Genomic bacterial load associated with bacterial genotypes and clinical characteristics in patients with scrub typhus in Hainan Island, Southern China

Scrub typhus, caused by mite-borne Orientia tsutsugamushi (O. tsutsugamushi), is a major febrile disease in the Asia-Pacific region. The DNA load of O. tsutsugamushi in the blood was previously found to be significantly higher in patients with fatal disease than those with non-fatal disease and correlated with the duration of illness, presence of eschar, and hepatic enzyme levels. In this prospective observation study, we analyzed the association of bacterial DNA load with clinical features, disease severity, and genotype using real-time PCR targeting the 56 kDa TSA gene of O. tsutsugamushi in the blood samples of 117 surviving patients with scrub typhus who had not received appropriate antibiotic treatment. The median O. tsutsugamushi DNA load was 3.11×103 copies/mL (range, 44 to 3.3×106 copies/mL). The severity of patients was categorized as mild, moderate, and severe based on the number of dysfunctional organs, and no significant difference in O. tsutsugamushi DNA load was found among these groups. Patients infected with the Karp group showed a significantly higher O. tsutsugamushi DNA load than those in the Gilliam (P < 0.05) and TA763 (P < 0.01) groups. Patients belonging to the Li ethnic group showed a significantly higher DNA load than those in the Han ethnic groups. The blood bacterial DNA load of patients showed no significant difference between groups divided by gender, age, with or without eschar, or the season of disease onset. The highest body temperature recorded during fever onset was positively correlated with O. tsutsugamushi DNA load (ρ  =  0.272, P  =  0.022). Correlation analyses indicated that the serum total bilirubin level was positively correlated with O. tsutsugamushi DNA load. In conclusion, the findings in this study demonstrated the association of DNA load of O. tsutsugamushi with the severity and genotype in patients with scrub typhus in Hainan, China.


Introduction
Orientia tsutsugamushi (O. tsutsugamushi) is an obligate intracellular bacterium that causes the mite-borne acute febrile disease scrub typhus, which accounts for up to 20% of all febrile cases in the Asia-Pacific region (the Tsutsugamushi Triangle). The disease was also discovered in the Arabian Peninsula, Chile, and possibly Kenya, suggesting that it may be more widespread than previously thought [1][2][3][4]. The widespread re-emergence of scrub typhus in India, Micronesia, and the Maldives, and the rapid emergence of cases in northern China and South Korea have raised considerable concern in endemic countries [5,6]. The most common clinical features of scrub typhus are fever and headache beginning 7-14 days after inoculation with an infected larval mite. Other symptoms include rash, myalgia, cough, lymphadenopathy, nausea, vomiting, abdominal pain, and eschar at the site of mite bite [7][8][9]. Additionally, multiple organ failures including acute renal failure, acute respiratory distress, meningoencephalitis, myocarditis, septic shock, and disseminated intravascular coagulation have been reported in severe cases [10,11]. The fatality rate is 30-50% in untreated cases and 4% in cases treated with antibiotics [12,13]. O. tsutsugamushi is subdivided into 3 classes: high virulence (Karp, Kato, and KN-3 serotypes), intermediate virulence (Gilliam serotype), and low virulence (Kuroki, Kawasaki, and KN-2 serotypes) [14]. Although scrub typhus is a major public health issue, it remains a neglected disease worldwide [5].
The severity and fatality of scrub typhus show geographic variability [13]. The basis for this variability remains largely unknown and was proposed to be influenced by multiple factors, including differences in host genetic susceptibility and immune response, variability in the virulence of the pathogen, and vector species. Additionally, recent studies have investigated the relationship between O. tsutsugamushi genomic DNA load and clinical characteristics of scrub typhus. The results indicated that the patients with fatal disease had a significantly higher bacterial load than patients with non-fatal disease, and that duration of illness, presence of eschar, and hepatic enzyme levels were positively correlated with the O. tsutsugamushi DNA load [15,16] High genomic DNA load is a marker for the virulence of the pathogen and is associated with disease severity in bacterial infectious diseases. In severe sepsis, the genomic bacterial load has been proposed to be a useful marker for assessing severity, predicting prognosis, and monitoring clinical response [17]. Patients with a high bacterial load of Mycobacterium tuberculosis complex (MTBC) showed a lower rate of conversion to negative by the 8th week of treatment than those with a lower bacterial burden, and the virulence of MTBC isolates in the macrophage lysis model was positively associated with the MTBC DNA load in sputum of patients [18,19]. In meningococcal disease, a higher DNA load of Neisseria meningitidis is associated with an increased risk of death, prolonged hospitalization, and infection with the more virulent serogroup C [20].
Although studies have examined the relationship of DNA load and disease severity for surviving and non-surviving patients with scrub typhus, the association between DNA load and severity in surviving patients has not been investigated [15,16]. In this study, the O. tsutsugamushi genomic load in blood samples of patients with scrub typhus taken at the time of hospital admission was evaluated, and its association with disease severity, O. tsutsugamushi genotypes, and other clinical features was analyzed.

Ethics statement
This study was approved by the Ethics Committee of Hainan Medical University. Written informed consent to participate in this study was obtained from adult patients. For patients younger than 18 years, the written informed consent was obtained from their parent or legal guardian.

Patient recruitment and case definition
Patients presenting at the infectious disease departments with acute undifferentiated fever (FUO) were recruited from 4 hospitals in Hainan Province from July 2018 to November 2021. The hospitals included the First Affiliated Hospital of Hainan Medical University, the Second Affiliated Hospital of Hainan Medical University, the People's Hospital of Haikou City from Haikou, which is the provincial capital, and the People's Hospital of Qiongzhong Li Miao Autonomous County, located at the center of Hainan Province. Patients who fulfilled the following criteria were enrolled in this study: 1) axillary temperature � 37.5˚C accompanied by at least 1 of these symptoms (eschar, skin rash, lymphadenopathy, hepatomegaly, and/or splenomegaly); and 2) history of field exposure within 3 weeks before the onset of symptoms.
Whole blood and serum samples were collected immediately following admission. The samples were transported to the laboratory within 48 h using cold chain transportation and stored at -80˚C before testing. The clinical characteristics and demographic features of the enrolled patients were also collected.

Nucleotide sequence and phylogenetic analyses
Genomic DNA was extracted from whole blood samples using a QIAamp DNA Mini Kit (QIAGEN, Hilden, Germany) according to the manufacturer's instructions. Nested PCR was performed using primers targeting a 483bp fragment of the 56-kDa TSA gene as previously reported [23]]. The amplified product of nested PCR was purified using a QIAquick gel extraction kit (Qiagen) and sent to Sangon Biotech (Sangon, Shanghai) for Sanger sequencing with a 3730×1 DNA analyzer (Applied Biosystems, Foster, CA, USA) in both forward and reverse directions. The 2 obtained sequences of the 56 kDa TSA gene from the same samples were aligned using DNASTAR SeqMan II to obtain a confirmed sequence for further analyses. The resulting sequences (GenBank accession No. ON862382-ON862498) were aligned with the basic local alignment search tool (BLAST) using the National Center for Biotechnology Information (NCBI) database to determine the closest sequences. The newly obtained nucleotide sequences were aligned with their closest related sequences and representative sequences from the worldwide NCBI database using CLUSTAL X software. Phylogenetic analysis was performed using the neighbor-joining method with a bootstrap of 1000.

Statistical analysis
Statistical analyses were performed using SPSS software version 19.0, and figures were generated using GraphPad Prism 6.0. Nonparametric tests (Kruskal-Wallis test) were used to compare the differences between bacterial DNA copies and the indicated groups. The duration of illness and the highest body temperature recorded during fever among the groups with different disease severity were compared by one-way analysis of variance (ANOVA) tests. Statistical significance was set as P < 0.05. Correlation analyses were performed using Spearman correlation.

Relationship between O. tsutsugamushi DNA load and genotype
Phylogenic analysis was performed based on truncated 56 kDa TSA protein gene sequences amplified by nested PCR (S1 Fig). The results indicated that among the 117 cases, 64 were closely related to the Karp group, with sub-genotypes A, B, and C contributing to 29, 15, and  (Fig 2A). No significant difference was observed between the Gilliam and TA763 (P = 0.07) groups (Fig 2A). The DNA loads of the samples between sub-genotypes were further analyzed, and there was no significant difference between sub-genotypes within the Karp (P = 0.21) or Gilliam groups (P = 0.059) (Fig 2B and 2C).  Among the patients, 43% were 41-60 years old, which was the largest age group, followed by 41% of patients aged > 61 years. Eighty-three of the cases belonged to the Han Chinese ethnic group, 21 belonged to the Li ethnic group. The patients were grouped according to sex, age, and ethnicity, and the DNA load was compared among the groups (Fig 3A, 3B and 3C). The DNA load of patients in the Li group was significantly higher than that of the Han ethnic groups (median DNA load of 8145 copies/mL versus 2177 copies/mL, P = 0.018) (Fig 3C).
There was no significant difference in the O. tsutsugamushi DNA load among patients of different sexes (P = 0.48) and ages (P = 0.60) (Fig 3A and 3B).

The association between O. tsutsugamushi DNA load and clinical characteristics
Detailed medical documentation was available for 77 patients with positive RT-PCR, based on which these cases were categorized as mild (N = 54), moderate (N = 15), or severe (N = 8 .41] copies/mL), these differences were not statistically significant (P = 0.11 and P = 0.21, respectively) ( Fig 4A). The duration of illness and the highest body temperature recorded during fever was also compared and no significant difference was found among the three groups with different disease severity. In addition, we compared the differences in DNA load between patients with (N = 44) and without eschars (N = 67), and the results showed no significant differences between the 2 groups ( Fig 4B). Patients were divided into 4 groups based on the season of disease onset: spring (March-May), summer (June-August), autumn (September-November), and winter (December-February). A comparison of the O. tsutsugamushi DNA load in patients indicated that there was no significant difference among the 4 groups (P = 0.09) (Fig 4C). The correlation between DNA load and duration of illness (N = 117) and the highest body temperature recorded during fever (N = 77) was also evaluated using correlation analyses. The results indicated that only the highest body temperature (ρ = 0.272, P = 0.022) was positively correlated with O. tsutsugamushi DNA load ( Table 2).  (Table 2).

Discussion
High bacterial load is an indicator of disease severity in many infections; however, the association between O. tsutsugamushi DNA load and disease severity has not been widely investigated. In this study, we analyzed the association of DNA load with disease severity, O. tsutsugamushi genotype and clinical factors in 117 patients who had not received appropriate antibiotic treatment. Our results indicated that 95.07% of nested PCR positive patients had detectable bacterial loads in real-time PCR assays. Several real-time PCR assays have been   [15,16,26]. In this study, we developed a real-time PCR assay for the quantification of O. tsutsugamushi targeting the 56-kDa TSA gene of O. tsutsugamushi, and the assay showed good concordance with the nested PCR assay. The median O. tsutsugamushi DNA load in this study was 3.11×10 3 copies/mL, ranging between 44 and 3.2×10 6 copies/mL. However, this load varies among studies; Singhsilarak et al. reported a median number of 4.2×10 6 copies/mL (range, 1×10 6 -2.88 ×10 7 copies/mL) using real-time PCR targeting the 47 kDa gene in 7 patients with scrub typhus and positive RT-PCR, whereas Kim et al. reported a median number of 7.8×10 4 copies/mL (range 3,960) in surviving patients, and a median number of 8.38×10 7 copies/mL (range 244,600) in 3 non-surviving patients using the same primers and probes [15,16]. Sonthayanon et al reported a median number of 284 copies/mL of O. tsutsugamushi DNA (range, 124-943 copies/mL) in 81 patients with positive quantitative RT-PCR results targeting 16 S rRNA [15]. In addition to population, chigger species, and O. tsutsugamushi strains, variation in bacterial loads was proposed to be partially affected by the target genes selected for the RT-PCR assay and inclusion criteria of these studies [15].
The mortality rate of scrub typhus varies greatly among regions studied, and the cause of this variation remains largely unknown; however, the distribution of O. tsutsugamushi genotypes was proposed to play an important role in this variation [16]. Karp and closely related strains are found in most endemic countries, including Japan, Korea, China, Thailand and Southeast Asia [12,13,27]. The dominance of sub-genotypes varies among countries: Boryong is predominant in South Korea, Karp_C in Japan, and Karp_A in Taiwan, Thailand and Cambodia [13,27]. Several sub-genotypes of the Gilliam group are also prevalent in endemic countries: the Kawasaki genotype is prevalent in South Korea, and the JG-C genotype is primarily found in Taiwan, Thailand and Cambodia [27]. Isolates belonging to the TA763 and Kato groups were mainly reported in Taiwan [27]. The characteristics of the genetic components of each genotype can influence the ability of O. tsutsugamushi to invade the host, grow in large numbers, and disseminate. The severity and clinical presentation of scrub typhus were found to be strain-dependent in both humans and laboratory animals [7,28]. Differences in clinical features between prototypes of O. tsutsugamushi have also been reported. Patients infected by the Boryoung cluster had significantly more generalized weakness, eschars, skin rashes, conjunctival injection, high albumin levels, and greater ESR and fibrinogen levels than those infected with the Karp cluster; however, the treatment response to current antibiotics was significantly slower in the Karp cluster than in the Boryoung cluster [14]. Additionally, the Karp genotype was proposed as a severe genotype. Consistently, our results indicate that Karp is correlated with the severe clinical characteristics of patients. The DNA load of the Karp genotype was significantly higher than that of the Gilliam and TA763 genotypes, indicating a more efficient propagation of O. tsutsugamushi and a slower immune clearance in infections with the Karp genotype. Similar results were reported in C57BL/6NJcl mice showing a higher bacterial load of the Karp genotype compared to the less virulent strain UT176 [29].
Genetic differences in host populations may influence susceptibility and responses to O. tsutsugamushi infection and are proposed to be another factor that influences DNA load. Single-nucleotide polymorphisms (SNPs) in TLR2, TLR4, and HSP70 genes, which lead to variations in immune response and alter susceptibility to various infectious diseases, were analyzed in the Tamil-speaking Dravidian ethnic group in India, and the prevalence of the TLR4D299G SNP was found to be significantly higher in patients with scrub typhus than in controls [30]. A genome-wide association study (GWAS) in Korean patients with scrub typhus identified 8 potent scrub typhus-related SNPs located on PRMT6, PLGLB2, DTWD2, BATF, JDP2, ONECUT1, WDR72, KLK, MAP3K7, and TGFBR2 genes [31]. A study in mice using the Gilliam strain found that O. tsutsugamushi infections showed significantly different degrees of virulence due to genetic differences between mouse strains [32]. Additionally, a comparison of lethal and nonlethal mouse models of O. tsutsugamushi infection revealed differences in CD4 + and CD8+ T-cell population-associated cytokines between the 2 groups [33]. Notably, in the present study, patients in the Li ethnic group showed significantly higher DNA load than the Han groups. To verify whether the difference was caused by the infecting O. tsutsugamushi genotypes, the composition of the genotypes in Han and Li ethnic groups was analyzed. The main genotypes of Han ethnic groups (N = 51) were Karp (N = 30, 59%), Gilliam (N = 18, 35%), and TA763 (N = 3, 6%), while those of Li ethnic groups (N = 21) were Karp (N = 13, 62%) and Gilliam (N = 8, 38%). Statistical analysis showed that the genotypes of the two ethnic groups were not significantly different. Additionally, 95% (20/21) of the Li ethnic patients were recruited in the Qiongzhong Li Miao Autonomous County. The county is located at the center of Hainan Island and most of the area is mountainous with forests. Most patients of the area are forest workers or rubber workers. In contrast, patients recruited in Haikou were mostly from the coastal cities and the vegetation varied from Qiongzhong. Patients from Haikou also had various occupations including farmers and factory workers. These differences in ecological and behavioral factors may also contribute to the difference in O. tsutsugamushi DNA load between patients of Li and Han ethnic groups. It would be interesting to assess the existence of SNPs in these patients that might explain the variations in the immune response and susceptibility to O. tsutsugamushi infection.
Sonthayanon et al. reported that hepatic enzymes are positively correlated with O. tsutsugamushi DNA load in patients recruited in northeast Thailand [15]. In contrast, we found no significant difference in O. tsutsugamushi DNA load between the severe, moderate, and mild groups. This is consistent with the finding of no correlation between hematological indicators and DNA load, with the exception of TBil being positively related to DNA load, which is indicative of hepatic impairment. Eschar is a typical clinical sign of scrub typhus, and the prevalence of eschar in patients has been reported to be highly variable (7%-80%) and largely dependent on the O. tsutsugamushi genotype [14,34]. The presence of eschar was reported to be correlated with DNA load in the study by Sonthayanon et al.; however, although eschar was observed in 40.5% (45/111) of patients with physical examination data available in this study, no significant difference was found between patients with and without eschar [15]. The differences in the correlation analyses of DNA load with clinical characteristics between this study and that by Sonthayanon et al. may be due to multiple factors. The population in the study reported by Sonthayanon et al. in northeast Thailand was predominantly Thai, whereas our study population was predominantly Han and Li ethnic groups. Additionally, the differences could be related to the chigger. The main vectors of scrub typhus in Thailand were reported to be Leptotrombidium imphalum, L. deliense, and L. chiangraiensis. In contrast, although the vectors in Hainan Island were not studied widely, the mainly reported vector was L. deliense [35,36]. Bitten by different species of chiggers could result in the difference in clinical characteristics in both human and mice, which may also related to O. tsutsugamushi DNA load [37,38].
The major limitation of the study was that a small number of cases with scrub typhus were include. In addition, only 77 patients had detailed medical documentation. Another limitation of the study was that only some of the cases had a detailed blood profile comprising all the indicators shown in Table 2, as the clinical test items varied among patients with different symptoms. This may have led to bias in the results obtained from correlation analyses. A larger sample cohort with more clinical data is recommended in future studies to explore the relationship between O. tsutsugamushi DNA load and clinical characteristics in patients with scrub typhus.

Conclusion
In conclusion, our study demonstrated that genomic DNA load of O. tsutsugamushi was associated with its genotypes, ethnic group of the patients, the highest body temperature recorded during fever onset and the total bilirubin (TBil) in the blood. However, no significant difference in O. tsutsugamushi DNA load was found when comparing mild, moderate, and severe patient groups nor when comparing the groups of patients with or without eschar. This research contributes to the understanding of the pathogenic mechanism of O. tsutsugamushi.