Project Details
Projekt
AI-based identification and functional characterization of human gut microbiome-derived natural products
Subject Area
Metabolism, Biochemistry and Genetics of Microorganisms
Term
since 2025
Project identifier
Deutsche Forschungsgemeinschaft (DFG) - Project number 564428762
The human gut microbiome plays a critical role in health and disease. Many conditions, both within the gastrointestinal tract and in distant organs, are associated with the presence, absence, or altered abundance of specific gut bacteria. NPs produced by gut bacteria are thought to influence both microbiome composition and host physiology. However, the natural products (NPs) mediating these effects remain largely unknown. Genome mining—an in silico NP discovery strategy— offers a powerful approach to identify these bioactive molecules. It involves analyzing bacterial genomes to predict their biosynthetic potential. Currently existing tools excel at identifying biosynthetic gene clusters (BGCs) for assembly-line biosynthetic pathways, which rely on large, scaffold-forming enzyme complexes. However, such pathways are extremely energy-demanding and may not be optimal for the anaerobic lifestyle of the human gut microbiome. Unfortunately, small BGCs involved in alkaloid, ribosomally synthesized and post-translationally modified peptide (RiPP), and terpene biosynthesis often escape unrecognized by existing genome mining tools, as they do not encode scaffold-forming enzymes conserved across the respective NP class. To address this problem, we aim to chart the "biosynthetic dark matter" of the gut microbiome by focusing on these under-explored classes of BGCs. Specifically, we will adapt an artificial intelligence (AI)-based algorithm, that we recently developed (unpublished), to mine gut bacterial genomes for their full biosynthetic potential, with an emphasis on synthetic consortia studied within SPP2474. Our TailEnza algorithm detects NP BGCs through the identification of genes encoding NP scaffold-modifying enzymes and predicting their substrate specificity. This strategy complements current genome mining approaches and enables the discovery of BGCs overlooked by other tools. We will prioritize alkaloid, RiPP, and terpene BGCs from key members of the human gut microbiome that are relevant to the SPP2474 consortium. The BGCs will be functionally studied in the native producer or through heterologous expression, followed by the structural characterization of the associated NPs. A wide array of bioactivity assays will be employed to assess NP functions, including bacterial interaction assays, assays examining synthetic community dynamics, and in vitro eukaryotic cell-based assays. The latter assays will evaluate the impact of the identified NPs on intestinal function, immune responses, host metabolism, and the nervous system. In summary, our study leverages an advanced AI algorithm to discover gut microbiome-associated NPs produced by key microbiome members. By integrating comprehensive bioactivity assays, we aim to elucidate the roles of these NPs in shaping community dynamics and influencing the human host. The insights could inform the development of therapeutic interventions or probiotic strategies.
DFG Programme
Priority Programmes
