Targeting Human Bromodomain Proteins for the Treatment of Medulloblastoma

Medulloblastoma (MB) is a malignant pediatric brain tumor, where MYC overexpression is associated with poor prognosis and reduced survival. Directly targeting MYC remains difficult due to the lack of a druggable site. Targeting epigenetic readers such as BET bromodomain proteins, particularly BRD4, has emerged as a promising strategy. In MYC-driven MB, the BD1 and BD2 domains of BRD4 recognize acetylatedlysine residues, and the C-terminal motif recruits transcriptional machinery to promote MYC transcription. Most pan-BET inhibitors (iBET) bind equipotently to all bromodomains (BRD2/3/4/T), exhibiting no domain selectivity and dose-limiting toxicity. While these compounds highlight the therapeutic potential of iBETs, domain selectivity remains a challenge. We structurally characterized two ligands SRX3305, a BRD4 inhibitor exhibiting nanomolar-range affinity but lacked structural characterization, and MDP20, a structural analog of MDP5, a known iBET. We solved the crystal structure of a BRD4-BD2/SRX3305 complex, which reveals a ligand-induced conformational change within the acetyl-lysine binding pocket that could be leveraged to improve BD2 selectivity. Several Fluorescence Polarization-based probes have been developed to aid iBET drug discovery, but few have been characterized. To address this, we developed and characterized a fluorescent chemical probe based on structural insights from BRD4-BD2/SRX3305 complex. The probe demonstrated an affinity in the 2-25 µM range across BD1/BD2 of BRD2 and BRD4, with preference for BD1. To further assess the possible selectivity of these iBETs, we determined the crystal structure of the BRD2-BD2/SRX3305 complex along with MDP20 in complex with BRD2-BD2 and BRD4-BD2. These structural insights will guide the design of domain-selective iBET.