Functional modulation of the human gut microbiome by bacteria vehicled by cheese

Abstract

Since cheese is one of the most commonly and globally consumed fermented foods, scientific investigations in recent decades have focused on determining the impact of this dairy product on human health and well-being. However, the modulatory effect exerted by the autochthonous cheese microbial community on the taxonomic composition and associated functional potential of the gut microbiota of human is still far from being fully dissected or understood. Here, through the use of an in vitro human gut-simulating cultivation model in combination with multi-omics approaches, we have shown that minor rather than dominant bacterial players of the cheese microbiota are responsible for gut microbiota modulation of cheese consumers. These include taxa from the genera Enterococcus, Bacillus, Clostridium, and Hafnia. Indeed, they contribute to expand the functional potential of the intestinal microbial ecosystem by introducing genes responsible for the production of metabolites with relevant biological activity, including genes involved in the synthesis of vitamins, short-chain fatty acids, and amino acids. Furthermore, tracing of cheese microbiota-associated bacterial strains in fecal samples from cheese consumers provided evidence of horizontal transmission events, enabling the detection of particular bacterial strains transferred from cheese to humans. Moreover, transcriptomic and metabolomic analyses of a horizontally transmitted (cheese-to-consumer) bacterial strain, i.e., Hafnia paralvei T10, cultivated in a human gut environment-simulating medium, confirmed the concept that cheese-derived bacteria may expand the functional arsenal of the consumer's gut microbiota. This highlights the functional and biologically relevant contributions of food microbes acquired through cheese consumption on the human health.IMPORTANCEDiet is universally recognized as the primary factor influencing and modulating the human intestinal microbiota both taxonomically and functionally. In this context, cheese, being a fermented food with its own microbiota, serves not only as a source of nourishment for humans, but also as a source of nutrients for the consumer's gut microbiota. Additionally, it may act as a vehicle for autochthonous food-associated microorganisms which undergo transfer from cheese to the consumer, potentially influencing host gut health. The current study highlights not only that cheese microbiota-associated bacteria can be traced in the human gut microbiota, but also that they may expand the functional repertoire of the human gut microbiota, with potentially significant implications for human health.

Keywords: food; human diet; metabolomics; metagenomics; metatranscriptomics; microbiota.

Fig 1

Microbial analysis workflow and taxonomic profile of cheese after cultivation in GESM. Panel a depicts the workflow followed to investigate the taxonomic, transcriptomic, and metabolomic outcomes obtained after a batch cultivation of raw milk cheese in a simulated gut environment. Panel b shows the hierarchical clustering approach based on Bray-Curtis matrix calculated on taxonomical profiles of cheese samples after GESM cultivation. In the center, the heat map displays the Bray-Curtis distance value matrix, highlighting the substantial taxonomic variability among cheese samples after GESM cultivation. Bar plots flanking the heat map illustrate the dominant bacterial species of raw cheese samples (left) and those cultivated on GESM medium (right). Only bacterial species with a maximum relative abundance of at least 10% across metagenomic samples were included.

Fig 2

Differences in enzyme-coding gene expression levels among EXCs. The figure depicts an extract of all enzymatic functions with biological activity for which we identified statistically significant variances in the expression of their associated genes between the three EXCs. Kruskal-Wallis P-value column explains the significance obtained through Kruskal-Wallis statistical analysis, with the green-highlighted cells containing statistically significant values (P-value <0.05). The fitting FDR with Benjamini-Hochberg correction P-value column shows the statistical significance of the fitting analysis, which identified enzymatic functions (classified by their EC number) that best explain gene expression biodiversity in the studied data. Instead, the variability explained by the EC numbers found to be significant from the fitting analysis is shown in column R², with values that range from lowest (gray) to highest (blue). The EXC columns depict the average gene expression values relative to the selected EC numbers, with the maximum values in red and the lowest values in gray.

Fig 3

Metabolomic analysis of GESM cultivations. Panels a and b show the violin plot of log₂ distribution profiles of metabolites with significant signal following cheese microbiota growth in GESM (log₂ >1 or log₂ <−1) with respect to the medium without inoculum. Panel c depicts the bar plot of metabolite signals increased (or appeared de novo) and decreased in all GESM cultivation samples compared to those modified in a single GESM cultivation sample.

Fig 4

Multi-omics evaluation of cheese-to-consumer horizontal transmission of bacterial strains. Panel a illustrates the workflow followed to validate transmission of metabolic functions from the microbiota present in cheeses to the gut microbiome of consumers through metagenomics, culturomics, genomics, and transcriptomics approaches. Panel b shows a series of heat maps related to metagenomic strain tracking performed using BBMap and BWA methods, with green cells indicating positive hits with more than 10,000 covered bases. Panel c provides the completeness level scores of the isolated strains and their total genome length based on CheckM analysis.