UC Berkeley UC Berkeley Electronic Theses and Dissertations

Title Primary resistance to ATP-competitive mTOR inhibitors for the treatment of solid tumors

Permalink https://escholarship.org/uc/item/2vw456fk

Author Ducker, Gregory Stuart

Publication Date 2013

Peer reviewed|Thesis/dissertation

eScholarship.org Powered by the California Digital Library University of California

Primary resistance to ATP-competitive mTOR inhibitors for the treatment of solid tumors

Gregory Stuart Ducker

A dissertation submitted in partial satisfaction of the

requirements for the degree of

Doctor of Philosophy

Chemistry

in the

Graduate Division

of the

University of California, Berkeley

Committee in charge:

Professor Kevan M. Shokat, Chair Professor Christopher J. Chang Professor Karsten Weis

Spring 2013

Primary resistance to ATP-competitive mTOR inhibitors for the treatment of solid tumors

Gregory Stuart Ducker Abstract

Primary resistance to ATP-competitive mTOR inhibitors for the treatment of solid tumors

Gregory Stuart Ducker

Doctor of Philosophy in Chemistry

University of California, Berkeley

Professor Kevan M. Shokat, Chair

The mammalian target of rapamycin (mTOR) functions to integrate nutrient and energy availability with extracellular growth factor signals to regulate macromolecular biosynthesis including protein translation and lipogenesis. This essential gene in conserved from yeast to humans and is a core metabolic regulatory element. Aberrant regulation of metabolism is a phenotype recognized as a hallmark of cancer, and it has been shown that increases in translation alone can be oncogenic. Additionally, many of the most common oncogenes in cancer alter the growth factor signaling network upstream of mTOR and these lesions likely activate mTOR in many cancers. Thus, repression of mTOR activity is an emerging therapeutic strategy for human cancer and there is both strong mechanistic and epidemiological evidence to suggest that attenuation of mTOR signaling may be broadly applicable. As a kinase with a defined small- molecule binding pocket, mTOR presents an attractive target for therapeutic intervention because of its conserved role in integrating many different oncogenic lesions and its central requirement for metabolic regulation. The successful application of mTOR inhibitors to clinical oncology will require the development of potent and selective inhibitors of this kinase and equally importantly, an understanding of which oncogenic lesions mark tumors as specifically sensitive to mTOR inhibition as well as independent biomarkers for in vivo efficacy. Fortuitously, mTOR is inhibited with near perfect selectivity by the natural product rapamycin, and analogs of this compound have been approved for the treatment of solid tumors. Rapamycin inhibits mTOR through a non-conserved non-competitive mechanism of action and only blocks the phosphorylation of certain substrates. ATP-competitive inhibitors of mTOR have recently been invented that occupy the kinase active site and block all substrate phosphorylation. In many preclinical models they are significantly more potent than rapamycin and as they enter clinical trials, questions about how to maximize their therapeutic index naturally arise. This may be challenging to ask for mTOR because it is not mutated in cancer and what genetic lesions mark cancers as susceptible or not to mTOR inhibitors has not been determined. To address the question of how best to apply ATP-competitive mTOR inhibitors to human cancer, I performed a large (~650) unbiased cell screen to identify markers for sensitivity and resistance to PP242, a tool compound that has been developed into the phase I clinical drug, MLN0128. In comparison to rapamycin, PP242 was a more effective compound and more cell

1 lines were inhibited. Colon origin was significantly associated with resistance to both drugs. For PP242, mutations in the gene PIK3CA were a marker of sensitivity. I subsequently focused on colon cancer because it gave the strongest mTOR drug dependent signature, and one of resistance. A panel of mTOR and PI3K pathway inhibitors differentiated colon cancer cell lines based on RAS and PIK3CA genotypes. Ordering of colon cancer cell lines by sensitivity to PP242 revealed a striking resistance to mutations in KRAS within the already resistant colon cancer set. I identified that this KRAS specific resistance was due to a specific failure to inhibit phosphorylation of the translational repressor 4E-BP1 even when other mTOR substrates were inhibited. Resistance correlated with the amount of KRAS in the active GTP-bound form and was independent of canonical mitogen activated protein kinase signaling. Finally, introduction of mutant PIK3CA can sensitize even KRAS mutant colon cancer cells to PP242 and the mechanism is correlated with 4E-PB1 phosphorylation. In colon cancer cell lines I identified predictive markers of sensitivity and resistance and a biomarker that reported upon functional inhibition of mTOR in vivo. Cell lines have well documented shortcomings, and I worked to characterize a colon cancer patient-derived xenograft (PDX) model and apply it to early drug discovery to validate these findings. The PDX model uses metastatic colon cancer removed from patients that is then propagated in nude mice. Each patient tumor can be expanded into a cohort of identical tumors, and a drug trial can be conducted on an individual’s tumor. In the colon cancer PDX model, PP242 was orally bioavailable and acutely inhibited mTOR signaling in many tumors. In a continuous dosing trial however, sensitivity to the drug paralleled what was observed in cell lines. Single KRAS mutant tumors were refractory to treatment. Overall, failure to durably inhibit 4E-BP1 phosphorylation correlated with failure to respond to treatment. I discovered a set of cell lines that were resistant to the ATP-competitive mTOR inhibitor PP242 and identified a defect in inhibition of 4E-BP1 phosphorylation as giving rise to this phenotype. This mechanism of resistance appears to be common in KRAS mutant colon cancer cell lines as well as patients. The differential inhibition of distinct mTOR substrates I discovered reveals an additional layer of as yet uncharacterized biological control in this kinase signaling pathway. 4E-BP1 is a robust biomarker for ATP-competitive drugs and should be strongly considered for clinical use in trials of these agents.

2 Table of Contents

Preliminary Pages

Table of Contents i List of Figures iii , List of Tables iv List of Abbreviations v Acknowledgments vi

Chapter 1: Introduction- Inhibition of mTOR for the treatment of malignant neoplasms 1

1.1 The mammalian target of rapamycin (mTOR) signaling complex 2 1.2 Nutrient sensing functions of mTOR 4 4 1.3 The role of mTOR in oncogenic growth factor signaling pathways 4 4 1.4 Pharmacological inhibition of mTOR 5 5 1.5 ATP-competitive inhibitors in the clinic and patient selection 7 7 1.6 Summary and significance 8 8

Chapter 2: Cell screening for the identification of markers of sensitivity and resistance to mTOR inhibitors 10 10

2.1 Abstract 11 11 2.2 Introduction-High throughput cell screening to identify resistance and sensitivity to kinase inhibitors 11 11 2.3 PP242 is more potent inhibitor of cell growth than rapamycin 12 12 2.4 Colon and pancreas are mTOR inhibitor resistant organ types 14 14 2.5 PIK3CA and RAS mutations are markers for response to PP242 but not rapamycin 16 2.6 PIK3CA mutations are predictors of sensitivity to PP242 in multiple tumor types 20 2.7 Cluster analysis reveals that PI3K/mTOR inhibitor segregate colon cancer cell lines by genotype 21 21 2.8 Discussion 22 22 2.9 Data and Methods 23 23

Chapter 3: Patient-derived xenografts in preclinical drug development 25

3.1 Abstract 26 3.2 Introduction- Xenografts in colon cancer 26 3.3 Establishment of human metastatic colorectal cancers xenografts 26 3.4 Synthesis of ATP-competitive mTOR inhibitor PP242 29 3.5 Oral dosing of PP242 is effective at inhibiting mTOR phosphorylation in vivo 29 3.6 Discussion 33 3.7 Materials and Methods 33

Chapter 4: Incomplete inhibition of phosphorylation of 4E-BP1 as a mechanism of primary resistance to ATP-competitive mTOR inhibitors 36

4.1 Abstract 37 4.2 Introduction 37 4.3 Screening of cancer cell lines 38 4.4 mTORC1 substrates 4E-BP1 and rpS6 are differentially inhibited by PP242 40 4.5 MAPK signaling differences do not alter mTORC1 substrate specificity 43 4.6 Mutant PIK3CA but not PTEN loss leads to mTOR inhibitor sensitization 45 4.7 KRAS mutation status predicts response to PP242 in human primary xenografts 48 4.8 Inhibition of 4E-BP1 and not rpS6 correlates with anti-tumor effect of PP242 50 4.9 Discussion 50 4.10 Materials and Methods 52

Chapter 5: Conclusions and future perspectives 55

5.1 The genetic landscape of cancer and targeted therapies 56 5.2 Combining anti-mTOR therapy with other agents 57 5.3 Conclusion 57

References 59

Appendix 1: A cell growth screen of mTOR inhibitors rapamycin and PP242 70

ii List of Figures

Figure 1.1 The mTOR signaling pathway 3 Figure 1.2 Clinical mTOR inhibitors 6 Figure 1.3 mTORC1 inhibition induced feedback activation of Akt 7 Figure 2.1 Distribution of cell growth responses to treatment with mTOR inhibitors 13 Figure 2.2 Distribution of cell growth responses by organ type 15 Figure 2.3 Residuals from linear regression models of genotype and mTOR inhibitor response 19 Figure 2.4 Effect of genotypes in PP242 sensitivity in select organ types 20 Figure 2.5 A panel of PI3K/mTOR inhibitors distinguishes colon cancer cell lines by genotype 21 Figure 3.1 Patient-derived xenografts maintain key morphological features of human colon cancer 27 Figure 3.2 Synthetic scheme for PP242 and MLN0128 30 Figure 3.3 PP242 inhibits mTOR outputs in different colorectal cancer genetic backgrounds 31 Figure 3.4 Tumor morphology of PP242 treated tumors 32 Figure 4.1 An unbiased cell screen reveals factors leading to resistance and sensitivity to the ATP-competitive mTOR inhibitor PP242 39 Figure 4.2 mTORC1 substrates are differentially inhibited in PP242 resistant versus sensitive cell lines 41 Figure 4.3 PP242, but not rapamycin inhibits mTORC1 substrates in colon cancer cell lines 42 Figure 4.4 Quantification of phosphorylated mTORC1 substrates upon PP242 treatment 44 Figure 4.5 Inhibition of MAPK signaling does not alter mTORC1 substrate phosphorylation 46 Figure 4.6 PIK3CA mutation but not PTEN loss sensitizes KRAS mutant cells to PP242 47 Figure 4.7 KRAS mutant patient-derived xenografts are resistant to PP242 by incomplete inhibition of 4E-BP1 phosphorylation 49

iii List of Tables

Table 2.1 Linear regression models for genotypes and PP242 17 Table 2.2 Linear regression models for genotypes and rapamycin 18 Table 3.1 Genotypes and patient characteristics of tumors used in this study 28 Table 4.1 KRAS and PIK3CA mutations modulate sensitivity to mTOR inhibition 40

iv List of Abbreviations

4E-BP1 eIF4E binding protein 1 AMPK 5’ adenosine monophosphate-activated protein kinase ATP adenosine triphosphate CML chronic mylogenous leukemia EGFR epidermal growth factor receptor ER estrogen receptor GST glutathione S-transferase IC50 inhibitory constant 50 IRS insulin receptor substrate LKB1 liver kinase B1 MAPK mitogen activated protein kinase mLST8 mammalian lethal with SEC13 protein 8 mTOR mammalian target of rapamycin mTORC1 mTOR complex 1 mTORC2 mTOR complex 2 mSin1 mammalian stress-activated map kinase-interacting protein 1 NSCLC non-small cell lung cancer OLS ordinary least squares PARP poly (ADP-robose) polymerase PDX patient-dervied xenograft PH pleckstrin homology PKCα protein kinase C alpha PIP3 phosphatidylinositol (3,4,5)-triphosphate PTEN phosphatase and tensin homolog raptor regulatory-associated protein of mTOR rictor raptor independent companion of mTOR complex 2 rpS6 ribosomal protein S6 RTK receptor tyrosine kinase S6K p70S6 Kinase S or Ser serine SGK1 serum and glucocorticoid protein kinase 1 SREBP Sterol regulatory element-binding protein T or Thr threonine TSC tuberous sclerosis protein

v Acknowledgements

Completing my doctorate required the support and assistance of a great many people, for help both scientific and moral. Here I would like to name them and their contributions to my efforts, without which I would surely not have made it to completion. I cannot overstate how important their faith in me, belief in me, and friendship to me has been and how these have given me the strength and tools necessary to realize my academic pursuits. First I need to acknowledge the huge debt of gratitude I owe my advisor, Professor Kevan Shokat. I have had an amazing experience in graduate school, and so much of it is due to the fact that I was able to complete my research in the Shokat lab. And it was not a straightforward path for me- I wasn’t even enrolled at the right school. I met Kevan on the Berkeley admitted students day in spring 2007, not even realizing how unusual it was for him to meet with a prospective Cal student. It was at that initial meeting that I first saw the qualities of Kevan that I have sought to emulate and I knew then that he was my first choice of advisor from everyone I had met. Kevan has an ability to spin the complex threads of experimental science into a whole narrative cloth that reveals a greater truth. His clear presentation of his uniquely creative science gave me much to aspire to and I am forever grateful that I made the commitment to come across the bay and accept his invitation to work at the University of California, San Francisco. It has been a pleasure and an honor to work for Kevan and I am proud to have been trained by him. At Berkeley, I have to acknowledge the chemical biology graduate program for admitting me. I thank former chemistry department chair Professor Michael Marletta for helping to convince me to go to Berkeley and sell me on chemical biology at Cal. I thank my rotation advisors at Berkeley during my first year. I learned an immense amount that year and it was due in large part to them. Professor Christopher Chang was not only my first rotation advisor, but also my qualifying exam committee chair and served on my thesis committee. I thank Professors Rebecca Heald and Karsten Weis of molecular and cell biology for my third rotation. Furthermore, Professor Weis served as the outside member of my qualifying exam committee and thesis committees. I also thank Professors Carolyn Bertozzi and John Kuriyan for serving on my qualifying exam committee. My thesis work touched on many different fields and I was very dependent upon collaborators for the success of my project. In particular, I have had a deep multi-year collaboration with the surgical oncology group at the University of California, San Francisco. Led by Dr. Robert Warren, a fellow Saint Paul native, this group developed the mouse xenograft models and asked the patient driven questions that drove my research. I’d like to thank Dr. Warren for his support and assistance. David Donner Ph.D. has been a constant supporter of my work and been extremely generous in his time to help edit and improve my manuscripts. Mary Matli located patient samples, performed staining and coordinated much of the essential logistics. Medical oncologist Dr. Emily Bergsland met with us frequently during the gestation of my project and made key insights into the importance of treating KRAS mutant colon cancer patients that really drove the research forward. Dr. Jeffry Simko performed pathology on the weekends for my mouse samples. And lastly, I need to thank Dr. Byron Hann of the UCSF preclinical therapeutic core facility for managing, caring for and treating the mice used in my experiments. My ability to easily interface with Byron, and his staff scientist, Don Hom Ph.D., allowed me to execute my collaboration with the surgical oncology group with a minimum of difficulty. They both showed interest in the project and provided valuable help in shaping experimental design and ensuring a successful outcome.

vi What made graduate school so enjoyable and successful for me was the immense amount of support I received from my fellow students and post docs in the Shokat lab. For this I need to first thank our lab administrator, Valerie Ohman. She made this process so smooth for me and took on all sorts of work on my behalf so that I could concentrate only on research. Her organizational skills and leadership make this lab work as well as it does. I owe much to the senior graduate students who worked on mTOR and PI3K before me and provided me with so much. I thank Eli Zunder Ph.D. and Morri Feldman Ph.D. for their great assistance. Eli and Morri are great biologists and they provided me with an amazing start. My closest collaborator in the Shokat lab has been a clinical fellow and now assistant faculty member, Dr. Chloe Atreya. Chloe trained as a medical oncologist in gastro-intestinal cancers (GI), and her expertise was invaluable for me. She gave me the medical knowledge I needed, told me about conferences to attend including the ASCO GI conference in San Francisco and went with me to the AACR conference on molecular targeted therapies in San Diego in 2012. She has been a major supporter of my research and I am very happy to have worked so closely with her. My rotation in the Shokat lab was overseen by the most talented synthetic chemist I have ever met, Tatsuya Okuzumi Ph.D. I had virtually no experience in medicinal chemistry, and still am by no means a skilled organic chemist, but nearly everything I know I learned from him in 10 short weeks that were bisected by Christmas and New Years. He was extremely generous and understanding of my poor skills, and a patient teacher. I joined Kevan’s lab in the same class as three other graduate students with chemistry backgrounds, Joseph Kliegman, Michael Lopez and Nicholas Hertz. Suffice it to say, graduate school would have not have been the same without them and our collective friendship is one of the core defining features of this entire experience for me. Somehow we all ended up in the same bay and have been a constant presence together on the bench for 5 years. I have learned more about the day-to-day work of being a scientist and more importantly how to think like a scientist from them than any other single source. Each of them knows how to combine serious scholarship and questioning with an infectious joie-de-vie that makes science a joy. I liked coming into lab because I liked talking with Nick, Mike and Joe and I knew that they would always have my back if things weren’t going well. Research is hard, but it need not be painful or solitary and one can have a full life in and out of lab. The greatest sorrow in leaving the Shokat lab is the dissolution of this group of researchers but I’m confident I will remain friends with them for the rest of my career. My family has always valued education and scholarship and their support of my work has been invaluable. I would like to thank my father and mother, Dr. Thomas Ducker and Suzanne Ducker. I cannot easily summarize what they’ve given me, short of saying everything. They have always challenged me with high expectations and their support has been extraordinary. I want to thank my best friends, my brothers Michael, Laurence and Erik Ducker, for their support and friendship. Lastly, I want to thank my friends outside of lab for putting up with my trials and tribulations and remaining close while I was in California. I would like to thank old friends Meg Pain, Kaisa Taipale, Ethan Mooar, and Mark Huberty who have been with me for many years now despite now being far apart. I would like to thank my undergraduate advisor Professor Joseph Chihade for turning me on to biological chemistry and being my greatest advocate. I thank Emily Johnson for her support during my first years here and the transition to the west coast and Austin Pitcher for helping me to find my footing after moving to the city. Thanks to Katyn Chmielewski for perspective and moral support in gritting out the degree. And thanks to

vii Squaw Valley and Alpine Meadows for providing much needed relief in the long dark winter months and reminding me that I am from the north country and truly at home in the snow.

viii

CHAPTER 1:

Introduction- Inhibition of mTOR for the treatment of malignant neoplasms

1 1.1 The mammalian target of rapamycin (mTOR) signaling complex

The mammalian target of rapamycin (mTOR) is a serine/threonine protein kinase member of the phosphoinositide 3-kinase (PI3K) related kinase (PIKK) family. An evolutionarily conserved regulator of cell metabolism, mTOR integrates external growth factor signaling with nutrient availability to control protein translation and influence cell growth and proliferation (1,2). mTOR function is essential and conserved in eukaryotes; homozygous mTOR knockouts in mice are embryonic lethal (3). In humans mTOR is active in the catalytic core of two related heteromeric protein complexes, mTORC1 and mTORC2 (4-6). mTORC1 regulates protein translation by phosphorylating critical regulatory proteins, of which the best described are p70S6 kinase (S6K) which controls ribosome biogenesis through ribosomal protein S6 (rpS6) and eIF4E binding protein 1 (4E-BP1), a repressor of cap-dependent translation (7,8). The kinase activity of mTOR is tightly regulated by the protein complexes in which it exclusively functions. These two defined signaling complexes are multimeric, related and evolutionarily conserved (Figure 1.1a) (9). Common components of the two complexes include the proteins mammalian lethal with SEC13 protein 8 (mLST8) and deptor (1,2,10). mTOR complex 1 (mTORC1) is defined by the additional presence of regulatory-associated protein of mTOR (raptor) and PRAS40 (3-5). In mTOR complex 2 (mTORC2) the proteins mammalian stress-activated map kinase-interacting protein 1 (mSin1), raptor independent companion of mTOR complex 2 (rictor) define the complex while protor has been recently identified as a component (4-6,11). These distinct sets of component proteins shape the substrate specificity of the two mTOR complexes. These differences also lead to distinct effects of pharmacological agents. Most notably, mTORC1 is acutely inhibited by the natural product macrolide rapamycin via a unique allosteric mechanism that selectively inhibits substrate access to mTOR only when found as a member of mTORC1. Rapamycin forms a ternary complex between the FRB domain of mTOR, itself and the prolyl isomerase FKB12 (Figure 1.1b) (7,8,12,13). When mTOR is bound with rictor in mTORC2, the FRB domain of mTOR is occluded and rapamycin is ineffective. However, in some cell types and likely in humans, chronic rapamycin treatment may be able to titrate out mTOR from mTORC2 (9,14). The two complexes have distinct and non-overlapping sets of substrates. The best characterized phosphorylation targets of mTORC1 control protein synthesis via the protein kinase p70S6 (S6K) and the translational repressor protein 4E-BP (15). mTORC1 mediated control of translation is specific and certain genes are selectivity inhibited (16). One pathway notably affected by mTOR-regulated translation is lipid synthesis. The biosynthetic genes in this pathway are controlled by the transcription factors sterol regulatory element binding protein1/2 (SREBP1/2) which themselves are regulated by mTORC1 (17,18). Independent of translation, mTORC1 critically regulates autophagy and acute inhibition induces this cell scavenger program. mTORC1 phosphorylates regulatory components ATG13 and ULK1 to regulate this activity (19- 21). In contrast mTORC2 targets are less well described and all seem to be members of the AGC kinase family. Serine (S) 473 of AKT is the best-described target in mammalian systems (22), but serum and glucocorticoid protein kinase 1 (SGK1) and protein kinase C alpha (PKCα) have also been described at mTORC2 specific targets. SGK1 enhances sodium ion channel expression through phosphorylation that is mTORC2 activity specific (23). Phosphorylation of PKCα mediates a cytoskeleton specific effect of mTORC2 activity in yeast and perhaps certain mammalian cell lineages (11).

2 ab mTORC1 mTORC2 FKBP12 RAP mTOR mTOR mTOR Pras40 Raptor Sin1 Rictor Pras40 Raptor mLST8 deptor mLST8 deptor mLST8 deptor Protor

R T HO O HO O K OH PTEN OH P P P P S I P T308 P P P P P P O R P S473 PI3Kα Akt P S S p85 ATP RAS mTORC2 P TSC1 TSC2 AMPK Raf AAs

P P Mek S6K mTORC1 RAGs

P P Erk P P P P P S6 4E-BP1

PROLIFERATION TRANSLATION eIF4E

Figure 1.1 The mTOR signaling pathway. (a) The mTORC1 and mTORC2 complexes are independent conserved signaling complexes. (b) Rapamycin engages FKBP12 to inhibit mTOR in a non-competitive manner only when it is a member of mTORC1. (c) Human growth factor signaling pathways that impact mTOR. Frequently mutated oncogenes and tumor suppressors are colored in red.

3 The x-ray crystal structure of mTOR solved in complex with mLST8 provides mechanistic insight into how these different complexes regulate access to the active site (24). Previous work had identified a conserved TOR signaling motif (TOS) present in mTORC1 substrates that was shown to interact with raptor (25). But work with rapamycin and PP242 has shown that different substrates of mTORC1 are treated differently and additional regulation was necessary. mTOR was crystalized in the active confirmation suggesting that regulation of phosphorylation occurs entirely via substrate accessibility. The rapamycin binding FRB domain of mTOR is an insert in the N-lobe of the kinase domain and creates a crowded active site. This domain acts as a second mTORC1 substrate recognition site (after raptor) and also binds FKB12 with rapamycin to alter access to the active site as was seen earlier in a cryogenic electron microscopy structure of the complete complex (26). The difficult accessibility of the constitutively active mTOR active site, coupled with the prime positioning of the FRB domain to interact with substrates allow for a mechanistic understanding of how rapamycin inhibits some but not all mTORC1 substrates and mTORC2 functions completely differently.

1.2 Nutrient sensing functions of mTOR

The conserved role of mTOR in eukaryotes is that of an integrator of different nutrient states to regulate cell growth and metabolic homeostasis (27). These nutrient sensing pathways persist in humans and have recently been elucidated in great detail. mTOR senses adenosine triphosphate (ATP) levels via liver kinase B1 (LKB1) which signals through 5’ adenosine monophosphate-activated protein kinase (AMPK) (28,29). Amino acid levels appear to be sensed via a set of Rag GTPases that bring the entire mTORC1 complex to the lysosome and are able to sense amino acid levels in the lysosomal lumen (30,31). An unexpected role for these amino acid sensors was uncovered in mice homozygous for a constitutively active mutant of RagA. Mice developed normally but died immediately after birth, in an effect traced to impaired glucose sensing by mTORC1 (32). These results demonstrate that mTOR senses levels of both available and potential cellular energy (ATP and glucose) as well as core metabolic building blocks (amino acids) and that by integrating these signals it is a master regulator of cell growth.

1.3 The role of mTOR in oncogenic growth factor signaling pathways

mTORC1 is regulated by growth factor signaling via PI3K signaling through AKT (Figure 1.1c) (33,34). Activated receptor tyrosine kinases (RTKs) recruit PI3K to the cell membrane via phospho-tyrosine binding SH2 domains contained in the regulatory subunits of PI3K (35). PI3K is made up of a catalytic and regulatory subunit. Each catalytic unit has preferred regulatory partners. These regulatory units repress the catalytic function and mutations impair this control mechanism (36). There are 4 isoforms of catalytic PI3K subunits (3 class 1A- α, β, δ and 1 class IB, γ). PI3K α and β are universally expressed whereas γ and δ are restricted to myeloid lineages. All PI3K perform the same chemical reaction, the phosphorylation of the 3- position of phosphatidylinositol (4,5)-bisphosphate to form phosphatidylinositol (3,4,5)- triphosphate (PIP3). PIP3 recruits AKT via a pleckstrin homology (PH) domain to the membrane where it can be phosphorylated by PDK1 at threonine (T) 308 and mTORC2 at S473. AKT phosphorylates many substrates critical for cellular metabolism and transcription. It also controls mTOR via phosphorylation-induced repression of tuberin, a member of the tuberous sclerosis complex (TSC1/2), and the repressor of the mTORC1 activator Rheb.

4 Dysregulated mTORC1 signaling is present in human diseases that alter metabolism, including both diabetes and cancer (37,38). In cancer, tumor suppressor loss leading to elevated mTOR protein levels has been observed with some frequency, and functional mTOR mutations are observed in 1% of tumors but no conserved mutations have been identified (39,40). Instead, mTOR is more commonly subject to many variations of aberrant activation as part of the mTORC1 complex (41). mTORC1 is activated by mutations in upstream signaling networks, and in cancer the best described alterations are in the PI3K/AKT/TSC pathway (33). Many of the most commonly mutated oncogenes and tumor suppressors in cancer signal to mTOR including diverse receptor tyrosine kinases, PI3K, phosphatase and tensin homolog (PTEN) and AKT. Ras, the most mutated oncogene in cancer can activate the p110α catalytic subunit of PI3K and it may be considered to signal upstream of mTOR as well (42). It has been well-documented that sporadically occurring solid tumors display mTOR activation as evidenced by immunohistochemical analysis of the most common mTOR substrates S6K and 4E-BP1 (43). The activation of aberrant mTOR signaling by a specific upstream mutation is most clearly seen in tuberous sclerosis. This disease is characterized by inherited mutations in tuberin that constitutively activate mTOR signaling and drive tumor growth; treatment with rapamycin analogs that block mTORC1 signaling is now clinically approved for these patients (44). Rare sporadic TSC1 mutations in cancer are also highly rapamycin responsive (45). However, whether mutations in genes further upstream, including PIK3CA and AKT are also predicative of sensitivity to mTOR targeted therapy and which downstream biomarkers of mTOR activity are most correlated with clinical response is not yet clear. One recent study assessing the efficacy of rapamycin in pre-clinical models used both upstream and downstream signals (p-AKT and p- rpS6) as biomarkers to predict sensitivity to the drug suggesting that both signals are independently important in predicting response (46).

1.4 Pharmacological inhibition of mTOR

Rapamycin (clinically known as sirolimus) first found clinical efficacy as a immunosuppressant (47). Through T-cell specific mechanisms, rapamycin is able to impair T- cell mediated responses without affected the humoral response (48). Long-term treatment with rapamycin is generally well tolerated, although hyperglycemia is commonly seen leading to the onset of diabetes (49-51). Provocatively, this insulin resistance phenotype was shown to be due to inhibition of the rapamycin insensitive mTORC2 complex in vivo, and that this on-target sensitivity was separable from rapamycin’s life extension benefits (52). Rapamycin’s low toxicity, high selectivity, and history of successful clinical application make it an attractive drug for several indications, but it may have only limited efficacy in the oncology setting. There are currently only two mTOR inhibitors, the rapamycin derivatives temsirolimus and everolimus, approved for the treatment of solid tumors (Figure 1.2a). Both are indicated for advanced renal cancer although the activity of these agents is somewhat limited and the mechanism of action is debated (53,54). New phase III data suggest that everolimus also delays pancreatic neuroendocrine tumor progression, and it recently gained approval for this indication, although overt tumor shrinkage is rare (55,56). Recent data suggest a role for everolimus in the treatment of estrogen receptor (ER) positive breast cancer due to selective mTOR pathway activation downstream of this amplification(57). And yet both clinical and cell line data strongly support that notion that rapamycin is only a partial mTOR inhibitor and as such is unable to

5 inhibit hyper-activated mTOR signaling in such a way as to cause complete growth arrest or cell death. A new and potentially more efficacious class of mTOR inhibitors has been developed specifically to target cancer (58-60). These small molecule drugs are not natural products and directly bind the ATP binding pocket of the mTOR kinase domain. The discovery of these molecules allowed for the division of canonical mTORC1 substrates into classes: those sensitive to inhibition by rapamycin (S6K) and those that were relatively insensitive to rapamycin treatment (4E-BP1). This discovery has provided mechanistic insight into the poor success of rapamycin derivatives in clinical trials and how more complete inhibitors of mTOR signaling may be significantly more effective in inhibiting neoplastic growth.

R= Rapamycin (Sirolimus) Everolimus

Temsirolimus

BEZ-235

MLN0128 WYE-125132 OSI-027 AZD8055

Figure 1.2 Clinical mTOR inhibitors. (a) Rapamycin (marketed as sirolimus) its derivatives everolimus and temsirolimus. These pharmacokinetic (PK) optimized derivatives of the natural product are FDA approved for the treatment of certain human solid tumors. (b) BEZ-235 is a dual mTOR/ PI3K inhibitor that reached phase II clinical trials. (c) Selective ATP-competitive mTOR inhibitors now in phase I clinical trials for the treatment of human solid tumors.

Of special note is the differing effects of rapamycin and ATP-competitive mTOR inhibitors on AKT phosphorylation. Acute inhibition of mTOR with rapamycin leads to a time dependent increase in phosphorylation of T308 phosphorylation on AKT in many cell lines, including colon (Figure 1.3a). This effect is due to a negative feedback loop involving S6K and the insulin receptor substrate (IRS) (Figure 1.3b) (61). Inhibition with an ATP-competitive 6 mTOR inhibitor such as PP242 abrogates this phenomenon in a dose-dependent fashion (Figure 1.3c). This data supports a theory that this S473 mTORC2 dependent phosphorylation stabilizes T308 phosphorylation and without active mTORC2, AKT cannot be in the active state (62). This may have clinical impact as this feedback loop was associated with poor response to rapamycin in a glioblastoma trial (63).

a b Rapamycin (20nM) R Time (hrs) 0 .25 .5 1 6 24 T HO O HO O K OH p-Akt T308 PTEN OH P P P P S I P T308 P P P P P P p-rpS6 (240/244) O R P S473 PI3Kα Akt P S S p85 β-actin mTORC2 P c TSC1 TSC2 Rapamycin PP242 00

p-Akt T308 P S6K mTORC1 p-Akt S473 p-rpS6 (240/244) P P P P P P β-actin S6 4E-BP1

Figure 1.3 mTORC1 inhibition induced feedback activation of Akt. (a) Rapamycin treatment induces a profound hyper-phosphorylation of Akt T308. HCT 15 cells were treated with 20 nM rapamycin for the indicated time, lysed and western blotted with anti-p-T308 antibodies. (b) A negative feedback loop exists between mTORC1 signaling and the insulin receptor substrate (IRS). (c) Inhibition of both mTORC1 and mTORC2 abrogates feedback induced Akt activation. The ATP-competitive inhibitor PP242 blocks phosphorylation of Akt completely 24 hours after initiation of treatment.

1.5 ATP-competitive inhibitors in the clinic and patient selection

ATP-competitive inhibitors designed to be anti-cancer agents have entered clinical trials for a variety of malignant neoplasms (64,65). Their advancement to human trials was based on pre-clinical data showing significant effects in both solid and hematologic malignancies. But as discussed earlier, mTOR is not mutated in cancer, and common upstream mutations are actually several mechanistic steps away from mTOR itself and for what tumors inhibition will be most successful remains an open question. This difference distinguishes ATP-competitive mTOR inhibitors from other clinical kinase inhibitors developed to treat cancer. Drugs such as imatinib, vemurafenib, and lapatinib have been developed specifically for patients with mutations in their target kinases and represent classic cases of what is now termed “oncogene addiction” (66-69). This model posits that mutated oncogenes alter the signaling pathways of the cell and that tumors become addicted to the aberrant growth for continued survival and proliferation. However, without a clear molecular signature, mTOR inhibitors are 7 being tested in a diverse patient population without pre-selection on genetic markers. As mTOR is such a central kinase in control of cell growth, changes to multiple inputs, independently or in concert, may increase dependence upon mTOR signaling. This challenges the paradigm of oncogene addiction and calls for new techniques that take into account the entire signaling network. ATP-competitive mTOR inhibitors have now been in clinical trials for over 4 years (Figure 1.2b-c). The first inhibitors to enter trials were dual-PI3K/mTOR inhibitors. BEZ-235 was the most potent of this first class of inhibitors and reached clinical trials first, but it has not advanced to FDA approval (70). To both lower off-targets effects and to better validate inhibition of mTOR as a target in human oncology, more selective kinase inhibitors were needed (71). This second class of more selective mTOR inhibitors without PI3K inhibitory activity are earlier in development and include the pyrazolopyrimidines MLN0128 and WYE-125132, the imidazo[1,5-a]pyrazine OSI-127 and the pyridopyrimidine AZD8055 (Figure 1.2b) (72-75). As of yet, no results from clinical trials of these second-generation more selective inhibitors have been published. Pre-clinical data from these inhibitors and preclinical tool compounds suggest various markers for efficacy. Apoptosis has been observed in multiple different hematological malignancies upon treatment with selective ATP-competitive inhibitors. OSI-027 treatment lead to apoptosis in acute lymphoblastic leukemia, mantle cell lymphoma and marginal zone lymphoma (74). PP242, the precursor to MLN0128 induced apoptosis in Philadelphia chromosome harboring acute leukemia cell lines as well as multiple myeloma (76,77). In many cases these effects were specific to malignant cell lines, but the general observation is that white blood derived cell lines may be especially sensitive to mTOR inhibition. Observations from solid tumor models is less impressive with regards to apoptosis, but growth arrest is profound and dependent upon upstream aberrations in activated oncogenes (72,78). These studies differed however in the role of the PI3K activator and tumor suppressor PTEN. In a PTEN null prostate mouse model, inhibition with MLN0128 was highly effective whereas in breast cancer cell lines, PTEN loss was not correlated with response to PP242. This contradiction highlights the context dependence of mutations in predicting sensitivity and that large data sets covering many different tissue types will be necessary to understand how upstream network changes will effect mTOR inhibition.

1.6 Summary and significance

The mTOR kinase functions within two key signaling complexes that are critical for regulating cell growth via control of macromolecule biosynthesis. The mTORC1 complex lies downstream of many of the most common mutations identified in cancer and as such has generated significant interest as a drug target. Existing approved small molecule inhibitors are based off an unusual natural product, rapamycin, that acts through a unique mechanism of action. This mechanism results in only incomplete inhibition and is insufficient for strong anti-tumor responses. New inhibitors of mTOR that are ATP-competitive have now been developed. They are superior to rapamycin in pre-clinical trials and have advanced to patient trials. To maximize chances that these drugs can be approved and useful for cancer, careful patient selection will be necessary. However, there is not an understanding of which patients may be sensitive or resistant to these agents. Identifying mutations in signaling pathways that are associated with resistance and sensitivity, as well as the mechanism underlying these effects, will be critical to successfully

8 developing ATP-competitive mTOR inhibitors. This knowledge will allow both the selection of patients likely to respond as well as the characterization of biomarkers to track response in individuals in real-time.

CHAPTER 2:

Cell line screening for the identification of markers of sensitivity and resistance to mTOR inhibitors

10 2.1 Abstract

The identification of select patient sub-populations that will benefit from new targeted therapies for cancer is essential to maximizing their therapeutic potential. The large degree of patient heterogeneity makes preclinical predictions of efficacy very difficult. While targeted drugs are often developed with specific mutations in mind, unexpected resistance and sensitivity can emerge from other signaling nodes. Large format cell screening can be of use in overcoming these obstacles. We screened representatives (PP242 and rapamycin) of 2 independent classes of inhibitor of the mammalian target of rapamycin (mTOR) against a greater than 650 cell set of human solid tumor cell lines. The conserved kinase mTOR is not itself mutated in cancer and it has yet to be validated what markers for sensitivity and resistance exist. We discovered significant differences in efficacy between the two inhibitors and little correlation in cellular response between the drugs. PP242 was more effective and we analyzed the effect of several mutations in predicting response to PP242. Screening a subset of colon cancer cells with an expanded set of PI3K/mTOR inhibitors revealed significant similarities in response to these agents and allowed for the clustering of cell lines with similar activating mutations validating their role in predicting efficacy to these agents.

2.2 Introduction-High throughput cell screening to identify resistance and sensitivity to kinase inhibitors

The discovery of oncogenes (genes that when hyperactive lead to a cancer phenotype) and tumor suppressors (genes the when deleted or inactivated lead to a cancer phenotype) has allowed for the development of small molecule and antibody drugs that target specific mutated proteins found in tumors and not normal tissue. The development of imatinib, a small molecule kinase inhibitor against the oncogenic fusion protein BCR-ABL, has most clearly demonstrated the utility of this advance, transforming a once fatal diagnosis of chronic mylogenous leukemia (CML) into a manageable chronic condition with greater than 90% survival (66). Drugs have been approved or are in development against dozens of oncogenes, but no combination of target and drug have shown to be as successful as BCR-ABL and imatinib (79,80). Unlike CML, a hematological malignancy defined by a single diagnostic mutation, solid tumors have complex etiologies and are characterized by dozens of non-overlapping mutations. While certain oncogenes are thought to be “driver mutations,” the impact of inhibiting the same mutant gene can vary widely from one tumor to the next (81). Coincident with this problem, the number of newly discovered frequently mutated kinases in cancer is decreasing, even as cancer genome sequencing accelerates, making it important to target non-mutated kinases that exist in critical common pathways (82,83). In response to this, interest has recently been renewed in targeting key cellular pathways that are deregulated in common among many cancers, chief among them the cellular growth and nutrient sensing pathway regulated by mammalian target of rapamycin (mTOR) (1). mTOR is a protein kinase conserved throughout eukaryotic life and is critical for integrating nutrient availability with growth factor signaling to make cell growth decisions (reviewed in chapter 1) (2). Many of the most common oncogenes are components of growth factor signaling and direct biological connections can be drawn between aberrant signaling in these proteins and mTOR. Upstream of mTOR lies the epidermal growth factor receptor (EGFR), and its intracellular partners, RAS and phosphoinositide 3-kinase catalytic subunit alpha

11 (PIK3CA). These three genes are commonly activated in solid tumors. RAS (the most mutated oncogene in solid tumors) activity is further mediated by bRAF, the most common mutated gene in melanoma. RAS and bRAF are thought to carry signals in a pathway parallel to mTOR and their activation may bypass mTOR but significant cross-talk exists. PIK3CA activity is opposed by the tumor suppressor phosphatase and tensin homolog (PTEN), and both types of mutations lead to directly to the activation of mTOR. With this biological framework, new small-molecule inhibitors of mTOR have been developed and have recently entered phase I clinical trials for the treatment of diverse solid tumors(58,60,84). Evidence from other agents that target proteins in the mTOR pathway and preclinical models of mTOR inhibitors suggests that there will be a large variety of responses, and successful approval of mTOR inhibitors may depend upon the identification of a selected patient population that will benefit from these agents. Identifying the characteristics of a tumor that will render it sensitive to a specific therapy is one of the greatest challenges in cancer drug discovery. Data driven methods such as SNP arrays, copy number arrays and complete genome sequencing provide the most complete analysis of a tumor’s variability, but their clinical application remains a long way off (85). Instead, clinically actionable criteria for patient selection are limited to either tissue of origin or mutations and/or expression changes in one or maybe two genes. Importantly, these can be both positive and negative markers for selection. EGFR inhibitors are indicated for patients with mutations in EGFR in the context of lung cancer, while in colorectal cancer, EGFR antibodies are excluded from patients with RAS mutations (86). The proceeding EGFR examples of selection criteria were discovered and validated in patients and experimental models after the drug had entered the clinic (87). To prospectively identify criteria to select patients for mTOR inhibitors is the goal of this analysis. The prospective identification of criteria predicting sensitivity and resistance to a specific therapy requires a large screening set that can capture the variability seen in actual patient populations. A suitably large panel of human cancer cell lines may contain enough diversity to predict sensitivity and resistance to a therapeutic agent. The research group of Jeffrey Settleman has collected a large (>700) set of solid tumor human cell lines and developed the robotic handling equipment necessary to screen these cell lines for growth inhibition against new drugs (88). This method was validated by testing several inhibitors and showing that the cell lines most potently inhibited contained the mutant oncogene targeted by the drug of interest (e.g. EGFR and the EGFR inhibitor lapatinib) (89). We sought to expand this method to the analysis of two representative mTOR inhibitors, the natural product rapamycin and the ATP-competitive inhibitor PP242. PP242 was a more effective agent and there was a stronger genetic signature or resistance and sensitivity. Rapamycin was a weak inhibitor and little trend was observed. Interestingly, there was almost no correlation between rapamycin and PP242 response, showing that the mechanism and completeness of mTOR inhibition strongly differentiate these drugs.

2.3 PP242 is more potent inhibitor of cell growth than rapamycin

PP242 and rapamycin were screened against a panel of human solid tumor cell lines. For PP242 there were 666 cell lines with data for three concentrations tested. PP242 was screened at 5000, 500 and 50 nM for 72 hrs (complete data available in appendix 1). For rapamycin, there were 707 cell lines with growth information at three concentration of drug: 10, 1 and 0.1 µM. Although data was collected for both drugs at all concentrations, the analysis was conducted at

12 the medium concentration for both drugs to maximize the ability to discover both resistance and sensitivity in the same analysis.

a b PP242 500 nM Rapamycin 50 nM 15 15 10 10 Percent Percent 5 5 0 0 0 .2 .4 .6 .8 1 1.2 1.4 .2 .4 .6 .8 1 1.2 1.4 Proportion of ctrl cell growth Proportion of ctrl growth

c d Rapamycin vs PP242 Growth 1.4 1.2 PP242 Rapamycin R2= 0.032 1 Obs 666 707 .8 Mean 0.5792 0.7348 .6

Std. Dev. 0.1902 0.182 .4 PP242 cell growth

Percentiles .2

1 0.148 0.294 0 .2 .4 .6 .8 1 1.2 1.4 1.6 5 0.255 0.447 Rapamycin cell growth 10 0.338 0.504

e 1.4 25 0.453 0.619 1.2 R2= 0.42 50 0.584 0.735 1 75 0.709 0.847 .8 90 0.819 0.966 .6 95 0.873 1.048 500 nM PP242 .4 99 1.024 1.179 .2 0

.4 .6 .8 1 1.2 1.4 50 nM PP242

Figure 2.1 Distribution of cell growth responses to treatment with mTOR inhibitors. (a) A histogram of the growth of 666 human solid tumor cell lines treated with the 500 nM of the ATP-competitive mTOR inhibitor PP242 for 72 hrs. (b) A histogram of the same cell line set treated with 1 µM rapamycin. (c) Summary statistics for PP242 and rapamycin datasets. (d) A scatterplot of PP242 cell growth vs rapamycin cell growth shows no correlation. (e) A scatterplot between two different concentrations of PP242 shows a moderate level of correlation.

Cell growth was scaled to a no treatment control, so that a value of 1 was equivalent to no treatment, and 0 would be the value if all cells were killed. Values of greater than 1 were

13 possible due to both measurement error and the chance that some cells could actually grow better with the drug, but those cases were rare with PP242, and more common with rapamycin (1.2% vs. 8.1%). The values of cell growth in response to 500 nM PP242 (PP242_med) were normally distributed about 0.579 with a standard deviation of 0.192 (Figure 2.1a and c). The Shapiro- Wilk test of normality was not significant for PP242_med (p = .150) and we could safely proceed with the normal curve assumption for the outcome data. Cell growth in response to rapamycin was less affected by the treatment and the distribution more skewed (Figure 2.1b and c). The distribution of cell growth was not normal (Shapiro-Wilk test, p= .045), and the average response was only 0.735 (±0.182). While we proceeded with our analysis, the interpretation of the rapamycin dataset must necessarily be done carefully. There was no correlation between cellular responses to rapamycin and PP242 (Figure 2.1d). A scatter plot of the cell growth effects of 1 µM rapamycin vs 500 nM PP242 gave no trend and the R2 for the linear regression was only 0.03. This lack of correlation highlights the significant differences in mechanism between these two inhibitors. For PP242, there was a strong correlation between the responses of cell lines to different concentrations of the drug (Figure 2.1e). A plot of 500 nM PP242 against 50 nM PP242 shows a clear relationship and the R2 for the linear regression was 0.41. We believe that this significant but still moderate level of correlation between doses of the same drug can be understood in terms of varying difference in drug sensitivity. Cell growth responses to changes in drug concentration are not linear but bounded by the design of the assay, the shape of the dose-response curve, and where on this curve for a specific cell line the tested drug concentration lay.

2.4 Colon and pancreas are mTOR inhibitor resistant organ types

The organ of origin is a critical variable in describing the outcome of drug treatment on a particular tumor or cell lines. Each organ has its own set of common mutations and inherent properties, and the organ type captures large amounts of information. A box plot of the response of each organ type to rapamycin and PP242 shows significant variance among types (Figure 2.2a and b). A one-way ANOVA was very significant for both rapamycin (F=5.36, p<0.0005) and PP242 (F=2.89, p<0.0005). Several organ types were significantly resistant or sensitive to PP242 and rapamycin treatment. Figure 2.2c lists the treatment average of each organ type as a standard deviation from the complete screen treatment mean and the p-value associated with the comparison of means (Student’s T-test). To correct for multiple comparisons, an adjusted p- value of 0.0023 was used for the significance level of α=0.05 using the conservative Bonferroni adjustment. The organ types that had the significant p-values for mTOR inhibitor response were colon and pancreas and both were resistant to inhibitor treatment. For rapamycin treatment, average cell growth for colon and pancreas cell lines were 0.863 and 0.662 standard deviations above the entire treatment set and both comparisons of meeting had a p-value of <0.0005. For PP242, colon cell lines were 0.588 standard deviations above the group mean and again this was very significant (p value <0.0005). Several cell types showed sensitivity to mTOR inhibitors but the magnitude of the effect was smaller than observed in the resistant cases with the exception of uterus and rapamycin (p value <0.0005).

14 ab Rapamycin by Organ PP242 by Organ 1.5 1.5 1 1 .5 .5 PP242 cell growth Rapamycin cell growth 0 0

Skin BoneBrain LiverLung Ovary LiverLung Skin BreastCervix Kidney Muscle TestesThyroidUterus BoneBrain Cervix Ovary Bladder Intestine ProstateStomach Bladder Breast Kidney Muscle TestesThyroidUterus Pancreas Intestine PancreasProstateStomach Esophagus Esophagus Head & Neck Lung:NSCLCMiscellaneous Head & Neck Lung:NSCLCMiscellaneouservous System Nervous System N

c Rapamycin PP242 Stdev mean p value Stdev mean p value Std. dev. mean bladder 0.241 0.262 0.031 0.884 x>0.5 bone 0.011 0.954 0.361 0.045 0.5>x>0.25 brain -0.480 0.005 0.317 0.066 0.25>x>0.15 breast 0.174 0.228 -0.067 0.636 -0.25

Figure 2.2 Distribution of cell growth responses by organ type. (a) A boxplot of cell growth distributions by organ type after treatment with rapamycin. (b) A boxplot of cell growth distributions by organ type after treatment with PP242. (c) Standard deviations (Stdev mean) and p values for each organ type. P values less than 0.0023 are highlighted as being significant at the level of α=0.05 using the Bonferroni adjustment for multiple comparisons.

15 2.5 PIK3CA and RAS mutations are markers for response to PP242 but not rapamycin

Upstream mutations in signaling pathways are thought to sensitize cells to mTOR inhibition. To analyze the relative contribution of common oncogenic mutations to the observed sensitivity and resistance to mTOR inhibitors, multiple regression using ordinary least squares (OLS) was used. For both mTOR treatments, the resulting models fit the data very poorly with R2 values of less than 0.1. For PP242 treatment, the model using all 15 gene indicator variables fit the complete 357 observation data set very poorly (Model 1, Table 2.1). The R2 was 0.073 and adjusted R2 0. 0323. The genetic data gathered for the cell lines only explained 7% of the variance in growth effect due to treatment with PP242. Analysis of the residuals from the model showed that the poor model performance was not due to errors in functional form or to extreme outliers (Figure 2.3a and b). The plot of standardized residuals versus fitted values showed no bias in residuals as fitted value increased and standardizes residuals were evenly distributed both above and below zero (Figure 2.3a). There were 4 outliers (std. dev. residual> 2.5, but that would be expected for a data set of this size. These outliers were all cell lines that were particularly insensitive to PP242, and one in particular, SW620, had a growth value of significantly greater than 1. A qnorm plot of the standardized residuals versus a normal distribution showed similar behavior with the only deviation from normality observed at the upper end of the fitted values, consistent with the known outlier data (Figure 2.3b). As would be expected with such a low R2, only two of the genes were significant, PIK3CA and RAS. For PIK3CA the coefficient was negative with a p-value .017 and for RAS it was positive with a p-value of 0.045. However, given 15 variables, it would be expected that one would be significant at a p-value of .05 just by chance. With that in mind, it is difficult to be confident in the significance of these results other that they suggest that these variables might be significant but it warrants further analysis to be definitive. To further test whether these variables might be significant, we constructed additional models using restricted genotype dataset information. We employed both a biased and unbiased approaches to reduce the number of genotype variables analyzed. A stepwise function employs an algorithm to eliminate variables that are not significant until a consistent set of significant variables remains. Using information about the signaling pathways in cancer, we tested the 4 genes whose mutations we believed would be most likely to affect mTOR (BRAF, RAS, PTEN and PIK3CA). The stepwise backwards elimination model was constructed using a generous significance cutoff of p=0.15 knowing that few of the genes had been significant in the full model. The model reduced the number of gene variables from 15 to 5 (Model 2, Table 2.1). These genes were RAS, PIK3CA, RB1, CTNNB and APC. The R2 was lower than in the full model, 0.0466, but the adjusted R2 was nearly the same, .033. In this model, 2 genes were again significant at the p=.05 level, but they were PIK3CA and RB1, and not RAS as in the full model. The coefficient on PIK3CA was again negative and very similar to the full model (-.079 restrict versus -.083 full) and the p-value decreased to 0.01. The newly significant gene RB1 had a positive coefficient of 0.08 and an associated p-value of 0.03. Finally, a model testing 4 known mTOR effector genes was constructed (Model 3, Table 2.1). The fit of the model using only BRAF, PTEN, RAS and PIK3CA was the worst of the three tested. The R2 was 0.028 and the adjusted R2 0.17. However, these differences were not statistically different from the full model. A F-test comparison of this restricted model 3 versus

16 0.0284 0.0174 0.19813 Model 3 Targeted Genotypes Targeted 0.0108303 -0.0474112 0.0690718 0.715 0.0643242 0.0149117 0.1137367 0.011 0.5443545 0.5149724 0.5737366 <.001 0.0432598 -0.0207021 0.1072218 0.184 -0.0536238 -0.1147824 0.0075347 0.086 0.033 0.0466 0.1965 Model 2 Stepwise (backwards) p=.15 0.544978 0.5191821 0.570774 <.001 0.0453988 -0.0037941 0.0945917 0.07 0.0823679 0.0073595 0.1573762 0.031 0.2 0.0654173 -0.0109643 0.1417988 0.093 0.0731 0.1966 0.0323 Model 1 Full Model 0.5331 0.48658 0.5796635 <.001 Coeff 95% CI (lo) 95% CI (hi) value P Coeff 95% CI (lo) 95% CI (hi) value P Coeff 95% CI (lo) 95% CI (hi) value P 0.0527757 0.0010839 0.1044676 0.045 0.0201867 -0.0259111 0.06628450.0603213 -0.0191611 0.148 0.1398037 0.136 0.0246088 -0.0362066 0.0854241 0.39 0.0295041 -0.0571149 0.116123 0.503 0.0457224 -0.0202706 0.1117154 0.174 -0.0644642 -0.2276261 0.0986977 0.438 -0.1728356 -0.4544219 0.1087506 0.228 -0.069989 -0.2491448 0.1091668 0.443 0.0511624 -0.0272317 0.1295565 Linear regression models for genotypes and PP242 Table 2.1 Table RAS Variables r2 adjusted r2 root MSE EFGR SMAD4Constant 0.0465895 -0.038948 0.1321269 0.285 CTNNBSTK11 0.1441375 -0.0173956 0.3056705 0.08 0.1361597 -0.0203472 0.2926666 0.088 P53 CDKN2ARB1 APC -0.0323916MAP2K4 -0.0763386VHL 0.0115554NF2 0.100252 -0.0602774 0.148 0.2607815 0.22 PIK3CABRAF -0.0787299 -0.143082 -0.0143778 0.017 -0.0832714 -0.1466242 -0.0199186 0.01 PTEN

17 0.0245 0.0133 0.17821 Model 6 Targeted Genotypes Targeted 0.0526423 0.00813080.0195384 0.09715380.0278358 -0.0355037 -0.0246012 0.0745806 0.0802728 0.021 0.486 0.297 0.7125323 0.685921 0.7391125 <.001 0.0646 0.0512 0.17476 Model 5 Stepwise (backwards) p=.15 0.712766 0.6910957 0.734464 <.001 0.1465582 0.0149063 0.2782102 0.029 0.7 0.91 0.208 0.233 -0.045194 -1.011029 0.0171490.0990.002 0.0488216 0.113 0.1124262 -0.0176804 -0.0372037 0.0459172 0.1153236 -0.0947661 0.1789351 0.0203586 0.15 0.001 0.205 0.1310115 0.0759475 0.0462572 0.0231369 0.1827975 0.2340518 0.0807 0.0401 0.17577 Model 4 Full Model 0.296831 -0.0165812 0.030989 -0.0234143 0.0853924 0.263 0.073882 -0.0862878 Coeff 95% CI (lo) 95% CI (hi) value P Coeff 95% CI (lo) 95% CI (hi) value P Coeff 95% CI (lo) 95% CI (hi) value P 0.1126814 0.0425652 0.0598614 -0.0112886 0.1352721 -0.00916240.0159212 0.2797067 -0.1299879 0.1618303 0.066 0.83 0.0246406 -0.0528162 0.1020975 0.532 -0.0358843 -0.0949056 -0.0050494 -0.0463503 0.0362516 0.81 -0.0722735 -0.3240307 0.1794836 0.573 Linear regression models for genotypes and rapamycin Table 2.2 Table Variables r2 adjusted r2 root MSE RAS PTEN BRAF P53 CDKN2ARB1 0.0012895 -0.0381112 0.0406903NF2 0.949 PIK3CA -0.0112927 -0.0688426 APC MAP2K4VHL 0.0700722 -0.07345STK11 0.2135944 0.338 EFGR CTNNB SMAD4Constant 0.0674764 -0.0090046 0.6990625 0.1439575 0.6573569 0.7407681 <.001 0.084 0.0679243 -0.0069137 0.1427623 0.075

18 a PP242 b Rapamycin 4 4 2 2 0 0 -2 -2 Standardized residuals Standardized residuals -4 -4

.4 .5 .6 .7 .8 .6 .7 .8 .9 Fitted values Fitted values

c d 4 4 2 2 0 0 Standardized residuals -2 -2 Standardized residuals -4 -4

-4 -2 0 2 4 -4 -2 0 2 4 Inverse Normal Inverse Normal

Figure 2.3 Residuals from linear regression models of genotype and mTOR inhibitor response. (a) Standardized residuals vs fitted values for the linear regression of PP242 against genotype. (b) A q-norm plot of the standardized residuals for PP242. (c) Standardized residuals vs fitted values for the linear regression of rapamycin against genotype. (d) A q-norm plot of the standardized residuals for rapamycin.

the complete model 1 was not significant (F= 1.49 p=0.1318). While the F test for the full model was significant (p=.0345), implying that a least one variable is better than chance alone, the restricted model describes the data as well (or equally poorly in the context) as the full model. Only the RAS variable was significant in model 3 with a coefficient of 0.06 and a p-value of .011. For rapamycin, there was even less of a relationship between genotype and cellular response to inhibitor. In the complete 15 gene regression, only one genotype was significant, APC (Model 4, Table 2.2). The R2 for the model was 0.081 and adjusted R2 0.04. APC mutations are highly coincident with colon cancer and so the fact that both colon and APC were resistance markers in rapamycin is not surprising. The lack of significance for APC mutations in PP242 treatment is more puzzling. The residuals were not as evenly distributed about zero as in the PP242 data set and a preponderance of them appeared to be negative, but the qnorm plot showed minimal departure from linearity (Figure 2.3c and d). Simplified models of the rapamycin genotype data did not uncover new genotypes that would be robust predictors of sensitivity and resistance. The stepwise backwards elimination of variables left 5 variables in the model, but only one additional variable was significant (CTNNB) (Model 5, Table 2.2). This variable was also significant in PP242 case in the stepwise model. 19 But there are only 7 cell lines mutant for this protein and it is also highly coincident with colon cancer. When the analysis was restricted to the targeted genotypes (RAS, PTEN, PIK3CA and BRAF) only RAS was significant as was the case with PP242 (Model 6, Table 2.2).

2.6 PIK3CA mutations are predictors of sensitivity to PP242 in multiple tumor types

Cell signaling pathways are tissue specific and the role of mutations in different tumor backgrounds may not be conserved. The impact of specific mutations can be dependent upon organ. We analyzed the role that mutations would have within tumor types in a subset of tumor types, focusing on those that were most represented as well as select cell lines with high numbers of mutations. Because the numbers were much smaller in these subsets, we did not perform multiple regressions to avoid problems with over fitting, but instead performed unpaired T-tests. In all, 5 genotypes were analyzed in 8 organ backgrounds using both the rapamycin and PP242 datasets. For no combination of organ type and genotype were the results significant with the rapamycin set. But for PP242, several associations were markers of sensitivity and resistance.

a NSCLC Breast Cervix

1.25 1.25 * 1.25 * * 1.00 1.00 1.00

PIK3CA 0.75 0.75 0.75 0.50 0.50 0.50 Cell Growth Cell Cell Growth Cell Cell Growth Cell

Cell growth 0.25 Cell growth 0.25 Cell growth 0.25

0.00 0.00 0.00 MUT WT MUT WT MUT WT

Colon 1.50 * b n RAS PIK3CA BRAF PTEN RB1 1.25 Breast (coef) 30 -0.217 0.396 1.00 p value 0.015 n.s. 0.003 Ras 0.75

Cell Growth Cell 0.50 Colon (coef) 23 0.177 Cell growth 0.25 p value 0.036 n.s. 0.00 MUT WT NSCLC (coef) 64 -0.211 Breast p value n.s. 0.015 n.s. ** Cervix (coef) 11 -0.365 1.25

1.00 p value 0.048

0.75 Uterus 8 n.s. Rb1 0.50 Pancreas 15 n.s. n.s. Cell Growth Cell Cell growth 0.25 Skin 50 n.s. n.s. n.s. 0.00 MUT WT Brain 22 n.s. n.s.

Figure 2.4 Effect of genotypes in PP242 sensitivity in select organ types (a) Cell growth effect of PP242 on mutant vs wildtype PIK3CA, RAS and RB1 in select organ types. All comparisons were made using unpaired Student’s T- tests. * p <0.05. (b) Coefficients and p values for comparisons in (a).

In 5 genotype organ combinations the presence of a mutation was significant for PP242 response (Figure 2.4a and b). The combinations are PIK3CA with breast, NSCLC and cervix, 20 RAS with colon, and RB1 with breast. As expected from the earlier regression data, PIK3CA mutations predict sensitivity to PP242 and RB1 and RAS resistance. For PIK3CA, the coefficient for a mutation was -0.22 for Breast cancer (p=0.015), -0.365 for cervix (p=0.048) and -0.21 for NSCLC (p=0.015). In colon cancer, the coefficient for RAS was large (0.18) and significant as well (p=.035). a mTOR Inhibitors PI3K Inhibitors Dual Inhibitors PP30 PP242 Rapa PIK 93 TGX 221 PIK 90 PP102 PI 103 PP121 BEZ 235

mTORC1/2 mTORC1/2 mTORC1 p110 , , , p110 p110 , , p110 , , , p110 , , PI3Ks p110 , , Cell Line VPS34 mTORC1/2 mTORC1/2 mTORC1/2 RTKs VPS34 HCT 116 >10 1.56 >10 >10 0.316 >10 5.6 0.979 0.032 HCT 15 >10 0.41 >10 >10 >10 0.329 >10 1.17 0.634 0.006 DLD-1 5.59 1.69 >10 >10 7.62 0.752 >10 1.28 1.67 0.005 HT-29 2.56 3.35 >10 2.39 0.111 0.692 0.382 0.008 SW620 >10 11 >10 >10 >10 0.937 >10 1.37 >10 0.096 SW1116 >10 15 >10 >10 >10 1.97 >10 1.31 >10 0.061 LS174T 7.14 0.835 >10 >10 >10 4.17 4.68 1.08 0.034 KM12 >10 2.57 >10 >10 >10 >10 >10 4.78 0.174 Hela >10 >10 >10 >10 >10 >10 >10 >10 >10 0.105 COLO 201 >10 0.696 >10 >10 >10 0.861 >10 >10 1.95 0.017 COLO 205 8.97 >10 6.35 0.272 0.694 >10 0.938 0.079 b PP30 PP242 PIK 90 PI 103 TGX 221 Rapamycin PP102 BA121 PIK 93 BEZ 235 Sorafenib SW13 SW620 KRAS Mut SW1116 HCT 116 KRAS Mut/ HCT 15 PI3K Mut DLD-1 COLO 201 LS174T MSI unstable KM12 HT-29 BRAF V600E COLO 205 HeLa

Figure 2.5 A panel of PI3K/mTOR inhibitors distinguishes colon cancer cell lines by genotype. (a) Inhibitory constant 50 (IC50) values for PI3K/mTOR inhibitors against colon cancer cell lines. Values were calculated from 6- point drug dilution series measure with resazurin. Chemical structures of each inhibitor are shown. (b) Clustering analysis of the data from panel (a). Cell lines were clustered by drug response. Mutational similarities among grouped cell lines are shown at right.

2.7 Cluster analysis reveals that PI3K/mTOR inhibitor segregate colon cancer cell lines by genotype 21

We next asked if a cell screen starting with multiple inhibitors of the mTOR pathway would be able to discriminate cell lines based upon genotype. This would validate our hypothesis that genotypes were important to understanding the efficacy of mTOR pathway inhibitors. This pilot cell screen was conducted in only a single tissue background, colon, and with 11 cell lines. Colon was chosen because it was particularly resistant to both mTOR inhibitors tested in the large screen and also harbored large numbers of mutations (90). 10 inhibitors of either mTOR, the upstream signaling kinase PI3K or dual inhibitors of both proteins were profiled against the cell lines in 6-point growth inhibition assays. IC50 (inhibitory constant 50%) values were calculated from fitted sigmoidal dose-response equations and reported in micromolar (Figure 2.5a). Unsurprisingly, the most promiscuous inhibitors led to the most severe growth defects. BEZ-235 inhibits all main isoforms of PI3K as well as mTOR and average IC50 values were less than 100 nM. The pan-PI3K kinase inhibitor PIK90 was also particularly effective across cell types. Two compounds stood out as potently inhibiting some cell lines while having little effect on others. PP102 is a pan PI3K inhibitor that was less potent against PI3Kα than PIK90, but inhibited two BRAF V600E cell lines at 600 nM (91). And PP242 was effective against many cell lines, but not Hela cells or SW620 and SW1116 cell lines. With these data in hand, we asked how the profiles of each cell line across the group of similar inhibitors would cluster. We preformed cluster analysis to determine the similarities between cell lines and visualized the results (Figure 2.5b). The analysis divided the cell lines into 5 groups with readily identifiable genotypes. The two closest cell lines were SW1116 and SW620, two PP242 resistance cell lines that are both mutant for KRAS and WT for PIK3CA, PTEN and BRAF. A group of three cell lines (HCT 15, HCT 116 and DLD-1) that are mutant for KRAS and PIK3CA clustered together as well. Two known microsatellite unstable (MSI unstable) cells grouped together as did the BRAF V600E containing cells. Finally, as a control, it was gratifying to see the Hela cells, which are not of colon origin, did not cluster with any colon cells. This analysis strongly supports the hypothesis that PI3K/ mTOR pathway inhibitors are sensitive to common mutations in colon cancer and these should be carefully considered in any trial testing their efficacy.

2.8 Discussion

We screened two different mTOR inhibitors to uncover determinants of resistance and sensitivity to this method of intervention. Rapamycin and PP242 act thought different modes of action and have been shown to have significant differences in vitro (60). These differences were very apparent in the large format cell screen experiments we conducted. Rapamycin was less effective than PP242, responses between the two drugs did not correlate, and markers of resistance and sensitivity were not conserved. These significant differences add to the developing understanding that ATP-competitive mTOR inhibitors are fundamentally different from natural product analogs of rapamycin, and their biologic effects are likely to be profoundly different as well. The cell screen analysis uncovered two organ types that were strongly resistance to mTOR inhibition by both inhibitors tested. Colon and pancreas were markers of resistance. Interestingly, both tumor types are strongly associated with mutations in KRAS, and yet RAS mutation is not a predictor of resistance in the rapamycin dataset, and only weakly predictive in PP242. This can be explained by the observation that RAS appeared to only be a marker of

22 resistance in select cell types. In non-small cell long cancer (NSCLC), the most represented cell type, RAS mutation was not predictive of resistance to PP242 or rapamycin. Brain and uterus were sensitive to rapamycin, but not PP242. Multiple regression by OLS was used in three sets of models to uncover significant markers of resistance and sensitivity by genotype. The first model examined all genotypes and subsequent models focused on a narrow set. The differences between rapamycin and PP242 were profound. The single largest marker for sensitivity and resistance to PP242 (PIK3CA and RAS respectively) were not significant in the rapamycin dataset. In fact, the only significant gene in the rapamycin dataset (APC) was a marker for colon cancer. Unfortunately, even for the PP242 datasets, the significance for any one genotype was very low. Among these common genotypes, we did not uncover a very strong marker, and these results do not appear very robust absent more data. Within specific organ types, the significance of genetic variables varied. For tumors with PIK3CA mutations, which in the entire data set predicted sensitivity to PP242, sensitivity was observed in 3 organ types examined (NSCLC and breast and cervix). For RAS mutations, resistance to PP242 was seen in 1 organ type examined (colon). This suggests that the effect of certain mutations in the whole data set was actually driven by their effect in specific organ subsets, but without greater enrichment for either the mutation or organ of interest, we were unable to address this further. All inhibitors have off-target effects and the validation of mechanisms of action requires multiple inhibitors of the same pathway to identify the common effects that are presumably due to on-target activities. To control for this effect, and to understand the sensitivity of cell lines to groups of common inhibitors, a group of colon cancer cells was treated with a panel of PI3K/mTOR inhibitors. From our cell screen results, their appeared to be a weak but potentially significant correlation between PIK3CA and RAS mutation and response. To further validate this, we sought to determine if cellular responses to inhibitors of these pathways depended upon mutations in these genes. We observed through clustering analysis that cell with common signaling mutations in these pathways responded to the inhibitors in common ways strongly suggesting to us that these mutations are important to understanding the efficacy of mTOR inhibitors. The goals of the analysis were to determine if any common gene mutation or organ type are associated with sensitivity or resistance to PP242 treatment. Furthermore, in the organ types that are more represented, we examined whether there are genetic effects specific to that background. The data are not very dense and do not show an obvious genetic signature for either resistance or sensitivity. However, there is evidence that RAS and PIK3CA mutations can alter sensitivity, and that the effects of PP242 are dependent upon organ type. Inclusion of significantly more descriptive data (including gene expression data and SNP arrays) about the specific cell lines tested in the assay may allow for the creation of a more predictive model for response to mTOR inhibition (88).

2.9 Data and Methods

Cell growth screen. For the high throughput growth screen, cell lines were grown and drug treated as previously described (89). Rapamycin was obtained from commercial sources. PP242 was synthesized according to previously published procedures as detailed in chapter 3 (92).

23 For the subsequent colon cancer cell line panel, cells were purchased from the ATCC (Manassas, VA). PI3K and mTOR pathway inhibitors were synthesized according to published protocols (PP30, PP242, PP102, PI 103, PP121) (92) or obtained from commercial suppliers (Rapamycin, PIK93, TGX221 BEZ 235). Cell growth was assessed in a 96-well plate format using a resazurin assay 48 to 96 hours after initiation of treatment with inhibitor in 0.1% DMSO final concentration. Drugs were diluted in a 6-point series 3-fold from 10 µM.

Analysis and statistics. Each cell line was coded for one of 22 common organ types or miscellaneous. The distribution of cell lines among organs was highly unequal reflecting both the unequal distribution of cancer but also the biases of the cell line set analyzed. A subset of the cell lines (n=357 for PP242, n=355 for Rapamycin) were annotated for common mutations in the Sanger COSMIC database (http://cancer.sanger.ac.uk/cancergenome/projects/cosmic/). For each observation, mutations in 15 common genes from the Sanger 50 set were recorded as indicator variables (1 for mutant, 0 for wild-type). The frequency of these mutations in the data set ranged from 2 (VHL) to 220 (P53). Overall, 733 mutations were recorded for the PP242 data set and each cell line averaged 2.05 mutations. 23 cell lines contained no mutations in the 15 genes analyzed, and 5 cell lines contained mutations in 5 genes. All statistical comparisons were conducted using the software suite Stata release 12 (Statacorp LLC, College Station, TX). Analysis was based on the treatment of growth as a continuous variable. This is in contrast to previous work by the Settleman group which converted the cell growth data from a continuous output to a categorical response. Cell lines were categorized as high, medium, low or non-responders. For agents that targeted specific mutant genes, the pattern of responses was a few responders (high, medium and low) while the majority of cell lines were non-responsive. From this highly skewed distribution of outcomes, simple comparison tests were preformed to assess whether a genotype or tissue of origin were more highly represented among the responders compared to the non-responders. To address the continuous growth data in this analysis, multiple regression using ordinary least squares was used. Clustering analysis was conducted separately using the open source programs Cluster and visualized with TreeView (http://www.eisenlab.org/eisen/?page_id=42).

CHAPTER 3:

Patient-derived xenografts in preclinical drug development

25 3.1 Abstract

Patient-derived xenografts (PDX) are a model of human cancer that is able to recapitulate key features of tumors better than standard cell line models. Sets of these xenografts allow access to groups of human tumors in an organismal setting without having to go through the adaptation process to plastic tissue culture. The use of this model may be particularly useful in preclinical drug development. PDX models allow for the drug to be assayed for both tumor intrinsic and extrinsic mechanisms of action in the same model. Furthermore, the accessibly heterogeneity of different patients allows for class effects to be separated from model specific results and provide a more robust platform than cell lines. We characterize the histological and cell-signaling response of a set colorectal cancer patient derived xenografts to treatment with the ATP- competitive mTOR inhibitor PP242. The xenografts faithfully maintained the patient specific tumor architecture over many subsequent murine passages. Treatment with PP242 resulted in changes to cell signaling and repressed apoptosis but did not alter tumor architecture.

3.2 Introduction- Xenografts in colon cancer

Direct xenotransplantation of human tumors into immunocompromised mice allows for the maintenance of these cancers in vivo without undergoing the irreversible changes that occur upon in vitro culture (93). This strategy seeks to avoid many of the common problems encountered in the standard cell lines and cell line xenograft models typically used in preclinical testing which are poorly predicative of clinical response (94,95). The use of a patient-derived xenograft model for preclinical drug testing has been most thoroughly developed in primary pancreatic cancer (96,97). Here, it was shown that the xenografts were stable over multiple murine passages with respect to many different tumor characteristics including gene expression, protein expression, and drug sensitivity. In a separate small study of colorectal cancer, the treatment response of xenografts generated from primary surgical resections closely matched the actual patient response to the same agents (98). A large-scale colon cancer derived xenograft study validated the important clinical predictive features of this model by showing that KRAS mutant tumors were resistant to the anti-EGFR antibody cetuximab (87). Colorectal cancer is the fourth most common cancer diagnosis and the second most common cause of cancer death in the United States. Patients who present with local disease can be successfully treated with surgery and adjuvant chemotherapy but for those with distant metastases, prognosis is poor with a 5 year median survival rate of 11% ((99). Treatment options for metastatic disease are limited and significantly improved therapies are needed. However, for a subset of patients with metastatic disease confined to the liver, surgical resection of these tumors is possible and outcomes are dramatically improved (40% median 5 year survival) (100,101). These surgeries provide rare direct access to metastatic tumor; our goal was that such tumors could be used to produce xenografts in nude mice to generate a novel preclinical model in which to study the biology of advanced colorectal cancer.

3.3 Establishment of human metastatic colorectal cancers xenografts

Colorectal cancer liver metastases were surgically resected as part of standard treatment with curative intent. Portions of these metastases were directly implanted into female nude athymic mice and subsequently further passaged in mice (Figure 3.1a). The patients whose

26 a

b CR708

Patient P1 P5 P10 CR703

Patient P1 P3 P5 c d Patient-derived xenograft doubling times

30 CR 702 CR 698 25 CR 727 20 15 CS174T 10

5 Doubling time (days)

0 0 1 2 3 4 5 6 7 8 Passage #

DLD-1 e

DAPI Mouse Human Merge

Figure 3.1 Patient-derived xenografts maintain key morphological features of human colon cancer. (a) Schematic of patient-derived xenograft generation from surgical specimens (b) Sequential passage of human tumors in mice 27 maintains tumor architecture. The first panel in each row is an H&E stain of a patient surgical specimen from a liver resection and the later panels are subsequent passages of the same tumor in nude mice. Patient characteristics are presented in Table 1. Scale bar represents 50 µm. (c) Doubling times of select xenografts across multiple patients. Tumors were measured approximately every week by external calipers and a volume was calculated assuming an ellipsoid form. Doubling time was estimated by fitting an exponential growth function to the measured values. (d) Tumors derived from colorectal cell lines do not resemble resected metastases. CS174T (top) and DLD-1 (bottom) were subcutaneously injected and allowed to develop into tumors that are homogeneous and have little stromal component. (e) Following xenotransplantation, tumor architecture is maintained with murine-derived stroma. Genomic FISH analysis was performed using species-specific labeled probes for the COT-I gene to identify the origin of each cell. Labeling strictly segregates the contribution of each species and pathology confirms the stroma is uniquely labeled with the murine probe. Sample shown was from patient CR703 and is passage 5.

tumors were used in this xenotransplantation study presented with advanced (stage III or IV) disease at diagnosis and were treated with surgical resection of the primary tumor followed by systemic chemotherapy (predominantly 5-fluorouracil and oxaliplatin based regimens) (Table 3.1). Liver metastases were removed either when prior treatment made it possible to do, or soon upon relapse.

Table 3.1 Genotypes and patient characteristics of tumors used in this study

Patient CR 698 CR 702 CR 727 CR 736 CR 739 CR 700 CR 703 CR 708 Sex M F M M F M F F Age 66 55 72 59 56 40 33 73 Stage at IV IV III IV III III IV II DX Prior FOLFOX FOLFOX FOLFOX FOLFOX FOLFOX FOLFOX FOLFOX none therapy bev bev bev NED 3 TTP 6 TTP 2 TTP 9 TTP 14 NED 2 No F/U No F/U Outcome yrs mos mos mos mos yrs PIK3CA WT WT H1047R H1047L WT WT WT WT KRAS WT G12D G12V G13D WT G12D G12D WT BRAF WT WT WT WT WT WT WT WT p53 WT R248L L130P exon 6 del WT R175H WT R273C

Prior treatment before surgical resection of liver metastases is abbreviated as follows: FOLFOX: (5-fluorouracil/ leucovorin/ oxaliplatin), Bev: bevacizumab. Outcomes are defined as: NED: No evidence of disease; TTP: time to progression; No F/U: no follow-up at UCSF.

To establish whether the passaged xenografts retained the characteristics of hepatic colorectal metastases, histopathological analysis was performed. Original patient tumor samples removed during surgery were compared with tumors taken from each subsequent murine passage and compared to determine if tumors had undergone gross changes in mice or retained their original pathological features (Figure 3.1b). Comparisons of patient surgical samples to later xenografts of the same tumor were made by a pathologist. Xenograft tumors were pathologically nearly identical to their original patient source material. For example, patient tumor CR 703 was distinguished by containing significant necrotic areas, a high degree of extracellular mucin and

28 having pleomorphic nuclei. These same features were noted in similar ratios in the passages P1, P3 and P5 of the same tumor (Figure 3.1b-lower panels). To assess whether xenograft tumors were accumulating changes in gross growth characteristics, tumor-doubling times were calculated (Figure 3.1c). Within tumor lines, average tumor doubling times were roughly consistent and no clear trend of increasing growth rates was observed. In contrast to the primary xenografts, similar implantations of cultured tumor cell lines generated a tumor that was homogeneous and defined by solid growth of epithelial sheets with little stromal network (Figure 3.1d). We observed that the tumors had maintained their overall architecture through multiple murine passages and that stromal cells were consistently supporting the structure. To determine whether murine stromal cells had been recruited to the tumors in order to support its stromal networks through multiple passages and subsequent outgrowths, genomic fluorescence in-situ hybridization (FISH) was performed. Using species-specific probes against the human and murine COT-1 gene, the contribution of each species to the tumor could be ascertained (Figure 3.1e). FISH staining of a P5 xenograft showed unambiguously that while the epithelial cells were human in origin, the stroma in the xenograft tumors was entirely murine in origin.

3.4 Synthesis of ATP-competitive mTOR inhibitor PP242

The mTOR inhibitor PP242 is representative of a new class of ATP-competitive inhibitors that have shown greater efficacy in preclinical trials than allosteric inhibitors that are derivatives of the natural product rapamycin. It remains unknown for which cancer patients these inhibitors will be most effective for. To identify early biomarkers that may be indicative of response to PP242 in colorectal cancer xenografts, gram quantities of PP242 were synthesized for dosing in mice at 100 mg/ kg according to previous published protocols (60). PP242 and its clinical derivative MLN0128 were synthesized according to the scheme outlined in Figure 3.2 following published procedures (72,92). Oral administration of PP242 was effective at reaching xenografts and inhibiting phosphorylation of mTOR substrates within 2 hours of dosing. Quantitative mass spectrometry was performed on serum samples to determine the actual concentration of PP242 achieved in vivo. Two hours after administration, the PP242 serum concentration was 5.8 µM, a very comparable number to efficacious concentrations in cell culture.

3.5 Oral dosing of PP242 is effective at inhibiting mTOR phosphorylation in vivo

To investigate how different patient backgrounds would affect PP242’s ability to inhibit phosphorylation of downstream mTOR targets, multiple patient tumors were expanded into cohorts of nude mice. Of particular interest were the differences in mutations to common colorectal oncogenes, KRAS and PIK3CA. Six different patient-derived tumors representing three different combinations of mutant PIK3CA (p110α) and KRAS were analyzed: WT for KRAS and WT for PIK3CA (CR 698 and CR 739); Mut for KRAS and WT for PIK3CA (CR 702 and CR 649); Mut for KRAS and Mut for PIK3CA (CR 727 and CR 736). Mice were sub- cutaneous implanted with two tumors, one on each flank, and administered drug at 100 mg/kg orally once daily for three days. Four hours after the final dose, animals were euthanized and tumors harvested for western blot analysis. PP242 effectively inhibited phosphorylation of both mTORC1 and C2 substrates (Figure 3.3a). mTORC2 substrate AKT S473 was inhibited

29 significantly in all tumors compared to vehicle controls. The phosphorylation of targets of mTORC1, 4E-BP1 and rpS6 was inhibited without any change in protein abundance Figure 3.2 O NH2 I CN o O 180 C N DMF N N ++N I-N N O/N 80 oC O/N H N N H NH2 N N N N 2 H H H O 1 2

NH2 I NH2 I Tetrakis(Pd), K t-butoxide OMe Na2OH N DMF N EtOH, DMF N + I N + (HO)2B o N N 0 RT O/N N N N 90 C 4 hr H Boc

2 3

OMe OH

NH2 NH BBr3, DCM NH2 NH o N -80 C RT N N N N N N N

4 5 (PP242) NH2 O N

Pd(OAc)2, NH2 I Triphenylphosphine, NH2 (HO) B Na OH, EtOH, DMF N 2 N 2 N N + NH2 N o N N O 90 C 2 hr N N

3 6 7 (MLN0128) Figure 3.2 Synthetic scheme for PP242 and MLN0128

The variance in cell signaling between tumors was much higher between mice than within the same mouse as seen by the comparison of changes between left and right tumors from the same mouse versus the same tumor in different mice. We analyzed the overall effect of

30 PP242 in all six tumors by quantifying the phosphorylated forms of rpS6 (S240/44), AKT (S473) and 4EBP1 (T37/46) in each treated tumor compared to the vehicle control (Figure 3.3b). PP242 consistently inhibited phosphorylation of all measured mTOR substrates. At this early treatment time point, the inhibition of mTOR substrates in the three genetic backgrounds was consistent and no correlation existed with either KRAS or PIK3CA mutation status.

a Patient CR 698 CR 702 CR 727 Vehicle PP242 Vehicle PP242 Vehicle PP242

L R L R L R L R L R L R L R L R L R L R L R L R L R L R L R L R L R L R p-Akt S473 p-rpS6 S240/244 rpS6 p-4E-BP1 T37/46

4E-BP1

cleaved PARP β-Actin

CR 739 CR 700 CR 736 Vehicle PP242 Vehicle PP242 Vehicle PP242

L R L R L R L R L R L R L R L R L R L R L R L R L R L R L R L R L R L R p-Akt S473 p-rpS6 S240/244 rpS6 p-4E-BP1 T37/46

4E-BP1

cleaved PARP β-Actin

b c mTOR substrate phosphorylation Cleaved PARP 2.0 0.9 Vehicle KRAS PIK3CA 0.8 PP242 0.7 CR 698 WT WT 1.5 0.6 CR 739 WT WT CR 702 0.5 G12D WT 1.0 0.4 CR 700 G12D WT 0.3 CR 736 0.2 G13D H1047L 0.5 Treated/Ctrl CR 727 0.1 G12V H1047R 0.0 0.0 p-4EBP1 p-Akt S473 p-rpS6 Fluorescence Intensity (normalized to β - actin)

CR 698 CR 739 CR 702 CR 700 CR 736 CR 727 Figure 3.3 PP242 inhibits mTOR outputs in different colorectal cancer genetic backgrounds. (a) Western blots of six different patient tumors are shown. For each patient tumor, 6 mice were implanted with two tumors each (left (L) and right (R) flank) and then randomized to receive either PP242 (100 mg/kg) or vehicle (three mice each) for three days before being sacrificed. (b) Quantification of mTOR signaling outputs in patient tumors from different KRAS/ PIK3CA genetic backgrounds. Western blots (representative samples shown in Figure 2D) were quantified using fluorescent secondary antibodies and the ratio of treated phosphoprotein to control was plotted for each tumor. (c) Quantification of leaved PARP in PP242 treated tumors. Western blots for cleaved PARP were conducted and imaged as in (b).

We analyzed tumors for drug-induced changes in apoptosis using Poly ADP ribose polymerase (PARP) cleavage as a surrogate. Basal levels of cleaved PARP were patient tumor specific and detectable in most tumors (Figure 3.3c). Upon 3 days of PP242 treatment, 31 significant changes in PARP cleavage were only observed in one tumor, CR700. Higher PARP cleavage in this KRAS mutant/ PIK3CA WT tumor suggest that PP242 may be efficacious in this tumor despite no major differences in mTOR signaling compared to similar tumors. Long-term drug treatment with PP242 did result in significant differences in tumor growth during a drug treatment trial (See chapter 4). Compared to the significant differences observed in tumor growth between the treated and untreated samples, alterations in gross tumor pathology were minor. Only in tumor CR 702 were significant differences noted between treated and untreated tumors. PP242 treatment induced cystic necrosis in the tumor as compared to vehicle controls (Figure 3.4). Surprisingly, the growth inhibited tumors CR 727 and CR 698 (834-3veh, 834-14 PP) showed no identifiable pathological differences. In the trials conducted, short term treatment with PP242 did not result in identifiable pathological changes while long- term exposure led to increases in necrosis that were not accompanied by a reduction in tumor growth rate.

VEHICLE PP242 CR698 CR702 CR727

Figure 3.4 Tumor morphology of PP242 treated tumors. H&E stains of paired treated and untreated PDX tumors. Tumors were treated for a minimum of 21 days once daily.

32 3.6 Discussion

Metastatic cancer is a multi-faceted disease that has remained extremely recalcitrant to successful intervention despite decades of research. The reasons for this are likely extremely varied and based on the very defining features of the disease itself (102). One specific reason for the failure of treatments to transition to the clinic is the poor state of many preclinical models of cancer. Human cancer is a disease that likely takes decades to develop and thus implicates a large number of signaling pathways and many individual mutations. Within every single tumor is substantial heterogeneity that can confound both molecular diagnosis and treatment (103). Whereas cell line xenografts and genetically engineered mouse models are both unnatural presentations of cancer without the diversity in human patients, PDX models may be able to more faithfully represent the human disease (104). We developed and characterized a PDX model of human colon cancer derived from metastatic disease. This technique was originally developed over thirty years ago, but was not used in drug discovery at the time (105,106). Our model faithfully recapitulates key morphological characteristics of the original tumor across many passages in mice. The Bardelli group has also initiated work in this model of metastatic colon cancer and has obtained very similar results (87). Of particular interest to ourselves was how faithfully this model recreates the stromal and elements that make up the tissue microenvironment. While these elements are of course murine, their recruitment to make functional tissue adds an important element to this model of cancer. To test our model in the context of preclinical drug development, we synthesized the mTOR inhibitor PP242 and administered it for three days before harvesting tumors and examining them for inhibition. We were pleased to see consistent inhibition as assayed by western blot of known mTORC1 and mTORC2 targets across many individual patient tumors. We observed some variability in response to this early dose, but none that could be associated with the patient histories or tumor genotypes. While incomplete, this early drug data did validate that mTOR in these xenograft tumors was being inhibited by oral dosing of the drug. By expanding the use of this model, we hope to be able characterize a broader spectrum of primary tumor resistance and sensitivity to a specific agent. In chapter 4, we report upon tumor growth inhibition trials using PP242 in the same PDX model and characterize primary resistance to this inhibitor. This model is ideal for the discovery of secondary resistance as well. The heterogeneity inherent in the tumors should allow for the rapid discovery of resistant clones to any specific therapy and PDX may be uniquely suited to this task. Discovery of mechanisms of resistance before they arise in patients should allow for the development of strategies to combat these simultaneous to the application of the new drugs themselves.

3.7 Materials and Methods

Patient-derived xenografts. The research protocol was approved by the Committee on Human Research of the University of California, San Francisco and patient consent was obtained. All animal studies followed a protocol approved by the University of California, San Francisco Animal Care and Use Committee. Colorectal cancer liver metastases were resected as part of standard treatment with curative intent. To establish xenografts, excess tumor removed during hepatic resection was minced under sterile conditions to generate pieces approximately 4-8 mm3. These were dipped into

33 sterile matrigel (BD Biosciences, Sparks, MD) and implanted sub-cutaneously (SC) onto the flanks of female athymic mice (FOXN1 nude, Harlan, Indianapolis, IN). Non-implanted pieces were flash-frozen in liquid nitrogen and banked. For each tumor, 2-4 pieces were implanted into 2-3 mice to establish the initial xenograft passage. When tumor volume reached ~1,000 mm3, mice were sacrificed; tumors were divided and implanted SC into new animals.

Histologic analysis of xenograft tumors. 4 µm sections prepared from FFPE tissue were stained with hematoxylin & eosin (H & E) and submitted for review by the pathologist.

Fluorescence In Situ Hybridization (FISH). Tissue sections were assayed for the presence of mouse and human cells by using FISH with species-specific genomic probes (107). Briefly, 5 µm sections were deparaffinized in xylene, dehydrated in ethanol, treated with NaSCN, pepsin, and HCl, and dehydrated in 70%, 85%, and 100% ethanol in series. Sections were denatured and hybridized with probes against CY3-d-CTP (GE Life Sciences, Piscataway NJ) labeled mouse COT-1 DNA (Invitrogen, Carlsbad CA), or CY5-d-CTP (GE) labeled human COT-1 DNA (Invitrogen). Probe sizes ranged between 300 and 2,000 bp.

Chemical Synthesis. PP242 was prepared according to the scheme shown in figure 3.2 derived from previously published procedure (92). MLN0128 was synthesized from similar starting materials as shown in scheme 1 and outlined below. 1H NMR was recorded on a Varian Innova 400 MHz spectrometer. 1H Chemical shifts are reported as s (singlet), d (doublet), t (triplet), q (quartet) or m (multiplet). Low resolution electrospray ionization LC/MS (ESI-MS) was recorded on a Waters Micromass ZQ equipped with a Waters 2695 Separations module using an XTerra MS C18 3.6 µm column (Waters). Synthesis of 5-(4-amino-1-isopropyl-1H-pyrazolo[3,4-d]pyrimidin-3-yl)benzo[d]oxazol- 2-amine (7, MLN0128). To a solution of consisting of 10 mL DMF, 1.25 mL EtOH and 1.25 mL H20, degassed with argon in a 100 mL round bottom flask, 3-iodo-1-isopropyl-1H-pyrazolo[3,4- d]pyrimidin-4-amine (3, 1.0 g, 3.3 mmol), (2-aminobenzo[d]oxazol-5-yl)boronic acid (6, 0.847 g, 3.96 mmol), Pd(OAc)2 (225 mg, 1 mmol), PPh3 (0.52g, 1.98 mmol) and Na2Co3 (1,78 g, 16.83 mmol) were added with stirring. The mixture was degassed with argon, and heated to 90 for 2 hrs (Suzuki reaction). The reaction was acidified to pH 0 with HCl, filtered, neutralized and extracted 3 times with 200 mL of 5% MeOH in EtOAc. The resulting organic phase was dried, concentrated and purified over an 80g silica column on a 5-10% MeOH gradient in DCM over 35 minutes. Product was analyzed by 1H NMR and LC/MS (456 mg, 45% yield). 1H NMR (DMSO-δ6) δ 8.227 (s, 1H), 7.524 (br, 4H), 7.463 (d, J=8Hz, 1H), 7.406 (d J=1.6Hz, 1H), δ 7.237 (dd, J=8.4Hz, J=2Hz, 1H), 5.053 (m, 1H), 1.490 (d, 6H). ESI-MS (M+H)+ m/z calc. 310.14, found 310.26.

Drug dosing. For pharmacodynamic experiments, mice bearing bilateral tumors of approx. 200 mm3 were given three doses of PP242 or vehicle and sacrificed 4 hours after the final dose. PP242 was prepared as a 25 mg/mL suspension in 5% 1-methyl-2 pyrrolidone (NMP), 80% polyvinyl pyrrolidone (PVP) and 15% H2O for pharmacodynamic studies and then in 3.1% NMP, 81.6% PVP, 15.3% H2O for tumor regression studies to reduce toxicity. 100 µL of PP242 suspension (2.5 mg/ dose equal to 100mg/kg) or vehicle alone was given orally once daily.

34 Western Blotting. Frozen xenografts harvested from mice were homogenized then lysed in radio-immunoprecipitation assay buffer (RIPA); lysates were normalized for protein content using a Bradford assay (absorbance at 595 nm), resolved by SDS-PAGE, transferred to nitrocellulose and blotted. Phosphorylation-specific antibodies were purchased from Cell Signaling Technology (Danvers, MA). Quantitative western blotting was accomplished using fluorescent secondary antibodies (800 nm emission) for visualization using an Odyssey IR scanner from LI-COR Biosciences (Lincoln, NE).

Chapter 4:

Incomplete inhibition of phosphorylation of 4E-BP1 as a mechanism of primary resistance to ATP-competitive mTOR inhibitors*

*This chapter was adapted from “Ducker GS, Atreya CE, Simko JP, Hom YK, Matli MR, Benes CH, et al. Incomplete inhibition of phosphorylation of 4E-BP1 as a mechanism of primary resistance to ATP-competitive mTOR inhibitors. Oncogene. 2013. doi:10.1038/onc.2013.92

36 4.1 Abstract

The mammalian target of rapamycin (mTOR) regulates cell growth by integrating nutrient and growth factor signaling and is strongly implicated in cancer. But mTOR is not an oncogene, and which tumors will be resistant or sensitive to new adenosine triphosphate (ATP)- competitive mTOR inhibitors now in clinical trials remains unknown. We screened a panel of over 600 human cancer cell lines to identify markers of resistance and sensitivity to the mTOR inhibitor PP242. RAS and PIK3CA mutations were the most significant genetic markers for resistance and sensitivity to PP242, respectively; colon origin was the most significant marker for resistance based on tissue type. Among colon cancer cell lines, those with KRAS mutations were most resistant to PP242, while those without KRAS mutations most sensitive. Surprisingly, cell lines with co-mutation of PIK3CA and KRAS had intermediate sensitivity. Immunoblot analysis of the signaling targets downstream of mTOR revealed that the degree of cellular growth inhibition induced by PP242 was correlated with inhibition of phosphorylation of the translational repressor 4E-BP1, but not ribosomal protein S6. In a tumor growth inhibition trial of PP242 in patient-derived colon cancer xenografts, resistance to PP242 induced inhibition of 4E-BP1 phosphorylation and xenograft growth was again observed in KRAS mutant tumors without PIK3CA co-mutation, compared to KRAS WT controls. We show that, in the absence of PIK3CA co-mutation, KRAS mutations are associated with resistance to PP242 and that this is specifically linked to changes in the level of phosphorylation of 4E-BP1.

4.2 Introduction

Clinically approved kinase inhibitors such as imatinib, vemurafenib, and crizotinib show strong anti-tumor responses in patients with mutated forms of their target kinases, BCR-ABL, BRAF V600E, and EML4-ALK, respectively (66,67,108). The intrinsic sensitivity of cancer cells expressing these mutationally activated kinase alleles provides a template for patient selection and clinical trial design (109). However, many of the protein kinase targets currently being investigated in cancer such as MEK, ERK, AKT, and mTOR are not commonly mutated, but rather lie at critical nodes in conserved cancer signaling pathways. The design of clinical trials for experimental therapeutics that inhibit these targets is challenging as mutations upstream of these nodes may or may not predict sensitivity to inhibition of downstream kinases (110). Deregulated mammalian target of rapamycin (mTOR) signaling is present in human diseases that alter metabolism, including diabetes and cancer (2). An essential and evolutionarily conserved regulator of cell metabolism, mTOR is the catalytic core of two related heteromeric protein complexes, mTORC1 and mTORC2 (5,6,9). In cancer, conserved mTOR mutations or gene amplifications have not been identified; instead, mTORC1 is activated by mutations in upstream signaling networks (39,41). The network most implicated in oncogenic mTORC1 signaling is the phosphatidylinositol 3 kinase (PI3K)/AKT/TSC (tuberous sclerosis complex) pathway (111). Enhanced response to mTOR inhibition in patients with rare somatic tuberous sclerosis mutations supports the rationale of targeting this network (45). Temsirolimus and everolimus, derivatives of the natural product rapamycin, are the only mTOR inhibitors currently approved for the treatment of solid tumors but their activity is limited and mechanism of action debated (53,54). A new and potentially more efficacious class of mTOR inhibitors has been developed specifically to target cancer (58-60,92). These small molecule drugs competitively target the adenosine triphosphate (ATP) binding pocket of the 37 mTOR kinase domain and have now entered clinical trials (112). The discovery of these molecules allowed for the division of canonical mTORC1 substrates into classes: those sensitive to inhibition by rapamycin (p70S6 kinase, S6K; and its direct target ribosomal protein S6, rpS6) and those that were relatively insensitive to rapamycin (eIF4E Binding Proteins, 4E-BPs). Inhibition of 4E-BP activity downstream of mTORC1 is responsible for the anti-proliferative effects of PP242 in cell culture models (113). But it remains unclear whether rapamycin sensitive or insensitive targets downstream of mTORC1 are the clinically relevant biomarkers for treatment efficacy. Colon cancer contains many of the most prevalent aberrations in cancer including KRAS and PIK3CA (phosphatidylinositol 3-kinase catalytic subunit alpha) mutations and loss of PTEN (phosphatase and tensin homolog) expression (114). The frequency of these mutations makes it possible to study how each contributes to resistance and sensitivity to molecularly targeted therapies (86). Here we used a high-throughput cell-screening platform to identify a genetic signature for primary resistance of colon cancer to the ATP-competitive mTOR inhibitor PP242. We validated these observations in cultured cell lines and primary human tumor xenografts and identified a biomarker for PP242 efficacy.

4.3 Screening of cancer cell lines

We identified markers of resistance or sensitivity to the ATP-competitive mTOR inhibitor PP242 in solid tumor cell lines (detailed in chapter 2). Our approach relied on automated screening of growth of solid tumor cell lines that were annotated for common oncogenic mutations and tissue of origin (89). The cell line set (n=666) was treated with 500 nM PP242 and assayed for growth inhibition at 72 hours (Appendix 1). The PP242 treatment results were normally distributed (Shapiro-Wilk test, p= 0.145) and centered upon 57.9% of untreated control providing maximal sensitivity to detect both resistant and sensitive cell lines. Significant differences were observed in the response of cell lines grouped by tissue of origin (Figure 4.1a). Of all significantly resistant and sensitive cell types, colon was distinct for its combination of large magnitude of resistance (0.59 standard deviations higher than the population mean) and the significance of this difference (p=0.005). For a subset of cell lines (n=357), we used information from the Sanger COSMIC (http://www.sanger.ac.uk/genetics/CGP/cosmic/) database of cancer cell lines to annotate mutations in 15 of the most common oncogenes. We analyzed the role of these mutations in accounting for sensitivity or resistance to PP242 by two independent methods. Multivariate linear regression of all genotypes against PP242 growth response showed that RAS and PIK3CA were significant and independent predictors of resistance and sensitivity, respectively (test statistics from the regression analysis: RAS, p=0.045; PIK3CA, p=0.017). We also tested whether RAS and PIK3CA mutations were significantly enriched in the population of the most sensitive and resistant cell lines (Figure 4.1b). PIK3CA mutations were absent in the 10% most PP242 resistant cell lines while enriched in the 10% most sensitive ones (Fisher’s exact test: resistant cell lines, p=0.0048; sensitive cell lines, p=0.013). In the most PP242 sensitive cell lines, the reduced number of RAS mutations was significant (Fisher’s exact test: p=0.030), but the enrichment of mutants in resistant cell lines was not (p=0.35). Compared to prior studies that examined responsiveness to inhibitors targeting a specific genetic lesion (such as lapatinib and the EGFR mutant L858R) (89), the strength of the correlation between either PIK3CA or RAS mutation and resistance or sensitivity to PP242 was modest.

38 a b PP242 Response 0.0001 PIK3CA RAS Cell Line Wild-Type Mutant SW620 BT-549 colon DoTc2 4510 0.001 SW756 C-4 II p-value NCI-H2342 Panc 03.27 resistant SCC-9 K1 0.01 MDA-MB-435S stomach cervix HeLa NSCLC ABC-1 nervous COLO-678 kidney bone NCI-H1573 ovary 0.05 FTC-133 brain MKN28 thyroid KP-4 lung head/neck COLO-783 muscle Calu-1 pancreas SW837 skin prostate MDA-MB-468 sensitive resistant SW1116 COLO-792 -1.0 -0.5 0.0 0.5 1.0 MM455 HCC1143 NCI-H1651 Standard deviations from mean A431 HUP-T4 MOG-G-UVW M059J c RPMI 2650 1205Lu Colon cell lines IGR-1 MGH-MC-1 COLO-680 N KRAS mutant KYSE-70 (n= 13) PC-9 * 23132/87 p<.05 H4 KRAS WT EGI-1 ZR-75-30 (n= 10) CAL-12T UACC-893 NCI-H460 0.00 0.25 0.50 0.75 1.00 1.25 NAE TCCSUP NIH-6 Proportion of control cell growth MIA PaCa-2 EFM-19 G-401 OE33 A-204 d NCI-H1048 MCF7 2.5 MG-63 KRAS IGROV-1 CHL-1 KRAS/ PIK3CA MKN45 2.0 IGR-37 WT KRAS NCI-H1581 PFSK-1 MFE-296 1.5 ME-180 MSTO-211H

( μ M) BEN 50 1.0 LK-2 G-402 IC LU99C SW 13 0.5 AN3CA HGC-27

sensitive LU99A 0.0

HT-29 SW48 Caco-2 SW620 LS-513 HCT 15 SW480 SW1116 SK-CO-1 LS-174T HCT 116 COLO 201 COLO 205

Figure 4.1: An unbiased cell screen reveals factors leading to resistance and sensitivity to the ATP-competitive mTOR inhibitor PP242 (a) Colon cell origin is a strong predictor for resistance to PP242 treatment. The mean response of each cell type to 500 nM PP242 treatment was plotted as a standard deviation from the population mean. The y-axis indicates the significance of the test-statistic for each independent difference in means (cell type versus population). The size of the circle corresponds to the number of cell lines of each type analyzed. Colon origin (n=39) was the strongest single predictor of resistance or sensitivity to PP242 among all annotated organ types. (b) PIK3CA mutations are prevalent in cell lines sensitive to PP242 while RAS marks cell lines that are resistant to PP242. The set of 357 cell lines with known mutation status for PIK3CA and RAS were assayed for growth

39 inhibition and ranked according to the inhibition results. The 10% most resistant and most sensitive cell lines are shown here in order of increasing response to PP242. (c) KRAS mutant colorectal cells are more resistant to PP242 than cells harboring WT KRAS. Comparison of means was made using Student’s t-test. (d) IC50 values for PP242 in selected colon cancer cell lines. KRAS mutant cell lines with concomitant PIK3CA mutations are more sensitive to PP242 than KRAS mutants alone.

Knowing that RAS was a modest marker of resistance to PP242, we asked if the high frequency of RAS mutations in colon cancer might account for the degree of PP242 resistance observed in this tumor type. KRAS mutant colon cell lines were significantly more resistant than wild-type (WT) colon cancer cell lines (Figure 4.1c). To quantify the degree of resistance to PP242 imparted upon colon cancer cell lines by KRAS mutations, we determined the half maximal inhibitory concentration (IC50) for a subset of cell lines (Figure 4.1d). The potency (IC50) of PP242 in colon cell lines varied widely from 90 nM to 8 µM (Table 4.1). KRAS mutant colon cell lines were significantly more resistant to PP242 than WT cell lines (p=0.0036; unpaired t-test). The most PP242 resistant cell line, SW620 (IC50 =8 µM) is mutant for KRAS. Of special note were KRAS mutant cell lines with intermediate sensitivity to PP242. This group of cell lines had both KRAS mutations and PIK3CA mutations and their responsiveness to PP242 was closer to that of sensitive cell lines than the resistant ones. The most sensitive colon cancer cell lines were all WT for KRAS. To identify a mechanism for the varied responsiveness of colon cancer cells to PP242, analysis of the downstream effectors of mTOR was performed.

Table 4.1. KRAS and PIK3CA mutations modulate sensitivity to mTOR inhibition

Mutational Status Cell Line PP242 KRAS PI3K bRAF SW620 G12V WT WT 7.8 SW480 G12V WT WT 4.6 SK-CO-1 G12V WT WT 4 LS-513 G12D WT WT 3.9 SW1116 G12A WT WT 0.84 LS-174T G12D H1047R WT 0.84 HCT 116 G13D H1047R WT 0.41 HCT 15 G13D E545K WT 0.3 COLO 205 WT WT V600E 0.24 HT-29 WT WT V600E 0.23 COLO 201 WT WT V600E 0.23 SW48 WT WT WT 0.09

4.4 mTORC1 substrates 4E-BP1 and rpS6 are differentially inhibited by PP242

To ascertain the signaling alterations that lead to the spectrum of responses to PP242 in colon cancer cells, we examined whether mTOR substrate phosphorylation was differentially inhibited in resistant versus sensitive cell lines. Three representative cell lines were studied:

40 a SW620 HCT 15 SW48 b PP242 μM PP242 μM PP242 μM

.33 .33 SW620 SK-CO-1 HCT 116 HCT 15 HT-29 SW48 DMSO RAPA 3.0 1.0 0 0.11 DMSO RAPA 3.0 1.0 0 0.11 DMSO RAPA 3.0 1.0 0.33 0.11 p-AKT T308 PTEN p-AKT S473 rpS6 AKT p-p70S6K T389 p-4E-BP1 T37/46 p-rpS6 S240/244 rpS6 4E-BP1 p-4E-BP1 T37/46 β-Actin p-4E-BP1 T65 4E-BP1 β-Actin

SW620 HCT15 c SW480 LS513 d MLN0128 MLN0128 PP242 μM PP242 μM

DMSO PP242 1.0 0.2 0.04 DMSO PP242 1.0 0.2 0.04 DMSO RAPA 3.0 1.0 0.33 0.11 DMSO RAPA 3.0 1.0 0.33 0.11 p-AKT S473 p-AKT S473 p-rpS6 p-rpS6 S240/244 rpS6 rpS6

p-4E-BP1 T37/46 p-4E-BP1 T37/46

4E-BP1 4E-BP1 β-Actin β-Actin

e SW620 HCT 15 KU-0063794 KU-0063794

DMSO PP242 3.0 0.33 DMSO PP242 10 1.0 10 3.0 1.0 0.33 p-p70 S6K (T389) p-AKT S473

p-rpS6 S240/244 rpS6

p-4E-BP1 T37/46

4E-BP1

β-Actin

Figure 4.2 mTORC1 substrates are differentially inhibited in PP242 resistant versus sensitive cell lines (a) 4E-BP1 is differentially inhibited in PP242 resistant and sensitive colon cancer cell lines. Representative cell lines SW620, HCT 15 and SW48 were treated with PP242 or rapamycin (20 nM) for 1 hour and analyzed by western blotting. (b) Expression levels of 4E-BP1 and basal p-4E-BP1 do not correlate with sensitivity to PP242 treatment in colon cancer cell lines. Selected colon cancer cell lines arranged from left to right in order of increasing sensitivity to PP242 were grown in complete media and lysed without specific stimulation. (c) KRAS mutant colon cancer cell lines SW480 and LS513 are resistant to PP242 treatment and show incomplete inhibition of 4E-BP1

41 phosphorylation compared to sensitive cell lines. (d, e) HCT15 and SW620 cells were treated with ATP- competitive mTOR inhibitors MLN0128 and KU-0063794 for 1 hour prior to lysis.

SW620 (Mut for KRAS), HCT 15 (Mut for both KRAS and PIK3CA) and SW48 (WT for KRAS and PIK3CA) (Figure 4.2a). In PP242-sensitive HCT 15 and SW48 cell lines, 1 hour of treatment with PP242 similarly reduced the phosphorylation of mTORC1 substrates S6K and 4E- BP1, as well as mTORC2 substrate AKT S473. We were surprised to observe that in the PP242- resistant cell line SW620, mTORC1 substrates were differentially inhibited by PP242. Phosphorylation of S6K and its effector rpS6 were potently inhibited by PP242, as in the PP242- sensitive cell line HCT 15, but the phosphorylation of 4E-BP1 was poorly inhibited even at high drug concentrations. Expression levels of 4E-BP1 or basal amounts p-4E-BP1 did not correlate with response to PP242 in the cell lines examined (Figures 4.2a and 4.2b). Resistance of 4E- BP1 to dephosphorylation was observed in other KRAS mutant colon cancer cell lines as well (Figure 4.2c). We performed the same western blot analysis with 2 additional active site mTOR inhibitors: KU-0063794, which is based on a different chemical scaffold from PP242 (59), and MLN0128 (previously INK128), a clinical derivative of PP242 (72). In both cases, 4E-BP1 phosphorylation was less potently inhibited in SW620 cells compared to HCT 15 cells, as was observed with PP242 (Figures 4.2d and 4.2e).

a HCT 15 Cells Rapamycin nM PP242 μM b HCT 15 Cells

7 2 3 1 . 37 DMSO 33 1 3. 1. 10 .12 Rapamycin PP242 3 1.1 0. 100 0 p-AKT T308 Time (hrs) 0 1 6 24 0 1 6 24 p-AKT S473 p-AKT S473 AKT p-rpS6 S240/244 rpS6 p-AKT T308 p-4E-BP1 T37/46 p-rpS6 S240/244 4E-BP1 β-Actin p-4E-BP1 T37/46

c HCT 15 Cells

6 Rapamycin 5 PP242

1 Proliferation (RF) 0 0 1 0.1 10 0.001 0.01 Concentration [μM]

Figure 4.3 PP242, but not rapamycin inhibits mTORC1 substrates in colon cancer cell lines (a) Differential inhibition is reminiscent of incomplete mTORC1 inhibition by rapamycin. In HCT 15 cells, rapamycin only partially inhibits 4E-BP1 phosphorylation after a 1 hour treatment, despite potently blocking rpS6 phosphorylation. (b) HCT 15 cells were treated with rapamycin (20 nm) and PP242 (5 µM) for the times indicated. (c) Treatment with PP242 but not rapamycin fully inhibits cell growth.

Differential inhibition of mTOR substrates has previously been observed with rapamycin where inhibition of phosphorylation of mTORC1 substrates bifurcates between sensitive (rpS6) and insensitive (4E-BP1) targets. Rapamycin has no short-term activity against mTORC2 and

42 incompletely inhibits mTORC1; it fails to inhibit 4E-BP1 phosphorylation while paradoxically increasing AKT phosphorylation (61). All of these effects were visible in the KRAS mutant colon cancer cell line HCT 15 (Figure 4.3a). Feedback activation of AKT has also been observed in human trials with rapamycin derivatives (63). Unlike rapamycin, PP242 suppressed feedback activation of AKT over 24 hours and fully inhibited cell growth (Figures 4.3b and 4.3c). These findings add to work in colon and other cell types showing that the more complete inhibition of mTOR by PP242 results in greater inhibition of cell growth than that achieved by rapamycin (16,115,116). To quantify the sensitivity of 4E-BP1 phosphorylation to inhibition by PP242 and relate it to cell growth, phosphoprotein IC50 curves were constructed (Figure 4.4a). The IC50 for inhibition of phosphorylation of rpS6 was similar in HCT 15 and SW620 cell lines (72 and 212 nM respectively, a 3-fold difference). Conversely, the IC50 values for inhibition of 4E-BP1 phosphorylation were 0.43 µM for HCT 15 cells and 10 µM for SW620 cells. The IC50 values for 4E-BP1 phosphorylation and the difference between cell lines (23-fold) correspond closely to the growth IC50 values (0.30 µM and 7.8 µM, a 26-fold difference). Immunofluorescence imaging of p-4E-BP1 and p-rpS6 in treated SW620 and HCT 15 cells established that rpS6 is excluded from the nucleus in both cell lines and is similarly inhibited (Figure 4.4b). p-4E-BP1 partitions between the cytoplasm and nucleus as previously described (117). Differential inhibition of 4E-BP1 phosphorylation by PP242 between cell lines was not correlated with a difference in subcellular localization of 4E-BP1 (Figures 4.4b and 4.4c). Single cell analysis showed no PP242 resistant subpopulation within the SW620 cell line that would account for the differences in sensitivity (Figure 4.4c). The consistent inhibition of S6K target rpS6 in SW620 and HCT 15 cells demonstrates that PP242 enters the cells and binds mTOR with similar efficacies.

4.5 MAPK signaling differences do not alter mTORC1 substrate specificity

Our finding that KRAS mutant colon cancer cell lines exhibit distinct patterns of substrate inhibition by PP242 led us to question whether outputs of the mitogen activated protein kinase (MAPK) pathway were impacting downstream mTOR substrates. It is known that the mTORC1 component raptor is phosphorylated by the ERK substrate, p90RSK, and we examined whether inhibition of this effect would impact mTOR dependent phosphorylation of 4E-BP1 and rpS6 (118). Immunoblotting revealed that expression of raptor and the phosphorylation of mTOR is greater in the resistant SW620 cell line compared to sensitive HCT 15 cells, however phosphorylation of raptor appeared to be equivalent (Figure 4.5a). MAPK pathway activation as assayed by ERK phosphorylation was also comparable between the two KRAS mutant cell lines regardless of PIK3CA mutation status. Pharmacological inhibition of the MAPK pathway by the MEK inhibitor PD0325901 potently inhibited cell growth in KRAS mutant SW620 cells (119), without impacting mTORC1 signaling in either SW620 or HCT15 cells (Figures 4.5a and 4.5b). In contrast, inhibition of phosphorylation of p90RSK by FMK-MEA (120,121), did not affect mTORC1 substrate phosphorylation, (Figure 4.5a) or cell growth, even at high concentrations (Figure 4.5c). The MAPK pathway does not directly alter the phosphorylation of mTOR substrates in colon cancer. Combination treatment of PD0325901 and an AKT inhibitor leads to profound growth arrest and synergistic inhibition of 4E-BP1 phosphorylation (122). We asked whether the combination of PP242 and a MAPK inhibitor would similarly lead to enhanced suppression of

43 PP242 p-rpS6 S240/244 DMSO 5 μM 1 μM 0.2 μM 0.04 μM a 5 SW620 b

4 HCT 15

p-4E-BP1 (T37/46) 2x 10 μm 1 Relative Intensity SW620 0 0 1 10 0.01 0.1 100 PP242 (μM) 2x 10 μm p-rpS6 (S240/44) p-4E-BP1 T37/46

5 SW620

4 HCT 15

2 p-4E-BP1 (T37/46) 2x 10 μm

1 HCT 15 HCT Relative Intensity

1 0 10 0.01 0.1 100 PP242 (μM) 2x 10 μm p-rpS6 (S240/44)

c p-rpS6 S240/244 p-4E-BP1 T37/46 SW620 HCT 15 SW620 HCT 15

100 100 100 100

80 80 80 80 69 76 60 60 60 60

40 40 40 40

20 20 20 20

0 0 0 0 02004006008001000 02004006008001000 0500100015002000 0500100015002000

100 100 100 100

80 80 80 80 64 73 60 60 60 60

40 40 40 40

20 20 20 20

0 0 0 0 02004006008001000 02004006008001000 0500100015002000 0500100015002000

100 100 100 100

80 80 80 80 68 81 60 60 60 60

40 40 40 40 PP242

0.20 μ M20 1.0 μ M 5.0 μ M 20 20 20

0 0 0 0 02004006008001000 02004006008001000 0500100015002000 0500100015002000

100 100 100 100

80 80 80 80 68 77 60 60 60 60

40 40 40 40 0.04 μ M 20 20 20 20

0 0 0 0 02004006008001000 02004006008001000 0500100015002000 0500100015002000 100 100 100 100 80 80 80 80 60 67 70 60 60 60 40 40 40 40 20 DMSO 20 20 20

0 0 0 0 02004006008001000 02004006008001000 0500100015002000 0500100015002000

Figure 4.4 Quantification of phosphorylated mTORC1 substrates upon PP242 treatment (a) Quantification of mTORC1 substrate inhibition shows that inhibition of p-4E-BP1 and not inhibition of p-rpS6 tracks with growth inhibition. Quantification was performed on western blots of lysed cells after treatment for 1 hour with increasing PP242 concentrations in two independent experiments. (b) Immunofluorescence of mTORC1 substrates reveals consistent subcellular localization despite differential inhibition by PP242. SW620 and HCT 15 cells were treated with increasing concentrations of PP242 for 1 hour, formalin fixed and stained for either p-4E-BP1 or p-rpS6 (both

44 green) and counterstained with DAPI (blue). PP242 treatment does not alter the subcellular localization of either phosphorylated substrate. (c) SW620 and HCT 15 cells were treated with indicated doses of PP242 for 1 h before fixation and staining for immunofluorescence of p-rpS6 and p-4E-BP1. For each condition, average staining intensity for each cell was calculated using the software MetaMorph, and plotted as a normalized histogram. The x- axis is binned average staining intensity, and the y-axis percent of total counts. This analysis demonstrates that there is not a specific PP242 resistant subpopulation of cells but rather the entire distribution is resistant in the SW620 cell line. For p-4E-BP1, the subcellular localization of staining intensity was also calculated and plotted as a pie chart where the darker region represents co-localized staining with DAPI (nuclear) and the light the cytosolic fraction. Subcellular localization of p-4E-BP1 did not correlate with resistance to PP242.

phosphorylation of 4E-BP1. By itself, acute treatment with PP242 modestly increased phosphorylation of ERK, and significantly increased phosphorylation of p90RSK and consequently raptor, suggesting that MAPK combination therapy could be useful (Figure 4.5a). This effect was most prominent in the PP242 resistant SW620 cells. Combination treatment of PP242 and PD0325901 was additive and at least partly mTOR independent (Figure 4.5b). However, addition of PD0325901 did not increase the inhibition of phosphorylation of 4E-BP1 beyond what was achieved with PP242 in both resistant and sensitive cell lines (Figure 4.5a). There was no additive benefit in treatment combining FMK-MEA with PP242 (Figure 4.5c). Thus the enhanced inhibitory effects of combining PP242 with PD0325901 on cell growth are not due to increased inhibition of mTOR substrates but rather ERK substrates that are independent of p90RSK. MAPK signaling did not explain the differential sensitivity of KRAS mutant colon cancer cell lines to PP242, so we examined RAS directly. Mutations activate RAS by biasing the fraction in the active GTP-bound form. This conformation can be selectively pulled down using a GST tagged RAS binding domain (RBD) taken from c-RAF-1 (123). We observed that the amount of RAS-GTP pulled down differed among KRAS mutant cell lines and was inversely correlated with sensitivity to PP242 (Figure 4.5d). KRAS WT and PP242-sensitive cell line SW48 had non-detectable amounts of RAS-GTP in the basal state, yet KRAS mutant cell lines HCT 15, SW480 and SW620 had progressively more RAS-GTP. Our experiment differentiates KRAS mutant cell lines by levels of RAS-GTP and shows a correlation between this activation and resistance to mTOR inhibition by PP242. This was not accompanied by a similar trend in MAPK pathway activation, suggesting that other RAS effectors may be responsible for transmitting the mTOR resistance phenotype observed in high RAS-GTP cells.

4.6 Mutant PIK3CA but not PTEN loss leads to mTOR inhibitor sensitization

Phosphoinositide signaling in colon cancers is deregulated by both loss of PTEN expression (35%) and activation of PIK3CA (15%) (85,124). Our observation that KRAS mutant cell lines with PIK3CA mutations were sensitive to PP242 led us to investigate how changes in phosphoinositide signaling impact the sensitivity of mTOR to PP242. A previous analysis of PTEN status in breast cancer cell lines observed no correlation between PTEN expression and PP242 sensitivity (78). Our analysis of PTEN in the complete PP242 cell screen similarly showed no relation between PTEN mutations and PP242 sensitivity, but this analysis was confined to the small subset of PTEN mutant tumors whereas loss of expression is more often accomplished via epigenetic mechanisms. All of our cell lines express PTEN, and because loss of PTEN can activate AKT signaling, we wanted to test whether reduction of PTEN would sensitize cells to PP242. Using siRNA directed against PTEN, a 96-hour knockdown was

45 a SW620 HCT 15 d

1 W620 90 S SW480 HCT 15 SW48 5

32 PP242 IC50 (μM) 7.8 4.6 0.3 0.09 P242+PD D0 P242+PD KRAS MUT + ++- DMSO PD0325901 FMK-MEA PP242 P PP242+FMK DMSO P FMK-MEA PP242 P PP242+FMK PIK3CA MUT - - + - p-AKT S473 pull down GST AKT RAS p-ERK T202/Y204 GST-RBD p-p90RSK S380 RAS p-raptor S722 raptor p-MEK1/2 S217/221

p-mTOR S2448 MEK1/2 Lysate mTOR p-ERK T202/Y204 p-rpS6 S240/244 ERK rpS6 4E-BP1 p-4E-BP1 T37/46 KU-86

4E-BP1

β-Actin

b SW620 Cells c SW620 Cells

MEK inhibition RSK inhibition 5 5

4 4

3 3 PP242 2 2 PP242 PD0325901 FMK PP242 + 10 nm PD 1 1 PP242+1 µM FMK Fluorescence (arbitrary) Fluorescence (arbitrary) 0 0

0 0 1 0.01 0.1 1 10 0.1 10 100 Concentration (μM) Concentration (μM)

Figure 4.5 Inhibition of MAPK signaling does not alter mTORC1 substrate phosphorylation (a) Treatment of cell lines with MAPK inhibitors does not sensitize mTOR substrates to PP242 inhibition. Cells were treated with the MEK inhibitor PD0325901 (20 nM), the p90RSK inhibitor FMK-MEA (3 µM), PP242 (1 µM) singly or in combination for 1 hour. Neither FMK-MEA nor PD0325901 sensitize mTOR substrates to PP242 treatment. (b-c) Inhibition of ERK but not p90RSK augments the cell growth arrest induced by PP242. The MEK inhibitor PD0325901 is a potent inhibitor of cell growth, but selective inhibition of ERK substrate p90RSK by FMK-MEA does not recapitulate this phenotype in a three-day cell growth assay measured using a resazurin assay. (d) RAS- GTP loading, but not MAPK pathway activation, correlates with resistance to PP242. Glutathione pull down of RAS-GTP with GST-RBD shows high RAS-GTP levels in unstimulated KRAS mutant PP242 resistant cell lines. Significantly different levels of RAS-GTP activation were observed within KRAS mutant cells.

performed and the cells were analyzed for their response to PP242 (Figure 4.6a). PTEN knockdown modestly increased p-AKT S473 levels but did not change the dose response on

46 either mTORC1 substrate. Our results are consistent with prior studies showing that PTEN and PIK3CA activate mTOR signaling downstream of AKT in non-redundant ways. To determine if mutant PIK3CA is sufficient to sensitize SW620 cells to PP242, we transfected SW620 cells with constructs of the most common PIK3CA mutations. Mutations in PIK3CA cluster into two conserved “hotspots”: a kinase domain hotspot, most commonly H1047R, and helical domain mutations, most commonly E542K or E545K (125). As previously shown, the PIK3CA kinase domain mutant was more transforming than the helical domain mutations and conferred a substantial growth advantage to the cells (Figure 4.6b) (126). a b PP242 PP242 DMSO DMSO DMSO siRNA (PTEN) ------+ + + + + 12500 siRNA (scramble) - + + + + + - - - - - WT PTEN 10000 E542K p-AKT S473 H1047R 7500 AKT p-rpS6 S240/244 5000 p-4E-BP1 T37/46 2500 Relative Fluorescence 4E-BP1 0 0 1 2 3 4 5 6 β-Actin Day

c SW620 p110α WT H1047R E542K PP242 μM PP242 μM PP242 μM PP242

DMSO DMSO RAPA 0.33 1.0 3.0 DMSO RAPA 0.33 1.0 3.0 RAPA 0.33 1.0 3.0 DMSO RAPA p-AKT T308 p-AKT S473 p-rpS6 S240/244 p-4E-BP1 T37/46 β-Actin

d PIK3CA Genotype WT H1047R E542K PP242 IC (μM) 11 1.2 10 50

Figure 4.6 PIK3CA mutation but not PTEN loss sensitizes KRAS mutant cells to PP242 (a) siRNA against PTEN does not sensitize KRAS mutant CRC cells to PP242. SW620 cells were treated with siRNA against PTEN for 72 hours prior to 1 hour drug treatment with PP242 (3.0, 1.0, 0.33 and 0.1 µM). (b) Addition of mutant PIK3CA to the KRAS mutant cell line SW620 increases sensitivity to PP242. Retroviral insertion of either WT PIK3CA, helical (E542K) or kinase (H1047R) domain mutations only resulted in elevated basal AKT activation in the H1047R case.

Cells were treated with rapamycin (20 nM) or PP242 (3.0, 1.0, 0.3 µM) for 1 h before lysis. (c) IC50s for SW620 cell lines engineered to contain additional PIK3CA mutations.

47 We treated the PIK3CA and KRAS co-mutant isogenic cell lines with PP242 and analyzed phospho-signaling. The basal phosphorylation of both AKT T308 and S473 was significantly higher in the H1047R mutant than either the WT or the helical domain E542K mutant expressing cells (Figure 4.6c). The inhibition of phosphorylation of the mTOR substrate 4E-BP1 was only significantly altered in the H1047R mutant expressing cell line. Activated AKT signaling sensitized KRAS mutant cells to mTOR inhibition consistent with the response of mTORC1 substrate 4E-BP1 to acute pharmacological inhibition. To confirm the relationship between 4E-BP1 inhibition and cell growth, SW620 PIK3CA mutant cell growth was assayed and the PP242 IC50 was determined. The values of 10 and 11 µM for the E542K and WT lines were nearly identical to the measured SW620 parental cell line (8 µM) whereas the H1047R cell line was approximately 8-fold more sensitive to PP242 (Figure 4.6d).

4.7 KRAS mutation status predicts response to PP242 in human primary xenografts

We sought to validate our cell line observations about PP242 sensitivity and KRAS and PIK3CA mutation status in an in vivo model of human colon cancer, patient-derived xenografts. Such xenografts allow patient tumors to be maintained in vivo without undergoing the irreversible changes that occur upon in vitro culture (93). Patient-derived xenografts overcome many of the problems that render standard cell line and cell line derived xenografts models poorly predicative of clinical response (94,95). Their utility in colon cancer was recently demonstrated by the identification of a genetic marker of resistance to the anti-EGFR antibody cetuximab (87). Xenografts were established from liver metastases of patients with colon cancer resected with curative intent (100). Non-diagnostic portions of removed metastases were implanted, characterized and subsequently passaged in athymic nude mice (for complete characterization see chapter 3). To determine the effects of PP242 in patient-derived xenografts with genetic lesions common in colon cancer, three different patient-derived tumors representing three different combinations of mutant PIK3CA and KRAS were analyzed: WT KRAS and WT PIK3CA (CR 698); Mut KRAS and WT PIK3CA (CR 702); Mut KRAS and Mut PIK3CA (CR 727). Cohorts of single tumor-bearing mice were treated once daily with PP242 or vehicle for 30 days or until (control) tumor burden had reached protocol limits. Treatment was tolerated. PP242 slowed tumor growth compared to control (Figure 4.7a). In trials with either WT or double mutant tumors (CR 698 and CR 727, respectively), the decrease in tumor growth between treatment and control arms was apparent after seven days. This was in contrast to the more modest effect of PP242 in the KRAS single mutant tumor (CR 702), where the difference in tumor growth was only significant after 28 days. In no trial did PP242 lead to significant tumor regression (>50% in volume) in an individual mouse, but stable disease (final tumor volume of -50% to +20% of starting) was achieved in 26% of mice with CR 698 or CR 727 tumors (and no mice with CR 702 tumors). In PP242 responsive tumors, the growth inhibitory effects were not accompanied by a histological change in tumor characteristics. To directly compare the separate trials and better understand the differences between the resistant KRAS mutant tumor CR 702, and sensitive tumors CR 727 and CR 698, we fit the trial data to a linear mixed effects model. Using the model, we obtained a daily tumor growth rate for each treatment condition and compared the effect of PP242 on tumor growth rate (Figure 4.7b). Differences in tumor growth rate between xenografts could not be independently excluded as

48 CR 698 CR 702 CR 727 a KRAS WT/ PIK3CA WT KRAS G12D/ PIK3CA WT KRAS G12V/ PIK3CA H1047R

900 900 900 Vehicle Vehicle Vehicle ** ** ** PP242 PP242 700 PP242 700 * 700 ** * ** 500 ** 500 500 ** ** * *** * 300 300 300 ** *

100 100 100 Tumor Vol. Growth (%) Vol. Tumor Growth (%) Vol. Tumor Growth (%) Vol. Tumor 0 5 10 15 20 25 051015202530 0 5 10 15 20 25 30 35 Day Day Day

b c CRCR 698 CRCR 702 CR 727 e Growth rate diff. n.s. 0 -6 -5 -4 -3 -2 -1 0

*** ** Growth rate difference 9 Vehicle -1 8 0 % phospho inhibition PP242 7 -2 6 CR 702 -3 -25 5 CR 698 4 -4 CR 727 -50 3 -5 n.s. % Growth/ Day 2 -6 1 CR 702 -75 -7 0 * CR 727 CR 698 CR 698 CR 702 CR 727 -100 d CR 698 CR 702 CR 727 p-4E-BP1 p-rpS6 ** n.s. *** 1.0 1.0 1.0

0.75 0.75 0.75

0.50 0.50 0.50

0.25 0.25 0.25 p-4E-BP1 (T37/46)

0 0 0 Vehicle PP242 Vehicle PP242 Vehicle PP242

** 1.0 1.0 *** 1.0 ***

0.75 0.75 0.75

0.50 0.50 0.50

0.25 0.25 0.25 p-rpS6 (S240/244) 0 0 0 Vehicle PP242 Vehicle PP242 Vehicle PP242

Figure 4.7 KRAS mutant patient-derived xenografts are resistant to PP242 by incomplete inhibition of 4E-BP1 phosphorylation. (a) Percent growth curves of three xenografts show differences in response to PP242 treatment. KRAS and PIK3CA genotypes are as follows: CR 698 (KRAS WT/ PIK3CA WT), CR 702 (KRAS Mut/ PIK3CA WT), CR 727 (KRAS Mut/ PIK3CA Mut). Mice were given 100 mg/kg PP242 once daily or vehicle for the indicated time. Tumors were normalized to 100 percent at the beginning of dosing and percent growth ±SEM was plotted for each day when tumor volume measurements were taken. Asterisks indicate significant differences in tumor growth at each measurement point as determined by an unpaired t-test (* p< .05, ** p< .01, *** p< .001). (b) Treatment effect is significant in tumors CR 698 and CR 727. Tumor growth rates were calculated using a linear mixed effects model. PP242 led to a significant reduction in growth rate as calculated using a Wald test (asterisks represent the same p values as in A) in the KRAS WT tumor CR 698 and the double mutant tumor CR 727, but not the KRAS single-mutant tumor CR 727. (c) PP242 is most effective at inhibiting growth of the KRAS WT tumor CR 698. Comparison of the growth rate difference calculated from the model shows that PP242 is significantly more effective at inhibiting growth in CR 698 than in CR 702. The growth rate difference is the growth rate of the PP242 treated tumors minus the control growth rate. All other comparisons were not statistically significant. (d) Whole-tumor western blots show that p-4E-BP1 levels were significantly more reduced by PP242 treatment in

49 KRAS WT and KRAS/ PIK3CA double mutant tumors but not in KRAS single-mutant tumors. After treatment with either PP242 or vehicle for 30 days, tumors were removed and analyzed by western blot for phosphoprotein analysis. Bands were quantified by fluorescent antibodies and intensities internally normalized to those of β-actin. Intensities are reported as arbitrary normalized fluorescence units. Statistical comparisons were made using two- tailed t-tests as in Figure 3A. (e) Changes in p-4E-BP1 but not p-rpS6 correlate with changes in tumor growth. A plot of the tumor growth rate difference (Figure 4.7c) versus percent inhibition of p-4E-BP1 and p-rpS6 shows that the efficiency in inhibiting 4E-BP1 phosphorylation correlates linearly with the percent growth defect between treated and untreated tumors. Percent inhibition of p-rpS6 does not vary significantly with genotype or tumor growth defect as calculated from linear mixed effects model.

contributing to the treatment effect, but molecular data strongly suggest that the effects were related to mTOR inhibition. The effect of PP242 treatment on tumor growth was highly significant as determined by a Wald test for both the CR 698 trial (p <0.001) and CR 727 trial (p=0.001) but not for the CR 702 trial (p =0.123). Further comparison of the magnitude of the PP242 treatment effect showed that mTOR inhibition was significantly more effective in the CR 698 trial than the CR 702 trial, (p=0.04) but that the difference in treatment effect was not statistically significant between CR 702 and CR 727 (Figure 4.7c).

4.8 Inhibition of 4E-BP1 and not rpS6 correlates with anti-tumor effect of PP242

To identify a basis for the differential effect of PP242 on tumor growth, we conducted phospho-signaling analysis by quantitative western blotting (Figure 4.7d). Western blots showed an unambiguous inhibition of p-rpS6 in all PP242-treated samples indicating that mTOR was at least partially inhibited in all trials and that PP242 was able to access the tumors equally. The phosphorylation of substrate 4E-BP1 was differentially inhibited among the different trials. p- 4E-BP1 was significantly inhibited in the CR 698 and CR 727 trials, but not CR 702. Immunohistochemical staining for p-4E-BP1 was preformed on select tumors, and the results were consistent with the western blotting (data not shown). Percent inhibition of p-4E-BP1, not p-rpS6, linearly correlated with tumor growth inhibition (Figure 4.7e). These trials showed a striking primary resistance to PP242 treatment in a KRAS mutant tumor (CR 727) that was not evident in either a tumor with wild type KRAS or a tumor with a PIK3CA mutation in addition to KRAS.

4.9 Discussion

The ability to identify determinants of primary resistance to targeted therapies undergoing clinical development has the potential to provide a benefit to patients by guiding patient selection and development of effective biomarkers. We approached this study with the hypothesis that upstream inputs of mTOR that are mutated would affect sensitivity to mTOR inhibitors. Our large cell screen showed that the average sensitivity of different tumors types varied significantly, reflecting the different genetic make-ups of these cancers. Additionally, we were encouraged to find significant positive and negative correlations to PP242 efficacy with mutations in PIK3CA and KRAS, respectively. Newly published data validates our approach showing that PIK3CA H1047R mutations are predictive of response to rapamycin based mTOR inhibitors in patients (127). We chose to examine colon cancer in detail because both mutations

50 implicated in sensitivity and resistance are present and the extreme resistance to mTOR inhibitors in KRAS mutant colon cancer cell lines was unique. Differences in sensitivity to PP242 among tumors and cell lines were linked to differences in the sensitivity of mTOR substrates to pharmacological inhibition of their phosphorylation. What was unexpected about our findings was the decoupling of inhibition of mTORC1 substrates 4E-BP1 and rpS6 by an ATP-competitive mTOR inhibitor. The phosphorylation status of 4E-BP1 is a better indicator of the functional state of mTOR than S6K and closely correlates with the growth arrest caused by mTOR inhibition. In the single mutant KRAS patient-derived xenograft and cell lines, the inhibition of mTORC1 substrates by PP242 is similar to that of rapamycin, which is now well established to block p-rpS6 but not p-4E-BP1 by a still unknown mechanism. Our observation that in certain cell lines, an ATP-competitive inhibitor could produce these rapamycin like differential inhibitory effects, suggests that a common mechanism may underlie these phenomena. To test the hypothesis that activated PIK3CA signaling conferred sensitivity to mTOR inhibition, even in a mutant KRAS background, isogenic cell lines were created that differed only by their PIK3CA mutation status. Increased sensitivity to PP242 was only observed in the cell line containing an AKT activating PIK3CA mutation. This result adds to a significant set of data that shows that upstream AKT mutations sensitize tumors to mTOR inhibition. The resistance driven by mutant KRAS is not reversible by short-term MAPK inhibition, leading us to conclude that phosphorylation changes regulating mTORC1 do not affect substrate accessibility. Our observation that resistance is correlated with RAS-GTP loading functionally differentiates cell lines with KRAS mutations, and suggests that the mechanism of resistance is directly KRAS driven but independent of MAPK signaling. Concurrent to our work on primary resistance to ATP-competitive mTOR inhibitors, other groups have recently reported mechanisms of acquired resistance to these agents. In a study of human mammary epithelial cells treated at sublethal doses of the dual PI3K/mTOR inhibitor BEZ235, Roberts and coworkers identified MYC amplification in one cell line and eIF4E amplification in another cell line (128). Sonenberg and coworkers characterized PP242 resistant E1A/Ras-clones generated by growing the cells for two months in drug (129). The emergent clones exhibited down regulation of 4E-BP1 and 2 or overexpressed eIF4E, highlighting the importance of monitoring the ratio of 4E-BPs and eIF4E in tumors to assess likely responses to ATP-competitive mTOR inhibitors. These two studies on emergent resistance further highlight the importance of translational regulation in response to mTOR inhibitors. Owing to the lack of direct inhibitors of KRAS, considerable effort has been made to uncover druggable targets (such as kinases TBK1 and STK33) that might act in a synthetic lethal manner with KRAS mutant tumors (130). Our study reveals a different approach to the same goal, that of finding a small molecule that is effective even in a setting of KRAS mutant tumors. We find that the presence of a second oncogene, PI3KCA, provides a signal that sensitizes the KRAS mutant tumors to an ATP-competitive mTOR inhibitor. This result is surprising because the presence of a second lesion typically provides the cancer with a bypass mechanism to avoid kinase inhibitor sensitivity (131). Of note is that 3 of the 4 RAS mutant cell lines that were among the 10% most sensitive to PP242 also harbored a PIK3CA mutation. In the context of colon cancer, a solid tumor in which small molecule kinase inhibitors have yet to achieve significant clinical utility, our findings suggest subsets of patients who may benefit from targeted anti-mTOR therapy. First, mutations in PIK3CA are likely to be sensitive to mTOR inhibition. The importance of selectively treating mutant PIK3CA colon cancer

51 patients has been highlighted by a retrospective analysis demonstrating that low dose aspirin (by an as yet unclear mechanism) can dramatically prolong survival in those patients (132). Additionally, patients with WT KRAS are likely to be responsive to mTOR inhibitors. Furthermore, we believe that among colon cancer patients with mutant KRAS, those with concomitant hyperphosphorylation of AKT induced by mutant PIK3CA may benefit from ATP- competitive mTOR inhibitors. Finally, although rpS6 remains the default biomarker for mTOR inhibitors, our study shows that the phosphorylation status of 4E-BP1 may be a more relevant biomarker of treatment efficacy for ATP-competitive inhibitors.

4.10 Materials and Methods

Inhibitors. The mTOR inhibitors PP242 and MLN0128 were synthesized from commercially available starting materials as in chapter 3. Rapamycin and PD0325901 were purchased from EMD-Millipore chemicals (Billerica, MA). FMK-MEA was a gift of Jack Taunton (UC San Francisco, San Francisco, CA). \

Cell Screen. The automated cell screen was performed using PP242 (500 nM) as described in chapter 2. Complete results are available in Appendix 1.

Cell culture. Cell lines were purchased from the American Type Tissue Collection (ATCC, Manassas, VA) and cultured according to their recommendations in antibiotic-free media.

Cell proliferation assay. Cells were plated on 96-well plates at densities between 2500 and 5000 cells/well. Cell growth was assayed using resazurin sodium salt (Sigma, Saint Louis, MO) and measured using a Safire bottom-reading fluorescent plate reader with excitation at 530 nm and emission at 590 nm.

Western blot analysis. Cells were grown in 6 or 12-well plates and treated with inhibitor(s) or vehicle (0.1% or 0.2% DMSO for single or combination drug assays, respectively) for 1 h unless otherwise noted. Cells were then lysed in radio-immunoprecipitation assay buffer (RIPA); lysates were normalized for protein content using a Bradford assay (absorbance at 595 nm), resolved by SDS-PAGE, transferred to nitrocellulose and blotted. Phosphorylation-specific antibodies were purchased from Cell Signaling Technology (Danvers, MA), except for phospho- raptor S722 (Millipore) and visualized by HRP or fluorescent secondary antibodies. Quantitative western blotting was accomplished using fluorescent secondary antibodies (800 nm emission) for visualization using an Odyssey IR scanner from LI-COR Biosciences (Lincoln, NE). All reported band intensities were internally normalized to β-Actin and each experiment was done in independent biological duplicates.

Immunofluorescence. Cells were plated on fibronectin treated glass-bottom 12-well plates, drug treated and then fixed and stained following standard protocols (Cell Signaling Technology). An Alexa Fluor 488 conjugated goat anti-rabbit secondary antibody (Life Technologies, Carlsbad, CA) was used to visualize the cells using a Zeiss Axiovert 200M fluorescence microscope. Nuclei were counterstained with Hoechst 33342 dye (Pierce). Image analysis was conducted with the software suite MetaMorph. 52

GST-RBD Pull Down. Assay was adapted from a protocol published with the Pierce active RAS pull down and detection kit (#16117). Cells were lysed in HEPES lysis buffer (40 mM HEPES pH 7.4, 150 mM NaCl, 0.1% Tx-100). GST Pull down was conducted according to protocol using purified GST-RBD-c-Raf-1 and glutathione beads (GE Healthcare, Pittsburgh, PA). Pan RAS antibody was from Epitomics (Burlingame, CA).

siRNA. Pooled siRNA against PTEN (SMARTpool PTEN) or control scramble siRNA was purchased from Millipore. siRNA was transfected using DharmaFECT 2 reagent according to manufacture’s instructions (Thermo Scientific Dharmacon Products, Lafayette, CO). 96 h after transfection, cells were treated with drugs and western blotted.

Mutant PIK3CA SW620 cell lines. Mutant p110α expressing SW620 lines were created by retroviral infection using a pMIG-p110α plasmid as previously reported (133). Briefly, ecotropic p110α viral stocks were generated by transfecting pMIG-p110α plasmid DNA into the Phoenix Eco cell line. Human SW620 cells were pseudotyped for infection with murine ecotropic virus by transient transfection with pcDNA3-EcoR/MCAT1. 24 h after transfection with pcDNA3-EcoR/MCAT1, the SW620 cells were infected with thawed p110α viral stocks for 12 h, switched to growth medium for 6 h, and then expanded into 75 cm2 flasks. After 3 days, the infected cells were sorted by FACS for expression of green fluorescent protein (GFP). Each stable GFP-expressing cell population was confirmed to contain the desired mutation by allele specific sequencing.

Patient-derived xenografts. The research protocol was approved by the Committee on Human Research of the University of California, San Francisco (UCSF) and patient consent was obtained. All animal studies followed a protocol approved by the UCSF Animal Care and Use Committee. Xenografts are described in chapter 3.

Drug treatment. PP242 was prepared as a 25 mg/mL suspension in 3.1% NMP, 81.6% H2O, 15.3% PVP. 100 µL of PP242 suspension (2.5 mg/dose equal to 100mg/kg) or vehicle alone was given orally once daily. For drug efficacy trials, mice bearing single tumors were treated for 30 days or when tumors reached 3000 mm3 and then sacrificed and tumors were harvested.

Histologic analysis of xenograft tumors. 4 µm sections prepared from FFPE tissue were stained with hematoxylin & eosin (H & E) and submitted for review by the pathologist, J.P.S.

Sequencing. DNA extracted from surgical specimens with the Qiagen tissue kit (Qiagen, Valencia, CA) was sequenced using standard Sequenom platform protocols (Sequenom, San Dieg, CA) and colon cancer-specific mutational panel (ColoCarta) (134). iPLEX well 6, containing KRAS-Q61L and HRAS-Q61L primers was omitted.

Statistical analyses. Statistical tests and curve fitting was conducted using the programs Stata and Prism. Univariate and multivariate regressions were conducted using the default analysis models in Stata. Comparison of means was conducted using two-tailed Student’s t tests. Data from drug treatment trials was fit to a linear mixed effects model as follows: Tumor volume

53 was transformed to Y(t) = ln[V(t)/V(0)], where V(t) is the tumor volume at time t and V(0) is the volume at time 0, where time 0 is set equal to 100. We used the Stata program xtmixed to fit a linear mixed effects model to the transformed data: Y(tij) = (B1 + B2 × Itreatment + b1i) × tij + eij (1) where B1 is the fixed slope for untreated (vehicle), B2 is the change in slope due to treatment, Itreatment is an indicator for the treatment (=1 for treated, =0 for untreated), b1i is a random effect on the slope for the ith mouse, tij is the jth time that the volume was measured on the ith mouse, and eij is a random error term (135). The treatment effects were tested by pooling data from all experiments and introducing a separate change in slope fixed-effect parameters, B21, B22, and B23 for each treatment. Wald tests determined whether the slopes differed for different treatments.

Chapter 5:

Conclusions and future perspectives

55 5.1 The genetic landscape of cancer and targeted therapies

Genetic sequencing of human tumors has revealed the specific molecular lesions that give rise to cancer (136). A large number of these mutations are to members of the kinase family of intracellular signaling proteins. Kinases share a conserved active site and small-molecule inhibitors developed against this class of proteins have in select instances shown remarkable efficacy in the clinic. The most striking successes have been observed in certain hematological malignancies such as the case of chronic myelogenous leukemia (CML) that is driven by a single chromosomal translocation activating a single kinase. Inhibition leads to apoptosis in the affected cells and suppression is durable given continued treatment. Similar successes have not been observed in solid tumors. Genetically they are very complex with many independent mutations. Treatment with inhibitors targeted to mutated lesions can result in spectacular results, but unlike for CML, responses are rarely greater than 50% and even when tumor regression is observed, they are rarely durable even with continued treatment. If these solid tumors are not resectable, targeted kinase inhibitors have not appreciably improved survival. The greatest challenges in the field remain unmet. They include consolidating successes in patients with mutations who initially respond to targeted therapy, understanding why large numbers of patients with selected mutations fail to respond to targeted therapy, and finally developing agents that can target activated signaling pathways in patients without mutations in druggable kinases. Expanding the scope of targeted kinases beyond those mutated in cancer may help to address these major challenges. The vast majority of mutations in signaling kinases are to nodes that lie near the top of signaling cascades where significant redundancies exist. While these mutations can propagate signal to many different downstream pathways, it has become clear that in the complicated genetic environment of solid tumors, they are rarely truly essential and adaptive resistance to targeted treatment is common. Along a kinase cascade, mutations are less common near the bottom, where the targets are often functional proteins such as transcription factors and translation proteins. There is both great promise and risk in targeting these kinase nodes. The potential therapeutic index (the difference in drug effects on healthy versus tumor tissue) is much narrower, but at these points evolution around the pharmacological blockade is much more difficult. Additionally, these nodes may integrate signals of several different upstream mutations, including from non-druggable proteins. Despite the risks, these benefits have attracted much attention and inhibitors against these non-mutated kinases have entered the clinic. Inhibitors against MEK and mTOR both target an essential and evolutionary conserved signaling kinase that is itself not mutated in cancer. MEK is a kinase in the MAPK pathway downstream of the most common oncogene in cancer, RAS. But RAS is a GTPase, and it is likely impossible to develop a competitive inhibitor of this protein owing to the high affinity for its natural substrate. Consequently, efforts have been focused on pathway components downstream. MEK inhibitors are extremely potent inhibitors of cell growth, but also show high toxicity to normal tissue. The therapeutic inhibition of mTOR fits into this paradigm. Mutations to both kinases and tumor suppressors that lie upstream of mTOR activate it, while mutations in mTOR itself appear to be only rarely tolerated. Given its role in directly controlling cell growth and key metabolic pathways, inhibition of mTOR may be highly toxic. However this has not been the case with either of the classes of mTOR inhibitors used in animal and human trials to date. Rapamycin, with its unique mechanism of action, is well tolerated but that is likely largely due to

56 its only impartial inhibition of mTORC1. The more complete ATP-competitive inhibitors are also very well tolerated- at least in short-term trials. This is due in part to the natural feedback mechanisms the mTOR can engage. Acute mTOR inhibition engages cellular autophagy, a protective program that cells can use to avoid starvation. mTOR inhibition also activates some growth factor signaling pathways, in particular AKT. With these protective pathways in place, mTOR inhibition while effective in cancer cells, may not be apoptotic. To take advantage of these new agents, they will likely need to be combined with different therapeutic modalities in order to secure cell death and not just growth arrest.

5.2 Combining anti-mTOR therapy with other agents

The standard of care for metastatic cancer are regimens of cytotoxic chemotherapy. Based upon a variety of agents that targets primarily DNA synthesis and replication, these regimens have been refined over the years to extend life in patients but unfortunately only rarely result in cures in adult disease. Common agents target the DNA unpacking enzyme topoisomerase, DNA synthesis through the use of uracil analogs and DNA itself via alkylation. mTOR inhibitors are being tested in clinical trials after failure of these conventional therapies but there is likely significant potential in combining them. Understanding combination therapy will also help guide clinicians who are likely initially to use these drugs with older agents. In cell assays, PP242 has been combined with representative drugs in each of these classes of chemotherapy (116). In these experiments, PP242 was synergistic with irinotecan, a DNA topoisomerase inhibitor in multiple colon cancer cell lines tested. For 5-fluorouracil and oxaliplatin, synergy with PP242 was only observed in those cell lines already sensitive to PP242 treatment alone. Together, this data supports a model whereby the efficacy of mTOR inhibition in cells can be significantly augmented by the addition of chemotherapy. The most promising developments in the treatment of metastatic disease however have been in the field of immunotherapy. T-cell checkpoint antibodies such as ipilumimab have the potential to bring cures for many patients. While there is no clinical data about the combination of these therapies and ATP-competitive mTOR inhibitors, preclinical work suggests that counter intuitively, these drugs may synergize with immunotherapy. PP242 is not immunosuppressive, unlike rapamycin in cell lines models (76). Jiang and colleagues report that AZD8055, but not rapamycin can potentiate an immune response to an anti-tumor vaccine in mouse models (137). These very early results open up interesting possibilities in combination therapy with mTOR inhibitors going forward.

5.3 Conclusion

Cancer therapy is in the middle of a revolution fueled by massive advances in our understanding of cancer cell biology and the body’s own reaction to tumors. Kinase inhibitors have now been successfully applied clinically to a variety of relatively rare tumors with specific genetic lesions, but new drugs that target a broader spectrum of patients are now in development. Among these new targets is mTOR and ATP-competitive inhibitors that target this kinase have shown remarkable preclinical promise. The existing FDA approved anti-mTOR therapies, derivatives of the natural product rapamycin are unlikely to shown significant benefit in a broad variety of cancers. Their failure to inhibit oncogenic targets of mTORC1, such as 4E-BP1, will limit their applicability. It is still very early days, but the promise of mTOR inhibition by ATP-

57 competitive inhibitors is very high. They appear to act on a pathway that is deregulated in a large number of tumors and may in fact be thought of as a chemotherapeutic in that regards. However, it is clear that specific upstream mutations do sensitize cells to inhibitors of this pathway, and clinical trials should prioritize these patients, chief among them those with PIK3CA mutations for phase I/II testing. The use of these drugs will almost assuredly be in combination with other therapies and early data suggests that ATP-competitive mTOR inhibitors may synergize with some traditional chemotherapies and modulate the immune response in a positive way when combined with existing and future immunotherapies. Clinical trial data is eagerly awaited to further our understanding of the human effects of these agents and if they can replicate their very successful preclinical results.

58 References

1. Efeyan A, Sabatini DM. mTOR and cancer: many loops in one pathway. Current Opinion in Cell Biology. 2009 Nov 27.

2. Laplante M, Sabatini DM. mTOR Signaling in Growth Control and Disease. Cell. Elsevier Inc; 2012 Apr 13;149(2):274–93.

3. Murakami M, Ichisaka T, Maeda M, Oshiro N, Hara K, Edenhofer F, et al. mTOR is essential for growth and proliferation in early mouse embryos and embryonic stem cells. Molecular and Cellular Biology. 2004 Aug 1;24(15):6710–8.

4. Kim D-H, Sarbassov DD, Ali SM, King JE, Latek RR, Erdjument-Bromage H, et al. mTOR interacts with raptor to form a nutrient-sensitive complex that signals to the cell growth machinery. Cell. 2002 Jul 26;110(2):163–75.

5. Hara KK, Maruki YY, Long XX, Yoshino K-IK, Oshiro NN, Hidayat SS, et al. Raptor, a binding partner of target of rapamycin (TOR), mediates TOR action. Cell. 2002 Jul 26;110(2):177–89.

6. Sarbassov DD, Ali SM, Kim D-H, Guertin DA, Latek RR, Erdjument-Bromage H, et al. Rictor, a novel binding partner of mTOR, defines a rapamycin-insensitive and raptor- independent pathway that regulates the cytoskeleton. Curr Biol. 2004 Jul 27;14(14):1296–302.

7. Kuo CJ, Chung JK, Fiorentino DF, Flanagan JU, Blenis J, Crabtree GR. Rapamycin Selectively Inhibits Interleukin-2 Activation of P70 S6 Kinase. Nature. 1992;358(6381):70–3.

8. Gingras AC, Gygi SP, Raught B, Polakiewicz RD, Abraham RT, Hoekstra MF, et al. Regulation of 4E-BP1 phosphorylation: a novel two-step mechanism. Genes Dev. 1999 Jun 1;13(11):1422–37.

9. Loewith RR, Jacinto EE, Wullschleger SS, Lorberg AA, Crespo JLJ, Bonenfant DD, et al. Two TOR complexes, only one of which is rapamycin sensitive, have distinct roles in cell growth control. Mol Cell. 2002 Sep 1;10(3):457–68.

10. Peterson TR, Laplante M, Thoreen CC, Sancak Y, Kang SA, Kuehl WM, et al. DEPTOR Is an mTOR Inhibitor Frequently Overexpressed in Multiple Myeloma Cells and Required for Their Survival. Cell. 2009 May 12;:1–14.

11. Jacinto E, Loewith R, Schmidt A, Lin S, Rüegg MA, Hall A, et al. Mammalian TOR complex 2 controls the actin cytoskeleton and is rapamycin insensitive. Nat Cell Biol. 2004 Nov;6(11):1122–8.

12. Brown EJ, Albers MW, Shin TB, Ichikawa K, Keith CT, Lane WS, et al. A mammalian protein targeted by G1-arresting rapamycin-receptor complex. Nature. 1994 Jun 30;369(6483):756–8. 59 13. Choi J, Chen J, Schreiber SL, Clardy J. Structure of the FKBP12-rapamycin complex interacting with the binding domain of human FRAP. Science. 1996 Jul 12;273(5272):239–42.

14. D DD Sarbassov Dos, Ali SMS, Sengupta SS, Sheen J-HJ, Hsu PPP, Bagley AFA, et al. Prolonged Rapamycin Treatment Inhibits mTORC2 Assembly and Akt/PKB. Mol Cell. 2006 Apr 21;22(2):10–0.

15. Ma XM, Blenis J. Molecular mechanisms of mTOR-mediated translational control. Nat. Rev. Mol. Cell Biol. 2009 May;10(5):307–18.

16. Hsieh AC, Costa M, Zollo O, Davis C, Feldman ME, Testa JR, et al. Genetic Dissection of the Oncogenic mTOR Pathway Reveals Druggable Addiction to Translational Control via 4EBP-eIF4E. Cancer Cell. 2010 Mar 16;17(3):249–61.

17. Wang BT, Ducker GS, Barczak AJ, Barbeau R, Erle DJ, Shokat KM. The mammalian target of rapamycin regulates cholesterol biosynthetic gene expression and exhibits a rapamycin-resistant transcriptional profile. Proc Natl Acad Sci USA. 2011 Aug 29;108(37):15201–6.

18. Düvel K, Yecies JL, Menon S, Raman P, Lipovsky AI, Souza AL, et al. Activation of a metabolic gene regulatory network downstream of mTOR complex 1. Mol Cell. 2010 Jul 30;39(2):171–83.

19. Ganley IG, Lam DH, Wang J, Ding X, Chen S, Jiang X. ULK1.ATG13.FIP200 complex mediates mTOR signaling and is essential for autophagy. J Biol Chem. 2009 May 1;284(18):12297–305.

20. Jung CH, Jun CB, Ro SH, Kim YM, Otto NM, Cao J, et al. ULK-Atg13-FIP200 Complexes Mediate mTOR Signaling to the Autophagy Machinery. Mol Biol Cell. 2009 Feb 11;20(7):1992–2003.

21. Hosokawa N, Hara T, Kaizuka T, Kishi C, Takamura A, Miura Y, et al. Nutrient- dependent mTORC1 association with the ULK1-Atg13-FIP200 complex required for autophagy. Mol Biol Cell. 2009 Apr;20(7):1981–91.

22. Sarbassov DD, Guertin DA, Ali SM, Sabatini DM. Phosphorylation and regulation of Akt/PKB by the rictor-mTOR complex. Science. 2005;307(5712):1098–101.

23. García-Martínez JM, Alessi DR. mTOR complex 2 (mTORC2) controls hydrophobic motif phosphorylation and activation of serum- and glucocorticoid-induced protein kinase 1 (SGK1). Biochem J. 2008 Dec 15;416(3):375–85.

24. Yang H, Rudge DG, Koos JD, Vaidialingam B, Yang HJ, Pavletich NP. mTOR kinase structure, mechanism and regulation. Nature. Nature Publishing Group; 2013 May 1;:1– 8.

25. Schalm SSS, Blenis JJ. Identification of a Conserved Motif Required for mTOR

60 Signaling. Curr Biol. 2002 Apr 16;12(8):8–8.

26. Yip CK, Murata K, Walz T, Sabatini DM, Kang SA. Structure of the Human mTOR Complex I and Its Implications for Rapamycin Inhibition. Mol Cell. 2010 Jun;38(5):768–74.

27. André C, Cota D. Coupling nutrient sensing to metabolic homoeostasis: the role of the mammalian target of rapamycin complex 1 pathway. Proc Nutr Soc. 2012 Nov;71(4):502–10.

28. Shaw RJ, Bardeesy N, Manning BD, Lopez L, Kosmatka M, DePinho RA, et al. The LKB1 tumor suppressor negatively regulates mTOR signaling. Cancer Cell. 2004 Jul;6(1):91–9.

29. Gwinn DM, Shackelford DB, Egan DF, Mihaylova MM, Mery A, Vasquez DS, et al. AMPK phosphorylation of raptor mediates a metabolic checkpoint. Mol Cell. 2008 Apr 25;30(2):214–26.

30. Sancak Y, Bar-Peled L, Zoncu R, Markhard AL, Nada S, Sabatini DM. Ragulator-Rag complex targets mTORC1 to the lysosomal surface and is necessary for its activation by amino acids. Cell. 2010 Apr 16;141(2):290–303.

31. Zoncu R, Bar-Peled L, Efeyan A, Wang S, Sancak Y, Sabatini DM. mTORC1 Senses Lysosomal Amino Acids Through an Inside-Out Mechanism That Requires the Vacuolar H+-ATPase. Science. 2011 Nov 3;334(6056):678–83.

32. Efeyan A, Zoncu R, Chang S, Gumper I, Snitkin H, Wolfson RL, et al. Regulation of mTORC1 by the Rag GTPases is necessary for neonatal autophagy and survival. Nature. 2012 Dec 23.

33. Shaw RJ, Cantley LC. Ras, PI(3)K and mTOR signalling controls tumour cell growth. Nature. 2006 May 25;441(7092):424–30.

34. Manning BD, Cantley LC. AKT/PKB signaling: navigating downstream. Cell. 2007 Jun 29;129(7):1261–74.

35. Cantley LC. The Phosphoinositide 3-Kinase Pathway. Science. 2002 May 31;296(5573):1655–7.

36. Huang C-H, Mandelker D, Schmidt-Kittler O, Samuels Y, Velculescu VE, Kinzler KW, et al. The structure of a human p110alpha/p85alpha complex elucidates the effects of oncogenic PI3Kalpha mutations. Science. 2007 Dec 14;318(5857):1744–8.

37. Dann SG, Selvaraj A, Thomas G. mTOR Complex1-S6K1 signaling: at the crossroads of obesity, diabetes and cancer. Trends Mol Med. 2007 Jun 1;13(6):252–9.

38. Inoki K, Corradetti MN, Guan K-L. Dysregulation of the TSC-mTOR pathway in human disease. Nat Genet. 2005;37(1):19–24.

61 39. Sato T, Nakashima A, Guo L, Coffman K, Tamanoi F. Single amino-acid changes that confer constitutive activation of mTOR are discovered in human cancer. Oncogene. 2010 May 6;29(18):2746–52.

40. Mao J-H, Kim I-J, Wu D, Climent J, Kang HC, DelRosario R, et al. FBXW7 targets mTOR for degradation and cooperates with PTEN in tumor suppression. Science. 2008 Sep 12;321(5895):1499–502.

41. Menon S, Manning BD. Common corruption of the mTOR signaling network in human tumors. Oncogene. 2008 Dec 1;27 Suppl 2:S43–51.

42. Zhao L, Vogt PK. Hot-spot mutations in p110alpha of phosphatidylinositol 3-kinase (pI3K): differential interactions with the regulatory subunit p85 and with RAS. Cell Cycle. 2010 Feb 1;9(3):596–600.

43. Zhang Y-J, Dai Q, Sun D-F, Xiong H, Tian X-Q, Gao F-H, et al. mTOR signaling pathway is a target for the treatment of colorectal cancer. Ann Surg Oncol. 2009 Sep 1;16(9):2617–28.

44. Krueger DA, Care MM, Holland K, Agricola K, Tudor C, Mangeshkar P, et al. Everolimus for subependymal giant-cell astrocytomas in tuberous sclerosis. N Engl J Med. 2010 Nov 4;363(19):1801–11.

45. Iyer G, Hanrahan AJ, Milowsky MI, Al-Ahmadie H, Scott SN, Janakiraman M, et al. Genome Sequencing Identifies a Basis for Everolimus Sensitivity. Science. 2012 Aug 23;:–.

46. O'Reilly T, McSheehy PM. Biomarker Development for the Clinical Activity of the mTOR Inhibitor Everolimus (RAD001): Processes, Limitations, and Further Proposals. Transl Oncol. 2010 Apr 1;3(2):65–79.

47. Araki K, Ellebedy AH, Ahmed R. TOR in the immune system. Current Opinion in Cell Biology. 2011 Dec;23(6):707–15.

48. Struijk GH, Minnee RC, Koch SD, Zwinderman AH, van Donselaar-van der Pant KAMI, Idu MM, et al. Maintenance immunosuppressive therapy with everolimus preserves humoral immune responses. Kidney International. Nature Publishing Group; 2010 Aug 11;78(9):934–40.

49. Di Paolo S, Teutonico A, Leogrande D, Capobianco C, Schena PF. Chronic inhibition of mammalian target of rapamycin signaling downregulates insulin receptor substrates 1 and 2 and AKT activation: A crossroad between cancer and diabetes? J Am Soc Nephrol. 2006 Aug;17(8):2236–44.

50. Teutonico A, Schena PF, Di Paolo S. Glucose metabolism in renal transplant recipients: effect of calcineurin inhibitor withdrawal and conversion to sirolimus. J Am Soc Nephrol. 2005 Oct;16(10):3128–35.

62 51. Johnston O, Rose CL, Webster AC, Gill JS. Sirolimus is associated with new-onset diabetes in kidney transplant recipients. J Am Soc Nephrol. 2008 Jul;19(7):1411–8.

52. Lamming DW, Ye L, Katajisto P, Goncalves MD, Saitoh M, Stevens DM, et al. Rapamycin-induced insulin resistance is mediated by mTORC2 loss and uncoupled from longevity. Science. 2012 Mar 30;335(6076):1638–43.

53. Motzer RJ, Escudier B, Oudard S, Hutson TE, Porta C, Bracarda S, et al. Efficacy of everolimus in advanced renal cell carcinoma: a double-blind, randomised, placebo- controlled phase III trial. Lancet. 2008 Aug 9;372(9637):449–56.

54. Hudes G, Carducci M, Tomczak P, Dutcher J, Figlin R, Kapoor A, et al. Temsirolimus, interferon alfa, or both for advanced renal-cell carcinoma. N Engl J Med. 2007 May 31;356(22):2271–81.

55. Pavel MEM, Hainsworth JDJ, Baudin EE, Peeters MM, Hörsch DD, Winkler RER, et al. Everolimus plus octreotide long-acting repeatable for the treatment of advanced neuroendocrine tumours associated with carcinoid syndrome (RADIANT-2): a randomised, placebo-controlled, phase 3 study. The Lancet. 2011 Dec 10;378(9808):2005–12.

56. Yao JC, Phan AT, Chang DZ, Wolff RA, Hess K, Gupta S, et al. Efficacy of RAD001 (everolimus) and octreotide LAR in advanced low- to intermediate-grade neuroendocrine tumors: results of a phase II study. Journal of Clinical Oncology. 2008 Sep 10;26(26):4311–8.

57. Baselga J, Campone M, Piccart M, Burris HA, Rugo HS, Sahmoud T, et al. Everolimus in postmenopausal hormone-receptor-positive advanced breast cancer. N Engl J Med. 2012 Feb 9;366(6):520–9.

58. Thoreen C, Kang S, Chang J, Liu Q, Zhang J, Gao Y, et al. An ATP-competitive mTOR inhibitor reveals rapamycin-insensitive functions of mTORC1. J Biol Chem. 2009 Jan 15.

59. García Martínez JM, Moran J, Clarke RG, Gray A, Cosulich SC, Chresta CM, et al. Ku- 0063794 is a specific inhibitor of the mammalian target of rapamycin (mTOR). Biochem J. 2009 Jun 12;421(1):29–42.

60. Feldman ME, Apsel B, Uotila A, Loewith R, Knight ZA, Ruggero D, et al. Active-Site Inhibitors of mTOR Target Rapamycin-Resistant Outputs of mTORC1 and mTORC2. Plos Biol. 2009 Feb 1;7(2):e1000038–8.

61. O'Reilly KE, Rojo F, She Q-B, Solit D, Mills GB, Smith D, et al. mTOR inhibition induces upstream receptor tyrosine kinase signaling and activates Akt. Cancer Res. 2006 Feb 1;66(3):1500–8.

62. Rodrik-Outmezguine VS, Chandarlapaty S, Pagano NC, Poulikakos PI, Scaltriti M, Moskatel E, et al. mTOR Kinase Inhibition Causes Feedback-Dependent Biphasic 63 Regulation of AKT Signaling. Cancer Discovery. 2011 Aug 15;1(3):248–59.

63. Cloughesy TF, Yoshimoto K, Nghiemphu P, Brown K, Dang J, Zhu S, et al. Antitumor activity of rapamycin in a Phase I trial for patients with recurrent PTEN-deficient glioblastoma. PLoS Med. 2008 Jan 22;5(1):e8.

64. Zhang C, Habets G, Bollag G. Interrogating the kinome. Nat Biotechnol. 2011 Nov 1;29(11):981–3.

65. Liu Q, Kirubakaran S, Hur W, Niepel M, Westover K, Thoreen CC, et al. Kinome-wide selectivity profiling of ATP-competitive mTOR (mammalian target of rapamycin) inhibitors and characterization of their binding kinetics. J Biol Chem. 2012 Jan 5;:–.

66. Kantarjian H, Sawyers C, Hochhaus A, Guilhot F, Schiffer C, Gambacorti-Passerini C, et al. Hematologic and cytogenetic responses to imatinib mesylate in chronic myelogenous leukemia. N Engl J Med. 2002 Feb 28;346(9):645–52.

67. Chapman PB, Hauschild A, Robert C, Haanen JB, Ascierto P, Larkin J, et al. Improved survival with vemurafenib in melanoma with BRAF V600E mutation. N Engl J Med. 2011 Jun 30;364(26):2507–16.

68. Zhou C, Wu Y-L, Chen G, Feng J, Liu X-Q, Wang C, et al. Erlotinib versus chemotherapy as first-line treatment for patients with advanced EGFR mutation-positive non-small-cell lung cancer (OPTIMAL, CTONG-0802): a multicentre, open-label, randomised, phase 3 study. Lancet Oncology. Elsevier Ltd; 2011 Jul 27;12(8):735–42.

69. Weinstein IB. Cancer: Addiction to oncogenes - The Achilles heal of cancer. Science. 2002;297(5578):63–4.

70. Burris H, Rodon J, Sharma S, Herbst RS. First-in-human phase I study of the oral PI3K inhibitor BEZ235 in patients (pts) with advanced solid tumors. -- Burris et al. 28 (15): 3005 -- ASCO Meeting Abstracts. J Clin …. 2010.

71. Knight ZA, Shokat KM. Features of selective kinase inhibitors. Chem Biol. 2005 Jun;12(6):621–37.

72. Hsieh ACA, Liu YY, Edlind MPM, Ingolia NTN, Janes MRM, Sher AA, et al. The translational landscape of mTOR signalling steers cancer initiation and metastasis. Nature. 2012 May 3;485(7396):55–61.

73. Yu K, Shi C, Toral-Barza L, Lucas J, Shor B, Kim JE, et al. Beyond Rapalog Therapy: Preclinical Pharmacology and Antitumor Activity of WYE-125132, an ATP- Competitive and Specific Inhibitor of mTORC1 and mTORC2. Cancer Res. 2010 Jan 13;70(2):621–31.

74. Gupta M, Hendrickson AEW, Yun SS, Han JJ, Schneider PA, Koh BD, et al. Dual mTORC1/mTORC2 inhibition diminishes Akt activation and induces Puma-dependent apoptosis in lymphoid malignancies. Blood. 2012 Jan 12;119(2):476–87.

64 75. Chresta CM, Davies BR, Hickson I, Harding T, Cosulich S, Critchlow SE, et al. AZD8055 Is a Potent, Selective, and Orally Bioavailable ATP-Competitive Mammalian Target of Rapamycin Kinase Inhibitor with In vitro and In vivo Antitumor Activity. Cancer Res. 2010 Jan 4;70(1):288–98.

76. Janes MRM, Limon JJJ, So LL, Chen JJ, Lim RJR, Chavez MAM, et al. Effective and selective targeting of leukemia cells using a TORC1/2 kinase inhibitor. Nat Med. 2010 Feb 1;16(2):205–13.

77. Hoang B, Frost P, Shi Y, Belanger E, Benavides A, Pezeshkpour G, et al. Targeting TORC2 in multiple myeloma with a new mTOR kinase inhibitor. Blood. 2010 Nov 25;116(22):4560–8.

78. Weigelt B, Warne PH, Downward J. PIK3CA mutation, but not PTEN loss of function, determines the sensitivity of breast cancer cells to mTOR inhibitory drugs. Oncogene. 2011 Jul 21;30(29):3222–33.

79. Krause DS, Van Etten RA. Tyrosine kinases as targets for cancer therapy. N Engl J Med. 2005 Jul 14;353(2):172–87.

80. Zhang J, Yang PL, Gray NS. Targeting cancer with small molecule kinase inhibitors. Nat Rev Cancer. 2009;9(1):28–39.

81. Luo J, Solimini NL, Elledge SJ. Principles of cancer therapy: oncogene and non- oncogene addiction. Cell. 2009 Mar 6;136(5):823–37.

82. Bignell GR, Greenman CD, Davies H, Butler AP, Edkins S, Andrews JM, et al. Signatures of mutation and selection in the cancer genome. Nature. Nature Publishing Group; 2010 Feb 8;463(7283):893–8.

83. Yaffe MB. The Scientific Drunk and the Lamppost: Massive Sequencing Efforts in Cancer Discovery and Treatment. Science signaling. 2013 Apr 2;6(269):pe13–3.

84. García-Martínez J, Moran J, Clarke R, Gray A, Cosulich S, Chresta C, et al. Ku- 0063794 is a specific inhibitor of the mammalian target of rapamycin (mTOR). Biochem J. 2009 Apr 29.

85. Network TCGA. Comprehensive molecular characterization of human colon and rectal cancer. Nature. Nature Publishing Group; 2012 Jul 10;487(7407):330–7.

86. Lièvre A, Bachet J-B, Boige V, Cayre A, Le Corre D, Buc E, et al. KRAS mutations as an independent prognostic factor in patients with advanced colorectal cancer treated with cetuximab. J Clin Oncol. 2008 Jan 20;26(3):374–9.

87. Bertotti A, Migliardi G, Galimi F, Sassi F, Torti D, Isella C, et al. A Molecularly Annotated Platform of Patient-Derived Xenografts (“Xenopatients”) Identifies HER2 as an Effective Therapeutic Target in Cetuximab-Resistant Colorectal Cancer. Cancer Discovery. 2011 Nov 1;1(6):508–23.

65 88. Barretina J, Caponigro G, Stransky N, Venkatesan K, Margolin AA, Kim S, et al. The Cancer Cell Line Encyclopedia enables predictive modelling of anticancer drug sensitivity. Nature. Nature Publishing Group; 2012 Mar 19;483(7391):603–307.

89. McDermott U, Sharma SV, Dowell L, Greninger P, Montagut C, Lamb J, et al. Identification of genotype-correlated sensitivity to selective kinase inhibitors by using high-throughput tumor cell line profiling. Proc Natl Acad Sci USA. 2007 Dec 11;104(50):19936–41.

90. Purnak T, Ozaslan E, Efe C. Molecular basis of colorectal cancer. N Engl J Med. 2010 Apr 1;362(13):1246; authorreply1246–7.

91. Knight ZA, Gonzalez B, Feldman ME, Zunder ER, Goldenberg DD, Williams O, et al. A pharmacological map of the PI3-K family defines a role for p110alpha in insulin signaling. Cell. 2006 May 19;125(4):733–47.

92. Apsel B, Blair JA, Gonzalez B, Nazif TM, Feldman ME, Aizenstein B, et al. Targeted polypharmacology: discovery of dual inhibitors of tyrosine and phosphoinositide kinases. Nat Chem Biol. 2008 Oct 12;4(11):691–9.

93. Daniel VC, Marchionni L, Hierman JS, Rhodes JT, Devereux WL, Rudin CM, et al. A primary xenograft model of small-cell lung cancer reveals irreversible changes in gene expression imposed by culture in vitro. Cancer Res. 2009 Apr 15;69(8):3364–73.

94. Talmadge JE, Singh RK, Fidler IJ, Raz A. Murine models to evaluate novel and conventional therapeutic strategies for cancer. Am J Pathol. 2007 Mar 1;170(3):793–804.

95. Voskoglou-Nomikos T, Pater JL, Seymour L. Clinical predictive value of the in vitro cell line, human xenograft, and mouse allograft preclinical cancer models. Clin Cancer Res. 2003 Sep 15;9(11):4227–39.

96. Jimeno A, Feldmann G, Suárez-Gauthier A, Rasheed Z, Solomon A, Zou G-M, et al. A direct pancreatic cancer xenograft model as a platform for cancer stem cell therapeutic development. Mol Cancer Ther. 2009 Feb 1;8(2):310–4.

97. Rubio-Viqueira B, Jimeno A, Cusatis G, Zhang X, Iacobuzio-Donahue C, Karikari C, et al. An in vivo platform for translational drug development in pancreatic cancer. Clin Cancer Res. 2006 Aug 1;12(15):4652–61.

98. Fichtner I, Slisow W, Gill J, Becker M, Elbe B, Hillebrand T, et al. Anticancer drug response and expression of molecular markers in early-passage xenotransplanted colon carcinomas. Eur J Cancer. 2004;40(2):298–307.

99. Cancer Facts and Figures 2013. Atlanta: American Cancer Society; 2013. pp. 1–64.

100. House MG, Ito H, Gönen M, Fong Y, Allen PJ, DeMatteo RP, et al. Survival after hepatic resection for metastatic colorectal cancer: trends in outcomes for 1,600 patients during two decades at a single institution. J. Am. Coll. Surg. 2010 May;210(5):744–52–

66 752–5.

101. Tomlinson JSJ, Jarnagin WRW, DeMatteo RPR, Fong YY, Kornprat PP, Gonen MM, et al. Actual 10-year survival after resection of colorectal liver metastases defines cure. CORD Conference Proceedings. 2007 Oct 10;25(29):4575–80.

102. Hanahan D, Weinberg RA. Hallmarks of Cancer: The Next Generation. Cell. 2011 Mar 4;144(5):646–74.

103. Gerlinger M, Rowan AJ, Horswell S, Larkin J, Endesfelder D, Gronroos E, et al. Intratumor Heterogeneity and Branched Evolution Revealed by Multiregion Sequencing. N Engl J Med. 2012 Mar 8;366(10):883–92.

104. Tentler JJ, Tan AC, Weekes CD, Jimeno A, Leong S, Pitts TM, et al. Patient-derived tumour xenografts as models for oncology drug development. Nat Rev Clin Oncol. 2012 Jun;9(6):338–50.

105. Houghton JA, Taylor DM. Maintenance of biological and biochemical characteristics of human colorectal tumours during serial passage in immune-deprived mice. British Journal of Cancer. 1978 Feb 1;37(2):199–212.

106. Pickard RG, Cobb LM, Steel GG. The growth kinetics of xenografts of human colorectal tumours in immune deprived mice. British Journal of Cancer. 1975;31(1):36–45.

107. Kuperwasser C, Chavarria T, Wu M, Magrane G, Gray JW, Carey L, et al. Reconstruction of functionally normal and malignant human breast tissues in mice. Proc Natl Acad Sci USA. 2004 Apr 6;101(14):4966–71.

108. Kwak EL, Bang Y-J, Camidge DR, Shaw AT, Solomon B, Maki RG, et al. Anaplastic lymphoma kinase inhibition in non-small-cell lung cancer. N Engl J Med. 2010 Oct 28;363(18):1693–703.

109. Jänne PA, Gray N, Settleman J. Factors underlying sensitivity of cancers to small- molecule kinase inhibitors. Nat Rev Drug Discov. 2009 Mar 8;8(9):709–23.

110. Joseph EW, Pratilas CA, Poulikakos PI, Tadi M, Wang W, Taylor BS, et al. The RAF inhibitor PLX4032 inhibits ERK signaling and tumor cell proliferation in a V600E BRAF-selective manner. Proc Natl Acad Sci USA. 2010 Aug 17;107(33):14903–8.

111. Shaw RJ, Cantley LC. Ras, PI(3)K and mTOR signalling controls tumour cell growth. Nature. 2006 May 25;441(7092):424–30.

112. Zhang Y-J, Duan Y, Zheng XFS. Targeting the mTOR kinase domain: the second generation of mTOR inhibitors. Drug Discovery Today. 2011 Apr 1;16(7-8):325–31.

113. Dowling RJOR, Topisirovic II, Alain TT, Bidinosti MM, Fonseca BDB, Petroulakis EE, et al. mTORC1-mediated cell proliferation, but not cell growth, controlled by the 4E- BPs. Science. 2010 May 28;328(5982):1172–6.

67 114. Markowitz SD, Bertagnolli MM. Molecular Basis of Colorectal Cancer. N Engl J Med. 2009 Dec 17;361(25):2449–60.

115. Blaser B, Waselle L, Dormond-Meuwly A, Dufour M, Roulin D, Demartines N, et al. Antitumor activities of ATP-competitive inhibitors of mTOR in colon cancer cells. BMC Cancer. 2012;12:86.

116. Atreya CE, Ducker GS, Feldman ME, Bergsland EK, Warren RS, Shokat KM. Combination of ATP-competitive mammalian target of rapamycin inhibitors with standard chemotherapy for colorectal cancer. Invest New Drugs. 2012 Dec 1;30(6):2219–25.

117. Rong L, Livingstone M, Sukarieh R, Petroulakis E, Gingras A-C, Crosby K, et al. Control of eIF4E cellular localization by eIF4E-binding proteins, 4E-BPs. RNA. 2008 May 29;14(7):1318–27.

118. Carrière A, Cargnello M, Julien L-A, Gao H, Bonneil E, Thibault P, et al. Oncogenic MAPK signaling stimulates mTORC1 activity by promoting RSK-mediated raptor phosphorylation. Curr Biol. 2008 Sep 9;18(17):1269–77.

119. Barrett SD, Bridges AJ, Dudley DT, Saltiel AR, Fergus JH, Flamme CM, et al. The discovery of the benzhydroxamate MEK inhibitors CI-1040 and PD 0325901. Bioorg Med Chem Lett. 2008 Dec 15;18(24):6501–4.

120. Le N-T, Takei Y, Shishido T, Woo C-H, Chang E, Heo K-S, et al. p90RSK Targets the ERK5-CHIP Ubiquitin E3 Ligase Activity in Diabetic Hearts and Promotes Cardiac Apoptosis and Dysfunction. Circ Res. 2012 Feb 17;110(4):536–50.

121. Cohen MS, Zhang C, Shokat KM, Taunton J. Structural bioinformatics-based design of selective, irreversible kinase inhibitors. Science. 2005 May 27;308(5726):1318–21.

122. She Q-B, Halilovic E, Ye Q, Zhen W, Shirasawa S, Sasazuki T, et al. 4E-BP1 is a key effector of the oncogenic activation of the AKT and ERK signaling pathways that integrates their function in tumors. Cancer Cell. 2010 Jul 13;18(1):39–51.

123. Taylor SJ, Resnick RJ, Shalloway D. Nonradioactive determination of Ras-GTP levels using activated ras interaction assay. Meth Enzymol. 2001;333:333–42.

124. Bohanes P, LaBonte MJ, Winder T, Lenz H-J. Predictive molecular classifiers in colorectal cancer. Semin. Oncol. 2011 Aug;38(4):576–87.

125. Zhao L, Vogt PK. Helical domain and kinase domain mutations in p110alpha of phosphatidylinositol 3-kinase induce gain of function by different mechanisms. Proc Natl Acad Sci USA. 2008 Feb 19;105(7):2652–7.

126. Samuels Y, Diaz LA, Schmidt-Kittler O, Cummins JM, Delong L, Cheong I, et al. Mutant PIK3CA promotes cell growth and invasion of human cancer cells. Cancer Cell. 2005 Jun 1;7(6):561–73.

68 127. Janku F, Wheler JJ, Naing A, Falchook GS, Hong DS, Stepanek V, et al. PIK3CA mutation H1047R is associated with response to PI3K/AKT/mTOR signaling pathway inhibitors in early phase clinical trials. Cancer Res. 2012 Oct 12.

128. Ilic N, Utermark T, Widlund HR, Roberts TM. PI3K-targeted therapy can be evaded by gene amplification along the MYC-eukaryotic translation initiation factor 4E (eIF4E) axis. Proc Natl Acad Sci USA. 2011 Aug 29.

129. Alain T, Morita M, Fonseca BD, Yanagiya A, Siddiqui N, Bhat M, et al. eIF4E/4E-BP ratio predicts the efficacy of mTOR targeted therapies. Cancer Res. 2012 Oct 24.

130. Brough R, Frankum JR, Costa-Cabral S, Lord CJ, Ashworth A. Searching for synthetic lethality in cancer. Curr Opin Genet Dev. 2011 Feb 1;21(1):34–41.

131. Knight ZA, Lin H, Shokat KM. Targeting the cancer kinome through polypharmacology. Nat Rev Cancer. 2010 Feb 1;10(2):130–7.

132. Liao X, Lochhead P, Nishihara R, Morikawa T, Kuchiba A, Yamauchi M, et al. Aspirin Use, Tumor PIK3CAMutation, and Colorectal-Cancer Survival. N Engl J Med. 2012 Oct 25;367(17):1596–606.

133. Zunder ER, Knight ZA, Houseman BT, Apsel B, Shokat KM. Discovery of drug- resistant and drug-sensitizing mutations in the oncogenic PI3K isoform p110 alpha. Cancer Cell. 2008 Aug 12;14(2):180–92.

134. Fumagalli D, Gavin PG, Taniyama Y, Kim S-I, Choi H-J, Paik S, et al. A rapid, sensitive, reproducible and cost-effective method for mutation profiling of colon cancer and metastatic lymph nodes. BMC Cancer. 2010 Mar 16;10(1):101–1.

135. Pinheiro JC, Bates DM. Mixed Effects Models in S and S-Plus. New York, NY: Springer; 2000.

136. Vogelstein B, Papadopoulos N, Velculescu VE, Zhou S, Diaz LA, Kinzler KW. Cancer Genome Landscapes. Science. 2013 Mar 28;339(6127):1546–58.

137. Jiang Q, Weiss JM, Back T, Chan T, Ortaldo JR, Guichard S, et al. mTOR Kinase Inhibitor AZD8055 Enhances the Immunotherapeutic Activity of an Agonist CD40 Antibody in Cancer Treatment. Cancer Res. 2011 Jun 14;71(12):4074–84.

69 Appendix 1: A cell growth screen of mTOR inhibitors rapamycin and PP242

Cell Line Organ of Origin Mutations 1 µM Rapamycin 500 nM PP242 5637 Bladder TP53 : RB1 0.815 0.643 1108-MP2 Pancreas 0.545 1205Lu Skin BRAF 0.762 0.807 1321N1 Brain 0.685 0.567 143B Bone 0.816 0.69 143B PML BK TK Bone 0.693 0.945 1A6 Bladder 0.862 0.857 201T Lung:NSCLC 0.776 0.546 22RV1 Prostate PIK3CA 0.484 0.609 23132/87 Stomach (no mutations in Sanger 50) 0.561 0.294 273T Lung:NSCLC 0.89 0.497 42-MG-BA Brain 0.46 0.486 4244135 GCT 0.543 451Lu Skin 0.54 0.495 609MP9 Pancreas 0.731 0.783 617-MP17 Pancreas 0.527 621-101 Kidney 0.927 639-V Bladder RB1 : PTEN : PIK3CA : CDKN2A : TP53 0.95 0.761 647-V Bladder MAP2K4 : RB1 : TP53 0.745 0.662 722-MP14 Pancreas 0.845 769-P Kidney CDKN2A 0.805 0.611 786-O Kidney VHL : PTEN : TP53 : CDKN2A 0.669 0.353 8305C Thyroid TP53 : BRAF 0.7 0.403 8505C Thyroid BRAF : TP53 : NF2 : CDKN2A 0.792 0.539 950MP11 Pancreas 0.763 A-204 Muscle (No mutations in Sanger 50) 0.337 0.25 A-375 Skin BRAF : CDKN2A 0.998 0.519 A-427 Lung:NSCLC STK11 : KRAS : CDKN2A : CTNNB1 0.822 0.515 A13A Pancreas 0.293 0.717 A172 Brain PTEN : CDKN2A 0.542 0.655 A2.1 Pancreas 0.572 0.341 A2058 Skin PTEN : BRAF : TP53 : RB1 0.965 0.726 A2780 Ovary PTEN 1.067 0.425 A2780ADR Ovary 0.521 0.645 A373-C6 Skin BRAF 0.774 0.633 A375.S2 Skin BRAF : CDKN2A 0.961 0.635 A431 Skin TP53 0.926 0.824 A549 Lung:NSCLC CDKN2A : STK11 : KRAS 0.753 0.398 A673 Muscle BRAF : CDKN2A : TP53 0.48 0.444

70 Cell Line Organ of Origin Mutations 1 µM Rapamycin 500 nM PP242 ABC-1 Lung:NSCLC TP53 0.847 0.944 ACC112 Head & Neck 0.673 0.679 ACC2 Head & Neck 0.583 0.543 ACC3 Head & Neck 0.647 0.468 ACCS Head & Neck 0.65 0.981 ACHN Kidney NF2 : CDKN2A 0.857 0.307 AGS Stomach KRAS : CTNNB1 : CDH1 : PIK3CA 0.568 0.793 AN3CA Uterus PTEN : TP53 0.504 0.082 ASH-3 Thyroid 0.791 0.746 AsPC-1 Pancreas CDKN2A : MAP2K4 : KRAS : TP53 1.033 0.492 AU565 Breast TP53 0.878 0.453 AZ-521 Stomach 0.488 0.438 B-CPAP Thyroid TP53 : BRAF 0.677 0.359 BE(2)-C Nervous System 0.633 0.31 BEAS-2B 0.748 BEN Lung:NSCLC TP53 0.506 0.16 BeWo Miscellaneous 0.622 0.918 BFTC-905 Bladder CDKN2A : NRAS : TP53 0.868 0.428 BFTC-909 Kidney PIK3CA : TP53 : CDKN2A 0.725 0.459 BHT-101 Thyroid TP53 : BRAF 0.869 0.377 BHY Head & Neck TP53 : CDKN2A 0.772 0.344 BICR 10 Head & Neck 0.723 0.444 BICR 22 Head & Neck 0.599 0.764 BICR 31 Head & Neck 0.777 0.827 BICR 78 Head & Neck 0.624 0.634 BPH-1 Prostate 0.781 BT-20 Breast PIK3CA : CDKN2A : TP53 0.561 0.638 BT-474 0.793 BT-483 Breast 0.744 0.443 BT-549 Breast RB1 : TP53 : PTEN 0.935 1.242 BT-B 0.777 BU-ML Skin 0.468 0.485 Bu25 TK- 0.942 BxPC-3 Pancreas CDKN2A : SMAD4 : TP53 : MAP2K4 0.848 0.471 C-33 A Cervix RB1 : TP53 : PIK3CA : PTEN 0.77 0.636 C-4 I Cervix 0.697 0.692 C-4 II Cervix (No mutations in Sanger 50) 0.724 1.028 C170 Intestine 0.784 0.847 C2BBe1 Intestine 0.644 C32 Skin PTEN: BRAF : CDKN2A 0.58 0.63 C3A Liver NRAS 0.943 0.661

71 Cell Line Organ of Origin Mutations 1 µM Rapamycin 500 nM PP242 Ca Ski Cervix PIK3CA 0.679 0.523 Ca9-22 0.556 CA9-22 Head & Neck TP53 0.411 Caco-2 Intestine 0.799 0.663 Caki-1 Kidney CDKN2A 0.879 0.594 CAL 27 Head & Neck SMAD4 : TP53 : CDKN2A 0.795 0.623 CAL-120 Breast TP53 0.632 0.688 CAL-12T Lung:NSCLC CDKN2A : TP53 : BRAF 0.559 0.285 CAL-148 0.661 CAL-29 Bladder 0.539 0.697 CAL-33 Head & Neck TP53 : SMAD4 : CDKN2A : PIK3CA 0.78 0.712 CAL-39 Cervix CDKN2A : TP53 0.327 0.464 CAL-51 Breast PTEN : PIK3CA 0.77 0.634 CAL-54 Kidney CDKN2A 0.738 0.514 CAL-62 Thyroid NF2 : TP53 : KRAS : CDKN2A 0.968 0.387 CAL-72 Bone CDKN2A 0.648 0.761 CAL-78 Bone 0.775 0.812 CAL-85-1 Breast RB1 : TP53 0.685 0.743 Calu-1 Lung:NSCLC KRAS 0.632 0.875 Calu-3 Lung:NSCLC TP53 0.603 0.379 CAMA-1 Breast VHL : TP53 : PTEN 0.504 0.467 Caov-3 Ovary TP53 : STK11 0.629 0.592 Caov-4 Ovary 0.77 Capan-1 Pancreas SMAD4 : BRCA2 : CDKN2A : KRAS : TP53 0.877 0.557 Capan-2 Pancreas KRAS 1.031 0.704 CaR-1 Intestine CDKN2A : STK11 : TP53 0.612 0.499 CCF-STTG1 Brain PTEN 0.619 0.597 CCK-81 Intestine 0.956 0.596 CFPAC-1 Pancreas SMAD4 : KRAS : TP53 0.871 0.68 Ch8 Bone 0.733 ChaGo-K-1 Lung:NSCLC TP53 0.485 0.344 CHL-1 Skin TP53 : CDKN2A 0.352 0.206 CHP-212 Nervous System NRAS 0.691 0.634 CHSA 0011 Bone 0.698 0.769 CHSA 0108 Bone 0.53 0.568 CHSA8926 Bone 0.45 CL-11 0.772 CL-14 1.133 CL-34 0.729 CL-40 0.84 CoCM-1 Intestine APC : SMAD4 : TP53 : PIK3CA 0.76 0.571

72 Cell Line Organ of Origin Mutations 1 µM Rapamycin 500 nM PP242 COLO 201 Intestine 0.789 0.903 COLO 205 Intestine TP53 : BRAF : APC : SMAD4 0.816 0.475 COLO 320DM Intestine 1.209 0.686 COLO 741 Intestine BRAF : TP53 : CDKN2A 0.571 0.685 COLO 792 Skin (No mutations in Sanger 50) 0.659 0.867 COLO 853 Skin BRAF 0.617 0.467 COLO 857 Skin 0.832 0.442 COLO 858 Skin BRAF 0.857 0.597 COLO-206F Intestine 0.717 0.611 COLO-320 1.077 COLO-678 Intestine KRAS : CDKN2A : APC : SMAD4 0.904 0.928 COLO-679 Skin BRAF : CDKN2A 0.927 0.471 COLO-680N Esophagus TP53 : CDKN2A 0.655 0.8 COLO-699 Lung:NSCLC 0.663 0.36 COLO-783 Skin BRAF 1.165 0.883 COLO-818 Skin BRAF 0.782 0.501 COLO-824 0.917 COLO-849 Skin BRAF 1.047 0.447 COR-L 105 Lung:NSCLC (No mutations in Sanger 50) 0.896 0.463 COR-L 23/CPR Lung:NSCLC 0.836 0.487 COR-L23 Lung:NSCLC KRAS : CDKN2A 0.591 0.468 CS1 Bone 0.552 0.75 CS1R Bone 0.363 0.223 DAN-G Pancreas 0.74 0.257 Daoy Brain TP53 : CDKN2A 0.412 0.526 DBTRG-05MG Brain PTEN : CDKN2A : BRAF 0.452 0.691 Detroit 562 Head & Neck CDKN2A : TP53 : PIK3CA 0.441 0.334 DK-MG Brain PIK3CA : CDKN2A 0.674 0.703 DLD-1 Intestine 0.493 0.726 DMS 273 Lung RB1 : TP53 0.428 0.679 DMS 53 Lung STK11 : TP53 0.658 0.553 DOK Head & Neck TP53 0.653 0.454 DoTc2 4510 Cervix BRCA2 0.791 1.048 DU 145 Prostate RB1 : TP53 : STK11 : CDKN2A 0.451 0.678 DV-90 Lung:NSCLC KRAS : CDKN2A : APC 1.039 0.4 EBC-1 Lung:NSCLC 0.451 0.545 EFE-184 Uterus TP53 0.522 0.667 EFM-19 Breast PIK3CA : CDKN2A : MAP2K4 0.804 0.257 EFM-192A Breast 0.458 0.574 EFM-192B Breast 0.798 0.495 EFM-192C 0.793

73 Cell Line Organ of Origin Mutations 1 µM Rapamycin 500 nM PP242 EFO-21 Ovary TP53 0.854 0.654 EFO-27 Ovary PTEN: TP53 0.596 0.398 EGI-1 Liver TP53 : KRAS : CDKN2A 0.644 0.291 EJ138 Bladder 0.874 0.75 EN Uterus 0.574 0.341 EPLC-272H Lung:NSCLC TP53 0.749 0.758 ES-2 Ovary 0.654 ESS-1 Uterus FBXW7 : TP53 : PIK3CA : RB1 0.724 0.638 EVSA-T 0.743 EW 7476 Bone 0.688 0.77 FaDu Head & Neck CDKN2A : SMAD4 : TP53 0.761 0.798 FLYA13 Miscellaneous 0.393 0.451 fR2 1.061 FTC-133 Thyroid NF1 : FLCN : MSH6 : PTEN : TP53 0.666 0.923 FTC-238 Thyroid 0.824 0.654 FU-OV-1 Ovary 0.852 0.727 FU97 Stomach 0.752 0.357 G-292 Clone A141B1 Bone 0.913 0.399 G-361 Skin CDKN2A : BRAF : STK11 0.841 0.504 G-401 Kidney (No mutations in Sanger 50) 0.878 0.255 G-402 Kidney (No mutations in Sanger 50) 0.554 0.148 G-MEL Skin 0.504 0.509 GAMG Brain TP53 : CDKN2A : SMAD4 0.469 0.397 GCT Miscellaneous BRAF : TP53 0.72 0.413 GMS-10 Brain RB1 : TP53 0.837 0.7592 GOS-3 Brain 0.756 0.772 GP5d Intestine APC : KRAS : PIK3CA 1.094 0.618 GTL-16 Stomach 0.693 0.32 H-EMC-SS 0.515 H2052 Lung 0.654 0.402 H2369 Lung 0.559 0.542 H2373 Lung 0.624 0.364 H2461 Lung 0.572 0.451 H2591 Lung 0.379 H2595 Lung 0.651 0.642 H2596 Lung 0.755 0.47 H2691 0.409 H2722 Lung 0.544 0.599 H2731 Lung 0.632 0.587 H2795 Lung 0.517 0.391 H28 Lung 0.548 0.646

74 Cell Line Organ of Origin Mutations 1 µM Rapamycin 500 nM PP242 H2803 Lung 0.461 0.524 H2804 Lung 0.703 0.572 H2810 Lung 0.706 H2818 Lung 0.608 H2869 Lung 0.779 0.819 H290 Lung 0.713 0.216 H292 Lung 0.587 0.41 H3118 Head & Neck 0.4 0.553 H3255 Lung:NSCLC 0.898 0.637 H4 Brain PTEN : CDKN2A 0.695 0.292 H513 Lung 0.533 0.607 H69V Lung 0.602 0.817 HARA Lung:NSCLC 0.689 0.632 HBE135-E6E7 0.821 HBE4-E6/E7 0.74 HBE4-E6/E7-C1 0.752 HCC-15 Lung:NSCLC 0.964 0.646 HCC-366 Lung:NSCLC 0.579 0.253 HCC-44 Lung:NSCLC 0.564 0.375 HCC-56 Intestine 1.054 0.719 HCC-78 Lung:NSCLC 0.635 0.388 HCC-827 Lung:NSCLC 0.65 0.634 HCC1143 Breast TP53 0.814 0.832 HCC1395 Breast PTEN : CDKN2A : TP53 0.775 0.66 HCC1419 Breast TP53 0.507 0.356 HCC1428 Breast 0.874 0.475 HCC1569 Breast 0.633 0.518 HCC1806 Breast CDKN2A : TP53 1.027 0.642 HCC1937 Breast TP53 0.708 0.645 HCC1954 Breast PIK3CA : TP53 0.759 0.508 HCC202 Breast 0.709 0.42 HCC38 Breast CDKN2A : TP53 0.711 0.79 HCC70 Breast PTEN : TP53 0.754 0.614 HCE7 Esophagus 0.909 0.504 HCT 116 Intestine KRAS : CTNNB1 : CDKN2A : PIK3CA : BRCA2 : MLH1 0.886 0.763 HCT-15 Intestine APC : TP53 : PIK3CA : BRCA2 : KRAS 0.619 0.747 HCT-8 Intestine 0.881 0.616 HDQ-P1 Breast 0.944 0.697 HE8935 Miscellaneous 0.665 HEC-1 Uterus TP53 : KRAS : PIK3CA 0.668 0.478 HeLa Cervix STK11 0.861 0.964

75 Cell Line Organ of Origin Mutations 1 µM Rapamycin 500 nM PP242 Hep 3B2.1-7 Liver 0.72 0.683 Hep G2 0.895 HEP G2 Liver 0.59 HGC-27 Stomach TP53 : APC : PTEN : PIK3CA 0.447 0.076 HLE Liver TP53 : RB1 0.713 0.472 HLF Liver 0.68 0.492 HMCB Skin 0.779 0.658 HMVII Skin NRAS : BRAF : TP53 0.639 0.551 HN Head & Neck 0.405 0.676 HO-1-N-1 Head & Neck TP53 0.386 0.654 HO-1-u-1 Head & Neck 0.794 0.676 HOP92 0.888 HOS Bone TP53 : CDKN2A 0.892 0.68 HPAC Pancreas 0.936 0.61 HPAF-II Pancreas TP53 : KRAS : CDKN2A 1.009 0.56 HRT-18 1.067 Hs 257.T Intestine 1.01 0.7623 Hs 387.T Bone 0.905 0.807 Hs 417.Lu 0.396 Hs 578T Breast TP53 : CDKN2A : HRAS 0.728 0.661 Hs 588.T 0.816 HS 588.T Cervix 0.475 Hs 633T Miscellaneous 0.455 0.514 Hs 683 Brain 0.53 0.772 Hs 707(A).T 1.46 Hs 746T Stomach 0.884 0.611 Hs 766T Pancreas 0.71 0.549 Hs 894(B).T Bone 0.455 0.698 Hs 894(E).Lu 0.728 Hs 939.T Skin 0.643 0.707 Hs 940.T Skin NRAS 0.614 0.74 Hs 944.T Skin NRAS 0.804 0.571 HSC-2 Head & Neck PIK3CA : CDKN2A : TP53 0.563 0.622 HSC-3 Head & Neck TP53 : CDKN2A 0.666 0.785 HSC-4 Head & Neck TP53 : CDKN2A 0.745 0.519 HT 1080 Miscellaneous NRAS : CDKN2A 0.489 0.417 HT 1376 Bladder RB1 : TP53 0.868 0.784 HT-1197 Bladder 0.546 0.821 HT-29 Intestine PIK3CA : TP53 : BRAF : SMAD4 : APC 1.2 0.775 HT-3 Cervix TP53 0.606 0.677 HT115 Intestine 0.941 0.709

76 Cell Line Organ of Origin Mutations 1 µM Rapamycin 500 nM PP242 HT55 Intestine APC 0.875 0.515 HTC-C3 Thyroid TP53 : BRAF 0.513 0.368 HuCCT1 Liver TP53 : KRAS 0.669 0.548 huH-1 Liver 0.892 0.516 HUH-6 Clone 5 0.846 HuH-7 Liver TP53 0.841 0.46 HuH28 0.83 HuO-3N1 Bone RB1 : TP53 0.805 0.743 HuO9 Bone RB1 0.804 0.494 HuO9N2 Bone 0.783 0.635 HUP-T3 Pancreas TP53 : CDKN2A : KRAS 0.785 0.593 HUP-T4 Pancreas TP53 : CDKN2A : KRAS 0.777 0.82 IGR-1 Skin BRAF : CDKN2A 0.677 0.803 IGR-37 Skin BRAF 0.479 0.201 IGR-39 Skin BRAF 0.522 0.769 IGROV-1 Ovary SMAD4 : MLH1 : TP53 : BRCA1 : PTEN : MSH6 0.985 0.219 IHH-4 Thyroid 1.079 0.217 IM-95 Stomach 0.834 0.555 IM-95m Stomach 0.982 0.696 IMR-32 Nervous System 0.639 IMR-90 0.573 IPC-298 Skin TP53 : NRAS 0.754 0.474 Ishikawa Uterus 0.598 0.498 Ishikawa (Heraklio) 02 ER-Uterus 0.634 0.867 J82 Bladder TP53 : PIK3CA : RB1 : PTEN 0.779 0.661 JAR 0.705 JHH-1 Liver 0.551 0.49 JHH-2 Liver 0.872 0.558 JHH-4 Liver 0.848 0.52 JHH-6 Liver 0.83 0.531 JHH-7 Liver 0.811 0.387 JHU-011 Head & Neck 0.481 0.44 JHU-013 Head & Neck 0.818 0.826 JHU-019 Head & Neck 0.609 0.339 JHU-022 Head & Neck 0.612 0.612 JHU-028 Head & Neck 0.429 0.607 JHU-028EP Head & Neck 0.777 0.518 JHU-029 Head & Neck 0.662 0.545 JIMT-1 Breast 0.677 0.487 K1 Skin BRAF 0.858 0.978 K19 Skin NRAS 0.615 0.735

77 Cell Line Organ of Origin Mutations 1 µM Rapamycin 500 nM PP242 K2 Skin BRAF 0.817 0.607 K4 Skin BRAF 0.379 0.453 K8 Skin 0.956 0.843 KATO II Stomach 0.611 0.223 KATO III Stomach (No mutations in Sanger 50) 0.676 0.413 KELLY 0.906 KG-1-C Brain 0.784 0.835 KHM-3S 0.576 KHOS-240S Bone 1.003 0.669 KHOS-312H Bone 1.126 0.688 KHOS/NP Bone 0.746 0.579 KMH-2 Thyroid (No mutations in Sanger 50) 0.869 0.501 KMRC-1 Kidney 0.574 0.737 KMRC-20 Kidney 0.716 0.843 KMRM-M1 Kidney 0.903 0.335 KON Head & Neck 0.765 0.554 KOSC-2 cl3-43 Head & Neck TP53 : CDKN2A 1.072 0.523 KP-1N Pancreas 0.843 0.82 KP-1NL Pancreas 0.974 0.593 KP-2 Pancreas 1.073 0.82 KP-3 Pancreas 0.84 0.441 KP-3L Pancreas 0.875 0.471 KP-4 Pancreas CDKN2A : SMAD4 : KRAS 0.966 0.889 KPL-1 Breast 0.647 0.246 KU-19-19 Bladder CDKN2A : NRAS 0.676 0.334 KYSE-140 Esophagus CDKN2A : TP53 0.773 0.728 KYSE-150 Esophagus TP53 0.592 0.444 KYSE-180 Esophagus CDKN2A : TP53 0.742 0.45 KYSE-220 Esophagus 0.857 0.687 KYSE-270 Esophagus STK11 : TP53 0.75 0.664 KYSE-30 Esophagus 0.701 0.86 KYSE-410 Esophagus TP53 : KRAS 0.393 0.463 KYSE-450 Esophagus TP53 : CDKN2A 0.662 0.481 KYSE-50 Esophagus 0.809 0.8675 KYSE-510 Esophagus TP53 : CDKN2A : PIK3CA 0.661 0.559 KYSE-520 Esophagus TP53 0.628 0.541 KYSE-70 Esophagus CDKN2A : TP53 0.863 0.798 LC-1 sq Lung:NSCLC 1.017 0.549 LCLC-103H Lung:NSCLC TP53 : CDKN2A 0.663 0.74 LCLC-97TM1 Lung:NSCLC TP53 : CDKN2A : KRAS 0.805 0.605 LK-2 Lung:NSCLC TP53 : APC : CDKN2A 1.038 0.149

78 Cell Line Organ of Origin Mutations 1 µM Rapamycin 500 nM PP242 LN-18 Brain 0.714 0.751 LN-229 Brain 0.795 0.718 LN-405 Brain TP53 : PTEN : RB1 0.66 0.597 LNZTA3WT11 Brain 0.535 0.783 LNZTA3WT4 Brain 0.711 0.813 LOU-NH91 Lung:NSCLC 0.685 0.686 LoVo Intestine KRAS : APC 0.29 0.421 LS174T Intestine CTNNB1 : KRAS : PIK3CA 1.072 0.753 LS180 Intestine 1.029 0.74 Lu-134-A-H 0.821 Lu-135 Lung TP53 : RB1 0.878 0.386 LU65A Lung:NSCLC TP53 : KRAS : RB1 0.657 0.655 LU65B Lung:NSCLC TP53 : KRAS : RB1 0.742 0.462 LU65C Lung:NSCLC TP53 : KRAS : RB1 0.925 0.397 LU99A Lung:NSCLC CDKN2A : KRAS : PIK3CA 0.524 0.066 LU99B Lung:NSCLC CDKN2A : KRAS : PIK3CA 0.946 0.419 LU99C Lung:NSCLC CDKN2A : KRAS : PIK3CA 0.878 0.131 LUDLU-1 Lung:NSCLC 0.718 0.642 M-14 Skin BRAF : TP53 : CDKN2A 0.636 0.717 M059J Brain PTEN 0.511 0.8169 M059K Brain 1.044 0.645 MB 157 Breast 0.622 0.721 MC-IXC Nervous System (No mutations in Sanger 50) 0.88 0.457 MCAS Ovary 0.796 0.264 MCC13 Skin 0.742 MCC26 Skin 0.782 0.791 MCF7 Breast CDKN2A : PIK3CA 0.764 0.224 MDA-H2774 Ovary 1.055 0.939 MDA-MB-134-VI Breast 0.258 MDA-MB-157 Breast TP53 0.917 0.772 MDA-MB-175-VII Breast (No mutations in Sanger 50) 0.515 0.572 MDA-MB-231 Breast TP53 : NF2 : CDKN2A : KRAS : BRAF 0.789 0.504 MDA-MB-330 Breast 0.522 MDA-MB-361 Breast (No mutations in Sanger 50) 0.712 0.545 MDA-MB-415 Breast TP53 0.568 0.56 MDA-MB-435S Breast TP53 : BRAF : CDKN2A 1.232 0.974 MDA-MB-436 Breast 0.844 0.572 MDA-MB-453 Breast PIK3CA : CDH1 0.906 0.757 MDA-MB-468 Breast RB1 : PTEN : TP53 : SMAD4 0.953 0.872 MDST8 Intestine 0.445 0.406 ME-180 Cervix PIK3CA 0.746 0.176

79 Cell Line Organ of Origin Mutations 1 µM Rapamycin 500 nM PP242 MEL-HO Skin BRAF : CDKN2A 0.735 0.47 MEL-JUSO Skin HRAS : NRAS : CDKN2A 0.742 0.398 MES-SA Uterus CDKN2A 0.47 0.71 MES-SA/Dx-5 0.79 MEWO Skin 0.99 0.574 MFE-280 Uterus TP53 : PIK3CA : RB1 0.742 0.48 MFE-296 Uterus TP53 0.251 0.181 MFE-319 Uterus 0.249 0.812 MFM-223 Breast TP53 : PIK3CA 0.902 0.385 MG-63 Bone CDKN2A 0.815 0.221 MGH-BA-1 Skin 0.649 0.734 MGH-BO-1 Skin 0.526 0.684 MGH-MC-1 Skin BRAF 0.462 0.801 MGH-MCC-1 Skin 0.76 0.802 MGH-PO-1 Skin 0.699 0.414 MGH-QU-1 Skin 0.59 0.363 MGH-ST-1 Skin 0.587 0.63 MGH-SW-1 Skin 0.738 0.355 MGH-TH-1 Skin 0.626 0.517 MHH-ES-1 Bone CDKN2A : TP53 0.823 0.585 MIA PaCa-2 Pancreas TP53 : CDKN2A : KRAS 1.004 0.259 MKN1 Stomach TP53 : PIK3CA 0.793 0.56 MKN28 Stomach TP53 : APC 0.997 0.916 MKN45 Stomach SMAD4 0.684 0.203 MKN7 Stomach TP53 1.053 0.5778 MKN74 Stomach 1.125 0.832 ML-1 Thyroid 0.777 0.591 MM455 Skin BRAF; PTEN 0.615 0.859 MM608 Skin BRAF 0.434 0.453 MOG-G-CCM Brain TP53 : CDKN2A : PTEN : APC 0.652 0.599 MOG-G-UVW Brain TP53 : PTEN 0.753 0.819 MRC-9 0.708 MS751 Cervix 0.905 0.612 MSTO-211H Lung CDKN2A 0.914 0.169 MT-3 Breast 1.179 0.838 NAE Skin BRAF 0.444 0.267 NB69 Nervous System (No mutations in Sanger 50) 0.621 0.329 NCC-IT-A3 Testes 0.735 0.642 NCI-H1048 Lung TP53 : PIK3CA : RB1 0.8 0.247 NCI-H1299 Lung:NSCLC NRAS 0.725 0.422 NCI-H1435 Lung:NSCLC 0.824 0.82

80 Cell Line Organ of Origin Mutations 1 µM Rapamycin 500 nM PP242 NCI-H1437 Lung:NSCLC TP53 : CDKN2A 0.631 0.612 NCI-H1563 0.684 NCI-H1568 Lung:NSCLC 0.664 0.419 NCI-H1573 Lung:NSCLC KRAS : TP53 0.655 0.924 NCI-H1581 Lung:NSCLC TP53 : APC 0.871 0.193 NCI-H1623 Lung:NSCLC TP53 : STK11 0.864 0.642 NCI-H1648 Lung:NSCLC 0.8601 NCI-H1650 Lung:NSCLC CDKN2A : EGFR : TP53 0.873 0.633 NCI-H1651 Lung:NSCLC CDKN2A : TP53 0.833 0.826 NCI-H1666 Lung:NSCLC STK11 : BRAF : CDKN2A 1.103 0.479 NCI-H1688 Lung 0.911 0.7793 NCI-H1693 Lung:NSCLC TP53 0.64 0.363 NCI-H1703 Lung:NSCLC TP53 : APC : CDKN2A 0.643 0.512 NCI-H1734 Lung:NSCLC RB1 : STK11 : KRAS 0.774 0.443 NCI-H1755 Lung:NSCLC TP53 : CDKN2A : BRAF 0.579 0.413 NCI-H1781 Lung:NSCLC 0.744 0.645 NCI-H1792 Lung:NSCLC TP53 : KRAS 0.728 0.663 NCI-H1793 Lung:NSCLC TP53 : CDKN2A 0.495 0.361 NCI-H1869 Lung:NSCLC 0.721 0.595 NCI-H1876 Lung 0.632 NCI-H1915 Lung:NSCLC 0.837 0.581 NCI-H1944 Lung:NSCLC 0.705 0.485 NCI-H196 Lung 0.981 0.733 NCI-H1975 Lung:NSCLC TP53 : APC : CDKN2A : PIK3CA : EGFR 0.62 0.588 NCI-H1993 Lung:NSCLC STK11 : TP53 0.659 0.518 NCI-H2009 Lung:NSCLC KRAS : TP53 : RB1 0.582 0.723 NCI-H2023 Lung:NSCLC 0.518 0.431 NCI-H2029 0.967 NCI-H2030 Lung:NSCLC TP53 : KRAS 0.632 0.511 NCI-H2073 Lung:NSCLC 0.697 0.488 NCI-H2085 Lung:NSCLC 0.615 0.608 NCI-H2087 Lung:NSCLC BRAF : NRAS : TP53 0.765 0.512 NCI-H2110 Lung:NSCLC 0.649 0.726 NCI-H2122 Lung:NSCLC KRAS : TP53 : CDKN2A : CDH1 : STK11 0.701 0.783 NCI-H2135 1.237 NCI-H2170 Lung:NSCLC CDKN2A : TP53 0.294 0.517 NCI-H2172 Lung:NSCLC 0.707 0.717 NCI-H2195 0.476 NCI-H2196 0.868 NCI-H2198 Lung 0.587 NCI-H2228 Lung:NSCLC RB1 : CDKN2A : TP53 0.589 0.403

81 Cell Line Organ of Origin Mutations 1 µM Rapamycin 500 nM PP242 NCI-H2286 Lung 0.631 0.473 NCI-H2291 Lung:NSCLC 0.642 NCI-H23 Lung:NSCLC TP53 : KRAS : STK11 0.68 0.502 NCI-H2342 Lung:NSCLC TP53 0.81 1.024 NCI-H2347 Lung:NSCLC NRAS 0.454 0.481 NCI-H2405 Lung:NSCLC CDKN2A : MAP2K4 : SMAD4 : TP53 0.577 0.502 NCI-H2444 Lung:NSCLC 0.706 0.779 NCI-H2452 Lung CDKN2A 0.642 0.396 NCI-H3122 Lung:NSCLC 0.655 0.372 NCI-H322 Lung:NSCLC TP53 : CDKN2A 0.832 0.631 NCI-H358 Lung:NSCLC KRAS 0.651 0.755 NCI-H441 Lung:NSCLC TP53 : KRAS 0.791 0.762 NCI-H460 Lung:NSCLC KRAS : PIK3CA : STK11 : CDKN2A 0.814 0.27 NCI-H520 Lung:NSCLC CDKN2A : APC : TP53 0.716 0.338 NCI-H522 Lung:NSCLC TP53 0.684 0.313 NCI-H596 Lung:NSCLC TP53 : PIK3CA : RB1 0.573 0.486 NCI-H630 Intestine APC : TP53 0.951 0.663 NCI-H647 Lung:NSCLC 0.494 0.705 NCI-H650 Lung:NSCLC TP53 0.73 0.747 NCI-H661 Lung:NSCLC TP53 : CDKN2A 0.641 0.608 NCI-H727 Lung TP53 : KRAS 0.83 0.435 NCI-H810 Lung:NSCLC (No mutations in Sanger 50) 0.787 0.715 NCI-H838 Lung:NSCLC STK11 : TP53 : CDKN2A 0.891 0.325 NCI-H841 Lung 0.654 0.631 NCI-N87 Stomach TP53 : SMAD4 0.542 0.532 NCI/ADR-RES Ovary CDKN2A : PIK3CA 0.952 0.791 NH-6 Nervous System (No mutations in Sanger 50) 0.786 0.26 NIH:OVCAR-3 Ovary TP53 0.701 0.48 Nthy-ori 3-1 0.803 NUGC-2 Stomach 0.641 0.558 NUGC-3 Stomach TP53 1.024 0.444 NUGC-4 0.883 NY Bone (No mutations in Sanger 50) 0.698 0.742 OAW28 Ovary MAP2K4 0.792 0.579 OAW42 Ovary PIK3CA 0.423 0.438 OCUG-1 Liver 0.764 0.7 OCUM-1 Stomach 0.78 0.528 OE19 Esophagus TP53 0.711 0.764 OE21 Esophagus 0.561 0.25 OE33 Esophagus CDKN2A : TP53 0.877 0.251 ONCO-DG-1 Thyroid 0.568 0.304

82 Cell Line Organ of Origin Mutations 1 µM Rapamycin 500 nM PP242 OSA 80 Bone 0.647 0.749 OSC-19 Head & Neck 0.478 0.591 OSC-20 Head & Neck 0.693 0.451 OUMS-23 Intestine 0.82 0.749 OV-90 Ovary 0.348 0.512 OVCAR-5 Ovary CDKN2A : KRAS 0.845 0.431 OVCAR-8 Ovary ERBB2 : TP53 1.051 0.598 OVISE Ovary 0.82 0.582 OVKATE 0.788 OVMIU Ovary 0.698 0.521 OVSAYO Ovary 0.87 0.257 OVTOKO Ovary 0.573 0.331 PA-1 Ovary NRAS 0.66 0.539 PA-TU-8902 Pancreas 0.821 0.221 PA-TU-8988S Pancreas 1.111 0.823 PA-TU-8988T Pancreas 0.623 0.498 Panc 02.03 Pancreas 1.035 0.593 Panc 03.27 Pancreas KRAS : CDKN2A : TP53 : SMAD4 0.75 0.998 Panc 04.03 Pancreas 1.047 0.372 Panc 08.13 Pancreas CDKN2A : KRAS : SMAD4 0.782 0.428 Panc 10.05 Pancreas TP53 : KRAS 0.841 0.38 PANC-1 Pancreas 0.892 0.792 PC-14 Lung:NSCLC TP53 : EGFR 0.482 0.411 PC-3 Prostate PTEN : TP53 0.467 0.465 PC-3 [JPC-3] Lung:NSCLC 0.637 0.692 PC-9 Lung:NSCLC EGFR; P53 0.628 0.298 PCI-15 Head & Neck 0.588 0.576 PCI-15A Head & Neck 0.682 0.682 PCI-15B Head & Neck 0.676 0.639 PCI-30 Head & Neck 0.655 0.545 PCI-38 Head & Neck 0.825 0.506 PCI-4A Head & Neck 0.182 0.225 PCI-4B Head & Neck 0.444 0.505 PCI-6A Head & Neck 0.756 0.599 PCI-6B Head & Neck 0.74 0.504 PE/CA-PJ15 Head & Neck 0.675 1.002 PFSK-1 Brain TP53 0.624 0.182 PL18 Pancreas 0.62 0.532 PL4 Pancreas 0.563 0.669 PL45 Pancreas 0.724 0.5 PLC/PRF/5 Liver TP53 0.665 0.404

83 Cell Line Organ of Origin Mutations 1 µM Rapamycin 500 nM PP242 QGP-1 Pancreas 0.804 1.025 RCM-1 Intestine TP53 : KRAS : APC 0.847 0.723 RD Muscle 0.501 0.666 RERF-GC-1B Stomach 0.807 0.485 RERF-LC-Ad1 Lung:NSCLC 0.891 0.559 RERF-LC-Ad2 0.994 RERF-LC-KJ Lung:NSCLC 0.726 0.579 RERF-LC-MA Lung 0.423 RERF-LC-MS 0.486 RERF-LC-Sq1 Lung:NSCLC 0.712 0.766 RKN Ovary 0.563 0.538 RMG-I Ovary CDKN2A 0.707 0.4 RO82-W-1 Thyroid CDKN2A : TP53 0.842 0.483 RPMI 2650 Head & Neck STK11 0.68 0.816 RPMI-7951 Skin CDKN2A : TP53 : PTEN : BRAF 0.636 0.544 RT-112 Bladder TP53 : CDKN2A 0.532 0.362 RT112/84 Bladder TP53 : CDKN2A 0.687 0.359 RT4 Bladder CDKN2A 0.937 0.543 RVH-421 Skin PTEN : BRAF : CDKN2A 0.893 0.463 S-117 Thyroid NF2 : CDKN2A : TP53 0.596 0.705 SACC-83 Head & Neck 0.889 0.683 Saos-2 Bone RB1 0.637 0.708 Sarc9371 Bone 0.607 SAS Head & Neck TP53 0.942 0.592 SAT Head & Neck 0.801 0.392 SBC-3 Lung 0.683 0.954 SBC-5 Lung CDKN2A 0.743 0.548 SCaBER Bladder 0.795 0.65 SCC-15 Head & Neck TP53 0.578 SCC-25 Head & Neck 0.787 SCC-4 Head & Neck 0.695 0.603 SCC-9 Head & Neck TP53 : CDKN2A 0.551 0.991 SCC90 Head & Neck 0.821 0.641 SCCH-196 0.645 SCCH-26 0.778 SCH 0.65 SCLC-21H 0.718 SF-295 Brain TP53 : CDKN2A : PTEN 0.67 0.614 SH-SY5Y Nervous System 0.781 0.182 SHP-77 0.865 SiHa Cervix STK11 0.699 0.568

84 Cell Line Organ of Origin Mutations 1 µM Rapamycin 500 nM PP242 SISO Cervix 0.746 0.646 SJCRH30 Muscle 0.658 0.518 SK-BR-3 Breast 0.88 0.342 SK-CO-1 Intestine KRAS : APC 0.714 0.621 SK-ES-1 0.433 SK-HEP-1 Liver CDKN2A : BRAF 0.781 0.706 SK-LU-1 Lung:NSCLC CDKN2A : KRAS : TP53 0.522 0.393 SK-MEL-119 Skin NRAS 0.944 0.571 SK-MEL-131 Skin 0.638 0.627 SK-MEL-2 0.907 SK-MEL-28 Skin BRAF : EGFR : TP53; PTEN 0.816 0.592 SK-MEL-3 0.865 SK-MEL-30 Skin APC : NRAS : TP53 0.441 0.669 SK-MEL-37 Skin 0.851 0.764 SK-MEL-39 Skin BRAF; PTEN 0.625 0.367 SK-MEL-63 0.879 SK-MES Lung:NSCLC 0.852 0.559 SK-MES-1 Lung:NSCLC CDKN2A : TP53 0.64 0.557 SK-N-AS Nervous System NRAS 0.582 0.649 SK-N-DZ 0.944 SK-N-MC Bone 0.593 0.846 SK-N-SH Nervous System 1.004 0.635 SK-NEP-1 0.579 SK-OV-3 Ovary TP53 : PIK3CA : CDKN2A : APC 0.507 0.44 SKG-IIIa Cervix (No mutations in Sanger 50) 0.789 0.696 SKG-IIIb Cervix 0.736 0.906 SKN Uterus 0.455 0.593 SKN-3 Head & Neck 0.711 0.964 SN-12C Kidney TP53 0.532 0.503 SNG-M Uterus KRAS : PIK3CA : CTNNB1 : PTEN 0.862 0.766 SNU-1 0.651 SNU-16 0.948 SNU-182 Liver 0.833 0.495 SNU-387 Liver CDKN2A : NRAS : TP53 0.728 0.707 SNU-398 Liver 0.459 0.855 SNU-423 Liver (No mutation in Sanger 50) 0.578 0.555 SNU-449 Liver CDKN2A : TP53 0.751 0.457 SNU-475 Liver TP53 0.543 0.562 SNU-5 0.54 STC 1 0.552 STS 0421 Muscle 0.716 0.484

85 Cell Line Organ of Origin Mutations 1 µM Rapamycin 500 nM PP242 SU.86.86 Pancreas 0.962 0.447 SUIT-2 Pancreas 1.048 0.543 SVG p12 0.434 SVts-8 1.026 SW 1088 Brain CDKN2A : PTEN : TP53 0.587 0.75 SW 1116 Intestine TP53 : APC : KRAS : SMAD4 1.175 0.87 SW 1271 Lung 0.83 0.715 SW 13 Kidney TP53 1.206 0.126 SW 1417 Intestine APC : TP53 : BRAF 1.084 0.516 SW 1463 Intestine TP53 : KRAS : APC 1.179 0.707 SW 156 Kidney 0.607 0.614 SW 1573 Lung:NSCLC KRAS : CDKN2A : PIK3CA : SMAD4 : CTNNB1 0.748 0.521 SW 1783 Brain RB1 : PTEN : TP53 0.707 0.599 SW 1990 Pancreas CDKN2A : KRAS 0.799 0.758 SW 48 Intestine EGFR : CTNNB1 1.08 0.513 SW 626 Ovary TP53 : KRAS : APC 0.743 0.619 SW 780 Bladder CDKN2A 0.603 0.598 SW 900 Lung:NSCLC TP53 : CDKN2A : KRAS 0.766 0.668 SW-1710 Bladder TP53 : CDKN2A 0.773 0.537 SW-948 Intestine PIK3CA : KRAS : APC 0.844 0.558 SW527 Breast 0.905 0.868 SW620 Intestine APC : TP53 : KRAS : MAP2K4 0.877 1.384 SW756 Cervix KRAS : STK11 0.759 1.031 SW837 Intestine APC : TP53 : FBXW7 : KRAS 1.078 0.873 T.T Esophagus (No mutations in Sanger 50) 0.793 0.6 T.Tn Esophagus 0.741 0.814 T24 Bladder TP53 : HRAS 0.984 0.597 T47D Breast PIK3CA : TP53 0.87 0.317 T84 Intestine KRAS : PIK3CA : TP53 : APC 0.984 0.734 T98G Brain CDKN2A : TP53 0.439 0.41 Takigawa Stomach 1.038 0.46 TASK1 Nervous System 1.425 0.364 TCCSUP Bladder TP53 : PIK3CA : RB1 0.806 0.264 TCO-1 Thyroid TP53 : KRAS 0.756 0.712 TE7 Esophagus 0.811 0.439 TGW 1.056 THLE-3 0.985 TMK-1 Stomach 0.664 0.352 TOV-112D Ovary 0.814 0.077 TOV-21G Ovary 0.509 0.489 TT2609-C02 Thyroid 1.065 0.279

86 Cell Line Organ of Origin Mutations 1 µM Rapamycin 500 nM PP242 TYK-nu Ovary TP53 : NRAS : CDKN2A 0.575 0.627 U-118 MG Brain PTEN : CDKN2A : TP53 0.638 0.471 U-138 MG Brain 0.618 0.564 U-2 OS Bone (No mutations in Sanger 50) 0.578 0.483 U-251 MG Brain TP53 : CDKN2A : PTEN 0.584 0.661 U373 MG Brain 0.868 0.891 UACC-62 Skin PTEN : CDKN2A : BRAF 0.718 0.444 UACC-812 0.872 UACC-893 Breast TP53 : PIK3CA 0.239 0.28 UACC903 Skin BRAF; PTEN 0.249 0.328 UDSCC2 Head & Neck 0.773 0.965 UISO-MCC 1 Skin 0.831 0.745 UM-UC-3 Bladder CDKN2A : KRAS : TP53 : PTEN 0.823 0.602 UMC-11 Lung SMAD4 : CDKN2A : TP53 : STK11 0.604 0.448 UO-31 Kidney CDKN2A 0.433 0.408 VCaP 0.771 VM-CUB1 Bladder TP53 : CDKN2A : HRAS 0.734 0.38 VMRC-LCD Lung:NSCLC 0.781 0.572 VMRC-LCP Lung:NSCLC TP53 : STK11 : CDKN2A 0.851 0.723 VMRC-MELG 0.816 VMRC-RCW Kidney 0.673 0.577 VMRC-RCZ Kidney CDKN2A 0.537 0.526 WI-26 VA4 1.019 WI-38 VA13 sub 2 RA 0.885 WiDr Intestine 1.088 0.662 WM 266-4 Skin BRAF 0.672 0.399 WM-115 Skin BRAF : CDKN2A : PTEN 0.622 0.633 WM1158 Skin BRAF 0.468 0.642 WM1552C Skin 0.568 0.866 WM164 Skin BRAF 0.733 0.663 WM239A Skin BRAF; PTEN 0.67 0.46 WM278 Skin BRAF 0.918 0.719 WM35 Skin BRAF 0.861 0.717 WM793B Skin BRAF 0.711 0.635 WM902B Skin 0.778 0.635 XPA-1 Pancreas 0.843 0.582 XPA-3 Pancreas 1.038 0.683 XPA-4 Pancreas 0.838 0.873 YAPC Pancreas SMAD4 : CDKN2A : KRAS : TP53 0.776 0.665 YKG-1 Brain CDKN2A : PTEN : TP53 0.536 0.728 ZR-75-1 Breast 0.616 0.272

87 Cell Line Organ of Origin Mutations 1 µM Rapamycin 500 nM PP242 ZR-75-30 Breast CDH1 0.713 0.29