Metaproteomics to understand how microbiota function: The crystal ball predicts a promising future

Abstract

In the medical, environmental, and biotechnological fields, microbial communities have attracted much attention due to their roles and numerous possible applications. The study of these communities is challenging due to their diversity and complexity. Innovative methods are needed to identify the taxonomic components of individual microbiota, their changes over time, and to determine how microoorganisms interact and function. Metaproteomics is based on the identification and quantification of proteins, and can potentially provide this full picture. Due to the wide molecular panorama and functional insights it provides, metaproteomics is gaining momentum in microbiome and holobiont research. Its full potential should be unleashed in the coming years with progress in speed and cost of analyses. In this exploratory crystal ball exercise, I discuss the technical and conceptual advances in metaproteomics that I expect to drive innovative research over the next few years in microbiology. I also debate the concepts of 'microbial dark matter' and 'Metaproteomics-Assembled Proteomes (MAPs)' and present some long-term prospects for metaproteomics in clinical diagnostics and personalized medicine, environmental monitoring, agriculture, and biotechnology.

FIGURE 1

Estimation of the molecular complexity of a human faecal sample. The numbers indicated are derived from published results (Grenga et al., 2022), obtained from 39 human faecal samples analysed in triplicate. The different groups of identified organisms are indicated with the number of entities (×1 for the host) and a list of representative species per group is mentioned. The number of annotated protein‐coding sequences per group (20,383 for the host) is mentioned and their ratios are represented with the external circle. The inner circle shows the protein biomass of each group assessed experimentally by metaproteomics.

FIGURE 2

Tentative estimation of the protein and peptide spaces of a human faecal sample. The protein space is directly derived from the number of organisms estimated in Figure 1 and focussed on the most abundant organisms. Peptides are considered with equate I/L as these two residues are indistinguishable by simple mass spectrometry, with an average of 24 peptides per protein without missed‐cleavage and 70 peptides per protein when considering 1 possible missed‐cleavage. An average of 5 variants per peptide sequence is taken into account for synthesized proteins from the 200 most abundant species to obtain the number of peptides & variants.

FIGURE 3

Metaproteomics can identify and quantify uncharted branches of the tree of life. Tandem mass spectrometry‐based proteotyping of organisms present in the sample can be based on taxa‐spectrum matches (TSMs) and taxon‐specific peptides (Hardouin et al., ; Lozano et al., 2022). In this figure focused on a theoretical archaeal enriched microbial community, the presence of an uncharacterized organism belonging to the Candidatus Lokyarchaeota phylum and another belonging to the Halobacteriales order are indicated, while other organisms characterized at the genus taxonomical rank are also confirmed. These organisms are identified from taxon‐specific peptides and TSMs at the different taxonomical ranks when querying a generalist database such as NCBInr. The ratios of organisms are established based on protein biomass values derived from the TSMs parameter and are reported in percentage compared to the signal interpreted for the phylum taxonomical rank. The values presented are imaginary and serve only to explain the concept. A decrease of TSMs along the taxonomical ranks may be observed due to the phylogenetic distance between the proteins from the organisms present in the sample and those from the organisms listed in the database.

FIGURE 4

Clinical and environmental diagnostics by metaproteomics, timelines and applications. A sample‐to‐result timeline is proposed based on previously published results (Hardouin et al., 2022), along with likely optimizations over the next decade. The fields of application of metaproteomics for diagnosis or routine analysis are schematized.