We captured small mammals at 14 sites across Hokkaido in 2023 and 2025 (Appendix Figure). We identified the species of animal on the basis of morphologic features (11) and confirmed species by sequencing the mitochondrial genome (data not shown). We collected liver, spleen, lung, heart, kidney, brain, and serum samples of trapped animals. All animal experiments were approved by the Animal Care and Use Committee of Hokkaido University (approval nos. 23-0063, 23–0178, and 25-0119). Permission to capture wild mammals was obtained from the Hokkaido Government.

We captured a total of 420 wild small mammals, consisting of 252 mice (Apodemus argenteus and A. speciosus), 121 voles (Craseomys rufocanus and Clethrionomys rutilus), and 47 shrews (Sorex caecutiens, S. gracillimus, and S. unguiculatus). We collected serum samples from 199 animals and screened them for YEZV-specific antibodies by ELISA (1). Of the 199 serum samples, 14 (7.0%) samples were positive for YEZV antibodies (Table 1). The seroprevalence was not statistically different between mice and voles (p = 0.561 by Fisher exact test). We screened liver samples from all 420 animals for YEZV small (S) segment RNA using real-time quantitative reverse transcription PCR (qRT-PCR) (8). We detected YEZV RNA in 16 samples from mice and shrews, whereas all voles tested negative (Table 1). RNA detection rates in mice and shrews were significantly higher than those in voles (p = 0.038 [mice] and p = 0.004 [shrews] by pairwise comparison using Fisher exact test with Holm adjustment). Among the 199 individual animals tested by both real-time qRT-PCR and ELISA, 6 were RNA-positive/antibody-negative, 4 were RNA-positive/antibody-positive, and 10 were RNA-negative/antibody-positive, reflecting different stages of infection.

Figure 1

Figure 1. RNA copy numbers and viral titers in organs of wild small mammals in study of wild mice and shrews as natural reservoirs of emerging tickborne Yezo virus, Hokkaido, Japan, 2023–2025....

We further quantified the amount of S, medium (M), and large (L) segment RNA in the 16 viral RNA-positive samples using the following primers and probes: for the M segment, forward primer 5′-TGGAGTCAAGGGCTGTTATG-3′, reverse primer 5′-CACCAGGCATTTACCTCTACTT-3′, and probe 5′-/56-FAM/TAAGCAGGC/ZEN/TGACATGGAGCCTTT/3IABkFQ/-3′; for the L segment, forward primer 5′-CTCAACCTGCTTCCAACCTATC-3′, reverse primer 5′-GGTGTAAAGCCCAACATCCT-3′, and probe 5′-/5Cy5/CCAAGGAAG/TAO/CACACAGATGGGTACA/3IAbRQSp/-3′. The copy numbers of the S segment RNA were consistently higher than those of the M and L segment RNA (Figure 1, panel A). Of note, all 5 shrews tested positive for all 3 segments, whereas only 2 of the 11 mice were positive for all 3 segments, suggesting that the mode of viral replication differs between mice and shrews.

To confirm production of infectious virions, we attempted virus isolation using the 16 liver samples that tested positive for viral RNA (Table 2). We intraperitoneally inoculated AG129 mice, which showed higher susceptibility to YEZV infection than did cell lines (1,4), with each sample. Subsequently, we inoculated the serum and liver samples of AG129 mice to ISE6 cells. Finally, we successfully isolated YEZV from liver samples from HAs25-107 (A. speciosus mouse) and HSg23-013 (S. gracillimus shrew). To investigate the systemic infection and tissue tropism of YEZV in those wild mammals, we quantified RNA copy numbers and infectious titers in the serum and internal organs (Figure 1, panels B–E). We detected viral RNA in all examined organs except for the brain. Consistent with the RNA copy numbers, we did not detect infectious virus in the serum of HAs25-107 and multiple organs of both animals; the highest virus titers and RNA copy numbers were in the liver.

In 2025, ticks were also collected by a flagging method at the same sites where small mammals were captured for this study to examine local transmission cycles. Of the 2,286 ticks collected, 2,273 ticks were pooled into 288 samples and screened by real-time qRT-PCR (8). A total of 7 pools tested positive for YEZV RNA (Table 2).

Figure 2

Figure 2. Phylogenetic tree of Yezo virus from study of wild mice and shrews as natural reservoirs of emerging tickborne Yezo virus, Hokkaido, Japan, 2023–2025. Tree was constructed based on the nucleotide...

To examine the relationship among YEZV infections in mammals and ticks, we determined complete coding sequences for the S (n = 7), M (n = 5), and L (n = 5) segments from small mammal samples and for the S (n = 2), M (n = 1), and L (n = 1) segments for Ixodes persulcatus tick pools (Table 2). The sequences of YEZV isolates obtained from small mammals and questing ticks in this study shared high identity with previously reported sequences from human patients and ticks collected from humans and birds in Japan (93.2%–100.0% at the nucleotide level and 99.0%–100.0% at the amino acid level) (Appendix Tables 1–3). Phylogenetic analysis based on those sequences demonstrated that most YEZV isolates from small mammals and ticks clustered with isolates from Japan (Figure 2). Some sequences from the same sampling sites (i.e., isolates HSg23-005 and HSc23-012 and isolates HIpe24-390 and HAs25-201) formed site-specific clusters with high identity (>99.9% at the nucleotide level) (Appendix Tables 1, 3). In contrast, isolates HSc23-012 and HSg23-013 from the same site belong to phylogenetically distinct clusters (Table 2; Figure 2). One isolate (HAa25-262) from an A. argenteus mouse clustered with isolates reported in China rather than with isolates from Japan.

This study identified Apodemus spp. mice and Sorex spp. shrews as natural reservoir hosts for YEZV and confirmed local circulation of the virus in Hokkaido, Japan. Systemic infection with viremia indicates that wild mice and shrews serve as a source of infection for feeding ticks, which might subsequently transmit the virus to humans. In contrast, all the liver samples of voles (Cr. rufocanus and Cl. rutilus) were negative for viral RNA despite showing seropositivity comparable to mice, implying that voles contribute less to YEZV circulation. The cocirculation of the diverse isolates at local sites suggests the frequent movement of YEZV within Hokkaido as well as between Hokkaido island (Japan) and Eurasia (China and Russia). Given that wild mice and shrews have small home ranges (13), YEZV is likely dispersed by highly mobile species, such as migratory birds (8,9). In addition, because those shrew and mouse species and their close relatives, together with Ixodes ticks, are also distributed in areas considered as nonendemic for YEZV, such as the Korean Peninsula and the main island of Japan (Honshu) (11,12), YEZV might be more widely distributed than currently recognized. Our findings provide critical insights into the ecology of YEZV and highlight the need for efficient epidemiologic surveillance for this virus.