Neonatal microbiota development and the effect of early life antibiotics are determined by two distinct settler types

Abstract

The neonatal period, during which the initial gut microbiota is acquired, is a critical phase. The healthy development of the infant's microbiome can be interrupted by external perturbations, like antibiotics, which are associated with profound effects on the gut microbiome and various disorders later in life. The aim of this study was to investigate the development of intestinal microbiota and the effect of antibiotic exposure during the first three months of life in term infants. Fecal samples were collected from healthy infants and infants who received antibiotics in the first week of life at one week, one month, and three months after birth. Microbial composition was analyzed using IS-pro and compared between antibiotics-treated and untreated infants. In total, 98 infants, divided into four groups based on feeding type and delivery mode, were analyzed. At one week of age, samples clustered into two distinct groups, which were termed "settler types", based on their Bacteroidetes abundance. Caesarean-born infants belonged to the low-Bacteroidetes settler type, but vaginally-born infants were divided between the two groups. The antibiotics effect was assessed within a subgroup of 45 infants, vaginally-born and exclusively breastfed, to minimize the effect of other confounders. Antibiotics administration resulted in lower Bacteroidetes diversity and/or a delay in Bacteroidetes colonization, which persisted for three months, and in a differential development of the microbiota. Antibiotics resulted in pronounced effects on the Bacteroidetes composition and dynamics. Finally, we hypothesize that stratification of children's cohorts based on settler types may reveal group effects that might otherwise be masked.

Fig 1. Heat map of all profiles at T = 1 (one week), sorted and colored by phylum.

Abundance of OTUs in each sample at T = 1 (one week). Rows correspond to samples (antibiotic treatment: red; vaginal delivery: green; caesarean section: orange); columns correspond to OTUs. Phyla color shades represent the abundance of each OTU in a sample (Bacteroidetes: pink; FAFV: blue). Cosine correlations and hierarchical clustering were calculated on the raw data, but for a better interpretation we only present the called peaks in this visualization. Top cluster samples (pink) belong to the Bacteroidetes-dominant settler type. Bottom cluster samples (blue) belong to the low-Bacteroidetes settler type (with three exceptions). FAFV = Firmicutes, Actinobacteria, Fusobacteria, Verrucomicrobia; OTU = operational taxonomic units.

Fig 2. Convergence of settler types over time.

A principle coordinate analysis of intestinal microbiota of all untreated infants at week 1 (T = 1), month 1 (T = 2) and month 3 (T = 3). Samples are colored by settler type. (b) Cosine distances, as a measure of community dissimilarity between settler types, displayed over time by treatment group. Dots indicate median cosine distances, bars indicate IQR (inter quartile range). A higher distance indicates the settler types are less similar.

Fig 3. Abundance and diversity of Bacteroidetes and FAFV by settler type and treatment group.

(a) Average abundance of Bacteroidetes and FAFV over time by settler type (rows) and treatment group (columns). Values are log2 transformed intensities (measured as log2 RFU). Error bars depict the standard error. (b) Median Shannon diversity index of Bacteroidetes and FAFV over time by settler type (rows) and treatment group (columns). Error bars denote the upper and lower IQR (inter quartile range); RFU—relative fluorescent units; FAFV = Firmicutes, Actinobacteria, Fusobacteria, Verrucomicrobia.

Fig 4. Cosine distances as a measure of community dissimilarity within individuals over time by settler type and treatment group.

Cosine distances displayed by settler type in antibiotic treated infants (left) and controls (right). Distances shown are between one week of age (T = 1) and three months of age (T = 3) within each individual. Dots indicate median cosine distances (during the 3-month interval) of all individuals of each settler type; bars indicate IQR (inter quartile range).