Structure-based discovery and definition of RiPP recognition elements

Area of Science:

• Natural product discovery
• Bioinformatics
• Structural biology
• Genomics

Background:

• Ribosomally synthesized and post-translationally modified peptides (RiPPs) are diverse natural products crucial in various biological processes.
• RiPP Recognition Elements (RREs) are key peptide-binding domains essential for RiPP biosynthesis and genome mining.
• Existing RRE-Finder tools face limitations due to high false-positive rates and difficulty in identifying sequence-divergent RREs.

Purpose of the Study:

• To improve the accuracy and scope of RiPP genome mining by enhancing RRE identification.
• To leverage structure-based searching (Foldseek) and AlphaFold predictions to discover sequence-divergent RREs and their associated precursor peptides.
• To refine bioinformatic tools for more comprehensive identification of RRE-dependent biosynthetic pathways.

Main Methods:

• Employed Foldseek for structure-based searching of the AlphaFold database to identify divergent RREs.
• Developed 11 new Foldseek-derived Hidden Markov Models (HMMs) and refined existing models for RRE-Finder.
• Utilized AlphaFold 3 to predict RRE-peptide complexes, enabling the mapping of recognition sequences.

Main Results:

• The updated workflow identified over 90,000 high-confidence RREs, nearly doubling the retrieval rate from UniProt compared to original models.
• Discovered novel RRE domain fusions and 5,000 previously unidentified RRE domains, retaining canonical folds but offering new bioinformatic handles.
• Mapped 13 distinct recognition sequence motifs across RiPP classes by predicting RRE-precursor peptide interactions.

Conclusions:

• The integration of structure-based searching and advanced modeling significantly enhances the accuracy and efficiency of RiPP genome mining.
• This improved approach expands the known landscape of RRE-dependent biosynthetic pathways, providing access to previously hidden natural products.
• The findings streamline the identification of RiPP precursor peptides and their cognate RREs, facilitating further research into RiPP diversity and function.