Dynamic Circulation and Genetic Exchange of a Shrew-borne Hantavirus, Imjin virus, in the Republic of Korea

Hantaviruses (family Bunyaviridae) are enveloped negative-sense tripartite RNA viruses. The natural hosts of hantaviruses include rodents, shrews, moles, and bats. Imjin virus (MJNV) is a shrew-borne hantavirus identified from the Ussuri white-toothed shrews (Crocidura lasiura) in the Republic of Korea (ROK) and China. We have isolated MJNV and determined its prevalence and molecular diversity in Gyeonggi province, ROK. However, the distribution and phylogeography of MJNV in other regions of ROK remain unknown. A total of 96 C. lasiura were captured from Gangwon and Gyeonggi provinces, ROK, during 2011–2014. Among them, four (4.2%) shrews were positive for anti-MJNV IgG and MJNV RNA was detected from nine (9.4%), respectively. Based on the prevalence of MJNV RNA, the preponderance of infected shrews was male and adult, consistent with the gender- and weight-specific prevalence of hantaviruses in other species. We monitored the viral load of MJNV RNA in various tissues of shrews, which would reflect the dynamic infectious status and circulation of MJNV in nature. Our phylogeographic and genomic characterization of MJNV suggested natural occurrences of recombination and reassortment in the virus population. Thus, these findings provide significant insights into the epidemiology, phylogeographic diversity, and dynamic circulation and evolution of shrew-borne hantaviruses.

in infants and juveniles, suggesting that MJNV may be pathogenic to humans 19 . However, additional genomic sequences of MJNV strains are required to determine the geographic distribution and molecular prevalence in other areas of ROK, as well as the pathogenicity of MJNV in humans.
Genetic exchanges among viruses give rise to genetic diversities that are the basis for molecular evolution 20,21 . Recombination and reassortment are major molecular mechanisms for genetic exchange that results in divergent virus progeny. Previous studies have shown that these genetic events in both RNA and DNA viruses impact their molecular diversity, fitness, and pathogenicity [22][23][24] . Bunyaviruses have been reported to undergo recombination or reassortment in vitro and in nature [25][26][27] . Our recent study identified an S segment recombinant of Hantaan virus (HTNV) in an HFRS patient specimen 28 . In addition, L segment reassortment of HTNV has been shown to occur in nature and contribute to the geographic diversity of HTNV strains in the ROK 29 . However, whether the molecular genetic events of shrew-borne hantaviruses occur in nature have remained unknown.
This study described the distribution and phylogenetic diversity of MJNV in Gangwon province, ROK. The prevalence of MJNV from 96 shrews was comparable between Gangwon and Gyeonggi provinces. There was a clear preponderance of males and adults among MJNV-infected C. lasiura, showing the gender and weight-specific prevalence of hantavirus infection reported in other species of shrews and rodents. Genomic sequences of MJNV were obtained from three (75.0%) of four anti-MJNV IgG seropositive shrews and six (6.5%) of 92 seronegative shrews. Additionally, the entire coding sequence of MJNV 10-8 tripartite RNA was obtained since it was identified in 2010. In an analysis of MJNV RNA in the various tissues of seropositive and seronegative shrews, threshold cycles (Ct) values demonstrated a high viral load (low Ct-value) and divergent distribution of MJNV RNA in IFA + PCR + (seropositive and PCR positive) shrews by reverse-transcription quantitative PCR (RT-qPCR) using MJNV M segment-specific primers. Of the IFA − PCR + (seronegative but PCR positive) shrews, Cl 12-2 and 14-78 harboured a low amount (high Ct-value) of MJNV RNA in different tissues, whereas MJNV RNA was observed only in the lungs of Cl 14-37, 14-42, and 14-70. Phylogenetic analyses of partial genomic sequences of MJNV tripartite RNA demonstrated the well-supported genetic diversity of MJNV in Gangwon and Gyeonggi provinces. Analyses of MJNV open reading frames (ORFs) suggested that recombination and reassortment naturally occurred in the virus population. MJNV 12-2 showed an evidence of L segment recombination, and MJNV strains from Yeoncheon were very likely M segment reassortants. In conclusion, these observations provide better understandings of the epidemiology, phylogeographic diversity, and dynamic circulation and genetic exchanges of shrew-borne hantaviruses in natural reservoirs.

Methods
Ethics statement. Trapping of animals was approved by US Forces Korea (USFK) in accordance with USFK Regulation 40-1 "Prevention, Surveillance, and Treatment of Hemorrhagic Fever with Renal Syndrome". All procedures and handling of animals were conducted under an approved protocol by the Korea University Institutional Animal Care and Use Committee (KUIACUC, #2010-212).
Trapping and autopsy. Shrews  Indirect immunofluorescent antibody (IFA) test. Sera were initially diluted 1:32 in phosphate-buffered saline (PBS) and examined for anti-MJNV IgG antibody. The diluted sera were added to wells of acetone-fixed Vero E6 cells infected with MJNV, and the slides were incubated for 30 min at 37 °C. After washing twice, fluorescein isothiocyanate-conjugated goat antibody to rat and mouse IgG antibodies (MP Bio Inc., CA, USA) was added, and the slides were incubated at 37 °C for 30 min. After washing three times, virus-specific fluorescence was evaluated using a fluorescent microscope (Axioscope, Zeiss, Berlin, Germany).     33 . Recombination and reassortment events were significantly suggested by RDP4 if at least two criteria were satisfied; the P-value (p) was under 0.05 and the RDP recombination consensus score (RDPRCS) was over 0.6 27 . Recombination and reassortment events were considered possible when p was under 0.05 and the RDPRCS was between 0.4 and 0.6. The likelihood of recombination and reassortment events was considered insignificant when the RDPRCS was under 0.4 with p < 0.05. Subsequently, phylogenetic relationships were reconstructed for each genetic event using the ML method in MEGA 5.2.

Results
A total of 96 Ussuri white-toothed shrews (C. lasiura) were captured in Gangwon and Gyeonggi provinces from 2011 to 2014. Using sera from the shrews, serological tests showed that four (4.2%) of 96 samples were positive for anti-MJNV IgG, with three from Gangwon province and one from Gyeonggi province ( Table 2). There was no seropositive C. lasiura for rodent-borne hantaviruses including HTNV and Seoul virus (SEOV). Partial MJNV L (coordinates 962-1,593 nt) and M (coordinates 2,252-2,784 nt) sequences were detected in nine (9.4%) out of 96  To determine the amount of MJNV RNA in IFA + PCR + and IFA − PCR + C. lasiura, real-time PCR was performed on samples from various tissues, including lungs, livers, kidneys, and spleens (Fig. 2). The Ct values of the RT-qPCR to detect MJNV M segment from Cl 13-1, 14-21, and 14-73 (IFA + PCR + ) were overall low (high viral loads) and varied in the different tissues, demonstrating a dynamic MJNV infection in hosts. Cl 12-2 (IFA − PCR + ) showed the presence of MJNV RNA in all tissues, whereas all tissues, except for the liver, from Cl 14-78 (IFA − PCR + ) contained the comparable viral loads of MJNV RNA. The IFA − PCR + shrews, Cl 14-37, 14-42 and 14-70, showed the highest MJNV RNA loads in their lungs, whereas there were very low amounts of viral RNA in their livers, kidneys, and spleens. The amount of MJNV RNA in Cl 14-71 (IFA − PCR + ) was likely less than the limit of detection of the assay.    To investigate the genetic diversity of MJNV in Gangwon and Gyeonggi provinces, phylogenetic trees were generated by NJ, ML, and Bayesian methods. The geographic locations of MJNV strains collected in this study are shown in the Fig. 3a. The phylogenetic analysis of the partial MJNV L segment, coordinated to 962-1,593 nt, demonstrated well-supported geographical clusters (Fig. 3b). In Gangwon province, sequences of MJNV strains from Hwacheon clustered with each other. MJNV 12-2 from Pyeongchang formed an independent genetic lineage with MJNV 14-21 from Inje. MJNV 14-78 from Cheorwon grouped with the MJNV strains from Pocheon, Gyeonggi province. In Gyeonggi province, MJNV strains from Paju formed a distinct geographic cluster. MJNV strains from Yeoncheon seemed to be differentiated from all other MJNV strains in ROK. The L segment of MJNV strains differed by 0.0-13.0% and 0.0-4.4% at the nucleotide and amino acid levels, respectively (Table S1). The partial MJNV M sequences (coordinates 2,252-2,784 nt) detected in the shrews from Gangwon and Gyeonggi provinces showed that MJNV strains from Hwacheon and Pocheon formed geographical clusters (Fig. 3c). The partial M sequence of MJNV 12-2 from Pyeongchang formed a distinct genetic lineage with MJNV 14-21 from Inje. MJNV 04-3, 05-10, and 10-8 from Paju, Gyeonggi province, were included in a single genetic lineage, whereas MJNV 05-11, 09-3, and 14-37 grouped phylogenetically with the MJNV strains from Yeoncheon. The nucleotide and amino acid sequences of the MJNV M segment showed differences of 0.0-11.7% and 0.0-2.3%, respectively (Table S2). Phylogenetic analysis of the partial MJNV S segment, coordinated to 405-1,012 nt, indicated that MJNV strains from Hwacheon formed a geographic cluster (Fig. 3d). Consistent with the partial MJNV L and M sequences, partial S sequences of MJNV 12-2 from Pyeongchang grouped phylogenetically with MJNV 14-21 from Inje in Gangwon province. MJNV 14-78 from Cheorwon was closely related to MJNV 09-136 from Pocheon. In Gyeonggi province, MJNV 10-8 and 14-37 from Paju clustered together. The genetic lineage comprising MJNV strains from Yeoncheon was distinct from all other MJNV strains. The differences of nucleotide and amino acid sequences of the partial MJNV S segment differed by 0.0-12.0% and 0.0-1.5%, respectively (Table S3). To evaluate recombination and reassortment events, sequences of the tripartite ORFs from MJNV genomes were concatenated, aligned, and analysed using RDP 4.0. The partial sequence of MJNV 12-2 L segment, originated from Pyeongchang, was very likely to represent a recombination with MJNV 04-55 from Yeoncheon (Fig. 4a). The p-values of the analysis ranged from 2.824E-5 to 7.821E-30, and the RDPRCS of MJNV12-2 was 0.725. Using MJNV strains representing a recombinant, parents, and in-and out-groups, phylogenetic trees of the MJNV L, M, and S segments were generated. The recombined region (coordinated 1,775 to 2,287 nt) of the MJNV 12-2 L segment clustered with MJNV strains from Yeoncheon, whereas the rest of the MJNV 12-2 L segment (coordinated 45 to 1,774 nt and 2,288 to 6,494 nt) clustered with MJNV 14-21 from Inje ( Fig. 4b-d). The ORF sequences of MJNV 12-2 M and S segments clustered with those of MJNV 14-21 ( Fig. 4e and f). These results indicate that MJNV12-2 may be a recombinant because of a partial exchange of the L segment in nature.
The RDP4 analysis suggested that the genomic configuration of MJNV 04-55 and 13-1, identified from Yeoncheon, was likely to represent the reassortment of M segments with MJNV 05-11 and 14-37 from Paju (Fig. 5a). The p-value from the analysis was under 0.05, and the RDPRCS of MJNV strains from Yeoncheon was 0.548. The results satisfied the criteria to suggest a reassortment event in the MJNV M segment. The phylogenetic trees, including reassortants, parents, and in-and out-groups, demonstrate that the M segments of MJNV strains from Yeoncheon clustered with those from Paju, whereas the L and S segments formed a distinct genetic lineage (Fig. 5b).

Discussion
We have demonstrated the isolation, characterization, and molecular diversity of MJNV strains collected from C. lasiura in Gyeonggi province, ROK 9,17 . The previous study described serological and molecular prevalence year, demonstrating about 9.0%, 13.5%, 8.0%, and 2.5% for spring, summer, autumn, and winter, respectively. This observation suggests that the circulation of MJNV is enzootic rather than epizootic among shrews.
HTNV, a rodent-borne hantavirus, causes HFRS in the highly endemic areas of ROK and China 34,35 . The prevalence of HTNV in Gyeonggi province was found to be higher than that in Gangwon province, corresponding to higher HFRS incidences 36 . We reported that seronegative rodents in the highly HFRS-endemic area of Gyeonggi, but not Gangwon province, harboured HTNV-specific genes 37 [39][40][41][42][43] . We found that male and adult shrews showed significantly higher prevalence of MJNV infection than female (Fisher's exact test, p < 0.05) and youth shrews. The preferential male infection with MJNV suggests that the horizontal transmission of MJNV may occur via aggressive behaviors such as biting or fighting. The dominant prevalence of adult infection with MJNV indicates that reproductive activities may be associated with the transmission of MJNV. The breeding of C. lasiura occurs during summer and autumn 44 . The detection of MJNV RNA in shrews was observed in September and November, suggesting the breeding may be one of transmission modes for MJNV.
To examine the viral load of MJNV in shrews, the amount of MJNV was determined from various tissues of IFA + PCR + and IFA − PCR + animals. IFA + PCR + samples showed higher viral loads of MJNV RNA in the tissues, including lungs, kidneys, livers, and spleens. For the three IFA − PCR + samples, MJNV RNA was the highest in the lung, reflecting an early phase of infection with MJNV in the endemic areas during autumn (September and November) 45 . An early phase of infection may be explained by MJNV replication primarily in the lung before MJNV-specific IgG antibodies were generated 46,47 . The viral RNA loads in the different tissues may reflect the dynamic infectious status and circulation of MJNV among the shrews in autumn.
Genetic exchanges, e.g., recombination and reassortment, promote genetic diversities for viral evolution in nature 24 . These exchanges have influenced fitness, transmission, and pathogenicity of the virus 48 . A recent study demonstrated that the reassortment of PB1, PB2, and HA of influenza B virus contributed to its fitness and replication 49 . Recombination and reassortment of bunyaviruses have been observed in nature and in vitro [50][51][52] . A previous study described M segment recombinants in the HTNV population of southwestern China 53 . Other hantaviruses, such as Tula virus, PUUV, Andes virus, and SEOV, exhibited evidences of the S segment recombination in natural reservoirs or in vitro 23,26,[54][55][56] . SNV, a family of Bunyaviridae, has shown the evidence of an M segment reassortment in nature and in vitro 52 . Recently, HTNV L segment reassortment was found to occur naturally in Gyeonggi and Gangwon provinces, ROK 29 . In this study, both recombination and reassortment were detected in the MJNV population. Recombinant MJNV 12-2, identified from Pyeongchang, contained a heterogeneous partial L segment sequence (coordinated to 1,775-2,287) that clustered with MJNV 04-55 from Yeoncheon. However, the rest of the L segment and whole M and S segments of MJNV 12-2 formed a genetic lineage with MJNV 14-21 from Inje. This configuration of the MJNV tripartite genomes may exhibit the recombination of L segment in nature. In addition, the M segment of MJNV 04-55 and 13-1 from Yeoncheon grouped with that of MJNV 05-11 from Paju, whereas their L and S segments phylogenetically formed a segregated lineage. This genomic configuration suggested that MJNV strains in Yeoncheon contained a genome organization compatible with reassortment of the M segment. We observed the L segment reassortant and S segment recombinant of HTNV infected humans and caused HFRS, respectively 28 . Dobrava-Belgrade virus (DOBV) showed the reassortment of M segment between avirulent strain DOBV-Aa and virulent strain DOBV-Af in vitro 57 . Thus, further studies are required to understand the biological and ecological consequences of recombination or reassortment in the MJNV population.
In conclusion, this study first characterized the geographic prevalence, phylogenetic diversity, and dynamic circulation and molecular evolution of MJNV from C. lasiura collected in Gangwon and Gyeonggi provinces, ROK. The preferential infection of MJNV was described in male and adult shrews. Based on 28 partial MJNV sequences, including 24 ORF genomic sequences, the phylogenetic analyses indicated geographic diversity of MJNV in the regions. Profiling of MJNV RNA in various tissues of the shrew specimens may reflect the dynamic circulation of MJNV in natural reservoirs. Evidences for the occurrence of recombination and reassortment suggest active genetic exchanges of shrew-borne hantaviruses in nature. These findings provide broad and deep insights into the epidemiology, virus-host interaction, and molecular evolution of hantaviruses in soricid hosts.