Large-scale analysis of 12,238 gut metagenomes identifies microbial co-colonizers and co-excluders, offering insights into infection resistance and potential non-antibiotic therapies.
Study: Ecological dynamics of Enterobacteriaceae in the human gut microbiome across global populations. Image Credit: SewCreamStudio/Shutterstock.com
In a recent study published in Nature Microbiology, a group of researchers revealed global patterns and microbial interactions shaping Enterobacteriaceae colonization in the gut, informing infection resistance and therapeutic strategies.
Background
The human gut microbiota is a diverse community crucial for health, supporting digestion, immune regulation, and protection against pathogens. However, some gut microbes, particularly Enterobacteriaceae species like Escherichia coli (E. coli) and Klebsiella pneumoniae (K. pneumoniae), can cause severe infections and are linked to conditions like Crohn’s disease and higher mortality.
Their overabundance, coupled with rising multidrug resistance, poses a significant global health challenge. While microbiome-based therapies show promise, most research on Enterobacteriaceae has focused on clinical isolates, limiting ecological insights.
Further studies are essential to understand their dynamics within the gut microbiome, paving the way for innovative, non-antibiotic interventions.
About the study
Human gut metagenomic data was compiled from 12,238 samples across 45 countries and 65 studies, all sourced from the European Nucleotide Archive (ENA). Samples were included based on criteria such as a minimum of 500,000 paired-end metagenomic reads, availability of metadata on health state, age group, and country of origin, absence of acute infections, and no antibiotic use within a month prior to sample collection.
Metagenomic reads were processed for quality filtering using TrimGalore and for the removal of human contamination by aligning against the human genome GRCh38 using Burrows-Wheeler Aligner MEM (BWA MEM).
Filtered datasets were mapped against a curated database of 4,612 species from the Unified Human Gastrointestinal Genome (UHGG) catalog. Rigorous curation excluded low-quality genomes using criteria such as low completeness scores assessed by CheckM and potential chimeric classifications identified by Genome UNClutterer (GUNC).
Batch effects arising from variations in study methodologies were corrected using a conditional quantile normalization algorithm, enabling reliable comparisons across datasets.
Microbial community differences were analyzed through alpha diversity (Shannon index) and beta diversity (Aitchison distance) metrics. Supervised machine learning models, including ridge regression, random forest, and gradient boosting, were employed to classify Enterobacteriaceae colonization status using species abundance as features.
These models demonstrated high accuracy in distinguishing colonization patterns, offering critical insights into microbial interactions.
Differential abundance analyses, conducted with Analysis of Differential Abundance Using Compositional Data (ALDEx2) and Multivariable Association with Linear Models 2 (MaAsLin2), identified microbial species significantly associated with Enterobacteriaceae presence or absence.
Functional analyses linked co-colonizers and co-excluders to metabolic pathways involved in nutrient competition and ecological roles.
Study results
The datasets encompassed diverse geographic regions, with the majority of samples originating from Europe (35%) and North America (27.5%), followed by Asia (23.2%) and Africa (8.4%).
Most samples were collected from adults (67.6%) and healthy individuals (62.2%), providing a great foundation for analyzing microbial community dynamics. Using the UHGG catalog, 4,612 gut microbial species, including 113 from the Enterobacteriaceae family, were identified and quantified with stringent quality controls.
Enterobacteriaceae were detected in 66% of the samples, with E. coli, K. pneumoniae, and Enterobacter hormaechei being the most prevalent species. Their distribution varied across geographic locations, age groups, and health states, with notable prevalence among African samples (88%) and individuals with rheumatoid arthritis (96%).
Patterns of polymicrobial colonization revealed significant co-occurrence of E. coli with other species, such as K. pneumoniae, particularly in Asia and Africa. These findings suggest that environmental, dietary, and healthcare factors may influence Enterobacteriaceae colonization globally.
A deeper analysis of E. coli strain diversity using multilocus sequence typing (MLST) identified 585 sequence types (STs), with a majority being novel and overrepresented in African samples. This highlights the underexplored diversity of E. coli and the limited representation of non-clinical isolates in reference databases.
Machine learning models demonstrated that gut microbiome composition could predict Enterobacteriaceae colonization with high accuracy, revealing consistent associations across geographic regions and health states.
Differential abundance analysis identified 307 microbial species significantly associated with Enterobacteriaceae colonization, classified as co-colonizers or co-excluders. Co-excluders, such as Faecalibacterium species, were linked to short-chain fatty acid (SCFA) production and iron metabolism, mechanisms that may inhibit Enterobacteriaceae growth.
In contrast, co-colonizers exhibited greater metabolic diversity, suggesting a competitive advantage in colonization under certain gut conditions.
Functional analyses revealed that co-excluders were enriched in genes involved in quorum sensing and SCFA production, while co-colonizers were associated with drug resistance and nutrient metabolism. These differences underscore the role of interspecies interactions and metabolic competition in shaping gut colonization dynamics.
Conclusions
To summarize, this large-scale, global study explored gut microbiome signatures linked to Enterobacteriaceae colonization, revealing taxonomic and functional shifts related to co-colonization and co-exclusion.
Key findings include the identification of a significant uncharacterized subspecies diversity of E. coli in Africa, highlighting the role of species like Faecalibacterium in colonization resistance.
Co-colonization patterns involving taxa like Intestinibacter and Faecalimonas phoceensis were also uncovered. Additionally, co-excluders harbor uncharacterized biosynthetic gene clusters (BGCs) linked to quorum sensing, potentially modulating Enterobacteriaceae abundance.
