Design of high specificity and sensitive sequencing primers to detect the bifidobacterial population in biological samples

Abstract

Background : Bifidobacteria were poorly represented or even totally absent in many of the currently published microbiota/metagenomics datasets. This work designed Bifidobacterium -targeting primers that were suitable for use in high-throughput sequencing to characterized the feature of Bifidobacterium . Results : After multiple steps of in silico prediction, phylogenetic analysis, polymerase chain reaction and single-molecule real-time sequencing (SMRT) experiments, the Bif-6 primer pair was confirmed to be optimal for such purpose. Up to 99.42% and 99.32% of the detected sequences in human stool and breast milk samples represented up to 6 and 8 different Bifidobacterium species, respectively. Additionally, the Bif-6 primer pair detected the species Bifidobacterium catenulatum and Bifidobacterium kashiwanohense , which have not previously been reported in human breast milk to our knowledge. The high specificity of Bif-6 could be explained by the lower average nucleotide identity of 78.42% (versus 93.91% for the universal primers 27f/1492r). Moreover, to optimize the process of microbiota annotation, here a Bifidobacterium -specific database was constructed that largely reduced the time required for annotation compared with using the NCBI nucleotide (nt) database. The two databases generated highly consistent profiles of bifidobacteria when tested on the same samples. Conclusion : this work designed a highly reliable primer and database specific for detecting the bifidobacterial population in complex biological samples, which will facilitate future studies of this group of bacteria.