Massively parallel discovery of peptides to inhibit cellular protein interactions

Presented on November 13th 2024 by Andrew Savinov Abstract: Peptides can act as potent modulators of protein folding and binding interactions, allowing them to function as inhibitors and regulatory elements. Peptide fragments of larger proteins are particularly attractive candidates for achieving these functions due to their inherent potential to form native-like binding interactions. However, it was previously unclear how functional such fragments are in isolation. We have developed an experimental approach to perform high-throughput functional characterization of protein fragments in living cells. By tiling more than 10,000 peptide fragments across a variety of protein sequences, we found that inhibitory fragments map out key protein interaction sites across structurally and functionally diverse proteins, and determined principles of fragment-based inhibition (Savinov et al., PNAS 2022). Further, we have now developed a computational method, FragFold, that systematically employs AlphaFold to predict protein fragment binding to full-length protein targets, in order to predict which of the many possible protein fragments are inhibitory (Savinov and Swanson et al., bioRxiv 2023). We establish that our approach is a sensitive predictor of protein fragment function: Across full protein sequences, 68% of FragFold-predicted binding peaks match experimentally measured inhibitory peaks. The success rate of FragFold demonstrates that this computational approach should be broadly applicable for discovering inhibitory protein fragments across proteomes. Overall, we find that protein fragments are generalizable, genetically encodable inhibitors of protein interactions which can be readily designed with our FragFold framework.