Andrew Thorburn - US grants

DESCRIPTION (Provided by the applicant) Androgens functioning through the androgen receptor (AR) are required for survival of prostate epithelial cells and androgen deprivation is a common treatment for advanced prostate cancer. Androgen deprivation is successful in most cases. However, tumors often relapse and exhibit a hormone-independent phenotype where cancer cells are no longer dependent on androgens for survival. At this stage, therapeutic options are limited and most patients with androgen insensitive prostate cancer die from their disease. To improve treatment options for these patients, we need to understand the mechanisms by which the AR controls prostate cell apoptosis. The AR is a transcription factor and it has been thought that AR-induced gene expression is required for prostate epithelial cell survival. Conversely, it is hypothesized that removal of hormone results in loss of expression of these genes thus leading to prostate cell apoptosis. However, there are few convincing candidates for genes whose expression is AR-dependent and might be responsible for this effect. Recently, a quite different way for regulation of apoptosis by the AR was described. This mechanism involves proteolysis of the AR to release a toxic peptide. This mechanism is hormone-dependent because proteolysis does not occur in the presence of androgens. Our preliminary studies indicate that this mechanism can function in normal prostate epithelial cells. We hypothesize that this new mechanism contributes to androgen regulation of prostate cell survival and that defects in this mechanism may contribute to the progression of advanced prostate cancer. In this pilot project, we will test our hypothesis with the following aims: l) Identify toxic proteolytic fragments from the normal androgen receptor. 2) Determine the mechanism of androgen receptor proteolysis-induced cell death. The experiments proposed here will determine how this mechanism contributes to prostate epithelial cell survival and apoptosis. These studies may open a new avenue of research into androgen regulation of prostate cell survival and lead to novel therapeutic strategies to treat advanced prostate cancer.

DESCRIPTION (Provided By Applicant): Tumor necrosis factor (TNF) signaling via the receptors TNFRI and TNFR2 contributes to both cell death and protection after stroke or other brain trauma. We must understand how these different signaling pathways work if we are to develop ways to minimize tissue damage after stroke or other trauma. TNFRI causes cell death by a well-understood apoptosis pathway that activates a caspase cascade and is inhibited by caspase inhibitors such as zVAD.fink. However, TNF can also kill cells by a mechanism that cannot be inhibited by zVAD.fink, and which may also contribute to TNF-induced tissue damage. The molecular mechanism by which this pathway kills cells is unknown. We recently made the surprising discovery that TRADD, an adaptor protein that mediates downstream signaling through TNFR1 binding at the cell membrane, is not solely a cytoplasmic protein. Rather, TRADD shuttles into and out of the nucleus via active nuclear import and export. Furthermore, an isolated domain from TRADD is localized exclusively in the nucleus where it induces apoptosis via a mechanism that is not inhibited by zVAD.fmk. We therefore hypothesize: nuclear signaling from TRADD contributes to the resistant cell death that is caused by TNF. Here, we propose to test this hypothesis with the following specific aims. 1) To determine the role of TRADD nuclear import and export in inducing cell death. 2) To determine how TRADD induces cell death from the nucleus. Our studies relate to the purpose of the R21 mechanism for NS-00-01 I in the following ways. Our experiments will provide the first information regarding the significance and mechanism of action of this unexpected nuclear activity of TRADD and. establish the concept that nuclear shuttling of a receptor interacting protein activates a novel apostolic pathway. The work proposed here should also form a strong basis for future projects to determine how these activities contribute to the tissue damage that is caused by stroke or other brain trauma. This information may lead to improved strategies for therapeutic manipulation of TNF signaling pathways in damaged brain tissues.

[unreadable] DESCRIPTION (provided by applicant): The adapter protein TRADD regulates pro- and anti-apoptotic signaling from the Tumor Necrosis Factor alpha (TNF) receptor TNFR1. TNFR1 regulation of cell survival and apoptosis is implicated in diverse diseases including ischemic injury, neurodegeneration, infections, inflammatory disease and cancer. An important pathway leading to TNFR1-induced apoptosis requires binding of TRADD to another adapter called FADD, which then recruits and activates caspase 8 leading to activation of downstream caspases and cell death. This occurs in the cytoplasm when TRADD binds to the intracellular death domain of TNFR1. However, multiple lines of evidence indicate that this is not the only mechanism by which TNFR1 (and thus components of the TNFR1 complex such as TRADD) induce apoptosis. If we can understand the other ways that these molecules induce apoptosis we may be able to develop rational strategies to manipulate these responses to improve treatments for diseases where TNF signaling is important. [unreadable] [unreadable] With the support of an NINDS R21 (NS42662), we recently made the surprising discovery that TRADD is not just a cytoplasmic protein but instead can shuttle between the cytoplasm and the nucleus where it is associated with promyelocytic leukemia protein (PML) nuclear bodies. Nuclear TRADD stimulates an apoptosis pathway that is mechanistically distinct from the well-established cytoplasmic pathway. The nuclear pathway requires p53 and PML and is inhibited by Bcl-xL. We hypothesize that this pathway induces apoptosis through PML nuclear body- and p53-dependent signals that cause mitochondria dysfunction. We further hypothesize that this pathway explains some of the TNF-induced cell death that cannot be accounted for by the established cytoplasmic pathway. To test our hypothesis, we have the following specific aims. 1) Identify the regulators that are required for apoptosis by nuclear TRADD. 2) Determine the role of p53 in apoptosis by nuclear TRADD. 3) Determine the role of PML nuclear bodies in apoptosis induced by nuclear TRADD. 4) Determine the role of nuclear TRADD in mediating TNF-dependent signals. These studies should provide a detailed investigation of a novel apoptosis pathway that may have a role in various human diseases. [unreadable] [unreadable]

DESCRIPTION (provided by applicant): Defects in apoptosis regulation are thought to be a prerequisite for cancer development however the nature of the apoptotic pathways that are defective is poorly understood. The identification and understanding of apoptotic pathways that are selectively disrupted during cancer development may provide new insights into cancer development and identify novel therapeutic targets. We are studying a novel apoptosis pathway that is induced by the death domain of the adaptor protein FADD (FADD-DD). This pathway has unusual characteristics that suggest it is an example of an apoptosis pathway that has to be disrupted for breast or prostate cancer to develop. Our previous studies show that FADD-DD can induce apoptosis only in normal epithelial cells. This cell type-specific response works via a previously unrecognized mechanism that is separate from the established mode of action of FADD and may involve a novel FADD-binding protein called PL31. The novel pathway is specifically disrupted when epithelial cells become immortalized. However, this disruption is unrelated to inactivation of the known pathways (p53, Rb &telomerase) that are involved in immortalization. An oncogene (SV40 T antigen) can confer resistance to this apoptosis pathway in normal cells without affecting other apoptosis mechanisms, while a specific tumor suppressor (Bin1) can confer sensitivity to this pathway in cancer cells. Endogenous FADD protein can activate this pathway when it is stimulated by TRAIL. Thus, we have identified a new apoptosis pathway that is activated by TRAIL, involves FADD, may involve PL31 and Bin1 and is specifically disrupted by SV40 T antigen through a p53- and Rb independent mechanism that is associated with cell immortalization. This data will lead us to develop our hypothesis: FADD participates in a novel apoptotic pathway that is specifically disrupted during breast or prostate cancer development. Here, we test this hypothesis and determine the roles of PL31, Bin1, T antigen and TRAIL with the following aims: 1) Determine how FADD-DD induces apoptosis of normal epithelial cells. 2). Determine why immortal cells are resistant to FADD-DD-induced apoptosis. 3). Characterize the physiologic signal that activates the FADD-DD-dependent pathway in normal epithelial cells. These studies should provide a detailed understanding of a previously unrecognized apoptosis pathway that may be intimately involved in cancer development.

DESCRIPTION (provided by applicant): TNF-Related Apoptosis Inducing Ligand (TRAIL) kills tumor cells with little effect on normal tissues and recombinant TRAIL and antibodies that recognize TRAIL receptors are in clinical trials at the University of Colorado and elsewhere. In addition, TRAIL receptor signaling determines the efficiency with which other agents kill tumor cells. However, tumor cells are often resistant to TRAIL and the mechanisms by which this occurs and what this resistance means for tumor progression and clinical outcomes is poorly understood. We recently discovered that one mechanism by which breast and ovarian tumor cells can become selectively resistant to TRAIL is through the increased expression of the homeobox transcription factor Six1. We found that increased Six1 is common, occurring in >60% of metastatic ovarian cancers and 90% of metastatic breast cancers, and associated with poor clinical outcomes. We further found that Six1 expression is sufficient to make non-metastatic tumor cells metastasize in vivo. Because TRAIL signaling is known to suppress metastasis, we hypothesize that Six1 inhibits TRAIL by a specific mechanism and this leads to increased metastasis resulting in poor clinical outcomes in patients and resistance of patient's tumors to TRAIL. To test this hypothesis we propose an integrated project that will determine the molecular mechanism by which Six1 inhibits TRAIL receptor-induced apoptosis, test if these mechanisms are responsible for increased metastasis in mice and determine whether these effects apply in primary tumor cells from patients and if they lead to worse clinical outcomes. Because this work encompasses research on basic mechanisms at the cellular level, testing those mechanisms in animal models of cancer progression and metastasis and clinical and translational studies in ovarian and breast cancer patients'tumors, we have adopted a team approach that will involve a cell biologist with expertise in TRAIL signaling (Dr. Thorburn), a pioneer in the analysis of Six1 in cancer development and progression (Dr. Ford) and an oncologist with expertise in clinical and translational research (Dr. Behbakht). To achieve these goals we have the following aims: 1. Determine how Six1 alters signaling by TRAIL receptor-targeted therapeutic agonists in ovarian and breast cancer. 2. Test if Six1-induced metastasis involves the TRAIL resistance mechanism, and 3. Determine if Six1 expression predicts TRAIL sensitivity and prognosis in patient tumors. Together, these aims should allow us to understand how Six1 regulates TRAIL receptor signaling, determine the role of these mechanisms in tumor metastasis and test if the same mechanisms apply in breast and ovarian cancer patients and thus determines their clinical outcomes. PUBLIC HEALTH RELEVANCE: Breast and ovarian cancer are the second and fifth leading causes of cancer death in women. One exciting new approach to treating these (and other) cancers is to use therapies that directly activate TRAIL receptors, however many tumor cells are resistant to these therapies. This project focuses on one mechanism (Six1 expression) by which tumor cells become resistant to TRAIL that our data indicate affect a high proportion of metastatic cancer patients (~60-90%). By understanding how this mechanism causes resistance to TRAIL, how this affects disease progression and metastasis and how this alters clinical outcomes in patients, we should gain new insights into the value of Six1 as a prognostic marker and better understand how to use the TRAIL therapeutics that are being developed.

DESCRIPTION (provided by applicant): Fas Associated Death Domain (FADD) is an adaptor protein that is required for signaling by the Tumor Necrosis Factor-Related Apoptosis Inducing Ligand (TRAIL) receptors. TRAIL receptors are important therapeutic targets in cancer with six TRAIL receptor-targeted drugs in clinical trials at the current time, several others in pre-clinical development and accumulating evidence suggesting that signaling through endogenous TRAIL is important in the mechanism of action of other anti-cancer treatments including DNA damaging agents, anti-metabolites and histone deacetylase inhibitors. The molecular mechanisms by which FADD activates caspases upon TRAIL receptor stimulation are quite well understood, however mechanisms of TRAIL resistance are still poorly understood and this limits our ability to optimally use the TRAIL receptor-targeted drugs. In the previous funding period we analyzed mechanisms of TRAIL resistance and FADD signaling and made the unexpected discovery that TRAIL receptors induce autophagy and that a FADD inhibitor could induce autophagy implying that FADD negatively regulates autophagy. Because autophagy can affect apoptosis responses in tumor cells, we propose that these activities affect the efficiency by which TRAIL receptor signaling activates the apoptosis machinery and thus kills tumor cells. We have also found that autophagy controls the characteristics of dying cells, particularly the release of an immune regulator called HMGB1 and that this also occurs in a FADD-dependent manner in response to TRAIL. Based on these findings, this competitive renewal focuses on three complementary questions: How does FADD regulate autophagy? What effect does autophagy have on TRAIL receptor signaling? And, does manipulation of autophagy provide a way to improve the anti-tumor effect of TRAIL receptor-targeted drugs? To answer these questions we have the following aims. Aim 1. Determine the role of FADD in regulation of autophagy. This aim tests the hypothesis that FADD negatively regulates autophagy by interaction with autophagy regulators. Aim 2. Determine how autophagy affects signaling by TRAIL-R targeted drugs. This aim tests the hypothesis that FADD's ability to inhibit autophagy serves to coordinate competing signals and thus provide fine control over tumor cell death during treatment with TRAIL R-targeted drugs. Aim 3. Test if autophagy manipulation improves the effectiveness of TRAIL-R targeted drugs in vivo. This aim tests the hypothesis that autophagy inhibition will make TRAIL receptor-targeted drugs (lexatumumab, mapatumumab) more effective and uses a unique set of isogenic tumor cells in which we can determine the relative roles of exogenous and endogenous TRAIL receptor stimuli in the anti-tumor response and the role of autophagy in controlling these responses. These studies should provide new insights into FADD and TRAIL receptor signaling, the role of autophagy in determining the response to anti-cancer therapy and provide a basis for improving the use of TRAIL receptor- targeted drugs in treating people with cancer. PUBLIC HEALTH RELEVANCE: In the last few years it has become clear that a hitherto understudied cellular process called autophagy is an important regulator of cancer development and treatment. However there is considerable confusion about what we should try to do to autophagy to improve cancer therapy- in fact it is not clear whether we should try to inhibit autophagy or stimulate it during treatment of cancer. This grant examines signaling by a protein called FADD, which is required for tumor cell killing after activation of TRAIL receptors. This is important because TRAIL receptors are targeted by at least 6 anti-cancer agents and are also important for tumor cell killing by other drugs that work indirectly through TRAIL. In the previous funding period, we made several discoveries; first we found that FADD (and TRAIL) can regulate autophagy. Second, we found that autophagy can modulate the efficiency of tumor cell killing by various drugs including TRAIL. In this proposal we aim to answer the key questions that arose out of the previous work. We will work out how FADD regulates autophagy, how this activity alters signaling by TRAIL receptors and test whether manipulation of these processes alters the effectiveness of treatment by the TRAIL receptor-targeted drugs that are used in people. These studies should provide a way to improve the use of the various anti-cancer agents that target TRAIL receptors which are already in clinical trials and provide a rationale to allow us to manipulate autophagy during cancer treatment.

DESCRIPTION (provided by applicant): There is great interest in manipulating autophagy to improve cancer treatment but considerable confusion about how to do so. For example, it has become clear in the last few years that many anti-cancer treatments induce autophagy in tumor cells, however, because this autophagy has been reported to both protect and kill tumor cells, even such a basic question as whether we should try to increase or decrease autophagy during cancer treatment is unclear. Despite this uncertainty, clinical trials are being developed that combine autophagy inhibitors with other drugs while other patients are treated with drugs that induce autophagy but without any consideration for how this autophagy affects the outcome. Thus a major need in cancer research is to better understand the roles of autophagy in tumor cell death so that we can decide how best to manipulate autophagy in people with cancer. We recently discovered a previously unrecognized function for autophagy: autophagy controls the selective release of the nuclear protein HMGB1 from dying tumor cells. HMGB1 release from dying tumor cells is known to lead to a beneficial tumor-specific immune response through activation of Toll-like Receptors on dendritic cells but may also induce non-beneficial pro-tumorigenic activities. Therefore, our recent findings open up an entirely new issue that needs to be dealt with as we consider how to manipulate autophagy in people. We hypothesize that it is not just the efficiency of tumor cell killing that is important but that the characteristics of the dying tumor cells, exemplified by selective release of HMGB1, also determines the overall success of treatment. And, we propose that autophagy controls both these processes. To test this hypothesis we will use a well-characterized model of metastatic breast cancer to complete the following aims. Aim 1. Determine how autophagy controls the efficiency and characteristics of tumor cell killing by anti-cancer treatments. This aim will test the hypothesis that autophagy regulates both the efficiency of tumor cell killing and the release of HMGB1 from dying cells but does so differently for different kinds of anti-cancer treatments. Aim 2. Determine which aspects of the autophagic process are required for the different responses during tumor cell death. Autophagy is a dynamic process that involves multiple steps; in this aim, we test the hypothesis that different steps in the process are required for the different functions of autophagy and we use novel single cell imaging methods to determine how autophagy regulates the core apoptosis machinery. Aim 3. Determine how manipulation of autophagy alters the long-term efficacy of cancer treatment in vivo. In this aim, we test the hypothesis that autophagy manipulation can determine the long term effectiveness of breast cancer treatment using a model of adjuvant and neoadjuvant chemotherapy and we will determine which functions of autophagy are important for controlling the response to therapy. These studies will give us new insights into the role of autophagy during tumor cell death and should provide a rationale for developing autophagy manipulation strategies to improve the effectiveness of cancer treatment. PUBLIC HEALTH RELEVANCE: In the last few years it has become clear that a hitherto understudied cellular process called autophagy is an important regulator of cancer development and treatment. Although it is widely believed that autophagy is important during tumor cell death (e.g. after treatment with anti-cancer drugs), the precise roles for autophagy are unclear and it has been reported that autophagy may both protect tumor cells and cause their death under different circumstances. Our lack of understanding of what autophagy does during tumor cell death limits our ability to manipulate autophagy in order to improve cancer treatment. We recently discovered a new function for autophagy during tumor cell killing- autophagy regulates the release of a protein that controls the immune system. We propose that this function along with autophagy's ability to control the amount of tumor cell killing will together determine whether efforts to kill cancer cells are effective or not. The work proposed here will test these ideas and determine how autophagy controls the amount and characteristics of tumor cell death and should provide a framework to attempt to manipulate autophagy in people who are being treated for cancer in order to improve the benefits of the treatment.

DESCRIPTION (provided by applicant): The purpose of this Paul Calabresi Award in Clinical Oncology Research (PCACOR) is to prepare highly qualified cancer clinical researchers called Scholars who can independently design, manage and complete cancer clinical trials by learning to communicate and coordinate with multidisciplinary teams of clinical and basic scientists using state-of-the-art laboratory analyses to personalize cancer treatment. Specific objectives are to: 1) Provide a flexible Individual Development Program for oncology medical doctors and Ph.D. clinicians (nurses, pharmacists, clinical psychologists, epidemiologists, oral surgeons) who have completed their clinical training and basic research scientists who are committed to a translational clinical cancer research career in an academic setting; 2) Foster interdisciplinary training, communication and interaction through multiple mentoring of program scholars; 3) Create an ongoing mentorship to the completed PCACOR Scholars to ensure their successful transition to clinical research independence. Strong clinical and basic science mentors provide exceptional training and role models in all oncology clinical disciplines as well in cancer basic science. A multidisciplinary Advisory Committee oversees the PCACOR directed by Program Leaders, Madeleine A. Kane, MD, Ph.D., Professor of Medical Oncology, and Andrew Thorburn, Ph.D., Professor and Chair, Department of Pharmacology, Deputy Director, UCCC. Training period lasts from two to seven years and includes Core requirements such as Responsible Conduct of Research and coursework provided by CCTSI Clinical Science Program and the UCD Graduate School. Each Scholar completes a basic science project and also designs at least one cancer clinical trial. Scholars submit an NIH style grant application within the last yea of training. The ultimate success is demonstrated by our track record. A total of fifteen junior faculty have trained as PCACOR Scholars, nine during the current funding period (out of 42 applications). Thirteen Scholars remain faculty members at University of Colorado Denver Anschutz Medical Campus (UCD-AMC); six have been promoted, all have multiple peer-reviewed publications, both clinical and basic science. Eight have obtained peer-reviewed funding from NIH, DOD ACS and/or Komen Foundation. All have designed at least one cancer clinical trial, and several PCACOR graduates lead national cancer clinical trials, cancer research training programs and clinical cancer research programs. Our PCACOR successfully produces the translational cancer clinical researchers of the future.

The University of Colorado at Denver and Health Sciences Center (UCDHSC) Pharmacology Graduate Training Program, currently in its 30th year of NIGMS funding, requests annual support for nine predoctoral students during the next five years. This Graduate Training Program distinguishes itself by providing a highly interactive environment, on a new HSC campus, in which students can obtain a broadly-based integrative perspective of science, training in the foundation of knowledge that defines pharmacology, and sophistication in specialized, state-of-the-art research areas. The Training Program is directed by Nancy Zahniser, Ph.D., and co-directed by David Port, Ph.D., who also chairs the Graduate Training Committee, which provides the day-to-day oversightover this Training Program. The 38 members of the Training Program faculty are drawn both from within and from outside of the Department of Pharmacology, School of Medicine, and have been recruited to provide broad, multidisciplinary training opportunities in neuropharmacology, cell signaling and trafficking, molecular pharmacology, pharmacogenetics, cancer biology, genomics, proteomics, lipidomics, biomolecular structure, bioinformatics, as well as translational pharmacology. The Training Program faculty are all accomplished, committed researchers and mentors with significant extramural funding. The sources of students entering this Training Program include direct applicants to the Program, as well as students who transition from HSC 'feeder' programs (Biomedical Sciences Program and Medical Scientist Training Program). The Pharmacology Graduate Training Program has three curricular tracks: pharmacology (primary track), biomolecular structure and bioinformatics. Hallmarks of the Program are a comprehensive didactic component, three laboratory rotations, a strong emphasis on student presentations in seminar settings, and a wide choice of thesis research options. Career development in the pharmacological sciences and student initiative are also emphasized. Since its last review, 42 students (from a total of 58) have been supported by this Training Grant, of which 10% came from under-represented populations. The Training Program currently has 27 students. Twenty-three trainees have graduated with Ph.D. degrees, on average, in five years. The competitiveness of the students for individual national fellowships, high quality publications in peer-reviewed journals and invitationsto participate in national meetings are all measures by which the successful training of the students is gauged. Additionally,the retention of the graduates in academic, industry and government positions is another measure of the success of the Training Program. With renewal of funding, this Training Program will continue to thrive and meet the national demands for individuals, trained as pharmacologists, who are individually astute researchers, can be multidisciplinary research team members, and also have the breadth of knowledge to plan and communicate effectively across a spectrum of technologies.

DESCRIPTION (provided by applicant): The University of Colorado Denver | Anschutz Medical Campus (UCD|AMC) Pharmacology Graduate Training Program, currently in its 34th year of NIGMS funding, requests annual support for nine predoctoral students during the next five years. This Graduate Training Program distinguishes itself by providing a highly interactive environment, on a new health sciences campus, in which students can obtain a broadly based integrative perspective on science, training in the foundation of knowledge that defines pharmacology, and sophistication in specialized, state-of-the-art areas of research. The Principle Investigator for the Training Grant is Dr. Andrew Thorburn, Chair of the Department of Pharmacology. The Training Program Director is Dr. David Port, who also chairs the Graduate Training Committee (GTC), which provides the day-to-day oversight for this Training Program. The 39 members of the Training Program faculty are drawn both from within and from outside of the School of Medicine Department of Pharmacology, and have been recruited to provide broad, multidisciplinary training opportunities in neuropharmacology, cell signaling and trafficking, molecular pharmacology, pharmacogenetics, cancer biology, genomics, proteomics, lipidomics, biomolecular structure, bioinformatics, as well as translational pharmacology. The Training Program faculty are all accomplished, committed researchers and mentors with significant extramural funding. The sources of students entering this Training Program include direct applicants to the Program, as well as students who transition from Graduate School 'feeder' programs (Biomedical Sciences and Medical Scientist Training Programs). Hallmarks of the Program are a comprehensive didactic component, three laboratory rotations, a strong emphasis on student presentations in seminar settings, and a wide choice of thesis research options Career development in the pharmacological sciences and student initiative are also emphasized. Since its last review, 22 students (from a total of 47) have been supported by this Training Grant, of which ~9 percent came from under-represented populations. The Training Program currently has 16 students. Since 2008, 21 trainees have graduated with Ph.D. degrees in just over five years, on average. The competitiveness of the students for individual national fellowships, high quality publications in peer-reviewed journals and invitations to participate in national meetings are all measures by which the successful training of the students is gauged. Additionally, the retention of the graduates in academic, industry and government positions is another measure of the success of the Training Program. With renewal of funding, this Training Program will continue to thrive and meet the national demands for individuals, trained as pharmacologists, who are individually astute researchers, can be multidisciplinary research team members, and also have the breadth of knowledge to plan and communicate effectively across a spectrum of technologies.

? DESCRIPTION (provided by applicant): Autophagy is important in cancer development, progression and response to therapy. Although we are already trying to target autophagy in the clinic, there is currently no way to identify the tumors that will or will not benefit from autophay inhibition. We recently reported that some breast tumor cells are much more dependent on autophagy than others and our data suggest that it is only in the truly autophagy-dependent tumor cells (which we define as those that require autophagy to survive even in the absence of additional stress) where autophagy inhibition will be effective on its own or in combination with other agents. Indeed, our data suggest that combining autophagy inhibition with another agent in tumor cells that are not autophagy-dependent can be counterproductive. We also identified a potential mechanism to explain autophagy-dependent breast cancer- autophagy controls STAT3 activity in only some tumor cells and this is necessary and sufficient to explain autophagy-dependent survival; moreover recent studies have also indicated that tumor stem-like cell activity requires autophagy in these cells. These data led us to hypothesize: autophagy-dependency defines a specific tumor subtype whose cell survival and stem cell-like characteristics depend on autophagy: these are the tumors that will respond best to autophagy inhibition. To test our hypothesis, we have three specific aims. Aim 1. Identify autophagy-dependent tumors and determine response to autophagy inhibition. We hypothesize that we can identify autophagy-dependent tumors and will test a preliminary gene expression signature that we think can do this. We propose that autophagy inhibition will be effective but only for autophagy-dependent tumors and will test this in primary tumors and by neoadjuvant and adjuvant treatment of surgically resected metastatic tumors. Aim 2. Test if autophagy inhibition sensitizes to other treatments in autophagy-dependent breast cancer. To test the hypothesis that autophagy-dependent tumors will benefit most from combination treatment with autophagy inhibitors, experiments using large scale shRNA analysis well as a novel bioinformatics approach using the COXEN principle will work out how best to target autophagy in combination with other drugs. Aim 3. Determine the mechanism underlying autophagy-dependency in breast tumors. We hypothesize that autophagy regulation of autocrine signaling explains STAT3 activation, survival and stem cell like of autophagy-dependent tumor cells. We will test this by analyzing autophagy's ability to signal through cytokines like IL6, which we have shown is specifically controlled by autophagy only in the autophagy-dependent tumor cells. If our ideas are correct, our work will help define a novel subtype of autophagy-dependent breast cancer and provide a new way to identify, treat and improve combination treatments for these tumors while providing new insights into the roles of autophagy in cancer therapy and cancer biology.

? DESCRIPTION (provided by applicant): Autophagy is important in cancer development, progression and response to therapy. Although we are already trying to target autophagy in the clinic, there is currently no way to identify the tumors that will or will not benefit from autophay inhibition and it is unknown if tumors that are driven by different oncogenic driver mutations will differ in their response to autophagy inhibition or not. Because different tumor drivers have opposing effects on autophagy (e.g. mutant RAS is thought to promote autophagy while loss of PTEN activates the canonical PIK3C pathway which should inhibit autophagy), it is likely that tumors with different drivers will differ in their autophagy dependency and thus respond differently to autophagy inhibition. In animal models using conventional (albeit state of the art) genetically engineered mouse (GEM) models, accumulating evidence suggests that RAS-driven tumors benefit from autophagy inhibition. However the time and cost associated with making such GEMs makes testing multiple clinically relevant tumor drivers head to head along with multiple autophagy regulators very difficult and expensive. In this pilot project we aim to develop a new approach that will allow us to directly assess the effects of inhibiting autophagy by targeting different autophagy regulators and in tumor cells driven by five different clinically relevant prostate tumor drivers and to do so both in vitro and during cancer development and progression in vivo. We will do this by capitalizing on a novel approach pioneered by the Cramer lab that uses engineered purified prostate stem cells recombined with urogenital mesenchyme to grow tissue recombinants that mimic normal prostate tissue or aggressive tumors driven by specific tumor drivers (RAS activation, PTEN loss etc.). We will combine this new approach with new methods from the Thorburn lab to target and assess autophagy, allowing us to test the effects of inhibiting three essential autophagy regulators that control distinct steps in the autophagy process in tumor cells driven by five different combinations of tumor drivers all associated with aggressive human prostate cancer and predicted to have different and sometimes opposing effects on autophagy. We will perform both in vitro and in vivo analysis to test our central hypothesis: different tumor drivers cause tumor cells to display different degrees of autophagy dependence in vitro and respond to autophagy inhibition in vivo. We will do this with two aims: 1. Determine the effects of specific tumor-driving oncogenic mutations on autophagy and autophagy-dependence in vitro. And 2. Determine the role of autophagy in prostate tumor development & progression in vivo in the context of different oncogenic mutations/drivers. This high risk/ high return project will establish if our approach is feasible o not and thus allow us to know how to proceed with a larger research project to determine the importance and mechanism of these effects and assess whether analysis of tumor mutations can be used to select patients whose cancer is most likely to respond to autophagy inhibition therapy. We believe these characteristics of high risk but potential for high return with the abiliy to identify a clear way forward based on the results obtained makes this project ideal for the R21 mechanism.

Anti-cancer drugs usually work by inducing apoptosis. Unfortunately a significant body of evidence suggests that apoptosis may not be a good way to kill cancer cells? e.g. apoptosis causes rapid, caspase-dependent tumor repopulation of drug-resistant cells because apoptotic cells send growth promoting signals to neighboring cells that aren't killed and apoptosis causes non-immunogenic tumor cell killing thus reducing the likelihood of generating an effective anti-tumor immune response. Recent work from our lab suggests that it is feasible to regulate and change the mode of action of an apoptotic stimulus so that instead of killing by apoptosis, the cell dies by necroptosis. Necroptosis is a form of programmed necrosis, which does not involve caspases and is more immunogenic than apoptosis that may, therefore, be a better way to kill cancer cells. In this pilot grant we will test the central hypothesis that it is possible to manipulate a normal cell process, autophagy, in order to change the mode of action of anti-cancer drugs so that they kill human tumor cells by necroptosis instead of or as well as apoptosis. This work builds on previous studies from our group and is intended to develop pilot data establishing feasibility of our proposed approach. If we establish the feasibility of our central hypothesis, we should have a foundation to develop a larger project to test if it is possible and worthwhile to try to manipulate not only whether or not we kill cancer cells but also control how we kill them. To achieve these goals we have two aims. Aim 1. Test if the autophagy machinery regulates necroptosis in human cancer cells. Aim 2. Test if re-activation of necroptosis capacity in human tumor cells with RIPK3 silencing allows broadening of the autophagy regulation of necroptosis to more cancer cells. Upon completing this small pilot grant, we will be positioned to develop a larger project to test if we can change the current paradigm of cancer therapy to control the mechanism of tumor cell death as a way to improve cancer therapy.

An important unsolved question in cell death is to understand why different cells within a population vary in their responses. Which cells will live or die and what determines exactly how they die after exposure to a death stimulus? These questions underlie fundamental cell fate decisions and also have important practical ramifications, for example, during cancer therapy when non-heritable, heterogeneous responses to anti-cancer drugs underlie the eventual acquisition of resistance to therapy. Heterogeneity in cell responses can be driven by stable genetic differences between cells, which are easy to understand. However, such differences also occur even in genetically homogeneous cell populations. What underlies these differences? More important, can we manipulate these effects? Previous work supported by this grant discovered that even in a homogeneous population of cells under unstressed conditions, there is extensive variation in the amount of autophagic flux, which in turn predicts the outcome to future treatment with a death stimulus. And, in the last funding period, we discovered a specific mechanism by which autophagy controls the apoptosis threshold and a quite different mechanism by which the autophagy machinery can control necroptosis. Building on these previous studies, we hypothesize: autophagy controls apoptotic and necroptotic thresholds by regulating Mitochondrial Outer Membrane Permeabilization (MOMP). And, this explains cell death variation between cells in a population. We will test this hypothesis by completing the following aims using a variety of new approaches including the first method that allows optogenetic regulation of autophagy. Specific Aim 1. Test if autophagy variation before and after a death stimulus controls heterogeneity in apoptosis responses in a cell population. Specific Aim 2. Determine how autophagy regulates necroptosis. By completing these aims, we will gain new insights into the interplay between two major forms of programmed cell death (apoptosis and necroptosis) and uncover how autophagy regulates these processes.