INTRODUCTION

Top Abstract Introduction Materials and Methods Results Discussion References

Lyme disease is primarily caused by three genomic species, Borrelia burgdorferi sensu stricto, Borrelia garinii, and Borrelia afzelii (2, 4). B. garinii and B. afzelii are widely distributed from Europe to the Far East including Japan, while B. burgdorferi sensu stricto is prevalent in North America and has been confirmed in part of Europe. B. burgdorferi sensu lato is mainly transmitted by some tick species of the Ixodes ricinus complex, and these ticks infest both mammals and birds (1, 9, 10, 13, 19, 25).

Concerning the prevalence of Borrelia in ticks feeding on birds, B. garinii has been isolated from Ixodes persulcatus on Emberiza spodocephala and Turdus chrysolaus in Hokkaido, Japan (20, 24), and B. burgdorferi sensu stricto, B. garinii, B. afzelii, and Borrelia andersonii have been isolated from Ixodes dentatus, I. ricinus, Ixodes scapularis, Ixodes uriae, and other ticks, which infest a large number of bird species in Europe and North America (6, 13, 15, 27, 28, 30). Furthermore, Haemaphysalis leporispalustris, detected on some bird species, was reported to be a reservoir of B. burgdorferi sensu stricto in North America (13, 26).

B. afzelii is transmitted between I. persulcatus and field rodents, and B. garinii is transmitted between I. persulcatus and migratory birds or rodents, in Hokkaido, Japan (24, 25). However, there has not been a survey of Borrelia in migratory birds which travel directly between the Asiatic continent and Japan. In this survey, we examined the Borrelia prevalence in juvenile ticks removed from birds captured on the Japanese mainland.

	MATERIALS AND METHODS

Top Abstract Introduction Materials and Methods Results Discussion References

Survey site. From September to November 1995 to 1997, surveys were carried out at the Bird Banding Otayama Station, located in the mountainous area (maximum elevation, about 600 m above sea level) in Fukui Prefecture of central Japan along the coast of the Sea of Japan (Fig. 1). Additional surveys were carried out at the same station from late April to early May 1996 to 1997. 

View larger version (18K): [in this window] [in a new window]   FIG. 1.   Survey site (solid circle) in central Japan and migratory routes of T. pallidus (solid line) and E. spodocephala (dashed line) around the Japanese mainland during autumn.

Tick collection. We took part in the bird banding performed by the Yamashina Institute for Ornithology and Fukui Branch of the Wild Bird Society of Japan. Migratory birds were captured using about 40 Japanese mist nets (12 m in length, 36-mm mesh) at ground level. The species of all birds were identified, and, if possible, their sexes and ages were determined. Prior to banding and release, each bird was closely examined, particularly around the head and neck. Ticks were removed with forceps and placed in separate glass vials containing moist sanitary cotton. The stage and species of each tick were identified as described by Takada (34).

Spirochete isolation. The internal organs of live ticks that were insufficiently engorged were immediately removed for spirochete isolation and placed in Barbour-Stoenner-Kelly II (BSK) II medium as described previously (3, 7). Healthy specimens of fully engorged larvae and nymphs of ticks were kept in glass vials containing moist sanitary cotton in a thermostatic chamber at 25°C and then dissected to isolate spirochetes as soon as the ticks molted to the later stages. All cultures were incubated at 32°C and examined for spirochetes by phase-difference microscopy once weekly for 5 weeks. Dead ticks and the birds themselves were not examined for spirochetes.

PCR and RFLP analysis. Spirochete isolates were identified by 5S-23S rRNA intergenic spacer restriction fragment length polymorphism (RFLP) analysis. Primers corresponding to the 3' end of the rRNA (rrf [5'-CTGCGAGTTCGCGGGAGA-3']) and the 5' end of the 23S rRNA (rrl [5'-TCCTAGGCATTCACCATA-3']) as described previously (29) were synthesized using b-cyanoethyl phosphoramidite by a custom oligonucleotide synthesis service (Bex Co., Tokyo, Japan). Two-milliliter aliquots of culture were washed, and the cells were resuspended in 100 ml of water. The resultant cell suspensions were boiled at 100°C for 10 min. PCR was performed by a method previously described (18, 29). The amplicon obtained after PCR was digested with MseI and DraI according to the manufacturer's recommendations (New England Biolabs, Beverly, Mass.), and the digested DNA was electrophoresed through a 16% polyacrylamide gel and subsequently stained with ethidium bromide. Marker 10 purchased from Nippon Gene Co. (Toyama, Japan) was used as a molecular weight marker.

Sequencing of amplified products. Each PCR product was cloned into the pCR II plasmid vector, and the recombinant plasmids were transformed into Escherichia coli INV- F' using a TA cloning kit (Invitrogen Co., San Diego, Calif.) according to the manufacturer's instructions. The recombinant plasmids were extracted from E. coli cultures in Luria-Bertani broth using the Wizard 373 DNA purification system (Promega Co., Madison, Wis.) and sequenced by a dideoxy chain termination method using the dye terminator Taq cycle sequencing kit and a model 373A DNA sequencer (Applied Biosystems Inc., Foster City, Calif.). At least two clones were sequenced for determination of each strain.

Nucleotide sequence accession numbers. The intergenic spacer sequences were assigned the following accession numbers: strain Fi14f, AB015911; strain Fi24f, AB015912.

	RESULTS

Top Abstract Introduction Materials and Methods Results Discussion References

Tick collection. A total of 1,733 birds representing 40 species were examined for ticks in autumn, and 361 ticks were removed from 173 birds of 15 species. All ticks removed were juvenile ticks from six species of two genera: Haemaphysalis flava, 341; Haemaphysalis longicornis, 4; Ixodes columnae, 1; I. persulcatus, 9; Ixodes turdus, 4; unidentified species (related to I. persulcatus with a few morphological differences), 2. Of these, H. flava ticks were detected on 145 birds of 13 species and constituted 94.4% of all tick specimens. The tick prevalence rates on Turdus pallidus and Turdus cardis were 31.7 and 35.1%, respectively. The rate on E. spodocephala was 3.0% (Table 1). I. persulcatus was detected on nine birds of three species. The prevalence rate of I. persulcatus was 2.6% (7 of 271) on T. pallidus (Table 2). In additional surveys in spring, only 16 juvenile ticks (2 of H. flava, 2 of H. longicornis, 1 of Haemaphysalis phasiana, and 11 of I. turdus) were removed from 102 birds of 21 species (not shown).

Isolation of spirochetes from unmolted ticks. In autumn, 10 of 34 larvae and 15 of 61 nymphs of H. flava feeding on T. pallidus, 1 of 32 larvae of H. flava on T. cardis, and 1 of 14 larvae of H. flava on E. spodocephala were positive for spirochetes (Table 1), and also 1 larva and 1 of 6 nymphs of I. persulcatus on T. pallidus were positive for spirochetes (Table 2). All other tick species were negative for spirochetes. All of the 16 ticks obtained in spring were negative for the presence of spirochetes.

Isolation of spirochetes from molted ticks. One of 16 H. flava females molted from nymphs that fed on T. pallidus in autumn was positive for spirochetes, while spirochetes were not detected in 46 nymphs molted from larvae or 22 males molted from nymphs feeding on T. pallidus (Table 1). Some ticks on T. cardis and E. spodocephala were negative for spirochetes.

Identification of spirochete isolations. Of a total of 30 isolates, 15 strains were used for 5S-23S intergenic spacer RFLP: 10 of 25 from unmolted juveniles of H. flava that fed on T. pallidus (15 isolates from juvenile ticks that fed on the same bird were not used), one each from larval H. flava that fed on T. cardis and E. spodocephala, 1 from molted adults of H. flava, and 2 from unmolted juveniles of I. persulcatus. Figure 2 shows RFLP patterns observed among the isolates. Thirteen of 15 isolates analyzed showed pattern B and pattern B' by MseI and DraI digestion, respectively, according to the patterning in a previous report (29) and were consequently identified as B. garinii of pattern B/B'. However, two isolates, Fi14f and Fi24f, were designated variants (R_v1 and R_v2) of pattern R, because their patterns resembled those of strains 935T (accession no. L39081) from Korea and ChY13p (accession no. AB007450) from China, which had been classified as pattern R (14) (Table 3). To confirm the uniqueness of strains Fi14f and Fi24f, their intergenic spacer sequences were compared with those of some known strains (Table 4). The sequences of strains Fi14f and Fi24f were highly similar to that of strain 935T (99.6%) from Korea and that of strain ChY13p (97.9%) from Inner Mongolia, China (14, 34), respectively. Since strains belonging to the same species usually showed over 95% similarity values for 5S-23S rRNA intergenic spacer sequences in previous experiments (18, 29) and also since strains 935T and ChY13p had been clustered into the B. garinii group based on 16S rRNA sequences (11, 14, 17), strains Fi14f and Fi24f were identified as B. garinii.

View larger version (21K): [in this window] [in a new window]   FIG. 2.   Representative RFLP patterns of the 5S-23S rRNA intergenic spacer observed among Borrelia isolates from bird-feeding ticks. The PCR products were digested by DraI (A) or MseI (B). DNA was electrophoresed on a 16% polyacrylamide gel and stained with ethidium bromide. The molecular size standards are indicated on the left of the gel. Lane 1, Fi01f; lane 2, Fi10f; lane 4, Fi17f; lane 5, Fi19f; lane 6, Fi20f; lane 7, Fi22f; lane 8, Fi23f; lane 10, Fi26f; lane 11, Fi30f; lane 12, Fi71p; lane 13, Fi72p; lane 14, Fi03f; lane 15, Fi16f; all have pattern B. Lane 3, Fi14f; lane 9, Fi24f; both have pattern R.

	DISCUSSION

Top Abstract Introduction Materials and Methods Results Discussion References

The prevalence of ticks on examined birds was relatively high among ground-feeding birds, especially on T. pallidus and T. cardis, while it was generally low among arboreal birds. Some juvenile ticks of I. persulcatus, a northern species associated with Lyme disease borreliae, were detected in some birds examined. This species has not been found previously at this survey site by flagging vegetation (unpublished data) but is known to be widely distributed in mountainous areas over 1,000 m above sea level in the eastern part of Fukui Prefecture (7, 8).

The transovarial transmission of Borrelia does not occur in I. persulcatus or Ixodes ovatus in Japan (22), although it has been demonstrated partly in I. ricinus, Ixodes pacificus, and I. scapularis in Europe and North America (12, 21, 31). The transovarial transmission of Borrelia in H. flava has not yet been examined. However, it is well known that all stages of field-collected H. flava have usually been negative for Borrelia, although there were rare cases in which Borrelia spp. were isolated from field-collected nymphs or adults of genus Haemaphysalis ticks in China (35) and Japan (7, 8). Thus, the present isolations of B. garinii from juveniles of H. flava and I. persulcatus that fed on birds confirm that some species of migratory birds possess B. garinii, and especially that T. pallidus, from which B. garinii-positive ticks were predominantly found, may be one of the important reservoirs. Of course, it is difficult to determine the absolute positivity rate for Borrelia in ticks (or that in host birds) by BSK culture alone, since BSK culture may have a bias for genospecies of Borrelia and since the frequency of Borrelia transmission may vary by feeding time or the engorgement condition of the tick on the host bird. Nevertheless, BSK culture of ticks easily estimated the prevalence of live borreliae in birds examined.

Most of the present isolates showed pattern B and pattern B' by MseI and DraI digestion, respectively. These are common patterns in Borrelia from I. persulcatus in Eurasia and I. ricinus in Europe (16, 29). Nakao et al. (24) reported that isolates from bird-derived larvae of I. persulcatus were identified as ribotype II of B. garinii, the common subtype in Europe and far-eastern Asia, and most strains such as those of ribotype II generated pattern B on the 5S-23S rRNA intergenic spacer PCR-RFLP system (T. Masuzawa, unpublished data). Thus, our findings reconfirmed a strong affinity between birds and B. garinii.

Although not many Korean or Chinese borreliae have been clearly characterized, most Korean and northeastern Chinese strains isolated in previous surveys (11, 14, 34) were identified as B. garinii with pattern B or pattern C or B. afzelii with pattern D (14, T. Masuzawa, unpublished data); only two strains, 935T and ChY13p, were identified as having pattern R. Such a unique pattern had not been observed previously among isolates in Japan and far-eastern Russia. Our findings revealed that strains Fi14f and Fi24f, characterized in the pattern R group, were closely related to strains 935T in Korea and ChY13p in China, respectively. It has been reported that Turdus and E. spodocephala birds mainly migrate on the route shown in Fig. 1 in autumn. Therefore, our results strongly suggest that there is a gradual route of introducing Lyme disease-causing B. garinii from northeastern China via Korea to Japan by long-distance dispersal of ticks feeding on migratory birds. Our additional surveys reconfirmed that there were not many ticks on birds in spring, as the occurrence of juveniles of common ticks including I. persulcatus is known to drop from winter to spring. This suggests that migratory birds have not so many chances to be newly infected with Borrelia before leaving Japan and may not play a significant role in carrying Borrelia from Japan to the Asiatic continent.

Although the transstadial transmission of Borrelia under laboratory conditions has been experimentally demonstrated for I. persulcatus and I. scapularis (5, 23), that in Amblyomma americanum, Amblyomma andersonii, Dermacentor variabilis, and I. ovatus is unclear (5, 23, 32). In the present experiment, a female H. flava tick, which molted from a nymph that fed on T. pallidus, was positive for B. garinii, and the isolate showed the same PCR-RFLP pattern as most B. garinii isolates from unmolted partly engorged larvae or nymphs examined. This is the first study to show that H. flava transmits B. garinii transstadially. This suggests there is an eventual route of Borrelia transmission in nature, namely, the infected adult of H. flava may transmit Borrelia to large animals, although we hardly detected Borrelia in routine samples of field-collected H. flava. The probability of human cases of Lyme disease caused by H. flava is not yet established, although this species is well known to bite humans (33), and a few suspected cases associated with its bite have been seen in Japan (unpublished data).