Genome-resolved functional characterisation of the broiler caecal microbiome following commercial gut health interventions

Technical Summary: Genome-resolved functional characterisation of the broiler caecal microbiome following commercial gut health interventions

Problem Statement
The broiler caecal microbiome is a metabolically active community critical for nutrient utilization and host homeostasis. While Gut Health Interventions (GHIs)—including probiotics, prebiotics, essential oils, coccidiosis vaccines, and reduced crude protein diets—are widely used to modulate this ecosystem, their impact is often assessed through taxonomic shifts. However, taxonomic changes can be subtle or inconsistent in mature microbial ecosystems. Furthermore, existing gene-centric metagenomic approaches fail to resolve how functional traits, such as antimicrobial resistance (ARG), virulence (VAG), and stress adaptation (SAG), are distributed among specific microbial populations. There is a lack of genome-resolved data characterizing how commercial GHI combinations shape the functional gene repertoire of the broiler gut, particularly regarding biogeochemical cycling and the ecological adaptation of specific taxa to these interventions.

Methodology
This study employed a genome-resolved metagenomic approach using 307 high-quality Metagenome-Assembled Genomes (MAGs) recovered from 118 caecal content samples across two production cycles (C1 and C2). The samples originated from a commercial trial involving six treatment groups (T1–T6) receiving various GHI combinations, including ionophores, attenuated live coccidiosis vaccines, Bacillus-based probiotics (A and B), essential oils, prebiotics, and reduced crude protein diets.

• Genomic Reconstruction: MAGs were assembled from Illumina shotgun metagenomic data, filtered for ≥50% completeness and ≤10% contamination.
• Functional Annotation: Genes were annotated using the METABOLIC pipeline, AMRFinderPlus, and ABRicate against multiple databases (NDARO, CARD, MEGARes, VFDB, etc.) to identify ARGs, SAGs, and VAGs. Thresholds of ≥70% nucleotide identity and ≥50% coverage were applied.
• Phylogenetic Analysis: A phylogenetic tree was constructed using 175 MAGs containing sufficient single-copy genes (SCGs) to assess novelty and taxonomic distribution.
• Statistical Modeling: Diversity metrics (alpha and beta) were calculated using the vegan package. Associations between GHI components and functional gene abundance were modeled using Generalised Linear Latent Variable Models (GLLVM). Differential abundance analysis was performed using DESeq2, with the ionophore-only group (T1) serving as the reference.
• Metabolic Reconstruction: Pathways for carbon, nitrogen, and sulphur cycling were identified to assess potential greenhouse gas production (methane, nitrous oxide) and other metabolite synthesis (ammonium, hydrogen sulphide).

Key Results

• Functional Gene Landscape: A total of 208 putative functional genes were annotated. Virulence-associated genes (VAGs) were the most prevalent (51.92%), followed by antimicrobial resistance genes (ARGs, 43.75%) and stress-associated genes (SAGs, 4.33%).
  • Taxonomic Distribution: Firmicutes A carried the most diverse ARG classes. Glycopeptide resistance was the most widely distributed ARG, found in 65.08% of ARG-carrying MAGs. Notably, no ionophore resistance determinants were detected in any MAG.
  • Gene Concentration: Functional genes were not randomly distributed; specific MAGs, such as Escherichia flexneri (bin.48), harbored high numbers of ARGs, SAGs, and VAGs, suggesting that microbial risk is linked to the genetic repertoire of key community members rather than abundance alone.
• Impact of GHI Combinations:
  • Diversity Reduction: In Cycle 1, the combination of Bacillus-based probiotics A, essential oils, and coccidiosis vaccination (T5) significantly reduced the alpha diversity of the resistome and virulome compared to other treatments.
  • Gene Abundance: T5 was associated with significantly lower abundance of specific macrolide (mef(En2)) and lincosamide (lnuA, lnu(AN2)) resistance genes, as well as arsenic resistance (arsB). These genes were restricted to Phocaeicola vulgatus (bin.96).
  • Individual Interventions: Probiotics A was uniquely associated with reduced SAG and VAG abundance without significantly altering ARG abundance, indicating a functionally neutral-to-beneficial profile. In contrast, coccidiosis vaccination showed broad positive associations with ARG abundance.
• Biogeochemical Cycling:
  • Greenhouse Gases: No MAGs possessed the capacity for methanogenesis (methane production) or nitrous oxide production/consumption. However, 19 MAGs (primarily Firmicutes A) encoded pathways for methane consumption (methanotrophy).
  • Nutrient Cycling: MAGs were predicted to participate in ammonium production via nitrogen fixation and nitrite ammonification, and hydrogen sulphide synthesis via sulphite reduction. No MAGs were predicted to consume ammonium or hydrogen sulphide.
  • Intervention Effects: Probiotics A, probiotics B, finisher-stage ionophores, and prebiotics did not alter core biogeochemical functions (methanotrophy, ammonium, or hydrogen sulphide synthesis). Coccidiosis vaccination was associated with changes in all 14 eligible biogeochemical reaction steps.

Key Contributions

1. Genome-Resolved Baseline: This study provides the first systematic, genome-resolved characterization of how commercial GHI combinations shape the functional gene landscape of the broiler caecal microbiome, moving beyond taxonomic shifts to specific gene distributions.
2. Differentiation of Intervention Effects: It demonstrates that specific GHI combinations, particularly those containing probiotics, can reduce the diversity and abundance of resistance and virulence genes, whereas other interventions (like vaccination alone) may increase functional gene abundance.
3. Ecological Adaptation vs. Selective Pressure: The widespread distribution of glycopeptide resistance and the absence of ionophore resistance suggest that the observed resistome reflects ecological adaptation to the gut environment rather than selective pressure from the specific antimicrobials used in the trial.
4. Biogeochemical Potential: It establishes that the broiler caecal microbiome has limited intrinsic capacity for greenhouse gas production but possesses potential for methane consumption and nutrient cycling (ammonium and hydrogen sulphide synthesis), which remains largely stable across most GHI strategies.

Significance and Claims
The authors claim that this research advances the evidence base for rational GHI selection in poultry production. By demonstrating that strategic GHI selection—specifically combinations involving probiotics—can modulate the functional gene landscape to reduce resistome and virulome diversity, the study supports One Health initiatives aimed at mitigating microbial risks. The findings suggest that probiotic-containing interventions offer a favorable approach for maintaining a stable gut ecosystem without disrupting core biogeochemical functions. The study emphasizes that these findings represent genomic potential and serve as an exploratory step toward understanding how commercial interventions shape the functional microbiome at the population level, rather than confirming phenotypic resistance or transmission dynamics.