The goals of our project are to isolate prostate epithelial populations, including prostate stem cells, and to evaluate the tumorigenic and the metastatic capacity of such populations upon transformation by genetic mutations that are commonly associated with human prostate cancer. Here we describe our progress in characterizing a mouse model of prostate cancer initiated by prostate epithelial cell specific deletion of the tumor suppressors, Pten and Tp53. Loss of PTEN and loss of TP53 are common genetic aberrations occurring in prostate cancer. PTEN and TP53 contribute significantly to the regulation of adult stem/progenitor cells in somatic tissues, and we hypothesized that combined deletion of these tumor suppressors in prostate epithelium would provide insight into the pathophysiology of transformed prostate progenitor cells. Using the PB-Cre4; Ptenfl/fl;TP53fl/fl model of prostate cancer, we describe the histologic and metastatic properties of primary tumors, of transplanted primary tumor cells, and of clonal cell lines established from tumors. Acquired resistance following treatment of progressive prostate cancer with androgen deprivation therapy is a significant clinical problem. An adaptive mechanism of resistance hypothesizes that castration tolerant stem/progenitor cells are precursors of castration resistant prostate cancer (CRPC) that subsequently develops as a result of mutation and selection. Two properties, stem/progenitor characteristics and intrinsic androgen indifference, are of particular interest because of the mechanistic insight that may be applicable to some aspects of aggressive or advanced human prostate cancer. To address stem/progenitor properties, we analyzed clonal tumor initiating ability and lineage potential. Although classic, limiting-dilution tumor initiating cell assays previously have not been feasible in primary prostate cancer, we were able to show that rare tumor initiating cells exist in both basal and luminal fractions. This demonstrated for the first time in transplantation experiments the capacity for clonal growth from a self-renewing luminal cell. The development of organoid culture technology allowed us to analyze the lineage potential and self-renewing properties of fractionated luminal and basal populations. We determined that relative to WT prostate and Pten-/- tumors, Pten/Tp53 null tumors contained a greatly expanded luminal progenitor population, consisting of mainly luminal-committed cells but also a fraction of progenitors that showed bipotential (luminal and basal) differentiation potential. The basal fraction of primary tumors demonstrated a non-tumorigenic committed basal phenotype in organoid cultures, suggesting that a small fraction of basal cells, perhaps those with bipotential capacity, are the tumor initiating cells. The existence of luminal and bipotential progenitors is consistent with the production of adenocarcinoma and adenosquamous tumors from organoid cultures and clonal tumor-initiating cells. Finally and importantly, a proportion of self-renewing luminal progenitors are intrinsically resistant to inhibition of AR signaling. Neither normal nor transformed luminal progenitors are commonly observed giving rise to TP63+ basal cells using in vivo lineage tracing approaches. One possible explanation for the discrepancy between in situ and transplantation results is that the bipotential capacity of luminal progenitors is generally not expressed in the prostate but is revealed in other microenvironmental settings such as organoid culture and subcutaneous tumors. We have used orthotopic tumor models of Pten/Tp53 null cell lines to investigate specific questions related to progression of aggressive disease. To interrogate the metabolic alterations associated with survival following ADT, metabolite profiles and associated regulators were compared in tumors from androgen intact mice and in tumors surviving castration. AR inhibition led to changes in the levels of glycolysis and TCA cycle pathway intermediates. As anticipated for inhibitory feedback between AR and PI3K/AKT signaling pathways, pAKT levels were increased in androgen deprived tumors. Elevated mitochondrial HK2 levels and enzyme activities also were observed in androgen-deprived tumors, consistent with a role for pAKT dependent HK2 protein induction and mitochondrial association. Competitive inhibition of HK2 mitochondrial binding in prostate cancer cells led to decreased viability. These data argue for AKT-associated HK2-mediated metabolic reprogramming and mitochondrial association in PI3K driven prostate cancer as one survival mechanism downstream of AR inhibition, and in addition, potentially inform the use of biomarkers and metabolic imaging for monitoring the ADT responses of prostate cancers with documented PI3K/AKT activation. In summary, from our characterization and analysis of the Pb-Cre4;Ptenfl/fl;Tp53fl/fl model we conclude that loss of Tp53 and Pten in Probasin-targeted cells leads to luminal stem/progenitor cell amplification and lineage plasticity. Although the GEMM has limitations with respect to faithfully mirroring human prostate cancer progression, cell lines and primary cells from adenocarcinoma tumors have been useful for establishing concepts concerning a variety of properties expressed by Pten/Tp53 null, poorly-differentiated, androgen indifferent cells. Screening of such lines in high throughput viability assays has identified Hsp90 inhibitors as active compounds in growth inhibition. This finding has been further validated in xenograft tumors.
