The pioneering gut microbiome acquired via different delivery modes in neonates shapes distinct immune and metabolic environments

Abstract

Background: Cesarean section (CS) delivery is associated with an increased risk of inflammatory diseases, hypothesized to be driven by differences in the microbiome acquired at birth compared to vaginally delivered (VD) infants. How delivery mode associated differences in initial colonizers directly modulate early life immune education and metabolic development is poorly understood.

Objective: First, to investigate how differences in pioneering colonizers associated with delivery mode directly modulate early life immune education and metabolic programming. Second, to examine the effect of "vaginal seeding", an intervention aimed to restore the microbiome of CS infants to a VD state.

Design: Germ-free mice were colonized with transitional stool from VD, CS or CS-delivered and vaginally seeded neonates. Immune cell populations, serum immunoglobulin levels, and fecal microbiome and metabolome profiles were analyzed.

Results: Mice colonized with stool from VD neonates displayed increased numbers of myeloid cells at barrier tissues, whereas CS microbiome colonized mice exhibited decreased Th1/Th2 ratios and increased serum IgE levels. Key amino-acid pathways including tryptophan metabolism, riboflavin co-enzymes and carbohydrate metabolites were significantly enriched in the murine VD fecal metabolome and correlate with the increased abundance of Escherichia typically observed in the VD microbiome. Mice colonized with stool from CS neonates who received vaginal seeding, resulted in increased regulatory T cells and serum IgA in mice, suggesting potential benefits of vaginal seeding.

Conclusion: Collectively, our studies demonstrate the ability of pioneering colonizers to set the immune and metabolic tone that could have long-lasting effects and provide avenues for microbiome-mediated interventions.

Figure 1.. Composition of gut microbiomes in mice colonized with stool from neonates born via different delivery modes

A. Schematic of experimental design. Stool samples from vaginally delivered (VD), cesarean section (CS) delivered or CS-delivered and vaginally seeded (CS+VS) neonates were orally inoculated into germ-free mice at weaning. An additional group of germ-free mice received PBS as vehicle control. Three weeks post-inoculation fecal, serum and tissues samples were collected for downstream analysis. Microbiome profiling was performed on fecal samples collected from small intestinal and colon flushes, as well as fecal pellets, using 16S rRNA gene sequencing. B. Alpha diversity estimates of the microbiota from the small intestine, colon and fecal pellets were calculated using Observed ASVs and Shannon index. Each dot denotes a mouse with the shape and color representing a human donor subject and experimental group used to inoculate the mouse respectively, as shown in the key. Statistical significance was calculated using Kruskal-Wallis test (ns, p > 0.05; * p < 0.05; ** p < 0.01; ***p < 0.001; ****p < 0.0001) C. Principal coordinate analysis (PCoA) plots of Bray Curtis dissimilarity metrics colored by inoculation group. Statistical significance was assessed using Adonis with 999 permutations. D. Beta-diversity distances within and between groups using VD group as reference are plotted and statistical significance was calculated using PERMANOVA (***p <0.001) E. Differentially abundant taxa were identified using ANCOMBC for each pairwise comparison, as indicated along the x-asis and displayed as a dotplot. Genera that were signficantly different (q < 0.05) are grouped by phyla along the y-axis. ASVs not classified with a specific genus are indicated as a genus under the highest taxonmic classification reported or as unclassified. Colors represent differential abundance relative to the reference group (Ref) and size of the circles indicate signficance values following multiple-testing as shown in the key.

Figure 2.. Influence of different neonatal microbiomes acquired at birth in shaping the myeloid compartment at various barrier tissues in colonized germ-free mice

A-C. Quanitification of major myeloid cell populations in the siLP (A), lung (B) and skin (C). Each dot denotes a mouse with the shape and color representing a human donor subject and experimental group used to inoculate the mouse respectively, as shown in the key in C. Control group represents germ-free mice inoculated with PBS vehicle control. The gating strategy is as shown in Figure S2A. Statistical significance was assessed by one-way ANOVA with Tukey’s multiple comparison test and only statistically signficant comparisons are shown (not shown, p > 0.05; * p < 0.05; ** p < 0.01; ***p < 0.001; ****p < 0.0001)

Figure 3.. Comparison of the lymphoid compartment at barrier tissues in neonatal stool colonized mice

A-C. Cell numbers of major lymphocyte cell types in the siLP (A), lung (B) and skin (C). D. Frequency of CD4⁺ T cell subsets in siLP and lung. E. Ratio of Th1/Th2 cells. F. Ratio of Th1/Th17 cells. G. Frequency of Treg subsets. H. Ratio of Tbet⁺/RORγt⁺ γδ T cells. I. Concentration of IgA and IgE in serum. The gating strategy is as shown in Figure S2B. Each dot represents a mouse and is coded as shown in the key. Statistical signifcance was assessed by one-way ANOVA with Tukey’s multiple comparison test and only statistically signficant comparisons are shown (not shown, p > 0.05; * p < 0.05; ** p < 0.01; ***p < 0.001; ****p < 0.0001)

Figure 4.. Comparison of fecal metabolites of mice colonized with neonatal stool

A. PCA plot of fecal metabolite signals identified from mice colonized with VD, CS, or CS+VS neonate stool B. Volcano plots show metabolites that were differentially abundant in pairwise comparisons of groups as indicated. Statistical significance was assessed using Student’s t-test and metabolites with a log₂ fold change >1 and adjusted p < 0.05 are labeled and colored (with a log₂ fold change > 0.5 only colored) based on pathways or molecular families as shown in the key. C. Abundances of metabolites in the tryptophan pathway were compared between VD and CS mice using Student’s t-test. For each metabolite, log₂ fold change and p-value were calculated and mapped as shown. Red and blue circles indicate increases and decreases in VD mice in comparison to CS mice respectively. The size of each circle reflects statistical significance. D-E. z-scale values of histamine and acetylcholine (D) and arginine cycle metabolites (E). Each dot represents a mouse and is coded as shown in the key. Statistical signifcance was assessed by one-way ANOVA with Tukey’s multiple comparison test and only statistically signficant comparisons are shown (not shown, p > 0.05; * p < 0.05; ** p < 0.01; ***p < 0.001; ****p < 0.0001).

Figure 5.. Correlation analysis between microbial taxa, metabolites and immune cell types identified in colonized germ-free mice

A-B Spearman’s correlation coefficient between measurements of significantly altered fecal microbial taxa and immune cell types (A) and fecal microbial taxa and fecal metabolites (B) were calculated. Correlations were filtered for correlation coefficient r > |0.5| and adjusted p < 0.05 (A) or r > |0.6| and adjusted p < 0.01 (B) and displayed as a heatmap. Correlations shown are based on measurements from all subjects across the three groups of colonized mice. Taxa identified at order or family level indicate unclassified genera under those taxa, as represented in Figure 1E.