Experimental mRNA vaccine shields lactating dairy cows from H5N1 infection

An experimental mRNA–lipid nanoparticle vaccine against highly pathogenic avian influenza H5N1 protected lactating dairy cows from infection in a small controlled challenge study, cutting peak viral RNA in milk by roughly 1,000-fold and sharply reducing the mastitis and mammary fibrosis that the virus causes in infected udders. The vaccine, described in Nature Biotechnology, encodes the hemagglutinin of A/dairy cattle/Texas/24-008749-002/2024, a strain representative of the clade 2.3.4.4b H5N1 viruses that have spread through dairy herds in the United States since 2024, causing mastitis and steep drops in milk production.

The vaccine was well tolerated, with no obvious adverse effects reported. Because previous studies had detected trace amounts of vaccine-related mRNA in human milk after SARS-CoV-2 mRNA vaccination, the researchers also tested whether vaccine-derived mRNA could be found in bovine milk. None of the samples from vaccinated animals contained detectable H5 mRNA, indicating that vaccine-derived mRNA was not present at measurable levels in milk after intramuscular administration.

Immunologically, the vaccine performed as designed. H5-reactive IgG appeared in serum as early as 21 days after the first dose and rose significantly after the booster, and functional assays confirmed that serum antibodies could neutralize the challenge virus. Hemagglutination inhibition assays showed that antibodies from all four vaccinated animals blocked viral attachment to red blood cells before challenge. H5-reactive IgG was also detected in milk, appearing in only one vaccinated cow after the prime but in all four after the boost, although milk antibody levels were lower than serum levels and declined more rapidly. The researchers could not run neutralization tests on milk because nonspecific inhibitors in milk from unvaccinated animals interfered with the assay.

Deep mutational scanning of the viral hemagglutinin revealed that the neutralizing antibodies in vaccinated cow sera targeted antigenic sites A and B on the globular head of the HA protein, near the receptor-binding domain. The amino acid identities at these strongly targeted sites are not highly variable among circulating clade 2.3.4.4b viruses, suggesting the vaccine is well matched to the strains currently infecting dairy cattle. The authors noted that continued surveillance will be needed to detect any future antigenic changes at these sites.

After challenge, the differences between the groups were stark. In unvaccinated cows, H5N1 viral RNA in milk from inoculated mammary quarters was first detected one day after inoculation, peaked at two days, and remained detectable throughout the study period. In vaccinated cows, viral RNA levels were significantly lower at multiple time points, with peak titers about 1,000-fold lower than in unvaccinated animals. Most milk samples that tested positive by PCR with a cycle threshold below 35 contained infectious virus, underscoring the relevance of the RNA measurements. Notably, the researchers found evidence that virus migrated from inoculated to uninoculated mammary quarters in two unvaccinated cows, with viral RNA, infectious virus and viral antigen detected in milk and tissue from uninoculated quarters, whereas viral RNA in milk from uninoculated quarters of vaccinated animals remained below the limit of detection.

Clinical outcomes tracked the virological results. Unvaccinated cows showed decreased rumen motility and significant drops in milk production beginning two days after inoculation, with a dramatic decline at day three that persisted through much of the monitoring period. Their milk showed severe changes in color and consistency indicative of mastitis, and California mastitis test scores in inoculated quarters consistently indicated moderate to severe inflammation. Vaccinated cows were far less affected on all these measures, and no significant differences in rectal temperature were observed between the groups. At necropsy, inoculated mammary quarters from unvaccinated cows had macroscopic lesion scores of 95.4 percent compared with 8.1 percent in vaccinated cows, and microscopic scoring of collagen staining showed significantly less mammary fibrosis in the vaccinated group.

The authors emphasize that the vaccine reduced disease and shedding but did not sterilize infection. While it is reasonable to suspect that the large reduction in viral titers in milk would limit animal-to-animal transmission, they write that additional large-scale field experiments will be required to test that possibility directly. They also note that future work should optimize vaccination schedules and doses, explore whether a single immunization could elicit protective antibody levels in milk, and consider a differentiating infected from vaccinated animals strategy, since the vaccine elicits antibodies only to the hemagglutinin and detection of nucleoprotein antibodies could indicate natural infection.

The economic stakes are considerable. The authors cite a study estimating losses of 950 US dollars per clinically affected cow in an infected herd, totaling approximately 737,500 dollars for the herd studied, driven by decreased milk production, mortality and early herd removal. Because the main site of H5N1 replication in cattle is mammary tissue rather than the respiratory tract, and transmission between animals is thought to occur through viral shedding in milk, a vaccine that reduces virus in milk could be particularly valuable for controlling spread within and between farms, while also limiting potential spillovers into humans and other susceptible animals.

The authors also note that they did not assess T cell-mediated immunity in this study, although their earlier work in calves found that mRNA–LNP vaccination induces potent CD8+ T cell responses. They conclude that the H5 mRNA–LNP vaccine is safe and immunogenic in lactating dairy cows, eliciting robust antibody responses in serum and milk while reducing clinical disease, virus shedding in milk and mammary gland fibrosis, and they position it as an additional tool to be combined with existing control measures and robust viral testing programs rather than a standalone solution.

Subject of Research: An mRNA–lipid nanoparticle H5N1 vaccine tested in experimentally challenged lactating dairy cows

Article Title: An mRNA–lipid nanoparticle vaccine against H5N1 influenza virus protects lactating dairy cows from infection

Article References: Santos, J. J. S., Souza, C. K., Zanella, G. C., Goulart, D. B., Arruda, B., Boggiatto, P., Palmer, M. V., Snyder, C. A., Kristula, M. A., Dickens, C., Webb, T. L., Atkinson, R. K., Dadonaite, B., Dwivedi, G., Alameh, M.-G., Hunter, C. A., Murphy, L. A., Bloom, J. D., Weissman, D., … Hensley, S. E. (2026). An mRNA–lipid nanoparticle vaccine against H5N1 influenza virus protects lactating dairy cows from infection. Nature Biotechnology. https://doi.org/10.1038/s41587-026-03343-6

Image Credits: AI Generated

DOI: 10.1038/s41587-026-03343-6

Keywords: H5N1, highly pathogenic avian influenza, mRNA vaccine, lipid nanoparticle, dairy cattle, mastitis, milk production, viral shedding, hemagglutinin, neutralizing antibodies, mammary fibrosis, challenge study