Wild mice and shrews emerge as the field reservoirs for Yezo virus in Hokkaido
A four-year ecological sweep across Hokkaido has put two small mammal groups at the center of Yezo virus circulation: Apodemus mice and Sorex shrews. Investigators screened 420 wild small mammals and 2,273 ticks collected from 14 sites during 2023 to 2025, then tied the mammal findings to viral sequence data from local tick pools. The pattern was not subtle. Viral RNA appeared in mice and shrews, infectious virus was recovered from both, and the sequences matched closely with isolates from humans and ticks previously reported in Japan, China, and Russia.
The work matters because Yezo virus is not just a lab curiosity. It is a tickborne orthonairovirus first identified in febrile patients after tick exposure in Hokkaido, and subsequent human infections have already been documented across a broader swath of northern Asia. The missing piece has been the natural reservoir. Without a reservoir host, the field ecology remains speculative; with one, the transmission cycle becomes far easier to reconstruct, and public health surveillance becomes more targeted.
Infection patterns separate mice and shrews from voles
The study’s animal sampling was broad enough to compare three major mammal groups that share habitat with Ixodes ticks: mice, voles, and shrews. Of 199 serum samples tested for Yezo virus-specific antibodies, 14 were positive, yielding a seroprevalence of 7.0%. Seropositivity did not differ significantly between mice and voles. Viral RNA told a different story.
All 420 liver samples were tested by real-time quantitative reverse transcription PCR, and Yezo virus RNA was detected in 16 samples from mice and shrews, while all vole samples remained negative. RNA detection rates were significantly higher in mice and shrews than in voles. Among the 199 animals tested by both PCR and ELISA, six were RNA-positive and antibody-negative, four were positive by both methods, and 10 were RNA-negative and antibody-positive. That distribution fits a moving target of infection stage, with some animals sampled early, before a measurable antibody response, and others retaining serologic evidence after RNA had disappeared from the liver.
The contrast between mammals is the most telling result in the dataset. Voles carried antibodies but no detectable viral RNA in liver tissue, which argues against robust active infection and against a major role in maintaining local transmission. Mice and shrews, by contrast, carried the kind of virologic evidence expected in reservoir hosts: RNA in tissue, higher detection rates, and in selected cases, recoverable infectious virus.
Liver tropism and viremia point to active maintenance hosts
The viral burden in infected animals was not evenly distributed across organs. Copy numbers of the S segment consistently exceeded those of the M and L segments, and all five positive shrews carried all three segments, compared with only two of the 11 positive mice. In practical terms, that suggests more complete or more robust replication in shrews than in mice, although both groups supported the virus. The study also recovered infectious Yezo virus from liver samples of one mouse and one shrew after inoculation into AG129 mice and subsequent culture in ISE6 cells.
The organ data deepen the biologic case. Viral RNA was detected in all examined organs except brain, with the highest titers and copy numbers in liver. Serum from the infected animals did not yield infectious virus, but the systemic tissue distribution and measurable RNA in multiple organs indicate disseminated infection rather than a localized spillover event. A reservoir host does not have to look sick to sustain an arbovirus; it only has to maintain enough circulating virus for feeding ticks to acquire infection. The liver-heavy pattern here is consistent with that biology.
One sentence captures the field implication: these mammals can seed feeding ticks with virus.
Tick pools link the mammal findings to local transmission cycles
The tick survey completed the ecological chain. Of 2,273 ticks pooled into 288 samples, seven pools tested positive for Yezo virus RNA. Complete coding sequences were generated for mammal-derived and tick-derived S, M, and L segments, and those sequences shared high identity with previously reported human and tick isolates from Japan, with nucleotide identity ranging from 93.2% to 100.0% and amino acid identity from 99.0% to 100.0%.
Phylogenetic analysis placed most of the new isolates in clusters with Japanese sequences, but the pattern was not monolithic. Some sequences from the same sampling sites formed site-specific clusters with extremely high identity, while one isolate from an Apodemus argenteus mouse clustered more closely with sequences reported in China. Another pair from the same site landed in separate phylogenetic clusters. Those mixed patterns suggest repeated introduction, local evolution, and movement of the virus across ecological boundaries rather than a single static lineage confined to one pocket of Hokkaido.
The sequence data also weaken any assumption that YEZV is geographically narrow. When a virus from a mouse in Hokkaido aligns more closely with a Chinese lineage than with neighboring Japanese strains, the most plausible reading is broader regional circulation shaped by mobile vectors and host movement, not simply local spillover from a fixed focus.
Reservoir identification reshapes surveillance priorities
The investigators conclude that Apodemus mice and Sorex shrews function as natural reservoir hosts for Yezo virus in Hokkaido. That conclusion rests on converging lines of evidence: tissue RNA, serology, infectious virus recovery, and phylogenetic linkage to human and tick isolates. Voles, despite showing antibody positivity, did not show active tissue infection, which makes them a weaker candidate for maintaining the enzootic cycle.
The ecology matters beyond Hokkaido. The same or closely related rodent and shrew species, together with Ixodes ticks, are distributed in regions that have not yet been recognized as endemic for Yezo virus, including the Korean Peninsula and Honshu. Highly mobile species such as migratory birds may help disperse infected ticks or amplify geographic spread across islands and borders. Small home ranges in mice and shrews make them unlikely to explain long-distance movement on their own, which leaves a broader transport network to account for the observed phylogenetic diversity.
For clinicians and field investigators, the practical consequence is straightforward: surveillance cannot stop at human cases. Monitoring ticks alone will miss the vertebrate hosts that keep the virus in circulation between transmission events. Future work will need to define how long these mammals remain viremic, whether seasonality alters infection pressure, and how often the virus crosses between Hokkaido and neighboring regions. The current study identifies the reservoir side of the equation; the remaining task is to map how often infected ticks, infected mammals, and exposed humans intersect in real time.