Discovery and Development of Mithramycin Analogs for Prostate Cancer Treatment by Direct Inhibition of the Oncogenic Transcription Factor ERG (TMPRSS2-ERG)

Androgen ablation therapy can treat prostate cancer, but patients will often manifest hormone-refractory prostate cancer, for which there are limited treatments. Recently, it was discovered that a fusion of a transcription factor ERG with a transmembrane protein TMPRSS2 (TMPRSS2-ERG), present in 60% of prostate cancers (but not in normal cells), is: 1) correlated with high incidence of metastatic disease and cancer-related deaths, 2) results in androgen independence by interefing with androgen receptor function. The importance of ERG is critical in that it was shown that RNAi-induced depletion of ERG or its ablation via blocking a deubiquitinase enzyme decreases proliferation and invasiveness in prostate cancer cell lines and blocks tumor growth in xenografts. Also, ectopic expression or ERG in mice promotes prostate oncogenesis (through the functional loss of tumor suppressor gene PTEN). This crucial role of ERG and the high incidence of the TMPRSS2-ERG gene fusion in prostate cancer provide the impetus for developing chemical interventions to determine its potential as a therapeutic target. To date, direct targeting of transcription factors has not proven successful. Recently, mithramycin (MTM), a DNA binding natural product, was identified in a screen of 50,000 molecules as a potent inhibitor of the oncogenic transcription factor EWS-FLI1, another fusion that drives Ewing sarcoma. In fact, MTM entered clinical trials as a Ewing sarcoma therapeutic (ClinicalTrials.gov, ID#NCT01610570). FLI1 and ERG share a nearly identical DNA binding domain (DBD), a critical functional component of their respective fusions with EWS and TMPRSS2 and a likely target of MTM. Because MTM is quite toxic in general, selective MTM analogues are needed. In our own preliminary studies we showed that MTM and its innovative 3-side chain analogues are potent inhibitors of prostate cancer cells containing TMPRSS2-ERG and, more importantly, that these compounds display selectivity against prostate cancer cells expressing TMPRSS2-ERG. Furthermore, the potency and selectivity of MTM is modulated by our innovative 3-side chain analogues of MTM. Because MTM is quite toxic (due to its inhibition of Sp transcription factors, unrelated to EWS-FLI1), higher selectivity of MTM analogues would provide a robust therapeutic window for its use as a prostate cancer therapeutic. Our hypothesis is that MTM has a unique mode of action against ERG that can be exploited to develop potent and selective MTM analogues for use against prostate cancers that express TMPRSS2-ERG. Our crystal structures of MTM-SA-Trp and MTM-SA-Phe in complex with DNA together with the published ERG-DNA structure strongly suggest that the selective cytotoxicity arises from interactions between the 3-side chain of MTM and the DNA binding domain of ERG. Moreover, our biochemical data show that, instead of displacing FLI1 DBD from DNA (as MTM does with Sp1), MTM stabilizes FLI1-DNA complex. We propose to generate MTM analogues highly selective against prostate cancer cells expressing TMPRSS2-ERG. On the basis of our preliminary data, we hypothesize that MTM analogs will bind DNA cooperatively with TMPRSS2-ERG, leading to a hyper-stable complex that will disrupt its oncogenic transcriptional activity, which in turn will widen the efficacy-toxicity dosing range, and may avail this molecule as a treatment in prostate cancer. In Project 1 (Rohr) of this Synergistic Idea Development Award application, we will use semisynthetic methodology to generate MTM analogues for and suggested by studies in Projects 2 and 3. In Project 2 (Tsodikov) we will elucidate biochemically and structurally how MTM and its analogues bind DNA, interact with ERG and affect its DNA binding function drive rational MTM analogue design. In Project 3 (Leggas) we will determine the effect of MTM analogues on TMPRSS2-ERG-mediated transcription and their pharmacology in preclinical models of prostate cancer (Leggas). We expect that these studies will (in short- term) 1) elucidate the mechanism of MTM antagonism of TMPRSS2-ERG and 2) lead to the discovery of novel selective MTM analogues. This will set the stage for testing these analogues in the clinic as a longer-term impactful objective.