Patient-centric analysis of Orientia tsutsugamushi spatial diversity patterns across Hainan Island, China

Abstract

Background: Scrub typhus, traditionally caused by Orientia tsutsugamushi, is a re-emerging public health concern within the Tsutsugamushi Triangle. Despite growing awareness, prevention strategies remain inadequate on Hainan Island, China, where scrub typhus poses a significant threat, especially in field-related environments.

Methodology/principal findings: Gene flow analysis of the tsa56 gene and multilocus sequence typing (MLST) were conducted on 156 previously confirmed scrub typhus cases from 2018 to 2021 across Hainan Island. By integrating published datasets, we identified 12 major sub-genotypes and traced their origins, revealing that these sub-genotypes share origins with isolates from Southeast Asia and coastal provinces and island of China, but also demonstrate unique local adaptations across all isolates. Alpha diversity index analysis was applied across administrative regions to identify hotspot regions. This analysis showed that nine out of the detected fourteen administrative regions, particularly along the northern and western coastlines and inland areas, exhibited relatively high genetic diversity, with the highest incidence observed in Qiongzhong, a centrally located city. Related major sequence types were mapped, and distances between locations were estimated, showing that identical MLST sequence types were observed to transfer across distances of 23 to 125 km between different sites on the island. Pathogen density was analyzed using quantitative real-time PCR targeting the tsa56 gene. Without accounting for potential confounding factors or dataset limitations, the Karp_B_2 sub-genotype showed a significant increasing trend in pathogen density with prolonged fever duration, while Gilliam sub-genotypes exhibited a slower or even declining trend.

Conclusions/significance: These findings emphasize the urgent need for targeted public health interventions, particularly focusing on vulnerable populations in rural and agricultural areas of nine key administrative regions where high genetic diversity and pathogen spread were observed. Additionally, this study provides valuable insights into the transmission dynamics and infection progression of scrub typhus, using gene flow analysis and multilocus sequence typing to identify major sub-genotypes.

Fig 1. Population genetic structure of O. tsutsugamushi in Hainan Island and comparison with other regions using MLST analysis.

(A) O. tsutsugamushi population structure on Hainan Island (sample size n = 74) as revealed by geoBURST analysis, identifying newly reported sequence types (STs) and their associated clonal complexes (CCs). (B) Comparison of O. tsutsugamushi populations from Hainan Island (sample size n = 74) with those from other regions in the pubMLST database (n = 235), focusing only on clonal complexes (CCs). Regional breakdown: Thailand (n= 83, ST1-ST44); Papua New Guinea (n = 1, ST45); Burma (n = 1, ST46); Australia (n = 1, ST47); Japan (n = 2, ST20, ST49); Laos (n = 74, ST50-ST92, ST1, ST4, ST9, ST25, ST29, ST30, ST37); South Korea (n = 63, ST48, ST93-ST98); Malaysia (n = 2, ST99, ST100); Bangladesh (n = 8, ST101-ST108); Hainan Island, China (n=74, ST111-ST175).

Fig 2. Clonal complexes of O. tsutsugamushi population from Hainan Island, China, and other regions using phyloviz with full minimum spanning tree (MST).

The numbers in the figure represent unique clone groups based on sequence types (STs) defined in the MLST database. ST node colors represent different regions: Hainan Island (sample size n = 74), Australia (n = 1), Bangladesh (n = 8), Burma (n = 1), Japan (n = 2), Laos (n = 74), Malaysia (n = 2), Papua New Guinea (n = 1), South Korea (n = 63), and Thailand (n = 83).

Fig 3. Comparative and spatial analysis of genetic data.

(A) Tanglegram comparing tsa56 hypervariable regions with middle regions of conserved genes for MLST. (B) Map displaying locations of branches/subspecies derived from phylogenetic analysis.

Fig 4. Comparison of Shannon Diversity Index across cities on Hainan Island.

A. Distribution of Shannon Diversity Index; B. Regional comparison of Shannon Diversity Index.

Fig 5. Linear regression fit of infection days on pathogen density by genotype.

The linear relationship between infection days and pathogen density (log-transformed) for each genotype is illustrated. Pathogen density is quantified as the copy number of the tsa56 gene per milliliter (mL) of blood. Each panel represents a different genotype, with scatter points indicating pathogen density in individual samples and color coding the severity of the condition (Mild, Moderate, Severe). The solid black line represents the linear regression fit for each genotype, and P-values for each genotype’s regression model are annotated on the plot.