PROJECT SUMMARY/ABSTRACT: Immune checkpoint inhibitors (ICI) are active in only 10-25% of metastatic, castrate-resistant prostate cancer (mCRPC) patients. Co-clinical immune profiling studies of the tumor microenvironment (TME) in mCRPC patients and murine PC models demonstrated sparse immune infiltrates, with predominance of immunosuppressive myeloid cells, particularly tumor-associated macrophages (TAM). PTEN loss-of-function (LOF) occurs in multiple cancers, and specifically in 50-75% of mCRPC. It is associated with poor prognosis, therapeutic outcomes and de novo/acquired resistance to ICI in preclinical and clinical studies in multiple tumor types. In addition to the predominance of TAM within the sparse immune infiltrate in PC, we observed a 2-fold increase of myeloid derived suppressor cells within the PTEN LOF TME, relative to isogenic PTEN-proficient counterparts. These data reinforced our project goal of elucidating mechanisms underlying altered cancer/myeloid cell cross-talk in PTEN LOF mCRPC, that are likely similar across PTEN-deleted malignancies and contribute to ICI resistance. We have demonstrated that cyclic GMP-AMP Synthase (c-GAS)/Stimulator of INterferon Genes (STING) pathway, typically activated in response to cytosolic DNA double strand breaks (DSB), is frequently silenced in cancer cells, and cGAS/STING activation within myeloid compartment of the TME is critical for generating a robust immune infiltrate. However, generation of DNA DSB is insufficient to enhance T cell infiltration, which is consistent with lack of response to PARP inhibition (PARPi) plus ICI observed in our murine models and clinical studies. Probing more deeply into this disconnect, we found that failed STING pathway activation within TAM was responsible for this resistance. Furthermore, PI3K activity was preventing STING pathway activation in TAM, and PI3K inhibition (PI3Ki) plus DNA damage with PARPi led to ICI responsiveness in PTEN-proficient, but not in isogenic PTEN-knockout PC, suggesting that additional immunosuppressive mechanisms are driven by PTEN LOF. The central hypothesis of this proposal is that PTEN LOF PC are de novo resistant to DNA DSB sensing c-GAS/STING pathway activation, which can be overcome by direct STING agonist-based combinations, leading to ICI sensitization. To test this hypothesis, we propose the following specific aims. First, we will dissect the TME in a completed investigator-initiated clinical trial of PARPi/ICI in mCRPC patients, to determine whether the immune inhibitory mechanisms identified in preclinical models are relevant to patients. Second, we will elucidate the cancer cell extrinsic mechanism(s) by which PTEN-deficient PC are de novo resistant to DNA-sensing STING pathway activation. Third, we will interrogate the anti-cancer mechanism and therapeutic potential of direct STING agonist/PI3Ki combination therapy in sensitizing PTEN-knockout murine CRPC to ICI. Collectively, these “co-clinical” studies will pro...