Stalking The Mutating Monster

IT'S GRANT-WRITING TIME, and Pekosz's fifth floor office is covered with papers. Federal funding for research on influenza has steadily risen since 2003, when the H5N1 influenza virus struck poultry in Asia. Public health authorities feared that avian influenza—the "bird flu"—could mutate into a virus that could spread from person to person and spark a pandemic.

It hasn't. But if it did, the impact could be devastating because the human population has no immunity to the H5 strain. Various scientists and government agencies predict a pandemic could kill tens to hundreds of millions of people worldwide, and cost billions of dollars. Although the precise figures vary, what's not in dispute is the merciless nature of pandemic flu. Witnesses to the 1918 global influenza pandemic described an illness that progressed with violent speed—faces turning bluish-black as oxygen drained from the victims' blood; patients coughing so hard they tore abdominal muscles and rib cartilage; others gasping for breath as fluid filled their lungs.

Pekosz is acutely aware of the "15 minutes of fame" accorded new and newly emerging diseases. In addition to influenza, he also works on SARS, the virus that received a flurry of attention when it struck in 2003. "It made big headlines," he recalls. "There was tremendous worry and concern bordering on hysteria." These days, he asks, "when was the last time you heard about SARS?"

Likewise, Pekosz worries that concern about pandemic flu may dissolve from people's consciousness, although he and many other influenza experts believe the threat remains. He flips through the papers on his desk, apologizing for the clutter, then picks up a copy of an editorial that appeared recently in the journal Nature. It asserts that another flu pandemic is "inevitable."

In an office with few decorations, one figure on a shelf high above Pekosz's desk attracts the eye. It is the head of an Icelandic sheep with impressive horns, gazing outward. It came from Neal Nathanson, one of his two mentors at the University of Pennsylvania, where Pekosz earned his PhD. Nathanson, a legendary virologist, spent many years studying a retrovirus that is endemic in the breed. Because the virus does not infect people, some might question the merits of such research. But such basic science is essential to medicine, says Pekosz. Scientists can often infer how a virus will affect people based on studies that show how a similar virus interacts with an animal.

Pekosz's own research in graduate school focused on the mosquito-borne La Crosse virus (a major cause of pediatric encephalitis in North America). Since then, he has specialized in human respiratory viruses, such as influenza and SARS, always attacking questions from a basic science angle. Even his current research on an influenza vaccine grew out of earlier studies aimed at understanding the virus's basic components, its genes and proteins. His research motto might have been, Know your enemy—intimately, molecule by molecule.

UP CLOSE, THIS tiny organism—a mere 100 nanometers in diameter—looks like a spherical pincushion. This sphere contains the virus's genetic material, RNA—home to those eight genes—while the "pins" poking from this cushion consist of various types of protein macromolecules. These proteins are the reason scientists have struggled so hard to combat influenza.

Like the craftiest fighter in the ring who has unlimited moves for dodging an opponent, influenza has an extraordinary ability to change the composition of its surface proteins. This talent, the result of an exceptionally high genetic mutation rate, enables influenza repeatedly to evade the "enemy," whether that enemy be antiviral drugs or the human immune system.

"It's almost inevitable," says Pekosz. "You introduce one pressure, and the virus will find a way around it." To illustrate this phenomenon, he escorts a visitor down six flights of stairs to his basement laboratory. The cramped rooms contain the customary centrifuges, incubators and flow hoods. Stacked throughout the lab, on shelves and countertops, lie scores of Petri dishes. Pekosz indicates a stack of plates in a corner of the lab. Each is divided into 96 individual wells.

The wells contain influenza virus and mammalian cells that are prone to viral infection, along with various doses of the antiviral drug amantadine. In addition, each well contains a special dye used to gauge cell vitality; in the presence of live cells, the dye is activated and will reflect blue light.

Pekosz holds up one of the dishes. Some of its wells glow bright blue. Others are clear. The clear wells, he notes, indicate cell death due to the appearance of mutant viruses that have resistance to amantadine.

They took only three days to emerge.

THE CAPRICIOUS NATURE of influenza's protein surface also poses a challenge to vaccine developers. Two classes of proteins figure most significantly. One is called the hemagglutinins (HAs), which influenza uses to latch onto cells. Scientists have identified 16 different HAs, known as H1 through H16. They have also identified nine members of another class of surface proteins known as the neuraminadases (NAs), designated N1 through N9. Flu subtypes are named according to the HAs and NAs that populate their surface. So bird flu is influenza subtype H5N1. Subtypes now circulating in the population include H1N1 and H3N2.

One might suppose that a vaccine targeting the most prominent influenza subtypes would work well and last long. But influenza has not allowed such an easy solution. Thanks to its high rate of mutation, influenza's HA proteins constantly change or "drift" into subtly different forms. This process, known as antigenic drift, results in myriad strains. So, for example, there is not just one strain with the appellation H1N1; there are many.

To craft a seasonal flu vaccine, scientists predict which influenza strains are likely to circulate in the coming season. They then use killed or weakened versions of those strains to construct a vaccine that will stimulate an immune response targeting those particular strains. If all goes well, when a vaccinated person comes into contact with a live version of those influenza strains, their immune system launches an attack that neutralizes the virus.

But this approach has its limitations. First, because flu strains constantly drift, vaccine developers must reformulate their product every year. And even then, the process is not always perfect. Sometimes the experts err in their predictions, and fail to include a strain in the vaccine that ends up causing infections.

Influenza has another trick, however, with even graver consequences, a process called antigenic shift. It occurs when an animal strain of influenza mutates or when a human influenza strain and an animal influenza strain mix genes, giving rise to a novel influenza subtype. Introduced into a population with no baseline immunity against that subtype, the virus would spread extremely quickly and cause severe illness. It would launch a pandemic.

After the 2003 bird flu outbreak and the ensuing fear of a flu pandemic, researchers increased efforts to design a pandemic flu vaccine. That task has proven more difficult than researchers had anticipated, says Ruth Karron, MD, director of the School's Center for Immunization Research. She has overseen clinical trials of several of the candidate live, attenuated vaccines, which included components of viruses thought to have pandemic potential, including H5N1. While the live, attenuated vaccines performed well in animal studies, some were overattenuated (overly weakened) in human volunteers, and did not induce a strong immune response, says Karron, an International Health professor.

Research teams are now using those initial results to improve their vaccines. Several of the candidates look promising, says Karron. Pekosz, however, has taken a different strategy. Instead of focusing on influenza's HAs and NAs, he's turned his attention to a different protein called M2.

It turns out that the M2 protein, unlike the HAs and NAs, is virtually the same in all strains of influenza, including those with pandemic potential. In a biologist's parlance, M2 is "highly conserved." That fact has led Pekosz (and several other researchers in the field) to ponder whether a vaccine built around the M2 protein would stimulate immunity against all strains of influenza.

In other words, it would be a universal influenza vaccine.

With this hope, Pekosz is engineering two types of vaccines based on the M2 principle. One is an attenuated vaccine, a vaccine that contains a live but crippled virus. (The FluMist vaccine available for seasonal flu, for example, is an attenuated vaccine.) For this approach, Pekosz genetically manipulates the M2 gene to result in a virus that contains a longer-than-normal M2 protein. The goal is to craft a virus that can replicate just enough to provoke the immune system, but not enough to cause a full-blown infection.

So far, says Pekosz, animal studies conducted over the past few years have revealed promising results for several of his attenuated vaccines. The experiments involved inoculating laboratory mice with the experimental vaccines, and then exposing the mice to two different strains of influenza that have recently circulated in the human population. The mice remained healthy, while unvaccinated mice serving as a control group succumbed to the infection. In fact, says Pekosz, the experimental vaccines appear to provide even stronger immune protection than conventional flu vaccines.

Pekosz views those results as one of the small victories that sustain scientists. "If you're fortunate, those moments can feed you fuel to go another couple years."

While he's demonstrated a proof of principle, says Pekosz, much work likes ahead. Ultimately, he hopes his strategy will yield a vaccine that protects against three, six, twelve or more strains of the virus. He does not yet know whether his current experimental vaccines will confer such broad immunity. Achieving that goal may take several years. "It's a work in progress," he says.

But other challenges lie ahead. One relates to the fact that mice are not people. While mice can crank out a robust immune response to M2, people do not. No one yet understands why not. So in his second line of vaccine research, Pekosz is seeking a way to boost that weak immune response.

For this approach, the virologist is constructing vaccines that combine the M2 protein with a core protein from the hepatitis virus. Scientists have found that the hepatitis virus is especially skilled at "showing" protein antigen to the immune system. If his plan works as he hopes it will, the hepatitis protein will showcase the M2 protein, and the immune system will "see" the show and will respond by making a generous amount of M2 antibody. Initial tests of the vaccine in mice are encouraging, says Pekosz. Clinical tests of the vaccine may be years away.

The tricky part for Andy Pekosz is not knowing if he's pursuing a reasonable but ultimately futile path. "It is certainly possible to go down a dead-end," he says. "The challenge is knowing when to back up."

Some of Pekosz's colleagues underscore the difficulty of his task. They include his postdoctoral advisor, Northwestern University virologist Robert Lamb, who discovered the M2 gene and protein. If the immune system could target M2, he asks, "wouldn't nature have done it already?"

The ultimate test of his vaccines, however, will be to see if they protect against pandemic strains of influenza. Such studies require extra levels of safety precautions. So for the last leg of his tour, Pekosz takes his visitor to another laboratory that he has been retrofitting as a Biosafety Level 3 lab. The lab must meet the extra levels of safety precautions mandated by the CDC for labs handling potentially lethal respiratory viruses. So workers have been installing special air ventilation systems, door seals, autoclaves and high-level security systems. A box of disposable paper booties sits outside the entrance to the lab's main chamber. Entering the chamber requires full protective gear: foot coverings, coveralls, outer apron, gloves, as well as a hood and face mask connected to an air pressure monitor that sends a constant stream of air flowing away from the face to prevent the inhalation of airborne virus. The protective gear can get hot and uncomfortable. Researchers will have to work for several hours without a break to avoid repeated changes. (Pekosz has calculated each single-use outfit costs about $14.) "It's not easy work," he says.

The goal, however, supersedes the difficulties of the work and uncertainties of the results. "We have a tendency to forget how frightening pandemics are," says Pekosz. Most people have not experienced pandemic flu firsthand. But anyone who has suffered through a bout of seasonal flu has had a milder preview.

Pekosz himself came down with a textbook case of flu about three years ago—high fever, body aches, a dry hacking cough, aches and pains, and listlessness. He was miserable. A pandemic strain would cause a vastly more severe illness, and Pekosz can only imagine such agony.

"I'd been a good boy and had gotten my vaccination," says Pekosz. "I still got flu." Apparently, he'd gotten infected by an antigenic-drift variant. That knowledge compounded his misery at the time, says Pekosz, but also reaffirmed his conviction in his research. A universal flu vaccine might have prevented him from getting sick.

"It's an ongoing battle," he says. "There must be a way to prevent the virus from circulating. It's just a matter of finding the right approaches."