Could your previous flu infection protect you from bird flu? Scientists find that immunity to 2009 H1N1 slashes the risk of severe H5N1 illness, even after close contact exposure.
Study: Preexisting immunity to the 2009 pandemic H1N1 virus reduces susceptibility to H5N1 infection and disease in ferrets. Image Credit: digicomphoto / Shutterstock
In a recent study published in the journal Science Translational Medicine, a group of researchers tested whether infection-elicited immunity to influenza A virus (IAV) subtypes hemagglutinin 1 neuraminidase 1 (H1N1) or hemagglutinin 3 neuraminidase 2 (H3N2), and the order of exposure, lessens ferret susceptibility to and disease from clade 2.3.4.4b hemagglutinin 5 neuraminidase 1 (H5N1), including after direct contact with a dairy cow isolate.
Background
By age five, most people have already encountered IAVs, leaving an immune imprint that shapes their responses to later exposures. Since 2020, clade 2.3.4.4b H5N1 has swept through birds and spilled over into mammals, including an unprecedented outbreak in United States dairy cows, affecting 989 herds in 17 states within a year, and 41 laboratory-confirmed infections in farm workers by June 2025.
Most human cases have been mild, prompting a key question: can preexisting immunity to seasonal strains blunt H5N1 disease? Understanding the answer matters for workers, clinicians, and pandemic planners.
Further research is needed to clarify how imprinting and cross-reactive antibodies sustain protection over time.
About the study
The researchers used a well-established ferret model because ferrets recapitulate key features of human influenza pathogenesis and transmission. They first induced preexisting immunity by infecting animals with either B/Brisbane/60/2008 influenza B virus (IBV), A/Perth/16/2009 seasonal H3N2 IAV, or A/California/07/2009 pandemic H1N1 IAV.
In a separate experiment, ferrets received two sequential infections, 90 days apart, to test the effect of immune imprinting. Ninety days after the last seasonal infection, animals were either intranasally challenged with 10² tissue culture infectious dose 50% (TCID₅₀) of an A/mink/Spain/2022 highly pathogenic H5N1 virus or pair housed with a naïve cage mate that had been inoculated with 10¹⁰ TCID₅₀ of the A/dairy cattle/Texas/2024 H5N1 virus to model direct contact exposure.
Nasal washes were collected every other day and titrated on Madin-Darby canine kidney cells; body weight, clinical scores, and survival were monitored. Serum immunoglobulin G (IgG) binding, neutralizing, and neuraminidase-inhibiting antibodies were quantified by enzyme-linked immunosorbent assay (ELISA), microneutralization, and enzyme-linked lectin assay (ELLA), respectively.
Predetermined humane endpoints triggered euthanasia. Four ferrets were used per group; the experiments were not randomized or blinded. Statistical comparisons utilized analysis of variance or nonparametric tests as appropriate, and survival was analyzed using the Mantel-Cox test in Prism.
Study results
Preexisting immunity mattered, as compared with immunologically naïve or IBV-preimmune ferrets, animals previously infected with the 2009 pandemic H1N1 IAV shed markedly less virus after challenge with the A/mink/Spain/2022 H5N1 strain, maintained body weight, showed no clinical illness, and all survived.
Seasonal H3N2 IAV immunity offered partial protection: viral titers were intermediate, some animals lost notable weight, but all survived. Only H1N1 pre-immune ferrets mounted robust IgG responses to H1N1, and these antibodies neutralized H1N1 and inhibited the neuraminidase activity of H5N1; they also showed low, non-neutralizing cross-reactivity to H5 hemagglutinin.
When the team layered infections to probe immune imprinting, the pattern held. Regardless of whether H1N1 came first or second, any group that included the 2009 H1N1 virus shed minimal or no H5N1 and did not develop disease. In contrast, ferrets that experienced two rounds of IBV infection, or the sequence H3N2 followed by IBV, shed high titers, lost at least 15 percent of body weight, and up to half reached humane endpoints. Thus, order mattered less than simply having H1N1 in the infection history.
Finally, in a direct contact experiment modeling the high-dose exposures seen during dairy milking, naïve donor ferrets infected with a dairy cow H5N1 strain transmitted the virus to 100% of naïve cage mates, causing universally lethal disease. Prior H3N2 immunity did not prevent infection; every contact shed virus, and only half survived. Prior H1N1 immunity substantially raised the bar: only half of contacts had low, transient shedding on a single day, none developed clinical disease, and all survived.
Together, these data show that infection-elicited immunity to the 2009 H1N1 virus reduces susceptibility to, replication of, and disease caused by contemporary clade 2.3.4.4b H5N1 viruses in ferrets, whereas H3N2 immunity alone confers incomplete and variable protection.
Mechanistically, H1N1 exposure generated high titers of cross-reactive N1 neuraminidase–binding antibodies that inhibited the dairy cow and mink H5N1 neuraminidases, plus low titers of group 1 hemagglutinin stem-directed antibodies that bound but did not neutralize H5. Nucleoprotein antibodies, a proxy for responses to conserved internal antigens, were comparable between the H1N1 and H3N2 groups, suggesting that the superior protection in H1N1-primed animals likely reflected hemagglutinin or neuraminidase cross-reactivity rather than broader immunity to internal proteins. These immunologic patterns mirror recent human serology against clade 2.3.4.4b H5N1, reported in independent studies.
Conclusions
To summarize, preexisting, infection-elicited immunity to the 2009 pandemic H1N1 IAV acted as a strong barrier against clade 2.3.4.4b H5N1 infection, replication, and disease in ferrets, even after high-dose direct contact exposure. Immunity to seasonal H3N2 IAV attenuated but did not reliably prevent infection or illness. The order of earlier infections (immune imprinting) did not erode the protection conferred by H1N1 exposure.
While the authors acknowledge certain limitations, such as the fixed 90-day infection intervals not reflecting human timelines and an inability to assess cellular immunity, these findings help explain why most recent human H5N1 infections linked to dairy cows have been mild and argue that neuraminidase-focused cross-reactivity should be considered in preparedness strategies, vaccine design, risk assessment, and clinical surveillance policies.
Journal reference:
- Restori, K. H., Weaver, V., Patel, D. R., Merrbach, G. A., Septer, K. M., Field, C. J., Bernabe, M. J., Kronthal, E. M., Minns, A., Lindner, S. E., Lakdawala, S. S., Le Sage, V., & Sutton, T. C. (2025). Preexisting immunity to the 2009 pandemic H1N1 virus reduces susceptibility to H5N1 infection and disease in ferrets. Sci. Transl. Med. 17 (808). DOI: 10.1126/scitranslmed.adw4856, https://www.science.org/doi/10.1126/scitranslmed.adw4856