In silico prohormone processing

Abstract

Peptide hormones are processed from larger proteins by endopeptidases and have diverse biological functions. For example, glucagon-like peptide-1 (GLP-1) is well known for regulating blood glucose levels and has been used to treat diabetes and obesity. However, there are a vast number of peptides that are potentially processed by endopeptidases to create functional peptide hormones. Now, Coassolo et al. have developed a computational tool called Peptide Predictor for discovering unknown peptide hormones. The tool screens the human secretome for amino acid sequences containing at least four instances of dibasic motifs KR, RR, RK, and KK that are recognized and cleaved by peptidases from the prohormone convertase family to process at least five potential peptide hormones. From 2,082 proteins in the secretome, the team identified 373 proteins that generated 2,683 peptides. They used the existing tools PeptideRanker and MultiPep to predict that 21–23% of the peptides were bioactive. To validate their approach, Coassolo et al. screened a library of 100 of the peptides — each only homologous to their precursor protein — for expression of the gene Fos, which is associated with the bioactivity of peptides. They focused on the peptide BRP, which had increased Fos expression by greater than tenfold. BRP was found to be processed from the precursor BRINP2, primarily expressed in the brain. BRP suppressed food intake dose dependently in both lean mice that were fasted and in diet-induced obese mice. Mechanistically, BRP treatment led to the activation of protein kinase A, phosphorylation of the protein CREB, and increased expression of FOS in specific regions of the hypothalamus and other parts of the brain. An AlphaFold model and mutagenesis showed Leu8 as a key residue for BRP activity. Thus, this study shows the advantage of computational tools for proteome-wide scanning of cleavage sites by proteases for the discovery of therapeutic peptides.

Original reference: Nature https://doi.org/10.1038/s41586-025-08683-y (2025)