabella, A., Yang, J., Semaan, D.J., Chen, C., Fox, E.A., et al. (2011). Oncogenic PIK3CA-driven mammary tumors frequently recur via PI3K pathway-dependent and PI3K pathway-independent mechanisms. Nature medicine 17, 1116-1120.

Liu, P., Morrison, C., Wang, L., Xiong, D., Vedell, P., Cui, P., Hua, X., Ding, F., Lu, Y., James, M., et al. (2012). Identification of somatic mutations in non-small cell lung carcinomas using whole-exome sequencing. Carcinogenesis 33, 1270-1276.

Liu, X., Holstege, H., van der Gulden, H., Treur-Mulder, M., Zevenhoven, J., Velds, A., Kerkhoven, R.M., van Vliet, M.H., Wessels, L.F., Peterse, J.L., et al. (2007). Somatic loss of BRCA1 and p53 in mice induces mammary tumors with features of human BRCA1-mutated basal-like breast cancer. Proceedings of the National Academy of Sciences of the United States of America 104, 12111-12116.

Lu, N., Shen, Q., Mahoney, T.R., Neukomm, L.J., Wang, Y., and Zhou, Z. (2012). Two PI 3-kinases and one PI 3-phosphatase together establish the cyclic waves of phagosomal PtdIns(3)P critical for the degradation of apoptotic cells. PLoS biology 10, e1001245.

178

Lucas, C.L., Kuehn, H.S., Zhao, F., Niemela, J.E., Deenick, E.K., Palendira, U., Avery, D.T., Moens, L., Cannons, J.L., Biancalana, M., et al. (2014). Dominant-activating germline mutations in the gene encoding the PI(3)K catalytic subunit p110delta result in T cell senescence and human immunodeficiency. Nature immunology 15, 88-97.

Luo, J., and Cantley, L.C. (2005). The negative regulation of phosphoinositide 3-kinase signaling by p85 and it's implication in cancer. Cell cycle 4, 1309-1312.

Luo, J., Field, S.J., Lee, J.Y., Engelman, J.A., and Cantley, L.C. (2005a). The p85 regulatory subunit of phosphoinositide 3-kinase down-regulates IRS-1 signaling via the formation of a sequestration complex. The Journal of cell biology 170, 455-464.

Luo, J., McMullen, J.R., Sobkiw, C.L., Zhang, L., Dorfman, A.L., Sherwood, M.C., Logsdon, M.N., Horner, J.W., DePinho, R.A., Izumo, S., et al. (2005b). Class IA phosphoinositide 3-kinase regulates heart size and physiological cardiac hypertrophy. Molecular and cellular biology 25, 9491-9502.

Luo, J., Sobkiw, C.L., Logsdon, N.M., Watt, J.M., Signoretti, S., O'Connell, F., Shin, E., Shim, Y., Pao, L., Neel, B.G., et al. (2005c). Modulation of epithelial neoplasia and lymphoid hyperplasia in PTEN+/- mice by the p85 regulatory subunits of phosphoinositide 3-kinase. Proceedings of the National Academy of Sciences of the United States of America 102, 10238-10243.

Ma, X.J., Dahiya, S., Richardson, E., Erlander, M., and Sgroi, D.C. (2009). Gene expression profiling of the tumor microenvironment during breast cancer progression. Breast cancer research : BCR 11, R7.

MacLennan, M.B., Anderson, B.M., and Ma, D.W. (2011). Differential mammary gland development in FVB and C57Bl/6 mice: implications for breast cancer research. Nutrients 3, 929-936.

Maffucci, T., Cooke, F.T., Foster, F.M., Traer, C.J., Fry, M.J., and Falasca, M. (2005). Class II phosphoinositide 3-kinase defines a novel signaling pathway in cell migration. The Journal of cell biology 169, 789-799.

Maier, U., Babich, A., and Nurnberg, B. (1999). Roles of non-catalytic subunits in gbetagamma-induced activation of class I phosphoinositide 3-kinase isoforms beta and gamma. The Journal of biological chemistry 274, 29311-29317.

Maira, S.M., Stauffer, F., Brueggen, J., Furet, P., Schnell, C., Fritsch, C., Brachmann, S., Chene, P., De Pover, A., Schoemaker, K., et al. (2008). Identification and characterization of NVP-BEZ235, a new orally available dual phosphatidylinositol 3- kinase/mammalian target of rapamycin inhibitor with potent in vivo antitumor activity. Molecular cancer therapeutics 7, 1851-1863.

Mandelker, D., Gabelli, S.B., Schmidt-Kittler, O., Zhu, J., Cheong, I., Huang, C.H., Kinzler, K.W., Vogelstein, B., and Amzel, L.M. (2009). A frequent kinase domain mutation that changes the interaction between PI3Kalpha and the membrane. Proceedings of the National Academy of Sciences of the United States of America 106, 16996-17001.

179

Martin, A.L., Schwartz, M.D., Jameson, S.C., and Shimizu, Y. (2008). Selective regulation of CD8 effector T cell migration by the p110 gamma isoform of phosphatidylinositol 3-kinase. Journal of immunology 180, 2081-2088.

Mauvais-Jarvis, F., Ueki, K., Fruman, D.A., Hirshman, M.F., Sakamoto, K., Goodyear, L.J., Iannacone, M., Accili, D., Cantley, L.C., and Kahn, C.R. (2002). Reduced expression of the murine p85alpha subunit of phosphoinositide 3-kinase improves insulin signaling and ameliorates diabetes. The Journal of clinical investigation 109, 141-149.

McCurdy, C.E., Schenk, S., Holliday, M.J., Philp, A., Houck, J.A., Patsouris, D., MacLean, P.S., Majka, S.M., Klemm, D.J., and Friedman, J.E. (2012). Attenuated Pik3r1 expression prevents insulin resistance and adipose tissue macrophage accumulation in diet-induced obese mice. Diabetes 61, 2495-2505.

Meier, T.I., Cook, J.A., Thomas, J.E., Radding, J.A., Horn, C., Lingaraj, T., and Smith, M.C. (2004). Cloning, expression, purification, and characterization of the human Class Ia phosphoinositide 3-kinase isoforms. Protein expression and purification 35, 218-224.

Mellor, P., Furber, L.A., Nyarko, J.N., and Anderson, D.H. (2012). Multiple roles for the p85alpha isoform in the regulation and function of PI3K signalling and receptor trafficking. The Biochemical journal 441, 23-37.

Miaczynska, M. (2013). Effects of membrane trafficking on signaling by receptor tyrosine kinases. Cold Spring Harbor perspectives in biology 5, a009035.

Miled, N., Yan, Y., Hon, W.C., Perisic, O., Zvelebil, M., Inbar, Y., Schneidman-Duhovny, D., Wolfson, H.J., Backer, J.M., and Williams, R.L. (2007). Mechanism of two classes of cancer mutations in the phosphoinositide 3-kinase catalytic subunit. Science 317, 239- 242.

Muellner, M.K., Uras, I.Z., Gapp, B.V., Kerzendorfer, C., Smida, M., Lechtermann, H., Craig-Mueller, N., Colinge, J., Duernberger, G., and Nijman, S.M. (2011). A chemical- genetic screen reveals a mechanism of resistance to PI3K inhibitors in cancer. Nature chemical biology 7, 787-793.

Muranen, T., Selfors, L.M., Worster, D.T., Iwanicki, M.P., Song, L., Morales, F.C., Gao, S., Mills, G.B., and Brugge, J.S. (2012). Inhibition of PI3K/mTOR leads to adaptive resistance in matrix-attached cancer cells. Cancer cell 21, 227-239.

Murga, C., Fukuhara, S., and Gutkind, J.S. (2000). A novel role for phosphatidylinositol 3-kinase beta in signaling from G protein-coupled receptors to Akt. The Journal of biological chemistry 275, 12069-12073.

Murphy, G.A., Graham, S.M., Morita, S., Reks, S.E., Rogers-Graham, K., Vojtek, A., Kelley, G.G., and Der, C.J. (2002). Involvement of phosphatidylinositol 3-kinase, but not RalGDS, in TC21/R-Ras2-mediated transformation. The Journal of biological chemistry 277, 9966-9975.

Ni, J., Liu, Q., Xie, S., Carlson, C., Von, T., Vogel, K., Riddle, S., Benes, C., Eck, M., Roberts, T., et al. (2012). Functional characterization of an isoform-selective inhibitor of PI3K-p110beta as a potential anticancer agent. Cancer discovery 2, 425-433.

180

Nicoli, S., Knyphausen, C.P., Zhu, L.J., Lakshmanan, A., and Lawson, N.D. (2012). miR- 221 is required for endothelial tip cell behaviors during vascular development. Developmental cell 22, 418-429.

Nobusawa, S., Lachuer, J., Wierinckx, A., Kim, Y.H., Huang, J., Legras, C., Kleihues, P., and Ohgaki, H. (2010). Intratumoral patterns of genomic imbalance in glioblastomas. Brain pathology 20, 936-944.

Norris, F.A., Atkins, R.C., and Majerus, P.W. (1997). The cDNA cloning and characterization of inositol polyphosphate 4-phosphatase type II. Evidence for conserved alternative splicing in the 4-phosphatase family. The Journal of biological chemistry 272, 23859-23864.

O'Reilly, K.E., Rojo, F., She, Q.B., Solit, D., Mills, G.B., Smith, D., Lane, H., Hofmann, F., Hicklin, D.J., Ludwig, D.L., et al. (2006). mTOR inhibition induces upstream receptor tyrosine kinase signaling and activates Akt. Cancer research 66, 1500-1508.

Okkenhaug, K., Bilancio, A., Farjot, G., Priddle, H., Sancho, S., Peskett, E., Pearce, W., Meek, S.E., Salpekar, A., Waterfield, M.D., et al. (2002). Impaired B and T cell antigen receptor signaling in p110delta PI 3-kinase mutant mice. Science 297, 1031-1034.

Okkenhaug, K., and Vanhaesebroeck, B. (2003). PI3K in lymphocyte development, differentiation and activation. Nature reviews. Immunology 3, 317-330.

Orloff, M.S., He, X., Peterson, C., Chen, F., Chen, J.L., Mester, J.L., and Eng, C. (2013). Germline PIK3CA and AKT1 mutations in Cowden and Cowden-like syndromes. American journal of human genetics 92, 76-80.

Pacold, M.E., Suire, S., Perisic, O., Lara-Gonzalez, S., Davis, C.T., Walker, E.H., Hawkins, P.T., Stephens, L., Eccleston, J.F., and Williams, R.L. (2000). Crystal structure and functional analysis of Ras binding to its effector phosphoinositide 3-kinase gamma. Cell 103, 931-943.

Papakonstanti, E.A., Zwaenepoel, O., Bilancio, A., Burns, E., Nock, G.E., Houseman, B., Shokat, K., Ridley, A.J., and Vanhaesebroeck, B. (2008). Distinct roles of class IA PI3K isoforms in primary and immortalised macrophages. Journal of cell science 121, 4124- 4133.

Park, S.W., Zhou, Y., Lee, J., Lu, A., Sun, C., Chung, J., Ueki, K., and Ozcan, U. (2010). The regulatory subunits of PI3K, p85alpha and p85beta, interact with XBP-1 and increase its nuclear translocation. Nature medicine 16, 429-437.

Parsons, D.W., Jones, S., Zhang, X., Lin, J.C., Leary, R.J., Angenendt, P., Mankoo, P., Carter, H., Siu, I.M., Gallia, G.L., et al. (2008). An integrated genomic analysis of human glioblastoma multiforme. Science 321, 1807-1812.

Parsons, R. (2004). Human cancer, PTEN and the PI-3 kinase pathway. Seminars in cell & developmental biology 15, 171-176.

Peng, Y., Dai, Y., Hitchcock, C., Yang, X., Kassis, E.S., Liu, L., Luo, Z., Sun, H.L., Cui, R., Wei, H., et al. (2013). Insulin growth factor signaling is regulated by microRNA-486,

181

an underexpressed microRNA in lung cancer. Proceedings of the National Academy of Sciences of the United States of America 110, 15043-15048.

Philp, A.J., Campbell, I.G., Leet, C., Vincan, E., Rockman, S.P., Whitehead, R.H., Thomas, R.J., and Phillips, W.A. (2001). The phosphatidylinositol 3'-kinase p85alpha gene is an oncogene in human ovarian and colon tumors. Cancer research 61, 7426- 7429.

Pinkas-Kramarski, R., Soussan, L., Waterman, H., Levkowitz, G., Alroy, I., Klapper, L., Lavi, S., Seger, R., Ratzkin, B.J., Sela, M., et al. (1996). Diversification of Neu differentiation factor and epidermal growth factor signaling by combinatorial receptor interactions. The EMBO journal 15, 2452-2467.

Planchon, S.M., Waite, K.A., and Eng, C. (2008). The nuclear affairs of PTEN. Journal of cell science 121, 249-253.

Prigent, S.A., and Gullick, W.J. (1994). Identification of c-erbB-3 binding sites for phosphatidylinositol 3'-kinase and SHC using an EGF receptor/c-erbB-3 chimera. The EMBO journal 13, 2831-2841.

Rabinovsky, R., Pochanard, P., McNear, C., Brachmann, S.M., Duke-Cohan, J.S., Garraway, L.A., and Sellers, W.R. (2009). p85 Associates with unphosphorylated PTEN and the PTEN-associated complex. Molecular and cellular biology 29, 5377-5388.

Rahmani, M., Aust, M.M., Attkisson, E., Williams, D.C., Jr., Ferreira-Gonzalez, A., and Grant, S. (2013). Dual inhibition of Bcl-2 and Bcl-xL strikingly enhances PI3K inhibition- induced apoptosis in human myeloid leukemia cells through a GSK3- and Bim- dependent mechanism. Cancer research 73, 1340-1351.

Rameh, L.E., Chen, C.S., and Cantley, L.C. (1995). Phosphatidylinositol (3,4,5)P3 interacts with SH2 domains and modulates PI 3-kinase association with tyrosine- phosphorylated proteins. Cell 83, 821-830.

Rao, S.K., Edwards, J., Joshi, A.D., Siu, I.M., and Riggins, G.J. (2010). A survey of glioblastoma genomic amplifications and deletions. Journal of neuro-oncology 96, 169- 179.

Raynaud, F.I., Eccles, S.A., Patel, S., Alix, S., Box, G., Chuckowree, I., Folkes, A., Gowan, S., De Haven Brandon, A., Di Stefano, F., et al. (2009). Biological properties of potent inhibitors of class I phosphatidylinositide 3-kinases: from PI-103 through PI-540, PI-620 to the oral agent GDC-0941. Molecular cancer therapeutics 8, 1725-1738.

Reif, K., Okkenhaug, K., Sasaki, T., Penninger, J.M., Vanhaesebroeck, B., and Cyster, J.G. (2004). Cutting edge: differential roles for phosphoinositide 3-kinases, p110gamma and p110delta, in lymphocyte chemotaxis and homing. Journal of immunology 173, 2236-2240.

Rhodes, D.R., Kalyana-Sundaram, S., Mahavisno, V., Varambally, R., Yu, J., Briggs, B.B., Barrette, T.R., Anstet, M.J., Kincead-Beal, C., Kulkarni, P., et al. (2007). Oncomine 3.0: genes, pathways, and networks in a collection of 18,000 cancer gene expression profiles. Neoplasia 9, 166-180.

182

Rhodes, D.R., Yu, J., Shanker, K., Deshpande, N., Varambally, R., Ghosh, D., Barrette, T., Pandey, A., and Chinnaiyan, A.M. (2004). ONCOMINE: a cancer microarray database and integrated data-mining platform. Neoplasia 6, 1-6.

Rhodes, N., Heerding, D.A., Duckett, D.R., Eberwein, D.J., Knick, V.B., Lansing, T.J., McConnell, R.T., Gilmer, T.M., Zhang, S.Y., Robell, K., et al. (2008). Characterization of an Akt kinase inhibitor with potent pharmacodynamic and antitumor activity. Cancer research 68, 2366-2374.

Richardson, A.L., Wang, Z.C., De Nicolo, A., Lu, X., Brown, M., Miron, A., Liao, X., Iglehart, J.D., Livingston, D.M., and Ganesan, S. (2006). X chromosomal abnormalities in basal-like human breast cancer. Cancer cell 9, 121-132.

Rios, J.J., Paria, N., Burns, D.K., Israel, B.A., Cornelia, R., Wise, C.A., and Ezaki, M. (2013). Somatic gain-of-function mutations in PIK3CA in patients with macrodactyly. Human molecular genetics 22, 444-451.

Riviere, J.B., Mirzaa, G.M., O'Roak, B.J., Beddaoui, M., Alcantara, D., Conway, R.L., St- Onge, J., Schwartzentruber, J.A., Gripp, K.W., Nikkel, S.M., et al. (2012). De novo germline and postzygotic mutations in AKT3, PIK3R2 and PIK3CA cause a spectrum of related megalencephaly syndromes. Nature genetics 44, 934-940.

Rodon, J., Dienstmann, R., Serra, V., and Tabernero, J. (2013). Development of PI3K inhibitors: lessons learned from early clinical trials. Nature reviews. Clinical oncology 10, 143-153.

Rodriguez-Viciana, P., Sabatier, C., and McCormick, F. (2004). Signaling specificity by Ras family GTPases is determined by the full spectrum of effectors they regulate. Molecular and cellular biology 24, 4943-4954.

Rodriguez-Viciana, P., Warne, P.H., Dhand, R., Vanhaesebroeck, B., Gout, I., Fry, M.J., Waterfield, M.D., and Downward, J. (1994). Phosphatidylinositol-3-OH kinase as a direct target of Ras. Nature 370, 527-532.

Rodriguez-Viciana, P., Warne, P.H., Vanhaesebroeck, B., Waterfield, M.D., and Downward, J. (1996). Activation of phosphoinositide 3-kinase by interaction with Ras and by point mutation. The EMBO journal 15, 2442-2451.

Romanov, S.R., Kozakiewicz, B.K., Holst, C.R., Stampfer, M.R., Haupt, L.M., and Tlsty, T.D. (2001). Normal human mammary epithelial cells spontaneously escape senescence and acquire genomic changes. Nature 409, 633-637.

Rubio, I., Rodriguez-Viciana, P., Downward, J., and Wetzker, R. (1997). Interaction of Ras with phosphoinositide 3-kinase gamma. The Biochemical journal 326 ( Pt 3), 891- 895.

Sampath, D., Nannini, M., Jane, G., Hong, R., Parsons, K., Belvin, M., Friedman, L., and Wallin, J. (2013). Abstract P4-15-02: The PI3K inhibitor GDC-0032 enhances the efficacy of standard of care therapeutics in PI3K alpha mutant breast cancer models. Cancer research 73, P4-15-02.

183

Samuels, Y., Diaz, L.A., Jr., Schmidt-Kittler, O., Cummins, J.M., Delong, L., Cheong, I., Rago, C., Huso, D.L., Lengauer, C., Kinzler, K.W., et al. (2005). Mutant PIK3CA promotes cell growth and invasion of human cancer cells. Cancer cell 7, 561-573.

Samuels, Y., Wang, Z., Bardelli, A., Silliman, N., Ptak, J., Szabo, S., Yan, H., Gazdar, A., Powell, S.M., Riggins, G.J., et al. (2004). High frequency of mutations of the PIK3CA gene in human cancers. Science 304, 554.

Sasaki, T., Irie-Sasaki, J., Jones, R.G., Oliveira-dos-Santos, A.J., Stanford, W.L., Bolon, B., Wakeham, A., Itie, A., Bouchard, D., Kozieradzki, I., et al. (2000). Function of PI3Kgamma in thymocyte development, T cell activation, and neutrophil migration. Science 287, 1040-1046.

Saudemont, A., Garcon, F., Yadi, H., Roche-Molina, M., Kim, N., Segonds-Pichon, A., Martin-Fontecha, A., Okkenhaug, K., and Colucci, F. (2009). p110gamma and p110delta isoforms of phosphoinositide 3-kinase differentially regulate natural killer cell migration in health and disease. Proceedings of the National Academy of Sciences of the United States of America 106, 5795-5800.

Sawyer, C., Sturge, J., Bennett, D.C., O'Hare, M.J., Allen, W.E., Bain, J., Jones, G.E., and Vanhaesebroeck, B. (2003). Regulation of breast cancer cell chemotaxis by the phosphoinositide 3-kinase p110delta. Cancer research 63, 1667-1675.

Schade, B., Rao, T., Dourdin, N., Lesurf, R., Hallett, M., Cardiff, R.D., and Muller, W.J. (2009). PTEN deficiency in a luminal ErbB-2 mouse model results in dramatic acceleration of mammary tumorigenesis and metastasis. The Journal of biological chemistry 284, 19018-19026.

Schmid, M.C., Avraamides, C.J., Dippold, H.C., Franco, I., Foubert, P., Ellies, L.G., Acevedo, L.M., Manglicmot, J.R., Song, X., Wrasidlo, W., et al. (2011). Receptor tyrosine kinases and TLR/IL1Rs unexpectedly activate myeloid cell PI3kgamma, a single convergent point promoting tumor inflammation and progression. Cancer cell 19, 715- 727.

Schmid, M.C., Franco, I., Kang, S.W., Hirsch, E., Quilliam, L.A., and Varner, J.A. (2013). PI3-kinase gamma promotes Rap1a-mediated activation of myeloid cell integrin alpha4beta1, leading to tumor inflammation and growth. PloS one 8, e60226.

Schmit, F., Utermark, T., Zhang, S., Wang, Q., Von, T., Roberts, T.M., and Zhao, J.J. (2014). PI3K isoform dependence of PTEN-deficient tumors can be altered by the genetic context. Proceedings of the National Academy of Sciences of the United States of America 111, 6395-6400.

Schu, P.V., Takegawa, K., Fry, M.J., Stack, J.H., Waterfield, M.D., and Emr, S.D. (1993). Phosphatidylinositol 3-kinase encoded by yeast VPS34 gene essential for protein sorting. Science 260, 88-91.

Sergina, N.V., Rausch, M., Wang, D., Blair, J., Hann, B., Shokat, K.M., and Moasser, M.M. (2007). Escape from HER-family tyrosine kinase inhibitor therapy by the kinase- inactive HER3. Nature 445, 437-441.

184

Serra, V., Scaltriti, M., Prudkin, L., Eichhorn, P.J., Ibrahim, Y.H., Chandarlapaty, S., Markman, B., Rodriguez, O., Guzman, M., Rodriguez, S., et al. (2011). PI3K inhibition results in enhanced HER signaling and acquired ERK dependency in HER2- overexpressing breast cancer. Oncogene 30, 2547-2557.

Shi, F., Telesco, S.E., Liu, Y., Radhakrishnan, R., and Lemmon, M.A. (2010). ErbB3/HER3 intracellular domain is competent to bind ATP and catalyze autophosphorylation. Proceedings of the National Academy of Sciences of the United States of America 107, 7692-7697.

Shymanets, A., Prajwal, Bucher, K., Beer-Hammer, S., Harteneck, C., and Nurnberg, B. (2013). p87 and p101 subunits are distinct regulators determining class IB phosphoinositide 3-kinase (PI3K) specificity. The Journal of biological chemistry 288, 31059-31068.

Siegel, P.M., Ryan, E.D., Cardiff, R.D., and Muller, W.J. (1999). Elevated expression of activated forms of Neu/ErbB-2 and ErbB-3 are involved in the induction of mammary tumors in transgenic mice: implications for human breast cancer. The EMBO journal 18, 2149-2164.

Sierke, S.L., Cheng, K., Kim, H.H., and Koland, J.G. (1997). Biochemical characterization of the protein tyrosine kinase homology domain of the ErbB3 (HER3) receptor protein. The Biochemical journal 322 ( Pt 3), 757-763.

Soltoff, S.P., Carraway, K.L., 3rd, Prigent, S.A., Gullick, W.G., and Cantley, L.C. (1994). ErbB3 is involved in activation of phosphatidylinositol 3-kinase by epidermal growth factor. Molecular and cellular biology 14, 3550-3558.

Song, M.S., Carracedo, A., Salmena, L., Song, S.J., Egia, A., Malumbres, M., and Pandolfi, P.P. (2011). Nuclear PTEN regulates the APC-CDH1 tumor-suppressive complex in a phosphatase-independent manner. Cell 144, 187-199.

Song, M.S., Salmena, L., and Pandolfi, P.P. (2012). The functions and regulation of the PTEN tumour suppressor. Nature reviews. Molecular cell biology 13, 283-296.

Sopasakis, V.R., Liu, P., Suzuki, R., Kondo, T., Winnay, J., Tran, T.T., Asano, T., Smyth, G., Sajan, M.P., Farese, R.V., et al. (2010). Specific roles of the p110alpha isoform of phosphatidylinsositol 3-kinase in hepatic insulin signaling and metabolic regulation. Cell metabolism 11, 220-230.

Sorlie, T., Perou, C.M., Tibshirani, R., Aas, T., Geisler, S., Johnsen, H., Hastie, T., Eisen, M.B., van de Rijn, M., Jeffrey, S.S., et al. (2001). Gene expression patterns of breast carcinomas distinguish tumor subclasses with clinical implications. Proceedings of the National Academy of Sciences of the United States of America 98, 10869-10874.

Sorlie, T., Tibshirani, R., Parker, J., Hastie, T., Marron, J.S., Nobel, A., Deng, S., Johnsen, H., Pesich, R., Geisler, S., et al. (2003). Repeated observation of breast tumor subtypes in independent gene expression data sets. Proceedings of the National Academy of Sciences of the United States of America 100, 8418-8423.

185

Stambolic, V., Suzuki, A., de la Pompa, J.L., Brothers, G.M., Mirtsos, C., Sasaki, T., Ruland, J., Penninger, J.M., Siderovski, D.P., and Mak, T.W. (1998). Negative regulation of PKB/Akt-dependent cell survival by the tumor suppressor PTEN. Cell 95, 29-39.

Stenmark, H. (2009). Rab GTPases as coordinators of vesicle traffic. Nature reviews. Molecular cell biology 10, 513-525.

Stephens, L.R., Eguinoa, A., Erdjument-Bromage, H., Lui, M., Cooke, F., Coadwell, J., Smrcka, A.S., Thelen, M., Cadwallader, K., Tempst, P., et al. (1997). The G beta gamma sensitivity of a PI3K is dependent upon a tightly associated adaptor, p101. Cell 89, 105- 114.

Stjernstrom, A., Karlsson, C., Fernandez, O.J., Soderkvist, P., Karlsson, M.G., and Thunell, L.K. (2014). Alterations of INPP4B, PIK3CA and pAkt of the PI3K pathway are associated with squamous cell carcinoma of the lung. Cancer medicine.

Stoyanov, B., Volinia, S., Hanck, T., Rubio, I., Loubtchenkov, M., Malek, D., Stoyanova, S., Vanhaesebroeck, B., Dhand, R., Nurnberg, B., et al. (1995). Cloning and characterization of a G protein-activated human phosphoinositide-3 kinase. Science 269, 690-693.

Subramaniam, P.S., Whye, D.W., Efimenko, E., Chen, J., Tosello, V., De Keersmaecker, K., Kashishian, A., Thompson, M.A., Castillo, M., Cordon-Cardo, C., et al. (2012). Targeting nonclassical oncogenes for therapy in T-ALL. Cancer cell 21, 459-472.

Suire, S., Coadwell, J., Ferguson, G.J., Davidson, K., Hawkins, P., and Stephens, L. (2005). p84, a new Gbetagamma-activated regulatory subunit of the type IB phosphoinositide 3-kinase p110gamma. Current biology : CB 15, 566-570.

Suire, S., Condliffe, A.M., Ferguson, G.J., Ellson, C.D., Guillou, H., Davidson, K., Welch, H., Coadwell, J., Turner, M., Chilvers, E.R., et al. (2006). Gbetagammas and the Ras binding domain of p110gamma are both important regulators of PI(3)Kgamma signalling in neutrophils. Nature cell biology 8, 1303-1309.

Suire, S., Hawkins, P., and Stephens, L. (2002). Activation of phosphoinositide 3-kinase gamma by Ras. Current biology : CB 12, 1068-1075.

Sun, M., Hillmann, P., Hofmann, B.T., Hart, J.R., and Vogt, P.K. (2010). Cancer-derived mutations in the regulatory subunit p85alpha of phosphoinositide 3-kinase function through the catalytic subunit p110alpha. Proceedings of the National Academy of Sciences of the United States of America 107, 15547-15552.

Taniguchi, C.M., Tran, T.T., Kondo, T., Luo, J., Ueki, K., Cantley, L.C., and Kahn, C.R. (2006). Phosphoinositide 3-kinase regulatory subunit p85alpha suppresses insulin action via positive regulation of PTEN. Proceedings of the National Academy of Sciences of the United States of America 103, 12093-12097.

Taniguchi, C.M., Winnay, J., Kondo, T., Bronson, R.T., Guimaraes, A.R., Aleman, J.O., Luo, J., Stephanopoulos, G., Weissleder, R., Cantley, L.C., et al. (2010). The phosphoinositide 3-kinase regulatory subunit p85alpha can exert tumor suppressor

186

properties through negative regulation of growth factor signaling. Cancer research 70, 5305-5315.

Tao, J.J., Castel, P., Radosevic-Robin, N., Elkabets, M., Auricchio, N., Aceto, N., Weitsman, G., Barber, P., Vojnovic, B., Ellis, H., et al. (2014). Antagonism of EGFR and HER3 enhances the response to inhibitors of the PI3K-Akt pathway in triple-negative breast cancer. Science signaling 7, ra29.

Tenbaum, S.P., Ordonez-Moran, P., Puig, I., Chicote, I., Arques, O., Landolfi, S., Fernandez, Y., Herance, J.R., Gispert, J.D., Mendizabal, L., et al. (2012). beta-catenin confers resistance to PI3K and AKT inhibitors and subverts FOXO3a to promote metastasis in colon cancer. Nature medicine 18, 892-901.

Terauchi, Y., Tsuji, Y., Satoh, S., Minoura, H., Murakami, K., Okuno, A., Inukai, K., Asano, T., Kaburagi, Y., Ueki, K., et al. (1999). Increased insulin sensitivity and hypoglycaemia in mice lacking the p85 alpha subunit of phosphoinositide 3-kinase. Nature genetics 21, 230-235.

Thomas, R.K., Baker, A.C., Debiasi, R.M., Winckler, W., Laframboise, T., Lin, W.M., Wang, M., Feng, W., Zander, T., MacConaill, L., et al. (2007). High-throughput oncogene mutation profiling in human cancer. Nature genetics 39, 347-351.

Tian, L., Fang, Y.X., Xue, J.L., and Chen, J.Z. (2013). Four microRNAs promote prostate cell proliferation with regulation of PTEN and its downstream signals in vitro. PloS one 8, e75885.

Torbett, N.E., Luna-Moran, A., Knight, Z.A., Houk, A., Moasser, M., Weiss, W., Shokat, K.M., and Stokoe, D. (2008). A chemical screen in diverse breast cancer cell lines reveals genetic enhancers and suppressors of sensitivity to PI3K isoform-selective inhibition. The Biochemical journal 415, 97-110.

Tzahar, E., Waterman, H., Chen, X., Levkowitz, G., Karunagaran, D., Lavi, S., Ratzkin, B.J., and Yarden, Y. (1996). A hierarchical network of interreceptor interactions determines signal transduction by Neu differentiation factor/neuregulin and epidermal growth factor. Molecular and cellular biology 16, 5276-5287.

Ueki, K., Algenstaedt, P., Mauvais-Jarvis, F., and Kahn, C.R. (2000). Positive and negative regulation of phosphoinositide 3-kinase-dependent signaling pathways by three different gene products of the p85alpha regulatory subunit. Molecular and cellular biology 20, 8035-8046.

Ueki, K., Fruman, D.A., Brachmann, S.M., Tseng, Y.H., Cantley, L.C., and Kahn, C.R. (2002a). Molecular balance between the regulatory and catalytic subunits of phosphoinositide 3-kinase regulates cell signaling and survival. Molecular and cellular biology 22, 965-977.

Ueki, K., Fruman, D.A., Yballe, C.M., Fasshauer, M., Klein, J., Asano, T., Cantley, L.C., and Kahn, C.R. (2003). Positive and negative roles of p85 alpha and p85 beta regulatory subunits of phosphoinositide 3-kinase in insulin signaling. The Journal of biological chemistry 278, 48453-48466.

187

Ueki, K., Yballe, C.M., Brachmann, S.M., Vicent, D., Watt, J.M., Kahn, C.R., and Cantley, L.C. (2002b). Increased insulin sensitivity in mice lacking p85beta subunit of phosphoinositide 3-kinase. Proceedings of the National Academy of Sciences of the United States of America 99, 419-424.

Urick, M.E., Rudd, M.L., Godwin, A.K., Sgroi, D., Merino, M., and Bell, D.W. (2011). PIK3R1 (p85alpha) is somatically mutated at high frequency in primary endometrial cancer. Cancer research 71, 4061-4067.

Ursini-Siegel, J., Hardy, W.R., Zuo, D., Lam, S.H., Sanguin-Gendreau, V., Cardiff, R.D., Pawson, T., and Muller, W.J. (2008). ShcA signalling is essential for tumour progression in mouse models of human breast cancer. The EMBO journal 27, 910-920.

Utermark, T., Rao, T., Cheng, H., Wang, Q., Lee, S.H., Wang, Z.C., Iglehart, J.D., Roberts, T.M., Muller, W.J., and Zhao, J.J. (2012). The p110alpha and p110beta isoforms of PI3K play divergent roles in mammary gland development and tumorigenesis. Genes & development 26, 1573-1586.

Utermark, T., Schaffhausen, B.S., Roberts, T.M., and Zhao, J.J. (2007). The p110alpha isoform of phosphatidylinositol 3-kinase is essential for polyomavirus middle T antigen- mediated transformation. Journal of virology 81, 7069-7076.

Vadas, O., Dbouk, H.A., Shymanets, A., Perisic, O., Burke, J.E., Abi Saab, W.F., Khalil, B.D., Harteneck, C., Bresnick, A.R., Nurnberg, B., et al. (2013). Molecular determinants of PI3Kgamma-mediated activation downstream of G-protein-coupled receptors (GPCRs). Proceedings of the National Academy of Sciences of the United States of America 110, 18862-18867.

Vanhaesebroeck, B., Guillermet-Guibert, J., Graupera, M., and Bilanges, B. (2010). The emerging mechanisms of isoform-specific PI3K signalling. Nature reviews. Molecular cell biology 11, 329-341.

Vanhaesebroeck, B., Jones, G.E., Allen, W.E., Zicha, D., Hooshmand-Rad, R., Sawyer, C., Wells, C., Waterfield, M.D., and Ridley, A.J. (1999). Distinct PI(3)Ks mediate mitogenic signalling and cell migration in macrophages. Nature cell biology 1, 69-71.

Vanhaesebroeck, B., and Khwaja, A. (2014). PI3Kdelta inhibition hits a sensitive spot in B cell malignancies. Cancer cell 25, 269-271.

Vanhaesebroeck, B., Welham, M.J., Kotani, K., Stein, R., Warne, P.H., Zvelebil, M.J., Higashi, K., Volinia, S., Downward, J., and Waterfield, M.D. (1997). P110delta, a novel phosphoinositide 3-kinase in leukocytes. Proceedings of the National Academy of Sciences of the United States of America 94, 4330-4335.

Velichkova, M., Juan, J., Kadandale, P., Jean, S., Ribeiro, I., Raman, V., Stefan, C., and Kiger, A.A. (2010). Drosophila Mtm and class II PI3K coregulate a PI(3)P pool with cortical and endolysosomal functions. The Journal of cell biology 190, 407-425.

Vogt, P.K. (2011). PI3K p110beta: more tightly controlled or constitutively active? Molecular cell 41, 499-501.

188

Voigt, P., Dorner, M.B., and Schaefer, M. (2006). Characterization of p87PIKAP, a novel regulatory subunit of phosphoinositide 3-kinase gamma that is highly expressed in heart and interacts with PDE3B. The Journal of biological chemistry 281, 9977-9986.

Volinia, S., Dhand, R., Vanhaesebroeck, B., MacDougall, L.K., Stein, R., Zvelebil, M.J., Domin, J., Panaretou, C., and Waterfield, M.D. (1995). A human phosphatidylinositol 3- kinase complex related to the yeast Vps34p-Vps15p protein sorting system. The EMBO journal 14, 3339-3348.

Vora, S.R., Juric, D., Kim, N., Mino-Kenudson, M., Huynh, T., Costa, C., Lockerman, E.L., Pollack, S.F., Liu, M., Li, X., et al. (2014). CDK 4/6 Inhibitors Sensitize PIK3CA Mutant Breast Cancer to PI3K Inhibitors. Cancer cell 26, 136-149.

Wagner, K.U., Wall, R.J., St-Onge, L., Gruss, P., Wynshaw-Boris, A., Garrett, L., Li, M., Furth, P.A., and Hennighausen, L. (1997). Cre-mediated gene deletion in the mammary gland. Nucleic acids research 25, 4323-4330.

Wang, J., Hannon, G.J., and Beach, D.H. (2000). Risky immortalization by telomerase. Nature 405, 755-756.

Wang, Q., Von, T., Bronson, R., Ruan, M., Mu, W., Huang, A., Maira, S.M., and Zhao, J.J. (2013a). Spatially distinct roles of class Ia PI3K isoforms in the development and maintenance of PTEN hamartoma tumor syndrome. Genes & development 27, 1568- 1580.

Wang, Q., Weisberg, E., and Zhao, J.J. (2013b). The gene dosage of class Ia PI3K dictates the development of PTEN hamartoma tumor syndrome. Cell cycle 12, 3589- 3593.

Wang, S., Gao, J., Lei, Q., Rozengurt, N., Pritchard, C., Jiao, J., Thomas, G.V., Li, G., Roy-Burman, P., Nelson, P.S., et al. (2003). Prostate-specific deletion of the murine Pten tumor suppressor gene leads to metastatic prostate cancer. Cancer cell 4, 209- 221.

Watson, C.J., and Khaled, W.T. (2008). Mammary development in the embryo and adult: a journey of morphogenesis and commitment. Development 135, 995-1003.

Wee, S., Wiederschain, D., Maira, S.M., Loo, A., Miller, C., deBeaumont, R., Stegmeier, F., Yao, Y.M., and Lengauer, C. (2008). PTEN-deficient cancers depend on PIK3CB. Proceedings of the National Academy of Sciences of the United States of America 105, 13057-13062.

Wei, Y., Zou, Z., Becker, N., Anderson, M., Sumpter, R., Xiao, G., Kinch, L., Koduru, P., Christudass, C.S., Veltri, R.W., et al. (2013). EGFR-mediated Beclin 1 phosphorylation in autophagy suppression, tumor progression, and tumor chemoresistance. Cell 154, 1269-1284.

Weigelt, B., Warne, P.H., Lambros, M.B., Reis-Filho, J.S., and Downward, J. (2013). PI3K pathway dependencies in endometrioid endometrial cancer cell lines. Clinical cancer research : an official journal of the American Association for Cancer Research 19, 3533-3544.

189

Welch, H.C., Coadwell, W.J., Ellson, C.D., Ferguson, G.J., Andrews, S.R., Erdjument- Bromage, H., Tempst, P., Hawkins, P.T., and Stephens, L.R. (2002). P-Rex1, a PtdIns(3,4,5)P3- and Gbetagamma-regulated guanine-nucleotide exchange factor for Rac. Cell 108, 809-821.

Will, M., Qin, A.C., Toy, W., Yao, Z., Rodrik-Outmezguine, V., Schneider, C., Huang, X., Monian, P., Jiang, X., de Stanchina, E., et al. (2014). Rapid induction of apoptosis by PI3K inhibitors is dependent upon their transient inhibition of RAS-ERK signaling. Cancer discovery 4, 334-347.

Winnay, J.N., Boucher, J., Mori, M.A., Ueki, K., and Kahn, C.R. (2010). A regulatory subunit of phosphoinositide 3-kinase increases the nuclear accumulation of X-box- binding protein-1 to modulate the unfolded protein response. Nature medicine 16, 438- 445.

Wong, K.K., Engelman, J.A., and Cantley, L.C. (2010). Targeting the PI3K signaling pathway in cancer. Current opinion in genetics & development 20, 87-90.

Wu, H., Shekar, S.C., Flinn, R.J., El-Sibai, M., Jaiswal, B.S., Sen, K.I., Janakiraman, V., Seshagiri, S., Gerfen, G.J., Girvin, M.E., et al. (2009). Regulation of Class IA PI 3- kinases: C2 domain-iSH2 domain contacts inhibit p85/p110alpha and are disrupted in oncogenic p85 mutants. Proceedings of the National Academy of Sciences of the United States of America 106, 20258-20263.

Wu, R., Baker, S.J., Hu, T.C., Norman, K.M., Fearon, E.R., and Cho, K.R. (2013). Type I to type II ovarian carcinoma progression: mutant Trp53 or Pik3ca confers a more aggressive tumor phenotype in a mouse model of ovarian cancer. The American journal of pathology 182, 1391-1399.

Yeung, W.W., and Wong, Y.H. (2010). Galpha16 interacts with Class IA phosphatidylinositol 3-kinases and inhibits Akt signaling. Cellular signalling 22, 1379- 1387.

Yoshioka, K., Yoshida, K., Cui, H., Wakayama, T., Takuwa, N., Okamoto, Y., Du, W., Qi, X., Asanuma, K., Sugihara, K., et al. (2012). Endothelial PI3K-C2alpha, a class II PI3K, has an essential role in angiogenesis and vascular barrier function. Nature medicine 18, 1560-1569.

Yu, J., Wjasow, C., and Backer, J.M. (1998a). Regulation of the p85/p110alpha phosphatidylinositol 3'-kinase. Distinct roles for the n-terminal and c-terminal SH2 domains. The Journal of biological chemistry 273, 30199-30203.

Yu, J., Zhang, Y., McIlroy, J., Rordorf-Nikolic, T., Orr, G.A., and Backer, J.M. (1998b). Regulation of the p85/p110 phosphatidylinositol 3'-kinase: stabilization and inhibition of the p110alpha catalytic subunit by the p85 regulatory subunit. Molecular and cellular biology 18, 1379-1387.

Yuan, W., Stawiski, E., Janakiraman, V., Chan, E., Durinck, S., Edgar, K.A., Kljavin, N.M., Rivers, C.S., Gnad, F., Roose-Girma, M., et al. (2013). Conditional activation of Pik3ca(H1047R) in a knock-in mouse model promotes mammary tumorigenesis and emergence of mutations. Oncogene 32, 318-326.

190

Yum, H.K., Arcaroli, J., Kupfner, J., Shenkar, R., Penninger, J.M., Sasaki, T., Yang, K.Y., Park, J.S., and Abraham, E. (2001). Involvement of phosphoinositide 3-kinases in neutrophil activation and the development of acute lung injury. Journal of immunology 167, 6601-6608.

Zhang, X., Vadas, O., Perisic, O., Anderson, K.E., Clark, J., Hawkins, P.T., Stephens, L.R., and Williams, R.L. (2011). Structure of lipid kinase p110beta/p85beta elucidates an unusual SH2-domain-mediated inhibitory mechanism. Molecular cell 41, 567-578.

Zhao, J.J., Cheng, H., Jia, S., Wang, L., Gjoerup, O.V., Mikami, A., and Roberts, T.M. (2006). The p110alpha isoform of PI3K is essential for proper growth factor signaling and oncogenic transformation. Proceedings of the National Academy of Sciences of the United States of America 103, 16296-16300.

Zhao, J.J., Gjoerup, O.V., Subramanian, R.R., Cheng, Y., Chen, W., Roberts, T.M., and Hahn, W.C. (2003). Human mammary epithelial cell transformation through the activation of phosphatidylinositol 3-kinase. Cancer cell 3, 483-495.

Zhao, J.J., Liu, Z., Wang, L., Shin, E., Loda, M.F., and Roberts, T.M. (2005). The oncogenic properties of mutant p110alpha and p110beta phosphatidylinositol 3-kinases in human mammary epithelial cells. Proceedings of the National Academy of Sciences of the United States of America 102, 18443-18448.

Zhao, L., and Vogt, P.K. (2008). Helical domain and kinase domain mutations in p110alpha of phosphatidylinositol 3-kinase induce gain of function by different mechanisms. Proceedings of the National Academy of Sciences of the United States of America 105, 2652-2657.

Zhao, L., and Vogt, P.K. (2010). Hot-spot mutations in p110alpha of phosphatidylinositol 3-kinase (pI3K): differential interactions with the regulatory subunit p85 and with RAS. Cell cycle 9, 596-600.

Zheng, Y., Yin, L., Chen, H., Yang, S., Pan, C., Lu, S., Miao, M., and Jiao, B. (2012). miR-376a suppresses proliferation and induces apoptosis in hepatocellular carcinoma. FEBS letters 586, 2396-2403.

191

Appendix A: Supplemental table of class I PI3K alterations in cancer, with complete references

Alteration Type Cancer Type Frequency of Sample References Alteration Size Range Class IA PIK3CA (p110) Mutation Endometrial 10.3-53.0% 29-232 1, 2 Breast 7.1-35.5% 65-507 3-9 Ovarian (CC) 33.0% 97 10 Colorectal 16.9†-30.6% 72-195 11, 12 Bladder 5.0-20.0% 20-130 13-16 Lung (SCC) 20.0% 5 17 Lung (SQCC) 2.9-16.8% 35-178 17, 18 Lung (LCC) 11.9% 9 17 Lung (ADC) 0.6-4.3% 57-183 3, 17, 19, 20 Cervical 13.6% 22 2 Glioblastoma 4.3-11.0% 91-291 21-24 Head and neck 8.1-9.4% 32-74 25, 26 Esophageal 5.5% 145 27 Melanoma 5.0% 121 28 Prostate 1.3-3.6% 55-156 3, 29-31 Sarcoma 2.9% 207 32 Renal (CC) 1.0-2.9% 98-417 33, 34 Liver (HCC) 1.6% 125 35 Megalencephaly‡ 48.0% 50 36 Copy number Head and neck 9.1-100% 11-117 37-39 gain/amplification Cervical 9.1-76.4% 22-55 2, 40 Lung (SQCC) 42.9-69.6% 28-52 17, 41, 42 Lung (SCC) 33.3-66.7% 3-12 17, 41 Lung (LCC) 16.7-37.5% 6-16 17, 41 Lung (ADC) 9.5-19.1% 47-74 17, 41 Lung (NSCLC) 12.0% 92 43 Lymphoma (MCL) 68.2% 22 44 Lymphoma (DLBCL) 16.7% 60 45 Ovarian 39.8% 93 46 Ovarian (Serous) 13.3-24.3% 60-74 47, 48 Gastric 36.4% 55 49 Thyroid 30.0% 110 50 Prostate 28.1% 32 16 Breast 8.7-13.4% 92-209 8, 9 Glioblastoma 1.9-12.2% 139-206 21, 22 Endometrial 10.3% 29 2 Thyroid 9.4% 128 51 Esophageal 5.7% 87 52 Leukemia (CLL) 5.6% 161 53 Increased expression Prostate 40.0% 25 16

192

Appendix A: Supplemental table of class I PI3K alterations in cancer, with complete references (continued)

Alteration Type Cancer Type Frequency of Sample References Alteration Size Range Class IA PIK3CB (p110) Mutation Breast 0.5% 183 3, 54 Copy number Lung (SQCC) 56.5% 46 42 gain/amplification Thyroid 42.3% 97 50 Ovarian 5-26.9% NA-93 46, 55 Lymphoma (DLBCL) 20.0% 60 45 Glioblastoma 5.8% 103 56 Breast 4.9-5% NA-81 55, 57 Increased expression Prostate 46.7% 30 58 Glioblastoma 3.9% 103 56 PIK3CD (p110) Copy number gain Glioblastoma 40.0% 10 59 Increased expression Neuroblastoma 52.6% 19 60 Glioblastoma 5.8% 103 56 PIK3R1 (p85, p55, p50) Mutation Endometrial 19.8-32.8% 108-243 1, 61, 62 Pancreatic 16.7% 6 63 Glioblastoma 7.6-11.3% 91-291 22-24 Colorectal 4.6†-8.3% 108-195 11, 63 Melanoma 4.4% 68 64 Ovarian 3.8% 80 65 Esophageal 3.4% 145 27 Breast 1.1-2.8% 62-507 3, 4, 63, 66 Colon 1.7% 60 65 Decreased expression Breast 61.8% 458 66 Prostate 17-75%* NA 67 Lung 19-46%* NA 67 Ovarian 22%* NA 67 Breast 18%* NA 67 Bladder 18%* NA 67 Copy number loss Ovarian 21.5% 93 46 PIK3R2 (p85) Mutation Endometrial 4.9% 243 61 Colorectal 0.9% 108 63 Megalencephaly‡ 22.0% 50 36 Amplification Lymphoma (DLBCL) 23.3% 60 45 Increased expression Colon 55.0% 20 68 Breast 45.7% 35 68 PIK3R3 (p55) Copy number gain Ovarian 15.0% 93 46

193

Appendix A: Supplemental table of class I PI3K alterations in cancer, with complete references (continued)

Alteration Type Cancer Type Frequency of Sample References Alteration Size Range Class IB PIK3CG (p110) Copy number gain Ovarian 19.3% 93 46 Increased expression Breast 77.5% 40 69 Prostate 72.4% 29 70 Medulloblastoma 52.9% 17 71 PIK3R5 (p101) Mutation Melanoma 38.2% 68 64 Gastric 2.7% 37 63

CC, clear cell; SCC, small cell carcinoma; SQCC, squamous cell carcinoma; ADC, adenocarcinoma; LCC, large cell carcinoma; NSCLC, non-small cell lung carcinoma; MCL, mantle cell lymphoma; CLL, chronic lymphocytic leukemia; DLBCL, diffuse large B cell lymphoma; HCC, hepatocellular carcinoma.

‡ Megalencephaly syndromes are a collection of sporadic overgrowth disorders characterized by enlarged brain size and other distinct features.

† Combined number of hypermutated and non-hypermutated colon and colorectal patient samples with mutations in the indicated gene.

* Represents the percent reduction in gene expression.

NA Sample size not available for this study.

194

Appendix A references:

1. Cancer Genome Atlas Research, N. et al. Integrated genomic characterization of endometrial carcinoma. Nature 497, 67-73 (2013).

2. Miyake, T. et al. PIK3CA gene mutations and amplifications in uterine cancers, identified by methods that avoid confounding by PIK3CA pseudogene sequences. Cancer Lett 261, 120-6 (2008).

3. Kan, Z. et al. Diverse somatic mutation patterns and pathway alterations in human cancers. Nature 466, 869-73 (2010).

4. Cancer Genome Atlas, N. Comprehensive molecular portraits of human breast tumours. Nature 490, 61-70 (2012).

5. Shah, S.P. et al. The clonal and mutational evolution spectrum of primary triple-negative breast cancers. Nature 486, 395-9 (2012).

6. Banerji, S. et al. Sequence analysis of mutations and translocations across breast cancer subtypes. Nature 486, 405-9 (2012).

7. Stephens, P.J. et al. The landscape of cancer genes and mutational processes in breast cancer. Nature 486, 400-4 (2012).

8. Lopez-Knowles, E. et al. PI3K pathway activation in breast cancer is associated with the basal-like phenotype and cancer-specific mortality. Int J Cancer 126, 1121-31 (2010).

9. Wu, G. et al. Somatic mutation and gain of copy number of PIK3CA in human breast cancer. Breast Cancer Res 7, R609-16 (2005).

10. Kuo, K.T. et al. Frequent activating mutations of PIK3CA in ovarian clear cell carcinoma. Am J Pathol 174, 1597-601 (2009).

11. Cancer Genome Atlas, N. Comprehensive molecular characterization of human colon and rectal cancer. Nature 487, 330-7 (2012).

12. Seshagiri, S. et al. Recurrent R-spondin fusions in colon cancer. Nature 488, 660-4 (2012).

13. The Cancer Genome Atlas Research, N. Comprehensive molecular characterization of urothelial bladder carcinoma. Nature (2014).

14. Iyer, G. et al. Prevalence and co-occurrence of actionable genomic alterations in high- grade bladder cancer. J Clin Oncol 31, 3133-40 (2013).

15. Guo, G. et al. Whole-genome and whole-exome sequencing of bladder cancer identifies frequent alterations in genes involved in sister chromatid cohesion and segregation. Nat Genet 45, 1459-63 (2013).

16. Agell, L. et al. PI3K signaling pathway is activated by PIK3CA mRNA overexpression and copy gain in prostate tumors, but PIK3CA, BRAF, KRAS and AKT1 mutations are infrequent events. Mod Pathol 24, 443-52 (2011).

195

17. Okudela, K. et al. PIK3CA mutation and amplification in human lung cancer. Pathol Int 57, 664-71 (2007).

18. Cancer Genome Atlas Research, N. Comprehensive genomic characterization of squamous cell lung cancers. Nature 489, 519-25 (2012).

19. Ding, L. et al. Somatic mutations affect key pathways in lung adenocarcinoma. Nature 455, 1069-75 (2008).

20. Imielinski, M. et al. Mapping the hallmarks of lung adenocarcinoma with massively parallel sequencing. Cell 150, 1107-20 (2012).

21. Kita, D., Yonekawa, Y., Weller, M. & Ohgaki, H. PIK3CA alterations in primary (de novo) and secondary glioblastomas. Acta Neuropathol 113, 295-302 (2007).

22. Cancer Genome Atlas Research, N. Comprehensive genomic characterization defines human glioblastoma genes and core pathways. Nature 455, 1061-8 (2008).

23. Parsons, D.W. et al. An integrated genomic analysis of human glioblastoma multiforme. Science 321, 1807-12 (2008).

24. Brennan, C.W. et al. The somatic genomic landscape of glioblastoma. Cell 155, 462-77 (2013).

25. Stransky, N. et al. The mutational landscape of head and neck squamous cell carcinoma. Science 333, 1157-60 (2011).

26. Agrawal, N. et al. Exome sequencing of head and neck squamous cell carcinoma reveals inactivating mutations in NOTCH1. Science 333, 1154-7 (2011).

27. Dulak, A.M. et al. Exome and whole-genome sequencing of esophageal adenocarcinoma identifies recurrent driver events and mutational complexity. Nat Genet 45, 478-86 (2013).

28. Hodis, E. et al. A landscape of driver mutations in melanoma. Cell 150, 251-63 (2012).

29. Taylor, B.S. et al. Integrative genomic profiling of human prostate cancer. Cancer Cell 18, 11-22 (2010).

30. Baca, S.C. et al. Punctuated evolution of prostate cancer genomes. Cell 153, 666-77 (2013).

31. Barbieri, C.E. et al. Exome sequencing identifies recurrent SPOP, FOXA1 and MED12 mutations in prostate cancer. Nat Genet 44, 685-9 (2012).

32. Barretina, J. et al. Subtype-specific genomic alterations define new targets for soft-tissue sarcoma therapy. Nat Genet 42, 715-21 (2010).

33. Cancer Genome Atlas Research, N. Comprehensive molecular characterization of clear cell renal cell carcinoma. Nature 499, 43-9 (2013).

34. Guo, G. et al. Frequent mutations of genes encoding ubiquitin-mediated proteolysis pathway components in clear cell renal cell carcinoma. Nat Genet 44, 17-9 (2012).

196

35. Guichard, C. et al. Integrated analysis of somatic mutations and focal copy-number changes identifies key genes and pathways in hepatocellular carcinoma. Nat Genet 44, 694-8 (2012).

36. Riviere, J.B. et al. De novo germline and postzygotic mutations in AKT3, PIK3R2 and PIK3CA cause a spectrum of related megalencephaly syndromes. Nat Genet 44, 934-40 (2012).

37. Fenic, I., Steger, K., Gruber, C., Arens, C. & Woenckhaus, J. Analysis of PIK3CA and Akt/protein kinase B in head and neck squamous cell carcinoma. Oncol Rep 18, 253-9 (2007).

38. Pedrero, J.M. et al. Frequent genetic and biochemical alterations of the PI 3- K/AKT/PTEN pathway in head and neck squamous cell carcinoma. Int J Cancer 114, 242-8 (2005).

39. Woenckhaus, J. et al. Genomic gain of PIK3CA and increased expression of p110alpha are associated with progression of dysplasia into invasive squamous cell carcinoma. J Pathol 198, 335-42 (2002).

40. Ma, Y.Y. et al. PIK3CA as an oncogene in cervical cancer. Oncogene 19, 2739-44 (2000).

41. Massion, P.P. et al. Early involvement of the phosphatidylinositol 3-kinase/Akt pathway in lung cancer progression. Am J Respir Crit Care Med 170, 1088-94 (2004).

42. Massion, P.P. et al. Genomic copy number analysis of non-small cell lung cancer using array comparative genomic hybridization: implications of the phosphatidylinositol 3- kinase pathway. Cancer Res 62, 3636-40 (2002).

43. Kawano, O. et al. PIK3CA gene amplification in Japanese non-small cell lung cancer. Lung Cancer 58, 159-60 (2007).

44. Psyrri, A. et al. Phosphatidylinositol 3'-kinase catalytic subunit alpha gene amplification contributes to the pathogenesis of mantle cell lymphoma. Clin Cancer Res 15, 5724-32 (2009).

45. Cui, W. et al. Frequent copy number variations of PI3K/AKT pathway and aberrant protein expressions of PI3K subunits are associated with inferior survival in diffuse large B cell lymphoma. J Transl Med 12, 10 (2014).

46. Huang, J. et al. Frequent genetic abnormalities of the PI3K/AKT pathway in primary ovarian cancer predict patient outcome. Genes Chromosomes Cancer 50, 606-18 (2011).

47. Nakayama, K. et al. Sequence mutations and amplification of PIK3CA and AKT2 genes in purified ovarian serous neoplasms. Cancer Biol Ther 5, 779-85 (2006).

48. Nakayama, K. et al. Amplicon profiles in ovarian serous carcinomas. Int J Cancer 120, 2613-7 (2007).

49. Byun, D.S. et al. Frequent monoallelic deletion of PTEN and its reciprocal associatioin with PIK3CA amplification in gastric carcinoma. Int J Cancer 104, 318-27 (2003).

197

50. Liu, Z. et al. Highly prevalent genetic alterations in receptor tyrosine kinases and phosphatidylinositol 3-kinase/akt and mitogen-activated protein kinase pathways in anaplastic and follicular thyroid cancers. J Clin Endocrinol Metab 93, 3106-16 (2008).

51. Wu, G. et al. Uncommon mutation, but common amplifications, of the PIK3CA gene in thyroid tumors. J Clin Endocrinol Metab 90, 4688-93 (2005).

52. Miller, C.T. et al. Gene amplification in esophageal adenocarcinomas and Barrett's with high-grade dysplasia. Clin Cancer Res 9, 4819-25 (2003).

53. Brown, J.R. et al. Integrative genomic analysis implicates gain of PIK3CA at 3q26 and MYC at 8q24 in chronic lymphocytic leukemia. Clin Cancer Res 18, 3791-802 (2012).

54. Dbouk, H.A. et al. Characterization of a tumor-associated activating mutation of the p110beta PI 3-kinase. PLoS One 8, e63833 (2013).

55. Brugge, J., Hung, M.C. & Mills, G.B. A new mutational AKTivation in the PI3K pathway. Cancer Cell 12, 104-7 (2007).

56. Knobbe, C.B. & Reifenberger, G. Genetic alterations and aberrant expression of genes related to the phosphatidyl-inositol-3'-kinase/protein kinase B (Akt) signal transduction pathway in glioblastomas. Brain Pathol 13, 507-18 (2003).

57. Crowder, R.J. et al. PIK3CA and PIK3CB inhibition produce synthetic lethality when combined with estrogen deprivation in estrogen receptor-positive breast cancer. Cancer Res 69, 3955-62 (2009).

58. Zhu, Q. et al. Phosphoinositide 3-OH kinase p85alpha and p110beta are essential for androgen receptor transactivation and tumor progression in prostate cancers. Oncogene 27, 4569-79 (2008).

59. Mizoguchi, M., Nutt, C.L., Mohapatra, G. & Louis, D.N. Genetic alterations of phosphoinositide 3-kinase subunit genes in human glioblastomas. Brain Pathol 14, 372- 7 (2004).

60. Boller, D. et al. Targeting the phosphoinositide 3-kinase isoform p110delta impairs growth and survival in neuroblastoma cells. Clin Cancer Res 14, 1172-81 (2008).

61. Cheung, L.W. et al. High frequency of PIK3R1 and PIK3R2 mutations in endometrial cancer elucidates a novel mechanism for regulation of PTEN protein stability. Cancer Discov 1, 170-85 (2011).

62. Urick, M.E. et al. PIK3R1 (p85alpha) is somatically mutated at high frequency in primary endometrial cancer. Cancer Res 71, 4061-7 (2011).

63. Jaiswal, B.S. et al. Somatic mutations in p85alpha promote tumorigenesis through class IA PI3K activation. Cancer Cell 16, 463-74 (2009).

64. Shull, A.Y. et al. Novel somatic mutations to PI3K pathway genes in metastatic melanoma. PLoS One 7, e43369 (2012).

65. Philp, A.J. et al. The phosphatidylinositol 3'-kinase p85alpha gene is an oncogene in human ovarian and colon tumors. Cancer Res 61, 7426-9 (2001).

198

66. Cizkova, M. et al. PIK3R1 underexpression is an independent prognostic marker in breast cancer. BMC Cancer 13, 545 (2013).

67. Taniguchi, C.M. et al. The phosphoinositide 3-kinase regulatory subunit p85alpha can exert tumor suppressor properties through negative regulation of growth factor signaling. Cancer Res 70, 5305-15 (2010).

68. Cortes, I. et al. p85beta phosphoinositide 3-kinase subunit regulates tumor progression. Proc Natl Acad Sci U S A 109, 11318-23 (2012).

69. Xie, Y. et al. Identification of upregulated phosphoinositide 3-kinase gamma as a target to suppress breast cancer cell migration and invasion. Biochem Pharmacol 85, 1454-62 (2013).

70. Edling, C.E. et al. Key role of phosphoinositide 3-kinase class IB in pancreatic cancer. Clin Cancer Res 16, 4928-37 (2010).

71. Guerreiro, A.S. et al. A sensitized RNA interference screen identifies a novel role for the PI3K p110gamma isoform in medulloblastoma cell proliferation and chemoresistance. Mol Cancer Res 9, 925-35 (2011).

199

Appendix B: Supplemental table of genetically engineered mouse models of PI3K isoforms in cancer, with complete references.

Genotype Phenotype Ref PIK3CA (p110) LA2 RBD/RBD Protected from KRas-induced lung KRas ; Pik3ca 1 tumors Partial regression of KRas-induced Rosa26-Cre; KRasLA2; Pik3caRBD/flox 2 lung tumors Protected from Her2/neu-driven MMTV-Neu-IRES-Cre; Pik3caflox/flox 3 mammary tumors No effect on high-grade PIN driven by Pb-Cre; Ptenflox/flox; Pik3caflox/flox 4 Pten loss G12D flox/flox Protection from MPN induced by Mx1-Cre; KRas ; Pik3ca 5 oncogenic KRas No effect on MPN induced by Shp2 Mx1-Cre, LSL-Shp2GOF/+; Pik3caflox/flox 6 GOF Increased endometrial hyperplasia; Pten−/+; Pik3caKD/+ reduced pheochromocytoma and 7 thyroid tumors CCSP-rtTA; Tet-op-PIK3CAH1047R Develop lung tumors within 3 months 8 Develop mammary tumors within 7 MMTV-rtTA; tetO-PIK3CAH1047R 9 months Surviving mice develop mammary MMTV-Cre; LSL-PIK3CAH1047R 10 tumors within 7 months Develop mammary tumors within 16 MMTV-Cre; Pik3cae20H1047R/+ 11 months Develop mammary tumors within 36 WAP-Cre; LSL-PIK3CAH1047R 10 days post-partum Develop mammary tumors within 80 WAP-Cre; LSL-PIK3CAE545K 12 days post-partum Develop mammary tumors within 5 MMTV-Cre; p53flox/+; Rosa26-Pik3caH1047R 13 months Delayed mammary hyperplasia but no MMTV-rtTA; tetO-Cre; ErbB3flox/flox; tetO-PIK3CAH1047R effect on mammary tumor formation 14 driven by PIK3CAH1047R Accelerated mammary tumor formation and increased lung MMTV-rtTA; MMTV-Her2; tetO-PIK3CAH1047R 15 metastasis compared to Her2 or PIK3CAH1047R alone Develop ovarian tumors within 16 Ptenflox/flox; Pik3caLat-H1047R/+ 16 weeks Accelerated development of intestinal Gpa33-CrePR2; APCLOF/LOF; Pik3caLat-H1047R/+ tumors compared to Pik3caH1047R or 17 APCLOF alone Increased number and size of Fabp1-Cre; ApcMin/+; Rosa26-Pik3ca* intestinal tumors compared to Pik3ca* 18 or ApcMin/+ alone Increased number and size of Fabp1-Cre; Apcflox/+; Rosa26-Pik3ca* intestinal tumors compared to Pik3ca* 18 or Apcflox/+ alone

200

Appendix B: Supplemental table of genetically engineered mouse models of PI3K isoforms in cancer, with complete references (continued).

Genotype Phenotype Ref PIK3CB (p110) Reduced number of mammary tumors MMTV-Her2/neuT; Pik3cbKD/KD 19 driven by Her2/neuT Protection from high-grade PIN driven Pb-Cre; Ptenflox/flox; Pik3cbflox/flox 4 by Pten loss Reduced PIN and prostate cancer Pten−/+; Pik3cbKD/+ 7 driven by Pten loss Develop VP PIN by 10 weeks and (ARR)2PB-Pik3cbCA 20 DLP PIN by 60 weeks Accelerated mammary tumor MMTV-Neu-IRES-Cre; Pik3cbflox/flox formation and increased tumor burden 3 driven by Her2/neu PIK3CA (p110) and PIK3CB (p110) Loss of ¾ alleles of Pik3ca and Pik3cb 21, K14-Cre; Ptenflox/flox; Pik3caflox/flox; Pik3cbflox/flox blocks skin lesions and mammary 22 hyperplasia driven by Pten loss PIK3CD (p110) Reduced trafficking of NK cells; Pik3cdKD reduced NK cell extravasation to 23 tumor cells Mx1-Cre, LSL-Shp2GOF/+;Pik3cdKD/KD Reduced MPN induced by Shp2 GOF 6 No effect on development of T-ALL Lck-Cre; Ptenflox/flox; Pik3cd−/− 24 driven by Pten loss PIK3CG (p110) No effect on development of T-ALL Lck-Cre; Ptenflox/flox; Pik3cg−/− 24 driven by Pten loss PIK3CD (p110) and PIK3CG (p110) Delayed development of T-ALL driven Lck-Cre; Ptenflox/flox; Pik3cd−/−; Pik3cg−/− 24 by Pten loss PIK3R1 (p85, p55, p50) Reduced B-cell leukemia development CD19-Cre; Pik3r1flox/flox driven by ex vivo infection with BCR- 25 ABL Albumin-Cre; Pik3r1flox/flox Develop liver tumors within 20 months 26 Increased intestinal polyps but no Pten−/+; Pik3r1−/+ 27 change in PIN driven by Pten loss PIK3R2 (p85) Decreased number of colon tumors Pik3r2−/− 28 induced by AOM/DSS

−/+ −/− No change in intestinal polyps or PIN Pten ; Pik3r2 27 driven by Pten loss No effect on B-cell leukemia CD19-Cre; Pik3r2−/− development driven by ex vivo 25 infection with BCR-ABL

201

Appendix B: Supplemental table of genetically engineered mouse models of PI3K isoforms in cancer, with complete references (continued).

Genotype Phenotype Ref PIK3R1 (p85, p55, p50) and PIK3R2 (p85) CCSP-rtTA; tetO-KRasG12D; Pik3r1flox/flox; Pik3r2−/− Decreased incidence of lung tumors 8 LSL-KRasG12D; Pik3r1flox/flox; Pik3r2−/− driven by KRas Blocked B-cell leukemia development CD19-Cre; Pik3r1flox/flox; Pik3r2−/− driven by ex vivo infection with BCR- 25 ABL CCSP-rtTA; tetO-KRasG12D; Pik3r1flox/+; Pik3r2−/− Increased incidence of lung tumors 8 LSL-KRasG12D; Pik3r1flox/+; Pik3r2−/− driven by KRas PIK3C2A (PI3K-Cβ) Decreased microvessel density and Cdh5(PAC)-CreERT2; Pik3c2aflox/flox 29 tumor burden of implanted tumors

RBD, Ras binding domain mutant; KD, kinase dead mutant; CA, constitutively active; Tg, transgene; PIN, prostate intraepithelial neoplasia; AOM/DSS, azoxymethane/dextran sodium sulfate; LOF, loss of function; GOF, gain of function; VP, ventral prostate; DLP, dorsal/lateral prostate; MPN, myoproliferative neoplasia.

202

Appendix B references:

1. Gupta, S. et al. Binding of ras to phosphoinositide 3-kinase p110alpha is required for ras-driven tumorigenesis in mice. Cell 129, 957-68 (2007).

2. Castellano, E. et al. Requirement for interaction of PI3-kinase p110alpha with RAS in lung tumor maintenance. Cancer Cell 24, 617-30 (2013).

3. Utermark, T. et al. The p110alpha and p110beta isoforms of PI3K play divergent roles in mammary gland development and tumorigenesis. Genes Dev 26, 1573-86 (2012).

4. Jia, S. et al. Essential roles of PI(3)K-p110beta in cell growth, metabolism and tumorigenesis. Nature 454, 776-9 (2008).

5. Gritsman, K. et al. Hematopoiesis and RAS-driven myeloid leukemia differentially require PI3K isoform p110alpha. J Clin Invest 124, 1794-809 (2014).

6. Goodwin, C.B. et al. PI3K p110delta uniquely promotes gain-of-function Shp2-induced GM-CSF hypersensitivity in a model of JMML. Blood 123, 2838-42 (2014).

7. Berenjeno, I.M. et al. Both p110alpha and p110beta isoforms of PI3K can modulate the impact of loss-of-function of the PTEN tumour suppressor. Biochem J 442, 151-9 (2012).

8. Engelman, J.A. et al. Effective use of PI3K and MEK inhibitors to treat mutant Kras G12D and PIK3CA H1047R murine lung cancers. Nat Med 14, 1351-6 (2008).

9. Liu, P. et al. Oncogenic PIK3CA-driven mammary tumors frequently recur via PI3K pathway-dependent and PI3K pathway-independent mechanisms. Nat Med 17, 1116-20 (2011).

10. Meyer, D.S. et al. Luminal expression of PIK3CA mutant H1047R in the mammary gland induces heterogeneous tumors. Cancer Res 71, 4344-51 (2011).

11. Yuan, W. et al. Conditional activation of Pik3ca(H1047R) in a knock-in mouse model promotes mammary tumorigenesis and emergence of mutations. Oncogene 32, 318-26 (2013).

12. Meyer, D.S. et al. Expression of PIK3CA mutant E545K in the mammary gland induces heterogeneous tumors but is less potent than mutant H1047R. Oncogenesis 2, e74 (2013).

13. Adams, J.R. et al. Cooperation between Pik3ca and p53 mutations in mouse mammary tumor formation. Cancer Res 71, 2706-17 (2011).

14. Young, C.D. et al. Conditional loss of ErbB3 delays mammary gland hyperplasia induced by mutant PIK3CA without affecting mammary tumor latency, gene expression, or signaling. Cancer Res 73, 4075-85 (2013).

15. Hanker, A.B. et al. Mutant PIK3CA accelerates HER2-driven transgenic mammary tumors and induces resistance to combinations of anti-HER2 therapies. Proc Natl Acad Sci U S A 110, 14372-7 (2013).

16. Kinross, K.M. et al. An activating Pik3ca mutation coupled with Pten loss is sufficient to initiate ovarian tumorigenesis in mice. J Clin Invest 122, 553-7 (2012).

203

17. Hare, L.M. et al. Physiological expression of the PI3K-activating mutation Pik3ca(H1047R) combines with Apc loss to promote development of invasive intestinal adenocarcinomas in mice. Biochem J 458, 251-8 (2014).

18. Deming, D.A. et al. PIK3CA and APC mutations are synergistic in the development of intestinal cancers. Oncogene 33, 2245-54 (2014).

19. Ciraolo, E. et al. Phosphoinositide 3-kinase p110beta activity: key role in metabolism and mammary gland cancer but not development. Sci Signal 1, ra3 (2008).

20. Lee, S.H. et al. A constitutively activated form of the p110beta isoform of PI3-kinase induces prostatic intraepithelial neoplasia in mice. Proc Natl Acad Sci U S A 107, 11002- 7 (2010).

21. Wang, Q. et al. Spatially distinct roles of class Ia PI3K isoforms in the development and maintenance of PTEN hamartoma tumor syndrome. Genes Dev 27, 1568-80 (2013).

22. Wang, Q., Weisberg, E. & Zhao, J.J. The gene dosage of class Ia PI3K dictates the development of PTEN hamartoma tumor syndrome. Cell Cycle 12, 3589-93 (2013).

23. Saudemont, A. et al. p110gamma and p110delta isoforms of phosphoinositide 3-kinase differentially regulate natural killer cell migration in health and disease. Proc Natl Acad Sci U S A 106, 5795-800 (2009).

24. Subramaniam, P.S. et al. Targeting nonclassical oncogenes for therapy in T-ALL. Cancer Cell 21, 459-72 (2012).

25. Kharas, M.G. et al. Ablation of PI3K blocks BCR-ABL leukemogenesis in mice, and a dual PI3K/mTOR inhibitor prevents expansion of human BCR-ABL+ leukemia cells. J Clin Invest 118, 3038-50 (2008).

26. Taniguchi, C.M. et al. The phosphoinositide 3-kinase regulatory subunit p85alpha can exert tumor suppressor properties through negative regulation of growth factor signaling. Cancer Res 70, 5305-15 (2010).

27. Luo, J. et al. Modulation of epithelial neoplasia and lymphoid hyperplasia in PTEN+/- mice by the p85 regulatory subunits of phosphoinositide 3-kinase. Proc Natl Acad Sci U S A 102, 10238-43 (2005).

28. Cortes, I. et al. p85beta phosphoinositide 3-kinase subunit regulates tumor progression. Proc Natl Acad Sci U S A 109, 11318-23 (2012).

29. Yoshioka, K. et al. Endothelial PI3K-C2alpha, a class II PI3K, has an essential role in angiogenesis and vascular barrier function. Nat Med 18, 1560-9 (2012).

204