The virus had learned to target the udders almost exclusively
H5N1 caused severe mastitis in cattle udders rather than respiratory illness, evading early detection as veterinarians searched for bacterial causes. Specific glycan receptors in udder tissue—absent in airways—created a perfect breeding ground, allowing the virus to spread through contaminated milk.
- H5N1 infected cattle udders, causing severe mastitis, not respiratory illness
- Only N-linked sialic acid receptors bind H5N1; these were absent in cow airways but abundant in udders
- Infected cows shed large quantities of virus into milk, creating occupational risk for farm workers
- Outbreak began in Texas Panhandle dairy herds in early 2024
- Cats died from consuming raw milk from infected cattle
University of Pittsburgh researchers explain why H5N1 bird flu infected dairy cattle udders instead of lungs, revealing a framework to predict how the virus may surprise us next.
In early 2024, dairy farmers across the Texas Panhandle began reporting a strange affliction in their herds. Cows were developing severe mastitis—a painful inflammation of the udders that damaged mammary tissue and caused milk production to plummet. Veterinarians cycled through the usual suspects: bacterial infections, environmental pathogens, the standard culprits of industrial dairy farming. No one suspected bird flu. When H5N1 was finally identified as the cause, it upended everything the field thought it knew about how this virus behaves.
The discovery was jarring because H5N1 had never done this before. In birds, in humans, in other mammals that had contracted the virus, it attacked the respiratory system—lungs, trachea, nasal passages. But in cattle, it was different. The virus had somehow learned to target the udders almost exclusively, leaving the lungs largely untouched. This peculiarity meant that early detection was nearly impossible. Farmers and veterinarians were looking in the wrong place, searching for bacterial mastitis while a novel viral infection spread silently from herd to herd.
The delay in diagnosis had real consequences. Infected cows shed enormous quantities of virus into their milk. Farm workers handling the animals and their milk faced occupational exposure. Some farmers had the habit of feeding raw milk to domestic pets—cats in particular—and several cats died from the infection. The virus was moving faster than the response, contaminating environments and crossing species barriers in ways no one had anticipated.
A team at the University of Pittsburgh School of Public Health, led by Suresh Kuchipudi, set out to understand why H5N1 behaved so differently in cattle. Kuchipudi had spent his career studying how influenza viruses bind to different species and tissues, focusing on the role of glycans—sugar-based molecules that act like locks on cell surfaces, with the virus as the key. Earlier research had shown that the glycan receptors known to bind flu viruses were present in cow airways and lungs, just as they were in other mammals. Yet cattle weren't developing respiratory infections. Something else was happening.
Working with Lauren Pepi, a glycomics expert from Harvard Medical School, Kuchipudi's team deployed advanced imaging and binding experiments to map the architecture of glycan receptors with unprecedented detail. What they discovered was that not all glycan receptors function identically. Only one specific subtype—N-linked sialic acid receptors—could actually bind to H5N1. These receptors were virtually absent in cow airway tissue but abundant in udder tissue, making the mammary glands a perfect breeding ground for the virus while the lungs remained largely protected.
This finding opened a new door for public health. If scientists could understand which receptors a virus needs and where those receptors are located in different species and tissues, they could predict not just whether H5N1 might jump to a new host, but how it would behave once it did. Would it cause respiratory disease? Mastitis? Neurological symptoms, as the team had documented in cats? The framework transforms H5N1 from an unpredictable threat into something that can be anticipated.
Kuchipudi emphasized that the research provides a tool for preemptive screening. Scientists can now examine different species and different tissues within them to assess susceptibility before a crisis emerges. The goal is to avoid being caught by surprise again—to spot the virus's next unexpected move quickly enough that public health authorities can mount a response before the pathogen gains a foothold. The study, published in Science Advances, represents not just an explanation of what happened in 2024, but a map for preventing the next diagnostic blindness.
Notable Quotes
When the real culprit turned out to be bird flu, everyone in the field was caught completely by surprise. We hadn't even remotely considered that cattle could be a host for H5N1.— Suresh Kuchipudi, chair of Infectious Diseases and Microbiology at Pitt Public Health
We can preemptively screen different species and different tissues within them for susceptibility. The lessons learned could potentially help prevent us from being caught by surprise again.— Suresh Kuchipudi
The Hearth Conversation Another angle on the story
Why did it take so long to figure out that bird flu was infecting cattle udders instead of lungs?
Because veterinarians were looking for bacterial mastitis, which is the normal cause of that disease in dairy cattle. H5N1 had never done this before—it had never targeted udders in any species. The virus was hiding in plain sight, disguised as a familiar disease.
So the glycan receptors were there in the lungs all along, but the virus couldn't use them?
Exactly. The receptors were present, but only one specific type—N-linked sialic acid receptors—actually works as a functional lock for H5N1. The lungs had the wrong kind of lock. The udders had the right one.
What does this mean for the next time the virus jumps to a new animal?
It means we don't have to wait for an outbreak to figure out what's happening. We can screen tissues in advance, map their receptors, and predict where the virus will strike and how it will behave. We can be ready instead of reactive.
Is raw milk still a risk?
Yes, which is why pasteurization matters so much. The virus dies in heat. But raw milk—whether from infected cattle or fed to pets—remains dangerous. That's how cats got infected.
How many cattle were affected before they figured it out?
The source doesn't specify a total number, but it was enough to spread across multiple herds in the Texas Panhandle before identification. The delay meant the virus had time to establish itself and contaminate environments.
What's the next step?
Using this framework to screen other species and tissues for H5N1 susceptibility. The goal is to build a map of vulnerability before the virus exploits it.