Establishment of a lethal mouse model of emerging tick-borne orthonairovirus infections

Emerging and reemerging tick-borne virus infections caused by orthonairoviruses (family Nairoviridae), which are genetically distinct from Crimean-Congo hemorrhagic fever virus, have been recently reported in East Asia. Here, we have established a mouse infection model using type-I/II interferon receptor-knockout mice (AG129 mice) both for a better understanding of the pathogenesis of these infections and validation of antiviral agents using Yezo virus (YEZV), a novel orthonairovirus causing febrile illnesses associated with tick bites in Japan and China. YEZV-inoculated AG129 mice developed hepatitis with body weight loss and died by 6 days post infection. Blood biochemistry tests showed elevated liver enzyme levels, similar to YEZV-infected human patients. AG129 mice treated with favipiravir survived lethal YEZV infection, demonstrating the anti-YEZV effect of this drug. The present mouse model will help us better understand the pathogenicity of the emerging tick-borne orthonairoviruses and the development of specific antiviral agents for their treatment.

YEZV, an emerging orthonairovirus in the Sulina group was originally isolated from febrile patients who had a history of tick bites in Hokkaido, Japan.To date, eight infected patients have been reported, including retrospectively identified patients in both Japan and China [12,13].In two clinical cases, acute fever with thrombocytopenia, leukopenia, and elevation of liver enzymes, creatine kinase, and ferritin levels were reported.But, no pathological studies have been conducted in human patients infected with YEZV or those with the other emerging orthonairoviruses including Tamdy virus, Ta ˇche ´ng tick virus-1, and Songling virus.
Previously reported small animal infection models of orthonairoviruses, such as CCHFV have been established using immunodeficient mice, such as interferon receptor knock-out mice [14][15][16][17][18][19][20][21].In these models, the mice exhibit a variety of pathogenic features including liver damage, gastrointestinal disorders, and neurovirulence.Notably liver damage is a representative pathological feature commonly found in a variety of orthonairovirus-infected mice [15][16][17][18][20][21][22] and also in human CCHF patients [2,[23][24][25][26].Elevated levels of liver enzymes in YEZV patients suggest that YEZV infection may also potentially cause liver damage in humans.To investigate the pathology of YEZV infection, we have attempted to establish an animal model for YEZV infection.Specifically, we have developed a lethal mouse model of YEZV infection using type-I/II interferon receptor-knockout mice (AG129 mice) [27] for pathological analysis and for validation of potential antiviral agents.Histopathological and virological analysis using this model revealed that YEZV replicates primarily in liver and spleen resulting in a fatal hepatitis.Furthermore, AG129 mice treated with favipiravir (T-705) were rescued from lethal YEZV infection thereby successfully demonstrating the antiviral effect of T-705 in this mouse model.The present study is the first small animal model for YEZV and will help our understanding of the pathogenetic features of these emerging infections and allow us to evaluate new antiviral agents.

Clinical findings in YEZV-infected mice
To confirm the pathogenicity of YEZV in mice, 10 4 focus formation units (FFU) of YEZV isolated from a human patient were intraperitoneally inoculated into BALB/c, C57BL/6, and
To investigate the pathogenesis of lethal YEZV infection in AG129 mice, AG129 mice intraperitoneally inoculated with 10 4 FFU of YEZV were sacrificed for blood and organ collection at 2, 4, and 6 dpi.Blood biochemistry tests showed elevated levels of alanine transaminase (ALT), aspartate transaminase (AST), total bilirubin and decreased levels of total protein, albumin, and glucose in the YEZV-infected mice (Fig 1C).Marked elevation of ALT and AST levels were observed at 2 and 6 dpi, and interestingly, the increase in ALT and AST at 4 dpi was moderate.Serum total protein and albumin levels decreased significantly at 4 and 6 dpi, and blood glucose levels decreased remarkably at 2, 4, and 6 dpi (Fig 1C).Total bilirubin levels were found to be elevated only at 6 dpi (Fig 1C).Blood cell counts revealed increased white blood cells after 4 dpi (S2A Fig) .White blood cell and platelet counts decreased slightly at 2 dpi, but there was no obvious leukopenia or thrombocytopenia as has been reported in human patients (S2A and S2B Fig)

Pathological findings in YEZV-infected mice
To identify the mechanism underlying fatal infection of YEZV in AG129 mice, histopathological analysis was conducted at 2, 4, and 6 dpi.Macroscopically, the livers and spleens of YEZVinfected AG129 mice were enlarged and whitish at 4 and 6 dpi compared to those of mockinfected mice (Fig 2A).Microscopically, the liver tissue structure was disrupted with hepatocyte degeneration and infiltration of inflammatory cells at 4 and 6 dpi (Fig 2A).The splenic follicular structures were obscured at 4 and 6 dpi (Fig 2A).No clear histological changes were observed in the kidney, heart or small intestine (S3 Fig) .To characterize the distribution of the virus, YEZV antigens in the tissues were examined by immunohistochemical staining (IHC).Positive signals of YEZV antigen were diffusely observed in liver and spleen at 2 and 4 dpi and decreased at 6 dpi (Fig 2B).Viral antigens were detected mainly in hepatocytes in the liver, whereas Iba1(+) macrophages were positive for viral antigen in the spleen (Fig 2B and 2C).YEZV antigens were also detected in the kidney and heart, but at much lower levels than observed in the liver and spleen (S4 Fig) .In contrast, in C57BL/6 and BALB/c mice intraperitoneally inoculated with 10 4 FFU of YEZV, very few viral antigens were detected only in hepatocytes at 2 and 4 dpi (S5 Fig).Slight inflammatory cell infiltration was observed in the livers of these mice, but no clear pathological changes were observed in any of the organs including the liver (S6 Fig) .We next examined whether the observed pathological changes in AG129 mice were dependent on the inoculation route.Similar pathological observations were also found in AG129 mice subcutaneously inoculated with YEZV (S7 and S8 Figs).These results suggest that YEZV replicates primarily in the liver and spleen and causes an acute hepatitis in AG129 mice.
To further confirm viral replication in the tissues, infectious virus and viral RNA loads were quantified.Viral titers in liver, spleen, kidney, heart peaked at 2 dpi and decreased in a time-dependent manner (Figs 3A and S9).Similarly, copy numbers of viral RNA were also detected in liver, spleen, kidney, and heart through 2 to 6 dpi (Figs 3B and S9).In contrast to the AG129 model, infectious virus particles were not detected in the serum, liver and spleen of either BALB/c and C57BL/6 mice at 2 dpi and 4 dpi, but viral RNA was detected in those samples (S10 Fig).
To rationalize the gap between the viral titer and the viral RNA copy number in the AG129 model, viral protein production in the liver was measured by Western Blotting.Consistent with the results of the IHC, the nucleoprotein (N) signal in the liver was not different between 2 dpi and 4 dpi (S11 Fig) .On the other hand, at 4 dpi, the production of the envelope glycoprotein (Gn) was reduced, which may have resulted in an inability to form complete viral particles and a concomitant reduction of viral titer (S11 Fig).

Verification of the anti-YEZV effects of T-705 in the YEZV-infected AG129 mouse model
The nucleic acid analogs, T-705 and ribavirin have been shown to have antiviral effects against a broad spectrum of RNA viruses [28,29], including human pathogenic tick-born bunyaviruses, severe fever with thrombocytopenia syndrome virus (SFTSV) and CCHFV [30][31][32][33].The anti-YEZV effects of T-705 and ribavirin were evaluated using Vero cells in vitro.Both T-705 and ribavirin inhibited YEZV replication in Vero cells (Fig 4).The 50% and 90% inhibitory concentrations (IC 50 and IC 90 , respectively) of T-705 were 3.51 μM and 8.79 μM, respectively, and those of ribavirin were 12.32 μM and 25.47 μM, respectively.Since T-705 showed a stronger antiviral effect than ribavirin in vitro, the antiviral effect of T-705 was further analyzed in vivo using our established animal model.
To evaluate this model for the investigation of antiviral agents, YEZV-infected AG129 mice were treated with 100 mg/kg or 300 mg/kg of T-705 immediately after the viral challenge.In

PLOS PATHOGENS
Establishment of a lethal mouse model of emerging tick-borne orthonairovirus infections considered as an etiological agent of emerging orthonairovirus infections in East Asia and thus, has the potential to be a burden on public health [8][9][10][11].However, no pathological analyses of the orthonairoviruses belonging to the Sulina and Tamdy genogroups have been conducted.Our mouse model for YEZV infection is the first for emerging tick-borne orthonairoviruses involving the Sulina and Tamdy genogroup viruses.
Previous studies have used type-I interferon receptor-knockout mice with either C57BL/6 or 129 backbone, STAT-1 knockout mice and mice treated with cyclophosphamide to establish infection models of human pathogenic orthonairoviruses such as CCHFV and Dugbe virus [14][15][16].Here, we have demonstrated that AG129 mice, which are widely utilized in the field of flavivirus research as lethal infection models [34][35][36][37][38], can also be utilized for establishing a model of tick-borne orthonairovirus infections.Because YEZV was isolated from a human patient using AG129 mice that showed swelling of the spleen [12], AG129 mice was our first choice for establishing a mouse model of YEZV infection.
In the present study, we have characterized in detail the pathogenesis of lethal infection of YEZV in AG129 mice.Consistent with previous reports of the other human pathogenic orthonairoviruses, YEZV was lethal in immunodeficient but not in wild-type mice.Wild-type mice did not show any clinical symptoms but potentially developed viremia.Therefore, they may be important to utilize them as a relatively mild disease model in future studies.In comparison to the immunodeficient mouse models of orthonairoviruses, such as CCHFV, Hazara virus, and Tofla virus models, whose LD 50 were usually below 10 FFU or plaque forming unit regardless of the inoculation route, the LD 50 of YEZV infection in AG129 mice was compatible [16,17,19,21].YEZV replicated mainly in the liver and spleen in AG129 mice, resulting in severe liver inflammation.Infectious virus was rapidly eliminated before the AG129 mice died, but the severe hepatitis remained even after virus clearance and could have contributed to the observed mortality.Elevated liver enzyme levels, including AST and ALT, consistent with reported human cases [12,13], also suggests serious liver damage as found in mouse models of other orthonairovirus infection, including CCHFV, Hazara virus, Leopards Hill virus, and Keterah virus [15][16][17][18][20][21][22]39].Interestingly, the present mouse model showed a biphasic elevation of AST and ALT value consistent with biphasic-like increase of AST in a human patient [12], and which has not been reported in the other orthonairovirus infections in humans and also animal models.In the studies of CCHFV, Kupffer cells, hepatocytes, and endothelial cells have been identified as the primary target cells in mouse and macaque infection models and in human patients [2,15,20,[40][41][42][43][44].Kupffer cells have been proposed to be important for viral spread and replication in the liver during the early stages of CCHFV infection [20,41,42].

PLOS PATHOGENS
Establishment of a lethal mouse model of emerging tick-borne orthonairovirus infections However, in YEZV-infected mice, most of the virus antigen-positive cells in the liver were hepatocytes and only a small number of virus antigen-positive Kupffer cells were detected, even though virus antigen-positive signals were detected in Iba1(+) macrophages in the spleen.While infectious viral particles and/or virus RNAs were detected in the kidneys and heart, no tissue damage or severe inflammation was observed histopathologically.In addition, blood urea nitrogen and creatinine levels, indicators of renal function, were normal.Gastrointestinal disturbances as reported in mouse models of Leopards hill virus and Tofla virus [19,22] were not observed.These results suggest that YEZV is a hepatotropic virus like the other orthonairoviruses, although the less tissue damage other than in hepatocytes might be one of the unique features of YEZV that distinguishes it from the other orthonairoviruses, such as CCHFV.
In this model, even though the viral N was detected in the liver at 2 dpi and 4 dpi by both IHC and immunoblot analysis, the viral titer at 4 dpi was significantly lower than that at 2 dpi.On the other hand, the expression of viral Gn was significantly reduced at 4 dpi compared to 2 dpi.Therefore, defective expression of the glycoprotein could be an additional factor limiting virus elimination.Furthermore, although viral titers in tissue and serum declined after 4 dpi, the copy numbers of viral RNA in tissue and serum remained stable until 6 dpi.Together with the temporal reduction of AST and ALT levels at 4 dpi, later liver injury could be caused without the production of infectious virus particles but may still be associated with RNA replication.
Since there are not currently active therapeutic agents for treatment of these newly emerging tick-borne orthoanirovirus infections, discovery of new antiviral agents are required.In this study, the anti-YEZV effects of T-705 and ribavirin were evaluated, and both agents inhibited YEZV replication in vitro.Consistent with other human pathogenic tick-borne bunyaviruses including CCHFV and SFTSV, T-705 showed smaller IC 50 and IC 90 than ribavirin against YEZV in vitro [30,32,45].In addition to our in vitro results, since T-705 showed better therapeutic effect than ribavirin in mouse infection models of CCHFV, SFTSV, and HRTV [30,33,46], we evaluated the anti-viral effect of T-705 using the YEZV infection mouse model.Oral administration of T-705 rescued YEZV-infected AG129 mice from lethal infection in a dose-dependent manner, and viral titers in tissues of T-705-treated mice were lower than the limit of detection.In addition, severe inflammation was not observed in tissues of mice treated with T-705.These results suggest that T-705 inhibited YEZV replication in the early stages of infection, thus preventing lethal inflammation.
Recently reported novel pathogenic tick-borne orthonairovirus infections associated with the Tamdy and Sulina genogroups [8][9][10][11][12][13] have not been fully analyzed for their pathogenesis in humans as patients have been mainly identified by retrospective studies.Analysis using the present mouse model will contribute to our understanding of viral pathogenesis.In the future, it will be important to conduct detailed analyses in both patients and animal models to identify common pathological features.It will also be important to consider co-infections with other dpi (represented by gray color on the figure).All mice were monitored for survival (left panel) and body weight changes (right panel) for 14 days.Fourteen-day survival rates and body weight changes were compared with the vehicle-treated group, 100 mg/kg/day of T-705-treated group and 300 mg/kg/day of T-705-treated group using a log-rank test.(B-E) Vehicle-or 300 mg/kg/day of T705-treated YEZV-infected AG129 mice were sacrificed at 2 and 4 dpi to collect blood and tissue samples.(B) Representative gross images and H&E-stained tissue images of the liver and spleen of mice are shown.Each side of a square in the gross images is 5 mm.The scale bars on the histological images of the liver and spleen are 100 μm and 500 μm respectively.(C) YEZV N proteins were detected by IHC in the liver and spleen of vehicle-treated and T-705-treated mice, and representative images are shown.The scale bars on the images of the liver and spleen tissues are 100 μm and 500 μm, respectively.(D) Viral titers and (E) the copy number of viral RNA (YEZV L-segment) in the liver and spleen of vehicletreated and T-705-treated mice were quantified.Dotted lines indicated the lower limit of detection (L.O.D.) for the assay (4 FFU/mg of liver and spleen and 0.4 FFU/μL of serum).Plots lower than the L.O.D. are indicated by the half value of L.O.D. on the graph.Statistical differences in viral RNA copy number between the vehicle-treated group and T-705-group were examined by the Mann-Whitney U test, and the significant difference (p < 0.01) is indicated as double asterisks.
https://doi.org/10.1371/journal.ppat.1012101.g005pathogens with tick-borne virus, which has been recently highlighted [47].In fact, four of the seven YEZV-infected patients reported in Japan were infected with Borrelia spp.[12].Furthermore, symptoms such as fever and malaise seen in YEZV-infected patients are atypical and might be difficult to distinguish from other infections.A co-infection mouse model of tickborne pathogens may facilitate the development of broad-spectrum pan-tick-borne disease treatment.

Ethics statement
All animal experiments were approved by the Animal Care and Use Committee of Hokkaido University (Approval numbers: 18-0149 and 23-0063) and were performed in accordance with the committee's guidelines.
YEZV strain HH003-2019 was isolated from infected patients as previously reported [12].Briefly, the virus was isolated from the supernatant of Vero E6 cells inoculated the AG129 mouse serum collected at 5 days after intraperitoneally inoculation of infected patient's plasma.The virus was stored at -80˚C until use.

Mouse infection experiments
All animal experiments were conducted in a BSL-2 laboratory at the International Institute for Zoonosis Control, Hokkaido University with the above-described ethical approval.Female 5-week-old BALB/cCrSlc and C57BL/6NCrSlc mice were obtained from Japan SLC and used for animal experiments.AG129 mice, double knockout immunocompromised mice lacking both type I and type II interferon receptors (S129 background), were obtained from Marshall BioResources and maintained in our laboratory.Sex-mixed 5-weeks-old AG129 mice were used for the animal infection experiments.
BALB/cCrSlc, C57BL/6NCrSlc and AG129 mice were intraperitoneally inoculated with 100 μL of DMEM with or without 10 4 FFU of YEZV under anesthesia.All mice were monitored for body weight changes and survival for 14 days, and in order to collect blood and tissue samples, five mice were mock infected or infected with YEZV and sacrificed at 2, 4, and 6 days post-infection (dpi), respectively.A portion of the harvested tissues was homogenized in DMEM to prepare a 10% tissue emulsion for virus titration and RNA extraction.The other tissues were stored in phosphate-buffer with 10% formalin for histopathological analyses.Blood biochemistry tests were conducted using VetScan VS2 with Multi Roter I Preventive Care Panel (Zoetis Japan, Tokyo, Japan).For determining LD 50 , AG129 mice were intraperitoneally or subcutaneously inoculated with 100 μL of DMEM with or without 10 4 , 10 3 , 10 2 , 10 or 1 FFU of YEZV under anesthesia and observed daily for body weight changes and survival for 14 days.

Virus titration
The titers of virus used in this study were defined as FFU.The samples diluted in DMEM were inoculated into Vero cells on 24-well plates and incubated at 37˚C for 1 hour.The supernatants were removed, and the treated cells were cultured in Minimum Essential Medium (MEM) containing 1% methylcellulose and 2% FBS for 6 days.The cells were then fixed with methanol for 20 minutes at -30˚C.After blocking with PBS containing 1% bovine serum albumin for 30 minutes, the cells were incubated with 1,000-time-diluted YEZV-infected mouse serum for 1 hour.After washing three times with PBS, the cells were incubated with Goat Anti-Mouse IgG H&L-HRP (1:1,000, Abcam, Cambridge, U.K.) for 1 hour.After washing in the same manner, the cells were incubated with 3,3'-Diaminobenzidine (DAB), tetrahydrochloride (Nacalai twsque, Kyoto, Japan) for 10 to 20 minutes.Stained foci were counted, and the titers were determined.The staining procedure was performed at room temperature.

RNA extraction and Quantitative real-time PCR
Total RNA was extracted from the 10% tissue emulsions, mice serum or cell culture supernatant using TRIzol LS Reagent (Thermo Fisher Scientific, U.S.) and used to quantify YEZV viral RNA expression.Quantities of YEZV viral RNA were measured using RT-qPCR with One Step PrimeScript III RT-qPCR Mix (Takara Bio, Shiga, Japan) and the specific primers (5'-GGTGTAAAGCCCAACATCCT-3' and 5'-CTCAACCTGCTTCCAACCTATC-3') and probe (5'-/5Cy5/CCAAGGAAG/TAO/CACACAGATGGGTACA/3IAbRQSp/-3') for the L gene.The RT-qPCR reaction protocol included 5 minutes at 52˚C, 10 seconds at 95˚C and then 40 cycles of 5 seconds at 95˚C and 30 seconds at 60˚C.The signals were measured at the elongation step.A cloned plasmid was used for the standard to estimate the copy number.RT-qPCR was performed using qTower 3 G (Analytik Jena, Jena, Germany), and the obtained data were analyzed by qPCRsoft (Analytik Jena, Jena, Germany).

PLOS PATHOGENS
Establishment of a lethal mouse model of emerging tick-borne orthonairovirus infections with Immobilon Western Chemiluminescent HRP Substrate (Merck Millipore) to visualize the peroxidase signal.Beta-actin was detected by the same protocol using Anti-β-Actin pAb-HRP-DirecT (MEDICAL & BIOLOGICAL LABORATORIES CO., LTD., Tokyo, Japan) and HRP substrates.Chemical luminescent signals were detected by Amersham imageQuant 800 (Cytiva, Massachusetts, USA) and signal intensities were quantified using ImageJ software [48].

Verification of the efficacy of antiviral agents
Favipiravir (T-705) (Angene International, Nanjing, China) and ribavirin (Fujifilm Wako Pure Chemical, Osaka, Japan) was purchased from Namiki Shoji (Tokyo, Japan) and Fujifilm Wako Pure Chemical Industries, respectively.T-705 was dissolved in dimethyl sulfoxide (DMSO) for in vitro experiments and in 0.5% methyl cellulose 400 solution for in vivo experiments.Ribavirin was dissolved in DMSO for in vitro experiments.
In the cultured cell infection experiments, Vero cells were infected with YEZV at a multiplicity of infection (MOI) of 0.001 per cell in the presence of various concentration of T-705 (0.01, 0.1, 1, 2, 4, 8, 16, 32, 64, 128 μM) or ribavirin (0.01, 0.1, 2, 4, 8, 16, 32, 64, 128, 256 μM).Supernatants of cell culture were then collected at 6 dpi and used for virus titration.The 50% inhibitory concentration (IC 50 ) and the 90% inhibitory concentration (IC 90 ) were calculated by regression analysis using ImageJ software [48].The MTT assay was performed to evaluate cell viability according to previously reported methods [49].Vero cells were cultured for 6 days in the presence of the designated concentration of agents without viral infection.Cell viability was calculated as follows: [(absorbance of cell culture supernatant in the presence of the agent − absorbance of culture medium without cells in the absence of the agent)/(absorbance of cell culture supernatant in the absence of the agent − absorbance of culture medium without cells in the absence of the agent)] × 100.Viral replication reduction assays and cell viability assays were performed in triplicate.
In the animal experiments, each AG129 mouse was intraperitoneally inoculated with 100 μL of YEZV (10 4 FFU).The infected mice were orally dosed with T-705 (100 or 300 mg/ kg/day) or vehicle using a stomach probe at 1 hour, 1 day, 2 days and 3 days post-infection.The body weight change and survival of 6 vehicle-treated-mice, 7 mice treated with 100 mg/ kg/day of T-705 and 6 mice treated with 300 mg/kg/day of T-705 were monitored for 14 days.In order to collect blood and tissue samples, 5 mice treated with 300 mg/kg/day T-705 and 5 mice treated with vehicle were sacrificed at 2 and 4 dpi, respectively.Viral titration, RT-qPCR analysis and histopathological analysis were conducted as described above.

Statistical analysis
Survival rates were compared using log-rank test.Blood biochemistry levels were compared between uninfected and infected mice by the steel test.Statistical differences in viral RNA copy number between the two groups were examined by the Mann-Whitney U test.All statistical analyses were performed using EZR [50].Statistical significance was set at P < 0.05.
To examine the possibility of viremia, viral titers and viral RNA levels of YEZV in serum were quantified by focus formation assays and RT-quantitative PCR (RT-qPCR), respectively.Viral titers in serum peaked at 2 dpi and remarkably decreased after 4 dpi (Fig 1D).In contrast YEZV viral RNA in serum was detected at similar levels at 2, 4, and 6 dpi (Fig 1E).

Fig 4 .Fig 5 .
Fig 4. The effects of antivirals on YEZV infection in vitro.Inhibitory effects of T-705 and ribavirin on YEZV replication in Vero cells were verified.Viral titers (white circles and solid lines) in the supernatant and cell viability (white triangles and dotted lines) at 6 dpi were compared with those in the absence of the chemical compounds.https://doi.org/10.1371/journal.ppat.1012101.g004