The glucocorticoid receptor (GR) is a cortisol-activated transcription factor and chromatin remodeling protein that is variably expressed in ovarian cancer (OvCa). GR-mediated gene expression encoding anti-apoptotic pathway proteins is associated with chemotherapy-resistance in OvCa, while pre-treatment with a selective GR antagonist enhances chemotherapy sensitivity. Our group also recently discovered that amongst a range of OvCa histologic subtypes, both higher GR (NR3C1) mRNA and higher GR protein levels are associated with overall worse clinical outcome. We now propose to develop a ?GR transcriptional activity signature? (GRSig) for OvCa to better assess GR activity in OvCa with the goal of identifying patients most likely to benefit from addition of GR antagonism to chemotherapy. Using cell line models of GR+ OvCa representing a variety of histologies (Aim 1), we will identify genes whose expression is either significantly upregulated or repressed following GR activation by physiological glucocorticoid concentrations as well as significantly reversed in expression by GR antagonism. GR ChIP-seq will be employed to determine which of these genes are putative direct (rather than indirect) GR target genes with promoter or enhancer region GR chromatin association, further refining the GRSig to represent canonical OvCa GR activity. The resulting GRSig will then be used to assess relative GR activity in existing OvCa samples from a pooled public GEO dataset (Discovery set) containing normalized RNA expression and outcome (PFS and OS). A GRSig score cutoff will be calculated that identifies patients with worse prognosis in the GEO dataset. A second, well-annotated Mayo Clinic OvCa cohort (Validation set) will then be used to validate the cutoff. We will also test the hypothesis that higher GRSig scores (representing higher GR transcriptional activity) will associate with PFS and OS more strongly than GR (NR3C1) mRNA expression alone.
In Aim 2, we will examine the GRSig in N=131 OvCa PDX models with chemotherapy response data to test the hypothesis that higher GRSig score associates with relative chemotherapy resistance. Furthermore, we will explore the relationship of GRSig score to GR protein expression by performing GR IHC in the PDX tumors, and then evaluating the strength of association of the GRSig score relative to GR IHC score with respect to chemotherapy response. In an exploratory analysis, GRSig-scored PDX models will be treated with chemotherapy +/- a GR antagonist (GRA) to test the hypothesis that pre-treatment with a GRA improves tumor shrinkage and/or lengthens time to tumor regrowth in GRSig high PDX models. Completion of this project will identify a GR transcriptional signature designed to identify women with OvCa who have a worse prognosis and to for whom addition of a selective GR antagonist to chemotherapy is expected to improve outcome compared to standard treatment with chemotherapy alone.
Our laboratory has discovered that the body's stress hormone receptor, termed the glucocorticoid receptor (?GR?), is expressed at particularly high levels in ovarian cancers that have a high risk of relapse. We have begun to identify the genes that are increased or decreased in expression as a result of GR activity associated with overexpression, and found that these genes indeed can contribute to tumor aggressiveness and resistance to chemotherapy. Here we propose to develop an ovarian cancer-specific GR ?gene signature score? that can be measured in a patient's tumor sample; eventually we hope this assigned tumor score can be used to determine which patients will benefit from receiving GR blockers designed to reduce the risk of ovarian cancer recurrence.
