Metagenomic features of traditional fermented milk products

This study analyzed the microbiomes of 23 Chinese koumiss samples using deep metagenomic sequencing, as well as metagenome datasets of 20 nunu, 18 kefir, and 12 yogurt samples retrieved from public genome databases. There were large variations in the microbiota composition between sample types, reflected by the great differences in their profiles of dominant species. One-hundred fifty-three metagenome-assembled genomes were assembled, including four novel species belonging to the Lactobacillus, Streptococcus, Acetobacter, Rothia genera. The novel genomes contained various gene elements, encoding carbohydrate-active enzymes, secondary metabolite clusters, and bacteriocins, which might influence food flavor, quality and/or safety. Bacteriocin-encoding genes were common, which might help control undesirable microorganisms. Our metagenomic analysis also revealed the content of bacteriophage sequences of these products, meanwhile uncovering complex interactions between fermented milk-associated bacteriophages and bacteria, including the history of bacteriophage infection and relationship between Lactobacillaceae and bacteriophages. This study highlighted the microbial diversity and heterogeneity across fermented milk products. Our data demonstrated the power of deep metagenomic sequencing in expanding our understanding of low-complexity microbial communities.