Premium
Discovery of Antibiotic Peptides from Novelty‐Prioritized Natural Product Genome Mining
Author(s) -
Schwalen Christopher J,
Mitchell Douglas
Publication year - 2017
Publication title -
the faseb journal
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.709
H-Index - 277
eISSN - 1530-6860
pISSN - 0892-6638
DOI - 10.1096/fasebj.31.1_supplement.939.8
Subject(s) - novelty , natural product , product (mathematics) , antibiotics , natural (archaeology) , computational biology , genome , biology , genetics , gene , mathematics , biochemistry , psychology , social psychology , paleontology , geometry
Natural products remain the greatest asset to combat infectious disease, with the majority of approved antibiotics being either of natural origin or directly derived thereof. Despite decades of successful medical use, two challenges to this paradigm have arisen: the rise of antibiotic‐resistant strains of clinical pathogens is decreasing antibiotic efficacy, and the approval of novel antimicrobial compounds is declining. A potential solution is to increase the efficiency of traditional natural product chemistry methods to streamline the discovery process of new antibiotics. Modern sequencing techniques have expanded databases of genomic information, giving birth to genome‐mining as a methodology to locate and predict biosynthetic gene clusters of new natural products. However, the ever‐increasing rate at which new sequences are deposited and the size of entries makes managing and searching through such databases difficult. To address this, we have developed the genome‐mining platform Rapid ORF Description and Evaluation Online (RODEO) to efficiently analyze multiple biosynthetic gene clusters in a high‐throughput and modular fashion with an emphasis on natural product discovery. RODEO prediction software was trained using lasso peptides, members of the ribosomally synthesized post‐translationally modified peptides (RiPP) category of structurally diverse natural products with intriguing bioactivities. Metagenomic analysis of biosynthetic genes as well as machine‐learning precursor analysis generated the most comprehensive mapping of lasso peptide genetic space to date, revealing >1300 lasso peptides and >1400 clusters, (an order of magnitude increase over previous estimates). RODEO's data set was used to connect gene clusters to lasso peptides with no previously‐known biosynthetic information, to identify common structural motifs in lasso peptide biosynthesis and to trace phylogenetic trends in this class of natural products. Moreover, RODEO's predictions guided the identification and isolation of six novel lasso peptides with diverse antibacterial profiles and unprecedented structural characteristics. The modular design of RODEO is enabling characterization of other families of compounds, and the program is being integrated into discovery pipelines to better connect the chemistry and biosynthesis of natural products with available genomic data, further enabling genomics‐based natural product discovery. Support or Funding Information C.J.S. is a member of the NIH Chemistry‐Biology Interface Training Program (Grant NRSA 1‐T32‐GM070421)RODEO's genome‐mining pipeline was used to annotate the lasso peptide family of ribosomal natural products. Rapid analysis of biosynthetic gene clusters and machine‐learning precursor identification guided the isolation of new lasso peptides with antibacterial activity and several unprecedented structural features, as well as global trends in this natural product family.