Delayed gut microbiota development in high-risk for asthma infants is temporarily modifiable by Lactobacillus supplementation

Abstract

Gut microbiota dysbiosis and metabolic dysfunction in infancy precedes childhood atopy and asthma development. Here we examined gut microbiota maturation over the first year of life in infants at high risk for asthma (HR), and whether it is modifiable by early-life Lactobacillus supplementation. We performed a longitudinal comparison of stool samples collected from HR infants randomized to daily oral Lactobacillus rhamnosus GG (HRLGG) or placebo (HRP) for 6 months, and healthy (HC) infants. Meconium microbiota of HRP participants is distinct, follows a delayed developmental trajectory, and is primarily glycolytic and depleted of a range of anti-inflammatory lipids at 6 months of age. These deficits are partly rescued in HRLGG infants, but this effect was lost at 12 months of age, 6 months after cessation of supplementation. Thus we show that early-life gut microbial development is distinct, but plastic, in HR infants. Our findings offer a novel strategy for early-life preventative interventions.

Fig. 1

Gut microbiota maturation of high risk for asthma (HR) subjects is significantly distinct from that of healthy controls (HC), and influenced by oral Lactobacillus rhamnosus (LGG) supplementation. a Compared to HC (n = 29) subjects, HR placebo supplemented (HRP; n = 15) participants exhibit a significant delay in bacterial alpha diversification over the first year of life (β ANCOVA, p = 0.02), which is partly rescued by LGG supplementation (HRLGG; n = 10; β ANCOVA, p = 0.63). Error bars indicate standard-deviation from the mean. b HRP (n = 15) and c HRLGG (n = 10) infants exhibit significant differences in bacterial beta diversity across all of the time points assessed (Unweighted UniFrac distance; R² and p-values calculated using PEMANOVA). d HRLGG (n = 10) infant gut microbiota is significantly (though marginally) more similar to HC (n = 29) subjects during the period of LGG supplementation (Unweighted UniFrac distance; Bonferroni corrected t-test). Whiskers extend to 95% confidence interval

Fig. 2

Meconium microbiota and persistent founder taxa significantly differ in high risk for asthma (HR) and healthy control (HC) subjects. a Principal Coordinates analysis (PCoA) of meconium microbiota indicates significantly different composition based on asthma risk status (HR n = 18; HC n = 17; Unweighted UniFrac; PERMANOVA R² = 0.062, p = 0.001). b Taxon summary plot indicates expansion of Enterobacteriaceae and Bacteroidaceae in vaginally born HR (n = 17) meconium compared to HC (n = 14) neonates. c Persistent founder microbiota differ in HR (n = 14) and HC (n = 9) participants (Unweighted UniFrac; PERMANOVA R² = 0.081, p = 0.018). d While a number of persistent founder taxa are common to both HC (n = 9) and HR (n = 14) subjects, several differ (Fisher’s exact test; p < 0.05; black dots), including a number whose depletion has previously been associated with atopy and asthma development in childhood

Fig. 3

Three-way comparison of 6 month fecal samples identifies metabolites enriched in healthy controls (HC) and Lactobacillus rhamnosus-supplemented high risk for asthma (HRLGG) subjects compared to high risk for asthma placebo (HRP) group. HC (n = 15) and HRLGG (n = 7) participants share a number of common enriched metabolites compared with HRP (n = 11) group, including anti-inflammatory androgenic steroids, long-chain and polyunsaturated fatty acids. In comparison HRP subjects exhibit evidence of increased glycolysis and for distinct products of lipid metabolism (Welch’s two-sample t-test; p ≤ 0.05)

Fig. 4

Lactobacillus rhamnosus (LGG)-associated fecal products promote T_reg cell expansion and IL10 production ex vivo at 6 months of age. Sterile fecal water derived from 6-month-old LGG-supplemented high risk for asthma (HRLGG; n = 5) stool induced an increase in the proportion of a. T_reg cells and b. IL10, compared with high risk for asthma placebo (HRP; n = 5) or healthy (HC; n = 5) participants. At 12 months of age, 6 months following cessation of LGG supplementation, no difference in the proportion of c. T_reg cells or d. IL10 expression was observed between the three groups. Linear mixed effects (LME) adjusted for blood-donor was used to test for between group differences