A ligandable PNT domain establishes ERG as a directly targetable oncogenic driver in prostate cancer

The TMPRSS2:ERG gene fusion, present in approximately 50% of prostate cancers in patients of European ancestry, drives oncogenesis through aberrant overexpression of the ERG transcription factor. Despite its role as a truncal oncogenic driver, ERG has been considered undruggable due to the absence of enzymatic activity and apparent lack of ligandable pockets. Here, we demonstrate continued dependency on ERG in metastatic prostate cancer and identify a druggable pocket within its N-terminal Pointed (PNT) domain. Using an inducible shRNA system in TMPRSS2:ERG-positive VCaP cells, we show that ERG depletion causes profound growth inhibition. To therapeutically exploit this vulnerability, we conducted a domain-focused differential scanning fluorimetry screen targeting the ERG PNT domain, followed by structure-activity relationship optimization. This approach yielded PBITE-1 (PNT-Binding Inhibitor of the Transcription factor ERG), a small molecule that selectively binds the ERG PNT domain. NMR chemical-shift perturbation mapping and molecular docking revealed that PBITE-1 engages a discrete, solvent-exposed surface comprising two α-helices and an adjacent flexible loop, defining a ligand-binding pocket within the PNT domain. In cellular models, PBITE-1 directly engaged ERG, selectively inhibited proliferation and invasion, and induced apoptosis in ERG-driven prostate and hematologic malignancies. PBITE-1 potently suppressed growth of ERG-positive mouse and human-derived prostate cancer organoids. Furthermore, PBITE-1 treatment significantly induced tumor cell apoptosis in VCaP xenograft models. These findings establish the ERG PNT domain as ligandable and provide preclinical evidence that ERG is directly targetable by small molecules, enabling future development of ERG-directed inhibitors and targeted protein degraders.