Age-associated gut microbiome succession, colonization resistance, and relative resistome patterns in an antibiotic-restricted infant cohort.

BACKGROUND

Early infancy is critical for gut microbiome assembly and the establishment of colonization resistance against pathobionts. Whether age-associated microbiome maturation is accompanied by changes in colonization-resistance proxies and relative antimicrobial resistance gene profiles under low infant antibiotic exposure remains unclear.

METHODS

We analyzed shotgun metagenomes from 82 fecal samples collected from 54 healthy infants (54 at 1 month and 28 at 6 months). Taxonomic and functional profiles were generated using MetaPhlAn 4 and HUMAnN3, and AMR genes were annotated using RGI/CARD. Age-associated taxa were screened by LEfSe and tested using MaAsLin2 with adjustment for key perinatal covariates.

RESULTS

Age group was associated with modest but statistically significant differences in community structure (Bray-Curtis PERMANOVA R 2 = 0.03, p = 0.005) and higher species richness at 6 months (p < 0.001), with no statistically significant difference in Shannon or Simpson indices. In adjusted models, skin-associated pioneer taxa, including Staphylococcus epidermidis, were lower at 6 months, whereas several anaerobic or oral-associated taxa were higher, including Flavonifractor plautii. Enterobacteriaceae relative abundance was lower at 6 months than at 1 month (median 16.64 vs. 1.86%, p < 0.001), and Bifidobacterium-Enterobacteriaceae antagonism indices were higher. However, Escherichia coli and Klebsiella spp. did not show significant genus-level reductions. Copies per million (CPM)-normalized β-lactamase (bla) relative abundance showed no statistically significant timepoint difference and was positively correlated with selected Bifidobacterium species.

CONCLUSIONS

In this infant antibiotic-restricted cohort, microbiome profiles at 6 months were associated with lower relative abundance of potential pathobionts and higher colonization-resistance proxy indices. CPM-normalized bla relative abundance showed no statistically significant timepoint difference. These observational findings do not establish the genomic host or mobility of bla genes. Quantitative and host-resolved studies are needed to distinguish compositional shifts from absolute resistome trajectories.