Priority effects, nutrition and milk glycan-metabolic potential drive <i>Bifidobacterium longum</i> subspecies dynamics in the infant gut microbiome
Abstract
<h4>Background</h4>The initial colonization of the infant gut is a complex process that defines the foundation for a healthy microbiome development. <i>Bifidobacterium longum</i> is one of the first colonizers of newborns' gut, playing a crucial role in the healthy development of both the host and its microbiome. However, <i>B. longum</i> exhibits significant genomic diversity, with subspecies (<i>e.g., Bifidobacterium longum</i> subsp. <i>infantis</i> and subsp. <i>longum</i>) displaying distinct ecological and metabolic strategies including differential capabilities to break down human milk glycans (HMGs). To promote healthy infant microbiome development, a good understanding of the factors governing infant microbiome dynamics is required.<h4>Methodology</h4>We analyzed newly sequenced gut microbiome samples of mother-infant pairs from the Amsterdam Infant Microbiome Study (AIMS) and four publicly available datasets to identify important environmental and bifidobacterial features associated with the colonization success and succession outcomes of <i>B. longum</i> subspecies. Metagenome-assembled genomes (MAGs) were generated and assessed to identify characteristics of <i>B. longum</i> subspecies in relation to early-life gut colonization. We further implemented machine learning tools to identify significant features associated with <i>B. longum</i> subspecies abundance.<h4>Results</h4><i>B. longum</i> subsp. <i>longum</i> was the most abundant and prevalent gut <i>Bifidobacterium</i> at one month, being replaced by <i>B. longum</i> subsp. <i>infantis</i> at six months of age. By utilizing metagenome-assembled genomes (MAGs), we reveal significant differences between and within <i>B. longum</i> subspecies in their potential to break down HMGs. We further combined strain-tracking, meta-pangenomics and machine learning to understand these abundance dynamics and found an interplay of priority effects, milk-feeding type and HMG-utilization potential to govern them across the first six months of life. We find higher abundances of <i>B. longum</i> subsp. <i>longum</i> in the maternal gut microbiome, vertical transmission, breast milk and a broader range of HMG-utilizing genes to promote its abundance at one month of age. Eventually, we find <i>B. longum</i> subsp. <i>longum</i> to be replaced by <i>B. longum</i> subsp. <i>infantis</i> at six months of age due to a combination of nutritional intake, HMG-utilization potential and a diminishment of priority effects.<h4>Discussion</h4>Our results establish a strain-level ecological framework explaining early-life abundance dynamics of <i>B. longum</i> subspecies. We highlight the role of priority effects, nutrition and significant variability in HMG-utilization potential in determining the predictable colonization and succession trajectories of <i>B. longum</i> subspecies, with potential implications for promoting infant health and well-being.