One of the people who closely watched the wildfire-like spread of H5N1 was Cummings School professor Jonathan Runstadler, then an assistant biology professor at the University of Alaska Fairbanks. In 2004, Runstadler had just started his lab and wasn’t focused on influenza; he was looking more broadly at wildlife disease patterns, examining ducks, geese, sea birds, and other species and hoping to determine why certain animals are more susceptible to illness than others. But Runstadler couldn’t ignore the H5N1 outbreak, which quickly spread to more than 50 other countries in Asia, Africa, and Europe and would eventually kill an estimated 700 people. “It was a huge cause for concern,” Runstadler recalled, “because nothing like that had happened with influenza for some time.”

The last global flu pandemic had taken place more than three decades earlier, when the “Hong Kong flu” spread from China to other countries in Asia, Europe, and North America, killing one million people over 1968 and 1969. And that was nothing compared with the estimated 50 million people who died from the “Spanish flu” of 1918 and 1919, which killed more people over two years than the Black Death killed over a century. Since those pandemics, researchers had come to understand that the common link among the flu viruses—indeed, among all 116 types of flu known to infect humans and animals—is that all can be carried by wild birds. Runstadler’s fear about H5N1 around 2004, he recalled, was that if it wasn’t contained, “migratory birds were going to eventually move it all around the world.”

Thai Children

Thai children wear masks to protect against influenza in Bangkok in 2009. Photo: PORNCHAI KITTIWONGSAKUL/AFP/Getty Images

At the time, this was not a concern shared by all in the scientific community. The traditional theory was that avian influenza probably couldn’t even survive in hostile environments like Alaska, much less be carried all the way to North America by seemingly healthy birds. This line of thinking was reinforced when H5N1 didn’t become a full-blown pandemic—the only human case of it ever detected in the Americas was in a traveler from China. To those experts, viruses like H5N1 really had only two ways of invading: from infected human travelers or from importing infected birds. But Runstadler wondered if migrating birds could, in time, prove a third, hidden travel route for the virus.

From his perch atop North America—at wild birds’ migratory crossroads between Southeast Asia and the continental U.S.—Runstadler recognized that he was uniquely positioned to detect the influenza virus if it was going to come through the Arctic. If that new pathway could be proved possible, the knowledge would be essential in helping prepare researchers and doctors for the future. Because even though H5N1 didn’t become a pandemic, most experts agreed then, and now, that we will see one again.


Gathering Eggs

A farmer collects eggs at an Illinois poultry farm in 2015, after a devastating outbreak of avian influenza in the midwestern United States caused prices to skyrocket. Photo: Daniel Acker/Bloomberg via Getty Images

Many of us think of the seasonal flu as a particularly bad cold: You feel miserable for a while, and then you get over it. Often, that is more or less true. Although seasonal human influenza sickens an estimated 25 million Americans every year, most have acquired a degree of immunity to the regularly circulating strains of flu and recover without even needing to visit a hospital. Yet in those with weaker immune systems—the very young, the old, and others—it can turn deadly. Twelve thousand people died during the 2015-16 flu season; if not for the flu vaccine, the CDC estimates the death toll would have been 3,000 lives higher.

But every so often a wholly new strain of influenza surfaces that kills even young, previously healthy people. Such new strains are very rare, but far more dangerous than the seasonal flu, because large numbers of people haven’t built up any immunity to them. And unlike the winter arrival of seasonal flu, so far the only thing predictable about a deadly strain is that it will emerge unpredictably. Jonathan Runstadler and his team of researchers are on a mission to help change that.

Every so often a wholly new strain of influenza surfaces that kills even young, previously healthy people.

After the emergence of H5N1 in 2004, Runstadler pivoted his research. Allying with ornithologists and ecologists doing field work, plus a handful of laboratory scientists down in the Lower 48 states, he wanted to see whether they could find avian influenza in Alaska. At the time, there had only been one report of it that far north, from the 1990s. Testing large numbers of samples in a then novel way, with a molecular method referred to as RT-PCR (for “Reverse Transcriptase – Polymerase Chain Reaction”), Runstadler searched for evidence of influenza in the birds and found it. Lots of it.

Now, Runstadler thought to himself, the interesting part starts.

Over two years, he built up a lab equipped to grow flu strains he collected from fieldwork and analyze their genomes. The viruses they had found in Alaska, Runstadler and his fellow investigators showed, were intertwined with the flu’s circulation throughout the rest of the world. “There were enormous populations of influenza viruses moving around up north, where they were amplified or originated in birds hatched in a breeding season,” he said. “The migration of birds would then rain these new viruses down on the southern latitudes.”

Holding a duck

Nichola Hill (above) and the other Cummings researchers hope their data will help scientists identify parts of the viral particles that remain consistent across all flu strains—so they can develop a universal vaccine. Photo: Lillie Paquette

The discovery shifted the way scientists viewed the flu lifecycle (Runstandler’s paper on RT-PCR detection of influenza, meanwhile, helped establish it as the gold standard for detecting flu viruses). And when the National Institutes of Health in 2007 announced plans to establish a network of influenza centers working to track the flu—an initiative called the National Institutes of Health Centers of Excellence in Influenza Research and Surveillance—Runstadler answered the call.

In fall of 2011, Runstadler moved his lab to the Massachusetts Institute of Technology, where he hired Wendy Puryear and Nichola Hill. They study and sample animals out in the wild, including in Alaska and coastal Massachusetts, and then return to the lab to analyze the samples for influenza viruses. In 2017, Runstadler and his team moved to Cummings School, attracted by the active group of researchers studying wildlife and infectious disease there, as well as by the Tufts New England Regional Biosafety Laboratory, a level-three biosafety facility where the team can test for viruses, which can’t be spotted by eye in the birds carrying them. “We can hold a bird in our hands, and it will be completely asymptomatic to our eyes,” Runstadler said. All the clinical signs of influenza “don’t apply to a wild animal that has coevolved with the virus for so long and figured out a defense strategy to coexist.”

‘The evolution of flu happens at an unimaginably rapid scale,’ said Cummings School researcher Nichola Hill.

While much of the team’s research involves how the flu infects and evolves within animal hosts, these insights can help better safeguard human health. One avenue toward this goal is tracking and analyzing the many strains of flu found in nature to help develop an even more effective vaccine. A major challenge to a vaccine is the fact that influenza is constantly evolving. When it infects a host’s cells, it reproduces by quickly copying itself thousands of times over in a matter of hours, rather than years or decades—the timescale of human generations. “The evolution of flu happens at an unimaginably rapid scale,” said Hill, a senior research associate at Cummings School, “making it an ideal system for studying evolutionary processes, such as natural selection.” Even though the speedy replication process is sloppy, with the virus making many mistakes, some of those duplication errors can lead to mutations that are evolutionarily advantageous. The flu strain may change in such a way that it can appear new to a host’s immune system, escape its defenses and live on to cause new infections.

Examine Duck

Nichola Hill (left) and the other Cummings School researchers head out into the field every year to collect thousands of samples from water birds and other wild animals. Photo: Jonathan Runstadler

To track how flu viruses are changing involves gathering immense volumes of data. The Cummings researchers head out into the field every summer to collect thousands of samples from all manner of ducks, geese and other water birds, as well as other animals. They then feed their data into a genomic database funded by the National Institutes of Health. “NIH would like to know how much diversity there is versus how many of these strains are stable in a population,” Hill said. The team hopes that scientists will one day be able to identify parts of the viral particles that remain consistent across all strains to help develop a universal vaccine.

Another avenue of research for the Runstadler Lab is uncovering what environmental and other factors increase the flu’s pandemic potential. The flu is unique in that it can jump across species. “It’s usually difficult for viruses, and even for some other types of pathogens, to cross species boundaries,” Runstadler said. “But influenza seems to have been very successful in infecting lots of different species and lots of different groups of animals.”

A pandemic strain of flu can result when this species-hopping ability combines with a genetic-mixing process called “reassortment” (Hill calls it “virus sex”). Reassortment takes place when two viruses coinfect the same host cell, swap segments of their genetic code, and produce a virus that’s altogether new. For example, if a deadly avian influenza virus crossbreeds with a strain easily spread among humans, the resulting offspring could be a virus that infects many people, all of whom would have no immune history to fight it off.

This happened in April 2009 when an avian influenza spread from wild birds to poultry, mutated in a pig, and then made the leap into humans. From its emergence in Mexico, the H1N1 “swine flu” spread rapidly around the world and, within three months, the World Health Organization declared it the first pandemic in more than 40 years. It is estimated to have killed more than 240,000 people, most of them children and younger adults.

Since most strains of H1N1 circulate regularly (and relatively harmlessly) through humans, the pandemic caught many by surprise. “It was probably the last on everybody’s list of what to expect as the next emerging threat,” Runstadler said. His team’s goal is to reduce that element of surprise in the future.


Even after the H5N1 bird flu that savaged Thailand and other countries in 2004 and 2005 receded, it never stopped evolving. Over time, “the original H5N1 became endemic in some wild bird and poultry populations in Southeast Asia and China,” Runstadler said, and H5N1 also produced two new lines of offspring. In December 2014, one of those H5N1 descendants—the H5N8 avian flu—crossed from British Columbia into Washington State.

H5N8 tore through Midwestern poultry farms with alarming speed. On March 4, 2015, it was detected in Minnesota, five days later it was in Missouri, and it crossed into Arkansas and Kansas less than a week after that. By mid-April, the H5N8 had infected an Iowa farm that was the nation’s third-largest producer of eggs. The cost of stopping the spread, to prevent even worse outcomes, was astronomical: more than 8 million birds needed to be destroyed. “Fifteen years of work,” the farm’s president told the New York Times. “Gone in a week.” The newspaper reported that H5N8 caused the loss of more than 15,000 jobs and cost U.S. businesses $2.6 billion in sales.

For all its destruction to the poultry industry and the birds themselves, the virus did not affect people. “Given the information at this time, the risk of human infection [from H5N8] is low, but cannot be excluded,” concluded a 2016 WHO update that called for continued surveillance. After all, the “H5N8 strains belong to a group of viruses that, in the right instances, can cause human infections and severe pathology, if not death,” Runstadler said.

As the poultry pandemic raged on, Runstadler’s team was still studying birds in Alaska and, with additional funding from NIH to expand their work, they worked to understand the origins of the H5N8 virus in North America. After doing a kind of genetics forensic investigation, Hill and her colleagues showed that the H5N8 virus had traveled with migrating birds from Southeast Asia into North America. Their analysis, which mapped the highly pathogenic poultry viruses’ evolutionary family tree in Emerging Infectious Diseases in April 2017, revealed that reassortment between viruses in wild birds breeding in Alaska had then helped the highly deadly H5N8 virus get a foothold in North America and spread through Canada and the U.S.

Seals on beach

Seals had periodically been struck by influenza, and it had always just been assumed that they got the flu from wild birds. Photo: Margot Madden (Permit # NMSF 17670)

Alaska attracts huge numbers of migratory birds from Eurasia and North America that fly north to breed, making it a unique spot for cross-breeding viruses. “We see many viruses that are partly North American and partly Eurasian in their genetic composition,” Hill said. The new flu strains that result are then able to spread rapidly outward into North America and Eurasia, aided each year by an eruption of susceptible young birds to infect. Dangerous H5 subtype strains from Southeast Asia could travel across the Bering Strait and through North America via migrating birds, Runstadler’s team had shown—the very pathway he had theorized about a decade earlier with the H5N1 outbreak.

This essential new insight requires researchers to broaden the pathways of infection they are tracking. “It’s important for biosecurity that we understand that this bird flu was not something that came from a bird imported from Asia or someone traveling here with the virus,” Runstadler said. How well we predict the next introduction of a harmful strain of influenza, he continued, will hinge on our understanding of the ways that different animal populations, including humans, coming together in one place can create conditions ripe for the virus to evolve and spread.

After seals died from the flu, Runstadler’s team decided to challenge another tenet of conventional wisdom.

For years, Runstadler and his team have demonstrated that the flu living and mutating in populations of migratory birds could be carried to other parts of the world by way of Alaska, something many didn’t think was possible. After 162 harbor seals on the New England coast died from the flu in 2011, Runstadler’s team decided to challenge another tenet of conventional wisdom.

Seals had periodically been struck by influenza, and it had always just been assumed that they got the flu from wild birds, said Wendy Puryear, senior research associate in Runstadler’s lab. “We came into the picture saying, ‘Well, nobody’s really looked, so how do we know that it’s not something that’s circulating all the time?’”

To answer that question, Puryear every winter takes a team by boat to Muskeget and Monomoy islands off Cape Cod, the second and third largest seal-pupping colonies in the world. The temperatures are frigid. “We wear these big, crazy survival suits,” Puryear said. “You feel like the Sta-Puf man.” Working in pairs, Puryear and her colleagues carefully sneak up on a pup, scoop it up in a special bag, and carry it out of sight of the rest of the animals. The team takes photographs, measurements, and various kinds of samples to help support as many other marine mammal research projects as possible. Then they release the pup back where they found it. After that, they return to the lab to look for evidence of the flu.

Blood vials

Wendy Puryear takes a team to islands off Cape Cod to test seals for flu viruses. They’ve found influenza in 5 to 15 percent of the gray seals, which is comparable to what they find in wild birds. Photo: Milton Levin

So far, Runstadler’s team has found influenza in 5 to 15 percent of the gray seals it has sampled, which is comparable to the 5 to 20 percent it finds in wild birds. The findings are intriguing, said Puryear, because if the seals are contracting influenza from the birds, that may mimic how it has often evolved to pass between species and into humans. In the 2009 pandemic, it was close contact among wild birds, chickens, and pigs that led to an avian form of the virus spilling over into swine. That mammalian host provided just the right launching pad for swine flu to jump into humans. “Now,” Puryear said, “we’ve got the influenza virus potentially bouncing back and forth between sea birds and another mammal: seals.” The Cummings team is looking for new ways to grow enough of the virus sampled from seals to genetically sequence it and compare it with types found in other mammals.

Straddling a seal

In the 2009 pandemic, it was close contact among wild birds, chickens, and pigs that led to an avian form of the virus spilling over into swine. “Now, we’ve got the influenza virus potentially bouncing back and forth between sea birds and another mammal: seals,” said Wendy Puryear (above). Photo: Margot Madden (Permit # NMSF 17670).

What seems to be true from other studies Runstadler’s team has done is that reassortment events appear critical to flu viruses’ ability to move from one species to another. And knowing the hazards of reassortment events can also point to human behavior that is changing the environment in a way that might inadvertently help a future pandemic. For example, during the November to March scallop season on Cape Cod and the islands, fishermen shuck and discard the shellfish waste onto designated piles onshore, not far from seals’ pupping grounds. “It’s like something out of a Hitchcock movie,” Puryear said. “Hundreds of birds can flock down to the scallop pile for a feeding frenzy.” The Tufts researchers have found far more influenza around these waste piles than they’d normally expect. “We’re artificially congregating these animals, which may allow these transmission spikes to happen, and doing this is in close proximity to where seal pups are being born,” Puryear said. Understanding the relationship of human activity to influenza circulation and transmission in animals may prove critical to tracking and preventing future outbreaks.

The Cummings School researchers aren’t about to rest until they’ve figured it out. “The gray seal flu viruses appear to be circulating in a host that’s not only undergoing tremendous population growth, but also increasingly mixing with other seal populations,” Runstadler said. Meanwhile, humans are coming into closer contact with seals, both through the fishing industry and on beaches. And it’s possible that only one interaction between a seal host and a human can transfer a virus that could have dangerous repercussions, Runstadler said. “These are all reasons we need to keep working on this puzzle.”

Contact Genevieve Rajewski, the editor of Cummings Veterinary Medicine, at