A Taxonomically-informed Mass Spectrometry Search Tool for Microbial Metabolomics Data

Abstract

MicrobeMASST, a taxonomically-informed mass spectrometry (MS) search tool, tackles limited microbial metabolite annotation in untargeted metabolomics experiments. Leveraging a curated database of >60,000 microbial monocultures, users can search known and unknown MS/MS spectra and link them to their respective microbial producers via MS/MS fragmentation patterns. Identification of microbial-derived metabolites and relative producers, without a priori knowledge, will vastly enhance the understanding of microorganisms' role in ecology and human health.

Keywords: bacteria; fungi; metabolomics; microbiome; search tool.

Figure 1.. The microbeMASST search tool and reference database.

a) Community contributions of data and knowledge to GNPS, ReDU, and MassIVE from 2014 to 2022 were used to generate the microbeMASST reference database. Additionally, a public invitation to deposit data in June 2022 resulted in the further deposition of LC-MS/MS files from 25 different laboratories from 15 different countries across the globe, leading to the curation of a total of 60,781 LC-MS/MS files of microbial monoculture extracts. b) MicrobeMASST comprises 1,858 unique lineages, across three different domains of life, mapped to 541 unique strains, 1,336 species, 539 genera, 264 families, 109 orders, 41 classes, and 16 phyla. c) Examples of medically-relevant small molecules known to be produced by bacteria or fungi. Lovastatin, a cholesterol lowering drug originally isolated from Aspergillus genus, salinosporamide A, a Phase III candidate to treat glioblastoma produced by Salinispora tropica, and commendamide, a human G-protein–coupled receptor agonist. d) MicrobeMASST search outputs of the three different molecules of interest confirm that they were exclusively found in monocultures of the only known producers. Pie charts display the proportion of MS/MS matches found in the deposited reference database. Blue indicates a match with a monoculture, while yellow represents a nonmatch. Searches were performed using MS/MS spectra deposited in the GNPS reference library: lovastatin (CCMSLIB00005435737), salinosporamide A (CCMSLIB00010013003), and commendamide (CCMSLIB00004679239).

Figure 2.. MicrobeMASST can identify microbial MS/MS spectra within mouse and human datasets.

a) Workflow to extract microbial MS/MS spectra from biochemical profiles of 29 different tissues and biofluids of SPF mice that are not observed in GF mice. The MS/MS spectra unique to SPF mice (10,047) were searched with microbeMASST. A total of 3,262 MS/MS spectra had a match; those MS/MS also matching human cell lines were removed, leaving a total of 2,425 putative microbial MS/MS spectra (see Methods to download .mgf file). b) The presence of the 2,425 MS/MS spectra was evaluated in an additional animal study looking at antibiotics usage. A total of 512 MS/MS spectra, out of the 621 overlapping, was exclusively found in animals not receiving antibiotics. c) UpSet plot of the distribution of the detected MS/MS spectra (512) across bacterial phyla. Terpenoids were more commonly observed across phyla while amino acids and peptides appeared to be more phylum specific d) The 512 MS/MS spectra were searched in human datasets and 455 were found to have a match. These MS/MS spectra were present in both healthy individuals and individuals affected by different diseases. e) Most of the MS/MS spectra (n=411) matched fecal samples (n=110,973 matches), followed by blood, oral cavity, breast milk, urine, and several other organs. Abbreviations: CSF, cerebral spinal fluid; HIV, human immunodeficiency virus; PBI, primary bacterial infectious disease; SD, sleep disorder; AD, Alzheimer’s disease; IS, ischemic stroke; KD, Kawasaki disease; IBD, inflammatory bowel disease; T2D, type II diabetes.