Comparative virulence analysis of seven diverse strains of Orientia tsutsugamushi reveals a multifaceted and complex interplay of virulence factors responsible for disease.

Abstract

Orientia tsutsugamushi is an obligate intracellular bacterium found in Leptotrombidium mites that causes the human disease scrub typhus. A distinguishing feature of O. tsutsugamushi is its extensive strain diversity, yet differences in virulence between strains are not well defined nor well understood. We sought to determine the bacterial drivers of pathogenicity by comparing murine infections using seven strains combined with epidemiological human data to rank each strain in terms of relative virulence. Murine cytokine expression data revealed that the two most virulent strains, Ikeda and Kato, induced higher levels of IL-6, IL-10, IFN-{gamma} and MCP-1 than other strains, consistent with increased levels of these cytokines in severe human scrub typhus patients. We sought to identify the mechanistic basis of the observed differential virulence between strains by comparing their genomes, in vitro growth properties and cytokine/chemokine induction in host cells. We found that there was no single gene or gene group that correlated with virulence, and no clear pattern of in vitro growth rate that predicted disease. However, microscopy-based analysis of the intracellular infection cycle revealed that the only fully avirulent strain in our study, TA686, differed from all the virulent strains in its subcellular localisation and expression of its surface protein ScaC. We conclude that drivers of pathogenicity in Orientia tsutsugamushi are distributed throughout the genome, likely in the large and varying arsenal of effector proteins encoded by different strains, and that these interact in complex ways to induce differing immune responses and thus differing disease outcomes in mammalian hosts. Author SummaryScrub typhus is a vector-borne human disease caused by the bacterium Orientia tsutsugmushi and spread by mites. There are numerous different strains of this bacterium with some causing more severe disease in humans than others, and some that do not cause any illness at all. The factors driving these differences are not yet understood, and gaining insight into them could aid in vaccine development and help predict the severity of disease caused by new isolates. To better determine the mechanistic basis of pathogenicity in scrub typhus, we carried out experiments in which we compared seven diverse strains for virulence in animals. We measured their ability to cause disease in mice, so that we could reliably classify them as virulent or avirulent in this model. We then analysed various genomic and biological aspects to identify disease markers in both mice and humans. We report here that there is no single factor that predicts whether a strain will be pathogenic or not, but that disease in scrub typhus is a complex process resulting from the activity of multiple bacterial genes working together to drive different immune responses in the host, resulting in either clearance of the bacteria from the host, or escalating disease. Future work exploring the relationship of bacterial effector proteins will help to disentangle this complex relationship in mechanistic detail.