Parker, J. K. | Search bioRxiv

bioRxiv2023

Semantic search using protein large language models detects class II microcins in bacterial genomes

Class II microcins are antimicrobial peptides that have shown some potential as novel antibiotics. However, to date only ten class II microcins have been described, and discovery of novel microcins has been hampered by their short length and high sequence divergence. Here, we ask if we can use numerical embeddings generated by protein large language models to detect microcins in bacterial genome assemblies and whether this method can outperform sequence-based methods such as BLAST. We find that embeddings detect known class II microcins much more reliably than does BLAST and that any two microcins tend to have a small distance in embedding space even though they typically are highly diverged at the sequence level. In datasets of Escherichia coli, Klebsiella spp., and Enterobacter spp. genomes, we further find novel putative microcins that were previously missed by sequence-based search methods.

microbiology↗

bioRxiv2023

Export of diverse and bioactive peptides through a type I secretion system.

Microcins are peptide antibiotics secreted by Gram-negative bacteria that inhibit the growth of neighboring microbes. They are exported from the cytosol to the environment in a one-step process through a specific type I secretion system (T1SS). While the rules governing export of natural or non-native substrates have been resolved for T1SSs that secrete large proteins, relatively little is known about substrate requirements for peptides exported through T1SSs that secrete microcins. Here, we investigate the prototypic microcin V T1SS from Escherichia coli and show it can export a remarkably wide range of natural and synthetic peptides. We demonstrate that secretion through this system is not affected by peptide charge or hydrophobicity and appears only constrained by peptide length. A varied range of bioactive peptides, including an antibacterial peptide, a microbial signaling factor, a protease inhibitor, and a human hormone, can all be secreted and elicit their intended biological effect. Secretion through this system is not limited to E. coli, and we demonstrate its function in additional Gram-negative species that can inhabit the gastrointestinal tract. Our findings uncover the highly promiscuous nature of peptide export thorough the microcin V T1SS, which has implications for native cargo capacity and use of Gram-negative bacteria for peptide research and delivery. ImportanceMicrocin type I secretion systems in Gram-negative bacteria transport antibacterial peptides from the cytoplasm to the extracellular environment in single step. In nature, each microcin secretion system is generally paired with a specific peptide. We know little about the export capacity of these transporters and how peptide sequence influences secretion. Here, we investigate the microcin V type I secretion system. Remarkably, our studies show this system can export diverse peptides and is only limited by peptide length. Furthermore, we demonstrate that various bioactive peptides can be secreted, and this system can be used in Gram-negative species that colonize the gastrointestinal tract. These finding expand our understanding of secretion through type I systems and their potential uses in peptide applications.

microbiology↗

bioRxiv2022

A Principal Odor Map Unifies Diverse Tasks in Human Olfactory Perception

Mapping molecular structure to odor perception is a key challenge in olfaction. Here, we use graph neural networks (GNN) to generate a Principal Odor Map (POM) that preserves perceptual relationships and enables odor quality prediction for novel odorants. The model is as reliable as a human in describing odor quality: on a prospective validation set of 400 novel odorants, the model-generated odor profile more closely matched the trained panel mean (n=15) than did the median panelist. Applying simple, interpretable, theoretically-rooted transformations, the POM outperformed chemoinformatic models on several other odor prediction tasks, indicating that the POM successfully encoded a generalized map of structure-odor relationships. This approach broadly enables odor prediction and paves the way toward digitizing odors. One-Sentence SummaryAn odor map achieves human-level odor description performance and generalizes to diverse odor-prediction tasks.

neuroscience↗