In silico and experimental assessment of the 16S-ITS-23S operon as a genetic marker for high-resolution bacterial species identification.

Abstract

Accurate species-level identification of bacteria is crucial for public health surveillance, but standard methods like 16S rRNA hypervariable region (HVR) sequencing often lack sufficient resolution. This study systematically evaluates the taxonomic resolving power of the full 16S-ITS-23S ribosomal RNA (rrn) operon, enabled by high-accuracy long-read technology, compared to the full-length 16S gene and HVRs. A dual-validation framework was employed, combining an in silico analysis of a curated reference database with an experimental validation using rat faecal samples. In silico, the rrn operon demonstrated unequivocally superior accuracy across the full range of tested thresholds for both nucleotide mismatches and pairwise identity, consistently maintaining the highest proportion of monospecies clusters (i.e. clusters containing a single species). For instance, at a relaxed threshold of 30 mismatches, the operon maintained 96.8% monospecies clusters, compared to 80.5% for the full-length 16S gene and <66% for all HVRs. This theoretical advantage was confirmed experimentally through three parallel sequencing strategies: operon sequencing enabled species-level taxonomic assignment for 75.2% of unique sequences, significantly outperforming full-length 16S (62.1%) and dramatically surpassing the standard V4 HVR approach (18.5%). An internal bias-control analysis confirmed that these differences were due to the superior information content of the operon marker itself. Our findings provide robust evidence that high-accuracy long-read sequencing of the rrn operon is a superior method for culture-free bacterial surveillance, offering a new gold standard for high-resolution taxonomic profiling in complex environmental and host-associated samples.