UC San Diego UC San Diego Electronic Theses and Dissertations

Title Regulation of PHLPP and Akt signaling

Permalink https://escholarship.org/uc/item/1ff6w0r6

Author Warfel, Noel Andrew

Publication Date 2011

Peer reviewed|Thesis/dissertation

eScholarship.org Powered by the California Digital Library University of California UNIVERSITY OF CALIFORNIA, SAN DIEGO

Regulation of PHLPP and Akt signaling

A dissertation submitted in partial satisfaction of the requirements for the degree Doctor of Philosophy

Biomedical Sciences

Noel Andrew Warfel

Committee in charge:

Professor Alexandra C. Newton, Chair Professor Steve F. Dowdy Professor Kun-Liang Guan Professor Tony Hunter Professor Jing Yang

2011

The Dissertation of Noel Andrew Warfel is approved, and it is acceptable in quality and form for publication on microfilm and electronically:

Chair

University of California, San Diego

2011

iii

DEDICATION

Dedicated to Meredith Warfel for her endless dedication, love, and support and being a constant source of inspiration

Dedicated to Stephen and Barbara Warfel for their guidance throughout my life. Without their support and advice this would not have been possible.

TABLE OF CONTENTS

Signature Page……………………………………………………...…………...…….iii

Dedication……………………………………………………………………………..iv

Table of Contents…………………………………...………………………...………..v

List of Figures……………………………………………………………..…..……....vi

List of Tables…………………………………...………………………..…..……...... ix

Acknowledgements…………………………………………………….……..…...….x

Vita…………………………………………………………………………....…...... xii

Abstract of the Dissertation….…………………..…………………………..……....xiv

Chapter 1: An Introduction to Akt and PHLPP………………….….....……..……1

Chapter 2: Identification of a Negative Feedback Loop through which Akt Activity Controls the Stability of PHLPP1.……..…….…...... ….…19

Chapter 3: Mislocalization of the E3 ligase, β-TrCP1, in Glioblastoma Uncouples Negative feedback between PHLPP1 and Akt...... ……..…47

Chapter 4: Disruption of the Interface Between the PH and Kinase Domains of Akt is Sufficient for Hydrophobic Motif Phosphorylation Site in the Absence of mTORC2.…………………………………….....……..…89

Chapter 5: Unraveling a Complex Network of Signal Transduction Pathways...120

Appendix A: Regulation of PHLPP Activity and Evidence of PHLPP Phosphorylation……………………………….……..………….…..141

LIST OF FIGURES

Chapter 1

Figure 1.1: Akt activation by mitogen stimulation.………………..………...…....17

Figure 1.2: The PHLPP family of phosphatases.……………………………..…...18

Chapter 2

Figure 2.1: PHLPP1α and PHLPP1β are degraded more rapidly than PHLPP2.....40

Figure 2.2: Chronic inhibition of Akt decreases PHLPP1 protein levels.…...... 41

Figure 2.3: PHLPP1β degradation is accelerated in cells lacking mTORC2.…….42

Figure 2.4: Akt activity directly correlates with PHLPP1 stability.…………...….43

Figure 2.5: Regulation of PHLPP1 stability by Akt is dependent on the activity of

the proteasome and GSK-3β..…………………………………..…….45

Chapter 3

Figure 3.1: Akt-mediated feedback loop enhancing PHLPP1 levels is preferentially

lost in CNS tumors.………………………..………………………….81

Figure 3.2: Loss of the feedback loop in glioblastoma is independent of PHLPP1,

CK1, and GSK-3.……..……………………...... 83

Figure 3.3: β-TrCP1 is confined to the nucleus in glioblastoma cell lines and

patient samples.…………………………………………………….…84

Figure 3.4: PHLPP1 turnover is slower and mRNA levels are reduced in

glioblastoma cell lines..…………………………………………....….85

Figure 3.5: β-catenin turnover is impaired in glioblastoma cell lines……...……..86

Figure 3.6: Expression of β-TrCP1 in the cytosol rescues Akt-mediated regulation

of PHLPP1 in glioblastoma.…………………………………...….....87

Chapter 4

Figure 4.1: Akt can be phosphorylated at S473 in the absence of mTOR kinase

activity.………………………………………………………………113

Figure 4.2: Activation of PI3K is sufficient to restore Akt (S473) phosphorylation

in cells lacking mTORC2……………………………………………114

Figure 4.3: Disruption of the PH and kinase domain interface of Akt is sufficient

for S473 phosphorylation in the absence of

mTORC2…………………………………………………………….115

Figure 4.4: Disruption of the interaction between the PH and kinase domains of

Akt overcomes the necessity for mTORC2 for hydrophobic motif

phosphorylation.…………………………………..………………....116

Supplemental Figure 4.1: Inhibition of S6K is necessary for rapamycin-induced

phosphorylation of S473 in Sin1 -/- MEFs.…………………………117

Supplemental Figure 4.2: Rescue of S473 phosphorylation in Sin1 -/- MEFs is

dependent on PI3K activity and the conformation of Akt.……...... 118

Supplemental Figure 4.3: Regulation of Akt phosphorylation is specifically mediated

by mTORC2.………………………………………………………...119

Chapter 5

vii

Figure 5.1: The ability of Akt to regulate PHLPP1 stability is lost in glioblastoma

due to the nuclear confinement of β-TrCP1.………………………...139

Figure 5.2: PHLPP expression levels are regulated at multiple levels downstream

of PI3K.…………………………………………………………..….140

Appendix A

Figure A.1: The phosphatase activity of PHLPP1β and PHLPP2, but not PHLPP1α,

is increased in response to EGF stimulation.………………………..153

Figure A.2: Active Ras does not bind to the predicted RA domain of PHLPP1β, but

binds to a separate region.…………………………………………...154

Figure A.3: Identification phosphorylation sites on PHLPP1 and PHLPP2 under

basal and agonist stimulated conditions.…………………………….155

Figure A.3: PHLPP is phosphorylated in a EGF dependent manner at a site that is

recognized by the p-p38 (T180/Y182) antibody.…………………....156

Figure A.4: PHLPP is phosphorylated by its substrates, Akt and PKC.…………157

viii

LIST OF TABLES

Chapter 3

Supplemental Table 3.1: Numerical values corresponding to each of the NCI60 tumor

cell lines analyzed in Figure 3.1..………………..……………….…..89

Supplemental Table 3.2: Experimental identifiers and institutions responsible for

generating microarray data made available by the Developmental

Therapeutics Program NCI/NIH

(http://dtp.cancer.gov/mtargets/mt_index.html)……………………...89

ACKNOWLEDGEMENTS

I would like to acknowledge Professor Alexandra Newton, Ph.D. for her guidance as my research mentor and chair of my thesis committee. Her sense of optimism and passion for science were a constant source of support and motivation.

Also deserving of special recognition for playing a direct role in my training are the members of my thesis committee for their support and insight. I would also like to thank my classmates and friends in San Diego and elsewhere for their support throughout my graduate career. Special thanks to Drs. Maya Kunkel, Matt Niederst, and Charles King for their insight and technical support and the Newton lab, past and present, for making graduate school an enjoyable experience.

Chapter 1 is, in part, a review article to be submitted to the Journal of

Biological Chemistry, 2011. It was co-authored by Alexandra C. Newton.

Figure 5 of Chapter 2 has been previously published as supplemental data in the Journal of Biological Chemistry, 2010.

Chapter 3 is, in part, a reprint of the material as it appears in the Journal of

Biological Chemistry, 2010. It was co-authored by Matt Niederst, Michael W.

Stevens, Paul M. Brennan, Margaret C. Frame, and the research was supported under the guidance of Alexandra C. Newton. I was the primary investigator and author of this research.

Chapter 4 is, in full, a research article accepted for publication in the Journal of

Biological Chemistry, 2011. It was co-authored by Matt Niederst and supported under the guidance of Alexandra C. Newton. I was the primary investigator and author of this research.

VITA

2004 Bachelor of Science, James Madison University

2006 Master of Science, The Johns Hopkins University

2011 Doctor of Philosophy, University of California, San Diego

PUBLICATIONS

1. Warfel NA and Newton AC. PH domain Leucine-rich Repeat Protein Phosphatase (PHLPP): Structure, Function, Regulation, and Role in Disease. Manuscript to be submitted. J Biol Chem.

2. N. A. Warfel, M. Niederst, A. C. Newton, Disruption of the interface between the PH and kinase domains of Akt is sufficient for hydrophobic motif site phosphorylation in the absence of mTORC2. J Biol Chem, (2011).

3. Warfel NA, Niederst M, Stevens MW, Brennan PM, Frame MC, Newton AC. Mislocalization of the E3 Ligase, β-Transducin Repeat-containing Protein 1 (β-TrCP1), in Glioblastoma Uncouples Negative Feedback between the Pleckstrin Homology Domain Leucine-rich Repeat Protein Phosphatase 1 (PHLPP1) and Akt. J Biol Chem 2010;286:19777-88.

4. Brognard J, Niederst M, Reyes G, Warfel N, Newton AC. Common polymorphism in the phosphatase PHLPP2 results in reduced regulation of Akt and protein kinase C. J Biol Chem 2009;284:15215-23.

5. Gills JJ, Lopiccolo J, Tsurutani J, Shoemaker RH, Best CJ, Abu-Asab MS, Borojerdi J, Warfel NA, Gardner ER, Danish M, Hollander MC, Kawabata S, Tsokos M, Figg WD, Steeg PS, Dennis PA. Nelfinavir, A lead HIV protease inhibitor, is a broad-spectrum, anticancer agent that induces endoplasmic reticulum stress, autophagy, and apoptosis in vitro and in vivo. Clin Cancer Res 2007;13:5183-94.

6. Gills JJ, Castillo SS, Zhang C, Petukhov PA, Memmott, RM, Hollingshead M, Warfel N, Han J, Kozikowski AP, Dennis PA. Phosphatidylinositol ether lipid analogues that inhibit AKT also independently activate the stress kinase, p38alpha, through MKK3/6-independent and -dependent mechanisms. J Biol Chem 2007;282:27020-9.

7. Granville CA, Warfel N, Tsurutani J, Hollander MC, Robertson M, Fox S, Veenstra TD, Issaq HJ, Linnoila RI, Dennis PA. Identification of a highly

xii

effective rapamycin schedule that markedly reduces the size, multiplicity, and phenotypic progression of tobacco carcinogen-induced murine lung tumors. Clin Cancer Res 2007;13:2281-9.

8. Warfel NA, Lepper ER, Zhang C, Figg WD, Dennis PA. Importance of the stress kinase p38alpha in mediating the direct cytotoxic effects of the thalidomide analogue, CPS49, in cancer cells and endothelial cells. Clin Cancer Res 2006;12:3502-9.

xiii

ABSTRACT OF THE DISSERTATION

Regulation of PHLPP and Akt Signaling

Noel Andrew Warfel

Doctor of Philosophy in Biomedical Sciences

University of California, San Diego, 2011

Professor Alexandra C. Newton, Chair

Phosphatidylinositol 3-kinase (PI3K) is a key signaling pathway commonly dysregulated in various human diseases. This thesis addresses two questions relating to the regulation of PI3K signaling: 1] is mTORC2 necessary for Akt phosphorylation and activation, and 2] what are the mechanisms regulating PHLPP stability.

The PH domain Leucine-rich repeat Protein Phosphatase, PHLPP, plays a central role in controlling the amplitude of growth factor signaling by directly dephosphorylating and thereby inactivating Akt. Here we show that the cellular levels of PHLPP1 are enhanced by their substrate, activated Akt, providing a negative feedback loop to tightly control cellular Akt output. Furthermore, we found that this feedback loop is lost in high-grade glioblastomas because the E3 ligase, β-TrCP1, is confined to the nuclear compartment, where it can no longer target its cytosolic

xiv

substrates for degradation, providing a novel mechanism for the dysregulation of

PHLPP1 levels in this disease.

Akt is activated by phosphorylation at two crucial sites, the activation loop

(T308) and the hydrophobic motif (S473), both of which are necessary for maximal activation of the kinase. In recent years, mTORC2 has been widely accepted as the kinase responsible for phosphorylation of the hydrophobic motif of Akt. However, our results indicate that disruption of the interface between the PH and kinase domains of Akt, by genetic means or through prolonged activation of PI3K, is sufficient for phosphorylation of S473 in the absence of mTORC2; supporting a model in which the role of mTORC2 is not only to phosphorylate the hydrophobic motif of Akt but also to facilitate S473 phosphorylation by alternative mechanisms.

CHAPTER 1:

An Introduction to PHLPP and Akt signaling

Akt signaling:

The protein kinase B/Akt family of Ser/Thr protein kinases play a central role in signaling downstream of phosphatidylinositol 3-kinase (PI3K) (1). Akt is a member of the AGC kinase family, which is known to mediate a wide array of important cellular functions and whose dysregulation is strongly associated with the pathogenesis of many human diseases. The Akt family is composed of three isoforms, encoded by three genes (Akt1, Akt2, and Akt3) that share a similar domain structure: an amino-terminal pleckstrin homology (PH) domain, followed by an α-helical linker, and a carboxy-terminal kinase domain (2). Once activated, Akt phosphorylates defined substrates throughout the cell, ultimately activating pro-proliferation and anti- apoptotic signaling pathways (1).

The activation state of Akt is tightly controlled, and its dysregulation is implicated in the development of a variety of diseases, most notably cancer (3).

Phosphorylation of Akt, both chronically and acutely, controls its function (Figure 1).

Upon biosynthesis, the nascent Akt polypeptide is phosphorylated at the ribosome by mTORC2 (4-6) at a conserved C-terminal site originally identified in protein kinase C

1 2 and named the turn motif (TM) (7). Phosphorylation of this residue (T450 in Akt1)

(8) is important for the stability of AGC kinases (9, 10). Thus, Akt is heavily ubiquitinylated and degraded in cells where TM phosphorylation is absent (4, 5). This processing phosphorylation of Akt is constitutive and its dephosphorylation has not been reported. Once processed by phosphorylation at the turn motif, Akt localizes to the cytosol, where it is maintained in an inactive conformation through the interaction between its PH and kinase domains (11). In the presence of proliferative signals, phosphatidylinositol 3-kinase (PI3K) is activated and generates the lipid second messenger phosphatidylinositol 3,4,5-trisphosphate (PIP3). Akt is subsequently recruited from the cytosol to the plasma membrane through binding of its PH domain to PIP3, resulting in a conformational change that separates the PH and kinase domains and unmasks two key regulatory residues whose phosphorylations are required for maximal activation of the kinase. The first site, known as the activation loop site

(T308 on Akt1), is phosphorylated by phosphoinositide dependent kinase -1 (PDK-1)

(12). The second site is termed the hydrophobic motif (S473 on Akt1), and several candidates have been identified as potential HM kinases, including Akt itself (13),

DNA-dependent protein kinase (DNA-PK) (14), and integrin-linked kinase (ILK) (15), among others. Recent evidence supports a key role for mTORC2 in modulating the phosphorylation of this site on Akt, as genetic ablation of mTORC2-specific subunits,

Rictor or Sin1, results in the loss of Akt phosphorylation at both the TM and HM and impaired signaling toward certain substrates (16-18). Signaling by Akt is terminated by two primary mechanisms: removal of the activating lipid second messenger by the

3 phosphatase PTEN (phosphatase and tensin homolog on chromosome ten) (19) and direct dephosphorylation of the kinase by phosphatases, including PHLPP (20, 21).

The mammalian target of rapamycin (mTOR) is an evolutionarily conserved

Ser/Thr protein kinase that forms two distinct protein complexes in cells. These complexes are differentially regulated and perform distinct cellular functions (22).

The mTOR complex 1 (mTORC1) is composed of mTOR, Raptor, and mLST8 and is sensitive to inhibition by rapamycin. mTORC1 is a crucial regulator of cell growth in response to nutrients, stress, and growth factors (23, 24). A second complex, mTOR complex 2 (mTORC2) consists of mTOR, Rictor, mLST8, and Sin1 and is generally considered to be rapamycin-insensitive (25, 26). Its key targets are Akt and protein kinase C (5, 27).

While it has been well established that mTORC2 controls the phosphorylation of Akt at the TM and HM, it remains unclear whether this is solely due to direct phosphorylation of these sites. It is also a distinct possibility that mTORC2 functions independently of its catalytic activity to control HM phosphorylation. Thus, a focus of my thesis research is to determine whether mTORC2 is necessary for Akt phosphorylation at S473 and if it can be bypassed in response to cellular stimuli or through genetic manipulation of Akt.

The PHLPP phosphatases:

Structure. Given that many AGC kinase family members, such as PDK-1 and

Akt, contain a PH domain and their activity is regulated by phosphorylation, it was

4 logical to hypothesize that signal transduction by the AGC kinases would be opposed by cellular phosphatases that also contain lipid-binding domains, which could dephosphorylate and inactivate these enzymes. Thus, the Newton lab performed a rational search of the NCBI database for proteins containing both a PH domain and a phosphatase domain that could potentially terminate signaling downstream of PDK-1.

One such family of phosphatases was identified and subsequently named PHLPP (PH domain leucine-rich repeat protein phosphatase; pronounced ‘flip’) after its domain structure (20). The PHLPP phosphatases are novel members of the PP2C family of

Ser/Thr phosphatases, which are classified as PPM (protein phosphatase magnesium/manganese dependent) family of Ser/Thr protein phosphatases that require

Mg2+ or Mn2+ for their catalytic activity. PHLPP is an evolutionarily conserved phosphatase tracing back to a orthologue in yeast termed Cyr1 (28). The PHLPP family comprises three isozymes, the alternatively spliced PHLPP1α and PHLPP1β

(which differ in an amino-terminal extension on PHLPP1β), and a separate gene product, PHLPP2 (Figure 2). All three isozymes share a similar structure: an N- terminal PH domain, leucine-rich repeat region, PP2C phosphatase domain, and a C- terminal PDZ binding motif. In addition, both PHLPP1β and PHLPP2 contain an N- terminal Ras Association (RA) domain that is not present in PHLPP1α (29). PHLPP isoforms are ubiquitously expressed, and their levels vary between tissues and cell lines.

Molecular modeling of the phosphatase domain of PHLPP2, based on the structure of PP2Cα, identified several residues that are likely to be important for its

5 catalytic activity (Figure 1b). By measuring the ability of proven inhibitors to bind the active site, it was determined that the active site of PHLPP2 contains two Mn2+ ions coordinated by the following residues: D806, E989, and D1024 (30). To date, a majority of the research into the function of PHLPP has focused on the role of PHLPP as a phosphatase. However, unlike a majority of protein phosphatases, PHLPP contains multiple regulatory domains, which raises the likely possibility that it can control various effectors through mechanisms independent of its phosphatase activity.

Function. Initial studies demonstrated a critical role for PHLPP in opposing pro-survival signaling pathways. PHLPP isoforms selectively dephosphorylate the hydrophobic motif of Akt and protein kinase C (PKC) isozymes (20, 31). In the case of Akt, dephosphorylation at this site (S473 of Akt1) reduces its intrinsic catalytic activity, leading to increased apoptosis and decreased proliferation (20, 21). In the case of PKC, dephosphorylation at this site (S657 of PKCα) destabilizes PKC and shunts it to degradation pathways. The ability of PHLPP to dephosphorylate PKC is dependent upon the PH domain, as deletion of this domain results in increased PKC levels (31). Alternatively, the ability of PHLPP to properly regulate Akt depends on an intact PDZ binding motif, as deletion of the four amino acids delineating this motif greatly impairs the ability of PHLPP to terminate Akt signaling (20). Furthermore, knockdown of PHLPP1 and PHLPP2 revealed isoform specific interactions with individual Akt isoforms to regulate their activity toward downstream substrates (21).

Specifically, PHLPP1 controls the activity of Akt2 and Akt3 (but not Akt1), while

PHLPP2 regulates Akt1 and Akt3 (but not Akt2). As a result, the PHLPP family

6 constitutes a unique signaling axis to regulate both the amplitude and direction of Akt signaling.

Binding to protein scaffolds is also important for PHLPP-mediated inactivation of Akt. PHLPP is targeted to Akt through interaction with three distinct scaffolding proteins, FK506-binding protein 51 (FKBP51) (32), Scribble (33), and Na+/H+ exchanger regulatory factor 1 (NHERF1) (34), all of which enhance the ability of

PHLPP to dephosphorylate Akt by co-localizing the two enzymes. Additionally, overexpression of the membrane proteins, Scribble or NHERF1, forces the re- localization of PHLPP1, which is primarily found in the cytosol, to the plasma membrane. Using both purified proteins and immunoprecipitation studies, it was determined that that PDZ binding motif of PHLPP1 and PHLPP2 is necessary for their interaction with NHERF1. Specifically, PHLPP1 binds to both PDZ domains of

NHERF1, while PHLPP2 binds primarily to NHERF1-PDZ2 domain. Alternatively, the interaction between PHLPP1 and Scribble is independent of the PDZ binding motif of PHLPP1, but is lost upon deletion of the entire C-terminal region. These scaffolding proteins are commonly lost in tumorigenesis, which impairs the ability of

PHLPP to terminate Akt signaling and promotes resistance to chemotherapy and cell survival. Thus, poising PHLPP near its substrates through scaffolding interactions is crucial for its cellular function.

Another substrate that PHLPP acts on to inhibit is mammalian sterile 20-like kinase 1 (MST-1) (35). The MST-1 signaling pathway is known to be critical in the regulation of cell survival by activating MAPK signaling pathways and is frequently inactivated in tumorogenesis. PHLPP directly interacts with and dephosphorylates an

7 inhibitory site (Thr387) on MST-1, activating the kinase and allowing it to signal to downstream effectors, namely p38 and JNK, to induce apoptosis. Therefore, the antiproliferative and apoptosis-inducing effects of PHLPP are mediated through inactivation of numerous pro-survival signaling pathways.

In addition to its influence on cell survival, PHLPP is involved in a diverse array of physiological processes, including the circadian system, memory formation, and T cell development. PHLPP1β was identified by Shimizu, et al. and originally termed SCOP (SCN circadian oscillatory protein) because its expression was demonstrated to oscillate in a circadian manner, suggestive of a role in controlling the central clock (36). Subsequent studies in the Sassone-Corsi lab determined that in response to light-induced phase shift, PHLPP1 -/- mice display delayed shortening of circadian period length (tau) and impaired capacity to stabilize the circadian rhythm

(37). Additionally, in neuronal cell systems, PHLPP1β (SCOP) binds the nucleotide- free form of K-Ras, through interaction within LRR region of PHLPP1, to prevent

GTP binding to K-Ras and ultimately diminish signaling through the MEK/ERK pathway (38). Consequently, overexpression of PHLPP1 in the hippocampus of mice reduces ERK1/2 signaling, which inhibits CRE-mediated transcription and impairs long-term memory formation (39). These findings implicate a critical role for PHLPP in maintaining neural plasticity and the resetting of the circadian clock in response to changing stimuli.

Through their opposition of Akt, PHLPP isoforms also serve a critical role in the development and function of regulatory T cells (Tregs) (40). Inactivation of Akt is known to be a functional requirement of Tregs, and both PHLPP1 and PHLPP2

8 mRNA are significantly upregulated in Tregs compared to conventional T cells

(Tconv). Knockdown or genetic deletion of PHLPP1 results in the activation of Akt, an effect that dramatically impairs the development and function of Tregs.

Interestingly, stimulation of Tconv with the immunosuppressive cytokine, TGF-β, enhanced Smad-3 binding to the PHLPP1 promoter and induced PHLPP1 expression, which indicates that upregulation of PHLPP1 may be critical aspect of the Treg developmental program.

While a majority of the existing research on PHLPP has focused on its opposition of Akt and other pro-survival signaling pathways, there is growing evidence that PHLPP is a central player in a broad range of cellular processes. Unlike a majority of protein phosphatases, PHLPP contains multiple regulatory domains, which raises the likely possibility that it can control various effectors through mechanisms independent of its phosphatase activity. Further study will surely reveal novel interacting proteins and unique functions for the PHLPP phosphatases.

Regulation. Maintenance of PHLPP activity and expression are crucial to sustain cellular homeostasis in response to various stimuli and preserve the balance between cell survival and apoptosis. While the mechanisms governing PHLPP activity remain largely unknown, recent reports in retinal cells and cardiomyocytes indicate that PHLPP can be acutely activated by Akt (41) and adenylyl cyclase type 6

(42), respectively. Also, a polymorphism affecting the activity of PHLPP2 has been identified through sequencing of tissue samples from breast cancer patients (43). This polymorphism results in a single amino acid change (Leu 1016 Ser) within the

9 phosphatase domain of PHLPP2 that reduces its activity toward Akt and PKC and does not induce apoptosis as efficiently as wild-type PHLPP2.

Recent drug discovery efforts using chemical and virtual screening methods have identified small molecules that specifically inhibit PHLPP at micromolar concentrations (30). These compounds are selective for PHLPP1 and PHLPP2 compared to other PP2C family members, including PP2Cα. Biochemical and cellular assays demonstrate that the two most promising compounds, which are structurally diverse, effectively inhibit PHLPP in vitro, increase Akt signaling, and prevent etoposide-induced apoptosis in cells. Thus, the identification of PHLPP-specific inhibitors provides a powerful tool to enhance the understanding of PHLPP at the molecular level, as well as for potential therapeutic application. Further work is necessary to improve the efficacy of PHLPP inhibitors, identify compounds that activate PHLPP, and resolve the molecular mechanisms controlling PHLPP activity.

Role in Disease. Growing evidence supports a crucial role for PHLPP in the development and progression of various disease states. The most well documented role of PHLPP is that of a tumor suppressor protein in cancer. Also, several reports suggest that PHLPP not only blocks tumorogenesis by inactivation of oncogenic pathways, but also sensitizes cancer cells to chemotherapy (32, 44, 45).

Reductions in PHLPP1 and PHLPP2 mRNA and protein levels are associated with broad spectrum of cancers, including CLL (46), breast carcinomas (47, 48), glioblastoma (49), and melanoma (50) among others. Indicative of its importance as a tumor suppressor, PHLPP is deleted as often as PTEN (approximately 40%) in metastatic prostate cancers (51). A detailed role for PHLPP as a tumor suppressor has

10 been described in colon cancer (52). Immunohistochemical staining of colon tumors revealed a 78 and 86% decrease in PHLPP1 and PHLPP2 expression, respectively compared to normal tissue. Furthermore, stable overexpression of PHLPP in colon cancer cells decreases proliferation and sensitizes cells to growth inhibition induced by

PI3K inhibition. Reduced Akt signaling was largely responsible for the inhibition of cell growth observed upon PHLPP overexpression, as a constitutively active Akt construct negated these effects. Importantly, reconstitution of either PHLPP1 or

PHLPP2 into colon cancer cells strongly inhibits tumor growth in vivo.

As opposed to its role in promoting tumorigenesis, enhancing Akt activity can be advantageous to disease states such as ischemic heart disease and metabolic disorders. For example, a study by Miyamoto et al. showed that PHLPP1 is an important regulator of Akt signaling in cardiomyocytes (53). Knockdown or genetic depletion of PHLPP1 enhances Akt activity in cardiac myocytes and, in turn, provides protection against ischemic injury. Thus, acute inhibition of PHLPP1 after cardiac injury may be of therapeutic benefit. Similarly, Akt is a key modulator of insulin signaling. Impaired activation of Akt isoforms, specifically Akt2, reduces glucose transport and ultimately leads to insulin resistance, which is commonly associated with obesity and type 2 diabetes (54, 55). It has been shown that skeletal muscle and adipose tissue from the morbidly obese display higher levels of PHLPP1 compared to non-obese subjects, correlating with lower Akt (S473) phosphorylation (56).

Similarly, PHLPP1 mRNA is enhanced in type 2 diabetic individuals, correlating with decreased hydrophobic motif phosphorylation of Akt2 (57). Overexpression of

PHLPP1 in insulin-responsive cell lines results in a decrease in insulin-induced Akt

(S473) phosphorylation, diminished glycogen synthesis, and reduced glucose transport. Therefore, increased PHLPP1 expression may promote insulin resistance associated with diabetes and obesity. These reports suggest that mechanisms controlling the amount of PHLPP in the cell are often lost in disease, and maintenance of PHLPP levels is important for cellular homeostasis.

How is PHLPP stability regulated? PHLPP is poised in a position to greatly influence the balance between cell survival and apoptosis and, consequently, is a central player in the development and progression of numerous diseases. The ability of PHLPP to terminate signaling by Akt, PKC, and MST-1 depends on the amount of

PHLPP present in cells, and PHLPP expression levels are commonly dysregulated in cancer. The regulation of PHLPP stability and the mechanisms responsible for its degradation remain largely unknown. Therefore, alterations in the rate of PHLPP turnover could account for the markedly altered levels of PHLPP observed in cancer cell lines and, in turn, constitutive activation of Akt and its other substrates.

Elucidating the mechanisms controlling PHLPP stability and how these processes are dysregulated in cancer will be the primary focus of my thesis.

Acknowledgments: Chapter 1 is, in part, a review article to be submitted to the

Journal of Biological Chemistry, 2011. It was co-authored by Alexandra C. Newton.

References

1. L. C. Cantley, The phosphoinositide 3-kinase pathway. Science 296, 1655 (2002).

2. D. R. Alessi, P. Cohen, Mechanism of activation and function of protein kinase B. Curr Opin Genet Dev 8, 55 (1998).

3. J. R. Testa, P. N. Tsichlis, AKT signaling in normal and malignant cells. Oncogene 24, 7391 (2005).

4. W. J. Oh, C. C. Wu, S. J. Kim, V. Facchinetti, L. A. Julien, M. Finlan, P. P. Roux, B. Su, E. Jacinto, mTORC2 can associate with ribosomes to promote cotranslational phosphorylation and stability of nascent Akt polypeptide. EMBO J 29, 3939 (2010).

5. V. Facchinetti, W. Ouyang, H. Wei, N. Soto, A. Lazorchak, C. Gould, C. Lowry, A. C. Newton, Y. Mao, R. Q. Miao, W. C. Sessa, J. Qin, P. Zhang, B. Su, E. Jacinto, The mammalian target of rapamycin complex 2 controls folding and stability of Akt and protein kinase C. EMBO J 27, 1932 (2008).

6. V. Zinzalla, D. Stracka, W. Oppliger, M. N. Hall, Activation of mTORC2 by association with the ribosome. Cell 144, 757 (2011).

7. L. M. Keranen, E. M. Dutil, A. C. Newton, Protein kinase C is regulated in vivo by three functionally distinct phosphorylations. Curr Biol 5, 1394 (1995).

8. A. Bellacosa, T. O. Chan, N. N. Ahmed, K. Datta, S. Malstrom, D. Stokoe, F. McCormick, J. Feng, P. Tsichlis, Akt activation by growth factors is a multiple-step process: the role of the PH domain. Oncogene 17, 313 (1998).

9. A. S. Edwards, M. C. Faux, J. D. Scott, A. C. Newton, Carboxyl-terminal phosphorylation regulates the function and subcellular localization of protein kinase C betaII. J Biol Chem 274, 6461 (1999).

10. F. Bornancin, P. J. Parker, Phosphorylation of threonine 638 critically controls the dephosphorylation and inactivation of protein kinase C α. Curr. Biol. 6, 1114 (1996).

11. V. Calleja, M. Laguerre, P. J. Parker, B. Larijani, Role of a novel PH-kinase domain interface in PKB/Akt regulation: structural mechanism for allosteric inhibition. PLoS Biol 7, e17 (2009).

12. V. Calleja, D. Alcor, M. Laguerre, J. Park, B. Vojnovic, B. A. Hemmings, J. Downward, P. J. Parker, B. Larijani, Intramolecular and intermolecular interactions of protein kinase B define its activation in vivo. PLoS Biol 5, e95 (2007).

13. A. Toker, A. C. Newton, Akt/protein kinase B is regulated by autophosphorylation at the hypothetical PDK-2 site. J Biol Chem 275, 8271 (2000).

14. J. Feng, J. Park, P. Cron, D. Hess, B. A. Hemmings, Identification of a PKB/Akt hydrophobic motif Ser-473 kinase as DNA-dependent protein kinase. J Biol Chem 279, 41189 (2004).

15. M. Delcommenne, C. Tan, V. Gray, L. Rue, J. Woodgett, S. Dedhar, Phosphoinositide-3-OH kinase-dependent regulation of glycogen synthase kinase 3 and protein kinase B/AKT by the integrin-linked kinase. Proc Natl Acad Sci U S A 95, 11211 (1998).

16. E. Jacinto, V. Facchinetti, D. Liu, N. Soto, S. Wei, S. Y. Jung, Q. Huang, J. Qin, B. Su, SIN1/MIP1 maintains rictor-mTOR complex integrity and regulates Akt phosphorylation and substrate specificity. Cell 127, 125 (2006).

17. D. D. Sarbassov, D. A. Guertin, S. M. Ali, D. M. Sabatini, Phosphorylation and regulation of Akt/PKB by the rictor-mTOR complex. Science 307, 1098 (2005).

18. D. A. Guertin, D. M. Stevens, C. C. Thoreen, A. A. Burds, N. Y. Kalaany, J. Moffat, M. Brown, K. J. Fitzgerald, D. M. Sabatini, Ablation in mice of the mTORC components raptor, rictor, or mLST8 reveals that mTORC2 is required for signaling to Akt-FOXO and PKCalpha, but not S6K1. Dev Cell 11, 859 (2006).

19. T. Maehama, J. E. Dixon, The tumor suppressor, PTEN/MMAC1, dephosphorylates the lipid second messenger, phosphatidylinositol 3,4,5- trisphosphate. J Biol Chem 273, 13375 (1998).

20. T. Gao, F. Furnari, A. C. Newton, PHLPP: a phosphatase that directly dephosphorylates Akt, promotes apoptosis, and suppresses tumor growth. Mol Cell 18, 13 (2005).

21. J. Brognard, E. Sierecki, T. Gao, A. C. Newton, PHLPP and a second isoform, PHLPP2, differentially attenuate the amplitude of Akt signaling by regulating distinct Akt isoforms. Mol Cell 25, 917 (2007).

22. D. D. Sarbassov, S. M. Ali, D. M. Sabatini, Growing roles for the mTOR pathway. Curr Opin Cell Biol 17, 596 (2005).

23. S. Sengupta, T. R. Peterson, D. M. Sabatini, Regulation of the mTOR complex 1 pathway by nutrients, growth factors, and stress. Mol Cell 40, 310 (2010).

24. D. A. Guertin, D. M. Sabatini, An expanding role for mTOR in cancer. Trends Mol Med 11, 353 (2005).

25. D. D. Sarbassov, S. M. Ali, D. H. Kim, D. A. Guertin, R. R. Latek, H. Erdjument-Bromage, P. Tempst, D. M. Sabatini, Rictor, a novel binding partner of

14 mTOR, defines a rapamycin-insensitive and raptor-independent pathway that regulates the cytoskeleton. Curr Biol 14, 1296 (2004).

26. Q. Yang, K. Inoki, T. Ikenoue, K. L. Guan, Identification of Sin1 as an essential TORC2 component required for complex formation and kinase activity. Genes Dev 20, 2820 (2006).

27. T. Ikenoue, K. Inoki, Q. Yang, X. Zhou, K. L. Guan, Essential function of TORC2 in PKC and Akt turn motif phosphorylation, maturation and signalling. EMBO J 27, 1919 (2008).

28. K. Matsumoto, I. Uno, Y. Oshima, T. Ishikawa, Isolation and characterization of yeast mutants deficient in adenylate cyclase and cAMP-dependent protein kinase. Proc Natl Acad Sci U S A 79, 2355 (1982).

29. J. Brognard, A. C. Newton, PHLiPPing the switch on Akt and protein kinase C signaling. Trends Endocrinol Metab 19, 223 (2008).

30. E. Sierecki, W. Sinko, J. A. McCammon, A. C. Newton, Discovery of small molecule inhibitors of the PH domain leucine-rich repeat protein phosphatase (PHLPP) by chemical and virtual screening. J Med Chem 53, 6899 (2010).

31. T. Gao, J. Brognard, A. C. Newton, The phosphatase PHLPP controls the cellular levels of protein kinase C. J Biol Chem 283, 6300 (2008).

32. H. Pei, L. Li, B. L. Fridley, G. D. Jenkins, K. R. Kalari, W. Lingle, G. Petersen, Z. Lou, L. Wang, FKBP51 affects cancer cell response to chemotherapy by negatively regulating Akt. Cancer Cell 16, 259 (2009).

33. X. Li, H. Yang, J. Liu, M. D. Schmidt, T. Gao, Scribble-mediated membrant targeting of PHLPP1 is required for its negative regulation of Akt. EMBO Reports, (2011).

34. J. R. Molina, N. K. Agarwal, F. C. Morales, Y. Hayashi, K. D. Aldape, G. Cote, M. M. Georgescu, PTEN, NHERF1 and PHLPP form a tumor suppressor network that is disabled in glioblastoma. Oncogene, (2011).

35. M. Qiao, Y. Wang, X. Xu, J. Lu, Y. Dong, W. Tao, J. Stein, G. S. Stein, J. D. Iglehart, Q. Shi, A. B. Pardee, Mst1 is an interacting protein that mediates PHLPPs' induced apoptosis. Mol Cell 38, 512 (2008).

36. K. Shimizu, M. Okada, A. Takano, K. Nagai, SCOP, a novel gene product expressed in a circadian manner in rat suprachiasmatic nucleus. FEBS Lett 458, 363 (1999).

37. S. Masubuchi, T. Gao, A. O'Neill, K. Eckel-Mahan, A. C. Newton, P. Sassone- Corsi, Protein phosphatase PHLPP1 controls the light-induced resetting of the

15 circadian clock. Proc Natl Acad Sci U S A 107, 1642 (2010).

38. K. Shimizu, M. Okada, K. Nagai, Y. Fukada, Suprachiasmatic nucleus circadian oscillatory protein, a novel binding partner of K-Ras in the membrane rafts, negatively regulates MAPK pathway. J Biol Chem 278, 14920 (2003).

39. K. Shimizu, T. Phan, I. M. Mansuy, D. R. Storm, Proteolytic degradation of SCOP in the hippocampus contributes to activation of MAP kinase and memory. Cell 128, 1219 (2007).

40. S. J. Patterson, J. M. Han, R. Garcia, K. Assi, T. Gao, A. O'Neill, A. C. Newton, M. K. Levings, Cutting edge: PHLPP regulates the development, function, and molecular signaling pathways of regulatory T cells. J Immunol 186, 5533 (2011).

41. Y. Kanan, H. Matsumoto, H. Song, M. Sokolov, R. E. Anderson, R. V. Rajala, Serine/threonine kinase akt activation regulates the activity of retinal serine/threonine phosphatases, PHLPP and PHLPPL. J Neurochem 113, 477 (2010).

42. M. H. Gao, A. Miyanohara, J. R. Feramisco, T. Tang, Activation of PH- domain leucine-rich protein phosphatase 2 (PHLPP2) by agonist stimulation in cardiac myocytes expressing adenylyl cyclase type 6. Biochem Biophys Res Commun 384, 193 (2009).

43. J. Brognard, M. Niederst, G. Reyes, N. Warfel, A. C. Newton, Common polymorphism in the phosphatase PHLPP2 results in reduced regulation of Akt and protein kinase C. J Biol Chem 284, 15215 (2009).

44. B. S. Carver, C. Chapinski, J. Wongvipat, H. Hieronymus, Y. Chen, S. Chandarlapaty, V. K. Arora, C. Le, J. Koutcher, H. Scher, P. T. Scardino, N. Rosen, C. L. Sawyers, Reciprocal feedback regulation of PI3K and androgen receptor signaling in PTEN-deficient prostate cancer. Cancer Cell 19, 575 (2011).

45. D. J. Mulholland, L. M. Tran, Y. Li, H. Cai, A. Morim, S. Wang, S. Plaisier, I. P. Garraway, J. Huang, T. G. Graeber, H. Wu, Cell Autonomous Role of PTEN in Regulating Castration-Resistant Prostate Cancer Growth. Cancer Cell 19, 792 (2011).

46. K. Basso, A. A. Margolin, G. Stolovitzky, U. Klein, R. Dalla-Favera, A. Califano, Reverse engineering of regulatory networks in human B cells. Nat Genet 37, 382 (2005).

47. A. E. Karnoub, A. B. Dash, A. P. Vo, A. Sullivan, M. W. Brooks, G. W. Bell, A. L. Richardson, K. Polyak, R. Tubo, R. A. Weinberg, Mesenchymal stem cells within tumour stroma promote breast cancer metastasis. Nature 449, 557 (2007).

48. A. L. Richardson, Z. C. Wang, A. De Nicolo, X. Lu, M. Brown, A. Miron, X. Liao, J. D. Iglehart, D. M. Livingston, S. Ganesan, X chromosomal abnormalities in basal-like human breast cancer. Cancer Cell 9, 121 (2006).

49. M. Bredel, C. Bredel, D. Juric, G. R. Harsh, H. Vogel, L. D. Recht, B. I. Sikic, High-resolution genome-wide mapping of genetic alterations in human glial brain tumors. Cancer Res 65, 4088 (2005).

50. D. Talantov, A. Mazumder, J. X. Yu, T. Briggs, Y. Jiang, J. Backus, D. Atkins, Y. Wang, Novel genes associated with malignant melanoma but not benign melanocytic lesions. Clin Cancer Res 11, 7234 (2005).

51. B. S. Taylor, N. Schultz, H. Hieronymus, A. Gopalan, Y. Xiao, B. S. Carver, V. K. Arora, P. Kaushik, E. Cerami, B. Reva, Y. Antipin, N. Mitsiades, T. Landers, I. Dolgalev, J. E. Major, M. Wilson, N. D. Socci, A. E. Lash, A. Heguy, J. A. Eastham, H. I. Scher, V. E. Reuter, P. T. Scardino, C. Sander, C. L. Sawyers, W. L. Gerald, Integrative genomic profiling of human prostate cancer. Cancer Cell 18, 11 (2010).

52. J. Liu, H. L. Weiss, P. Rychahou, L. N. Jackson, B. M. Evers, T. Gao, Loss of PHLPP expression in colon cancer: role in proliferation and tumorigenesis. Oncogene 28, 994 (2009).

53. S. Miyamoto, N. H. Purcell, J. M. Smith, T. Gao, R. Whittaker, K. Huang, R. Castillo, C. C. Glembotski, M. A. Sussman, A. C. Newton, J. H. Brown, PHLPP-1 negatively regulates Akt activity and survival in the heart. Circ Res 107, 476 (2011).

54. D. A. Cross, D. R. Alessi, P. Cohen, M. Andjelkovich, B. A. Hemmings, Inhibition of glycogen synthase kinase-3 by insulin mediated by protein kinase B. Nature 378, 785 (1995).

55. A. D. Kohn, S. A. Summers, M. J. Birnbaum, R. A. Roth, Expression of a constitutively active Akt Ser/Thr kinase in 3T3-L1 adipocytes stimulates glucose uptake and glucose transporter 4 translocation. J Biol Chem 271, 31372 (1996).

56. F. Andreozzi, C. Procopio, A. Greco, G. C. Mannino, C. Miele, G. A. Raciti, C. Iadicicco, F. Beguinot, A. E. Pontiroli, M. L. Hribal, F. Folli, G. Sesti, Increased levels of the Akt-specific phosphatase PH domain leucine-rich repeat protein phosphatase (PHLPP)-1 in obese participants are associated with insulin resistance. Diabetologia 54, 1879 (2011).

57. D. Cozzone, S. Frojdo, E. Disse, C. Debard, M. Laville, L. Pirola, H. Vidal, Isoform-specific defects of insulin stimulation of Akt/protein kinase B (PKB) in skeletal muscle cells from type 2 diabetic patients. Diabetologia 51, 512 (2008).

Figure 1.1: Akt activation by mitogen stimulation. Upon biosynthesis, Akt is constitutively phosphorylated at T450 (orange phosphate) by mTORC2. Akt remains in an inactive confirmation through the interaction of its PH domain and kinase domains. Upon mitogen stimulation, PI3K is activated and generates PIP3. Akt then translocates to the plasma membrane where its PH domain binds to PIP3, allowing for subsequent phosphorylation of T308 (pink phosphate) by PDK-1 and S473 (green phosphate). Once fully phosphorylated, Akt is active and signals to promote cell survival and proliferation. Akt can be inactivated by removal of PIP3, mediated by the lipid phosphatase PTEN or by direct dephosphorylation, mediated by protein phosphatases, such as PHLPP.

Figure 1.2: The PHLPP family of phosphatases. The domain structures of the PHLPP isoforms consist of a Ras association domain (RA), pleckstrin homology domain (PH), Leucine-rich repeat region (LRR), PP2C domain, and PDZ-binding motif.

CHAPTER 2:

Identification of a Negative Feedback Loop Through which Akt Controls the

Stability of PHLPP1

Abstract

The PH domain Leucine-rich repeat Protein Phosphatase, PHLPP, plays a central role in controlling the amplitude of growth factor signaling by directly dephosphorylating and thereby inactivating Akt. Here we show that the cellular levels of the two splice variants, PHLPP1α and PHLPP1β, but not PHLPP2, are enhanced by their substrate, activated Akt, providing a negative feedback loop to tightly control cellular Akt output. Specifically, we show that the steady-state level of PHLPP1β, but not PHLPP2, is markedly reduced in cells in which Akt has been pharmacologically inhibited or in mTORC2-deficient cells, which lack fully-functional Akt. In these cells, introduction of a constitutively active Akt, but not catalytically-inactive Akt, stabilizes PHLPP1. Furthermore, pharmacological inhibition of Akt causes an increase in the ubiquitination of PHLPP1 and its degradation is blocked in response to inhibitors of the proteasome and GSK-3. Thus, Akt activity stabilizes its negative regulator, PHLPP1, creating a negative feedback loop to ultimately dampen Akt signaling.

19 20

Introduction

Tumor initiation and development is often the result of aberrant activation of cellular signaling pathways that drive cell growth, proliferation, and survival.

Constitutive activation of these pathways is a common mechanism promoting tumorigenesis (1). A key signal transduction pathway that is often augmented in cancer is the phosphatidylinositol 3-kinase (PI3K)/Akt pathway. In the presence of cellular stimuli, PI3K is activated by receptor protein tyrosine kinases (RTKs) and G- protein coupled receptors (GPCRs) (2). Upon activation, PI3K phosphorylates the phosphoinositide PI(4,5)P2 to generate the lipid second messenger PI(3,4,5)P3. Akt then translocates to the membrane and binds PI(3,4,5)P3 via its pleckstrin homology

(PH) domain, exposing two crucial phosphorylation sites. The first, known as the activation loop site (T308 on Akt1), is phosphorylated by PDK-1 (3). The second, termed the hydrophobic motif site (S473 on Akt1), is phosphorylated through a mechanism that depends on the mTORC2 protein complex (4, 5). Once activated, Akt phosphorylates defined substrates in the cytosol and nucleus, ultimately inducing proliferation and anti-apoptotic signaling pathways (2). Signaling by Akt can be terminated by two mechanisms: removal of the activating lipid second messenger by the phosphatase PTEN (phosphatase and tensin homolog on chromosome ten) (6) and direct dephosphorylation of the kinase by protein phosphatases, including PHLPP (7,

8). Failure to terminate Akt signaling results in increased cell growth, proliferation, and inhibition of apoptosis.

The PHLPP phosphatases are a novel family of Ser/Thr phosphatases composed of three isozymes: the alternatively spliced PHLPP1α and PHLPP1β, and a separate gene product, PHLPP2 (9). Previous studies have established that PHLPP selectively dephosphorylates the hydrophobic motif sites on Akt (S473) and protein kinase C (PKC) isozymes (7, 10). In the case of Akt, dephosphorylation at this site reduces the intrinsic catalytic activity of Akt, corresponding with increased apoptosis and decreased proliferation (7, 8). In the case of PKC, dephosphorylation destabilizes

PKC and shunts it to degradation pathways (10). There are three isoforms of Akt

(Akt1, Akt2, and Akt3) and PHLPP1 and PHLPP2 form distinct signaling complexes with each of the Akt isoforms to differentially regulate specific downstream substrates

(8). Thus, by regulating the phosphorylation state of Akt isoforms, PHLPP acutely controls both the amplitude and direction of agonist-evoked Akt signaling.

There is growing evidence that PHLPP may serve as a putative tumor suppressor protein in cancer. First, overexpression of PHLPP in normal or cancer cells decreases proliferation and induces apoptosis in concert with inactivation of Akt signaling. Moreover, overexpression of PHLPP1α in a glioblastoma cell line was shown to greatly reduce tumor growth in a nude mouse model (7). Second, analysis of samples from breast cancer patients reveal that PHLPP is lost or decreased in a high percentage of tumors, with loss of PHLPP being more prevalent in late stage disease

(11, 12). Third, PHLPP expression is lost in a high percentage of colon tumor tissues

(13). It is also noteworthy that PHLPP mRNA is dramatically reduced in chronic lymphocytic leukemia (14). These findings suggest that regulatory inputs controlling

22 the amount of PHLPP in the cell are frequently lost in tumorigenesis, resulting in amplified Akt signaling.

Constitutive activation of Akt is important to the initiation and progression of many cancer types. Elevated Akt kinase activity aids in survival when cells are exposed to various apoptotic stimuli, and heightened Akt activation in tumors correlates with poor patient prognosis (15). Previous studies have shown that

PI3K/Akt signaling can be activated by a number of mechanisms, including gene amplification or gain of function mutations in upstream receptor tyrosine kinases and hormone receptors (the most common mechanism in CNS tumors), activating mutations in PI3K, or loss of function mutations in the regulatory phosphatase PTEN

(16-19). However, these mechanisms alone cannot account for all instances where

Akt is found to be constitutively active in tumors, suggesting that alterations to the activity or levels of another protein is responsible for regulating Akt activity in these cases.

Here we identify a negative feedback loop in which activated Akt enhances its negative regulator, PHLPP1β (but not PHLPP2). Specifically, analyses of steady-state levels, protein turn-over, and protein ubiquitinylation reveal that PHLPP1β is stabilized when the signaling output of Akt is high and destabilized when the signaling output of Akt is low. Thus, Akt-mediated feedback on PHLPP1 suggests a novel mechanism to negatively regulate Akt activation.

Results

PHLPP1α and PHLPP1β are degraded significantly faster than PHLPP2.

The PHLPP family of phosphatases comprises three isozymes: PHLPP1α, PHLPP1β, and PHLPP2. As a first step in understanding how the cellular levels of PHLPP isoforms are regulated, we measured the turn-over rate of each isozyme by pulse-chase analysis. COS7 cells transiently overexpressing HA-tagged PHLPP1α, PHLPP1β, or

PHLPP2 were pulsed for one hour with [35S] Met/Cys to metabolically label a pool of newly-synthesized protein and then chased with unlabeled medium for up to 48 hours.

Each PHLPP isozyme was immunoprecipitated, analyzed by SDS-PAGE, and the 35S- labeled phosphatase was detected by autoradiography of transfer membranes

(Figure1a, upper panel). The same membrane was then probed with an anti-HA antibody to assess the amount of total PHLPP in each immunoprecipitate. To determine protein half-times, the amount of radiolabeled PHLPP divided by total

PHLPP was plotted as a function of time and the data were fitted to a first-order decay.

The graph in Figure 1b reveals that the pools of labeled PHLPP1α and PHLPP1β decayed with similar kinetics (half-times were 7.0 ± 0.5 h and 7.7 ± 0.3 h, respectively), whereas PHLPP2 was approximately 4 times longer-lived, decaying ! with a half-time of 27.8 ± 0.7 h. The decreased! stability of PHLPP1α/b compared to PHLPP2 suggests different regulatory mechanisms for controlling the stability of ! PHLPP1 versus PHLPP2.

Inhibition of PI3K/Akt decreases PHLPP1β, but not PHLPP2, protein levels.

To investigate which signaling pathways regulate the stability of PHLPP, we treated

Hs578t cells with inhibitors against various growth factor-regulated protein kinases and asked whether the levels of PHLPP1β or PHLPP2 were altered. Hs578t breast cancer cells were chosen for this experiment because they have relatively high levels

24 of both PHLPP1β and PHLPP2, allowing detection of changes in endogenous PHLPP protein levels. Hs578t cells were treated for 24 hours with DMSO or inhibitors to

PI3K (LY294002), Akt (Akti VIII), PKC (Gö6983), EGFR kinase (AG1478), JNK

(SP00125), or MEK (U0126) (Figure 2a). Quantitative analysis of Western blots revealed that treatment with the PI3K or Akt inhibitor resulted in a marked reduction

(5- and 3-fold, respectively) in the steady-state level of PHLPP1β, with no significant effect on PHLPP2. No other inhibitors affected PHLPP levels significantly. To verify that efficacy of each inhibitor, the phosphorylation of downstream substrates of each kinase target was examined. The Western blot in Figure 2a shows that treatment with

1] either the PI3K or Akt inhibitor effectively abrogated phosphorylation of Akt

(S473) (lanes 2 and 3), 2] the JNK inhibitor decreased phosphorylation of c-Jun (lane

6), and 3] the MEK inhibitor decreased phosphorylation of ERK (lane 7). Separate experiments confirmed that, under conditions similar to those in this experiment, the

PKC inhibitor prevented phorbol ester-stimulated activation of PKC and the EGFR inhibitor prevented EGF-mediated activation of the EGFR kinase (data not shown).

Only treatment with LY294002 and Akti VIII, two mechanistically distinct inhibitors of the Akt signaling pathway, led to a robust reduction in PHLPP1β protein levels.

The finding that PHLPP1β, but not PHLPP2, stability is controlled by inhibitors of the

PI3K/Akt pathway is further evidence that PHLPP1β and PHLPP2 are differentially regulated.

We next asked whether direct inhibition of Akt versus indirect inhibition (via

PI3K, the enzyme that catalyzes the formation of its activator) had similar effects on the half-time of PHLPP1β. Hs578t cells were treated with DMSO, LY294002, or Akti

VIII for up to 24 hours, and PHLPP1β levels were examined by Western blot (Figure

2b, right). Quantitative analysis (Figure 2b, right) revealed that the steady-state levels of PHLPP1β were relatively constant over a 24 hour incubation period with DMSO

(circles). In contrast, treatment of cells with the PI3K (diamonds) or Akt inhibitor

(squares) caused a rapid reduction in the levels of PHLPP1β, with half-times of degradation of 5.2 ± 0.3 h and 7.1 ± 0.4 h, respectively. It is of note that LY294002, which blocked phosphorylation of Akt (S473) more efficiently than Akti VIII, was slightly more effective at reducing PHLPP1β levels; this suggests that the extent of

Akt inhibition directly correlates with PHLPP1β protein levels. The finding that inhibitors of either PI3K or Akt cause steady-state levels of PHLPP1β to decrease with similar kinetics suggests that their effect on PHLPP1β stability is likely occurring through an analogous mechanism.

PHLPP1β is less stable in cells lacking TORC2. To further address the role of

Akt in regulating the turnover of PHLPP, we examined PHLPP1β stability in cells with or without fully-functional Akt. Specifically, we examined the turn-over of

PHLPP1β in wild-type MEFs or MEFs lacking Sin1, an essential component of the

TORC2 complex (4, 5, 20). TORC2 is essential for phosphorylation of the turn motif and hydrophobic motif of both PKC and Akt. Thus, in cells lacking Sin1, Akt is not phosphorylated at either of these two key sites (T450 and S473) and is inactive towards most substrates (20, 21). Sin1 -/- MEFs or Sin1 +/+ MEFs were treated with cycloheximide (CHX), an inhibitor of global protein synthesis, and lysates were collected over a 48 hour time course (Figure 3). Quantitative analysis (Figure 3, right)

of Western blots (Figure 3, left) revealed that PHLPP1β was degraded at a

significantly faster rate in the Sin1 -/- cells, which display impaired Akt activity (t1/2 =

5.1 ± 0.4 h), compared to Sin1 +/+ MEFs (t1/2 =12.1 ± 0.4 h). Consistent with

previous results from pulse-chase experiments, PHLPP2 protein was more stable than

! PHLPP1 protein. Importantly, its degradation! was the same in Sin1 +/+ and Sin1 -/-

MEFs. These data indicate that the stability of PHLPP1β, but not PHLPP2, is reduced

in cells lacking mTORC2, and in turn fully-active Akt.

Akt activity directly correlates with PHLPP1β stability. To determine whether

the decreased stability of PHLPP1β in Sin1 -/- MEFs resulted specifically from

defects in Akt activity, we manipulated the activation state of Akt in both Sin1 +/+ and

Sin1 -/- MEFs and examined its effect on PHLPP1β degradation. To this end, we

examined the effect of inhibiting PI3K on the rate of PHLPP1β degradation following

cycloheximide treatment of Sin1 +/+ or Sin1 -/- MEFs. The Western blot in Figure 4a

shows that treatment of Sin1 +/+ MEFs with LY294002 increased the degradation rate

of PHLPP1β from a half-time of 12.2 ± 0.2 h to 4.2 ± 0.1 h (Figure 4d). In contrast,

LY294002 treatment had no significant effect on the half-life of PHLPP1β in the Sin1 ! ! -/- MEFs (Figure 4d). Furthermore, the half-time of degradation in the Sin1 -/- cells

was comparable to the accelerated degradation observed in the Sin1 +/+ MEFS treated

with the PI3K inhibitor (5.0 ± 0.2 h versus 4.2 ± 0.1 h). LY294002 treatment

abolished Akt activity in both cell lines, as determined by lack of phosphorylation of ! ! Akt on T308. These data suggest that the accelerated degradation of PHLPP1β in the

mTORC2-deficient cells results from impaired Akt activity.

To further test whether the catalytic activity of Akt is required to stabilize

PHLPP1β, we introduced either a constitutively-active or a kinase-inactive construct of Akt into cells. The constitutively-active construct contains a phospho-mimetic at the hydrophobic motif phosphorylation site (S473D) and the kinase-inactive construct has a point mutation introduced in the ATP binding pocket (K179M), which perturbs

ATP binding (Figure 4b and 4c). Sin1 +/+ or Sin1 -/- MEFs were transfected with

HA-tagged constructs of Akt for 30-36 hours prior to treatment with cycloheximide.

Both cell lines expressed similar levels of exogenous Akt as detected with an anti-HA antibody, with a transfection efficiency of 30 - 60% (data not shown). Introduction of the S473D construct of Akt into the Sin1 +/+ MEFs moderately enhanced the stability of PHLPP1β; notably, approximately 40% of the PHLPP1β became resistant to degradation (Figure 4d, open diamonds). In contrast, introduction of the kinase- inactive construct of Akt did not significantly affect the stability of PHLPP1β (Figure

4d, filled circles). Importantly, reconstitution of active Akt (S473D) into the Sin1 -/-

MEFs caused a marked increase in the stability of PHLPP1β (Figure 4d); the half-time for degradation increased from 5.0 ± 0.2 hrs to 9.6 ± 0.2 h in cells transfected with constitutively-active Akt as compared to vector (Figure 4d). The increased stability ! ! observed in the Sin1 -/- MEFS reconstituted with constitutively-active Akt was comparable to that of the Sin1 +/+ MEFs. As observed in the Sin1 +/+ MEFs, introduction of the kinase-inactive construct of Akt into the Sin1 -/- MEFs had no significant effect on the degradation of PHLPP1β (Figure 4d, filled circles). The finding that introduction of constitutively-active Akt, but not kinase-inactive Akt, into

Sin1 -/- cells rescues the accelerated degradation of PHLPP1β is consistent with the catalytic activity of Akt controlling the stability of PHLPP1β.

Regulation of PHLPP1β stability by Akt is dependent on the activity of the proteasome and GSK-3β. To further explore the mechanism(s) responsible for

PHLPP1β degradation, we examined the effect of inhibiting several common protein degradation pathways on the turnover of PHLPP1β and PHLPP2. H157 cells were chosen for this experiment because they have a high basal level of PHLPP1β and

PHLPP2. In a manner similar to the previous experiments, cells were treated with cycloheximide in combination with either vehicle (DMSO) or inhibitors of the lysosome (pepstatin A), proteasome (MG-132), or calpains and cathepsins (ALLN) for up to 48 hours and PHLPP levels were analyzed (Figure 5a). Quantitative analysis of

Western blots revealed that the half-time for PHLPP1β degradation was not affected by inhibition of calpains/cathepsins (Figure 5a, open diamonds) or the lysosome (filled circles). In contrast, inhibition of the proteasome resulted in more than a 6-fold increase in the half-time of PHLPP1β degradation. The proteasome inhibitor, but not

ALLN or pepstatin, also increased the stability of PHLPP2 compared to vehicle control (Figure 5a, Western blot). These data reveal that PHLPP1β and PHLPP2 are primarily degraded by the proteasome.

We next explored whether the accelerated degradation of PHLPP1β caused by

Akt inhibition was accompanied by increased ubiquitination. H157 cells, co- transfected with RFP-PHLPP1β and HA-ubiquitin, were incubated with MG-132 for

30 minutes prior to treatment with either DMSO or LY294002 for 0, 1, or 2 hours.

PHLPP1β was immunoprecipitated from lysates using an anti-RFP antibody and the amount of ubiquitin linked to PHLPP1β was detected by immunoblotting with an anti-

HA antibody. The Western blot in Figure 5b reveals a robust increase in HA-ubiquitin bound to PHLPP1β in cells treated with the PI3K inhibitor compared to vehicle treated cells (lanes 5 and 6). Quantitation of the signal in the region of the immunoblot above

PHLPP1β (indicated by a bracket), normalized to the total amount of PHLPP1β in the immunoprecipitates, revealed a 4- and 6-fold higher increase in ubiquitination following 1 and 2 hours of treatment with LY294002, respectively. Accordingly, the ubiquitination of PHLPP1β increases significantly when PI3K/Akt signaling is blocked.

Gao and coworkers recently reported that the degradation of PHLPP1α is enhanced following direct phosphorylation by GSK-3β, a kinase that is inactivated by

Akt-catalyzed phosphorylation (22). Thus, we asked whether the enhanced degradation of PHLPP1β observed in response to Akt inhibition is prevented by treatment of cells with a GSK-3 inhibitor (GSK-3i IX). H157 cells were treated for 24 hours with LY294002 and either DMSO, MG-132, or GSK-3i IX. The Western blot in Figure 5c shows that, as described above, treatment with LY294002 caused a significant reduction in the steady-state levels of PHLPP1β, and to a lesser extent

PHLPP2 levels (compare lanes 1 and 2). Note the magnitude of the effect in H157 cells is slightly less than was observed in Hs578t cells (compare Figures 2a and 5b), but was still readily detectable. Co-treatment with either the proteasome inhibitor

(lane 3) or the GSK-3 inhibitor (lane 5) prevented the LY294002-induced reduction in

30 steady-state levels of PHLPP1β. The modest reduction in PHLPP2 levels following inhibition of PI3K also required proteasome activity, but, in contrast to PHLPP1β, was independent of GSK-3β activity. Treatment with the GSK-3 inhibitor alone had no significant effect on the levels of either PHLPP isozyme. These data reveal that, similar to the recently reported regulation of PHLPP1α, the regulation of PHLPP1β

(but not PHLPP2) by Akt is mediated by GSK-3β and requires proteasome activity.

Discussion

The loss or dysregulation of a tumor suppressor often leads to the activation of cell survival pathways that contribute to tumorigenesis. One mechanism for controlling the levels of tumor suppressors is a classic negative feedback loop between the tumor suppressor and its substrate. Here we describe a feedback loop where the substrate, phosphorylated Akt, sets the level of enzyme, PHLPP1: this feedforward stimulation of its negative regulator, PHLPP1, serves to inhibit Akt activity and thus maintain homeostasis.

The basal phosphorylation state of Akt is chronically controlled by the level of

PHLPP in the cell: overexpression of PHLPP reduces basal phosphorylation and genetic depletion increases basal phosphorylation specifically on S473 (7). Similarly, the amplitude of the agonist-evoked activity of Akt is suppressed by PHLPP; indeed, under agonist-stimulated conditions, PHLPP controls the phosphorylation of both

S473 and T308 (8). Because PHLPP is constitutively active, we reasoned that dysregulation of mechanisms controlling the cellular levels of the phosphatase might contribute to the elevated Akt activity that is observed in many cancers. To this end,

31 we searched for inputs that accelerated the turnover of PHLPP, thus reducing the steady-state levels of the protein. This led us to the finding that the level of phosphorylated Akt in the cell controls the levels of PHLPP1α and PHLPP1β, but not

PHLPP2. Our data are supported by the recent report by a Gao and co-workers showing that Akt activity increases the levels of PHLPP1α. Specifically, they showed that PHLPP1α is phosphorylated by GSK-3β, an Akt substrate, to initiate recognition by the E3 ligase β-TrCP, ubiquitination, and subsequent degradation (22). Our work is consistent with this model for negative feedback regulation because inhibition of

GSK-3 effectively prevented the destabilization of PHLPP1β in cells treated with the

PI3K inhibitor LY294002. Taken together, these data establish the presence of an

Akt-mediated feedback loop responsible for regulating the stability of both PHLPP1α and PHLPP1β, but not PHLPP2.

Negative feedback loops play central roles in maintaining cellular homeostasis.

Adjusting PHLPP levels in response to the amount of its key substrate, phosphorylated

Akt, is a highly effective mechanism to ensure that Akt signaling is buffered. Under basal conditions, Akt is relatively inactive, and steady-state levels of PHLPP1 are relatively low because its degradation is accelerated via GSK-3β. However, in cells where Akt becomes highly active, steady-state levels of PHLPP1 will increase because its degradation is slowed, ultimately increasing the amount of PHLPP1 available to dephosphorylate Akt and terminate signaling. It is noteworthy that this feedforward stimulation of the protein phosphatase PHLPP1 by its substrate, phosphorylated Akt, is analogous to the feedforward stimulation between the lipid phosphatase PTEN and its substrate, 3’-phosphoinositides. Specifically, Mustelin and coworkers showed that

32 the levels of 3’-phosphoinositides set the levels of PTEN by slowing the degradation of PTEN, likely through a phosphorylation-dependent mechanism (23). Thus, the two major negative regulators of the PI3K pathway, the lipid phosphatase that removes the activating lipids, and the protein phosphatase that removes the activating phosphorylations, are both under feedforward control by their substrates; resulting in feedback inhibition of the PI3K/Akt pathway. This tight control by substrates likely reflects how precise regulation of Akt activity is critical in order to maintain homeostasis.

The PI3K pathway, which contains archetypal negative feedback loops, is one of the most tightly-regulated pathways in cell signaling. Perhaps best characterized is the negative feedback loop from activated Akt to p70S6 kinase to insulin receptor substrate (IRS), such that high Akt activity inhibits the formation of 3’- phosphoinositides via downregulating IRS1 and IRS2 (24). Second, PTEN is arguably the most highly-regulated phosphatase characterized to date (25, 26). Most noteworthy is the negative feedback loop from 3’-phosphoinositides to PTEN levels.

Thus, it is perhaps not surprising that there is a negative feedback loop between activated Akt and the other major phosphatase that suppresses Akt signaling, PHLPP1.

It remains to be seen whether loss of this feedback loop promotes the constitutive activation of Akt and tumor progression.

Experimental Procedures

Plasmids. The cloning of HA-PHLPP1α and HA-PHLPP2 into mammalian expression vectors has been previously described (7, 8). The additional 1536 base

33 pairs of PHLPP1β were amplified by one-step RT-PCR (Qiagen) from RNA isolated from human brain and subcloned into the HA-PHLPP1α vector to generate the full length PHLPP1β gene product. RFP-PHLPP1β was constructed by PCR amplification and subsequent cloning into a pcDNA3 vector with monomeric RFP as an N-terminal tag. HA-Akt S473D and HA-Akt K179M were generated by QuikChange Site-

Directed Mutagenesis (Stratagene).

Materials and antibodies. Cycloheximide (CHX), LY294002, Akt Inhibitor

VIII (Akti VIII), Gö6983, AG1478, SP600125, U0126, GSK-3 inhibitor IX (GSK-3i

IX), ALLN, Pepstatin A, and MG-132 were purchased from Calbiochem and dissolved in dimethyl sulfoxide (DMSO). Antibodies to PHLPP1 and PHLPP2 were purchased from Bethyl Laboratory. The following antibodies were purchased from

Cell Signaling: phospho-antibodies for T308 (P308) and S473 (P473) of Akt, phospho-p44/42 MAPK (Thr202/Tyr204), phospho-c-Jun (Ser73), phospho-GSK-3

α/β (Ser21/9), and total Akt Antibody. An anti-HA monoclonal antibody was purchased from Covance. A DsRed (anti-RFP) antibody was purchased from

Clontech. Easy Tag [35S]Met/Cys (1175 Ci mmol–1) was purchased from PerkinElmer

Life Sciences. Met/Cys-deficient DMEM was purchased from Invitrogen. Ultra-Link protein A/G beads were obtained from Thermo Scientific. Electrophoresis reagents were obtained from Bio-Rad. All other materials and chemicals were reagent-grade.

Cell transfection and immunoblotting. All cell lines were maintained in

DMEM (Cellgro) containing 10% FBS (Hyclone) and 1% penicillin/streptomycin at

37°C in 5% CO2. Transient transfection of Sin1 +/+ MEFs, Sin1 -/- MEFs, and COS7 cells was carried out using FuGENE transfection reagent (Roche Applied Science).

Transient transfection of H157 cells was carried out using Effectene reagents

(Qiagen). Transfection efficiency for Sin1 +/+ and Sin1 -/- MEFs (determined by

FACS sorting of GFP positive cells 30-36 h after transfection) averaged between 30% and 60% respectively for each experiment. For immunoblotting, cells were lysed in buffer A, (50 mM Na2HPO4, 1 mM sodium pyrophosphate, 20 mM NaF, 2 mM

EDTA, 2 mM EGTA, 1% Triton X-100, 1 mM DTT, 200 µM benzamidine, 40 µg ml−1 leupeptin, and 1 mM PMSF) sonicated for 5 s, and protein yield was determined using Coomassie BCA Protein Assay (Pierce). Lysates containing equal protein were analyzed by SDS-PAGE, and individual blots were probed using the indicated antibody. Densitometric analysis was performed with the NIH Image J analysis software (version 1.63).

Immunoprecipitation. COS7 or H157 cells were transiently transfected with the constructs indicated in figure legends. Approximately 36 h post-transfection, cells were lysed in buffer B (50 mM Tris-HCl, pH 7.4, 100 mM NaCl, 5 mM EDTA, 1%

Triton X-100, 10 mM sodium pyrophosphate, 1 mM PMSF, and 1 mM sodium vanadate). Ten percent of the total detergent-solubilized cell lysate was quenched in

SDS sample buffer for further analysis, and the remaining detergent-solubilized cell lysate was cleared by centrifugation and then equal amounts of protein for each sample were incubated with the appropriate antibody and Ultra-link protein A/G- agarose (Pierce) overnight at 4°C. The immunoprecipitates were washed three times in buffer B and proteins were separated by SDS-PAGE and analyzed by immunoblotting.

Pulse-chase analysis. COS7 cells were transfected with DNA encoding HA-

PHLPP1α, HA-PHLPP1β, or HA-PHLPP2. 24–30 h after transfection, cells were incubated with Met/Cys-deficient DMEM for 30 min at 37 °C. The cells were then pulse-labeled with 0.5 mCi ml–1 [35S] Met/Cys in Met/Cys-deficient DMEM for 60 min at 37 °C, medium was removed, and cells were chased with DMEM containing 5 mM unlabeled methionine and cysteine (27). At the indicated times, cells were lysed in Buffer B and centrifuged at 13,000 x g for 5 min at 22 °C, and PHLPP isoforms in the supernatant were immunoprecipitated with an anti-HA monoclonal antibody overnight at 4 °C. The immune complexes were collected with Ultra-Link protein

A/G beads, washed with buffer B, separated by SDS-PAGE, and analyzed by autoradiography. Densitometric analysis of scanned autoradiograms was performed using NIH Image J analysis software and the kinetic analysis was performed using

Kaleidograph software (version 4.0).

Cellular Ubiquitination Assays. H157 cells were transfected with DNA encoding RFP-PHLPP1β (1 µg) and HA-ubiquitin (0.5 µg). Cells were pretreated for

30 minutes with MG-132 (10 µM) prior to the addition of either LY294002 (20 µM) or DMSO for the indicated times before harvest with buffer B containing 10 mM N- ethylmaleimide to preserve ubiquitinated species. After centrifugation at 13,000 x g for 5 min, the supernatants were incubated with DsRed (anti-RFP) antibody overnight at 4 °C and then Ultralink protein A/G beads for 2 h. The immunocomplexes were washed three times with buffer B containing 10 mM N-ethylmaleimide, proteins were separated by SDS-PAGE, and analyzed by immunoblotting.

Acknowledgements: Figure 5 of Chapter 2 has been previously published as

36 supplemental data in the Journal of Biological Chemistry, 2010.

References

1. P. Blume-Jensen, T. Hunter, Oncogenic kinase signalling. Nature 411, 355 (2001).

2. L. C. Cantley, The phosphoinositide 3-kinase pathway. Science 296, 1655 (2002).

3. V. Calleja, D. Alcor, M. Laguerre, J. Park, B. Vojnovic, B. A. Hemmings, J. Downward, P. J. Parker, B. Larijani, Intramolecular and intermolecular interactions of protein kinase B define its activation in vivo. PLoS Biol 5, e95 (2007).

4. D. D. Sarbassov, D. A. Guertin, S. M. Ali, D. M. Sabatini, Phosphorylation and regulation of Akt/PKB by the rictor-mTOR complex. Science 307, 1098 (2005).

5. D. A. Guertin, D. M. Stevens, C. C. Thoreen, A. A. Burds, N. Y. Kalaany, J. Moffat, M. Brown, K. J. Fitzgerald, D. M. Sabatini, Ablation in mice of the mTORC components raptor, rictor, or mLST8 reveals that mTORC2 is required for signaling to Akt-FOXO and PKCalpha, but not S6K1. Dev Cell 11, 859 (2006).

6. T. Maehama, J. E. Dixon, The tumor suppressor, PTEN/MMAC1, dephosphorylates the lipid second messenger, phosphatidylinositol 3,4,5- trisphosphate. J Biol Chem 273, 13375 (1998).

7. T. Gao, F. Furnari, A. C. Newton, PHLPP: a phosphatase that directly dephosphorylates Akt, promotes apoptosis, and suppresses tumor growth. Mol Cell 18, 13 (2005).

8. J. Brognard, E. Sierecki, T. Gao, A. C. Newton, PHLPP and a second isoform, PHLPP2, differentially attenuate the amplitude of Akt signaling by regulating distinct Akt isoforms. Mol Cell 25, 917 (2007).

9. J. Brognard, A. C. Newton, PHLiPPing the switch on Akt and protein kinase C signaling. Trends Endocrinol Metab 19, 223 (2008).

10. T. Gao, J. Brognard, A. C. Newton, The phosphatase PHLPP controls the cellular levels of protein kinase C. J Biol Chem 283, 6300 (2008).

11. M. Qiao, J. D. Iglehart, A. B. Pardee, Metastatic potential of 21T human breast cancer cells depends on Akt/protein kinase B activation. Cancer Res 67, 5293 (2007).

12. C. Desmedt, F. Piette, S. Loi, Y. Wang, F. Lallemand, B. Haibe-Kains, G. Viale, M. Delorenzi, Y. Zhang, M. S. d'Assignies, J. Bergh, R. Lidereau, P. Ellis, A. L. Harris, J. G. Klijn, J. A. Foekens, F. Cardoso, M. J. Piccart, M. Buyse, C. Sotiriou, Strong time dependence of the 76-gene prognostic signature for node-negative breast cancer patients in the TRANSBIG multicenter independent validation series. Clin Cancer Res 13, 3207 (2007).

13. J. Liu, H. L. Weiss, P. Rychahou, L. N. Jackson, B. M. Evers, T. Gao, Loss of PHLPP expression in colon cancer: role in proliferation and tumorigenesis. Oncogene, (2008).

14. P. Ouillette, H. Erba, L. Kujawski, M. Kaminski, K. Shedden, S. N. Malek, Integrated genomic profiling of chronic lymphocytic leukemia identifies subtypes of deletion 13q14. Cancer Res 68, 1012 (2008).

15. A. S. Clark, K. West, S. Streicher, P. A. Dennis, Constitutive and inducible Akt activity promotes resistance to chemotherapy, trastuzumab, or tamoxifen in breast cancer cells. Mol Cancer Ther 1, 707 (2002).

16. P. K. Vogt, S. Kang, M. A. Elsliger, M. Gymnopoulos, Cancer-specific mutations in phosphatidylinositol 3-kinase. Trends Biochem Sci 32, 342 (2007).

17. J. Li, C. Yen, D. Liaw, K. Podsypanina, S. Bose, S. I. Wang, J. Puc, C. Miliaresis, L. Rodgers, R. McCombie, S. H. Bigner, B. C. Giovanella, M. Ittmann, B. Tycko, H. Hibshoosh, M. H. Wigler, R. Parsons, PTEN, a putative protein tyrosine phosphatase gene mutated in human brain, breast, and prostate cancer. Science 275, 1943 (1997).

18. Y. Samuels, L. A. Diaz, Jr., O. Schmidt-Kittler, J. M. Cummins, L. Delong, I. Cheong, C. Rago, D. L. Huso, C. Lengauer, K. W. Kinzler, B. Vogelstein, V. E. Velculescu, Mutant PIK3CA promotes cell growth and invasion of human cancer cells. Cancer Cell 7, 561 (2005).

19. D. J. Slamon, G. M. Clark, S. G. Wong, W. J. Levin, A. Ullrich, W. L. McGuire, Human breast cancer: correlation of relapse and survival with amplification of the HER-2/neu oncogene. Science 235, 177 (1987).

20. E. Jacinto, V. Facchinetti, D. Liu, N. Soto, S. Wei, S. Y. Jung, Q. Huang, J. Qin, B. Su, SIN1/MIP1 maintains rictor-mTOR complex integrity and regulates Akt phosphorylation and substrate specificity. Cell 127, 125 (2006).

21. Q. Yang, K. Inoki, T. Ikenoue, K. L. Guan, Identification of Sin1 as an essential TORC2 component required for complex formation and kinase activity. Genes Dev 20, 2820 (2006).

22. X. Li, J. Liu, T. Gao, {beta}-TrCP-mediated ubiquitination and degradation of PHLPP1 is negatively regulated by Akt. Mol Cell Biol, (2009).

23. D. Birle, N. Bottini, S. Williams, H. Huynh, I. deBelle, E. Adamson, T. Mustelin, Negative feedback regulation of the tumor suppressor PTEN by phosphoinositide-induced serine phosphorylation. J Immunol 169, 286 (2002).

24. L. S. Harrington, G. M. Findlay, A. Gray, T. Tolkacheva, S. Wigfield, H. Rebholz, J. Barnett, N. R. Leslie, S. Cheng, P. R. Shepherd, I. Gout, C. P. Downes, R.

F. Lamb, The TSC1-2 tumor suppressor controls insulin-PI3K signaling via regulation of IRS proteins. J Cell Biol 166, 213 (2004).

25. C. B. Knobbe, A. Merlo, G. Reifenberger, Pten signaling in gliomas. Neuro Oncol 4, 196 (2002).

26. Y. Shi, Serine/threonine phosphatases: mechanism through structure. Cell 139, 468 (2009).

27. E. D. Sonnenburg, T. Gao, A. C. Newton, The phosphoinositide-dependent kinase, PDK-1, phosphorylates conventional protein kinase C isozymes by a mechanism that is independent of phosphoinositide 3-kinase. J Biol Chem 276, 45289 (2001).

Figure 2.1: PHLPP1α and PHLPP1β are degraded more rapidly than PHLPP2. (a) Schematic representation of the domain structure of each of the PHLPP isoforms. The PHLPP phosphatases share a similar domain structure consisting of a PH domain (blue), Leucine rich repeat region (LRR, red), PP2C phosphatase domain (yellow), and a PDZ binding motif (pink). PHLPP1β and PHLPP2 contain an N-terminal Ras association domain (RA, purple). (b) Autoradiogram showing 35S-labeling of HA- PHLPP1α (lanes 1-6), HA-PHLPP1β (lanes 7-12), or HA-PHLPP2 (lanes 13-18) immunoprecipitated from transfected COS7 cells pulsed for 60 min and then chased with unlabeled medium for the indicated times. Total HA-PHLPP in the immunoprecipitates was detected using an anti-HA antibody. (c) Graph showing the relative amount of 35S-labeled PHLPP isoforms, normalized to the total amount of HA-PHLPP in the immunoprecipitate, at each time point in (b). Data represent the average ± SEM of three independent experiments and were fit to an exponential decay, yielding the following half-times: t1/2 = 7.0 ± 0.5 h for PHLPP1α, 7.7 ± 0.3 h for PHLPP1β, and 27.8 ± 0.7 h PHLPP2.

Figure 2.2: Chronic inhibition of Akt decreases PHLPP1β protein levels. (a) Hs578t cells were treated for 24 hours with vehicle (DMSO) or the following kinase inhibitors: LY294002 (20 µM), Akt inhibitor VIII (5 µM), Gö6983 (250 nM), AG1478 (100 nM), SP600125 (20 µM), or U0126 (10 µM). Whole-cell lysates containing approximately 30 µg of protein were analyzed by SDS-PAGE and PHLPP1β and PHLPP2 detected by immunoblotting. The efficiency of LY294002 and Akt inhibitor VIII, SP600125, and U0126 were determined using phospho- specific antibodies against Akt (P473), phospho-c-Jun (p-c-Jun), and phospho-ERK1/2 (pERK). Graph shows average relative PHLPP levels normalized to Akt obtained by densitometric analysis of three individual experiments. Error represents SEM; * P < .05. (b) Hs578t cells were treated with Akt inhibitor VIII (5 µM) or LY294002 (20 µM) for the indicated times and PHLPP1β, the phosphorylation of Akt on S473 (P473), and total Akt were detected by immunoblotting. Data from three independent experiments were quantified and fit to an exponential decay to obtain half-times of degradation; points represent average ± SEM. * P < .05.

Figure 2.3: PHLPP1β degradation is accelerated in cells lacking TORC2. Sin1 +/+ MEFs or Sin1 -/- MEFs were treated with 5 µM cycloheximide (CHX) for up to 48 hours and the steady-state levels of PHLPP1β and PHLPP2 were monitored by immunoblotting. Sin1 -/- MEFs lack fully-active Akt as evidenced by lack of phosphorylation of Akt (S473). The relative amount of PHLPP1β, normalized to actin, was quantified from three independent experiments and the data fitted to an exponential decay to obtain half-times of degradation; points represent average ± SEM. * P < .05.

Figure 2.4: Akt activity directly correlates with PHLPP1β stability. (a) Inhibition of PI3K decreases the stability of PHLPP1β. SIN1 +/+ MEFs and SIN1 -/- MEFs were pretreated with DMSO or LY294002 (20 µM) for 20 minutes prior to the addition of CHX (5 µM). Lysates were collected over a 48 hour time course and immunoblotting was used to monitor the levels of PHLPP1β and PHLPP2, and the phosphorylation of Akt on T308 (P308). Actin was used as a loading control throughout this experiment. (b) Constitutively-active Akt stabilizes PHLPP1β. SIN1 +/+ and SIN1 -/- MEFs were transfected with vector control or a constitutively-active Akt mutant (HA-Akt S473D). 36 hours after transfection, cells were treated with 5 µM CHX and lysates were collected over a 48 hour time course. Expression of constitutively-active Akt was detected with an anti-HA antibody; PHLPP1β and PHLPP2 were detected with isozyme-specific antibodies. (c) Kinase-dead Akt does not affect the rate of PHLPP1β degradation. SIN1 +/+ and SIN1 -/- MEFs were transfected with vector control or a catalytically-inactive Akt mutant (HA-Akt K179M). 36 hours after transfection, cells were treated with 5 µM CHX and lysates were collected over a 48 hour time course. Expression of kinase-dead Akt was detected with an anti-HA antibody; PHLPP1β and PHLPP2 were detected with isozyme-specific antibodies. (d) PHLPP1β levels from three separate experiments as in (a,b,c) were quantified, normalized to actin, and fit to an exponential decay. Data represent the average ± SEM; * P < .05 of experimental sample compared to DMSO.

Figure 2.5: Regulation of PHLPP1β stability by Akt is dependent on the activity of the proteasome and GSK3β. (a) PHLPP1β and PHLPP2 are degraded by the proteasome. H157 cells were treated with cycloheximide (CHX) (5 µM) in combination with DMSO, MG-132 (10 µM), ALLN (10 µM), or pepstatin A (10 µM). Lysates were collected over a 48 hour time period and PHLPP1β and PHLPP2 detected by immunoblotting. PHLPP levels were normalized to actin and data fit to a first-order decay; data represent the average ± SEM of 3 separate experiments.

Figure 2.5: (b) Akt inhibition increases the rate of PHLPP1 ubiquitination. 293T cells were co-transfected with RFP-PHLPP1β and HA-ubiquitin. Cells were pretreated for 30 minutes with MG-132 prior to the addition of LY294002 (20 µM) or DMSO and lysates were collected at the indicated time points. RFP-PHLPP1β was immunoprecipitated with anti-RFP antibody and ubiquitination detected using the anti- HA antibody. The blot was stripped and reprobed using the anti-RFP antibody to determine the total amount of RFP-PHLPP1β immunoprecipitated in each sample. The phosphorylation of Akt on Ser473 in whole cell lysates (10% input) was detected with phospho-specific antibody (P473) and served as measure of efficiency of LY294002 inhibition. Graph shows the fold-increase in PHLPP1β ubiquitination determined from quantifying the HA-Ub labeling in the bracketed region indicated next to the Western blot, normalized to the amount of PHLPP1 in the immunoprecipitate. (c) Degradation of PHLPP1β in response to Akt inhibition depends on the proteasome and GSK-3β activity. 293T cells were treated for 24 hours with LY294002 (20 µM) in combination with DMSO, MG-132 (5 µM), or GSK3i IX (5 µM). The efficacy of LY294002 and GSK3i IX was determined by immunoblotting for Akt (P473) and phospho-GSK3 (Ser21/9), respectively. Data represent the mean ± SEM of three separate experiments. * P < .05 of inhibitor-treated sample compared to DMSO for corresponding PHLPP1 or PHLPP2 amounts.

CHAPTER 3:

Mislocalization of the E3 ligase, β-TrCP1, in Glioblastoma Uncouples Negative

Feedback Between PHLPP1 and Akt

Abstract

The PH domain Leucine-rich repeat Protein Phosphatase, PHLPP, plays a central role in controlling the amplitude of growth factor signaling by directly dephosphorylating and thereby inactivating Akt. The cellular levels of PHLPP1 have recently been shown to be enhanced by its substrate, activated Akt, via modulation of a phosphodegron recognized by the E3 ligase β-TrCP1, thus providing a negative feedback loop to tightly control cellular Akt output. Here we show that this feedback loop is lost in aggressive glioblastoma but not less aggressive astrocytoma.

Overexpression and pharmacological studies reveal that loss of the feedback loop does not result from a defect in PHLPP1 protein or in the upstream kinases that control its phosphodegron. Rather, the defect arises from altered localization of β-TrCP1: in astrocytoma cell lines and in normal brain tissue, the E3 ligase is predominantly cytoplasmic, whereas in glioblastoma cell lines and patient-derived tumor neurospheres, the E3 ligase is confined to the nucleus and thus spatially separated from PHLPP1, which is cytoplasmic. Restoring the localization of β-TrCP1 to the cytosol of glioblastoma cells rescues the ability of Akt to regulate PHLPP1 stability.

Additionally, we show that the degradation of another β-TrCP1 substrate, β-catenin, is

47 48 impaired and accumulates in the cytosol of glioblastoma cell lines. Our findings reveal that the cellular localization of β-TrCP1 is altered in glioblastoma, resulting in dysregulation of PHLPP1 and other substrates such as β-catenin.

Introduction

Glioblastomas (WHO Grade IV) account for approximately 70% of all tumors of the central nervous system (CNS), making them the most common type of malignant brain tumor (1). Primary glioblastoma, commonly referred to as

Glioblastoma Multiforme (GBM), accounts for the vast majority of glioblastoma cases and presents in an acute de novo manner with no prior evidence of lower grade pathology. Secondary glioblastomas are less common and are derived from the progression of lower grade astrocytomas (WHO I-III) (2). Loss or dysregulation of a tumor suppressor can result in the activation of signaling pathways that drive cell growth, proliferation, and survival and aid tumor initiation and development (3).

One signal transduction pathway that is important to the initiation and progression of many cancer types, including those of the CNS, is the phosphatidylinositol 3-kinase (PI3K)/Akt pathway. In the presence of proliferative signals, Akt is activated by phosphorylation at two crucial sites. The first site, known as the activation loop (T308 on Akt1), is phosphorylated by PDK-1 (4). The second site, termed the hydrophobic motif (S473 on Akt1), is phosphorylated through a mechanism regulated by the TORC2 protein complex (5, 6). Once activated, Akt phosphorylates defined substrates in the cytosol and nucleus, ultimately inducing proliferation and anti-apoptotic signaling pathways (7). Signaling by Akt is terminated by two primary mechanisms: removal of the activating lipid second messenger by the phosphatase PTEN (phosphatase and tensin homolog on chromosome ten) (8) and direct dephosphorylation of the kinase by phosphatases, including PHLPP (9, 10).

A second signaling pathway that is often amplified in cancer is the Wnt/β- catenin signaling pathway, which primarily functions to regulate cell proliferation and apoptosis. Under basal conditions, levels of free, cytosolic β-catenin are suppressed by proteasomal degradation. This process is regulated by a protein complex composed of axin, adenomatous polyposis coli (APC), casein kinase 1 (CK1) and glycogen synthase kinase-3 (GSK-3) (11). Accumulation and nuclear translocation of cytosolic

β-catenin activates various oncogenic substrates including c-Myc, cyclin D1, and members of the AP-1 family (12, 13).

Previous studies have shown that both PI3K/Akt and β-catenin signaling can be upregulated in tumorigenesis through a number of mechanisms. In the case of Akt, these include gene amplification or gain of function mutations in upstream receptor tyrosine kinase and hormone receptors (the most common mechanism in CNS tumors), activating mutations in PI3K, or loss of function mutations in the regulatory phosphatase PTEN (14-17). In the case of β-catenin, activation can result from amplification of upstream components of the Wnt pathway such as Dishevelled, as well as mutations to β-catenin itself and regulatory proteins such as APC and axin (18-

20). However, these mechanisms alone do not account for all instances where these signaling pathways are constitutively active in tumors, suggesting that alterations in additional proteins is responsible for activation.

The PHLPP phosphatases are members of a novel family of Ser/Thr phosphatases composed of three isozymes: the alternatively spliced PHLPP1α and

PHLPP1β, and a separate gene product, PHLPP2 (21). Our lab has previously established that PHLPP selectively dephosphorylates the hydrophobic motif of Akt

51 and protein kinase C (PKC) isozymes (9, 22). In the case of Akt, dephosphorylation at this site reduces its intrinsic catalytic activity, leading to increased apoptosis and decreased proliferation (9, 10). In the case of PKC, dephosphorylation destabilizes

PKC and shunts it to degradation pathways (22). There is mounting evidence that

PHLPP serves as a tumor suppressor protein in cancer. First, overexpression of

PHLPP in normal or cancer cells decreases proliferation and induces apoptosis in concert with inactivation of Akt signaling. Furthermore, overexpression of PHLPP1α in a glioblastoma cell line was shown to greatly reduce tumor growth in a nude mouse model (9). Second, PHLPP1 and PHLPP2 are frequently absent or reduced in cancer.

Notably, PHLPP1 mRNA has been shown to be reduced by an order of magnitude in chronic lymphocytic leukemia (CLL) (23) and, in fact, was absent in >50% of CLL tumors in one study (24); PHLPP1 and PHLPP2 mRNA levels were found to be 5-fold and 4-fold lower, respectively, in esophageal adenocarcinomas (25), and several studies have identified significant reduction in PHLPP1 and PHLPP2 mRNA in colon cancer (26-28). PHLPP1 mRNA has also been shown to be 2-fold lower in glioblastoma (29), melanoma (30), and breast carcinomas (31, 32). Additionally,

PHLPP1 protein levels are reduced in cancers of the liver, pancreas, and stomach (33).

Furthermore, PHLPP1 and PHLPP2 protein expression is significantly decreased or lost in 78% and 86%, respectively, of colon tumors (34). Thus, reductions in PHLPP accompany a broad spectrum of cancers, consistent with loss of PHLPP conferring a survival advantage to cancer cells. Indeed, PHLPP is deleted as often as PTEN

(approximately 40%) in metastatic prostate cancers (35). Aberrations in PHLPP may be particularly prevalent in glioblastoma, where several mutations have been found in

52 human tumors and where higher levels of PHLPP correlate with higher survival rates

(36, 37). These findings suggest that maintenance of PHLPP levels is important for inhibition of tumor growth, and mechanisms that control the amount of PHLPP in the cell are often lost in tumorigenesis.

Ubiquitin-mediated degradation is important for the proper regulation of many cellular processes including transcription and cell cycle progression. E3 ubiquitin protein ligases confer the specificity of this system by binding and targeting certain substrates for ubiquitination. The levels of E3 ligases can dramatically alter signal transduction pathways involved in tumorigenesis by influencing the levels of their substrate (38). One such E3 ligase family, Beta-Transducin repeat containing proteins

(β-TrCPs), are F-box proteins that recognize phosphorylated serine residues of target substrates that are processed by the ubiquitin/proteasome pathway (39, 40). β-TrCP1 has been demonstrated to mediate the phosphorylation-dependent degradation of numerous proteins involved in tumorigenesis, the most well studied being β-catenin

(41, 42) and IκB (43). Recently, β-TrCP1 was identified as the E3 ligase responsible for ubiquitin-mediated degradation of PHLPP1 in a phosphorylation-dependent manner (44). Similar to β-catenin, PHLPP1 is directly phosphorylated by CK1 and

GSK-3β on multiple serine residues to create a phosphodegron motif. β-TrCP1 then binds PHLPP1 through recognition of this destruction motif and PHLPP1 is subsequently poly-ubiquitinated and then degraded by the proteasome (44).

Phosphorylation of GSK-3β by Akt inhibits its activity and thus directly influences the stability of PHLPP1; this results in a feedback loop through which active Akt stabilizes its negative regulator, PHLPP1, ultimately serving to dampen Akt signaling.

Here we identify a subset of tumors in which the negative feedback loop by which the level of active Akt determines the expression of its negative regulator

PHLPP1, is lost. Specifically, we have found that this feedback loop is broken in high-grade glioblastomas because of the confinement of β-TrCP1 to the nuclear compartment, where it can no longer target its cytosolic substrates for degradation, providing a novel mechanism for the dysregulation of PHLPP1 and β-catenin levels.

Experimental Procedures

Plasmids: HA-PHLPP1α has been previously described (9). The additional

1536 base pairs of PHLPP1β were amplified by OneStep RT-PCR (Qiagen) from

RNA isolated from human brain and subcloned into the HA-PHLPP1α vector to generate the full-length PHLPP1β gene product (NM_194449) (21). Myc-epitope tagged β-TrCP1 was purchased from Addgene and has been described previously (45).

β-TrCP1 was subcloned in-frame C-terminal to CFP in pcDNA3 vector using

Gateway cloning techniques (Invitrogen).

Materials and Antibodies: Cycloheximide (CHX) and LY294002 were purchased from Calbiochem and dissolved in dimethyl sulfoxide (DMSO).

Antibodies to PHLPP1 and PHLPP2 were purchased from Bethyl Laboratory. The following antibodies were purchased from Cell Signaling: phospho-antibodies for

T308 (P308) and S473 (P473) of Akt, phospho-GSK-3α/β (Ser21/9), phospho- glycogen synthase (Ser641), phospho-β-catenin (Ser45), and total Akt Antibody. An anti-HA monoclonal antibody was purchased from Covance. A β-TrCP1 specific

54 antibody was purchased from Invitrogen. Monoclonal antibodies to actin and Myc were purchased from Sigma. Antibodies to Annexin1, Lamin A, and β-catenin were purchased from Santa Cruz Biotech. The antibody to VDAC was obtained from

Affinity BioReagents. All other materials and chemicals were reagent-grade. Non- viable cell pellets from the NCI-60 panel of tumor cell lines were provided by the

Developmental Therapeutics Program (NCI/NIH).

Cell Transfection and Immunoblotting: SNB-75, SF-268, SF-295, SF-539,

SNB-19, and U251 cell lines were maintained in RPMI (Cellgro) containing 10% FBS

(Hyclone) and 1% penicillin/streptomycin at 37°C in 5% CO2. Transient transfection of all cell lines was carried out using jetPRIME transfection reagent (Polyplus

Transfection) following the manufacturer’s protocol. For immunoblotting, cultured cells and non-viable cell pellets were lysed in buffer A (50 mM Na2HPO4, 1 mM sodium pyrophosphate, 20 mM NaF, 2 mM EDTA, 2 mM EGTA, 1% Triton X-100, 1 mM DTT, 200 µM benzamidine, 40 µg ml−1 leupeptin, and 1 mM PMSF, pH 7.4), sonicated for 5s, and protein yield was determined using Coomassie BCA Protein

Assay (Pierce). Lysates containing equal protein were analyzed by SDS-PAGE, and individual blots were probed using the indicated antibody. Densitometric analysis was performed with AlphaView analysis software (version 1.3.0.6) by Alpha Innotech

Corporation.

Cellular Fractionation: Tissue samples from healthy human brain (case no.

1505 and 1530) were homogenized at low speed and fractionated using the Qproteome

Cell Compartment kit (Qiagen) according to the manufacturer’s protocol.

Astrocytoma, glioblastoma, and patient-derived glioblastoma tumor neurosphere cell

55 lines were fractionated by centrifugation as follows. Approximately 1x106 cells were lysed in 200 µl of hypotonic lysis buffer B (10 mM Tris-HCl, pH 7.5, 50 mM NaCl, 1 mM sodium pyrophosphate, 20 mM NaF, 200 µM benzamidine, 40 µg/ml leupeptin, 1 mM PMSF, and 2 mM sodium vanadate). Lysates were passed through a 25G needle

12 times and a portion of the lysate was retained; the remainder was centrifuged at 800 x g for 5 minute at 4 °C to pellet the nuclear material. The nuclear pellet was then resuspended in 100 µl of buffer C (buffer B + 1% Trition X-100), mixed by vortex and placed on ice for 10 minutes prior to centrifugation at 16,000 x g for 15 minutes. The resulting supernatant was collected as the nuclear fraction. The supernatant taken after isolation of the nuclear pellet was subjected to centrifugation for 20 minutes at

110,000 x g. The resulting supernatant was collected as the cytosolic fraction. The remaining pellet was resuspended in 100µl of buffer C, mixed by vortex and placed on ice for 10 minutes prior to centrifugation at 110,000 x g for 20 minutes. The resulting supernatant was collected as the membrane fraction. The remaining pellet was resuspended in 100µl of buffer C and collected as the detergent-insoluble pellet. An equal volume of each fraction was run on an SDS-PAGE gel and analyzed by immunoblotting.

Generation of Neurospheres: Eligible patients were those undergoing craniotomy for therapeutic management of a brain tumor. Written informed consent was obtained from all patients. The study was approved by the Lothian (Scotland, UK) regional ethics committee (LREC 2004/4/16). Tumor biopsies were processed within

1 hour. Cell lines were expanded from gliomas as described previously (46, 47).

Briefly, tissue was homogenized and trypsinized. The single cell suspension was

56 pelleted, then resuspended in expansion media (Advanced DMEM:F12 (1:1) 1% B27

(10x) 0.5% N2 (100x) 1% Glutamax 100 mM, 1% Penicillin-Streptomycin, 1%

Fungizone, EGF 10 ng/ml (R&D Systems, Abingdon, UK), basic FGF 10 ng/ml (R&D

Systems), Heparin 5 µg/ml (Sigma)), in a matrigel-coated (BD Biosciences, UK) flask. All reagents and products are from Gibco unless otherwise stated. Experiments were performed on low passage number cells (less than passage 10). Patient-derived neurosphere cell lines (named O, U, and P) were designated GBM1, GBM2, and

GBM3, respectively, and two additional neurosphere cell lines (TS600 and TS576) have been described previously (48).

Results

Akt-mediated enhancement of PHLPP1 is preferentially lost in glioblastoma.

The frequent elevation of Akt activity in cancer, often via unexplained mechanisms, led us to explore whether the feedback loop between Akt and PHLPP1 is lost in certain tumors. To this end, we compared the activity of Akt, assessed via the phosphorylation state of T308, with the levels of PHLPP1 in the NCI60 panel of tumor cell lines (Figure 1a). Based on the enhancement of PHLPP1 levels by active

Akt (thus creating a feedback loop to suppress Akt activity) (44), PHLPP1 levels were predicted to track with Akt activity in cell lines containing an intact feedback loop.

However, in cells where this feedback loop is broken, an inverse correlation between

PHLPP1 levels and Akt activity would be expected. The Western blots in Figure 1a

57 present the levels of PHLPP11 and PHLPP2, the phosphorylation state of Akt on T308 and S473, and the relative amount of Akt in lysates from each of the NCI60 cell lines.

The first lane of each gel was loaded with an equal amount of lysate from COS7 cells overexpressing PHLPP1 to serve as an internal control for differences in exposure between immunoblots. To compare the level of PHLPP1 to Akt activity in each cell line, we plotted the ratio of PHLPP1 to T308 phosphorylation (Figure 1b). Note that the phosphorylation of Akt on T308 was chosen to represent Akt activity because phospho-S473 is a direct substrate of PHLPP and could potentially be regulated by the feedback loop described here or be elevated because of a cellular defect in PHLPP itself. However, in most cases, the relative amount of phospho-T308 mirrored that of phospho-S473. We identified five tumor cell lines (indicated by * in Fig 1b) in which the ratio of PHLPP1:Akt (T308) was at least 1 standard deviation below the median value (2 ± 1) of all 60 cell lines. These tumor cells have low levels of PHLPP1 despite very high basal Akt activity, consistent with potential loss of the Akt-mediated feedback loop on PHLPP1. Four of the five identified cell lines are defined as glioblastomas (SF-295 (no. 56), SF-539 (no. 57), SNB-19 (no. 58), and U251 (no. 60); designated cell line numbers are listed in Supplementary Table 1), and the other is an adenocarcinoma derived from a renal tumor (786-0 (no. 16)). Interestingly, the four

CNS cell lines with a low PHLPP1:Akt (T308) ratio (nos. 56, 57, 58, 60) are all classified as glioblastomas, whereas the two CNS lines predicted to have an intact feedback loop (nos. 55, 59) are less aggressive astrocytomas.

1 This is the first reported cloning of the human PHLPP1β isoform; throughout this study PHLPP1 refers to PHLPP1β because it is the predominant isoform expressed in all cell lines examined.

If the feedback loop is lost in the candidate glioblastoma cell lines with elevated Akt activity and low PHLPP1 expression (Figure 1b), we reasoned that Akt activity would no longer affect the stability of PHLPP1 in these cell lines. However, in the less aggressive astrocytoma cell lines in which the ratio of PHLPP1 to Akt

(T308) would indicate that the feedback loop is intact, PHLPP1 levels would be expected to retain sensitivity to pharmacological manipulation of Akt activity. To test this hypothesis, we treated both astrocytoma cell lines and the four candidate glioblastoma cell lines with DMSO or a PI3K inhibitor (LY294002) for 24 hours and then analyzed PHLPP1 levels. Figure 1c reveals that LY294002 effectively decreased the steady-state levels of PHLPP1 by over 50% in both astrocytoma cell lines (e.g. compare lanes 1 and 2) but had no significant effect on the steady-state levels of

PHLPP1 in the glioblastoma cell lines tested (e.g. compare lanes 5 and 6). Thus, we have identified a subset of aggressive CNS tumors that have lost Akt-mediated control on the stability of PHLPP1, a key regulatory mechanism for homeostasis of the Akt signaling output.

Loss of the feedback loop in glioblastoma is independent of PHLPP1, CK1, and GSK-3β. As previously reported (44), the effects of Akt activity on PHLPP1 stability are mediated through GSK-3β-dependent phosphorylation of PHLPP1, resulting in its ubiquitination and proteasome-mediated degradation. Consistent with this mechanism, inhibition of PI3K by LY294002 caused a robust increase in the ubiquitination of PHLPP1 (Supplemental Figure 1), whereas inhibition of GSK-3 or the proteasome prevented the LY294002-induced reduction in PHLPP1 steady-state levels (Supplemental Figure 2).

In order to determine whether a defect in PHLPP1 itself is responsible for the loss of Akt-mediated regulation of its stability, we monitored the stability of an exogenously expressed construct of PHLPP1 in response to PI3K inhibition. HA- tagged PHLPP1 was transfected into an astrocytoma cell line (SF-268) and several of the candidate glioblastoma cell lines (SF-295, SNB-19, U251) for 24 hours prior to treatment with DMSO or LY294002 for 24 hours; anti-HA antibody was used to monitor levels of exogenous PHLPP1 (Figure 2a). The activity of Akt was effectively blocked by treatment with LY294002 in all cell lines, as indicated by the reduced phosphorylation of Akt (S473). Similar to endogenous PHLPP1 in the astrocytoma cell line containing an intact feedback loop, exogenous PHLPP1 levels in these cells were also significantly reduced following Akt inhibition (compare lanes 1 and 2).

However, the level of exogenous PHLPP1 was refractory to Akt inhibition in all three of the glioblastoma lines, which harbor a broken feedback loop (e.g. compare lanes 3 and 4). The inability to rescue the defect by overexpression of exogenous PHLPP1 reveals the loss of Akt-mediated regulation of PHLPP1 in glioblastoma is not caused by a defect in PHLPP1 itself.

We next sequentially investigated the functionality of each player involved in the described phosphorylation-dependent degradation pathway from Akt to PHLPP1 to identify at which step the loop is broken in the indicated glioblastoma cell lines

(Figure 2b). First, we tested whether Akt activity was being effectively blocked by inhibition of PI3K in glioblastomas and astrocytomas. In addition to loss of phosphorylation at S473, phosphorylation of an Akt substrate, GSK-3α/β (S21/9), was reduced in response to Akt inhibition by similar amounts in both astrocytoma

(compare lanes 1 and 2) and glioblastoma cell lines (e.g., compare lanes 3 and 4).

These data demonstrate that inhibition of PI3K is able to decrease Akt activity in both astrocytoma and glioblastoma cells (Figure 2b). Second, we tested whether GSK-3β is able to properly phosphorylate its substrates in cells with a broken feedback loop.

To this end, we examined the phosphorylation state of glycogen synthase, an established target of GSK-3β, after 24 hours of PI3K inhibition. Following relief of the Akt-mediated inhibition of GSK-3β, phosphorylation of glycogen synthase (S641) was increased in both glioblastoma and astrocytoma cell lines, indicating that GSK-3β is functional toward its substrates (Figure 2b). Third, we investigated whether CK1 is able to phosphorylate its substrates in glioblastoma. The stability of a well- documented CK1 substrate, β-catenin, is regulated in a manner analogous to that of

PHLPP1, where a constitutive ‘priming’ phosphorylation by CK1 is necessary for subsequent phosphorylation within the phosphodegron motif. Therefore, to determine whether CK1 is functional in glioblastoma, we probed the CK1 site on β-catenin (S45) with a phospho-specific antibody. We observed a basal level of β-catenin (S45) phosphorylation in all of the cell lines tested, indicating that CK1 is functional in both intact and broken loop cell lines (Figure 2b). Taken together, these data indicate that loss of Akt-mediated regulation of PHLPP1 stability does not result from a defect in

PHLPP1 or the kinases responsible for phosphorylating and promoting its degradation.

β-TrCP1 is confined to the nucleus in glioblastoma cell lines and patient samples. Having established that the upstream kinases responsible for regulation of

PHLPP1 stability are functional, we next asked whether a defect in the E3 ligase and/or proteasomal degradation machinery could be responsible for the broken

61 feedback loop in glioblastoma. One possible mechanism for the loss of Akt-mediated regulation of PHLPP1 in glioblastoma is that the cellular localization of PHLPP1 or its

E3 ligase, β-TrCP1, could be altered in these cells. To examine the localization of these proteins, astrocytoma and glioblastoma cell lines were lysed and the membrane, cytoplasmic, nuclear, and detergent-insoluble fractions were isolated. Equal portions of starting material were then analyzed by immunoblotting to determine the subcellular location of PHLPP1 and β-TrCP1. In order to verify that we effectively isolated the indicated cellular fractions, antibodies specific to proteins that reside in each fraction were used as a control (Figure 3a). Though the fractionation was clean for the most part, we did observe some contamination between our nuclear and membrane fractions, as revealed by the presence of Lamin A, a nuclear protein, in the membrane fraction. However, it is important to note that there was no contamination between the cytosolic and nuclear fractions. Cellular fractionation revealed that

PHLPP1 was located primarily in the cytosol of both astrocytoma and glioblastoma cells. Strikingly, β-TrCP1 was primarily cytosolic in the astrocytoma cell lines

(Figure 3a, lanes 3 and 8), but partitioned almost exclusively in the nuclear fraction of the glioblastoma cells (Figure 3a, lanes 14, 19, and 24). These data identify a key difference in the cellular localization of β-TrCP1 between cell lines containing an intact versus broken feedback loop: PHLPP1 and β-TrCP1 are both located in the cytosol of cells harboring an intact feedback loop, whereas β-TrCP1 is confined to the nucleus, sequestered from PHLPP1, in cells where the feedback loop is broken. Thus, the impaired degradation of PHLPP1 observed in glioblastoma correlates with altered, nuclear localization of β-TrCP1.

To confirm whether our results in cell lines held true in glioblastoma, we obtained tissue samples from humans who died of causes unrelated to CNS tumors

(normal brain: frontal cortex, over 50% astrocytes) and cell lines derived from neurospheres generated from biopsies of human glioblastoma patients (46, 47). These cells are ideal for studying GBM biology because, unlike cultured cell lines, their genetic profiles are very similar to primary gliomas. Tissue from two healthy donors and five patient-derived glioblastoma tumor neurospheres were homogenized, and cytoplasmic and nuclear fractions were isolated as described in the methods section

(Figure 3b). Strikingly, β-TrCP1 was in the nucleus of cells derived from human glioblastoma tumors (n=5; lanes 6, 8, and 10) but in the cytosolic fraction of normal brain from humans who died of other causes (n=2; lanes 1 and 2). To verify the efficacy of our fractionation, Annexin 1 and Lamin A were used as markers of the cytoplasm and nucleus, respectively. These data confirm that the cellular localization of β-TrCP1 is altered in glioblastoma as compared to healthy human brain.

Turnover of PHLPP1 is slower in glioblastoma compared to astrocytoma cell lines. The finding that the localization of β-TrCP1 is altered in glioblastoma cell lines, led us to ask whether the basal rate of PHLPP1 turnover (i.e. independently of Akt inhibition) would be impaired as well in these cells compared to astrocytoma cell lines. SF-268 (intact loop) and U251 (broken loop) cells were treated with cycloheximide (CHX), an inhibitor of global protein synthesis, in combination with

DMSO or LY294002, and lysates were collected over a 24 hour time course (Figure

4a). Quantitative analysis (Figure 4a, bottom) of Western blots (Figure 4a, top) revealed that under basal conditions, PHLPP1 was degraded at a significantly faster

rate in the SF-268 cells (t1/2 = 9.5 0.2 h) compared to the U251 cell line (t1/2 = 19.5

0.3 h). As expected, inhibition of PI3K/Akt significantly increased the rate of

PHLPP1 turnover in SF-268 cells (t1/2 = 3.1 0.2 h), but had virtually no effect on the rate of turnover in the U251 glioblastoma cell line (t1/2 = 19.4 0.3 h). Furthermore, the rate of basal PHLPP1 turnover in U251 cells was significantly slower than that previously reported, whereas the half-time of PHLPP1 in SF-268 cells was similar to that reported for other cell lines and, additionally, was comparable to that determined for overexpressed PHLPP1 by pulse chase analysis (data not shown) (44). These data not only indicate that PHLPP1 stability is insensitive to Akt inhibition in glioblastoma, but that the basal rate of PHLPP1 turnover is also impaired.

Despite the decreased rate of PHLPP1 turnover in glioblastoma compared to astrocytoma cell lines, the absolute levels of PHLPP1 are similar in both types of cell lines (Figure 1a). Steady-state levels are dictated by the rate of biosynthesis and the rate of degradation. Because the rate of degradation is markedly reduced in the glioblastoma, yet steady-state levels are not significantly increased, we reasoned that the rate of biosynthesis may be slower. This led us to examine whether the PHLPP1 mRNA levels differed in glioblastoma versus astrocytoma cell lines. To this end, we analyzed data generated from six separate microarray experiments measured on

Affymetrix U95Av2 and U133 arrays (supplemental Table 2). Raw data from each experiment were normalized to the level of PHLPP1 mRNA present in the SNB-75 astrocytoma cell line, and relative mRNA levels from each experiment were used to calculate the average level of PHLPP1 mRNA in each cell line. Three of the four glioblastoma cell lines (nos. 56, 57, and 60 in Figure 1a) have at least 40% lower

PHLPP1 mRNA levels compared to those in the astrocytoma cell lines (Figure 4b).

SNB-19 glioblastoma cells have only slightly less PHLPP1 mRNA compared to SNB-

75 cells. Consistent with this cell line having higher PHLPP1 mRNA levels, and thus a greater rate of PHLPP1 synthesis, the steady-state levels of PHLPP1 are approximately 40% higher in the SNB-19 cells compared to the other glioblastoma cell lines (no. 58 Figure 1a). Taken together, these data indicate that degradation of

PHLPP1 is slower in glioblastoma, but steady-state levels of PHLPP1 protein are not significantly increased because of lower mRNA levels.

Turnover of β-catenin is impaired in glioblastoma cell lines. Because the basal rate of PHLPP1 turnover is impaired in glioblastoma cell lines, we asked whether other substrates of β-TrCP1 that are known to be degraded primarily in the cytosol are also stabilized. Previous reports reveal that the level and activity of one such substrate, β-catenin, are frequently enhanced in glioblastoma and correlate with increasing tumor grade and poor patient prognosis (49). Thus, we asked whether β- catenin stability was altered in cell lines where β-TrCP1 is confined to the nucleus.

We monitored the rate of degradation of β-catenin in SNB-75 (intact loop) and U251

(broken loop) cell lines by cycloheximide chase over 24 hours. Quantitative analysis

(Figure 5a, bottom) of Western blots (Figure 5a, top) revealed that under basal conditions β-catenin was degraded at a significantly faster rate in the SNB-75 cells compared to U251 cells. These data indicate that the degradation of another β-TrCP1 substrate is also impaired in glioblastoma cell lines.

Next, we sought to determine whether the amount of β-catenin in the cytosolic pool is enhanced in glioblastoma cell lines. Indeed, each of the glioblastoma cell lines

65 examined had higher levels of total β-catenin compared to astrocytoma cells despite having similar mRNA levels (data not shown and Figure 2b, assayed by densitometry). To measure the relative amount of β-catenin present in each cellular compartment, SNB-75 cells and U251 cells were fractionated as described previously and equal portions of starting material were analyzed by Western blotting. Cytosolic levels of β-catenin were approximately 4-fold higher in U251 cells compared to SNB-

75 cells (Figure 5b). Thus, the mislocalization of β-TrCP1 in glioblastoma decreases the rate of degradation and leads to accumulation of β-catenin in the cytosol.

Expression of cytoplasmic β-TrCP1 in glioblastoma rescues Akt-mediated regulation of PHLPP1. The finding that β-TrCP1 is differentially localized in cell lines that have an intact feedback loop compared to those with a broken feedback loop led us to test whether re-introducing β-TrCP1 to the cytoplasm of glioblastoma cells was sufficient to restore the ability of Akt activity to regulate PHLPP1. First, we examined the subcellular location of CFP-tagged β-TrCP1 to determine whether the protein partitions to the cytosol when overexpressed in glioblastoma cells.

Fluorescence microscopy and biochemical fractionation revealed exogenously expressed β-TrCP1 in both the cytosol and nucleus of all cell lines. A representative image of U251 glioblastoma cells transfected with CFP-β-TrCP1 illustrates that this protein was localized throughout the cell, with a majority present in the cytosol

(Figure 6a). It is noteworthy that the level of exogenous β-TrCP1 was approximately three fold higher than the endogenous protein, which may account for the artificial cytosolic localization observed with overexpression in glioblastoma (data not shown).

To test whether cytosolic localization of β-TrCP1 would rescue the ability of Akt activity to regulate PHLPP1 levels, both astrocytoma and glioblastoma cell lines were transfected with vector or Myc-β-TrCP1 for 24 hours prior to treatment with DMSO or LY294002 for an additional 24 hours (Figure 6b). As expected, treatment with

LY294002 significantly decreased PHLPP1 levels in the astrocytoma cell lines (e.g. compare lane 1 to 2). Also, overexpression of β-TrCP1 in cells with an intact feedback loop led to a modest decrease in the level of PHLPP1 (compare lane 1 and lane 3). Moreover, inhibition of PI3K in combination with overexpression of β-TrCP1 reduced PHLPP1 levels below that observed with either β-TrCP1 or LY294002 alone

(compare lanes 2 through 4). Next, focusing on the glioblastoma cell lines, we verified that treatment with LY294002 alone had no effect on the stability of PHLPP1

(e.g. compare lanes 9 and 10). However, overexpression of β-TrCP1 alone in cells with a broken feedback loop resulted in a modest decrease in PHLPP1 levels, comparable to that observed in the astrocytoma cell lines (e.g. compare lanes 9 and

11). Importantly, expression of cytoplasmic β-TrCP1 in combination with PI3K inhibition in the glioblastoma cell lines decreased PHLPP1 protein levels to a similar extent as that observed in astrocytoma cells under the same conditions (e.g. compare lanes 3 and 4 to 11 and 12). Therefore, restoring the cytoplasmic localization of β-

TrCP1 in glioblastoma is sufficient to rescue the defective regulatory feedback loop from Akt to PHLPP1. Taken together these data indicate: 1) β-TrCP1 must be present in the cytosol to effectively degrade PHLPP1 and 2) nuclear confinement of β-TrCP1

67 is responsible for the loss of the feedback loop between Akt and PHLPP1 in glioblastoma.

DISCUSSION

Tightly controlled regulation of protein biosynthesis and degradation in multiple signal transduction pathways plays a central role in maintaining cellular homeostasis. Here we show that β-TrCP1, an E3 ligase crucial to maintaining homeostasis in the PI3K/Akt and Wnt/β-catenin signaling pathways, is confined to the nucleus in glioblastoma and thus spatially segregated from, and unable to properly target, cytosolic substrates for degradation. The altered localization of β-TrCP1 in glioblastoma uncouples the level of a negative regulator, PHLPP, from the level of activated substrate, phospho-Akt. Furthermore, we show that the degradation of β- catenin, a critical mediator of cell survival, is also impaired in glioblastoma cell lines.

The finding that β-TrCP1 is localized differently in glioblastoma versus normal brain tissue is significant because this E3 ligase controls the levels of several crucial mediators of cell survival and may contribute to the pathogenesis of glioblastoma, a disease for which effective treatment options remain elusive.

The feedback loop that sets the level of PHLPP1 to match the level of phospho-Akt was identified recently by Gao and coworkers (44). Specifically, they showed that GSK-3β, which is inhibited by Akt phosphorylation, directly phosphorylates PHLPP1α to initiate ubiquitination by the SCFβ-TrCP1 complex and subsequent degradation of ubiquitinated PHLPP1. Thus, the level of substrate, phosphorylated Akt, sets the level of enzyme, PHLPP1: this feedforward stimulation

68 of its negative regulator, PHLPP1, serves to inhibit its own activity and thus maintain homeostasis. We reasoned that disruption of the Akt-mediated feedback loop on

PHLPP1 is a potential mechanism to promote the constitutive activation of Akt that is common in cancer. To test this possibility, we examined levels of PHLPP1 and phospho-Akt in tumor lysates from the NCI60 panel of tumor cell lines to identify candidate lines in which this feedback loop might be broken (Figure 1a). Consistent with Akt activity controlling PHLPP1 levels, we observed a modest correlation between the relative levels of basal Akt activity (assessed by monitoring the phosphorylation of T308) and PHLPP1 protein levels. Interestingly, this correlation was lost in four out of six CNS tumor cell lines. Furthermore, the loss was associated with tumor grade: the four lines in which PHLPP1:Akt (T308) ratio was unusually low were from Grade IV glioblastomas, whereas the remaining two CNS tumor lines, with an average or above-average PHLPP1:Akt (T308) ratio, were derived from lower-grade astrocytomas (Figure 1b). One possible explanation for the dramatic divergence between Akt activity and PHLPP1 levels is that the feedback loop was lost in the cancer cell lines derived from more aggressive glioblastomas. We tested this by examining whether PHLPP1 levels were sensitive to inhibition of PI3K in all of the

CNS tumor cell lines. Indeed, PHLPP1 levels were unchanged following inhibition of

PI3K/Akt in all four lines derived from glioblastoma but were sensitive to inhibition of PI3K in the astrocytoma cell lines in which phospho-Akt and PHLPP1 levels predicted an intact loop (Figure 1c). These data are consistent with a model in which

Akt-mediated regulation of PHLPP1 stability is lost in aggressive glioblastomas but not in less aggressive astrocytomas. Thus, we have identified a subset of high-grade

69 tumors that have lost the negative feedback loop between PHLPP1 and Akt, implying that this event may be an important factor in the progression of this disease.

Interestingly, we found that the turnover of PHLPP1 is noticeably slower in

U251 cells (broken loop) compared to that in SF-268 (intact loop) in which the half- time of PHLPP1 turnover is similar to that previously reported (Figure 4a) (44). This suggests that there is a defect in basal PHLPP1 degradation that is independent of agonist-evoked signaling. We next investigated each step in the process of PHLPP1 degradation and determined that neither mutation of PHLPP1 nor inactivation of the kinases responsible for preparing it for ubiquitination is observed in glioblastoma

(Figure 2b). Instead, cellular fractionation revealed that β-TrCP1, the E3-ligase responsible for targeting PHLPP1 for degradation, is differentially localized in astrocytomas versus glioblastomas. In astrocytoma cell lines containing an intact feedback loop, both β-TrCP1 and PHLPP1 are primarily located in the cytosol.

However, although PHLPP1 is in the cytosol of glioblastoma cells, β-TrCP1 is confined to the nucleus. These data suggest that under normal circumstances (intact feedback loop) when PHLPP1 is phosphorylated by CK1 and GSK-3β, it is then bound and ubiquitinated by the SCFβ-TrCP1 complex, leading to degradation by the proteasome. However, in glioblastoma, although PHLPP1 is properly phosphorylated by upstream kinases, it is spatially segregated from β-TrCP1, and thus PHLPP1 can no longer interact with β-TrCP1 and is not as readily degraded. Consistent with this, reintroduction of β-TrCP1 to the cytosol of glioblastoma cells is sufficient to restore the ability of Akt to control PHLPP1 levels. Therefore, PHLPP1 levels are not

70 properly regulated in glioblastoma under basal or agonist-stimulated conditions because of the altered, nuclear localization of β-TrCP1.

Curiously, our findings reveal that PHLPP1, although uncoupled from Akt, is actually more stable in glioblastoma compared to astrocytoma. This suggests that loss of the feedback loop does not modulate PHLPP levels in a manner that could account for the constitutive activation of Akt, as was our initial hypothesis. Thus, our work establishes that PHLPP1 levels are insensitive to PI3 kinase inhibitors in glioblastoma, but a compensating mechanism stabilizes the protein. Interestingly, analysis of

PHLPP1 mRNA levels showed that in a majority of the glioblastoma cell lines tested there is significantly less PHLPP1 mRNA compared to the astrocytoma cell lines.

These data explain why, despite the decreased rate of degradation in glioblastoma, steady state levels of PHLPP1 protein are similar to those detected in the astrocytoma cell lines (Figure 4) and ultimately why these cell lines were identified from our screen of the NCI60. To date little is known about how PHLPP is regulated at the transcriptional level. β-TrCP1 is known to affect the transcription of various proteins through modulation of its substrates, most notably the TCF/LEF, ATF/CREB, and NF-

κB transcription factors (50-52). Therefore, it is an intriguing possibility that β-TrCP1 could influence both the biosynthesis (RNA levels) and degradation of PHLPP1.

Importantly, our data reveal that inhibition of the PI3 kinase pathway in glioblastoma, in contrast to astrocytoma, will not reduce PHLPP1 levels.

Previous reports from both immunohistochemical and overexpression studies have revealed primarily cytoplasmic localization of β-TrCP1 (53-55), consistent with the cytosolic localization of the majority of its substrates. Here, we find that in

71 healthy human brain, β-TrCP1 is localized to the cytosol, while nuclear accumulation is associated with high-grade glioblastoma (Figure 3). To our knowledge this is the first report to show that endogenous β-TrCP1 is cytosolic in the human brain and mislocalized to the nucleus in human glioblastoma tumors. To date we have been unable to establish whether endogenous HOS/β-TrCP2, a functionally redundant β-

TrCP isoform, is also confined to the nucleus in glioblastoma due to a lack of suitable reagents. However, given the degradation of multiple shared substrates of β-TrCPs is impaired, it is clear that if β-TrCP2 is cytosolic, it is unable to compensate for the loss of β-TrCP1 in the cytosol. The mechanism responsible for the nuclear confinement of

β-TrCP1 in glioblastoma remains unknown, as there are numerous possible culprits.

β-TrCP1 does not contain a NLS and therefore must be localized to the nucleus through interactions with other proteins. It is possible that increased affinity or abundance of a nuclear protein could sequester β-TrCP1 in the nuclear compartment.

In fact, it has been reported that β-TrCP1 is primarily nuclear in some cell types because of constitutive binding to the nuclear phosphoprotein, hnRNP-U (56). Recent work has also shown that differential splicing of β-TrCP1 can result in dramatic changes in cellular localization and biological roles (57). Our model is consistent with nuclear localization of β-TrCP1 impairing its ability to target its cytosolic substrates for degradation. Thus, a more thorough exploration of the mechanisms governing β-

TrCP localization may be significant to understanding the activation of signal transduction pathways influenced by its substrates.

The finding that β-TrCP1 is confined to the nucleus and cannot properly regulate the cytosolic degradation of substrates is particularly intriguing considering that several of the proteins β-TrCP1 is known to regulate are linked to tumor progression (40). Here we demonstrate that the cytosolic pool of β-catenin is degraded at a slower rate in glioblastoma cell lines where β-TrCP1 is confined to the nucleus (Figure 4). There is precedent for the induction of β-catenin in the progression of tumors, specifically those of the CNS. Immunohistochemical studies comparing β- catenin levels in glioma patient samples to those in normal brain tissue reveal that β- catenin levels are higher in glioma samples, correlating with poor patient prognosis and outcome, yet relative mRNA levels remain unchanged, supporting a defect in protein degradation (49, 58). Another study of glioblastoma patient samples reported intense, homogenous staining of β-catenin in the cytoplasm of 81% (26/32) of samples

(59). Additionally, a study of 45 astrocytic gliomas showed a correlation between increased expression of β-catenin and ascending order of the tumor grade (60). Over

80% of glioblastoma (9/11) displayed high immunoreactivity for β-catenin, yet only two of these stained positively for nuclear accumulation indicating that a majority of the cytoplasmic β-catenin is not being properly degraded. It is also of note that mutations of β-catenin that would render it refractory to degradation are very rare in brain tumors, suggesting another aberration is responsible for accumulation of β- catenin in tumorigenesis (61, 62). Taken together with our findings, altered localization of β-TrCP1 in glioblastoma is potentially responsible for the unusually high levels of β-catenin observed in high-grade glioma patient samples. Therefore, in

73 addition to loss of PHLPP1 regulation, the stabilization of oncogenic factors arising from nuclear localization of β-TrCP1 may be favorably selected for in late stage tumors.

In summary, we have identified a subset of CNS tumors that display a distinct cellular localization of β-TrCP1. Our finding that the cellular localization of β-TrCP1 is altered in glioblastoma contributes to the dysregulation of two signal transduction pathways that are critical in tumorigenesis. Mislocalization of β-TrCP1 in glioblastoma not only impairs the ability of Akt to regulate PHLPP1 stability

(rendering PHLPP levels insensitive to inhibition of the PI3 kinase pathway) but also provides a novel mechanism for the amplification of β-catenin often observed in this disease.

Acknowledgements: We thank the Developmental Therapeutics Program

NCI/NIH for providing us with the NCI60 panel of tumor cell lines, Dr. Eliezer

Masliah (UCSD) and the Alzheimer's Disease Research Center (ADRC) at UCSD

(NIH-AG5131) for the human brain tissue, Dr. Cameron Brennan (Memorial Sloan-

Kettering Cancer Center) for providing TS600 and TS576 neurosphere cell lines, and

Kenny McLeod (University of Edinburgh) and Maria Del-Mar Inda (UCSD) for preparation of glioblastoma neurosphere cell lines. This work was supported by NIH

GM067946 (A.C.N.) and DoD BCRP award BC093021 (N.W.). Regarding grant

BC093021, the U.S. Army Medical Research Acquisition Activity, 820 Chandler

Street, Fort Detrick MD 21702-5014 is the awarding and administering acquisition office. The content of this article does not necessarily reflect the position or the policy

74 of the U.S. Government. Noel Warfel and Matt Niederst were supported in part by the

UCSD Graduate Training Program in Cellular and Molecular Pharmacology through an institutional training grant from the National Institute of General Medical Sciences,

T32 GM007752. Chapter 3 is, in part, a reprint of the material as it appears in the

Journal of Biological Chemistry, 2010. It was co‐authored by Matt Niederst,

Michael W. Stevens, Paul M. Brennan, Margaret C. Frame, and the research was supported under the guidance of Alexandra C. Newton. I was the primary investigator and author of this research.

References

1. H. Ohgaki, P. Kleihues, Genetic pathways to primary and secondary glioblastoma. Am J Pathol 170, 1445 (2007).

2. F. B. Furnari, T. Fenton, R. M. Bachoo, A. Mukasa, J. M. Stommel, A. Stegh, W. C. Hahn, K. L. Ligon, D. N. Louis, C. Brennan, L. Chin, R. A. DePinho, W. K. Cavenee, Malignant astrocytic glioma: genetics, biology, and paths to treatment. Genes Dev 21, 2683 (2007).

3. P. Blume-Jensen, T. Hunter, Oncogenic kinase signalling. Nature 411, 355 (2001).

4. V. Calleja, D. Alcor, M. Laguerre, J. Park, B. Vojnovic, B. A. Hemmings, J. Downward, P. J. Parker, B. Larijani, Intramolecular and intermolecular interactions of protein kinase B define its activation in vivo. PLoS Biol 5, e95 (2007).

5. D. D. Sarbassov, D. A. Guertin, S. M. Ali, D. M. Sabatini, Phosphorylation and regulation of Akt/PKB by the rictor-mTOR complex. Science 307, 1098 (2005).

6. D. A. Guertin, D. M. Stevens, C. C. Thoreen, A. A. Burds, N. Y. Kalaany, J. Moffat, M. Brown, K. J. Fitzgerald, D. M. Sabatini, Ablation in mice of the mTORC components raptor, rictor, or mLST8 reveals that mTORC2 is required for signaling to Akt-FOXO and PKCalpha, but not S6K1. Dev Cell 11, 859 (2006).

7. L. C. Cantley, The phosphoinositide 3-kinase pathway. Science 296, 1655 (2002).

8. T. Maehama, J. E. Dixon, The tumor suppressor, PTEN/MMAC1, dephosphorylates the lipid second messenger, phosphatidylinositol 3,4,5- trisphosphate. J Biol Chem 273, 13375 (1998).

9. T. Gao, F. Furnari, A. C. Newton, PHLPP: a phosphatase that directly dephosphorylates Akt, promotes apoptosis, and suppresses tumor growth. Mol Cell 18, 13 (2005).

10. J. Brognard, E. Sierecki, T. Gao, A. C. Newton, PHLPP and a second isoform, PHLPP2, differentially attenuate the amplitude of Akt signaling by regulating distinct Akt isoforms. Mol Cell 25, 917 (2007).

11. R. H. Giles, J. H. van Es, H. Clevers, Caught up in a Wnt storm: Wnt signaling in cancer. Biochim Biophys Acta 1653, 1 (2003).

12. P. Hasselblatt, L. Gresh, H. Kudo, J. Guinea-Viniegra, E. F. Wagner, The role of the transcription factor AP-1 in colitis-associated and beta-catenin-dependent intestinal tumorigenesis in mice. Oncogene 27, 6102 (2008).

13. R. T. Moon, A. D. Kohn, G. V. De Ferrari, A. Kaykas, WNT and beta-catenin signalling: diseases and therapies. Nat Rev Genet 5, 691 (2004).

14. P. K. Vogt, S. Kang, M. A. Elsliger, M. Gymnopoulos, Cancer-specific mutations in phosphatidylinositol 3-kinase. Trends Biochem Sci 32, 342 (2007).

15. J. Li, C. Yen, D. Liaw, K. Podsypanina, S. Bose, S. I. Wang, J. Puc, C. Miliaresis, L. Rodgers, R. McCombie, S. H. Bigner, B. C. Giovanella, M. Ittmann, B. Tycko, H. Hibshoosh, M. H. Wigler, R. Parsons, PTEN, a putative protein tyrosine phosphatase gene mutated in human brain, breast, and prostate cancer. Science 275, 1943 (1997).

16. Y. Samuels, L. A. Diaz, Jr., O. Schmidt-Kittler, J. M. Cummins, L. Delong, I. Cheong, C. Rago, D. L. Huso, C. Lengauer, K. W. Kinzler, B. Vogelstein, V. E. Velculescu, Mutant PIK3CA promotes cell growth and invasion of human cancer cells. Cancer Cell 7, 561 (2005).

17. D. J. Slamon, G. M. Clark, S. G. Wong, W. J. Levin, A. Ullrich, W. L. McGuire, Human breast cancer: correlation of relapse and survival with amplification of the HER-2/neu oncogene. Science 235, 177 (1987).

18. W. S. Park, R. R. Oh, J. Y. Park, S. H. Lee, M. S. Shin, Y. S. Kim, S. Y. Kim, H. K. Lee, P. J. Kim, S. T. Oh, N. J. Yoo, J. Y. Lee, Frequent somatic mutations of the beta-catenin gene in intestinal-type gastric cancer. Cancer Res 59, 4257 (1999).

19. S. M. Powell, N. Zilz, Y. Beazer-Barclay, T. M. Bryan, S. R. Hamilton, S. N. Thibodeau, B. Vogelstein, K. W. Kinzler, APC mutations occur early during colorectal tumorigenesis. Nature 359, 235 (1992).

20. K. Uematsu, B. He, L. You, Z. Xu, F. McCormick, D. M. Jablons, Activation of the Wnt pathway in non small cell lung cancer: evidence of dishevelled overexpression. Oncogene 22, 7218 (2003).

21. J. Brognard, A. C. Newton, PHLiPPing the switch on Akt and protein kinase C signaling. Trends Endocrinol Metab 19, 223 (2008).

22. T. Gao, J. Brognard, A. C. Newton, The phosphatase PHLPP controls the cellular levels of protein kinase C. J Biol Chem 283, 6300 (2008).

23. K. Basso, A. A. Margolin, G. Stolovitzky, U. Klein, R. Dalla-Favera, A. Califano, Reverse engineering of regulatory networks in human B cells. Nat Genet 37, 382 (2005).

24. P. Ouillette, H. Erba, L. Kujawski, M. Kaminski, K. Shedden, S. N. Malek, Integrated genomic profiling of chronic lymphocytic leukemia identifies subtypes of deletion 13q14. Cancer Res 68, 1012 (2008).

25. Y. Hao, G. Triadafilopoulos, P. Sahbaie, H. S. Young, M. B. Omary, A. W. Lowe, Gene expression profiling reveals stromal genes expressed in common between Barrett's esophagus and adenocarcinoma. Gastroenterology 131, 925 (2006).

26. J. Sabates-Bellver, L. G. Van der Flier, M. de Palo, E. Cattaneo, C. Maake, H. Rehrauer, E. Laczko, M. A. Kurowski, J. M. Bujnicki, M. Menigatti, J. Luz, T. V. Ranalli, V. Gomes, A. Pastorelli, R. Faggiani, M. Anti, J. Jiricny, H. Clevers, G. Marra, Transcriptome profile of human colorectal adenomas. Mol Cancer Res 5, 1263 (2007).

27. S. Kaiser, Y. K. Park, J. L. Franklin, R. B. Halberg, M. Yu, W. J. Jessen, J. Freudenberg, X. Chen, K. Haigis, A. G. Jegga, S. Kong, B. Sakthivel, H. Xu, T. Reichling, M. Azhar, G. P. Boivin, R. B. Roberts, A. C. Bissahoyo, F. Gonzales, G. C. Bloom, S. Eschrich, S. L. Carter, J. E. Aronow, J. Kleimeyer, M. Kleimeyer, V. Ramaswamy, S. H. Settle, B. Boone, S. Levy, J. M. Graff, T. Doetschman, J. Groden, W. F. Dove, D. W. Threadgill, T. J. Yeatman, R. J. Coffey, Jr., B. J. Aronow, Transcriptional recapitulation and subversion of embryonic colon development by mouse colon tumor models and human colon cancer. Genome Biol 8, R131 (2007).

28. C. Gaspar, J. Cardoso, P. Franken, L. Molenaar, H. Morreau, G. Moslein, J. Sampson, J. M. Boer, R. X. de Menezes, R. Fodde, Cross-species comparison of human and mouse intestinal polyps reveals conserved mechanisms in adenomatous polyposis coli (APC)-driven tumorigenesis. Am J Pathol 172, 1363 (2008).

29. M. Bredel, C. Bredel, D. Juric, G. R. Harsh, H. Vogel, L. D. Recht, B. I. Sikic, High-resolution genome-wide mapping of genetic alterations in human glial brain tumors. Cancer Res 65, 4088 (2005).

30. D. Talantov, A. Mazumder, J. X. Yu, T. Briggs, Y. Jiang, J. Backus, D. Atkins, Y. Wang, Novel genes associated with malignant melanoma but not benign melanocytic lesions. Clin Cancer Res 11, 7234 (2005).

31. A. E. Karnoub, A. B. Dash, A. P. Vo, A. Sullivan, M. W. Brooks, G. W. Bell, A. L. Richardson, K. Polyak, R. Tubo, R. A. Weinberg, Mesenchymal stem cells within tumour stroma promote breast cancer metastasis. Nature 449, 557 (2007).

32. A. L. Richardson, Z. C. Wang, A. De Nicolo, X. Lu, M. Brown, A. Miron, X. Liao, J. D. Iglehart, D. M. Livingston, S. Ganesan, X chromosomal abnormalities in basal-like human breast cancer. Cancer Cell 9, 121 (2006).

33. M. Qiao, Y. Wang, X. Xu, J. Lu, Y. Dong, W. Tao, J. Stein, G. S. Stein, J. D. Iglehart, Q. Shi, A. B. Pardee, Mst1 is an interacting protein that mediates PHLPPs' induced apoptosis. Mol Cell 38, 512

34. J. Liu, H. L. Weiss, P. Rychahou, L. N. Jackson, B. M. Evers, T. Gao, Loss of PHLPP expression in colon cancer: role in proliferation and tumorigenesis. Oncogene, (2008).

35. B. S. Taylor, N. Schultz, H. Hieronymus, A. Gopalan, Y. Xiao, B. S. Carver, V. K. Arora, P. Kaushik, E. Cerami, B. Reva, Y. Antipin, N. Mitsiades, T. Landers, I. Dolgalev, J. E. Major, M. Wilson, N. D. Socci, A. E. Lash, A. Heguy, J. A. Eastham, H. I. Scher, V. E. Reuter, P. T. Scardino, C. Sander, C. L. Sawyers, W. L. Gerald, Integrative genomic profiling of human prostate cancer. Cancer Cell 18, 11 (2010).

36. T. C. G. A. R. Network, Comprehensive genomic characterization defines human glioblastoma genes and core pathways. Nature 455, 1061 (2008).

37. D. W. Parsons, S. Jones, X. Zhang, J. C. Lin, R. J. Leary, P. Angenendt, P. Mankoo, H. Carter, I. M. Siu, G. L. Gallia, A. Olivi, R. McLendon, B. A. Rasheed, S. Keir, T. Nikolskaya, Y. Nikolsky, D. A. Busam, H. Tekleab, L. A. Diaz, Jr., J. Hartigan, D. R. Smith, R. L. Strausberg, S. K. Marie, S. M. Shinjo, H. Yan, G. J. Riggins, D. D. Bigner, R. Karchin, N. Papadopoulos, G. Parmigiani, B. Vogelstein, V. E. Velculescu, K. W. Kinzler, An integrated genomic analysis of human glioblastoma multiforme. Science 321, 1807 (2008).

38. C. A. Joazeiro, T. Hunter, Biochemistry. Ubiquitination--more than two to tango. Science 289, 2061 (2000).

39. D. Frescas, M. Pagano, Deregulated proteolysis by the F-box proteins SKP2 and beta-TrCP: tipping the scales of cancer. Nat Rev Cancer 8, 438 (2008).

40. S. Y. Fuchs, V. S. Spiegelman, K. G. Kumar, The many faces of beta-TrCP E3 ubiquitin ligases: reflections in the magic mirror of cancer. Oncogene 23, 2028 (2004).

41. H. Aberle, A. Bauer, J. Stappert, A. Kispert, R. Kemler, beta-catenin is a target for the ubiquitin-proteasome pathway. EMBO J 16, 3797 (1997).

42. M. Hart, J. P. Concordet, I. Lassot, I. Albert, R. del los Santos, H. Durand, C. Perret, B. Rubinfeld, F. Margottin, R. Benarous, P. Polakis, The F-box protein beta- TrCP associates with phosphorylated beta-catenin and regulates its activity in the cell. Curr Biol 9, 207 (1999).

43. A. Yaron, A. Hatzubai, M. Davis, I. Lavon, S. Amit, A. M. Manning, J. S. Andersen, M. Mann, F. Mercurio, Y. Ben-Neriah, Identification of the receptor component of the IkappaBalpha-ubiquitin ligase. Nature 396, 590 (1998).

44. X. Li, J. Liu, T. Gao, {beta}-TrCP-mediated ubiquitination and degradation of PHLPP1 is negatively regulated by Akt. Mol Cell Biol, (2009).

45. T. Ohta, Y. Xiong, Phosphorylation- and Skp1-independent in vitro ubiquitination of E2F1 by multiple ROC-cullin ligases. Cancer Res 61, 1347 (2001).

46. T. M. Fael Al-Mayhani, S. L. Ball, J. W. Zhao, J. Fawcett, K. Ichimura, P. V. Collins, C. Watts, An efficient method for derivation and propagation of glioblastoma cell lines that conserves the molecular profile of their original tumours. J Neurosci Methods 176, 192 (2009).

47. S. M. Pollard, K. Yoshikawa, I. D. Clarke, D. Danovi, S. Stricker, R. Russell, J. Bayani, R. Head, M. Lee, M. Bernstein, J. A. Squire, A. Smith, P. Dirks, Glioma stem cell lines expanded in adherent culture have tumor-specific phenotypes and are suitable for chemical and genetic screens. Cell Stem Cell 4, 568 (2009).

48. M. M. Inda, R. Bonavia, A. Mukasa, Y. Narita, D. W. Sah, S. Vandenberg, C. Brennan, T. G. Johns, R. Bachoo, P. Hadwiger, P. Tan, R. A. Depinho, W. Cavenee, F. Furnari, Tumor heterogeneity is an active process maintained by a mutant EGFR- induced cytokine circuit in glioblastoma. Genes Dev 24, 1731

49. C. Liu, Y. Tu, X. Sun, J. Jiang, X. Jin, X. Bo, Z. Li, A. Bian, X. Wang, D. Liu, Z. Wang, L. Ding, Wnt/beta-Catenin pathway in human glioma: expression pattern and clinical/prognostic correlations. Clin Exp Med, (2010).

50. I. Lassot, E. Segeral, C. Berlioz-Torrent, H. Durand, L. Groussin, T. Hai, R. Benarous, F. Margottin-Goguet, ATF4 degradation relies on a phosphorylation- dependent interaction with the SCF(betaTrCP) ubiquitin ligase. Mol Cell Biol 21, 2192 (2001).

51. M. Peifer, P. Polakis, Wnt signaling in oncogenesis and embryogenesis--a look outside the nucleus. Science 287, 1606 (2000).

52. M. Karin, The beginning of the end: IkappaB kinase (IKK) and NF-kappaB activation. J Biol Chem 274, 27339 (1999).

53. J. T. Winston, P. Strack, P. Beer-Romero, C. Y. Chu, S. J. Elledge, J. W. Harper, The SCFbeta-TRCP-ubiquitin ligase complex associates specifically with phosphorylated destruction motifs in IkappaBalpha and beta-catenin and stimulates IkappaBalpha ubiquitination in vitro. Genes Dev 13, 270 (1999).

54. F. Mantovani, L. Banks, Regulation of the discs large tumor suppressor by a phosphorylation-dependent interaction with the beta-TrCP ubiquitin ligase receptor. J Biol Chem 278, 42477 (2003).

55. A. Ougolkov, B. Zhang, K. Yamashita, V. Bilim, M. Mai, S. Y. Fuchs, T. Minamoto, Associations among beta-TrCP, an E3 ubiquitin ligase receptor, beta- catenin, and NF-kappaB in colorectal cancer. J Natl Cancer Inst 96, 1161 (2004).

56. M. Davis, A. Hatzubai, J. S. Andersen, E. Ben-Shushan, G. Z. Fisher, A. Yaron, A. Bauskin, F. Mercurio, M. Mann, Y. Ben-Neriah, Pseudosubstrate regulation of the SCF(beta-TrCP) ubiquitin ligase by hnRNP-U. Genes Dev 16, 439 (2002).

57. E. Seo, H. Kim, R. Kim, S. Yun, M. Kim, J. K. Han, F. Costantini, E. H. Jho, Multiple isoforms of beta-TrCP display differential activities in the regulation of Wnt signaling. Cell Signal 21, 43 (2009).

58. P. Pu, Z. Zhang, C. Kang, R. Jiang, Z. Jia, G. Wang, H. Jiang, Downregulation of Wnt2 and beta-catenin by siRNA suppresses malignant glioma cell growth. Cancer Gene Ther 16, 351 (2009).

59. H. Yano, A. Hara, K. Takenaka, K. Nakatani, J. Shinoda, K. Shimokawa, N. Yoshimi, H. Mori, N. Sakai, Differential expression of beta-catenin in human glioblastoma multiforme and normal brain tissue. Neurol Res 22, 650 (2000).

60. S. Utsuki, Y. Sato, H. Oka, B. Tsuchiya, S. Suzuki, K. Fujii, Relationship between the expression of E-, N-cadherins and beta-catenin and tumor grade in astrocytomas. J Neurooncol 57, 187 (2002).

61. N. Yokota, S. Nishizawa, S. Ohta, H. Date, H. Sugimura, H. Namba, M. Maekawa, Role of Wnt pathway in medulloblastoma oncogenesis. Int J Cancer 101, 198 (2002).

62. R. J. Gilbertson, Medulloblastoma: signalling a change in treatment. Lancet Oncol 5, 209 (2004).

Figure 3.1: Akt-mediated feedback loop enhancing PHLPP1 levels is preferentially lost in CNS tumors. (a) Lysates of non-viable cell pellets from the NCI60 panel of tumor cell lines were analyzed by Western blotting for PHLPP1, PHLPP2, Akt phosphorylation on T308 (P308) or S473 (P473), Akt, and actin. The first lane in each gel (C) contains lysate from COS7 cells overexpressing PHLPP1 serves to control for differences in exposure among different blots. (b) The ratio of PHLPP1 to active Akt (T308) was obtained by densitometric analysis of the blots in (a) and plotted across the panel of 60 cell lines. Horizontal lines indicate the value of the median (top line) and one standard deviation below the median value (bottom line). Cell lines with a ratio less than one standard deviation below the median line are indicated by an asterisk. (c) The indicated CNS cell lines were treated for 24 hours with DMSO or LY294002 (20 µM), and PHLPP1, Akt phosphorylation on T308 (P308) or S473 (P473), and actin were detected by Western blot analysis. The relative amount of PHLPP1, normalized to actin, is shown in the graph; data represent the mean ± SEM of three independent experiments.

Figure 3.2: Loss of the feedback loop in glioblastoma is independent of PHLPP1, CK1, and GSK-3. (a) Astrocytoma and glioblastoma cell lines were transfected with HA-PHLPP1 for 24 hours prior to 24 hour treatment with LY294002 (20 µM). Immunoblotting was used to determine levels of exogenous PHLPP1 (αHA) as well as the phosphorylation state of Akt (S473) and actin. The relative amount of exogenous HA-PHLPP1, normalized to actin, is shown in the graph; data represent the mean ± SEM of three independent experiments. * P < 0.05 compared to DMSO. (b) The indicated cell lines were treated with DMSO or LY294002 (20 µM) for 24 hours and immunoblotting was used to determine PHLPP1 levels, the phosphorylation state of Akt (S473), phospho-GSK-3α/β (S21/9), phospho-glycogen synthase (S641), and phospho-β-catenin (S45). Actin was used as a loading control.

Figure 3.3: β-TrCP1 is confined to the nucleus in glioblastoma cell lines and patient samples. (a) Astrocytoma and glioblastoma cell lines were fractionated as described in the Methods section. Fractions representing equal amounts of cell lysate were analyzed by immunoblotting to determine the cellular location of PHLPP1 and β- TrCP1. (b) Frontal cortex from healthy human brain and patient-derived glioblastoma tumor neurosphere specimens were fractionated as described in the Methods section. Fractions were analyzed by immunoblotting to determine the cellular location of PHLPP1 and β-TrCP1. VDAC (membrane), Annexin 1 (cytosol), and Lamin A (nucleus) were used as control markers for the indicated cellular fractions. W = whole cell lysate, M = membrane, C = cytosol, N = nuclear, D = detergent-insoluble pellet.

Figure 3.4: PHLPP1 turnover is slower and mRNA levels are reduced in glioblastoma cell lines. (a) SF-268 (intact loop) and U251 (broken loop) cell lines were treated with DMSO or LY294002 (20 µM) in addition to CHX (5 µM). Lysates were collected at the indicated times and levels of PHLPP1 and the phosphorylation of Akt on S473 (P473) were monitored by Western blot analysis. PHLPP1 levels were normalized to actin, which was used as a loading control. Data from three independent experiments were quantified and fit to an exponential decay to obtain half-times of degradation; points represent average ± SEM. * P < 0.05. (b) Relative PHLPP1 RNA levels were calculated from microarray experiments measured on Affymetrix U95Av2 and U133 arrays, and normalized to the level observed in SNB-75 cells. Data represent the mean ± SEM of six independent experiments.

Figure 3.5: β-catenin turnover is impaired in glioblastoma cell lines. (a) SNB-75 (intact) and U251 (broken) cell lines were treated with CHX (5 µM) and lysates were collected over 24 hours to monitor the levels of β-catenin and actin. β-catenin levels were normalized to actin and data from three independent experiments were quantified and fit to an exponential decay to obtain half-times of degradation; points represent average ± SEM. * P < 0.05. (b) SNB-75 and U251 cells were fractionated as described in the Methods section and immunoblotting was used to examine β-catenin. Densitometry was used to determine the relative levels of β-catenin in each cellular fraction. Graph represents the average ± SEM of three independent experiments. * P < 0.05.

Figure 3.6: Expression of β-TrCP1 in the cytosol rescues Akt-mediated regulation of PHLPP1 in glioblastoma. (a) Expression of a CFP-tagged construct of β-TrCP1 was expressed in U251 cells. DIC (top) and CFP (middle) images were merged (bottom) to show the cellular localization of exogenous β-TrCP1 in GBM cells. (b) Myc-β- TrCP1 was overexpressed in astrocytoma and glioblastoma cell lines for 24 hours prior to 24 hours treatment with DMSO or LY294002 (20 µM). Immunoblotting was used to monitor the levels of exogenous Myc-β-TrCP1 (αMyc), PHLPP1, Akt phosphorylation on S473 (P473), and actin. The relative amount of PHLPP1, normalized to actin, is shown in the graph; data represent the mean ± SEM of three independent experiments. * P < 0.05.

Supplemental Table 3.1: Numerical values corresponding to each of the NCI-60 tumor cell lines analyzed in Figure 1. Supplementary Table 1!

#! Cell Line! Tumor Type! #! Cell line! Tumor Type! 1! BT-549! Breast ! 31! PC-3! Prostate! 2! Hs-578T! Breast ! 32! COLO-205! Colon! 3! MCF-7! Breast ! 33! HCC-2998! Colon! 4! MDA-MB-231!Breast ! 34! HCT-116! Colon! 5! T-47D! Breast ! 35! HCT-15! Colon! 6! MDA-MB-468!Breast ! 36! HT29! Colon! 7! A549! NSCLC! 37! KM12! Colon! 8! EKVX! NSCLC! 38! SW-620! Colon! 9! HOP-62! NSCLC! 39! IGR-OV1! Ovarian! 10! HOP-92! NSCLC! 40! NCI/ADR-RES!Ovarian! 11! NCI-H226! NSCLC! 41! OVCAR-3! Ovarian! 12! NCI-H23! NSCLC! 42! OVCAR-4! Ovarian! 13! NCI-H322M! NSCLC! 43! OVCAR-5! Ovarian! 14! NCI-H460! NSCLC! 44! OVCAR-8! Ovarian! 15! NCI-H522! NSCLC! 45! SK-OV-3! Ovarian! 16! 786-0! Renal! 46! LOX IMVI! Melanoma! 17! A498! Renal! 47! M14! Melanoma! 18! ACHN! Renal! 48! MALME-3M! Melanoma! 19! CAKI-1! Renal! 49! MDA-MB-435! Melanoma! 20! RXF-393! Renal! 50! SK-MEL-2! Melanoma! 21! SN12C! Renal! 51! SK-MEL-28! Melanoma! 22! TK-10! Renal! 52! SK-MEL-5! Melanoma! 23! UO-31! Renal! 53! UACC-257! Melanoma! 24! CCRF-CEM! Leukemia! 54! UACC-62! Melanoma! 25! HL-60! Leukemia! 55! SF-268! CNS! 26! K562! Leukemia! 56! SF-295! CNS! 27! MOLT-4! Leukemia! 57! SF-539! CNS! 28! RMPI-8226! Leukemia! 58! SNB-19! CNS! 29! SR! Leukemia! 59! SNB-75! CNS! 30! DU145! Prostate! 60! U251! CNS!

Supplementary Table 3.2: Experimental identifiers and institutions responsible for generating microarray data made available by the Developmental Therapeutics Program NCI/NIH (http://dtp.cancer.gov/mtargets/mt_index.html).

Supplementary Table 2! Experiment ID! Pattern ID! Investigator! Institution! 26106! GC38735! Dr. Christian Lavedan!Novartis! 47871! GC54239! Dr. Eric Kaldjian! Gene Logic, Inc! 190533! GC150951! Dr. Eric Kaldjian! Gene Logic, Inc! 192695! GC153113! Dr. Eric Kaldjian! Gene Logic, Inc! 131713! GC226980! Dr. Eddie Moler! Chiron Corporation! 129959! GC225226! Dr. Eddie Moler! Chiron Corporation!

CHAPTER 4:

Disruption of the Interface between the PH and Kinase Domains of Akt is

Sufficient for Hydrophobic Motif Site Phosphorylation in the Absence of

mTORC2

Abstract

The pro-survival kinase Akt requires phosphorylation at two conserved residues, the activation loop site (T308) and the hydrophobic motif site (S473), for maximal activation. Previous reports indicate that mTORC2 is necessary for phosphorylation of the hydrophobic motif and that this site is not phosphorylated in cells lacking components of the mTORC2 complex, such as Sin1. Here we show that

Akt can be phosphorylated at the hydrophobic motif site (S473) in the absence of mTORC2. First, increasing the levels of PIP3 in Sin1 -/- MEFs by i] expression of a constitutively-active PI3K or ii] relief of a negative feedback loop on PI3K by prolonged inhibition of mTORC1 or S6K is sufficient to rescue hydrophobic motif phosphorylation of Akt. The resulting accumulation of PIP3 at the plasma membrane results in S473 phosphorylation. Second, constructs of Akt in which the PH domain is constitutively disengaged from the kinase domain are phosphorylated at the hydrophobic motif site in Sin1 -/- MEFs; both myristoylated-Akt and Akt lacking the

PH domain are phosphorylated at S473. Thus, disruption of the interface between the

PH and kinase domains of Akt bypasses the requirement for mTORC2. In summary,

89 90 these data support a model in which Akt can be phosphorylated at S473 and activated in the absence of mTORC2 by mechanisms that depend on removal of the PH domain from the kinase domain.

Introduction

The Akt/protein kinase B Ser/Thr protein kinases play a central role in signaling downstream of phosphatidylinositol 3-kinase (PI3K) (1). The three isoforms

(Akt1, Akt2, and Akt3) share a similar domain structure: an amino-terminal pleckstrin homology (PH) domain, followed by an α-helical linker, and a carboxy-terminal kinase domain that is controlled by phosphorylation (2). Once activated by phosphorylation, Akt phosphorylates defined substrates throughout the cell, ultimately inducing pro-proliferation and anti-apoptotic signaling pathways (1). The activation state of Akt is tightly controlled, and its dysregulation is implicated in the development of a variety of diseases, most notably cancer (3).

One key regulator of Akt phosphorylation is the mammalian target of rapamycin (mTOR), an evolutionarily conserved Ser/Thr protein kinase that forms two distinct protein complexes in cells. These complexes are differentially regulated and perform distinct cellular functions (4). The mTOR complex 1 (mTORC1) is composed of mTOR, Raptor, and mLST8 and is sensitive to inhibition by rapamycin. mTORC1 is a crucial regulator of cell growth in response to nutrients, stress, and growth factors (5, 6). A second complex, mTOR complex 2 (mTORC2), consists of mTOR, Rictor, mLST8, and Sin1 and is generally considered to be rapamycin- insensitive (7, 8). It is this complex, mTORC2, that regulates Akt and, additionally, protein kinase C (9, 10).

Upon biosynthesis, the nascent Akt polypeptide is phosphorylated at the ribosome by mTORC2 (10-12) at a conserved C-terminal site originally identified in protein kinase C and named the turn motif site (13). Phosphorylation of this residue

(T450 in Akt1) (14) is important for the stability of AGC kinases (15, 16). Thus, Akt is heavily ubiquitinated and degraded in cells lacking mTORC2 because the turn motif site is not phosphorylated (10, 11). This processing phosphorylation of Akt is constitutive and its dephosphorylation in cells has not been reported.

Once processed by phosphorylation at the turn motif site, Akt localizes to the cytosol, where it is maintained in an inactive conformation through the interaction between its PH and kinase domains (17). In the presence of proliferative signals, phosphatidylinositol 3-kinase (PI3K) is activated and generates the lipid second messenger phosphatidylinositol 3,4,5-trisphosphate (PIP3). Akt is subsequently recruited from the cytosol to the plasma membrane through binding of its PH domain to PIP3, resulting in a conformational change that separates the PH and kinase domains and unmasks two key regulatory residues whose phosphorylations are required for maximal activation of the kinase. The first site, the activation loop (T308 on Akt1), is phosphorylated by PDK-1 (18). The second site is termed the hydrophobic motif and corresponds to S473 on Akt1. Several candidates have been proposed as potential hydrophobic motif kinases, including Akt itself by autophosphorylation (19), DNA- dependent protein kinase (DNA-PK) (20), integrin-linked kinase (ILK) (21), and, most recently, mTORC2 (22-24). In support of the latter, genetic ablation of the mTORC2- specific subunits, Rictor or Sin1, results in the loss of Akt phosphorylation at both the turn motif and hydrophobic motif sites, resulting in impaired signaling toward certain substrates (22-24). However, it has not been clearly established whether the primary role of mTORC2 in regulation of the hydrophobic motif of Akt is through direct

93 phosphorylation of this site or by facilitating phosphorylation on S473 by other mechanisms.

In this report, we show that Akt can be phosphorylated at the hydrophobic motif site in the absence of mTORC2. In Sin1 -/- cells, activation of PI3K by overexpression of p110α or prolonged inhibition of mTORC1 and/or S6 kinase (S6K) rescues phosphorylation of Akt at S473 and restores its ability to signal to downstream effectors such as FOXO1/3a. Furthermore, expression of a myristoylated Akt or Akt lacking the PH domain is sufficient to rescue hydrophobic motif phosphorylation in

Sin1 -/- MEFs. Our results reveal that the hydrophobic motif site of Akt can be phosphorylated independent of mTORC2 by disrupting the interface between the PH and kinase domains.

Results

Akt can be phosphorylated at S473 in the absence of mTOR kinase activity –

To determine whether phosphorylation of Akt on S473 absolutely depends on mTOR kinase activity, we asked whether phosphorylation at this site was abolished by

Torin1, a selective ATP-competitive inhibitor of mTOR that blocks the activity of both mTOR complexes (25). 293T cells were treated with 50 nM Torin1 and the relative phosphorylation of Akt at each of its three regulatory sites (T450, T308, S473) was monitored using phospho-specific antibodies (Figure 1). Inhibition of mTOR was assessed by examining the phosphorylation of S6 ribosomal protein, a downstream substrate of mTORC1. Figure 1 shows that Torin1 treatment effectively abolished the phosphorylation of S6, with a halftime of 8.9 ± 0.1 minutes; this inhibition was

94 maintained throughout the course of the experiment, consistent with quantitative and sustained inhibition of mTOR. Phosphorylation of Akt at the turn motif (T450) was insensitive to acute mTOR inhibition (no significant decrease in phosphorylation following 120 minutes of Torin1 treatment, compare lanes 1 and 6). In contrast,

Torin1 treatment caused the dephosphorylation of the hydrophobic motif (S473) and activation loop (T308) with kinetics that mirrored the inactivation of mTORC1 (Figure

1, graph). However, unlike S6, this inhibition was transient, and the steady-state levels of S473 and T308 phosphorylation returned to half-maximal and basal levels, respectively, by the 120-minute time point, despite the absence of mTOR kinase activity (Figure 1). These data confirm a role for mTOR kinase activity in sustaining the steady-state phosphorylation of the hydrophobic motif of Akt, but also establish that this site can be phosphorylated in the absence of mTOR kinase activity.

Activation of PI3K in cells lacking mTORC2 restores Akt (S473) phosphorylation - One possible explanation for the recovery of Akt (S473) phosphorylation observed in response to Torin1 is that inhibition of mTORC1 relieves well-documented negative feedback inhibition of PI3K, thus promoting recruitment of

Akt to the plasma membrane and poising it for phosphorylation. Phosphorylation of insulin receptor substrates (IRS) by S6K decreases their function and stability, acting as a negative feedback loop to dampen PI3K signaling in response to mTORC1 activation (26-28). Previous studies in cell lines and clinical trials demonstrate that prolonged inhibition of mTORC1 by rapamycin and its analogs leads to activation of

PI3K, accumulation of PIP3 at the plasma membrane, and ultimately increased phosphorylation and activation of Akt (29-31). Thus, we hypothesized that relief of

95 this feedback loop would promote recruitment of Akt to the plasma membrane, poising the enzyme in an active conformation that enables phosphorylation of the hydrophobic motif site in the absence of mTORC2.

To test whether activation of PI3K via relief of negative feedback unmasks mTOR-independent S473 phosphorylation, we took advantage of a cell system in which mTORC2 is absent. Previous studies have shown that Sin1 is an essential component of mTORC2, and genetic ablation results in the loss of phosphorylation at the turn motif and hydrophobic motif sites of Akt as well as loss of activity toward downstream substrates such as FOXO1/3a (22). WT and Sin1 -/- MEFs were treated with 100 nM rapamycin for up to 24 hours to inhibit mTORC1 and relieve negative feedback on PI3K (Figure 2a). In both cell lines, rapamycin treatment effectively inhibited mTORC1 and S6K within 4 hours, as determined by loss of S6 phosphorylation. Simultaneously, we observed an increase in the mobility of IRS-2, reflecting the loss of phosphorylation by S6K (32). As expected, over time the total levels of IRS-2 protein increased due to a decrease in proteasomal degradation that is normally promoted by S6K-mediated phosphorylation (33, 34). As a direct readout of

PI3K activity, we monitored phosphorylation of p85 on Y458, a site that has previously been reported to track with the activation of this enzyme (35, 36): rapamycin treatment resulted in an approximately 2.5-fold increase in phosphorylation on Y458 within 16 hours (Figure 2a, p-PI3K p85). These data indicate that prolonged treatment with rapamycin relieves feedback inhibition and activates PI3K in WT and

Sin1 -/- MEFs (Figure 2a).

Under basal conditions in WT MEFs, Akt was fully phosphorylated at all three regulatory sites (Figure 2a). As previously reported, Akt was phosphorylated only on the activation loop site but not the turn motif or hydrophobic motif sites in Sin1 -/-

MEFs under basal conditions (8, 9, 22). However, we observed that prolonged rapamycin treatment rescued phosphorylation of Akt on the hydrophobic motif site in

Sin1 -/- MEFs but had no effect on phosphorylation of the turn motif site. Similar results were observed in Rictor -/- MEFs, indicating that the rescue of S473 phosphorylation is not specific to the loss of Sin1 (data not shown). The rescue of Akt hydrophobic motif site phosphorylation correlated with increased levels of IRS-2 and resultant activation of PI3K (Figure 2a). Importantly, while prolonged treatment of

Sin1 -/- MEFs with rapamycin did not restore hydrophobic motif site phosphorylation to WT levels (approximately 30%, compare lanes 5 and 10), phosphorylation of

FOXO1/3a was rescued in concert with phosphorylation of Akt (S473), indicating that this pool of dually phosphorylated Akt is capable of signaling to downstream substrates (Figure 2a). To verify that the rescue of hydrophobic motif site phosphorylation in response to long-term rapamycin treatment was a direct result of p70S6K inhibition, we asked whether a rapamycin-insensitive construct of p70S6K, containing a phospho-mimetic at the mTOR site (T389E), would prevent rescue. Sin1

-/- MEFs transfected with vector, WT p70S6K or p70S6K (T389E) were treated with rapamycin for 24 hours. As expected, S473 phosphorylation was restored by rapamycin treatment in cells expressing both vector and WT p70S6K. However, expression of p70S6K (T389E) negated the ability of rapamycin to cause an increase in IRS-2 levels and restore Akt (S473) phosphorylation, indicating that inactivation of

S6K is necessary to restore phosphorylation of the hydrophobic motif site

(Supplementary Figure 1). These data indicate that inhibition of mTORC1 is sufficient to rescue hydrophobic motif site phosphorylation of Akt in the absence of mTORC2.

Next, we reasoned that if relief of the negative feedback loop on PI3K in response to rapamycin was responsible for promoting hydrophobic motif phosphorylation of Akt, then direct inhibition of S6K should elicit the same effect as rapamycin. To this end, we treated WT and Sin1 -/- MEFs with PF-4708671, a potent, selective inhibitor of S6K, for 24 hours (37). As was observed with rapamycin, inhibition of S6K resulted in the activation and accumulation of IRS-2, which correlated with rescue of Akt phosphorylation at the hydrophobic motif site (Figure

2b). Thus, inhibition of S6K is able to restore S473 phosphorylation of Akt in the absence of mTORC2.

We have previously identified a novel phosphatase, PHLPP, which selectively dephosphorylates the hydrophobic motif sites on Akt and protein kinase C (38, 39). It was recently demonstrated that the expression of PHLPP1 is reduced by rapamycin via inhibition of S6K- and 4EBP-1-mediated protein translation (40). Thus, we asked whether the increased phosphorylation of S473 resulted from reduced expression of

PHLPP1. Figure 2b shows that PHLPP1 expression levels were unaffected by the

S6K inhibitor PF-4708671, under conditions in which a robust increase in S473 phosphorylation was observed. Furthermore, knockdown of PHLPP1 and PHLPP2 did not alter the biphasic kinetics of S473 phosphorylation following Torin1 treatment in 293T cells, although it did result in higher basal phosphorylation of S473 (data not

98 shown). These results indicate that the increase in S473 phosphorylation following

S6K inhibition is not due to the suppression of PHLPP phosphatase activity.

In order to verify that activation of PI3K, as is observed in response to rapamycin, is sufficient to restore hydrophobic motif phosphorylation of Akt in the absence of mTORC2, a myristoylated, and thus constitutively active, construct of the p110α subunit of PI3K (Myr-p110α) was expressed in WT and Sin1 -/- MEFs.

Overexpression of Myr-p110α in both cell lines increased phosphorylation of Akt on the activation loop site (T308) as well as S6 phosphorylation, indicative of heightened

PI3K signaling. Importantly, expression of a constitutively active PI3K in Sin1 -/-

MEFs was able to restore phosphorylation of Akt at the hydrophobic motif site (S473), but not the turn motif site (Figure 2c). These results indicate that activation of PI3K alone is sufficient to restore phosphorylation of the hydrophobic motif site of Akt in the absence of mTORC2.

Furthermore, to verify that phosphorylation of Akt on the hydrophobic motif site in cells lacking mTORC2 was independent of mTOR kinase activity, we treated

Sin1 -/- MEFs with increasing concentrations of Torin1 and monitored the relative phosphorylation state of Akt (S473) and S6 by immunoblotting. In response to increasing concentrations of Torin1, we observed a gradual decrease in S6 phosphorylation, which preceded a dose-dependent increase in phosphorylation of Akt

(S473) (Figure 2d, left). Quantification of Western blots revealed an inverse correlation between the phosphorylation state of Akt (S473) and S6. Akt (S473) phosphorylation was detected upon near complete inhibition of S6 phosphorylation, and maximal phosphorylation of S473 was only reached once S6K activity was

99 entirely blocked (Figure 2c, right). Taken together, these data indicate that activation of PI3K through multiple mechanisms, independent of mTOR kinase activity, is sufficient to restore hydrophobic motif site phosphorylation of Akt.

Rescue of S473 phosphorylation in Sin1 -/- MEFs is dependent upon PI3K activity and the conformation of Akt – The finding that PI3K activation is sufficient to rescue hydrophobic motif phosphorylation of Akt in cells lacking mTORC2 led us to ask whether binding to PIP3 at the plasma membrane mediated this effect.

Specifically, we tested whether rapamycin-induced phosphorylation of the hydrophobic motif site in Sin1 -/- MEFs was prevented by 1] an inhibitor of PI3K

(wortmannin) or 2] an allosteric inhibitor of Akt (Akti VIII) that locks Akt in an inactive conformation and prevents the PH domain from disengaging and binding to membranes (17). As reported above, treatment with rapamycin rescued S473 phosphorylation of Akt (Figure 3a, lane 4). However, when combined with wortmannin, rapamycin was no longer able to induce phosphorylation of Akt at the hydrophobic motif site (Figure 3a, lane 5), revealing that PI3K activity is essential for the rescue. Similarly, treatment with Akti VIII blocked phosphorylation of S473 in response to rapamycin (Figure 3a, lane 6), indicating that translocation of Akt to the plasma membrane and/or a conformational change disrupting the PH and kinase domain interface is also required (Figure 3a, lane 6). Similar to our results using

Torin1 (Figure 2c), we observed that the induction of S473 phosphorylation in Sin1 -/-

MEFs in response to rapamycin tracked with inactivation of S6K (Figure 3a, compare lanes 1 and 4). Furthermore, treatment with increasing concentrations of either

LY294002 or Akti VIII led to a dose-dependent decrease in rapamycin-induced

100 hydrophobic motif site phosphorylation, correlating with the extent of PI3K inactivation or Akt inhibition respectively (Supplemental Figure 1). Taken together, these data indicate that PI3K activity and the ability of Akt to attain an active conformation are essential for phosphorylation of the hydrophobic motif site in the absence of mTORC2.

Disruption of the PH and kinase domain interface of Akt is sufficient for S473 phosphorylation in the absence of mTORC2 – We next sought to determine whether we could bypass the necessity for mTORC2 in hydrophobic motif site phosphorylation of Akt by expression of Akt constructs having altered localization, activity, or structure. Sin1 -/- MEFs were transfected with the following: vector control, WT Akt, myristoylated Akt (Myr-Akt), myristoylated kinase-dead Akt (Myr-KD-Akt), Akt lacking the PH domain (Akt-ΔPH), or Akt with a phospho-mimetic at the turn motif site (Akt-T450D). The transfected Akt was then immunoprecipitated to examine the phosphorylation of S473 and T308. All constructs were phosphorylated to a similar extent on T308 (Figure 3b), with the exception of Akt-ΔPH, which exhibited markedly less phosphorylation at this site. Phosphorylation of T450 was not detected on any of these constructs (data not shown). WT-Akt, similar to our results for endogenous

Akt, was not phosphorylated on S473 (lane 2). Similarly, replacement of the turn motif Thr with a phospho-mimetic (T450D) failed to restore S473 phosphorylation.

Verifying previously published data; Myr-Akt was robustly phosphorylated on S473 in the absence of mTORC2 (10). This restoration of phosphorylation largely depended on the intrinsic catalytic activity of Akt, as significantly less phosphorylation (approximately 30%) accumulated at S473 on a myristoylated

101 construct of Akt rendered catalytically inactive by a point mutation (K179M) in its

ATP-binding pocket (lane 4). Note that two species are labeled with the p-T308 antibody, but only the slower mobility species is labeled with the p-S473 antibody, reflecting a minor pool of Akt that is dually phosphorylated (upper band lanes 3 and

4). Importantly, Akt was robustly phosphorylated at S473 when the PH domain was removed from the catalytic domain of Akt by genetic truncation (lane 5). Therefore, disruption of the PH domain from the kinase domain by either targeting Akt to the membrane and thus forcing the open conformation or by genetic truncation of the PH domain is sufficient to relieve autoinhibition and rescue hydrophobic motif site phosphorylation in the absence of mTORC2. However, phosphorylation of the hydrophobic motif site is severely compromised in an Akt construct with impaired catalytic activity.

Discussion

Akt is crucial to maintaining the proper balance between cell survival and apoptosis, and its hyperactivation is strongly correlated with the onset and progression of human tumors (3). Previous studies have clearly established that PDK-1 is responsible for phosphorylation of Akt at the activation loop (T308) site. Similarly, mTORC2 has been shown to serve a critical function in controlling the phosphorylation of Akt at the turn motif (T450) and hydrophobic motif (S473) sites

(7, 8, 10, 22). Phosphorylation at the turn motif site has recently been established to occur at the ribosome by direct phosphorylation of the nascent polypeptide by mTORC2 (11, 12). In this report we reveal that Akt can be phosphorylated and

102 activated in the absence of mTORC2. Specifically, the requirement for mTORC2 for hydrophobic motif site, but not turn motif site, phosphorylation can be bypassed by disengaging the PH domain from the kinase domain, suggesting a novel role for mTORC2 in controlling the hydrophobic motif site is to promote this disengagement, allowing phosphorylation by alternative mechanisms.

Treatment of Sin1 -/- MEFs with inhibitors of mTOR, mTORC1, or S6K relieves negative feedback on IRS and drives the activation of PI3K. Similar to overexpression of a constitutively-active PI3K, enhancing PIP3 signaling engages Akt at the plasma membrane, altering its conformation, exposing the hydrophobic motif site, and permitting phosphorylation through mTORC2-independent mechanisms.

Interestingly, phosphorylation of Akt (S473) in response to rapamycin was not observed until 8 hours after treatment in Sin1 -/- MEFs (Figure 2a). Given that IRS-2 levels continued to increase up to 24 hours, we speculate that a certain threshold of

PIP3 must be obtained in order rescue S473 phosphorylation in the absence of mTORC2. Indicative of the importance of a conformational change in Akt induced by binding to PIP3, inhibition of PI3K or treatment with an allosteric inhibitor of Akt, which locks Akt in an inactive conformation, prevented phosphorylation of the hydrophobic motif site in response to rapamycin (Figure 3a). Thus, binding to PIP3 and attaining an active conformation are essential for hydrophobic motif site phosphorylation of endogenous Akt in cells lacking mTORC2. Furthermore, under basal conditions, we observe S473 phosphorylation of both a myristoylated Akt and

Akt lacking the PH domain in Sin1 -/- MEFs (Figure 3b). Both of these Akt constructs have a common characteristic; the PH domain is no longer interacting with

103 the kinase domain. These data demonstrate that disruption of the inhibitory interaction between PH domain and the kinase domain of Akt is sufficient to rescue the defect in hydrophobic motif site phosphorylation observed in the absence of mTORC2 (Figure

4).

The finding that the hydrophobic motif site can be phosphorylated in an mTORC2-independent manner raises the question of which kinases modify this site.

A recent report identifies two membrane-associated kinases, IkappaB kinase epsilon and TANK-binding kinase 1, that are able to phosphorylate membrane-bound Akt at

S473 in the absence of mTORC2 (41). However, additional mechanisms are also likely to control Akt because constructs of Akt that do not readily localize to the membrane (e.g. Akt-ΔPH) are robustly phosphorylated at S473 in Sin1 -/- MEFs. An additional mechanism to account for the mTORC2-independent phosphorylation is autophosphorylation, which accounts for phosphorylation of protein kinase C at its hydrophobic motif site (13, 42). Indeed, Akt efficiently autophosphorylates at this site in vitro (19). In this report, the ability of a myristoylated Akt to bypass the mTORC2 requirement for S473 phosphorylation was in large part dependent on the intrinsic catalytic activity of Akt, a result previously noted by Jacinto and coworkers (10).

Thus, numerous kinases are likely to phosphorylate the hydrophobic motif site of Akt when the enzyme is poised in the proper conformation, an event that is facilitated by mTORC2.

Whether mTORC2 activity directly controls the phosphorylation of S473 of

Akt remains unresolved. Here we show that the requirement for mTORC2 can be bypassed by any of several methods to disengage the PH domain from the kinase

104 domain. However, we also show that acute inhibition of mTOR results in the dephosphorylation of Akt with kinetics that mirror the inactivation of mTOR.

Furthermore, regulation of S473 is specifically mediated by mTORC2 because Torin1 acutely reduces S473 phosphorylation on the constitutively phosphorylated Akt-ΔPH construct in Sin1+/+ cells but not in Sin1-/- cells (Supplemental Figure 2). Thus, mTORC2 kinase activity maintains phosphate on S473. The loss of phosphate on the hydrophobic motif following mTORC2 inhibition could reflect 1] suppression of the kinase component (e.g. if catalyzed directly by mTORC2 or any other mTORC2- controlled kinase) or 2] enhancement of the phosphatase component (e.g. if a hydrophobic motif phosphatase is suppressed by mTORC2) in maintaining steady- state levels of phosphate on the HM.

Note that the biphasic effect of mTOR inhibitors on Akt phosphorylation has recently been reported by Rosen and colleagues (43). Curiously, using different catalytic site inhibitors of mTOR, they observed dephosphorylation of both the activation loop and HM, but phosphate was only restored on the activation loop site; this was sufficient to rescue phosphorylation of FOXO1/3a, an Akt substrate previously shown to be dependent upon phosphorylation of the hydrophobic motif site

(22). One possibility is that a small pool of Akt is being phosphorylated at the hydrophobic motif site; consistent with this, we observed that rescue of S473 phosphorylation in several cell lines was minimal (below 10% of WT S473 phosphorylation), yet sufficient to restore signaling to FOXO1/3a. Thus, we find that phosphorylation of both the activation loop and hydrophobic motif sites is effectively restored following prolonged mTOR inhibition, regardless of mTORC2 status.

105

This study reveals a novel function for mTORC2 in the regulation of Akt phosphorylation and activation. Taken together with previous findings, our data are consistent with a model in which mTORC2 has two distinct roles: 1] co-translational phosphorylation of Akt (T450), which depends on mTORC2 (11, 12) and cannot be bypassed, and 2] facilitation of hydrophobic motif site phosphorylation of Akt, which can be effectively bypassed by disengaging the PH domain.

Acknowledgements: We thank Dr. Peter Vogt for the gift of the pCAGGS-Myr-p110α construct and Drs. Estella Jacinto and Bing Su for the gift of Sin1 +/+ MEFs and Sin1-

/- MEFs. This work was supported by NIH GM43154 and NIH GM067946 (A.C.N.) and DoD BCRP award BC093021 (N.W.). Regarding grant BC093021, the U.S.

Army Medical Research Acquisition Activity, 820 Chandler Street, Fort Detrick MD

21702-5014 is the awarding and administering acquisition office. The content of this article does not necessarily reflect the position or the policy of the U.S. Government.

Noel Warfel and Matt Niederst were supported in part by the UCSD Graduate

Training Program in Cellular and Molecular Pharmacology through an institutional training grant from the National Institute of General Medical Sciences, T32

GM007752.

Experimental Procedures

Plasmids: HA-tagged constructs of wild-type Akt (WT-Akt) and myristoylated Akt (Myr-Akt) were generous gifts from Dr. Alex Toker. HA-tagged constructs of WT-p70S6K and p70-S6K (T389E) were kind gifts from Drs. John

106

Blenis and Tianyan Gao. A QuikChange Site-Directed Mutagenesis kit (Qiagen) was used to make single amino acid changes in the following Akt constructs according to the manufacturer’s protocol. K179M was introduced into Myr-Akt to create a kinase- dead construct (Myr-KD-Akt), and T450D was introduced into WT-Akt to create a phospho-mimetic at the turn motif of Akt (Akt-T450D). Amino acids 1-107 of WT-

Akt were deleted to remove the PH domain (Akt-ΔPH). Myristoylated PI3K p110α was a kind gift from Dr. Peter Vogt.

Materials and Antibodies: LY294002, PF-4708671, wortmannin, Akt inhibitor

VIII were purchased from Calbiochem and dissolved in dimethyl sulfoxide (DMSO).

Torin1 was a kind gift from Drs. Nathanael Gray and David Sabatini. Antibodies to

PHLPP1 and PHLPP2 were purchased from Bethyl Laboratories. The following antibodies were purchased from Cell Signaling: phospho-antibodies for T308 (p-

T308), T450 (p-T450), and S473 (p-S473) of Akt, phospho-GSK-3α/β (Ser21/9), phospho-S6 ribosomal protein (Ser235/236), phospho-FOXO1/3a (Thr24/32), phospho-PI3K p85 (Y458), phospho-P70S6K (T389), total PI3K p85, total, p70S6K total S6, total IRS-2, PI3K p110α, and total Akt antibodies. An anti-HA monoclonal antibody was purchased from Covance. A monoclonal antibody to actin was purchased from Sigma. Protein A/G-agarose beads were obtained from Santa Cruz

Biotechnology. All other materials and chemicals were reagent grade.

Cell Transfection and Immunoblotting: 293T, Sin1 +/+ MEF, and Sin1 -/-

MEF cell lines were maintained in DMEM (Cellgro) containing 10% FBS (Hyclone) and 1% penicillin/streptomycin at 37°C in 5% CO2. Transient transfection of Sin1 -/-

MEFs was carried out using Lipofectamine PLUS transfection reagent (Invitrogen)

107 following the manufacturer’s protocol. For immunoblotting, cultured cells were lysed in Buffer A (50 mM Na2HPO4, 1 mM sodium pyrophosphate, 20 mM NaF, 2 mM

EDTA, 2 mM EGTA, 1% Triton X-100, 1 mM DTT, 200 µM benzamidine, 40 µg ml−1 leupeptin, and 1 mM PMSF, pH 7.4), sonicated for 5 s, and protein yield was determined using Coomassie BCA Protein Assay (Pierce). Lysates containing equal amounts of protein were analyzed by SDS-PAGE, and individual blots were probed using the indicated antibodies. Densitometric analysis was performed with

AlphaView analysis software (version 1.3.0.6) by Alpha Innotech Corporation.

Immunoprecipitation: Sin1 -/- MEFs were transfected with the indicated DNA constructs for approximately 30 hours prior to harvest in Buffer B (50 mM Tris-HCl, pH 7.4, 100 mM NaCl, 5 mM EDTA, 1% Triton X-100, 10 mM sodium pyrophosphate, 1 mM phenylmethylsulfonyl fluoride, 1 mM sodium vanadate). Ten percent of the total detergent-solubilized cell lysate was quenched in SDS sample buffer for further analysis, and the remaining lysate was cleared by centrifugation at

13,000 x g for 5 min. The resulting supernatants were incubated with anti-HA antibody overnight at 4 °C and then with Protein A/G-agarose beads for an additional

2 hours. The immunocomplexes were washed three times with Buffer B, separated by

SDS-PAGE, and analyzed by immunoblotting.

Acknowledgements: We thank Dr. Peter Vogt for the gift of the pCAGGS-

Myr-p110α construct and Drs. Estella Jacinto and Bing Su for the gift of Sin1 +/+

MEFs and Sin1-/- MEFs. This work was supported by NIH GM43154 and NIH

GM067946 (A.C.N.) and DoD BCRP award BC093021 (N.W.). Regarding grant

108

BC093021, the U.S. Army Medical Research Acquisition Activity, 820 Chandler

Street, Fort Detrick MD 21702-5014 is the awarding and administering acquisition office. The content of this article does not necessarily reflect the position or the policy of the U.S. Government. Noel Warfel and Matt Niederst were supported in part by the

UCSD Graduate Training Program in Cellular and Molecular Pharmacology through an institutional training grant from the National Institute of General Medical Sciences,

T32 GM007752. Chapter 4 is, in full, a research article accepted for publication in the

Journal of Biological Chemistry, 2011. It was co-authored by Matt Niederst and supported under the guidance of Alexandra C. Newton. I was the primary investigator and author of this research.

109

References

1. L. C. Cantley, The phosphoinositide 3-kinase pathway. Science 296, 1655 (2002).

2. D. R. Alessi, P. Cohen, Mechanism of activation and function of protein kinase B. Curr Opin Genet Dev 8, 55 (1998).

3. J. R. Testa, P. N. Tsichlis, AKT signaling in normal and malignant cells. Oncogene 24, 7391 (2005).

4. D. D. Sarbassov, S. M. Ali, D. M. Sabatini, Growing roles for the mTOR pathway. Curr Opin Cell Biol 17, 596 (2005).

5. S. Sengupta, T. R. Peterson, D. M. Sabatini, Regulation of the mTOR complex 1 pathway by nutrients, growth factors, and stress. Mol Cell 40, 310 (2010).

6. D. A. Guertin, D. M. Sabatini, An expanding role for mTOR in cancer. Trends Mol Med 11, 353 (2005).

7. D. D. Sarbassov, S. M. Ali, D. H. Kim, D. A. Guertin, R. R. Latek, H. Erdjument-Bromage, P. Tempst, D. M. Sabatini, Rictor, a novel binding partner of mTOR, defines a rapamycin-insensitive and raptor-independent pathway that regulates the cytoskeleton. Curr Biol 14, 1296 (2004).

8. Q. Yang, K. Inoki, T. Ikenoue, K. L. Guan, Identification of Sin1 as an essential TORC2 component required for complex formation and kinase activity. Genes Dev 20, 2820 (2006).

9. T. Ikenoue, K. Inoki, Q. Yang, X. Zhou, K. L. Guan, Essential function of TORC2 in PKC and Akt turn motif phosphorylation, maturation and signalling. EMBO J 27, 1919 (2008).

10. V. Facchinetti, W. Ouyang, H. Wei, N. Soto, A. Lazorchak, C. Gould, C. Lowry, A. C. Newton, Y. Mao, R. Q. Miao, W. C. Sessa, J. Qin, P. Zhang, B. Su, E. Jacinto, The mammalian target of rapamycin complex 2 controls folding and stability of Akt and protein kinase C. EMBO J 27, 1932 (2008).

11. W. J. Oh, C. C. Wu, S. J. Kim, V. Facchinetti, L. A. Julien, M. Finlan, P. P. Roux, B. Su, E. Jacinto, mTORC2 can associate with ribosomes to promote cotranslational phosphorylation and stability of nascent Akt polypeptide. EMBO J 29, 3939 (2010).

12. V. Zinzalla, D. Stracka, W. Oppliger, M. N. Hall, Activation of mTORC2 by association with the ribosome. Cell 144, 757 (2011).

110

13. L. M. Keranen, E. M. Dutil, A. C. Newton, Protein kinase C is regulated in vivo by three functionally distinct phosphorylations. Curr Biol 5, 1394 (1995).

14. A. Bellacosa, T. O. Chan, N. N. Ahmed, K. Datta, S. Malstrom, D. Stokoe, F. McCormick, J. Feng, P. Tsichlis, Akt activation by growth factors is a multiple-step process: the role of the PH domain. Oncogene 17, 313 (1998).

15. A. S. Edwards, M. C. Faux, J. D. Scott, A. C. Newton, Carboxyl-terminal phosphorylation regulates the function and subcellular localization of protein kinase C betaII. J Biol Chem 274, 6461 (1999).

16. F. Bornancin, P. J. Parker, Phosphorylation of threonine 638 critically controls the dephosphorylation and inactivation of protein kinase C α. Curr. Biol. 6, 1114 (1996).

17. V. Calleja, M. Laguerre, P. J. Parker, B. Larijani, Role of a novel PH-kinase domain interface in PKB/Akt regulation: structural mechanism for allosteric inhibition. PLoS Biol 7, e17 (2009).

18. V. Calleja, D. Alcor, M. Laguerre, J. Park, B. Vojnovic, B. A. Hemmings, J. Downward, P. J. Parker, B. Larijani, Intramolecular and intermolecular interactions of protein kinase B define its activation in vivo. PLoS Biol 5, e95 (2007).

19. A. Toker, A. C. Newton, Akt/protein kinase B is regulated by autophosphorylation at the hypothetical PDK-2 site. J Biol Chem 275, 8271 (2000).

20. J. Feng, J. Park, P. Cron, D. Hess, B. A. Hemmings, Identification of a PKB/Akt hydrophobic motif Ser-473 kinase as DNA-dependent protein kinase. J Biol Chem 279, 41189 (2004).

21. M. Delcommenne, C. Tan, V. Gray, L. Rue, J. Woodgett, S. Dedhar, Phosphoinositide-3-OH kinase-dependent regulation of glycogen synthase kinase 3 and protein kinase B/AKT by the integrin-linked kinase. Proc Natl Acad Sci U S A 95, 11211 (1998).

22. E. Jacinto, V. Facchinetti, D. Liu, N. Soto, S. Wei, S. Y. Jung, Q. Huang, J. Qin, B. Su, SIN1/MIP1 maintains rictor-mTOR complex integrity and regulates Akt phosphorylation and substrate specificity. Cell 127, 125 (2006).

23. D. D. Sarbassov, D. A. Guertin, S. M. Ali, D. M. Sabatini, Phosphorylation and regulation of Akt/PKB by the rictor-mTOR complex. Science 307, 1098 (2005).

24. D. A. Guertin, D. M. Stevens, C. C. Thoreen, A. A. Burds, N. Y. Kalaany, J. Moffat, M. Brown, K. J. Fitzgerald, D. M. Sabatini, Ablation in mice of the mTORC components raptor, rictor, or mLST8 reveals that mTORC2 is required for signaling to Akt-FOXO and PKCalpha, but not S6K1. Dev Cell 11, 859 (2006).

111

25. C. C. Thoreen, S. A. Kang, J. W. Chang, Q. Liu, J. Zhang, Y. Gao, L. J. Reichling, T. Sim, D. M. Sabatini, N. S. Gray, An ATP-competitive mammalian target of rapamycin inhibitor reveals rapamycin-resistant functions of mTORC1. J Biol Chem 284, 8023 (2009).

26. L. S. Harrington, G. M. Findlay, R. F. Lamb, Restraining PI3K: mTOR signalling goes back to the membrane. Trends Biochem Sci 30, 35 (2005).

27. L. S. Harrington, G. M. Findlay, A. Gray, T. Tolkacheva, S. Wigfield, H. Rebholz, J. Barnett, N. R. Leslie, S. Cheng, P. R. Shepherd, I. Gout, C. P. Downes, R. F. Lamb, The TSC1-2 tumor suppressor controls insulin-PI3K signaling via regulation of IRS proteins. J Cell Biol 166, 213 (2004).

28. O. J. Shah, Z. Wang, T. Hunter, Inappropriate activation of the TSC/Rheb/mTOR/S6K cassette induces IRS1/2 depletion, insulin resistance, and cell survival deficiencies. Curr Biol 14, 1650 (2004).

29. X. Wan, B. Harkavy, N. Shen, P. Grohar, L. J. Helman, Rapamycin induces feedback activation of Akt signaling through an IGF-1R-dependent mechanism. Oncogene 26, 1932 (2007).

30. K. E. O'Reilly, F. Rojo, Q. B. She, D. Solit, G. B. Mills, D. Smith, H. Lane, F. Hofmann, D. J. Hicklin, D. L. Ludwig, J. Baselga, N. Rosen, mTOR inhibition induces upstream receptor tyrosine kinase signaling and activates Akt. Cancer Res 66, 1500 (2006).

31. A. Carracedo, L. Ma, J. Teruya-Feldstein, F. Rojo, L. Salmena, A. Alimonti, A. Egia, A. T. Sasaki, G. Thomas, S. C. Kozma, A. Papa, C. Nardella, L. C. Cantley, J. Baselga, P. P. Pandolfi, Inhibition of mTORC1 leads to MAPK pathway activation through a PI3K-dependent feedback loop in human cancer. J Clin Invest 118, 3065 (2008).

32. D. Hartley, G. M. Cooper, Role of mTOR in the degradation of IRS-1: regulation of PP2A activity. J Cell Biochem 85, 304 (2002).

33. R. Zhande, J. J. Mitchell, J. Wu, X. J. Sun, Molecular mechanism of insulin- induced degradation of insulin receptor substrate 1. Mol Cell Biol 22, 1016 (2002).

34. T. Haruta, T. Uno, J. Kawahara, A. Takano, K. Egawa, P. M. Sharma, J. M. Olefsky, M. Kobayashi, A rapamycin-sensitive pathway down-regulates insulin signaling via phosphorylation and proteasomal degradation of insulin receptor substrate-1. Mol Endocrinol 14, 783 (2000).

35. J. H. Kim, C. Xu, Y. S. Keum, B. Reddy, A. Conney, A. N. Kong, Inhibition of EGFR signaling in human prostate cancer PC-3 cells by combination treatment with beta-phenylethyl isothiocyanate and curcumin. Carcinogenesis 27, 475 (2006).

112

36. C. Lau, X. Wang, L. Song, M. North, S. Wiehler, D. Proud, C. W. Chow, Syk associates with clathrin and mediates phosphatidylinositol 3-kinase activation during human rhinovirus internalization. J Immunol 180, 870 (2008).

37. L. R. Pearce, G. R. Alton, D. T. Richter, J. C. Kath, L. Lingardo, J. Chapman, C. Hwang, D. R. Alessi, Characterization of PF-4708671, a novel and highly specific inhibitor of p70 ribosomal S6 kinase (S6K1). Biochem J 431, 245 (2010).

38. T. Gao, F. Furnari, A. C. Newton, PHLPP: a phosphatase that directly dephosphorylates Akt, promotes apoptosis, and suppresses tumor growth. Mol Cell 18, 13 (2005).

39. T. Gao, J. Brognard, A. C. Newton, The phosphatase PHLPP controls the cellular levels of protein kinase C. J Biol Chem 283, 6300 (2008).

40. J. Liu, P. D. Stevens, T. Gao, mTOR-dependent regulation of PHLPP expression controls the rapamycin sensitivity in cancer cells. J Biol Chem 286, 6510 (2011).

41. X. Xie, D. Zhang, B. Zhao, M. K. Lu, M. You, G. Condorelli, C. Y. Wang, K. L. Guan, I{kappa}B kinase {varepsilon} and TANK-binding kinase 1 activate AKT by direct phosphorylation. Proc Natl Acad Sci U S A 108, 6474 (2011).

42. A. Behn-Krappa, A. C. Newton, The hydrophobic phosphorylation motif of conventional protein kinase C is regulated by autophosphorylation. Curr Biol 9, 728 (1999).

43. V. S. Rodik-Outmezguine, S. Chandarlapary, N. Pagano, P. I. Poulikakos, M. Scaltriti, E. Moskatel, J. Baselga, S. Guichard, N. Rosen, mTOR kinase inhibition causes feedback-dependent biphasic regulation of AKT signaling. Cancer Discovery, (2011).

113

Figure 4.1: Akt can be phosphorylated at S473 in the absence of mTOR kinase activity. 293T cells were treated with Torin1 (50 nM) and lysates were collected over 120 minutes. Immunoblotting with phospho-specific antibodies was used to determine the relative phosphorylation state of Akt on T450, T308, and S473 and of S6. The relative phosphorylation of S6 and Akt (S473) normalized to total protein levels is shown in the graph; data represent the mean ± SEM of three independent experiments.

114

Figure 4.2: Activation of PI3K is sufficient to restore Akt (S473) phosphorylation in cells lacking mTORC2. Sin1 -/- MEFs were (a) treated with rapamycin (100 nM) for the indicated times, (b) treated with DMSO or PF-4708671 (5 µM) for 24 hours, (c) transfected with Myr-p110α for approximately 30 hours, or (d) treated with the indicated doses of Torin1 for 24 hours prior to harvest. Immunoblotting was used to determine total and phosphorylated protein levels, and the relative phosphorylation of S6 and Akt (S473), normalized to actin, was determined by densitometry and fit to a sigmoidal dose response curve; data represent the mean ± SEM of three independent experiments.

115

Figure 4.3: Disruption of the PH and kinase domain interface of Akt is sufficient for S473 phosphorylation in the absence of mTORC2. (a) Sin1 -/- MEFs were treated with rapamycin (100 nM), wortmannin (500 nM), or Akt inhibitor VIII (5 µM), alone or in combination, for 24-hours. Immunoblotting was used to monitor the relative level of total and phosphorylated proteins. (b) Sin1 -/- MEFs were transfected with the indicated Akt constructs for approximately 30 hours prior to harvest. Exogenous Akt was immunoprecipitated and the phosphorylation state of Akt was determined using phospho-specific antibodies.

116

Figure 4.4: Disruption of the interaction between the PH and kinase domains of Akt overcomes the necessity for mTORC2 for hydrophobic motif phosphorylation. Model depicting two distinct roles for mTORC2 in controlling Akt phosphorylation: First, direct, co-translational, phosphorylation of the turn motif, and second, facilitation of disengagement of the PH domain from the kinase domain to expose S473 for phosphorylation by either mTORC2 itself or by other mechanisms (a) In wild-type cells, mTORC2 promotes an active confirmation of Akt, allowing S473 phosphorylation and downstream signaling. (b) In Sin1 -/- MEFs, Akt is only phosphorylated at T308, low levels of PIP3 are not sufficient to disrupt the PH and kinase domain interaction, and S473 is not phosphorylated (c) In Sin1 -/- MEFs, increased and/or sustained levels of PIP3 at the plasma membrane bind Akt through to elicit a conformational change allowing for phosphorylation at S473 in the absence of mTORC2.

117

Supplemental Figure 4.1: Inhibition of S6K is necessary for rapamycin-induced phosphorylation of S473 in Sin1 -/- MEFs. Sin1 -/- MEFs were transfected with vector, HA-p70S6K, or HA-p70S6K (T389E) prior to treatment with rapamycin (100 nM) for 24 hours. Immunoblotting was used to determine total and phosphorylated protein levels

118

Supplemental Figure 4.2: Rescue of Akt (S473) phosphorylation in cells lacking mTORC2 is dependent upon PI3K and the conformation of Akt. (a) Sin1 -/- MEFs were treated with increasing concentrations of rapamycin alone or in combination with increasing doses of LY294002 for 24 hours. (b) Sin1 -/- MEFs were treated with increasing concentrations of rapamycin alone or in combination with increasing doses of Akti VIII for 24 hours. The phosphorylation state of Akt, GSK-3, FOXO1/3a, and S6 were monitored by immunoblotting.

119

Supplemental Figure 4.3: Regulation of Akt (S473) phosphorylation is specifically mediated by mTORC2. Sin1 +/+ MEFs or Sin1 -/- MEFs were transfected with Akt- ΔPH for approximately 30 hours prior to treatment with Torin1 (50 nm) for 1-hour and the phosphorylation state of Akt (S473) and S6 were monitored by immunoblotting.

Chapter 5:

Unraveling a Complex Network of Signal Transduction Pathways

Abstract

Akt is Ser/Thr kinase that plays a vital role in the regulation of a diverse array of cellular functions including growth, survival, and proliferation (1, 2).

Dysregulation of the Akt signaling pathway is a causative factor associated with the development of cancer and therapeutic resistance (3). The PHLPP (PH domain

Leucine-rich repeat protein phosphatase) family of phosphatases has been identified as an important negative regulator of pro-survival cellular signaling pathways, including

Akt, PKC, and MST-1 (4-7). Recent reports indicate that PHLPP expression levels are commonly reduced or lost in cancer, promoting tumorogenesis (8-11). This chapter will review the important findings in this thesis regarding the regulation of Akt phosphorylation and mechanisms responsible for controlling PHLPP stability. This work provides novel insight into the role of mTORC2 in controlling Akt phosphorylation at the hydrophobic motif site, and in turn, its activation in cells. It will focus on the importance of negative feed back loops and mechanisms governing protein stability for maintaining homeostasis between cellular signaling pathways.

Lastly, the implications of these findings will be discussed in relation to their role in

120 121 tumorigenesis and the potential ramifications for targeting the PI3K signaling pathway as an anti-cancer strategy.

122 mTORC2: What does it do?

Akt is controlled by phosphorylation, both chronically and acutely, at three conserved residues: the turn motif site (T450), which is constitutively phosphorylated at the ribosome by mTORC2 (12, 13); the activation loop site (T308), which is phosphorylated by phosphoinositide-dependent kinase – 1 (PDK-1) (14), and the hydrophobic motif site (S473) by a mechanism that is controlled by mTORC2 (15-17).

Once dually phosphorylated at T308 and S473, Akt is fully activated and can phosphorylate all of its characterized substrates. The mammalian target of rapamycin

(mTOR) is an evolutionarily conserved Ser/Thr protein kinase crucial regulator of cell growth in response to nutrients, stress, and growth factors (18-20). mTOR forms two complexes, mTORC1 and mTORC2, the latter has been widely accepted as the direct kinase responsible for phosphorylation of a crucial residue, termed the hydrophobic motif site, that is necessary for full activation of AGC kinases, including Akt. Yet, it remains largely unknown how mTORC2 is activated by growth factors to promote hydrophobic motif phosphorylation of Akt.

Hydrophobic motif phosphorylation of Akt. It has been clearly established that mTORC2 is critical for maintaining phosphate at the turn motif and hydrophobic motif sites of Akt, and our results are consistent with previously published work indicating that the genetic ablation of mTORC2 or inhibition of mTOR activity is sufficient to decrease phosphate on S473 (17, 21). However, we reveal that Akt can be phosphorylated and activated in the absence of mTORC2. In cells lacking mTORC2, prolonged and/or sustained activation of PI3K or the expression of Akt constructs in which the PH domain is constitutively separated from the kinase domain is sufficient

123 to rescue S473 phosphorylation. Furthermore, both PI3K activity and the ability of

Akt to attain an active conformer are necessary for S473 phosphorylation in the absence of mTORC2. This report is the first to propose a mechanism by which mTORC2 regulates Akt phosphorylation by facilitating disengagement of the PH and kinase domains of Akt.

In the absence of proliferative signals, Akt is held in an inactive confirmation due to the autoinhibitory interaction of its PH domain and kinase domains (22). Upon mitogen stimulation, binding of the PH domain to PIP3 removes the PH domain, exposing the hydrophobic motif for phosphorylation (23). In WT cells this phosphorylation is likely to be controlled by mTORC2 directly, as mTOR inhibitors acutely reduce S473 phosphorylation on a constitutively phosphorylated Akt-ΔPH construct in Sin1+/+ cells but not in Sin1-/- cells (Supplemental Figure 4.3).

However, in the absence of mTORC2, the hydrophobic motif site remains masked in response to acute agonist stimulation and the ability of Akt to attain an active confirmation is dramatically impaired. Thus, mTORC2 may play a novel role in regulating Akt phosphorylation by promoting an active conformation, allowing hydrophobic motif phosphorylation by alternative mechanisms.

There is evidence indicating that several kinases may have the ability to directly phosphorylate Akt at S473 in cells (14, 24). Additionally, a recent report identifies two membrane-associated kinases, IkappaB kinase epsilon and TANK- binding kinase 1, that are able to phosphorylate membrane-bound Akt at S473 in the absence of mTORC2 (25). However, our results strongly support a role for autophosphorylation in regulation of hydrophobic motif phosphorylation of Akt, a

124 theory previously proposed by Toker and Newton (26). We find that the ability of a myristoylated Akt to bypass the mTORC2 requirement for S473 phosphorylation was in large part dependent on the intrinsic catalytic activity of Akt, verifying previously published results by Jacinto and coworkers (27). Taken in context with previous reports, numerous kinases, including Akt itself, are likely to phosphorylate the HM of

Akt when the enzyme is poised in the proper conformation, an event that is facilitated by mTORC2.

Therapeutic implications. The compound rapamycin is an mTORC1-specific inhibitor that has been well studied as both an immunosuppressive (28) and anti-tumor agent (29, 30). However, despite promising clinical results, there are many instances where rapamycin has failed to inhibit tumor growth, or even led to disease progression. One mechanism promoting tumor resistance in response to rapamycin is increased activation of the Akt pathway (31). It has been well documented that inhibition of mTORC1 leads to the inactivation of a negative feedback loop through which S6K inhibits IRS protein to dampen PI3K activity (32, 33). Thus, prolonged rapamycin treatment can cause an increase in PI3K activity and ultimately increase

Akt phosphorylation and activation. As a result, there has been increasing interest in the development of ATP-competitive compounds that inhibit the catalytic activity of mTOR, thus blocking signaling by both mTORC1 and mTORC2, with the idea that

Akt can no longer be phosphorylated on S473, and activated, when mTORC2 is not functional.

In contrast, our findings indicate that Akt can be phosphorylated in the absence of mTOR activity, restoring its ability to phosphorylate downstream substrates and

125 promote cell survival. Therefore, we suspect that treatment with mTOR-specific inhibitors alone would not sustainably inhibit Akt activation and are unlikely to negate the negative effects of enhanced PI3K/Akt signaling observed in response to treatment with rapamycin. Indeed, Rosen and colleagues recently demonstrated that prolonged treatment with several catalytic-site inhibitors of mTOR results in the reactivation of

Akt and its ability to phosphorylate substrates, such as FOXO1/3, that are thought to be mTORC2-dependent (34). While they hypothesize that FOXO1/3 is phosphorylated in the absence of S473, the feedback-dependent biphasic reactivation of Akt signaling reported is consistent with a small pool of Akt regaining phosphate on S473. Also, these findings have important implications for the use of mTOR specific inhibitors as a therapeutic strategy in distinct subsets of patients. For example, tumors that have lost PTEN or contain activating mutations in PI3K will likely have elevated basal levels of PIP3, which is sufficient to poise Akt in an active conformation. As a result these tumors may be largely resistant to Akt inhibition in response to mTOR inhibitors. Thus, knowledge on the status of PTEN and other

PI3K/AKT/mTOR-linked pathways may help to select for tumors that will respond well to mTOR-specific inhibitors.

While Akt is one of the most well characterized oncogenes, the exact mechanism controlling phosphorylation of the hydrophobic motif site (S473), and subsequent activation of the kinase, remains unclear. A more complete understanding of the mechanisms responsible for Akt activation at the molecular level is crucial for successful drug development. Therefore, we propose two distinct roles for mTORC2 in the regulation of Akt phosphorylation and activation. First, direct, co-translational,

126 phosphorylation of the turn motif site, and second, facilitating the disengagement of the PH domain from the kinase domain to expose S473 for phosphorylation by mTORC2 itself or alternative mechanisms (Figure 4.4).

PHLPP Feedback

Since I joined the Newton Lab, four years ago, a plethora of data have validated the critical role of PHLPP as tumor suppressor protein in cancer. Possibly the best evidence indicating the importance of PHLPP in tumorogenesis is that knockout mice lacking a single isoform, PHLPP1, develop high grade prostatic intepithelial neoplasia (PIN) and invasive carcinoma in the prostate (35).

Furthermore, loss of PHLPP expression is frequently observed in a broad range of human tumors, conferring a survival advantage to cancer cells. Therefore, we reasoned that mechanisms maintaining the steady state level of PHLPP in cells must also be lost in tumorigenesis.

Akt activity regulates PHLPP1 stability. The PHLPP family comprises three isozymes, the alternatively spliced PHLPP1α and PHLPP1β (which differ in an amino-terminal extension on PHLPP1β), and a separate gene product, PHLPP2. All three isozymes share a similar structure: an N-terminal PH domain, leucine-rich repeat region, PP2C phosphatase domain, and a C-terminal PDZ binding motif (36).

Despite their similarities, the stability PHLPP1 and PHLPP2 isoforms is differentially regulated. Pulse-chase experiments indicate that the turnover of PHLPP1 is much faster than that of PHLPP2. However, treatment with a proteasome inhibitor almost completely blocks the basal turnover of both PHLPP1 and PHLPP2 in cycloheximide

127 chase experiments, indicating that both of these isoforms are primarily degraded by the 26S proteasome.

While the structure, function, and biological output of PHLPP have been thoroughly studied, the mechanisms governing its expression in cells remain unclear.

Using biochemical and genetic methods, we discovered that steady-state levels of

PHLPP1, but not PHLPP2, are regulated by Akt activity. Thus, we have identified a novel feedback loop regulating the output of Akt signaling. Akt inhibition enhances the ubiquitinylation and subsequent degradation of PHLPP1, an effect that can be blocked by inhibiting the proteasome. Thus, under conditions where Akt is relatively inactive, the steady-state levels of PHLPP1 remain relatively low. Whereas, upon activation of Akt, PHLPP1 will increase because its degradation is slowed, ultimately increasing the amount of PHLPP1 available to dephosphorylate Akt and terminate signaling. Thus, Akt activity controls the steady state-levels of PHLPP1, revealing a highly effective mechanism to ensure that Akt signaling is buffered (Figure 2, left).

At the same time we identified that PHLPP1 stability was regulated by Akt signaling, Gao and colleagues identified PHLPP1 as a proteolytic target of β-TrCP, an

E3-ligase that serves as the substrate recognition subunit in the SCF (Skp1-Cullin 1-F- box protein) protein complex (37). Specifically, phosphorylation of PHLPP1, by casein kinase 1 (CK1) and glycogen synthase kinase 3 (GSK-3), at multiple residues within its PP2C domain generates a phosphodegron motif that promotes PHLPP degradation. CK1 is known to be a constitutively active kinase that phosphorylates

PHLPP1, an event that is essential for subsequent phosphorylation by GSK-3 at

Ser847. Upon phosphorylation by both CK1 and GSK-3, PHLPP1 is recognized and

128 bound by β-TrCP, poly-ubiquitinated, and degraded by the 26S proteasome.

Importantly, it has been previously established that the activity of GSK-3β is impaired upon phosphorylation by its upstream kinase, Akt (38). This work verified our findings and established a mechanism by which Akt controls PHLPP1 stability.

PHLPP feedback is lost in cancer. Constitutive activation of the PI3K/Akt pathway is a hallmark of tumorigenesis. Many mechanisms are known to promote constitutive activation of Akt, including gene amplification or gain of function mutations in upstream receptor tyrosine kinase receptors (RTKs), activating mutations in PI3K or Akt itself, or loss of function mutations in the regulatory phosphatase

PTEN (39-42). We reasoned that loss of the negative feedback loop by which

PHLPP1 is stabilized in response to Akt activation could provide a novel mechanism to account for constitutive activation of Akt frequently observed in cancer. A screen of the NCI60 panel of tumor cell lines identified a subset of tumors that presented highly active Akt, but relatively low PHLPP1 levels, indicating that the negative feedback loop between Akt and PHLPP1 is lost (43). Specifically, the cellular levels of PHLPP1 are insensitive to the manipulation of Akt activity in high-grade glioblastoma cell lines. Interestingly, this feedback loop was intact in cell lines derived from lower grade astrocytoma tumors. Thus, the ability of Akt activity to influence the steady-state level of PHLPP1 is specifically lost in glioblastoma (44).

Cellular fractionation revealed that in astrocytoma cell lines and normal brain tissue (intact feedback loop), β-TrCP1 is predominantly cytoplasmic, whereas in glioblastoma cell lines and patient-derived tumor neurospheres (broken feedback loop), the E3 ligase is confined to the nucleus and thus spatially separated from

129

PHLPP1, which is cytoplasmic. As a result, in glioblastoma, although PHLPP1 is properly phosphorylated by upstream kinases, it can no longer interact with β-TrCP1 and PHLPP1 expression levels are no longer sensitive to Akt activity (Figure 2).

Consistent with this, reintroduction of β-TrCP1 to the cytosol of GBM cells was sufficient to restore the ability of Akt to control PHLPP1 levels. This study indicates a novel mechanism for the dysregulation of PHLPP1 levels in cancer. While we identified the mechanism responsible for loss of the feedback loop in glioblastoma, the defect results in a more stable PHLPP1, making it an unlikely mechanism to promote constitutive activation of Akt. Interestingly, PHLPP1 mRNA levels were consistently reduced in a majority of the GBM cell lines tested compared to low grade astrocytomas, suggesting that dysregulation of PHLPP at the transcriptional level may be responsible for promoting Akt signaling in this disease.

Mislocalization of β-TrCP1 in cancer. The finding that β-TrCP1 is confined to the nucleus and cannot properly regulate the cytosolic degradation of substrates is particularly intriguing considering that several of the proteins β-TrCP1 is known to regulate are linked to tumor progression (45). The mislocalization of proteins within the cell is a common mechanism linked the aberrant activation signal transduction pathways. It has been previously reported that the differential subcellular localization of another E3 ligase, Cbl, alters its ability to properly target substrates for degradation

(46). Interestingly, confinement of β-TrCP1 to the nuclear compartment was specifically observed in high-grade glioblastomas, indicating that it may activate survival-signaling pathways that are important for tumor progression and is selected for in late stage disease. Indeed, in addition to PHLPP1, another substrate of β-TrCP1,

130

β-catenin, was stabilized in glioblastoma cell lines. β-catenin is an oncogene whose levels are frequently enhanced in glioblastoma, among other tumor types (47-49). β-

TrCP1 has been reported to regulate the stability of both cytosolic and nuclear substrates, adding another layer of complexity to determining its role in tumorogenesis

(50). Thus, confinement of β-TrCP1 to the nucleus in glioblastoma could promote tumorogenesis by stabilizing oncogenic pathways that drive tumor growth, or alternatively, degrading tumor suppressors in the nuclear compartment. Further investigation is necessary to determine whether the deregulation of proteolysis due to subcellular confinement of an E3 ligase in glioblastoma is common in tumors.

Therapeutic implications. Inhibitors of the PI3K/Akt/mTOR pathways have shown promise as a therapeutic strategy in clinical trials and are the subject of a great deal of anti-cancer research. However, due to the complex interactions governing signal transduction pathways, the activation of compensatory survival pathways often negates the anti-tumor effect of these drugs. In recent years, our knowledge of the mechanisms controlling PHLPP expression levels has greatly improved. Gao and colleagues recently reported that PHLPP1 and PHLPP2 expression are controlled by way of mTOR-dependent protein translation (51). This reports reveals that, at the protein level, Akt activity promotes the stabilization of PHLPP1. Thus, as an unintended side effect of Akt or mTOR inhibitors is that these compounds will likely reduce PHLPP levels, which may hinder their ability to stop tumor growth and induce apoptosis (Figure 2). For example, the use of mTOR inhibitors may have the unintended effect of decreasing the translation of PHLPP protein. Similarly, compounds targeting Akt will ultimately increase the GSK-3-mediated

131 phosphorylation and degradation of PHLPP1. In both of these circumstances, decreasing PHLPP levels will likely result in the induction of signaling through Akt,

PKC, and MST-1 (Figure 2). Alternatively, in glioblastoma cell lines, the efficacy of

Akt-specific inhibitors may be enhanced due to the fact that they have lost the feedback loop between Akt and PHLPP1 because PHLPP1 levels will remain constant despite inhibition of Akt activity. Thus, it will be important to determine whether the reduction of PHLPP expression in response to PI3K/Akt/mTOR pathway inhibitors negates some of their anti-tumorigenic effect.

Future directions

The work in this thesis contributes to the understanding of mechanisms that regulate PI3K signaling. Specifically, two novel findings are: 1] Akt can be phosphorylated at the hydrophobic motif independent of mTORC2, and 2] identification of a feedback loop by which Akt regulates PHLPP1 stability. This work also raises several important questions. In regards to Akt phosphorylation, the mechanism by which mTORC2 promotes hydrophobic phosphorylation remains unclear. We propose that mTORC2 facilitates the disengagement of the PH and kinase domains of Akt. Therefore, similar to studies performed by Calleja and Parker

(22, 23), imaging could be used measure how Akt responds differently to external stimuli in cells lacking mTORC2. Supporting previous work by Toker and Newton

(26), we provide further evidence that autophosphorylation plays a critical role in the regulation of S473. Thus, it may be useful to measure the ability of purified Akt to

132 incorporate phosphate at S473, in the presence or absence of a kinase-dead mTORC2, using in vitro kinase assays to test this claim.

Here, we clearly identify a mechanism controlling the stability of PHLPP1, however, the regulation of PHLPP2 stability remains undefined. Our results indicate that PHLPP2 is degraded in a proteasome-dependent manner. Therefore, high- throughput screening methods to identify the relevant E3 ligase(s) would be helpful in determining both how it is degraded as well as the cellular signaling pathways that influence its stability. While the identified defect in the ability of Akt to regulate

PHLPP does not account for constitutive Akt activation in glioblastoma, it is noteworthy that PHLPP1 mRNA levels are dramatically decreased in glioblastoma.

This finding may be useful in both understating whether PHLPP plays a role in promoting Akt activation in this cancer, as well as a system to investigate mechanisms leading to the reduced PHLPP expression in tumors.

Another key finding presented in this thesis is that β-TrCP1 is mislocalized in glioblastoma. Further work is necessary to determine of whether restoring cytosolic localization of β-TrCP1 in glioblastoma cell lines impedes biological processes associated with tumorogenesis. Furthermore, the mechanism responsible for its confinement to the nuclear compartment remains unsolved. Interestingly, overexpressing exogenous CFP-β-TrCP1 was observed throughout the cytosol of glioblastoma cell lines. Therefore, it is likely that there is a defect in the endogenous protein itself, although it cannot be ruled out that the cytosolic localization of exogenous β-TrCP1 is not an artifact of overexpression. The most likely scenarios leading to β-TrCP1 mislocalization are through a mutation in the protein itself or

133 sequestration in the nucleus due to differential altered protein-protein interactions.

Thus, sequencing β-TrCP1 from glioblastoma cell lines along with a proteomics approach to identify differential interactions with binding partners are promising strategies.

This work reinforces the complex nature of signal transduction that is necessary to maintain cellular homeostasis. A more complete understanding of how these cellular pathways operate is critical to resolving how they become dysregulated in disease.

134

References

1. D. R. Alessi, P. Cohen, Mechanism of activation and function of protein kinase B. Curr Opin Genet Dev 8, 55 (1998).

2. L. C. Cantley, The phosphoinositide 3-kinase pathway. Science 296, 1655 (2002).

3. A. S. Clark, K. West, S. Streicher, P. A. Dennis, Constitutive and inducible Akt activity promotes resistance to chemotherapy, trastuzumab, or tamoxifen in breast cancer cells. Mol Cancer Ther 1, 707 (2002).

4. J. Brognard, E. Sierecki, T. Gao, A. C. Newton, PHLPP and a second isoform, PHLPP2, differentially attenuate the amplitude of Akt signaling by regulating distinct Akt isoforms. Mol Cell 25, 917 (2007).

5. T. Gao, F. Furnari, A. C. Newton, PHLPP: a phosphatase that directly dephosphorylates Akt, promotes apoptosis, and suppresses tumor growth. Mol Cell 18, 13 (2005).

6. T. Gao, J. Brognard, A. C. Newton, The phosphatase PHLPP controls the cellular levels of protein kinase C. J Biol Chem 283, 6300 (2008).

7. M. Qiao, Y. Wang, X. Xu, J. Lu, Y. Dong, W. Tao, J. Stein, G. S. Stein, J. D. Iglehart, Q. Shi, A. B. Pardee, Mst1 is an interacting protein that mediates PHLPPs' induced apoptosis. Mol Cell 38, 512 (2008).

8. J. Liu, H. L. Weiss, P. Rychahou, L. N. Jackson, B. M. Evers, T. Gao, Loss of PHLPP expression in colon cancer: role in proliferation and tumorigenesis. Oncogene, (2008).

9. M. Qiao, J. D. Iglehart, A. B. Pardee, Metastatic potential of 21T human breast cancer cells depends on Akt/protein kinase B activation. Cancer Res 67, 5293 (2007).

10. C. Desmedt, F. Piette, S. Loi, Y. Wang, F. Lallemand, B. Haibe-Kains, G. Viale, M. Delorenzi, Y. Zhang, M. S. d'Assignies, J. Bergh, R. Lidereau, P. Ellis, A. L. Harris, J. G. Klijn, J. A. Foekens, F. Cardoso, M. J. Piccart, M. Buyse, C. Sotiriou, Strong time dependence of the 76-gene prognostic signature for node-negative breast cancer patients in the TRANSBIG multicenter independent validation series. Clin Cancer Res 13, 3207 (2007).

11. P. Ouillette, H. Erba, L. Kujawski, M. Kaminski, K. Shedden, S. N. Malek, Integrated genomic profiling of chronic lymphocytic leukemia identifies subtypes of deletion 13q14. Cancer Res 68, 1012 (2008).

12. W. J. Oh, C. C. Wu, S. J. Kim, V. Facchinetti, L. A. Julien, M. Finlan, P. P. Roux, B. Su, E. Jacinto, mTORC2 can associate with ribosomes to promote

135 cotranslational phosphorylation and stability of nascent Akt polypeptide. EMBO J 29, 3939 (2010).

13. V. Zinzalla, D. Stracka, W. Oppliger, M. N. Hall, Activation of mTORC2 by association with the ribosome. Cell 144, 757 (2011).

14. J. Feng, J. Park, P. Cron, D. Hess, B. A. Hemmings, Identification of a PKB/Akt hydrophobic motif Ser-473 kinase as DNA-dependent protein kinase. J Biol Chem 279, 41189 (2004).

15. E. Jacinto, V. Facchinetti, D. Liu, N. Soto, S. Wei, S. Y. Jung, Q. Huang, J. Qin, B. Su, SIN1/MIP1 maintains rictor-mTOR complex integrity and regulates Akt phosphorylation and substrate specificity. Cell 127, 125 (2006).

16. D. D. Sarbassov, D. A. Guertin, S. M. Ali, D. M. Sabatini, Phosphorylation and regulation of Akt/PKB by the rictor-mTOR complex. Science 307, 1098 (2005).

17. D. A. Guertin, D. M. Stevens, C. C. Thoreen, A. A. Burds, N. Y. Kalaany, J. Moffat, M. Brown, K. J. Fitzgerald, D. M. Sabatini, Ablation in mice of the mTORC components raptor, rictor, or mLST8 reveals that mTORC2 is required for signaling to Akt-FOXO and PKCalpha, but not S6K1. Dev Cell 11, 859 (2006).

18. D. D. Sarbassov, S. M. Ali, D. M. Sabatini, Growing roles for the mTOR pathway. Curr Opin Cell Biol 17, 596 (2005).

19. S. Sengupta, T. R. Peterson, D. M. Sabatini, Regulation of the mTOR complex 1 pathway by nutrients, growth factors, and stress. Mol Cell 40, 310 (2010).

20. D. A. Guertin, D. M. Sabatini, An expanding role for mTOR in cancer. Trends Mol Med 11, 353 (2005).

21. Q. Yang, K. Inoki, T. Ikenoue, K. L. Guan, Identification of Sin1 as an essential TORC2 component required for complex formation and kinase activity. Genes Dev 20, 2820 (2006).

22. V. Calleja, M. Laguerre, P. J. Parker, B. Larijani, Role of a novel PH-kinase domain interface in PKB/Akt regulation: structural mechanism for allosteric inhibition. PLoS Biol 7, e17 (2009).

23. V. Calleja, D. Alcor, M. Laguerre, J. Park, B. Vojnovic, B. A. Hemmings, J. Downward, P. J. Parker, B. Larijani, Intramolecular and intermolecular interactions of protein kinase B define its activation in vivo. PLoS Biol 5, e95 (2007).

24. M. Delcommenne, C. Tan, V. Gray, L. Rue, J. Woodgett, S. Dedhar, Phosphoinositide-3-OH kinase-dependent regulation of glycogen synthase kinase 3 and protein kinase B/AKT by the integrin-linked kinase. Proc Natl Acad Sci U S A 95, 11211 (1998).

136

25. X. Xie, D. Zhang, B. Zhao, M. K. Lu, M. You, G. Condorelli, C. Y. Wang, K. L. Guan, I{kappa}B kinase {varepsilon} and TANK-binding kinase 1 activate AKT by direct phosphorylation. Proc Natl Acad Sci U S A 108, 6474 (2011).

26. A. Toker, A. C. Newton, Akt/protein kinase B is regulated by autophosphorylation at the hypothetical PDK-2 site. J Biol Chem 275, 8271 (2000).

27. V. Facchinetti, W. Ouyang, H. Wei, N. Soto, A. Lazorchak, C. Gould, C. Lowry, A. C. Newton, Y. Mao, R. Q. Miao, W. C. Sessa, J. Qin, P. Zhang, B. Su, E. Jacinto, The mammalian target of rapamycin complex 2 controls folding and stability of Akt and protein kinase C. EMBO J 27, 1932 (2008).

28. S. N. Sehgal, K. Molnar-Kimber, T. D. Ocain, B. M. Weichman, Rapamycin: a novel immunosuppressive macrolide. Med Res Rev 14, 1 (1994).

29. T. F. Cloughesy, K. Yoshimoto, P. Nghiemphu, K. Brown, J. Dang, S. Zhu, T. Hsueh, Y. Chen, W. Wang, D. Youngkin, L. Liau, N. Martin, D. Becker, M. Bergsneider, A. Lai, R. Green, T. Oglesby, M. Koleto, J. Trent, S. Horvath, P. S. Mischel, I. K. Mellinghoff, C. L. Sawyers, Antitumor activity of rapamycin in a Phase I trial for patients with recurrent PTEN-deficient glioblastoma. PLoS Med 5, e8 (2008).

30. I. R. Konings, J. Verweij, E. A. Wiemer, S. Sleijfer, The applicability of mTOR inhibition in solid tumors. Curr Cancer Drug Targets 9, 439 (2009).

32. T. Haruta, T. Uno, J. Kawahara, A. Takano, K. Egawa, P. M. Sharma, J. M. Olefsky, M. Kobayashi, A rapamycin-sensitive pathway down-regulates insulin signaling via phosphorylation and proteasomal degradation of insulin receptor substrate-1. Mol Endocrinol 14, 783 (2000).

33. X. Wan, B. Harkavy, N. Shen, P. Grohar, L. J. Helman, Rapamycin induces feedback activation of Akt signaling through an IGF-1R-dependent mechanism. Oncogene 26, 1932 (2007).

34. V. S. Rodik-Outmezguine, S. Chandarlapary, N. Pagano, P. I. Poulikakos, M. Scaltriti, E. Moskatel, J. Baselga, S. Guichard, N. Rosen, mTOR kinase inhibition causes feedback-dependent biphasic regulation of AKT signaling. Cancer Discovery, (2011).

137

35. M. Chen, C. P. Pratt, M. E. Zeeman, N. Schultz, B. S. Taylor, A. O'Neill, M. Castillo-Martin, D. G. Nowak, A. Naguib, D. M. Grace, J. Murn, N. Navin, G. S. Atwal, C. Sander, W. L. Gerald, C. Cordon-Cardo, A. C. Newton, B. S. Carver, L. C. Trotman, Identification of PHLPP1 as a Tumor Suppressor Reveals the Role of Feedback Activation in PTEN-Mutant Prostate Cancer Progression. Cancer Cell 20, 173 (2011).

36. J. Brognard, A. C. Newton, PHLiPPing the switch on Akt and protein kinase C signaling. Trends Endocrinol Metab 19, 223 (2008).

37. X. Li, J. Liu, T. Gao, {beta}-TrCP-mediated ubiquitination and degradation of PHLPP1 is negatively regulated by Akt. Mol Cell Biol, (2009).

38. D. A. Cross, D. R. Alessi, P. Cohen, M. Andjelkovich, B. A. Hemmings, Inhibition of glycogen synthase kinase-3 by insulin mediated by protein kinase B. Nature 378, 785 (1995).

39. P. K. Vogt, S. Kang, M. A. Elsliger, M. Gymnopoulos, Cancer-specific mutations in phosphatidylinositol 3-kinase. Trends Biochem Sci 32, 342 (2007).

40. J. Li, C. Yen, D. Liaw, K. Podsypanina, S. Bose, S. I. Wang, J. Puc, C. Miliaresis, L. Rodgers, R. McCombie, S. H. Bigner, B. C. Giovanella, M. Ittmann, B. Tycko, H. Hibshoosh, M. H. Wigler, R. Parsons, PTEN, a putative protein tyrosine phosphatase gene mutated in human brain, breast, and prostate cancer. Science 275, 1943 (1997).

41. Y. Samuels, L. A. Diaz, Jr., O. Schmidt-Kittler, J. M. Cummins, L. Delong, I. Cheong, C. Rago, D. L. Huso, C. Lengauer, K. W. Kinzler, B. Vogelstein, V. E. Velculescu, Mutant PIK3CA promotes cell growth and invasion of human cancer cells. Cancer Cell 7, 561 (2005).

42. D. J. Slamon, G. M. Clark, S. G. Wong, W. J. Levin, A. Ullrich, W. L. McGuire, Human breast cancer: correlation of relapse and survival with amplification of the HER-2/neu oncogene. Science 235, 177 (1987).

43. N. A. Warfel, M. Niederst, M. W. Stevens, P. M. Brennan, M. C. Frame, A. C. Newton, Mislocalization of the E3 Ligase, {beta}-Transducin Repeat-containing Protein 1 ({beta}-TrCP1), in Glioblastoma Uncouples Negative Feedback between the Pleckstrin Homology Domain Leucine-rich Repeat Protein Phosphatase 1 (PHLPP1) and Akt. J Biol Chem 286, 19777 (2010).

44. N. A. Warfel, M. Niederst, M. W. Stevens, P. M. Brennan, M. C. Frame, A. C. Newton, Mislocalization of the E3 ligase, beta-transducin repeat-containing protein 1 (beta-TrCP1), in the pleckstrin homology domain leucine-rich repeat protein phosphatase 1 (PHLPP1) and Akt. J Biol Chem 286, 19777 (2011).

138

45. S. Y. Fuchs, V. S. Spiegelman, K. G. Kumar, The many faces of beta-TrCP E3 ubiquitin ligases: reflections in the magic mirror of cancer. Oncogene 23, 2028 (2004).

46. S. Checquolo, R. Palermo, S. Cialfi, G. Ferrara, C. Oliviero, C. Talora, D. Bellavia, A. Giovenco, P. Grazioli, L. Frati, A. Gulino, I. Screpanti, Differential subcellular localization regulates c-Cbl E3 ligase activity upon Notch3 protein in T- cell leukemia. Oncogene 29, 1463 (2009).

47. C. Liu, Y. Tu, X. Sun, J. Jiang, X. Jin, X. Bo, Z. Li, A. Bian, X. Wang, D. Liu, Z. Wang, L. Ding, Wnt/beta-Catenin pathway in human glioma: expression pattern and clinical/prognostic correlations. Clin Exp Med, (2010).

48. R. H. Giles, J. H. van Es, H. Clevers, Caught up in a Wnt storm: Wnt signaling in cancer. Biochim Biophys Acta 1653, 1 (2003).

49. R. T. Moon, A. D. Kohn, G. V. De Ferrari, A. Kaykas, WNT and beta-catenin signalling: diseases and therapies. Nat Rev Genet 5, 691 (2004).

50. D. Frescas, M. Pagano, Deregulated proteolysis by the F-box proteins SKP2 and beta-TrCP: tipping the scales of cancer. Nat Rev Cancer 8, 438 (2008).

51. J. Liu, P. D. Stevens, T. Gao, mTOR-dependent regulation of PHLPP expression controls the rapamycin sensitivity in cancer cells. J Biol Chem 286, 6510 (2011).

139

Figure 5.1: The ability of Akt to regulate PHLPP1 stability is lost in glioblastoma due to the nuclear confinement of β-TrCP1. (left) In normal brain and cell lines with an intact feedback loop, Akt inhibition leads to an increase in GSK-3 activity and subsequent phosphorylation of PHLPP1. Upon phosphorylation by GSK-3, PHLPP1 is recognized by β-TrCP1 and targeted for degradation by the proteasome. (right) In gliobloastoma and cell lines with a broken feedback loop, PHLPP1 is properly phosphorylated by its upstream kinases; however, β-TrCP1 is confined to the nucleus where it can no longer interact with PHLPP1 and target it for degradation.

140

Figure 5.2: PHLPP expression levels are regulated at multiple levels downstream of PI3K. At the protein level, Akt activity positively regulates stability of PHLPP1. PHLPP1 and PHLPP2 are regulated at the protein level by mTOR-dependent translation. PHLPP expression is necessary to dampen signaling through the pro- survival signaling pathways, Akt, PKC, and MST-1.

Appendix A:

Regulation of PHLPP Activity and Evidence of PHLPP Phosphorylation

Introduction

The PHLPP family comprises three isozymes, the alternatively spliced

PHLPP1α and PHLPP1β (which differ in an amino-terminal extension on PHLPP1β), and a separate gene product, PHLPP2 (1, 2). All three isozymes share a similar structure: an N-terminal PH domain, Leucine-rich repeat region, PP2C phosphatase domain, and a C-terminal PDZ binding motif. In addition, both PHLPP1β and

PHLPP2 contain an N-terminal Ras Association (RA) domain that is not present in

PHLPP1α. In contrast to PHLPP1α, PHLPP1β and PHLPP2 contain a predicted Ras

Association (RA) domain near their amino terminus (3). Other proteins harboring RA domains, known as Ras effector proteins, have been shown to interact specifically with the active, GTP-bound form of Ras, and this binding regulates protein function

(4, 5). Indicative of its potential importance, the RA domain of PHLPP is evolutionarily conserved, tracing back to the yeast protein, Cyr1. Previous research on

Cyr1 demonstrates that binding of Ras-GTP to the RA domain of CYR1 is required for the production of cAMP and ultimately the survival of the cell (6). Therefore, it is possible that the phosphatase domain of PHLPP1β and PHLPP2 may be regulated in a

141 142 comparable manner, with the binding of Ras-GTP to the RA domain altering its activity. Previous work in the Newton lab has shown that knockdown of either

PHLPP1 or PHLPP2 dramatically increases the agonist-evoked phosphorylation of

Akt in both normal mammary epithelial cells and breast cancer cells (2). Interestingly, genetic depletion of PHLPP isoforms in Hs578Bst, a normal breast cell line, dramatically increases (by almost two-orders of magnitude) the agonist-evoked phosphorylation of Akt, but has only a modest (two fold) effect on the basal phosphorylation state of Akt. Thus, the goal of this research was to determine if

PHLPP is phosphorylated and/or activated in response to mitogen-stimulation. These data primarily consist of preliminary results (n=1) and need to be further validated.

Results

Mitogen stimulation activates PHLPP1β and PHLPP2, but not PHLPP1α -

To determine whether mitogen stimulation can alter PHLPP activity, full-length

PHLPP1α, PHLPP1β, and PHLPP2 were overexpressed in COS7 cells, and, prior to lysis, these cells were serum starved for 4-hours and treated with either DMSO or EGF

(10ng/ml) for 15 minutes. PHLPP was then immunoprecipitated and phosphatase activity was measured as the ability of PHLPP to dephosphorylate a substrate peptide as previously described (7). Phosphatase activity was normalized to the amount of

PHLPP immunoprecipitated in each sample as determined by immunoblotting. When compared to vehicle treatment, the phosphatase activity of PHLPP1β and PHLPP2 are increased 2- and 4-fold, respectively after stimulation with EGF, while PHLPP1α

143 activity remained unchanged (Figure 1). Thus, the phosphatase activity of PHLPP1β and PHLPP2, but not PHLPP1α, is enhanced in response to mitogen stimulation.

The fundamental difference between the PHLPP isoforms is that PHLPP1α lacks the N-terminal RA domain present in PHLPP1β and PHLPP2. Therefore, only

PHLPP isoforms with a RA domain responded to agonist stimulation, suggesting that the RA domain may play a key role in regulating the phosphatase activity of PHLPP.

To determine whether the N-terminus of PHLPP1β, which contains the predicted RA domain, interacts with Ras, Flag-PHLPP1β-only and a constitutively active construct of Ras (GST-Ras (V12)) were overexpressed in COS7 and glutathione sepharose beads were used to immunoprecipitate Ras. Western blotting indicated that the

PHLPP1β-only region did co-immunoprecipitate with active Ras, indicating binding between these two proteins (N. Warfel, unpublished data). Next, we utilized peptide array technology to determine where Ras was binding to PHLPP1β. The amino sequence encoding the PHLPP1β-only region was spotted in 2 amino acid intervals and overlaid with purified GST-Ras (V12). The Ras-binding domain of Raf was also spotted on the membrane as a positive control (green box). An anti-GST antibody was then used to detect interaction between Ras and PHLPP1 (Figure 2). In a single peptide array experiment, one region of PHLPP1β (amino acid 378-434; red box), which was located outside of the predicted RA domain (yellow), was identified to directly bind RAS (Figure 2). However, we did not observe binding of Ras to the positive control, the Ras binding domain of Raf (FARKTFLKLAFC; blue box) in this experiment.

144

Next, we sought to determine whether PHLPP is phosphorylated in response to mitogen-stimulation, leading to the observed increase in phosphatase activity. To identify post-translational modifications to PHLPP, specifically phosphorylation, we used both biochemical techniques and mass spectrometry. In a manner similar to the phosphatase assays, HA-tagged constructs PHLPP1α and PHLPP2 were overexpressed in 293T cells and treated with either DMSO or 10 ng/ml EGF for 15 minutes prior to lysis. PHLPP was then immunoprecipitated, separated on an SDS- page gel, and silver staining was used to isolate PHLPP for mass spectrometry. The identified phosphorylation sites on PHLPP1α and PHLPP2 are listed in Figure 3. The only two identified on PHLPP1 were found to be mitogen-independent (blue), while several basal (blue) and EGF stimulated (green) sites were identified on PHLPP2. It is also of note that two residues on PHLPP2 showed basal phosphorylation that was lost upon stimulation with EGF (red). Thus, PHLPP2 is phosphorylated upon mitogen stimulation, while PHLPP1α is not.

Interestingly, the region surrounding two of the EGF-stimulated phosphorylations sites identified on PHLPP2 (T1266/Y1268) displays a resemblance to the activation loop motif of another member of the MAPK family, p38. It has previously been reported that phosphorylation of p38 on Thr180 and Tyr182 by

MKKs (3,6, and 4) displaces activation loop, increasing kinase activity and enhancing access to substrate (8). Thus, we sought to determine whether a phospho-specific antibody against the activation loop sites in p38 (T180/Y182) was also able to recognize PHLPP. To this end, exogenous PHLPP1 and PHLPP2 were expressed in

293T cells and immunoprecipitated using an anti-HA antibody. Western blotting

145 showed that the antibody against p-p38 (T180/Y182) strongly reacted with both

PHLPP1 and PHLPP2 (Figure 3, left). Furthermore, EGF stimulation of cells expressing PHLPP2 caused an increase in p-p38 (T180/Y182), which was effectively blocked by pretreatment with the MKK4/6 inhibitor SB203580 (Figure 4, right).

However, a similar increase was observed in a mutant of PHLPP2 in which the EGF stimulated phosphorylation sites have been mutated to Alanine residues

(T1266A/Y1268F/T1273A; HA-P2 (C4)), indicating that the p-p38 antibody may not recognize PHLPP2 at this region.

Second, given that PHLPP is known to directly interact with its substrates, Akt and PKC, these kinases are in position to directly phosphorylate PHLPP. To determine whether these kinases are able to phosphorylate PHLPP, we utilized phospho-specific antibodies that detect cellular proteins only when phosphorylated at the substrate consensus sequences of either Akt (p-Akt substrate; RxRxxS/T) or PKC

(p-PKC substrate; Ser) (9, 10). HA-tagged PHLPP1 and PHLPP2 were immunoprecipitated from 293T cells and immunoblotting was used to monitor p-Akt and p-PKC substrate phosphorylation. The p-Akt substrate antibody specifically recognized PHLPP2, but not PHLPP1 (Figure 5A, left). However, the amplitude of this phosphorylation was not affected by stimulation with agonists (10 ng/ml EGF) or inhibitors (LY294002) of the Akt signaling pathway, indicating that this antibody may not recognize sites that are truly modulated by Akt (Figure 5A, right). Second, the p-

PKC (Ser) substrate antibody reacted with both PHLPP1 and PHLPP2 (Figure 5B, left). Furthermore, stimulation with PdBU dramatically increased p-PKC substrate phosphorylation on PHLPP2, which was reversed by pretreatment with the PKC

146 inhibitors, Gö76 and Gö83, indicating that this phosphorylation is specific to PKC

(Figure 5B, right). Also, PHLPP1 and PHLPP2 reacted with another p-PKC substrate antibody against the consensus sequence TxR, and immunoprecipitation of the following isolated domains of PHLPP1: lacking the PH domain (ΔPH), the Leucine- rich repeat region alone (LRR), the PP2C domain (PP2C), and WT revealed that this antibody specifically recognizes the PP2C domain of PHLPP1 (Figure 5C). While there was very little recognition of the WT-PHLPP1 construct, phosphorylation of a p-

PKC (TxR) sequence in the PHLPP1-PP2C domain was found to be PKC-specific, as modulation of PKC activity by treatment with agonists (PdBU) and inhibitors (Gö83) altered the signal. Thus, this work suggests that PHLPP1 and PHLPP2 are phosphorylated by their substrates in cells.

Discussion

Phosphorylation is a well-documented mechanism used to control enzymatic activity. While the use of phospho-substrate antibodies is far from a direct assay for kinase-specific phosphorylation, it is a useful method for screening novel kinase substrates (9). Our results indicated that p-PKC (TxR and Ser) substrate antibodies react with PHLPP1 and PHLPP2, and p-PKC (TxR) specifically reacts with the PP2C domain of PHLPP1. It is interesting to note that WT PHLPP1 is not recognized by p-

PKC (TxR), while the isolated PP2C domain is responds in a PKC-dependent manner, suggesting that the PKC phosphorylation site in question is masked in the full-length protein. Importantly, these results were validated using PKC agonists and inhibitors.

Next, the p-Akt substrate antibody detected PHLPP2, and to a lesser extent, PHLPP1

147 but this signal remained unchanged following Akt activation or inhibition. Therefore, these findings suggest that PHLPP may be a substrate of Akt and PKC in cells.

Phosphatase assays indicate that the activity of PHLPP1β and PHLPP2, but not

PHLPP1α, is increased following EGF stimulation. One possible explanation for the differential activation of PHLPP isoforms is that in contrast to PHLPP1α, PHLPP1β and PHLPP2 contain a predicted Ras Association (RA) domain near their amino terminus. We found that the PHLPP1β-only region bound to active Ras in co- immunoprecipitation experiments. Furthermore, a peptide array revealed that Ras does not bind to the predicted RA domain, but instead at a different region within the

N-terminus of PHLPP1β. However, co-immunoprecipitation studies with PHLPP mutants lacking the predicted RA domain or the identified Ras binding region were never done to verify where Ras is truly binding at these sites. It has been previously reported that the LRR region of PHLPP1 can bind Ras and dampen signaling to MEK and ERK (11). Therefore, Ras may bind PHLPP at multiple sites, which have differential effects on signaling.

An alternative explanation for the increase in PHLPP activity following EGF treatment is post-translational modification. Mass spectrometry data revealed PHLPP isoforms are phosphorylated both under basal conditions and in response to EGF stimulation (Figure 2). Interestingly, phosphorylation sites were only identified within the C-Terminal linker region of both PHLPP1 and PHLPP2. The fact that these phosphorylation sites are located outside of the catalytic domain likely indicates an indirect mechanism for regulating activity, such as a conformational change.

Interestingly, two residues within PHLPP2 lost phosphate after EGF stimulation. It is

148 possible that upon EGF stimulation and subsequent activation, PHLPP2 autodephosphorylates at this site. Previous work indicates that autodephosphorylation is a common mechanism regulating the enzymatic activity of phosphatases (12).

Importantly, several agonist induced phosphorylation sites were identified in PHLPP2, but not PHLPP1, following EGF stimulation. This observation correlates our previous data indicating that PHLPP2, but not PHLPP1α, displayed increased phosphatase activity in response to EGF stimulation. It is also of note that PHLPP2 was recognized by an antibody directed against the activation loop site of p38

(T180/Y182). On PHLPP2 the p-p38 signal increased upon EGF treatment, a response that was completely blocked by pretreatment with SB203580, an inhibitor of

MKKs which are the upstream kinases responsible for phosphorylation of this site on p38 (8). This result indicates that PHLPP2 may be phosphorylated in a mitogen- dependent manner by MKKs. Mutations of the identified pT-G-pY motif on PHLPP2 did not abolish the p-p38 signal, indicating that this antibody may recognize another site within PHLPP2. A similarly strategy could be used to rule out the other three potential phosphorylation motifs in PHLPP2 that resemble the activation loop of p38

(a pT-G-pY sequence). In the future it will be important to make phosphomimetics at these sites on PHLPP2 to determine any are important for mediating PHLPP activity, specifically in response to EGF stimulation.

This work provides evidence that PHLPP is phosphorylated in cells and preliminary evidence that PHLPP isoforms may be phosphorylated by their substrates,

Akt and PKC. We show that PHLPP1β and PHLPP2, but not PHLPP1α isoforms are phosphorylated and activated in response to EGF stimulation. We also identify

149 several novel phosphorylation sites on PHLPP1 and PHLPP2. Further work is necessary to determine the role of these constitutive, and mitogen-stimulated phosphorylations on PHLPP function.

Acknowledgements: I would like to thank Dr. Charles King for sharing his expertise in protein detection and mass spectrometry and Dr. Emma Sierecki for technical support with the phosphatase assays.

Experimental Procedures:

Plasmids: HA-PHLPP1α, PHLPP1β, and PHLPP2 have been previously described (1, 2, 13). The PHLPP2 C4 mutant was made using a QuikChange Site-

Directed Mutagenesis kit (Qiagen) was used to make the following single amino acid changes: T1266A, Y1268F, and T1273A in a construct of WT PHLPP2.

Phosphatase Assay: PHLPP constructs were expressed in 293T cells for approximately 24 hours prior to treatment with DMSO or EGF (10 ng/ml) for 15 minutes. Exogenous PHLPP was immunoprecipitated and phosphatase assays were performed as previously described (7, 14).

Peptide Array: Purified GST-RasV12 The N-terminal region of PHLPP1β was spotted as 18-mer peptides onto a membrane. The peptide array membrane was incubated overnight at 4ºC with 1µM GST or GST-RasV12 in diluted in BSA. The membrane was then washed three times in PBST, blocked in 5% milk, and incubated with an anti-GST antibody overnight and GST was detected by immunoblotting.

150

Materials and Antibodies: LY294002, SB203580, Gö76, and Gö83 were purchased from Calbiochem. Epidermal growth factor (EGF) was purchased from

Peprotech Inc. and dissolved in Phosphate buffered saline. Substrate antibodies against p-Akt, p-PKC (Ser), p-PKC (TxR), and p-p38 (T180/Y182) were purchased from Cell signaling. An anti-HA monoclonal antibody was purchased from Covance.

Protein A/G-agarose beads were obtained from Santa Cruz Biotechnology. All other materials and chemicals were reagent grade.

Immunoprecipitation. 293T cells were transiently transfected using Effectene

(Qaigen) reagent according to the manufacturer’s protocol. Approximately 36 h post- transfection, cells were lysed in buffer B (50 mM Tris-HCl, pH 7.4, 100 mM NaCl, 5 mM EDTA, 1% Triton X-100, 10 mM sodium pyrophosphate, 1 mM PMSF, and 1 mM sodium vanadate, cleared by centrifugation, and equal amounts of protein for each sample were incubated with the anti-HA antibody and Ultra-link protein A/G- agarose (Pierce) overnight at 4°C. The immunoprecipitates were washed three times in buffer B and proteins were separated by SDS-PAGE and analyzed by immunoblotting.

151

References

1. T. Gao, F. Furnari, A. C. Newton, PHLPP: a phosphatase that directly dephosphorylates Akt, promotes apoptosis, and suppresses tumor growth. Mol Cell 18, 13 (2005).

2. J. Brognard, E. Sierecki, T. Gao, A. C. Newton, PHLPP and a second isoform, PHLPP2, differentially attenuate the amplitude of Akt signaling by regulating distinct Akt isoforms. Mol Cell 25, 917 (2007).

3. J. Brognard, A. C. Newton, PHLiPPing the switch on Akt and protein kinase C signaling. Trends Endocrinol Metab 19, 223 (2008).

4. P. Rodriguez-Viciana, C. Sabatier, F. McCormick, Signaling specificity by Ras family GTPases is determined by the full spectrum of effectors they regulate. Mol Cell Biol 24, 4943 (2004).

5. F. Shima, T. Okada, M. Kido, H. Sen, Y. Tanaka, M. Tamada, C. D. Hu, Y. Yamawaki-Kataoka, K. Kariya, T. Kataoka, Association of yeast adenylyl cyclase with cyclase-associated protein CAP forms a second Ras-binding site which mediates its Ras-dependent activation. Mol Cell Biol 20, 26 (2000).

6. M. Kido, F. Shima, T. Satoh, T. Asato, K. Kariya, T. Kataoka, Critical function of the Ras-associating domain as a primary Ras-binding site for regulation of Saccharomyces cerevisiae adenylyl cyclase. J Biol Chem 277, 3117 (2002).

7. J. Brognard, M. Niederst, G. Reyes, N. Warfel, A. C. Newton, Common polymorphism in the phosphatase PHLPP2 results in reduced regulation of Akt and protein kinase C. J Biol Chem 284, 15215 (2009).

8. J. D. Ashwell, The many paths to p38 mitogen-activated protein kinase activation in the immune system. Nat Rev Immunol 6, 532 (2006).

9. B. D. Manning, A. R. Tee, M. N. Logsdon, J. Blenis, L. C. Cantley, Identification of the tuberous sclerosis complex-2 tumor suppressor gene product tuberin as a target of the phosphoinositide 3-kinase/akt pathway. Mol Cell 10, 151 (2002).

10. J. Y. Kim, D. Saffen, Activation of M1 muscarinic acetylcholine receptors stimulates the formation of a multiprotein complex centered on TRPC6 channels. J Biol Chem 280, 32035 (2005).

11. K. Shimizu, M. Okada, K. Nagai, Y. Fukada, Suprachiasmatic nucleus circadian oscillatory protein, a novel binding partner of K-Ras in the membrane rafts, negatively regulates MAPK pathway. J Biol Chem 278, 14920 (2003).

152

12. J. Chen, B. L. Martin, D. L. Brautigan, Regulation of protein serine-threonine phosphatase type-2A by tyrosine phosphorylation. Science 257, 1261 (1992).

13. N. A. Warfel, M. Niederst, M. W. Stevens, P. M. Brennan, M. C. Frame, A. C. Newton, Mislocalization of the E3 ligase, beta-transducin repeat-containing protein 1 (beta-TrCP1), in the pleckstrin homology domain leucine-rich repeat protein phosphatase 1 (PHLPP1) and Akt. J Biol Chem 286, 19777 (2011).

14. E. Sierecki, W. Sinko, J. A. McCammon, A. C. Newton, Discovery of small molecule inhibitors of the PH domain leucine-rich repeat protein phosphatase (PHLPP) by chemical and virtual screening. J Med Chem 53, 6899 (2010).

153

Figure A.1: The phosphatase activity of PHLPP1β and PHLPP2, but not PHLPP1α, is increased in response to EGF stimulation. COS7 cells were transfected with HA- PHLPP isoforms and cells were treated with DMSO or EGF (10 ng/ml) for 15 minutes. Phosphatase activity was measured by a colorometric assay using pNPP as a substrate.

154

Figure A.2: Active Ras does not bind to the predicted RA domain of PHLPP1β, but binds to a separate region. (a) GST alone (top) or GST-RasV12 (bottom) was purified and overlayed on a peptide array of the N-Terminus of PHLPP1β and an immunoblotting with an anti-GST antibody was used to detect Ras binding. Indicated are the novel Ras binding region (red), the predicted RA domain (yellow), and the Ras binding domain of Raf (blue). (b) The identified Ras binding region is shown next to the corresponding peptide sequence for each spot.

155

Figure A.3: Identification phosphorylation sites on PHLPP1 and PHLPP2 under basal and agonist-stimulated conditions. 293T cells were transfected with HA- PHLPP1α and HA-PHLPP2 for approximately 24 hours prior to treatment with DMSO or EGF (10 ng/ml) for 15 minutes. PHLPP constructs were immunoprecipitated and mass spectrometry was used to identify phosphorylation sites. (top panel) Sequence alignment of C-Terminal portion of PHLPP1α and PHLPP2, delineating the end of the PP2C domain (yellow) and the PDZ binding motif (pink), and identified phosphorylation sites present under basal conditions (red), following EGF treatment (green), or under both conditions (blue). (bottom panel) A numerical list of the amino acid residues on PHLPP1β and PHLPP2 identified by mass spectrometry are indicated in the table.

156

Figure A.4: PHLPP is phosphorylated at a site that is recognized by the p-p38 antibody. 293T cells were transfected with HA-PHLPP1α and HA-PHLPP2 for approximately 24 hours prior to lysis. (left) PHLPP constructs were immunoprecipitated and Western blotting was used to detect p-p38 and HA. (right) 293T cells were transfected with PHLPP2 or a mutant of PHLPP2, termed C4 (*T1266A/Y1268F/T1273A). Prior to lysis cells were treated with DMSO, EGF (10 ng/ml), or pretreated with SB203580 (10 µM) for 15 minutes prior to EGF stimulation. Following immunoprecipitation, immunoblotting was used to detect p- p38 and HA.

157

Figure A.5: PHLPP is phosphorylated by its substrates, Akt and PKC. (a) 293T cells were transfected with HA-PHLPP1α and HA-PHLPP2 for approximately 24 hours prior to lysis (left), or treated with EGF (10 ng/ml) or LY294002 (20 µM) for 20 minutes prior to lysis (right). PHLPP was then immunoprecipitated and immunoblotting was used to detect p-Akt substrate. (b) 293T cells were transfected with HA-PHLPP1α and HA-PHLPP2 for approximately 24 hours prior to lysis (left), or treated with PdBU (200 nm) alone for 10 min or pretreated with DMSO, Gö76 (500 nM), or Gö83 (250 nM) for 20 minutes prior to the addition of PdBU (right). Following immunoprecipitation, immunoblotting was used to detect p-PKC (Ser) substrate and HA. (c) 293T cells were transfected with the indicated HA-PHLPP1 construct for approximately 24 hours prior to treatment with PdBU (200 nm) alone or following 20 min pretreatment with Gö83 (250 nM). Following immunoprecipitation, immunoblotting was used to detect p-PKC (TxR) substrate and HA (red = PP2C domain).