Altered Gut Microbiome and Fecal Immune Phenotype in Early Preterm Infants With Leaky Gut

Abstract

Intestinal barrier immaturity, or "leaky gut", is the proximate cause of susceptibility to necrotizing enterocolitis in preterm neonates. Exacerbated intestinal immune responses, gut microbiota dysbiosis, and heightened barrier injury are considered primary triggers of aberrant intestinal maturation in early life. Inordinate host immunity contributes to this process, but the precise elements remain largely uncharacterized, leaving a significant knowledge gap in the biological underpinnings of gut maturation. In this study, we investigated the fecal cytokine profile and gut microbiota in a cohort of 40 early preterm infants <33-weeks-gestation to identify immune markers of intestinal barrier maturation. Three distinct microbiota types were demonstrated to be differentially associated with intestinal permeability (IP), maternal breast milk feeding, and immunological profiles. The Staphylococcus epidermidis- and Enterobacteriaceae-predominant microbiota types were associated with an elevated IP, reduced breast milk feeding, and less defined fecal cytokine profile. On the other hand, a lower IP was associated with increased levels of fecal IL-1α/β and a microbiota type that included a wide array of anaerobes with expanded fermentative capacity. Our study demonstrated the critical role of both immunological and microbiological factors in the early development of intestinal barrier that collectively shape the intestinal microenvironment influencing gut homeostasis and postnatal intestinal maturation in early preterm newborns.

Keywords: fecal cytokines; gut microbiome; intestinal barrier maturation; local immune phenotype; preterm infants.

Figure 1

Preterm infants exhibited three distinct microbiome types. (A) Heatmap of the 25 most abundant intestinal bacterial taxa and their relative abundance in samples collected from 38 preterm infants enrolled in the study. The taxonomic composition of the microbiomes used the data set of whole community metagenomic sequencing and the profile was generated by MetaPhlAn version 2 (40). Statistical significance was calculated using Wilcoxon rank sum test using ggsignif R package (41). Ward linkage clustering was used to cluster samples based on their Jensen-Shannon distance calculated in vegan package in R (42). The number of clusters was validated using gap statistics implemented in the cluster package in R (43) by calculating the goodness of clustering measure. (B) Boxplots of IP, GA, MOM, and BW depicting distribution of microbiota types in fecal samples of preterm newborns. Significance value was calculated using Wilcoxon rank sum test. *p < 0.05, **p < 0.01. Plotted are interquartile ranges (IQRs, boxes), medians (line in box), and mean (red diamond). IP, intestinal permeability; GA, gestational age; NS., not significant.

Figure 2

Fecal cytokine profile associated with microbial diversity. Color map of microbial communities correlating with neonatal factors and barplot map of the 15 cytokines detected. Within-sample diversity was estimated using Shannon diversity index using Phyloseq R package (46). Plot was generated using R package ‘complexheatmap’ (47). *Value was scaled using square root. **Microbiota type was assigned according to the clustering pattern as shown in Figure 1 . IP, intestinal permeability; GA, gestational age; BW, birth weight; MOM, mother’s own breast milk cumulative volume use during the first week prior to IP measurement.

Figure 3

IL-1α and IL-1β linked to lower IP and improved neonatal factors. Boxplots of the cytokine levels: (A-D, I) IL-1β, (E–H) IL-1α, (J) IL-16, (K) IL-10, (L, O) IL-12p40, (M) GM-CSF, (N) IL-12p70, (P) IL-7 between different categories. Plotted are interquartile ranges (IQRs, boxes), medians (line in box), and mean (red diamond). Significance value was calculated using Wilcoxon rank sum test using ggsignif R package (41). Asterisk denotes the level of significance. Threshold for early or late gestational age is 28 weeks, for low or high birth weight is 1,500 g, for longer or shorter abx exposure is 3 days during the first week after birth prior to IP measurement, for maternal breast milk feeding is 150-180 ml/kg of cumulative intake of MOM by 7-10 days of age, according to clinical convention and previously validation (35, 36). IP, intestinal permeability; GA, gestational age; BW, birth weight; MOM, mother’s own breast milk (MOM) cumulative volume use during the first week prior to IP measurement; SVD, spontaneous vaginal delivery; CS, cesarean section.

Figure 4

Disparity in clustering patterns of taxonomic and immunological profiling among different microbiota types. Canonical Correspondence Analysis (CCA) of (A) microbial taxonomic groups; (B) cytokine profiles. CCA was based on Bray-Curtis distance. CA1 and CA2 selected as the major components based on the eigenvalue. A scaled eigenvalues was shown on the plot to represent the direction from the origin where a group has a larger than average value for the particular profile (42, 51). *Microbiota type was assigned according to the clustering pattern demonstrated in Figure 1 .

Figure 5

Illustration of the intestinal barrier maturation in early preterm neonates with different microbiota types and cytokine profiles. An immature, compromised gut barrier may render the mucosa susceptible to invasion by opportunistic pathogens in the gut lumen. IP was linked with a microbial community dominated by a single species S. epidermidis, K. pneumoniae, or E. coli, with less defined cytokine profiles among individuals. A functional intestinal barrier was associated with neonates with later GA; higher BW; greater microbial community biodiversity that encompasses a wide array of anaerobic and facultative microorganisms not dominated by S. epidermidis, K. pneumoniae, or E. coli; and a trend of increased levels of IL-1α/β, IL-7, IL-12p40, IL-15, and IL-16. Schematic representation illustrates the distal intestine (not drawn to scale). Created with BioRender.com.