Bifidobacterium longum subsp. infantis (B. infantis) is one of a few microorganisms capable of metabolizing human breast milk and is a pioneer colonizer in the guts of breastfed infants. One current challenge is differentiating B. infantis from its close relatives, B. longum and B. suis. All three organisms are classified in the same species group but only B. infantis can metabolize human milk oligosaccharides (HMOs). We compared HMO-metabolizing genes across different Bifidobacterium genomes and developed B. infantis-specific primers to determine if the genes alone or the primers can be used to quickly characterize B. infantis. We showed that B. infantis is uniquely identified by the presence of five HMO-metabolizing gene clusters, tested for its prevalence in infant gut metagenomes, and validated the results using the B. infantis-specific primers. We observed that only 15 of 203 (7.4%) children under 2 years old from a cohort of US children harbored B. infantis. These results highlight the importance of developing and improving approaches to identify B. infantis. A more accurate characterization may provide insights into regional differences of B. infantis prevalence in infant gut microbiota.

Bifidobacterium longum subsp. infantis (B. infantis) is one of few microorganisms capable of metabolizing human breast milk and is a pioneer colonizer in the guts of breastfed infants. One current challenge is differentiating B. infantis from its close relatives, B. longum and B. suis, by molecular methods. These two organisms are classified in the same species group as B. infantis but do not share the ability to metabolize human milk oligosaccharides (HMOs). Here, we compared HMO-metabolizing genes across different Bifidobacterium genomes to develop B. infantis specific primers and determine if they alone can be used to quickly characterize B. infantis with shotgun metagenomic sequencing data. We showed that B. infantis is uniquely identified by the presence of five HMO-metabolizing gene clusters, used this characterization to test for its prevalence in infants, and validated the results using the B. infantis-specific primers. By examining stool samples from a cohort of US children and pregnant women using shotgun metagenomic sequencing, we observed that only 18 of 204 (8.8%) of children under 2 years old harbored B. infantis. These results highlight the importance of developing and improving approaches to identify B. infantis. A more accurate characterization may provide insights into regional differences of B. infantis prevalence in infant gut microbiota.