Abstract A36: Identification of master regulators driving advanced prostate cancer and treatment response through the assembly of mouse and human prostate cancer interactomes

Abstract

Abstract Subsets of prostate tumors are highly aggressive and progress to a castration-resistant stage, which is metastatic and invariably lethal. Prostate tumors are associated with genetic perturbations such as chromosomal alterations, but rarely with mutational events. Loss of the NKX3.1 gene is an early and prevalent event while deletion of PTEN tumor suppressor, is prevalent in early disease and strongly associated with metastatic tumors. Ras is rarely mutated in prostate tumors but components of the RAS signaling pathway are frequently deregulated in primary tumors and profoundly altered in metastases. Additionally, translocations resulting in constitutive activation of Kras occur in a subset of aggressive prostate cancers. Our laboratory has developed new genetically-engineered mouse models (GEM) based on inducible prostate-specific deletion of Pten alone or together with activation of oncogenic Kras. These models recapitulate all stages of prostate tumorigenesis, including distant metastases with high penetrance. We have exploited these GEM models using reverse-engineering of regulatory programs to develop genome-wide molecular networks of transcriptional and post-translational interactions, namely interactomes. These approaches are unbiased and encompass the entire genome. We have focused on comparing regulatory networks for both mouse and human prostate cancer. Importantly, interactomes are predicated on the availability of large gene expression profile (GEP) datasets representative of the natural phenotypic variability and genetic perturbations of prostate cancer. We have generated GEP using a collection of prostate cancer mouse models from our laboratory, as well as from the mouse modeling community. Each model has been treated with a variety of targeting drugs against members of the androgen-signaling, MAPk, Akt/mTOR, JAK/STAT, NF-KB pathways and other kinase inhibitors as well as chemotherapeutic agents. These perturbations have been instrumental in increasing transcriptome variability within the same genetic background. In parallel, a human prostate cancer interactome has been generated using publicly available datasets. Interrogation of these interactomes represents a new paradigm for ‘human to mouse to human’ analyses. In particular, we have used these interactomes to identify molecular mechanisms and biomarkers of response to combinatorial targeting of Akt/mTOR and MAP kinase pathways in our metastatic prostate cancer GEM model of Kras activation and Pten loss-of-function. Molecular profiles from these tumors parallel those of human prostate cancers having poor prognosis, concomitant activation of Akt/mTOR and MAP kinase signaling and displaying robust phosphorylation (inactivation) of FoxO3a tumor suppressor. This combined therapy profoundly reduces tumor and metastatic burden, improves survival, decreases FoxO3a phosphorylation, and is significantly more effective than standard of care chemotherapy. Computational analyses show that combination therapy in mice shifts the malignant signature toward that of normal prostate, while cross-species analyses identify a two-gene signature downstream of FoxO3a, including FOXM1 and CENP-F that is expressed in human metastatic tumors and is predictive of clinical outcome. Our findings suggest that targeting the FoxO3a regulatory network may be particularly beneficial for patients who have failed conventional chemotherapy. Citation Format: Alvaro Aytes, Antonina Mitrofanova, Celine Lefebvre, Michael M. Shen, Andrea Califano, Cory Abate-Shen. Identification of master regulators driving advanced prostate cancer and treatment response through the assembly of mouse and human prostate cancer interactomes [abstract]. In: Proceedings of the AACR Special Conference on Advances in Prostate Cancer Research; 2012 Feb 6-9; Orlando, FL. Philadelphia (PA): AACR; Cancer Res 2012;72(4 Suppl):Abstract nr A36.