Biosynthetic Investigation, Synthesis, and Bioactivity Evaluation of Putative Peptide Aldehyde Natural Products From the Human Gut Microbiota

Abstract

Natural products produced by the human gut microbiota may mediate host health and disease. However, discovery of the biosynthetic gene clusters that generate these metabolites has far outpaced identification of the metabolites themselves. Employing traditional culturing and isolation-based natural product discovery techniques in this setting is difficult due to a variety of technical challenges. ‘Cryptic’ gene clusters identified in these microbes often remain unexpressed under standard laboratory culture conditions, and it can be cumbersome to cultivate the anaerobic bacteria on the scale required for isolation efforts. Several isolation-independent approaches have recently been used to access the products of gut microbial biosynthetic gene clusters, including functional metagenomics, heterologous expression, and synthesis of predicted natural product structures, and the continued development of such approaches may enable a greater understanding of the chemical ecology of the human gut. In this work, we used an isolation-independent approach that combined biosynthesis and chemical synthesis to access putative products of a family of nonribosomal peptide synthetase(NRPS)-encoding gene clusters from the human gut microbiota. These NRPS gene clusters all contain terminal reductase domains, indicating that they may produce peptide aldehyde products. Such natural products are known to act as inhibitors of serine, cysteine, and threonine proteases. We initially targeted an NRPS gene cluster in this family from the abundant gut commensal microbe Ruminococcus bromii (the rup gene cluster) for small molecule discovery. Using a combination of bioinformatic analyses and in vitro biochemical characterization of biosynthetic enzymes, we predicted that this gene cluster likely generates N-acylated dipeptide aldehyde(s), collectively referred to as the ruminopeptin(s), and gained insight about the biosynthetic building blocks likely incorporated by its main NRPS enzyme. These results demonstrate the utility of combining bioinformatic predictions with in vitro biochemical assays for predicting the structures of natural products. We next used a short solution-phase synthesis to access 12 predicted peptide aldehyde structures of ruminopeptin(s). Many of these compounds contain a glutamyl aldehyde at their C-terminus. Therefore, we predicted that they may target the glutamyl endopeptidases, a family of serine proteases found in several human opportunistic pathogens including Staphylococcus aureus and Enterococcus faecalis. We found that several putative ruminopeptins inhibited the S. aureus glutamyl endopeptidase SspA (also known as endoproteinase GluC or V8 protease). We also identified homologs of this protease encoded in the genomes of gut commensals and opportunistic pathogens as well as metagenomes from the human gut environment. We hypothesize that inhibition of this family of proteases by ruminopeptin(s) may be important for mediating microbe-microbe interactions in the human gut. Inspired by the success of this approach in discovering interesting small molecules with potentially physiologically relevant activity, we next expanded the scope of our study to investigate additional predicted peptide aldehydes from gut microbial genomic data. We selected four additional NRPS gene clusters of interest from gut commensal organisms and synthesized a comprehensive set of their bioinformatically predicted biosynthetic products to reach a total library size of 48 peptide aldehydes. We evaluated these compounds as potential inhibitors of human proteases, antibiotics against a set of prominent pathogens and commensals, and inhibitors of gut microbial secreted protease activity. Finally, we designed and synthesized an activity-based probe based on the structure of a peptide aldehyde and attempted to identify potential protein targets of this compound in Clostridioides difficile 630Δerm using an untargeted chemoproteomics workflow. Small molecule target identification and validation in the complex environment of the human gut is an emerging area of interest, and the approaches we have employed to investigate the biological activities of peptide aldehydes illustrate the current challenges and opportunities in this field.