Human Milk Archaea Associated with Neonatal Gut Colonization and Its Co-Occurrence with Bacteria

Abstract

Archaea have been identified as early colonizers of the human intestine, appearing from the first days of life. It is hypothesized that the origin of many of these archaea is through vertical transmission during breastfeeding. In this study, we aimed to characterize the archaeal composition in samples of mother-neonate pairs to observe the potential vertical transmission. We performed a cross-sectional study characterizing the archaeal diversity of 40 human colostrum-neonatal stool samples by next-generation sequencing of V5-V6 16S rDNA libraries. Intra- and inter-sample analyses were carried out to describe the Archaeal diversity in each sample type. Human colostrum and neonatal stools presented similar core microbiota, mainly composed of the methanogens Methanoculleus and Methanosarcina. Beta diversity and metabolic prediction results suggest homogeneity between sample types. Further, the co-occurrence network analysis showed associations between Archaea and Bacteria, which might be relevant for these organisms' presence in the human milk and neonatal stool ecosystems. According to relative abundance proportions, beta diversity, and co-occurrence analyses, the similarities found imply that there is vertical transmission of archaea through breastfeeding. Nonetheless, differential abundances between the sample types suggest other relevant sources for colonizing archaea to the neonatal gut.

Keywords: 16S rDNA; Archaea; breastfeeding; human milk; microbiota; neonatal gut; vertical transmission.

Figure 1

Archaeal composition in samples. (A) Stacked bar plots showing the archaeal relative mean abundance at the phylum level. Red and green colors indicate sample type and relative abundances are shown on the Y-axis. (B) Bar plots showing the square root relative mean abundance of archaea genera. The Y-axis labels indicate genus, while the X-axis shows the square root-relative and mean abundance. Error bars indicate the standard error of the mean.

Figure 2

Source tracker analysis of possible archaeal origin in neonatal stool samples. (A) Bar plots indicate the total source proportion. The Y-axis shows the proportion of the source for the neonatal stool samples, and the X-axis shows the source (colostrum or unknown). (B) Stacked bar plots showing genera composition for each source. The Y-axis shows the genera abundance, and the X-axis shows the source (colostrum or unknown). Color sectors indicate the source or the archaeal genus, according to the labels beside the graphics.

Figure 3

Archaeal diversity in samples. (A) Alpha diversity box plots. The Y-axis indicates values for the species richness (Observed), diversity indexes (Shannon, Simpson, and Inverse Simpson), and evenness (Fisher). The sample type is shown at the X-axis (see Table S4, for numerical data of indexes) Beta diversity Non-Metric Multidimensional Scaling (NMDS) scatter plots. The graphics show archaeal beta diversity calculated by (B) NMDS ordination based on weighted UniFrac and (C) Jensen–Shannon diverge (JSD) distances. Sample types (colostrum and neonatal stool) are similar according to ANOSIM (weighted UniFrac R = −0.02, p = 0.933 and JSD: R = −0.001 p = 0.494).

Figure 4

Correlation analysis of archaea between pairs and metadata. (A) Paired Pearson correlation test between colostrum and neonatal stool. Only p-values < 0.05 were included. The Y-axis shows the square root of mean relative abundance, and the X-axis, sample types. (B–D). Correlation analysis between archaea and metadata using aldex.cor module. No significant correlation was found with the mother’s (A) age, (B) pre-pregnancy BMI, or (D) weight gain during pregnancy.

Figure 5

Microbial co-occurrence network comparison between human milk and neonatal stools. (A) Venn diagram of edges between the networks of neonatal stool (NS) and human colostrum (HC). (B) Number distribution of taxa associated with the linked nodes of positive edges in networks of NS and HC. The number in the plot indicates the ratio of edges against all the positive edges in the network. (C) Microbial co-occurrence network of neonatal stool. A connection between nodes stands for a strong (Spearman’s ρ > 0.6) and significant (p > 0.01) correlation. (D) Microbial co-occurrence network of human colostrum. A connection stands for a strong (Spearman’s ρ > 0.6) and significant (p > 0.01).

Figure 6

Heatmap of functional microbial metabolic pathways using PICRUSt2 analysis with MetaCyc database. Columns show the abundance of main metabolic pathways with a prevalence of at least 10% in the samples and an abundance > 1%. Sample names are shown on the X-axis. The color scale from black (−2) to white (2) indicates the relative abundance of the predicted metabolic pathways; the green and red color tags indicate the sample type.