Orientia tsutsugamushi presents a formidable challenge to genomic analysis. Its genome is highly heterogeneous and repeat-rich, features that complicate both sequencing and assembly. For decades, molecular epidemiology of this bacterium has leaned on a small number of marker loci, short stretches of DNA assumed to serve as reliable proxies for the organism’s broader evolutionary relationships. The most widely used of these is tsa56, which encodes a major outer membrane antigen, alongside htrA and groEL, two more conserved housekeeping loci. Generations of genotype labels assigned to clinical and environmental isolates rest on sequences from these markers. Yet whether those labels correspond to true genome-wide relatedness had never been rigorously evaluated at scale.

The research team, led by Lele Zhao, Shiyu Li, and Shijie Chen, approached the question by assembling publicly available sequences of the three standard marker loci together with a set of 19 quality-controlled whole genomes. Their first test compared the two most widely used markers, tsa56 and htrA, across 35 isolates carrying both sequences. The phylogenetic trees built from these two markers were markedly incongruent. The Mantel correlation coefficient between the two distance matrices was just 0.103, with a P value of 0.433, indicating no statistically significant association. The normalized Robinson–Foulds distance, a measure of topological disagreement between phylogenetic trees, reached 0.938 on a scale where 1.0 represents complete discordance. In practical terms, the evolutionary picture painted by tsa56 bore almost no resemblance to the picture painted by htrA for the same set of isolates.

The team then evaluated whether any of the three marker loci could recover the topology of the core genome, the set of genes shared across all strains and therefore the best available proxy for the bacterium’s genuine vertical evolutionary history. None of the three markers succeeded. Normalized Robinson–Foulds distances between each marker tree and the core-genome tree ranged from 0.750 to 0.800, indicating substantial to severe topological discordance in every case. This result demonstrates that genotype assignments based on any single locus are strongly locus-dependent and cannot be considered directly interchangeable across markers. A strain labeled as one genotype according to tsa56 may occupy a very different position in the evolutionary tree when assessed with htrA or with genome-wide data.

Not all markers performed equally poorly, however. When the researchers examined whether marker-based pairwise distances correlated with genome-wide pairwise distances, the more conserved loci retained meaningful signal. The groEL marker produced a Mantel correlation coefficient of 0.845 with genome-wide distances, and htrA produced 0.613, both statistically significant at P less than 0.001. This means that although neither locus reproduces the core-genome tree faithfully, both carry usable information about how genetically distant strains are from one another. The tsa56 locus, by contrast, showed no significant correlation with genome-wide distance, producing a Mantel coefficient of just 0.146 with a P value of 0.313. For the most heavily used marker in scrub typhus molecular epidemiology, this is a sobering result. It suggests that tsa56, shaped perhaps by strong immune selection on the surface antigen it encodes, evolves in ways that decouple it from the rest of the genome.

The study also compared two different genome-scale approaches: a phylogeny built from core-genome single nucleotide polymorphisms and one built from core proteins. These showed a significant positive correlation in their distance matrices, with a Mantel coefficient of 0.625 and P less than 0.001, yet their topologies still differed, with a normalized Robinson–Foulds distance of 0.500. Even at the whole-genome level, nucleotide-based and protein-based reconstructions of Orientia’s evolution tell partially different stories, a reminder that analytical choices matter even when the data are rich.

Beyond the marker comparison, the team conducted a pan-genome analysis, cataloging the full repertoire of genes present across the sampled strains. They identified 4,350 gene clusters in total, of which only 13.8 percent constituted the core genome, meaning genes present in every strain. The remaining vast majority were accessory genes, present in some strains but not others, or strain-specific genes found in only a single isolate. The pan-genome accumulation curve remained open, rising steadily as more genomes were added without approaching a plateau. An open pan-genome indicates that the gene repertoire of Orientia tsutsugamushi is far from fully sampled and that each newly sequenced strain is likely to carry previously unseen genes. This pattern is typical of bacterial species with extensive horizontal gene transfer and recombination, and it underscores how much of this pathogen’s genetic repertoire remains undiscovered.

A striking feature of the pan-genome is how little of it can be functionally annotated. Most accessory genes, 60.1 percent, and an even larger proportion of strain-specific genes, 81.3 percent, were uncharacterized, meaning that no known function could be assigned to them. These genes of unknown function may include determinants of virulence, host adaptation, or antibiotic tolerance, but their roles remain speculative until functional studies catch up with the genomic data. The open, accessory-dominated structure of the pan-genome suggests that Orientia’s evolutionary success as a pathogen may depend heavily on this unexplored genetic reservoir.

The researchers also tested whether accessory gene composition correlated with either the geographic origin of the strains or their core-genome lineages. Despite the high heterogeneity of accessory-gene repertoires across isolates, no statistically detectable structuring by geography or core lineage emerged in the dataset. This absence of association suggests that horizontal gene transfer reshapes the accessory genome largely independently of the vertical evolutionary history recorded in the core genome and independently of where strains circulate. In other words, two strains isolated from the same region or belonging to the same core lineage may carry substantially different accessory gene sets, while distant strains may share accessory genes acquired through recombination.

The findings carry immediate practical implications. Genotype labels and strain relationships derived from single markers should be interpreted strictly within a locus-specific framework and should not be treated as interchangeable across loci, the authors conclude. In the small matched genome sets available, groEL and htrA better reflected genome-wide pairwise distances than tsa56, offering a hierarchy of marker reliability for situations where whole-genome sequencing is not feasible. More broadly, the study provides a curated overview of Orientia tsutsugamushi diversity in public datasets and a reproducible analytical framework for future surveillance. As scrub typhus continues to expand its recognized geographic range and as whole-genome sequencing becomes more accessible, this work lays the groundwork for a transition from marker-based classification toward genome-informed epidemiology, an essential step toward improved diagnostics, vaccine design, and therapeutic strategies against this neglected but dangerous pathogen.

Subject of Research: Genome-scale analysis of genetic diversity and pan-genome architecture in the scrub typhus bacterium Orientia tsutsugamushi

Article Title: Genome-scale insights into the genetic diversity and pan-genome architecture of Orientia tsutsugamushi

Article References: Zhao, L., Li, S., Chen, S., Liu, J., Li, R., Yu, Z., Yang, F., Shao, Z., Liu, K., & Lu, Z. (2026). Genome-scale insights into the genetic diversity and pan-genome architecture of Orientia tsutsugamushi. BMC Genomics. https://doi.org/10.1186/s12864-026-13351-0

Image Credits: AI Generated

DOI: 10.1186/s12864-026-13351-0

Keywords: Orientia tsutsugamushi, scrub typhus, pan-genome, genetic diversity, molecular epidemiology, phylogenetic discordance, tsa56, groEL, htrA, accessory genes, genomic surveillance, core genome