rpoB, a promising marker for analyzing the diversity of bacterial communities by amplicon sequencing

BackgroundMicrobiome composition is frequently studied by the amplification and high-throughput sequencing of specific molecular markers (metabarcoding). Various hypervariable regions of the 16S rRNA gene are classically used to estimate bacterial diversity, but other universal bacterial markers with a finer taxonomic resolution could be employed. We compared specificity and sensitivity between a portion of the rpoB gene and the V3V4 hypervariable region of the 16S rRNA gene.ResultsWe first designed universal primers for rpoB suitable for use with Illumina sequencing-based technology and constructed a reference rpoB database of 45,000 sequences. The rpoB and V3V4 markers were amplified and sequenced from (i) a mock community of 19 bacterial strains from both Gram-negative and Gram-positive lineages; (ii) bacterial assemblages associated with entomopathogenic nematodes. In metabarcoding analyses of mock communities with two analytical pipelines (FROGS and DADA2), the estimated diversity captured with the rpoB marker resembled the expected composition of these mock communities more closely than that captured with V3V4. The rpoB marker had a higher level of taxonomic affiliation, a higher sensitivity (detection of all the species present in the mock communities), and a higher specificity (low rates of spurious OTU detection) than V3V4. We compared the performance of the rpoB and V3V4 markers in an animal ecosystem model, the infective juveniles of the entomopathogenic nematode Steinernema glaseri carrying the symbiotic bacteria Xenorhabdus poinarii. Both markers showed the bacterial community associated with this nematode to be of low diversity (