In 2003 and 2004, a total of 273 unfed and adult Ixodes ticks (114 I. persulcatus and 159 I. ovatus) were collected in central Japan (Figure 1). Of these, 123 live ticks were dissected, and DNA was isolated from whole tissues of 73 ticks and salivary glands of 50 ticks by using the QIAamp DNA mini kit (Qiagen Inc., Valencia, CA, USA). For detection of A. phagocytophlilum DNA, a nested PCR using primers designed based on the highly conserved region of p44/msp2 paralogs of (p3726 [5´-GCTAAGGAGTTAGCTTATGA-3´], p3761, p4183, and p4257) was conducted (12–14). Four (12.1%) of 33 I. persulcatus ticks collected at the Utsukushinomori (UM) site in Yamanashi Prefecture were positive by PCR (Table). Sixteen (7 I. persulcatus and 9 I. ovatus) (32%) of 50 salivary glands from ticks collected at the Takabachi and Mizugazuka sites in Shizuoka Prefecture were positive by PCR. Data indicated that I. persulcatus and I. ovatus in Japan are naturally infected with A. phagocytophilum and that ticks at certain sites are highly infected.

We further examined the infection of immunocompromised mice with A. phagocytophilum in ticks by using the procedure described previously (12). Briefly, whole tissues from 150 live ticks (55 I. persulcatus and 95 I. ovatus) were pooled and intraperitoneally injected into 15 ddY male mice (6–15 pooled ticks per mouse) treated with the immunosuppressant cyclophosphamide. PCR was conducted with DNA isolated from blood and spleens of these mice. Only 1 of 9 spleens from I. ovatus-injected mice was positive by PCR (Table). We previously detected Ehrlichia spp. DNA in I. ovatus–injected mice, but did not detect A. phagocytophilum DNA in I. ovatus– or I. persulcatus–injected mice (12) because we used only a few immunocompromised mice, i.e., most had normal immune systems. Thus, we treated all 15 mice used in the present study with cyclophosphamide. Results indicate that A. phagocytophilum in I. ovatus can be infective for immunocompromised mice, although the efficiency of infection was low (1/95 [1.1%]).

The p44/msp2 amplicons from 8 PCR-positive ticks and 1 PCR-positive mouse were cloned into a pCR2.1 vector with the TA Cloning Kit (Invitrogen, Carlsbad, CA, USA). Recombinant clones were randomly selected and 28 recombinant p44/msp2 clones were sequenced with an ABI 3100-Avant Genetic Analyzer (Applied Biosystems, Foster City, CA, USA). A phylogenetic tree was constructed based on the alignment of Japanese p44/msp2 sequences and the most closely related paralogs (220–400 bp) by using ClustalX (http://www-igbmc.u-strasbg.fr/BioInfo/ClustalX/), followed by the neighbor-joining method with 1,000 bootstrap resamplings (Figure 2). In this tree, the p44/msp2 sequences obtained from I. ovatus were located mostly in clusters different from those where sequences from I. persulcatus were located, except for Tick41-1. This finding suggests that A. phagocytophilum in I. ovatus may encode p44/msp2 paralogs distinct from those of A. phagocytophilum in I. persulcatus. A previous study suggested that the p44/msp2 sequences from the United States and the United Kingdom can be divided into 27 similarity groups based on >90% similarities of DNA sequences, and most sequences from the United Kingdom are distinguishable from those from the United State because of the similarities <79% (15). Of 28 Japanese p44/msp2 sequences in this study, 11 sequences with similarities >85.6% to the previously identified paralogs were probably divided into 8 similarity groups (Figure 2). Of the remaining 17 sequences with similarities <73.1%, 11 members that were grouped into 2 distinctive clusters (Figure 2) and 6 members that were individually located (Figure 2, arrows) were distinguishable from the 8 similarity groups. Thus, some p44/msp2 paralogs of Japanese A. phagocytphilum are unique and distinct from those of A. phgocytophlium in other countries, although multiple copies of p44 in the genome of an organism should be considered (13).

A partial sequence of the 16S rRNA gene of A. phagocytophilum (1.4 kb) from a p44/msp2 PCR-positive mouse was amplified from spleen DNA with primers ER5-3, ER-R1, AP-F1, and AP-R1 (12), cloned, and sequenced. Similarities among 6 Japanese recombinant 16S rRNA sequences (GenBank accession nos. AY969010–AY969015) were 99.3%–99.6%. When compared with A. phgocytophilum human agent U02521, the similarities were 99.6%–99.8% between individual 16S rRNA cloned sequences and human agent U02521. Because we used pooled ticks to examine infection in mice, these sequence diversities may depend on genetic variants (or a heterogeneous population) of A. phagocytophilum from individual ticks. When the amplicon was directly sequenced, its sequence was identical with that of human agent U02521.

We demonstrated that A. phagocytophilum infects Ixodes ticks in Japan, that both I. persulcatus and I. ovatus ticks are naturally infected with A. phgocytophilum, that A. phagocytophilum may be transmitted by Ixodes ticks because of organisms in the salivary glands of unfed and female adult ticks, and that immunocompromised mice can be infected with A. phagocytophilum. This study provides new information on the ecologic, biologic, and public health significance of A. phagocytophilum and emphasizes the threat of anaplasmosis in Japan.