Highly pathogenic avian influenza A(H5N1) clade 2.3.4.4b viruses have displayed unprecedented global spread among wild birds leading to numerous spillover infections in mammalian species. Of note, outbreaks in dairy cattle and gallinaceous birds have resulted in human infections in the United States during 2024–2025 (1). Increased frequency of H5N1 viruses crossing species barriers has caused concern that the avian influenza viruses are adapting to mammals. A critical component of influenza pandemic preparedness is early identification of emerging novel influenza viruses that cause disease and transmit efficiently in humans. A clade 2.3.4.4b H5N1 virus, A/Michigan/90/2024 (MI90), genotype B3.13, was isolated from a conjunctival swab specimen collected from a human patient in Michigan with conjunctivitis after exposure to infected cattle (2,3). In this article, we report the pathogenesis, transmission, and airborne exhalation of MI90 virus in ferrets, the standard animal model for influenza virus risk assessments (4).

We inoculated 18 ferrets with MI90 virus as previously described (5,6). We euthanized 3 ferrets on 3 and 5 days postinoculation (dpi) to assess virus spread in tissues. We used 6 ferrets to assess transmission in a cohoused, direct contact setting as a direct contact transmission model and through the air in the absence of direct or indirect contact as a respiratory droplet transmission model. We paired each ferret with a naive contact, as previously described (4). We observed clinical manifestations daily and collected nasal wash (NW), conjunctival, and rectal swab samples every 2 days postinoculation or postcontact. We confirmed transmission by testing for seroconversion to homologous virus in the contact animals.

Although all MI90-infected ferrets survived the 21-day study, we noted moderate disease. In inoculated ferrets, the mean maximum weight loss was 9.8%, fever (1.8°C above baseline) and lethargy were transient, and nasal and ocular discharge and sneezing were evident on days 4–9 dpi (Table). We detected virus 3 dpi primarily in respiratory tract tissues; titers were highest in ethmoid turbinate samples (7.4 log₁₀ PFU/mL) and at low levels in brain and gastrointestinal tissues. We observed similar results in tissues collected 5 dpi.

Figure

Figure. Transmission and measurement of airborne avian influenza A(H5N1) virus isolated from dairy farm worker, Michigan. A, B) For DCT and RDT testing, ferrets (n = 12) were intranasally inoculated with...

During the direct contact transmission experiment, inoculated ferrets shed virus in NW that peaked at 4.7–5.4 log₁₀ PFU/mL at 1–5 dpi (Figure, panel A). Four of 6 cohoused contact animals had virus in NW (peak 2.5–4.9 log₁₀ PFU/mL) at 5–7 days postcontact, whereas all 6 contact animals had viral RNA detected (3.6–7.7 log₁₀ copies/mL) in NW (7) and seroconverted to MI90 virus, indicating that transmission was 100% (6/6 animals). In the respiratory droplet transmission experiment, NW collected from inoculated animals peaked 2.6–4.8 log₁₀ PFU/mL at 1–3 dpi, whereas 3/6 contact ferrets had detectable virus in NW by day 7 postcontact (peak 2.6–4.8 log₁₀ PFU/mL; days 9–11 postcontact) (Figure, panel B) as well as viral RNA (6.7–8.2 log₁₀ copies/mL), and seroconverted, confirming transmission through the air in 50% of ferrets (3/6). We also detected infectious virus in conjunctival and rectal samples from inoculated animals, but only from 2 contact animals (Table).

To further evaluate the level of virus exhaled by MI90-inoculated ferrets and the potential for airborne transmission, we collected aerosol samples 1 time each day at 1–5 dpi for 1 hour from the 3 ferrets that were euthanized at 5 dpi. Air samples were analyzed for infectious virus and viral RNA by using the BC251 cyclone-based sampler (kindly provided by Dr. William Lindsley, National Institute for Occupational Safety and Health) and the SPOT water condensation sampler (Aerosol Devices, https://aerosoldevices.com), as described previously (8) (Figure, panel D). The highest mean titer of virus was detected at 2 dpi in NW collected from all 3 inoculated ferrets (6.5 log₁₀ PFU/mL) (Figure, panel C). Airborne virus was highest at 3 dpi as measured in both samplers, up to 133 and 41 PFU/hour, supporting transmission observed in both contact models within 3–5 days after exposure.

Overall, MI90 virus displayed reduced virulence in ferrets compared to another H5N1 virus isolated from a dairy farm worker in Texas (8,9); the Texas virus possesses a genetic marker in the polymerase basic 2 protein (E627K), known for enhanced replication and pathogenesis in mammals. At this position, MI90 encodes 627E, like most other viruses isolated from cattle, and contains polymerase basic 2 M631L, which is associated with mammal adaptation (3,9). In addition, polymerase acidic 142N/E has been linked to increased virulence in mice (10); the Texas virus has an E and MI90 virus has a K at this position. Both viruses have identical hemagglutinin sequences associated with receptor binding and the multi-basic cleavage site. Despite differences in virulence, both viruses transmitted in the ferret model with similar proficiency and levels of airborne virus.

Because avian H5N1 viruses cross the species barrier and adapt to dairy cattle, each associated human infection presents further opportunity for mammal adaption. This potential poses an ongoing threat to public health and requires continual surveillance and risk assessment of emerging viruses to improve our ability to predict and prepare for the next influenza pandemic.

Top

Suggested citation for this article: Brock N, Pulit-Penaloza JA, Belser JA, Pappas C, Sun X, Kieran TJ, et al. Avian influenza A(H5N1) isolated from dairy farm worker, Michigan. Emerg Infect Dis. 2025 June [date cited]. https://doi.org/10.3201/eid3106.250386