Abstract 1537 Developing protein-based de novo-designed human ClpP inhibitors using deep-learning

Human ClpP is a mitochondrial protease that is overexpressed in certain cancers, including in acute myeloid leukemia. Unlike its bacterial homologs, human ClpP exists in two oligomeric states: a tetradecameric complex composed of two stacked homo-heptameric rings, and the individual heptameric rings. Only the tetradecameric complex is capable of catalyzing proteolysis. Leveraging recent advances in deep-learning based protein design, we aimed to design de novo protein binders that target human ClpP to inhibit its assembly into the active tetradecameric form.