Multi-pronged Approach to Score and Rank PROTAC-mediated Ternary Complex Models

Targeted protein degradation leverages the proteasome dependent proteolysis of ubiquitinated proteins as a therapeutic approach. The formation of a binary complex between the E3 ligase and the protein of interest is a key requirement for degradation mediated by ubiquitination. Proteolysis targeted chimeras (PROTACs) are bivalent molecules that recruit an E3 ligase and the target protein to the ternary complex. Consequently, the targeted degradation of proteins using bivalent degraders as well as molecular glues has emerged as a novel treatment modality. While the targeted protein degraders are rapidly realizing their promise of being a therapeutic, their rational design is not widely practiced. The reason for this includes the challenges involved in the accurate prediction of protein-ligase conformational complex which can productively bind to the PROTAC. Given the large surface area of the proteins, there can be several hundred ways in which two proteins can form a binary complex. Identifying a conformation from these solutions that can bind to the degrader is non-trivial. An additional limitation is the conformational sampling of the PROTAC linker in the context of the proteinligase binary complex. In this study, we seek to address this challenge of appropriately ranking the models. The pipeline developed by us generates models of PROTAC-mediated ternary complex from the 3D structures of the protein, E3 ligase and the PROTAC. The generated models are then ranked using a combination of energy terms, geometric and knowledge-based frameworks and contact features to identify plausible ternary complexes. This method was benchmarked on a set of 35 PmTC structures available in the PDB which use VHL as the ligase. For 80% of this set, the best complexes (within 5Å RMSD) identified by our method were ranked among the top 15 models out of several hundred solutions. By seeking to address the ranking challenge, our method takes an important step towards making prospective structureguided design of novel PROTACs for the targeted degradation of any protein a reality.