Perinatal environment shapes microbiota colonization and infant growth: impact on host response and intestinal function

Abstract

Background: Early microbial colonization triggers processes that result in intestinal maturation and immune priming. Perinatal factors, especially those associated with birth, including both mode and place of delivery are critical to shaping the infant gut microbiota with potential health consequences.

Methods: Gut microbiota profile of 180 healthy infants (n = 23 born at home and n = 157 born in hospital, 41.7% via cesarean section [CS]) was analyzed by 16S rRNA gene sequencing at birth, 7 days, and 1 month of life. Breastfeeding habits and infant clinical data, including length, weight, and antibiotic exposure, were collected up to 18 months of life. Long-term personalized in vitro models of the intestinal epithelium and innate immune system were used to assess the link between gut microbiota composition, intestinal function, and immune response.

Results: Microbiota profiles were shaped by the place and mode of delivery, and they had a distinct biological impact on the immune response and intestinal function in epithelial/immune cell models. Bacteroidetes and Bifidobacterium genus were decreased in C-section infants, who showed higher z-scores BMI and W/L during the first 18 months of life. Intestinal simulated epithelium had a stronger epithelial barrier function and intestinal maturation, alongside a higher immunological response (TLR4 route activation and pro-inflammatory cytokine release), when exposed to home-birth fecal supernatants, compared with CS. Distinct host response could be associated with different microbiota profiles.

Conclusions: Mode and place of birth influence the neonatal gut microbiota, likely shaping its interplay with the host through the maturation of the intestinal epithelium, regulation of the intestinal epithelial barrier, and control of the innate immune system during early life, which can affect the phenotypic responses linked to metabolic processes in infants.

Trial registration: NCT03552939 . Video Abstract.

Keywords: Antibiotics; Early programming; Environment; Epithelial barrier; Immune system; Microbiota; Mode of birth.

Conflict of interest statement

The authors report no potential conflict of interest.

Figures

Fig. 1

Factors affecting neonatal microbiota during the first month of life. a–c Discriminant analysis of principal components (DAPC) of the neonatal (a) and infant fecal microbiota at 7 days (b) and 31 days (c) at ASV level. Each point represented microbiota from a neonate. Adonis analysis was used to stablish the significance of studied variables. d Colonization patterns during the first moth of life. Neonatal microbiota composition at phylum level at birth (0d), 7 days (7d), and 1 month (31d). C-section (CS, n = 65), vaginal delivery at hospital (VAG, n = 92), and homebirth (HB, n = 23)

Fig. 2

Differences in relative abundance of most important and variable genera in fecal microbiota among the first month of life. Each point represented the mean and SEM of relative abundance of each genus in that point from the fecal samples of babies born by cesarean section (blue), vaginal delivery at hospital (green), and at home (orange). Kruskal-Wallis test with a Dunn’s post hoc test was performed to compare the different groups. Data not sharing the same letter in each point were significantly different (p < 0.05). Significant variations within the same group at different time points were marked by an asterisk (*). C-section (CS, n = 65), hospitalized vaginal delivery (VAG, n = 92), and homebirth (HB, n = 23)

Fig. 3

Place and mode of birth impact the infant growth. BMI z-scores (a) and weight for length (b) z-scores curves from delivery to 18 months of life according to mode of birth and place adjusted by covariates, breastfeeding duration, antibiotic intake during the first year of life, maternal pre-gestational BMI and infant BMI and weight for length (W/L) z-scores at birth. General linear model multivariate test adjusted by covariates was done and p < 0.05 was considered significant. Kruskal-Wallis was performed on the adjusted values (different letters indicate significant differences between three studied groups). C-section (CS, n = 58), hospitalized vaginal delivery (VAG, n = 85), and homebirth (HB, n = 23)

Fig. 4

Microbial functions computationally predicted present in neonatal microbiota along the first month of life. a, b Discriminant analysis of principal components (DAPC) of the neonatal (a) and infant fecal microbiota at 7 days and 31 days (b). Adonis analysis was used to stablish the significance of studied variables. c, d Computational analysis of lipopolysaccharide (LPS) biosynthesis (c) bacterial toxins (d) routs presents in the fecal microbiota of newborns along the first month of life. Results were expressed as percentage of total functional routs for each sample. *p < 0.05, **p < 0.01, ***p < 0.001. C-section (CS, n = 65), hospitalized vaginal delivery (VAG, n = 92), and homebirth (HB, n = 23)

Fig. 5

Effect of 1-month infant fecal water exposure in epithelial (a) and macrophages-like (b) cell lines after 24 h. a Cytokine production by HT-29 cells after exposure to fecal water from neonates born by C-Section (CS), vaginal delivery at hospital (VAG), and homebirth (HB). IL6 production in HT-29 cell line was below detection limit. b Cytokine production of THP1 cells after 24 h exposure to fecal water of each group. Data was presented as median and whiskers represented the 5–95 percentile. Kruskal-Wallis and Dunn’s post hoc (FDR adjustment) test was used to test the significance of the differences in cytokine response between the groups. *p < 0.05, **p < 0.01, ***p < 0.001. C-section (CS), hospitalized vaginal delivery (VAG), and homebirth (HB)

Fig. 6

Effect of fecal water long-term exposure (7 days) on the triple co-culture system. a, b Epithelial barrier function measured as trans-epithelial electric resistance (TEER) (a) and Lucifer yellow transport (LY) (b). c Mucus production by LSTH17 cells after the long-term exposure on the triple co-culture system measured by Bradford assay. d Interleukin (IL) 8 production by cells on the apical compartment (CacO-2 and LSTH17) measured by ELISA and expressed as increment respect to control condition. e, f Intestinal cells functional maturation degree measured as intestinal alkaline phosphatase activity (IAP) on apical compartment during the treatment (e) and at final time point (f). g, h Cytokine production in the basal compartment by THP-1 cells. IL-8 (g) and IL-6 (h) production after fecal supernatant long-term exposure expressed as increment respect to control condition. The treatments were fecal water from infants born by C-section (CS), vaginal delivery at hospital (VAG), and homebirth (HB). Non-normal data was presented as median and whiskers represented the 5–95 percentile while normal data was showed as mean and SD. Kruskal-Wallis/ANOVA and Dunn’s/Tukey’s post hoc (FDR adjustment) test was used to test the significance of the normal/non normal distributed variables between the groups. In the cytokine analysis, the symbol (*) represented variations between time within the same studied group according to the color. *p < 0.05, **p < 0.01, ***p < 0.001