To the Editor: Tick-borne relapsing fever (TBRF) is caused by infection with spirochetes belonging to the genus Borrelia. We previously reported a human case of febrile illness suspected to be TBRF on the basis of serologic examination results; the vector most likely was a genus Carios tick that had fed on a seabird colony (1). However, surveillance of ticks in the area did not identify Borrelia spp. in any of the Carios ticks sampled (2). In 2007 and 2008, a borreliosis investigation was conducted on Kutsujima Island (35.71′N, 135.44′E) because a bird-associated tick, genus Carios, inhabits this island. During the investigation, 77 Carios ticks (55 nymphs, 11 adult males, and 11 adult females) were collected from colonies of seabirds: Swinhoe's storm petrel (Oceanodroma monorhis) and streaked shearwater (Calonectris leucomelas). Identification of tick species as C. sawaii was based on tick morphology and rrs gene sequence analysis of the tick mitochondrion DNA (2). Total DNA was extracted from the ticks by using a DNeasy Tissue Kit (QIAGEN, Germantown, MD, USA). For the detection of Borrelia DNA, PCR designed was based on the flagellin gene (flaB) according to Sato et al. (3). To check for contamination and amplicon carryover, we used blank tubes as a negative control for each experiment. Of 77 C. sawaii ticks that were positive by PCR of tick genes (2), 25 (14 nymphs, 6 adult males, 5 adult females) were positive for Borrelia DNA by PCR of flaB.

To characterize the Borrelia spp., we sequenced amplified fragments of the flaB gene and the 16S ribosomal RNA (16SrRNA) gene of Borrelia spp. in a tick and compared the results with those of representative Borrelia spp. The primers BflaPBU and BflaPCR (3) for flaB and the 4 PCR primers (Technical Appendix) for 16SrRNA were used for direct sequencing and/or amplification. DNA sequence (GenBank accession no. AB491928) of a 294-bp amplified fragment of flaB showed the following nucleotide similarities with those of Borrelia spp.: B. turicatae (98.98%), B. parkeri (98.30%), Borrelia sp. Carios spiro-1 (98.64%), and Borrelia sp. Carios spiro-2 (98.30%). DNA sequence (GenBank accession no. AB491930) of a 1,490-bp amplified fragment of 16SrRNA showed the following nucleotide similarities with those of Borrelia spp.: B. turicatae (99.60%), B. parkeri (99.53%), and Borrelia sp. Carios spiro-2 (99.45%). Borrelia sp. Carios spiro-1 and Carios spiro-2, which were recently identified in C. kelleyi in the United States, have been classified into TBRF Borrelia (4,5). The Borrelia sp. found in this study, designated as Borrelia sp. K64, was closely related to B. turicatae but was distinct from other TBRF Borrelia spp. (Technical Appendix).

To observe Borrelia spp. in tick tissues, we performed an indirect fluorescence assay (IFA) according to methods described by Fisher et al. (6), with minor modifications. A tick that was negative by PCRs of flab and 16SrRNA was used as a negative control. The IFA of the tick salivary gland and midgut was conducted by using acetone for fixation, goat anti-Borrelia sp. polyclonal immunoglobulin (Ig) G (1:100; KPL, Inc., Gaithersburg, MD, USA) as the primary antibody, and Alexa fluor 488-labeled rabbit antigoat IgG (1:200, Invitrogen, Carlsbad, CA, USA) as the secondary antibody. Our analysis showed a spirochete, which was stained by anti-Borrelia spp. antibody, in salivary gland and midgut tissue (Technical Appendix). However, no spirochetes were detected by IFA in the negative control (data not shown).

We also attempted to isolate Borrelia spp. from tick specimens by using Barbour-Stoenner-Kelly medium (7). The motility of Borrelia-like organisms in the medium was initially observed by using dark-field microscopy. The Borrelia-like organisms were identified as Borrelia sp. K64 by sequencing of PCR-amplified fragments of flaB and 16SrRNA genes from the cultured medium. However, these Borrelia organisms were found for only 2 weeks after inoculation (data not shown).

The vertebrate reservoir hosts of TBRF Borrelia are usually rodents but can be a variety of other animals (8). Although competence as a reservoir has not been determined for birds, infection of an owl with a TBRF Borrelia sp. has been reported (9). The vertebrate host of the spirochete has not yet been determined. Given our results, it is possible that seabirds are potential vertebrate hosts for Borrelia spp.

In Japan, relapsing fever is a neglected infectious disease because it was not reported during 1956–1998 (10). In this study, we detected a Borrelia sp. in C. sawaii, and the spirochete we characterized is closely related to B. turicatae. Although the human health implications of infections caused by Borrelia spp. are not yet known, the findings from this study should contribute to the epidemiologic investigation of TBRF in Japan.