Significance: Immune checkpoint blockade (ICB) antibodies targeting CTLA-4 and PD-1/PD-L1 are having major impact in a wide range of cancers. However, efficacy in advanced prostate cancer (PC) has been limited. This limited efficacy is largely attributed to the non-T cell inflamed "immune desert" tumor microenvironment (TME) within advanced PCs. PTEN loss-of-function (LOF) occurs in 50-75% of advanced PC and is associated with poor clinical outcomes. More specifically, PTEN loss-of-function (LOF) leads to the well-known hyperactivation of the PI3K pathway, resulting in alterations in tumor metabolism. Moreover, PTEN LOF has also been shown to drive an immunosuppressive TME, which has been recently associated with lack of T cell infiltration and de novo and acquired resistance to ICB. The impact of metabolic alterations induced by PTEN LOF on the immune microenvironment remains poorly understood. Furthermore, the impact of dietary intervention strategies on immunomodulation in PC has been largely unexplored. The central hypothesis of this proposal is that ketogenic diet (KD) can reprogram fatty acid metabolism, thereby enhancing T cell infiltration and immunotherapy efficacy in PTEN-deficient PC. A deeper understanding of the impact of KD on immunometabolism within the TME of PTEN-deficient PC, will lead to the development of novel diet-based immune-oncology (IO) combination strategies to sensitize lethal PTEN-deficient cancers to ICB.

Innovation: This proposal seeks to elucidate the mechanistic basis for how KD alters intratumoral lipid metabolism and enhances T cell infiltration within the TME of PTEN-deficient murine PC models. Critically, we will utilize these mechanistic insights to develop novel diet-based immuno-oncology combination strategies to eradicate PTEN-deficient lethal PC. We believe that this integrative approach that leverages our team's collective expertise in cancer biology and metabolism, tumor immunology and experimental therapeutics, will lead to the generation of novel diet-based combinatorial immuno-oncology therapeutics to treat advanced prostate cancer. We envision that the results from this pilot proposal will have considerable impact in a relatively understudied area of translational cancer research. Given the metabolic parallels between ketogenic diet and intermittent fasting induced-ketosis, we anticipate that the mechanistic insights gained from this project, will address one of twenty NCI provocative questions (PQ2) i.e. how intermittent fasting can affect cancer incidence, treatment response, or outcome, and result in high-impact publications and future funding.

Aim 1. Elucidate the mechanism(s) by which ketogenic diet alters the immunometabolic profile within tumors derived from murine syngeneic and genetically engineered mouse PTEN-deficient models of prostate cancer.

Aim 2: Evaluate the therapeutic impact of ketogenic diet, singly and in combination with PI3K inhibitor, STING agonist or NLRP3 agonist, on sensitization of murine PTEN-deficient syngeneic and GEM models of PC to ICB.

In Aim 1, we will take a deep dive to elucidate the mechanistic link between enhanced energy stress pathway activation, suppression of fatty acid synthesis and enhanced T cell infiltration induced by KD. Using syngeneic Myc-CAP isogenic models -/+ PTEN (via CRISPR-Cas9 knockdown), we will measure serum ketones, glucose and adipokines (insulin, IGF-1, IL-6, TNF-a, adiponectin) from standard diet (SD)-, high-fat diet (HFD) and KD-fed mice, to confirm the expected physiologic states induced by dietary intervention. We will then perform RNA-seq analysis on the tumors harvested from these mice following 2 weeks of treatment, and utilize gene set enrichment analysis to dissect signaling, metabolic and immune pathways that are differentially expressed following KD vs. SD or HFD administration, comparing PTEN-proficient and PTEN-deficient syngeneic Myc-CAP models. Next, we will execute integrative bioinformatic analysis with analogous data obtained from PTEN-deficient and PTEN/p53-deficient GEMMs. For all these studies, we will compare with normal non-tumor bearing tissues (e.g. liver) from the same animals, to assess tumor-specific vs. global changes in metabolism/immune infiltration within other tissues. As a first step towards understanding the impact of KD on antigen specificity of tumor-infiltrating T cells, we will engraft C57BL6 mice with B6-Myc-PTEN knockout cells expressing chicken ovalbumin protein (OVA). Using flow cytometry and SIINFEKL/H-2Kb tetramer staining of single cell suspensions from harvested tumors, we will identify antigen-specific specific CD8 T cells within the TME. Furthermore, we will perform Granzyme B or CD44 co-staining by flow cytometry, to look at activation/cytolytic capacity of the tumor-infiltrating T cell populations following in vivo dietary manipulation. Critically, to directly assess the impact of dietary manipulation on T cell effector function, we will conduct adoptive T cell transfer of OT-I cells in KD vs. HFD vs. SD fed immunodeficient mice bearing OVA expressing B6-Myc-PTEN knockout tumors. In collaboration with Dr. Alexander Muir, a tumor metabolism expert at the University of Chicago Ben May Department for Cancer Research, we will isolate tumor interstitial fluid (TIF) by centrifugation (publication protocol link; https://bio-protocol.org/e3427), and perform unbiased metabolomics analysis by liquid chromatography-mass spectrometry, and measure absolute concentrations of targeted metabolites, including fatty acids and triglycerides, to assess whether the differential energy stress pathway regulation translates to changes in fatty acid metabolism within the TME of KD- vs. HFD or SD-fed PTEN-deficient syngeneic and GEMM mice. In addition, we will perform cytokine array profiling on TIFs isolated from KD vs. SD-fed mice, and correlate with the immune cell infiltrate analysis by flow cytometry. The absolute quantification of metabolite levels in TIF between tumor samples from KD- vs. HFD- or SD-fed mice, will then enable generation of tissue culture media that mimics physiological conditions found in a tumor, thus expanding the range of in vitro and ex vivo experiments that can be carried out under physiological nutrient conditions that mimic dietary manipulations in vivo. Using ex vivo co-culture systems under physiological nutrient conditions that mimic KD vs. HFD vs. SD-fed conditions, we will perform cytokine profiling and immune cell trafficking studies. We will also perform co-culture assays of splenic T cells with dendritic cells that are pulsed with tumor lysates from B6-Myc-PTEN knockout cells, under physiological nutrient conditions that mimic SD vs. HFD vs. KD. Collectively, these studies will shed light on the mechanism(s) by which KD selectively alters intratumoral metabolism and enhances immune cell trafficking and their functionality within murine PTEN-deficient cancers. The above studies will also allow assessment of relative contributions of tumor cell intrinsic vs. extrinsic mechanisms towards the generation of anti-tumor immune response elicited by KD.

One of the fundamental barriers to success of ICB in PC is the presence of an "immune-desert" TME, which is particularly prevalent in PTEN-deficient tumors. In vivo immune profiling studies in our lab demonstrated that KD resulted in a global increase in immune cell infiltration within the TME of myc-CAP syngeneic mice with CRISPR/CAS9-mediated knockdown of PTEN, relative to their PTEN-proficient counterparts. Furthermore, KD-fed PTEN-deficient GEMM mice demonstrate an increase in CD8 T cell infiltration relative to SD- or HFD-fed mice. These data demonstrate that KD can immunologically "warm" the TME within PTEN-deficient cancers. The central hypothesis of Aim 2 is that KD-induced immune cell infiltration, singly and/or in combination with other IO agents, will sensitize PTEN-deficient PC to ICB. In this aim, we propose rational immune-oncology (IO) combination strategies that capitalize on the observation that KD enhances T cell infiltration within PTEN-deficient prostate tumors. Our prior studies have demonstrated that copanlisib (pan-PI3K inhibitor, PI3Ki) treatment increases CD4 and CD8 T-cell infiltration and decreases Gr-MDSC and Tregs within the TME of PTEN/p53-deficient prostate GEMMs. In addition, activation of the evolutionarily conserved innate immune STING pathway with a STING agonist, resulted in decreased MDSC infiltration, and reduction in inhibitory VISTA and PD-L1 checkpoint expression on MDSCs, thus demonstrating that both PI3Ki and direct STING agonist can create an inflammatory TME poised for synergy with KD and ICB. Furthermore, recent studies have described the role of NOD-, LRR- and pyrin domain-containing protein-3 (NLRP3) as a sensor for pathogen-derived danger signals by antigen-presenting cells in the innate immune system. However, relatively little is known about the contribution of NLRP3 to modulation of immunotherapy responsiveness. Through a recent Bristol Myers Squibb International OncologyNetwork collaboration, we have access to the BMS systemic STING agonist and NLRP3 agonist for these mechanistic and therapeutic studies. We will evaluate the following parallel therapeutic strategies, singly and in combination with PD-1 and/or CTLA4 blockade, to enhance T cell infiltration within the TME and eradicate PTEN-deficient prostate cancers: 1) KD alone; 2) KD+copanlisib; 3) KD+BMS systemic STING agonist; 4) KD+BMS NLRP3 agonist; 5) KD+copanlisib+BMS systemic STING agonist; 6) KD+copanlisib+BMS NLRP3 agonist. KD- vs. SD-fed mice treated with the listed IO combinations, singly and in combination, will be tested in Myc-CAP-PTEN knockout syngeneic and PTEN- and PTEN/p53-deficient GEMM mice, under uncastrated and castrated conditions. For each cohort, we will perform acute (2-week) treatment for target modulation and immunometabolic studies, and long-term therapeutic studies to assess anti-tumor efficacy (4-weeks). For assessment of changes in the T cell repertoire under different treatment conditions, we will isolate T cells from tumors in mice, and perform TCR sequencing analysis. We will non-invasively monitor anti-tumor responses in syngeneic models and GEMMs by tumor caliper measurements and MRI, respectively, and mice will be euthanized when the tumors reach 2 cm in long-axis diameter. Tumor sections at the time of euthanasia will be harvested for flow cytometry, Western blot analysis, IHC and bulk-RNA-seq. We will also determine the impact of sequential vs. concomitant administration of combinations with ICB administration on therapeutic efficacy. Taken together, the pre-clinical credentialing of these novel KD-based IO combination strategies will provide the mechanistic foundation for the next wave of clinical trials to eradicate PTEN-deficient prostate cancer.

All required software for data analysis will be provided by the lab. Training sessions on how to use analysis software will be provided by the more senior members of the laboratory.

The student will present in various forums/conferences, including our laboratory and institutional working research group meetings, and larger international conferences organized by the American Association for Cancer Research and Prostate Cancer Foundation