OUP user menu

Characterization of Orientia tsutsugamushi isolated in Taiwan by immunofluorescence and restriction fragment length polymorphism analyses

Akira Tamura, Norio Ohashi, Yoko Koyama, Masahiro Fukuhara, Fumihiko Kawamori, Masamitsu Otsuru, Ping-Fuai Wu, Shen-Yu Lin
DOI: http://dx.doi.org/10.1111/j.1574-6968.1997.tb10374.x 225-231 First published online: 1 May 1997


A total of 10 strains of Orientia tsutsugamushi were isolated from field rodents and chiggers in Taiwan, and characterized by immunofluorescence analysis with monoclonal antibodies and by restriction fragment length polymorphism analysis of the 56-kilodalton (kDa) protein gene. The isolates were divided into two groups consisting of 1 and 9 strains which showed some relation to Gilliam and Karp type strains, respectively. However, all these isolates possessed characteristics distinct not only from those of known prototype strains including Gilliam and Karp but also from all isolates from Japan. These findings suggest the existence of a large number of immunotypic and genotypic variants among the strains of O. tsutsugamushi, and the distribution of distinguishable strains in each area to which this species is endemic.

  • Orientia tsutsugamushi
  • Antigenic subtype
  • Phenotyping

1 Introduction

Orientia tsutsugamushi which is transmitted to man by chiggers is the causative agent of scrub typhus, also known as tsutsugamushi disease. Recently, this microbe was transferred from the genus Rickettsia to the Orientia gen. nov. as Orientia tsutsugamushi, because it has characteristics different from those of other species of the genus Rickettsia[1]. Several antigenic types have been recognized in O. tsutsugamushi. Shishido [2] first described antigenic diversity among Gilliam, Karp, and Kato strains, and thereafter we found other types such as Shimokoshi [3], Kawasaki [4], and Kuroki [5, 6] strains, which were isolated from patients in Japan. These antigenic variations depend on antigenic diversity of the 56-kilodalton (kDa) major surface protein [7], and the protein genes of each antigenic prototype strain were cloned and sequenced in our laboratory [8, 9] and by Stover et al. [10]. Our previous study [11] demonstrated that the four antigenic types, Gilliam, Karp, Kawasaki and Kuroki isolated in Japan can be divided into 12 subtypes according to the reactivities with type-specific monoclonal antibodies (MAbs) reactive to the 56-kDa protein on immunofluorescence (IF) analysis, and to restriction fragment length polymorphism (RFLP) of the 56-kDa protein gene.

In the present study, 10 strains of O. tsutsugamushi were successfully isolated from field rodents and fed chiggers in Taiwan, and their reactivities with the same MAbs used previously [11] and RFLP patterns were examined. We will describe here the existence of Taiwan-specific types distinct from those found in Japan.

2 Materials and methods

2.1 O. tsutsugamushi and cells

Laboratory strains of Gilliam and Karp, and four isolates from Japan, 436T, LP-1, Matsuzawa and 402I, were used for comparison to isolates from Taiwan. All these strains including new isolates from Taiwan were propagated in monolayers of L929 cells as described previously [3].

2.2 Isolation of O. tsutsugamushi

O. tsutsugamushi was isolated essentially as reported previously [12]. Briefly, 29 wild rodents captured between 1986 and 1990 in Taiwan were killed, and the liver and spleen of each rodent were triturated and emulsified in 1.5 ml of SPG buffer (3.8 mM KH2PO4, 7.2 mM K2HPO4, 4.9 mM l-glutamic acid, 218 mM sucrose, pH 7.0) containing 200 U of penicillin G per ml in a mortar. Aliquots of 0.5 ml of each preparation were inoculated intraperitoneally into both immunosuppressant cyclophosphamide (CY)-treated and -untreated female ddY mice (5–8 week ages). For isolation from trombiculid chiggers, fed Leptotrombidium deliense collected from wild rodents were separated into groups of five, and the chiggers in each group were ground in SPG buffer with a tissue grinder. The homogenates were then centrifuged at 200 g for 5 min, and the supernatants were intraperitoneally injected into male ICR mice. After two weeks, the mice were killed, and smears of the peritoneum were examined by indirect IF test. Finally, 10 strains, designated as t1–t10, were isolated from rodents or pools of chiggers (Table 1). For preparation of IF antigens and DNAs of these isolates, all stocked spleens of infected mice were passaged once in CY-treated ddY mice, and then the isolates in the homogenate of peritoneal cells were propagated in L929 cell monolayers.

View this table:

Strains of O. tsutsugamushi isolated from field rodents and trombiculid chiggers in Taiwan

Isolated strainsYears of capturePlacesRodentsOrgans or fed chiggers for isolation
t11990Hungtou in Lan-Yu islandRattus norvegicusliver and spleen
t51990Langtao in Lan-Yu islandRattus rattusLeptotrombidium deliense
t61990Yehyin in Lan-Yu islandidemidem
t71990Yehyu in Lan-Yu islandRattus norvegicusliver and spleen
t91986Hsi-Yu in Yuweng island in Penghu island groupMus musculusidem
t101986Chengkung in Taitung PrefectureRattus rattusidem

2.3 MAbs and IF test

The MAbs used and method for indirect IF test were the same as those described previously [1116]. All the MAbs were shown to recognize the 56-kDa major surface protein and react with homologous strains with IF titers of 6,400–409,600. The MAbs were used at dilutions of 100- and 1,000-fold in the present study for the reactivity tests.

2.4 Preparation of DNA and polymerase chain reaction (PCR)-RFLP

DNA preparation and PCR-RFLP were performed as described previously [11]. Primers for PCR were as follows: forward primer A 5′-AAATTATGTTAATTGCTAGTG-CAATGTCTG-3′, corresponding to the sequence of 9–38 bp downstream from the initiation codon of the 56-kDa protein gene of the Gilliam strain [8]; and reverse primer B 5′-CTAGAAG-TTATAGCGTACACCTGCACTTGC-3′, corresponding to 1,538 to 1,567 bp. The PCR mixtures consisted of sample DNA, 200 mM each dNTP, 100 nM each primer A and B, 50 mM KCl, 1.5 mM MgCl2, 10 mM Tris (pH 8.3), and 1.25 units of Taq polymerase in a total volume of 50 μl. PCR was performed for 35 cycles of 94°C for 30 s, 42°C for 1 min, and 71°C for 2 min. For RFLP analysis, the PCR product was digested with each of HinfI, HhaI, HaeIII, DdeI, and PstI, and the digests were electrophoresed in 2.5% agarose gels. λ/HindIII and φX 174/HaeIII digests obtained from Toyobo Co., Tokyo were used as DNA size markers.

3 Results and discussion

The results of characterization of O. tsutsugamushi isolates from Taiwan by IF analysis and PCR-RFLP are shown in Table 2 and Fig. 1, respectively. As shown in Table 2, 9 isolates from Taiwan, designated as strains t1–t9, showed reactivity with 1, 2, or 3 anti-Karp MAbs. However, the prototype Karp strain, and 2 Karp subtype strains which were isolated in Japan and characterized as JP-1 and JP-2 subtypes in our previous study [11], reacted only with 3 anti-Karp MAbs, but t1–t9 strains isolated in Taiwan cross-reacted with some MAbs against other antigenic type strains. Similarly, strain t10 isolated in Taiwan was considered to be closely related to the Gilliam type strain, but the pattern of cross-reactivity with these MAbs in Table 2 was different from that of the prototype Gilliam strain and also from those of 2 Gilliam type strains which were isolated in Japan and characterized as JG-1 and JG-2 subtypes in our previous study [11]. Thus, all isolates from Taiwan were distinguished not only from the prototype Karp and Gilliam strains but also from the other known types.

View this table:

Cross-reactivities of O. tsutsugamushi isolates from Taiwan with MAbs by indirect IF analysis

Strains(MAbs) Anti-GilliamAnti-KarpAnti-KatoAnti-KawasakiAnti- KurokiClassificationa
t110010001000TWP I
t31001000100010001000TWP I and V
t410010001000TWP II
t510010001000100010001000TWP II and VI
t61000100100010001000TWP III
t81001000100010001000TWP IV
t1010001000100010001000100TWG I
  • 1000, positive at antibody dilution of 1000-fold; 100, positive at dilution of 100-fold; −, negative at dilution of 100-fold. a See text. b Data are cited from our previous report [11].

Figure 1

PCR-RFLP profiles of 56-kDa protein genes of isolates after digestion with endonucleases Hinf I, Hha I, Hae III, Dde I, and Pst I. The digests were electrophoresed in agarose gels and the bands were stained with ethidium bromide. λ and φ: DNA size markers. NT: PCR product of Gilliam strain (1,568 bp) without digestion with endonucleases. The other letters above each column indicate the strains of O. tsutsugamushi. Kp: Karp strain; p1: Matsuzawa; p2: 402I; t1–t10: isolates in Taiwan; G: Gilliam strain; g1: 436T; and L1: LP-2. The characters and Roman numerals at the bottom of each column indicate type of each strain as described in the text. JP and JG: Karp and Gilliam subtypes, respectively, isolated in Japan, TWP and TWG: type designation of isolates in Taiwan (see text). The numbers at the left side show the sizes in bp.

Furthermore, the t1–t9 isolates from Taiwan showed different cross-reactivity patterns with these MAbs, and the patterns were divided into six types as follows: pattern 1 of t1 and t2, pattern 2 of t3, pattern 3 of t4, pattern 4 of t5, pattern 5 of t6 and t7, and pattern 6 of t8 and t9.

In RFLP analysis shown in Fig. 1, the pairs of strains t1 and t2, t6 and t7, and t8 and t9 showed identical patterns on digestion with all endonucleases, indicating that these were pairs of the same type. Summation of molecular sizes of bands in each column, except for t3 and t5, gave a total size of about 1.5 kb which corresponded to that of the amplified PCR product, but summation of the bands from t3 and t5 after HaeIII digestion gave a total of about 3 kb, suggesting the presence of two different type strains in the t3 and t5 preparations. From comparison of RFLP profiles, t3 may be a mixture of t1–t2 type and another strain, and t5 may contain t4 and another type.

The RFLP patterns shown in Fig. 1 were compared with those of other type strains reported in our previous paper [11] and the patterns of t1, t2 and t3 HinfI digests, t4, t6 and t7 HhaI digests, t4, t6 and t7 PstI digests, and t6 and t7 HaeIII digests were identical with the corresponding enzyme digest patterns of prototype Karp and/or two Japanese Karp type strains of JP-1 and JP-2. Furthermore, DdeI digest patterns of t1, t2, t3, t4, t6 and t7 were all identical, and were similar but not completely identical to the DdeI digest patterns of Karp and JP-1 strains. The HinfI and HhaI digest patterns of Kuroki strain (refer our previous report [11]) were similar but not completely identical to the HinfI patterns of t1, t2 and t3, and HhaI patterns of t4, t6 and t7, respectively. In all cases, the RFLP patterns could be distinguished from other types such as Gilliam, Kato, Kawasaki, Kuroki and Shimokoshi. Strains t8 and t9 showed specific patterns different from those of other strains on RFLP analyses, but these strains reacted with 3 anti-Karp specific MAbs. The isolate of t5 was considered to be a mixture of more than two types as mentioned above, but this strain also reacted with all three anti-Karp Mabs. These observations indicate that the t1–t9 isolates from Taiwan were not identical to Karp or Karp type strains isolated in Japan, but are more closely related to Karp type strains than to other type strains. From the observations described above, t1 and t2 types were designated as belonging to type TWP1 (Taiwan Karp-related type 1), t4 as TWP2, t6 and t7 as TWP3, and t8 and t9 as TWP4. Furthermore, t3 may be a mixed population of TWP1 and an additional subtype designated as TWP5, and t5 may also be a mixture of TWP4 and another type, TWP6. Isolate t10 was designated as TWG1 type.

This classification into types based on RFLP patterns corresponded well with the six different cross-reactivity patterns with MAbs shown in Table 2. TWP1 type (t1 and t2) reacted with one anti-Karp 3C9 MAb and cross-reacted with anti-Kato 5D-3 and weakly with anti-Gilliam 1–18, TWP2 type (t4) reacted with two anti-Karp 1–19 and KP-7 MAbs and also weakly with anti-Gilliam 1–18 MAb, TWP3 (t6 and t7) reacted with two anti-Karp 1–19 and 3C9 MAbs and cross-reacted with anti-Gilliam 4–9 and 1–18 and also anti-Kato 5D-3 MAbs, TWP4 (t8 and t9) reacted with 3 anti-Karp MAbs and showed cross-reactivity with 2 anti-Gilliam 4–9 and 1–18 MAbs. Isolate t3, considered to be a mixed population of TWP1 and TWP5, showed reactivities with two anti-Karp Mabs in addition to those of TWP1, and isolate t5 which was thought to be a mixture of TWP2 and TWP6 reacted with anti-Karp 3C9, anti-Kuroki 55–12 and MAbs reactive with TWP2.

Fig. 2 shows the geographical areas in Taiwan from which these types were isolated. All TWP 1 to 6 types were detected in the Lan-Yu area which is a small island in the Pacific Ocean, and TWP4 was also detected in the Penghu island group. TWG1 was isolated from the Taiwan-Chengkung area.

Figure 2

Geographical areas in Taiwan from which types of O. tsutsugamushi were isolated.

In 1915, tsutsugamushi disease in Taiwan was first reported by Hatori [17], and thereafter more cases of this disease were found in different parts of this country [18]. In 1951, cases of the disease were reported in Penghu [19], and in 1970 an outbreak occurred in eastern Taiwan [20]. Recently, 58 cases in 1993, 103 cases in 1994 and 140 cases in 1995 were confirmed by either indirect IF antibody tests or isolation of organisms in Taiwan. Most cases of tsutsugamushi disease were reported in isolated islands such as Kinmen, Lienchiang, Lan-Yu, Penghu and mountain areas of Taiwan, and most cases occurred from June to October [21].

In this study, eight strains were isolated from rodents and chiggers from Lan-Yu island, suggesting a high prevalence of O. tsutsugamushi infection on this island. Indeed, two cases of infection of Taipei residents after visiting the Lan-Yu island were reported in 1985 [22]. Serum antibody positive rates to O. tsutsugamushi among the residents of this island were reported to be 80.6% in 1985 and 96% in 1988 [23, 24].

In conclusion, we demonstrated the existence of several Taiwan-specific types of O. tsutsugamushi by phenotyping with type-specific MAbs and genotyping by RFLP analysis. These findings suggest that new types or even new antigenic types may be present in other countries. The analytical procedures used here may be very effective for characterization of newly isolated O. tsutsugamushi.


This work was supported in part by grants-in-aid from the Ministry of Education, Science, and Culture, Japan, to N.O.(04770269, 05770198, 06770209), to A.T. (03454185, 06670307), and from the Uehara Memorial Foundation, the Sasakawa Scientific Research Grant from the Japan Science Society in 1993, and Ageo Chikuen Foundation in 1994 to N.O.


View Abstract