Distinct Changes Occur in the Human Breast Milk Microbiome Between Early and Established Lactation in Breastfeeding Guatemalan Mothers

Abstract

Human breast milk contains a diverse community of bacteria, but as breast milk microbiome studies have largely focused on mothers from high income countries where few women breastfeed to 6 months, the temporal changes in the breast milk microbiome that occur during later lactation stages have not been explored. For this cross-sectional study, microbiota from breast milk samples of Mam-Mayan mothers living in eight remote rural communities in the Western Highlands of Guatemala were analyzed. All mothers delivered vaginally and breastfed their infants for 6 months. Breast milk from 76 unrelated mothers was used to compare two lactation stages, either "early" (6-46 days post-partum, n = 33) or "late" (109-184 days post-partum, n = 43). Breast milk microbial communities were assessed using 16S ribosomal RNA gene sequencing and lactation stages were compared using DESeq2 differential abundance analysis. A total of 1,505 OTUs were identified, including 287 which could be annotated as putative species. Among several maternal factors, lactation stage explained microbiome variance and inertia in ordination with the most significance (p < 0.001). Differential abundance analysis identified 137 OTUs as significantly higher in either early or late lactation. These included a general shift from Staphylococcus and Streptococcus species in early lactation to Sphingobium and Pseudomonas species in late lactation. Species enriched in early lactation included putative commensal bacteria known to colonize the infant oral and intestinal tracts whereas species enriched in late lactation had a uniform functional trait associated with aromatic compound degradation. Differentially abundant species also included several species which have not previously been reported within breast milk, such as Janthinobacterium agaricidamnosum, Novosphingobium clariflavum, Ottowia beijingensis, and Flavobacterium cucumis. These discoveries describe temporal changes to the breast milk microbiome of healthy Guatemalan mothers from early to late lactation. Collectively, these findings illustrate how studying under-represented human populations might advance our understanding of factors that modulate the human milk microbiome in low and middle income countries (LMIC).

Keywords: 16S rRNA gene; human breast milk; lactation stage; metagenomics 16S; microbiome.

FIGURE 1

Breast milk community ordination and alpha diversity comparing early and late lactation microbiome. (A) Variation explained by the tested variables in constrained ordination (CAP). Significance testing using Monte Carlo permutation test (10,000 permutation; ***p < 0.001). (B) Lactation stage CAP, circles represent sample density relative to each group. (C) Alpha diversity indices were not significantly different (t-test, p > 0.05) between early (n = 33) and late (n = 43) lactation groups. See Supplementary File 1 for an overview of OTU taxonomy.

FIGURE 2

Differentially abundant bacteria associated to lactation stage. Differentially abundant OTUs between the early (n = 33) and late (n = 43) groups. (A) Fold change (FC log₂) in relative abundance of significantly different (DESeq2) OTUs between groups and their normalized mean counts. Species are grouped by phylum. The dashed red line indicates “infinite” fold change, where an OTU had detectable counts in samples from only a single group. (B) An illustration of species-level differentially abundant bacteria between milk of early (blue) and late (red) lactation stages. Raw abundance is represented. The complete OTU table including relative abundance, annotation, count distribution, blast statistics, alternative database hits, and sequences are provided in Supplementary File 1.

FIGURE 3

OTU-OTU correlations across all breast milk samples. Count correlation of all OTUs differentially abundant between early and late lactation stages which could be annotated at species level. Colors indicate positive (blue) or negative (red) correlation and black dots indicate significance (p < 0.05). See Supplementary File 1 correlations including non-differentially abundant of OTUs from mothers.