Introduction

Since the first cases of the highly pathogenic avian influenza (HPAI) H5N1 (clade 2.3.4.4b) virus were reported in US dairy cattle in early 2024, the outbreak has spread rapidly, affecting over 1000 dairy herds across 17 states (https://www.cdc.gov/bird-flu/situation-summary/mammals.html). In addition, over 50 human cases of infection with bovine HPAI H5N1 viruses have been reported across 10 states as of April 15, 20241 (https://www.cdc.gov/bird-flu/situation-summary/index.html). Most cases involved farm workers in contact with infected animals, with the majority exhibiting mild symptoms2,3. Although there is no evidence of human-to-human transmission, our recent data showed that bovine H5N1 virus has the potential to transmit via respiratory droplets in ferrets, suggesting the potential for transmission among humans4,5.

With no vaccines against bovine H5N1 virus ready to be deployed, antiviral drugs have been crucial to treat infected individuals. We previously examined the susceptibility of bovine H5N1 viruses to several of the antiviral drugs that are available to treat humans infected with influenza virus in mice5. Although oral oseltamivir phosphate or intranasal zanamivir, both of which are neuraminidase (NA) inhibitors, reduced viral replication in the nasal turbinate, lungs, and brain compared to control treatment, they were unable to prevent the lethality of bovine H5N1 virus infection in mice. In contrast, the polymerase inhibitors favipiravir and baloxavir marboxil (BXM) were effective against bovine H5N1 virus when treatment was initiated 1 h post-infection. BXM is more widely used globally than favipiravir as a treatment for influenza.

In this study, we conduct a detailed assessment of the therapeutic efficacy of BXM in mice infected with the bovine H5N1 virus. We show that early BXM treatment significantly suppresses viral replication and improves survival in mice, whereas delayed administration reduces efficacy and facilitates the emergence of resistance, highlighting the importance of timely antiviral intervention.

Results and discussion

Mice intranasally infected with 10 plaque-forming units (PFU) of A/dairy cattle /Texas/24-008749-001 (H5N1) virus, generated by reverse genetics6, were treated from 1 h, 24 h, 48 h, or 72 h post-infection by oral gavage twice daily (at 12 h intervals) for 5 days with BXM (50 mg/kg/12 h). A previous study showed that oral administration of BXM at 50 mg/kg in mice resulted in a plasma concentration of baloxavir acid of approximately 25.2 ng/mL at 12 h post-dose7. Because highly pathogenic avian influenza viruses exhibit markedly higher replicative capacity and pathogenicity compared to seasonal strains, the dose of BXM used in this study was several-fold higher than the human equivalent dose so that we could adequately assess the antiviral efficacy, based on the pharmacokinetics in mice8. Body weight and survival were monitored for up to 21 days. Animals were euthanized on day 3 or 5 post-infection, and the viral loads in their nasal turbinate, lungs and brain were examined to assess the inhibitory effect of BXM on viral replication in vivo.

Consistent with our previous study, treatment initiated 1 h post-infection drastically reduced viral titers in the lungs, nasal turbinate, and brain compared to mock treatment. No infectious virus was detected in these organs, and all BXM-treated mice survived the lethal bovine H5N1 challenge without body weight loss or seroconversion (Fig. 1A–C, Table 1). This finding suggests that early treatment effectively inhibits bovine H5N1 replication.

Fig. 1: Therapeutic effects of baloxavir marboxil on the replication of the bovine H5N1 virus.
figure 1

Female BALB/c mice (N = 15 per treatment condition) were deeply anaesthetized and intranasally infected with 10 PFU of A/dairy cattle/Texas/24-008749-001/2024 (H5N1) generated by reverse genetics. At 1 h (A–C), 24 h (D–F), 48 h (G–I), or 72 h (J–L) post-infection, mice were treated with baloxavir marboxil (BXM) daily for 5 days. Body weights (A, D, G, J) and survival (B, E, H, K) were monitored daily for 21 days (N = 5 mice per treatment condition). Data are mean percentages ± S.D. of the starting weight. (C, F, I, L) Virus titers in the lungs, nasal turbinate, or brain were determined by performing plaque assays in MDCK cells at day 3 (N = 5 mice per treatment condition) and day 5 (N = 5 mice per treatment condition) post-infection. PFU/g, plaque forming units per gram of tissue. Virus titers in the nasal turbinates and lungs were determined by using plaque assays. Data are means ±  S.D.; points represent data from individual mice. The lower limit of detection is indicated by the horizontal dashed line. Data are from one experiment.

Full size image
Table 1 Neutralizing titers of serum from mice that survived infection with bovine H5N1 virus
Full size table

When treatment was initiated at 24 or 48 h post-infection, viral titers in these organs were suppressed. However, infectious viruses were still detected in two of five mice (24 h) or all mice (48 h). In addition, 60% (24 h) and 100% (48 h) of infected mice succumbed to their infection with significant body weight loss by 21 days post-infection (Fig. 1D–I).

Treatment initiated at 72 h post-infection partially suppressed viral titers in the lungs but not in the nasal turbinate or brain. This late intervention had minimal effect on body weight loss or survival (Fig. 1J–L).

The emergence of drug-resistant viruses is a concern when using antiviral therapies. Previous studies have shown that the I38T substitution in the polymerase acidic (PA) protein of influenza viruses is an important marker of resistance to BXM9,10,11. To assess whether BXM-resistant virus emerged in BXM-treated mice, we attempted to isolate viruses from the organs of BXM-treated mice on day 3 or 5 post-infection, or upon death, and performed susceptibility testing and deep sequencing analysis of the isolated viruses. Virus isolation was first attempted from the lungs and, if unsuccessful, from the brain.

Since no virus was isolated from the group of mice treated with BXM from 1 h post-infection at 3 or 5 days post-infection, the risk of emergence of BXM-resistant virus would be low if treatment is initiated at 1 h post-infection (Supplementary data 1, Sample IDs #6–10, 21–25).

When BXM treatment was initiated from 24 h post-infection, viruses isolated from organs at 3 or 5 days post-infection exhibited similar susceptibility to BXM as the inoculated virus (Supplementary data 1, Sample IDs #11–15, 26–30). However, a virus isolated from one of three mice that succumbed to infection on day 14 showed a 44-fold reduction in susceptibility to BXM and carried the PA-I38T substitution, which is known to reduce BXT susceptibility9,10,11 (Supplementary data 1, Sample ID #52).

When BXM treatment was initiated 48 h post-infection, viruses isolated from organs at 3 days post-infection showed BXM susceptibility to similar that of the inoculated virus (Supplementary data 1, Sample IDs #16–20). However, a virus isolated from one of five mice at 5 days post-infection exhibited a 46-fold reduction in susceptibility to BXM due to the PA-I38T substitution (Supplementary data 1, Sample ID #34). Furthermore, viruses isolated from three of four mice that succumbed to infection on day 6 or 7 showed a 56- to 136-fold reduction in susceptibility to BXM, which was associated with the PA-I38T substitution (Supplementary data 1, Sample IDs #41, 48, and 49).

Treatment initiated 72 h post-infection rapidly led to the emergence of BXM-resistant viruses carrying the PA-I38T substitution; viruses isolated from two of five mice at 5 days post-infection, and from three of five mice that died at 6 days post-infection, exhibited significantly reduced susceptibility to BXM (Supplementary data 1, Sample IDs #36, 38, 43, 45, and 46).

We also attempted to isolate virus from nasal turbinates and found that none of the samples exhibited reduced susceptibility to BXM, and no resistance-associated mutations, including the I138T substitution, were detected (Supplementary data 1).

These results suggest that the timing of BXM treatment initiation is critical for its effectiveness and control of the emergence of BXM-resistant virus.

We note two key limitations of this study: (1) treatment was conducted for only five days, and it remains to be seen whether longer treatment could suppress the emergence of resistant viruses or improve survival; and (2) while the bovine H5N1 virus exhibits relatively high pathogenicity in mice and ferrets, it has not been associated with fatalities in humans, and the disease progression differs between rodents and humans. Although BXM would be expected to reduce viral loads in humans, clinical data are needed to determine the extent to which BXM alleviates clinical symptoms in human cases of bovine H5N1 infection.

Overall, our results demonstrate that while BXM is effective against bovine H5N1 infection, it is essential to initiate BXM treatment as early as possible, as delays in treatment increase the risk of emergence of BXM-resistant virus.

Methods

Ethics statement

All animal experiments and procedures were approved by the Institutional Care and Use Committees of the University of Tokyo (protocol # A2024IMS024). Animals were acclimated to the ambient conditions of the facilities (20–26 °C and 40–60% humidity) prior to the start of the experiments, allowed access to food and water ad libitum, kept on a 12 h on/off light cycle, and given enrichment. Humane endpoints for euthanasia included ≥ 25% body weight loss with no possibility of recovery.

Biosafety

Experiments with HPAI H5N1 viruses were carried out in Biosafety Level 3 (BSL-3) containment laboratories at the University of Tokyo. All experiments were approved by the Institutional Biosafety Committees (IBCs) of the University of Tokyo.

Cells

MDCK (Madin-Darby canine kidney) (obtained from Dr. Robert G. Webster) and 293 T (human embryonic kidney) cell lines (obtained from Dr. Tadashi Matsuda) were grown at 37 °C and 5% CO2 in Eagle’s minimum essential medium (MEM) containing 5% newborn calf serum or Dulbecco’s modified Eagle medium (DMEM) containing 10% fetal bovine serum, respectively. No authentication was performed for any of the cell lines used, and all were monitored regularly for mycoplasma contamination.

Viruses

A/dairy cattle/Texas/24-008749-001/2024 virus was generated by reverse genetics. Briefly, cDNAs encoding viral RNAs, based on published sequences (EPI_ISL_19014384)12, were synthesized by using a commercial vendor (Integrated DNA Technologies) and cloned into RNA polymerase I-based plasmids, followed by influenza virus generation as previously described5,6. The reverse genetics-generated TX001-H5N1 virus was sequenced to confirm the absence of unwanted mutations.

Antiviral susceptibility testing

For antiviral susceptibility testing in mice, six-week-old female BALB/c mice (Japan SLC, Inc.; N = 15 per group) were anaesthetized under isoflurane and intranasally inoculated with 10 PFU of the indicated viruses. At 1 h, 24 h, 48 h, or 72 h post-inoculation, mice were treated with baloxavir marboxil (50 mg per kg per 200 μl; obtained from Shionogi & Co., Ltd.) or methylcellulose (200 μl) administered orally twice daily for 5 days as described previously5. For some mice (N = 5 per treatment), body weights and survival were monitored daily for 21 days. For the remaining mice, tissues were collected on days 3 and 5 post-infection for virus titration (5 per time point).

Inhibitory concentration 50% (IC50) assay

MDCK cells in 6-well plates were infected with dilutions of influenza viruses sufficient to result in approximately 50 plaques per well. Infections were carried out for 1 h at 37 °C, and then the inoculum was removed, and cells were overlaid with MEM containing 0.3% bovine serum albumin, 1% low melting point agarose, 1 μg/ml TPCK-treated trypsin, and different concentrations of baloxavir acid. After 48 h at 37 °C, the cells were fixed with 10% formalin, overlays were removed, plates were dried, plaques were counted, and IC50 values were calculated by using Graphpad Prism. In accordance with the WHO criteria, a fold-change value greater than 3-fold was provisionally considered indicative of reduced susceptibility to baloxavir in this study (https://cdn.who.int/media/docs/default-source/influenza/laboratory---network/quality-assurance/antiviral-susceptibility-influenza/pa-marker-who-table_07-08-2024_updated_final-version.pdf?sfvrsn=5307d6fe_2).

Micro-neutralization assay

Virus neutralizing antibody titers against A/dairy cattle/Texas/24-008749-001 (H5N1) virus were evaluated in serum samples. Serum samples were treated with receptor-destroying enzyme (RDE; Denka Seiken) at 37 °C for 20 h, inactivated at 56 °C for 1 h, and diluted 1:10 in phosphate-buffered saline (PBS). Two-fold serial dilutions of sera were prepared in MEM, and each dilution was incubated with the same volume of virus diluent (100 TCID50/50 µL) in MEM containing 1 µg/mL of TPCK-trypsin at room temperature for 1 h. The serum/virus mixture was added to 100% confluent MDCK cells that were plated one day prior in 96-well plates. The cells were incubated for 2–3 days at 37 °C and then cytopathic effect (CPE) was microscopically assessed by eye. Virus neutralization titers were determined as the reciprocal of the highest serum dilution that completely prevented CPE. Each sample was analyzed in duplicate.

Deep sequence analysis

Viral RNA was extracted by using a QIAamp Viral RNA Mini Kit (QIAGEN). The All eight viral RNA segments of influenza A virus were simultaneously amplified by using the SuperScript IV one-step RT-PCR system (Thermo Fisher Scientific) with universal primers [MBTuni-12-R (5′-ACG CGT GAT CAG CRA AAG CAG G-3′) and MBTuni-13 (5′-ACG CGT GAT CAG TAG AAA CAA GG-3′)] targeting the highly conserved sequence of the viral RNA termini13. The multisegment RT-PCR (M-RTPCR) products were processed for sequencing libraries by using a QIAseq FX DNA Library Kit (QIAGEN) and were then analyzed by using the iSeq 100 System (Illumina) with a iSeq 100 i1 Reagent v2 (300-cycle) or the MiSeq i100 plus system (Illumina) with MiSeq i100 Series 25 M Reagent Kit (300 cycles). To determine the virus sequences, the reads were assembled by CLC Genomics Workbench (version 24, Qiagen) with the A/dairy cattle/Texas/24-008749-001/2024 sequences (EPI_ISL_19014384) as a reference. The average coverage depth was more than 4000.

Statistics and reproducibility

All animals were randomly allocated to experimental groups. No blinding was performed in any experiment. Sample sizes were based on our previous work. IC50 calculations were performed with GraphPad Prism software. All graphs were generated in GraphPad Prism. Survival curves were compared by using the log-rank Mantel-Cox test. All experiments with animals were performed once.

Reporting summary

Further information on research design is available in the Nature Portfolio Reporting Summary linked to this article.