The major goal of my laboratory is to understand how critical transcription factors mediate prostate cancer oncogenesis. Prostate cancers uniquely harbor very few mutations in well-known driver oncogenes (e.g., RAS, BRAF, etc.). Instead, they are characterized by aberrations in transcription factors, including the androgen receptor (AR) and ETS family transcription factors. These oncogenic transcription factors regulate a transcriptional program that determines cellular lineage, specifies the responses to external stimuli, and contributes to the oncogenic cellular context in prostate cancer. Our long-term goal is to generate an integrative understanding of how transcription factors normally guide the prostate lineage determination and how their deregulation leads to cancer. In addition, using biochemistry and genetic screens, we are pursuing ways to target the activity of these transcription factors that can lead to drug discovery.

One current focus of the laboratory is on ETS family transcription factors. Genomic translocation involving ERG and other ETS family members is found in more than 50 percent of all prostate cancers. The predominant translocation involves fusion of the untranslated region of the TMPRSS2 gene that is highly expressed in the prostate with ERG resulting in ERG overexpression. However, overexpression of ERG alone in non-transformed prostate cells and in the mouse prostate is not transforming. The oncogenic transformation potential of  ETS factors is exceptionally dependent on the cellular context. The cellular context can be determined by the cellular lineage and/or the presence of cooperative lesions. Thus, critical questions are: 1) What are the key oncogenic pathways regulated by ETS expression? 2) What are the optimal cellular contexts for ETS to mediate tumorigenesis? and 3) What other molecular lesions cooperate with ETS and what is the mechanism of cooperativity?

To address these questions in the endogenous tissue context we have built several genetically engineered mouse models that are designed to express the TMPRSS2-ERG fusion protein in a spatiotemporally controllable manner and in cooperation with other oncogenic drivers. These models have revealed that ERG expression alone in certain contexts can promote proliferation and inhibit cellular differentiation. Yet, in the prostate, ERG expression alone is insufficient to cause significant phenotype, but it strongly cooperates with loss of the tumor suppressor PTEN in tumorigenesis.

Going forward, we will utilize these mouse models to screen for other genetic lesions that cooperate with ERG expression and to define the cell of origin of ERG-positive prostate cancer. In addition, using massively parallel sequencing to map the chromatin state and ERG binding sites, we will study the mechanism of the context dependence of ERG-mediated oncogenesis and determine the factors that mediate both ERG binding to the genome and downstream transcriptional output after ERG binding.