OUP user menu

Genetic typing of the 56-kDa type-specific antigen gene of contemporary Orientia tsutsugamushi isolates causing human scrub typhus at two sites in north-eastern and western Thailand

Stuart D. Blacksell, Rungnapa Luksameetanasan, Thareerat Kalambaheti, Nuntipa Aukkanit, Daniel H. Paris, Rose McGready, François Nosten, Sharon J. Peacock, Nicholas P.J. Day
DOI: http://dx.doi.org/10.1111/j.1574-695X.2007.00375.x 335-342 First published online: 1 April 2008


Orientia tsutsugamushi is the causative agent of scrub typhus, a major cause of febrile illness in the rural areas of Southeast Asia. Twenty-three strains of O. tsutsugamushi were isolated from patients with scrub typhus in north-east (Udorn Thani province) and western Thailand (Tak province) between 2003 and 2005. The isolates were characterized by sequencing the entire ORF of the 56-kDa-type-specific antigen gene, followed by phylogenetic analysis. The majority (15/23) of isolates clustered with the Karp-type strain, six with a Gilliam-type strain and one each with the TA716- and TA763-type strains. Overall, there was considerable diversity in sequence, comparable to that seen in strains from across the rest of the scrub typhus-endemic world. There was no significant difference in the distributions of strains between the two provinces (P=0.08, Fisher's exact) nor a temporal change in distribution with year of isolation (P=0.80, Fisher's exact). Within this diversity there were also examples of isolates with identical 56-kDa genotypes that were cultured from patients from the same geographical areas.

  • Orientia tsutsugamushi
  • scrub typhus
  • 56-kDa
  • senotype
  • phylogenetic
  • Thailand


Orientia tsutsugamushi, formerly known as Rickettsia tsutsugamushi, is the cause of scrub typhus, a major cause of febrile illness in rural Southeast Asia (Leelarasamee et al., 2004; Phongmany et al., 2006; Suttinont et al., 2006). Infection is largely confined to the Asia-Pacific region, where the pathogen is vertically maintained in the Leptotrombidium spp. mite population and transmitted to humans by the bite of the larval stage (Watt & Parola, 2003). No vaccine is currently available, but antibiotic therapy with either doxycycline or azithromycin achieves an effective cure.

The 56-kDa type-specific antigen (TSA) of O. tsutsugamushi is located on the outer membrane surface, is the primary immunogen, and is responsible for eliciting neutralizing antibodies (Hanson, 1985; Tamura et al., 1985; Ohashi et al., 1989; Stover et al., 1990; Seong et al., 1997, 2000). The gene encoding the 56-kDa TSA has an ORF of c. 1600 bp (Ohashi et al., 1989, 1992; Stover et al., 1990). Orientia tsutsugamushi isolates are conventionally classified on the basis of reactivity with hyperimmune serum raised against prototype strains (e.g. Karp, Kato, and Gilliam), and the four hypervariable regions within the 56-kDa TSA are considered to play a significant role in type strain assignment.

Genetic and antigenic characterization of contemporary O. tsutsugamushi isolates causing human disease in scrub typhus-endemic regions is limited. A major contributing factor is that isolation involves growth in tissue culture lines within a BSL3 facility. However, variability of the 56-kDa TSA gene and its product within the population of natural isolates could have a major bearing on both the accuracy of diagnostic tests and vaccine development. Previous studies have noted a dominance of Karp-like strains in human and ecological studies in Thailand (Elisberg et al., 1968; Shirai et al., 1981; Kollars et al., 2003; Manosroi et al., 2006). The aim of this study was to determine the dominant contemporary strains and variability of the 56-kDa gene from O. tsutsugamushi isolated from scrub typhus patients at two sites in north-east and western Thailand to determine (1) distribution of strains compared with previous studies, (2) genetic relatedness of strains and (3) comparison with type strains from other countries.

Materials and methods

Patient specimens and isolates

Whole blood samples were collected from scrub typhus patients in Udon Thani (535 km north-east of Bangkok) and Tak (512 km north-west of Bangkok) provinces between September 2003 and August 2005. Samples were collected as part of studies investigating the causes of fever at these sites. Ethical clearance was obtained from the Faculty of Tropical Medicine, Mahidol University (Tak), the Thai Ministry of Public Health (Udon Thani) and the Oxford Tropical Research Ethics Committee (both studies). Patients provided informed written consent before sample collection. Twenty-three O. tsutsugamushi isolates were grown in vitro (Table 1) (Luksameetanasan et al., 2007). Nineteen isolates (19/23; 83%) were from Udon Thani patients and four (4/23; 17%) isolates were from Tak patients.

View this table:
Table 1

Description of Orientia tsutsugamushi isolates and reference strains examined in this study

IsolateMonth/ yearSourceDistrictProvince/ prefectureCountryBasesComplete or partial gene sequenceGenBank accession numberStrainIsolation/56-kDa gene references
UT7609/2003HumanMuangUdorn ThaniThailand1611CompleteEF213078KarpLuksameetanasan et al., (2007). This publication
UT12510/2003HumanMuangUdorn ThaniThailand1596CompleteEF213096GilliamLuksameetanasan et al., (2007). This publication
UT14406/2004HumanMuangUdorn ThaniThailand1596CompleteEF213091GilliamLuksameetanasan et al., (2007). This publication
UT15006/2004HumanMuangUdorn ThaniThailand1611CompleteEF213086KarpLuksameetanasan et al., (2007). This publication
UT16706/2004HumanPhenUdorn ThaniThailand1611CompleteEF213080KarpLuksameetanasan et al., (2007). This publication
UT16906/2004HumanMuangUdorn ThaniThailand1608CompleteEF213092KarpLuksameetanasan et al., (2007). This publication
UT17607/2004HumanBan PhuUdorn ThaniThailand1602CompleteEF213081KarpLuksameetanasan et al., (2007). This publication
UT17707/2004HumanMuangUdorn ThaniThailand1605CompleteEF213084KarpLuksameetanasan et al., (2007). This publication
UT19607/2004HumanMuangUdorn ThaniThailand1596CompleteEF213079GilliamLuksameetanasan et al., (2007). This publication
UT21307/2004HumanSang KhomUdorn ThaniThailand1611CompleteEF213088KarpLuksameetanasan et al., (2007). This publication
UT21907/2004HumanMuangUdorn ThaniThailand1611CompleteEF213100KarpLuksameetanasan et al., (2007). This publication
UT22108/2004HumanMuangUdorn ThaniThailand1614CompleteEF213097KarpLuksameetanasan et al., (2007). This publication
UT30208/2004HumanMuangUdorn ThaniThailand1587CompleteEF213095TA763Luksameetanasan et al., (2007). This publication
UT31610/2004HumanMuangUdorn ThaniThailand1611CompleteEF213082KarpLuksameetanasan et al., (2007). This publication
FPW201605/2004HumanPho PraTakThailand1608CompleteEF213085GilliamLuksameetanasan et al., (2007). This publication
FPW103810/2004HumanMae RamatTakThailand1593CompleteEF213087TA716Luksameetanasan et al., (2007). This publication
FPW203112/2004HumanPho PraTakThailand1614CompleteEF213098KarpLuksameetanasan et al., (2007). This publication
UT32907/2005HumanNa YangUdorn ThaniThailand1596CompleteEF213099GilliamLuksameetanasan et al., (2007). This publication
UT33207/2005HumanMuangUdorn ThaniThailand1611CompleteEF213083KarpLuksameetanasan et al., (2007). This publication
UT33607/2005HumanWang SamUdorn ThaniThailand1599CompleteEF213089KarpLuksameetanasan et al., (2007). This publication
UT39507/2005HumanMuangUdorn ThaniThailand1611CompleteEF213094KarpLuksameetanasan et al., (2007). This publication
FPW204907/2005HumanPho PraTakThailand1596CompleteEF213093GilliamLuksameetanasan et al., (2007). This publication
UT41808/2005HumanMuangUdorn ThaniThailand1605CompleteEF213090KarpLuksameetanasan et al., (2007). This publication
Gilliam1943HumanBurma1575CompleteGilliamOhashi et al., (1992)
Kawasaki1981HumanMiyazakiJapan1569CompleteM63383KawasakiOhashi et al., (1992)
Kato1955HumanNiigataJapan1590CompleteM63382KatoOhashi et al., (1992)
Karp1943HumanNew Guinea1599CompleteM33004KarpStover et al., (1990)
Boryong1998HumanSouth Korea1602CompleteL04956KurokiKim et al., (1993)
ShimogoshiJapan1566CompleteM63381ShimogoshiOhashi et al., (1992)
TA6861963Tupaia glisThailand1599CompleteU80635TA686Genbank submission
TA7631963Rattus rajahThailand1581CompleteU80636TA763Genbank submission
TA6781963Rattus rattusThailand1548CompleteU19904TA678Genbank submission
TA7161963Menetes berdmoreiThailand1575CompleteU19905TA716Genbank submission
MatsuzawaJapan1458PartialAF173043JP1Enatsu et al., (1999)
402IJapan1468PartialAF173047JP2Enatsu et al., (1999)
HirahataLeptotrombidium pallidumAichiJapan1474PartialAF201835JP2Tamura et al., (2001)
HSB11996–97RodentN/ASaitamaJapan1454PartialAF302983SaitamaTamura et al., (2001)
KurokiJapan1599CompleteM63380KurokiOhashi et al., (1992)
IkedaHumnNiigataJapan1424PartialAF173033JGTamura et al., (2001)
TW26-11990Rattus norvegicusLan-Yu IslandTaiwan1448PartialAY222636Karp-likeQiang et al., (2003)
TW73R1999Rattus spp.Kinmen IslandTaiwan1478PartialAY222628Karp-likeQiang et al., (2003)
TWYu8-11990Leptotrombidium pallidumLan-Yu IslandTaiwan1454PartialAY222640Karp-likeQiang et al., (2003)
TW12-11990Rattus norvegicusLan-Yu IslandTaiwan1517PartialAY222639Karp-likeQiang et al., (2003)
TW38-11990Rattus norvegicusLan-Yu IslandTaiwan1412PartialAY222635TA763Qiang et al., (2003)
TWYu1-11990Leptotrombidium pallidumLan-Yu IslandTaiwan1421PartialAY222641TA716Qiang et al., (2003)
TW46-11986Rattus rattusChengkungTaiwan1415PartialAY222631JGQiang et al., (2003)

Genotyping of the 56-kDa TSA gene

Orientia tsutsugamushi-infected VERO cells were harvested at the peak of infection as determined by the indirect immunofluorescence assay (IFA) (Luksameetanasan et al., 2007). Infected cells were scraped into the tissue culture supernatant (TCSN) and centrifuged to pellet the cells. The TCSN was removed by decanting and the cell pellet was washed twice with PBS. Genomic DNA was extracted from the cells using the Wizard SV Genomic DNA purification system (Promega Co.) according to the manufacturer's instructions. DNA was eluted in double-distilled, deionized H2O to a final volume of 200 µL and stored at −20 °C until use.

The complete ORF of 56-kDa TSA gene was amplified by PCR using previously described primers — Q1, Q2 and Q4 (Enatsu et al., 1999; Qiang et al., 2003) — and primers designed specifically for this study — R3 (5′-GCCTAATAGTGCATCTGTCG-3′), R4 (5′-GCCTATAAGTATAGCTGATCG)-3′, R5 (5′-GCTGCTGTGCTTGCTGCG-3′) and R6 (5′-GGCCAAGTTAAACTCTATGC-3′). The three primer sets gave sufficient overlap to enable determination of a contiguous ORF following nucleotide sequencing; Q1–Q2 (position 266–856), R4–R5 (position 454–1056) and R3–Q4 (position 909–1575). R6 was used as a sequencing primer at position 1318 for R3–Q4 amplicon products.

The PCR reaction mixture contained 1 µL extracted DNA, 2.5 µL 10 × PCR buffer (Promega Co.), 0.5 µL deoxynucleotide phosphate (dNTP) (10 mM each), 1.75 µL of each primer (5 µM), 0.25 µL (2.5 U) of Taq polymerase (Promega Co.), and distilled deionized water in a final volume of 25 µL. The cycling parameters used were denaturation at 95 °C for 30 s, annealing at 55 °C for 30 s and extension at 72 °C for 60 s for 30 cycles, with a final extension of 72 °C for 5 min. The PCR product (3 µL) was analyzed by 0.8–1.5% agarose gel electrophoresis in TBE buffer at 100 V. After completion the gel was stained with 2 µg mL−1 ethidium bromide for 5 min and destained in distilled water for 10 min. The DNA band was visualized and photographed under UV light and products were compared to the positive control sample and a known molecular size marker (EZ-load, Biorad Co., UK). Nucleotide sequencing was performed using the MegaBACE Model 1000 automated sequencer (Amersham Bioscience, UK).

Phylogenetic analysis

Multiple sequence alignment and analysis to determine genetic relationships of isolates and reference type strains was performed using bionumerics version 4.6 (Applied Maths, Belgium) and the clustal x program (Thompson et al., 1997). Complete and partial 56-kDa TSA ORF nucleotide sequences for reference and prototype strains were obtained from GenBank (Table 1). The exception was the Gilliam (M33267) 56-kDa gene nucleotide sequence, kindly provided by Dr Hiroshi Urakami, which had been referenced by a previous study (Enatsu et al., 1999). A percentage nucleotide identity matrix was constructed by the PAM250 method using the dnastar megalign 6.1 program. Phylogenetic analysis was performed and the resulting dendrogram was constructed by the mega 3.1 program (Kumar et al., 2004) using UPGMA and neighbor joining algorithms bootstrapped for 1000 replications. Fisher's exact test was performed to determine the significance of the relationship (P<0.05) between the strains of O. tsutsugamushi isolates and their geographical distribution (i.e. Tak and Udon Thani provinces). Fisher's exact test was also performed to determine whether there was significant association (P<0.05) between the strains of O. tsutsugamushi isolates and the annual distribution (i.e. 2003, 2004 and 2005).


Characteristics of contemporary strains of O. tsutsugamushi causing human disease in Thailand

Genetic analysis of the 56-kDa TSA ORF demonstrated that the 23 Thai O. tsutsugamushi isolates obtained during this study were related to either the Karp- or Gilliam-type strains, or to the historical Thai-type strains TA716 and TA763 (Fig. 1 and Table 2). The majority (65%; 15/23) of contemporary Thai O. tsutsugamushi isolates were related to the Karp-type strain [percentage nucleotide identity range (PNIR) with prototype Karp: 93.1–96.1% (Fig. 1 and Table 2)]. Six (26%; 6/23) isolates demonstrated highest similarity to the Gilliam strain cluster (Gilliam PNIR 88.9–91.8%). Isolate UT302 was most closely related to the TA763-type strain (TA763 PNIR: 90.1%) and FPW1038 grouped with type strain TA716 (95.9% identity). Isolates UT144, UT196 and UT125 (all clustering with the Gilliam strain) demonstrated 100% identity, as did UT150, UT167, UT316 and UT332 (in the Karp strain cluster).

Figure 1

Dendrogram representing the genetic relationships of Thai (represented in boldface type) and reference Orientia tsutsugamushi type strains based on the nucleotide sequence of the partial ORF (1412 bp) of the 56-kDa TSA gene demonstrating the relationships between Thai and non-Thai strains. The designations JP1, JP2, JG and Saitama are from the study by Tamura (2001).

View this table:
Table 2

Percentage nucleotide identity of the entire 56-kDa TSA gene ORF of contemporary Thai Orientia tsutsugamushi isolates compared with reference type strains

Reference O. tsutsugamushi type strains
  • Highest identity with reference strains is highlighted.

Geographical and temporal relationships between the contemporary Thai O. tsutsugamushi strains

The majority of O. tsutsugamushi isolates were from Udon Thani province (83%; 19/23). The majority of Udon Thani province isolates were related to the Karp-type strain (74%; 14/19) with a lower proportion of Gilliam-type strain-related isolates (21%; 4/19) and a single isolate that was related to type strain TA763 (5%; 1/19). The Tak province isolates were composed of two Gilliam-like strains (50%), a Karp-like strain (25%) and an isolate clustering with type strain TA716 (25%). There was no significant difference in the distributions of strains (i.e. Karp, Gilliam, TA716 and TA763) between the two provinces (P=0.08, Fisher's exact) nor a temporal change in distribution with year of isolation [2003 (Karp 1, Gilliam 1), 2004 (Karp 10, Gilliam 3) and 2005 (Karp 4, Gilliam 2) (P=0.80, Fisher's exact)].

Relationship between contemporary Thai O. tsutsugamushi isolates and strains in other countries

Of the Karp-related strains, the 56-kDa TSA sequence of Thai isolates was most similar in terms of Taiwanese strains from 1990 (TW26-1, 97.6% identity with UT336; TWYu8-1, 98.7% identity with UT76, 98.8% identity with UT150, UT167, UT316 and UT332) and 1999 (TW73R, 95.8% identity with UT176, 95.5% identity with UT177) and were distinct from the JP1, JP2 and Saitama Japanese Karp strains (Fig. 1). Similarly, within the Gilliam-related strains, the Thai isolates grouped with a 1986 Taiwan strain (TW46-1, 98.8% identity with UT125, UT144 and UT196) and were distinct from the Japanese JG Ikeda strain.


This study demonstrates the diverse nature of contemporary isolates of O. tsutsugamushi causing human disease at locations in north-eastern and western Thailand. Phylogenetic analysis clearly differentiated the isolates into Karp, Gilliam, TA763 and TA716 type strain-associated clusters based on conventional antigenic/genotypic classifications (Enatsu et al., 1999; Qiang et al., 2003; Tay et al., 2005).

This is the first study to genetically characterize O. tsutsugamushi isolates from a consecutive series of patients with scrub typhus infections from a defined geographical area. This is also the first study to genetically characterize the entire O. tsutsugamushi 56-kDa TSA gene ORF from Thai patients, which necessitated the design of new amplification and sequencing primers to amplify all Thai type strains that were isolated in vitro. The 56-kDa TSA gene ORF sequencing results (~1600 bp) presented here provide an improved level of phylogenetic detail than previous genetic studies that only examined selected hypervariable regions of the same gene (300 bp) (Kollars et al., 2003; Manosroi et al., 2006) and serologically based characterization methods, which can be difficult to standardize and require reference serum and antigens. The majority of O. tsutsugamushi isolates examined in this study belonged to Karp- and Gilliam-type strains. This study also describes the first reported cases of human disease caused by strains that are related to TA716- and TA763-type strains, which were originally isolated from small mammals in Thailand in 1963 (Elisberg et al., 1968). Karp-like strains were the dominant group of contemporary Thai isolates in this study, which concurs with a recent study of Thai O. tsutsugamushi samples (Manosroi et al., 2006). Previous antigenic studies of O. tsutsugamushi isolated from trombiculid mites in north-eastern Thailand demonstrated the dominance of Karp-like strains and, to a lesser extent, TA716- and TA763-like strains (Shirai et al., 1981). Interestingly, there were a higher proportion of Gilliam-like strains from Tak province in western Thailand, although this observation should be treated with caution because of the low number of samples analyzed. A previous study from northern Thailand that examined 12 patient isolates suggested no relationship of any members to the Karp strain group, although this was based on a 300-bp sequence of only one of four hypervariable regions of the 56-kDa TSA gene (Kollars et al., 2003). Nucleotide sequencing of three Karp (UT150/167/316) and three Gilliam (UT144/125/196) strain isolates from Udon Thani demonstrated 100% identity and raised the possibility of a common source of infection given the relatively close temporal and geographical relatedness of the patients. Japanese isolates (Mori, Kamimoto and Okazaki) from human scrub typhus cases in Tokushima in 1998 also demonstrated identical nucleotide sequences for 1455 bp (the extent of the sequence submitted to Genbank) of the 56-kDa gene, although this observation was not discussed in the original study (Enatsu et al., 1999). The geographic clustering of isolates and limited genetic variation may be attributable to the vertical transovarial transmission of the O. tsutsugamushi bacterium maintained within the mite vector and the existence of ‘mite islands’ (Audy & Harrison, 1951). The possibility of cross-contamination between cultures in this study was unlikely as the samples were not processed at the same time and stringent biocontainment measures were in place during in vitro propagation. Further evidence that cross-contamination is not the cause of the apparent clonality of the isolates is provided by significant (Coleman et al., 2002) high admission immunofluorescence antibody titers (i.e. ≥1 : 400 IgM) against O. tsutsugamushi in every patient suggesting true scrub typhus infection, and by the high homologous binding titers of patient sera when tested by indirect immunofluorescence (results not presented). The 56-kDa TSA sequences of contemporary Thai O. tsutsugamushi isolates were more similar to those of Taiwanese strains than Japanese strains. However, only a small number of entire 56-kDa TSA nucleotide sequences are currently deposited on genetic databases and further characterization of previously collected and future O. tsutsugamushi isolates from other scrub typhus- endemic countries is required for a complete understanding of the geographical diversity of this important immunogen.

Results from this study provide opportunities to improve serological and molecular diagnosis of scrub typhus infections as well as the raw materials for future studies. Sero-diagnosis for scrub typhus infections is performed by indirect IFA (gold standard), immunoperoxidase (IIP) assay, commercial enzyme-linked immunosorbent assays or rapid immunochromatographic assays that incorporate a mixture of antigen strain types. Strain results presented here demonstrate that antigen pools should contain at least Karp, Gilliam, TA716 and TA763 strain antigens, as well as the Kato strain, which has been previously recognized in Thailand (Shirai et al., 1981; Khuntirat et al., 2003; Manosroi et al., 2006). The nucleotide sequencing results from the entire 56-kDa TSA gene ORF of the studied isolates have been deposited on genetic databases and can be used for the design of improved diagnostic PCR primers. However, further genetic studies of the 56-kDa TSA and other diagnostically important genes are required in Thailand and other scrub typhus-endemic locations to make the information deposited on the databases more representative.


The authors wish to thank Thaksinaporn Thoujaikong and Wilairat Jedsadapanpong for excellent technical assistance. DHP was supported by the Swiss National Science Foundation (PBZHB-106270) and holds a Wellcome Trust Clinical Research Training Fellowship. S.J.P. was funded by a Wellcome Trust Career Development Fellowship.

This study was funded by the Wellcome Trust of Great Britain as part of the Mahidol University–Oxford Tropical Medicine Research Unit.

There are no known conflicts of interest.


  • Editor: Kai Man Kam


View Abstract