MADISON, WIS—New genetic sequence data from both old and recent influenza strains provide clues to what determines viral virulence and may help to guide strategies for addressing future outbreaks.

Influenza viruses have crossed from avian to mammalian hosts numerous times. In a 1997 outbreak in Hong Kong, a virulent H5N1 influenza A strain passed from chickens to humans: Of 18 infected poultry workers, six died. No transfer between humans was observed, and killing the city’s entire chicken population successfully attenuated the epidemic.

“This outbreak was unique in that … an avian virus was directly transmitted from birds to humans,” notes Yoshihiro Kawaoka, DVM, PhD, Professor of Pathobiological Sciences at the University of Wisconsin School of Veterinary Medicine in Madison. To understand how a bird virus could infect and kill mammals, Dr. Kawaoka and colleagues examined the genes of several H5N1 strains isolated from human victims of the 1997 outbreak and correlated genetic features with the strains’ virulence in mice.[1]

While strain HK486 produced only nonlethal respiratory infection in mice, strain HK483 caused lethal systemic infection. By mixing and matching viral genes from lethal and nonlethal strains, Dr. Kawaoka and colleagues could determine which of the eight viral gene segments were responsible for HK483’s enhanced virulence. Combining the viral PB2 gene, which encodes an internal polymerase, from HK483 with the remaining segments from HK486 produced a deadly virus that infected brain, among other, tissues. Using either the HK486 PB2 gene or one version of the HK486 hemagglutinin (HA) gene to replace their respective HK483 counterparts, on the other hand, rendered HK483 benign and apparently capable of infecting only respiratory tissue.

Features of both HK483 virulence genes were consistent with functions of their homologues in other influenza strains. “HA plays a vital role in infection: It’s the glue that binds to the sialic acid on cell receptors,” explained Elspeth Garman, PhD, a University Research Lecturer in the Department of Biochemistry at the University of Oxford in England. Infection of target cells by an avian influenza A virus also requires cleavage of HA into HA1 and HA2. “The amino- terminal end of HA2 is responsible for fusion and penetration, and has to be exposed to actually initiate infection,” noted Robert G. Webster, PhD, a commentator on the research.[2] Similarly, Dr. Kawaoka and colleagues found HA cleavability associated with extreme virulence in H5N1.

PB2 sequences revealed that a particular amino acid substitution distinguished non-virulent from virulent gene versions. Interestingly, the identical substitution determines the host range in another influenza A virus. Further, PB2s from avian influenza

A strains resemble the non-virulent HK486 PB2 at this residue, whereas PB2s from human strains resemble the HK483 PB2.

1918 “SPANISH” FLU

Remarkably, new molecular data from 83-year-old tissue samples now reveal a glimpse of one of the most infamous strains in history, the “Spanish” flu, which killed more than 20 million people in a 1918 pandemic. In a commentary, Dr. Webster, Chairman of the Department of Virology and Molecular Biology at St. Jude Children’s Research Hospital in Memphis, pointed out that determining “the origin of the 1918 Spanish influenza pandemic … is still a work in progress.”

In an interview with PULMONARY REVIEWS, Dr. Garman reviewed data on the 1918 virus: “So far, only three of eight viral genes have been fully sequenced: HA, neuraminidase, and nonstructural genes. When the detective job of getting all eight gene segments is done, we should know a bit more,” she suggested.

Although influenza infection is typically limited to the lungs, historical accounts suggest that “the 1918 virus was associated with a unique form of encephalopathy,” Dr. Webster told PULMONARY REVIEWS. “[T]he polymerase genes haven’t yet been sequenced,” he also noted. Because changes in the polymerase PB2 are associated with alterations in host and tissue specificity, sequencing PB2 from the 1918 virus may yet explain its deadliness.

VACCINES FLAWED

Vaccination is the standard precaution to limit influenza outbreaks, but the strategy is hampered by influenza’s rapid mutation. “HA is an important component of influenza vaccines. The problem is that the HA changes its sequence regularly,” remarked Graeme Laver, PhD, Professor of Biochemistry at Australian National University in Canberra. Vaccines are updated yearly to include new strains, but with an inherent and variable lag, according to Dr. Garman: “You’ve got to gear up, raise the vaccine, and get people to come in for inoculation. That means the vaccines are always a year behind” and, thus, quite inadequate, should another 1918-like virus emerge. “There was a problem last year—the vaccines didn’t grow well,” she pointed out. And again, this year’s vaccine is projected to be late. “The shortage of vaccines in two consecutive years indicates that the United States is ill prepared for a pandemic,” Dr. Webster suggested.

“There’s a lot of discussion about how best to prepare for pandemic influenza,” which can stretch public resources, said Dr. Garman. Drugs may be a better bet than vaccines, she and Dr. Laver argue in their commentary.[3] Inhibitors of the viral neuraminidase enzyme, which cleaves sialic acid on cell receptors to allow the infection to spread, offer promise; zanamivir (inhaled) and oseltamivir (oral) are two such drugs now in use. “Stockpile these drugs now in huge quantities” to prepare for the next influenza pandemic, Drs. Garman and Laver counsel public health administrators. “The neuraminidase inhibitors could be relatively cheap, and should have a long shelf life, although this is still unknown,” explained Dr. Garman.

However, Dr. Laver cautioned, “The drugs need to be started very early. They’re very effective at stopping the virus from replicating, but once you’re infected, there’s already damage.” Although prophylactic use of a neuraminidase inhibitor against ordinary strains may be excessive for healthy persons, “for people ... nursing those with the flu, it makes sense to take it,” said Dr. Garman. Because the drugs are highly specific for influenza neuraminidase, they are useless against other viruses. And because prescribing neuraminidase inhibitors to all patients with early influenza symptoms is costly and might delay treatment for other infections, Drs. Garman and Laver suggest adopting various low-cost, rapid influenza diagnostic tests now under development.

—Mimi Zucker, PhD

References
1. Hatta M, Gao P, Halfmann P, Kawaoka Y. Molecular basis for high virulence of Hong Kong H5N1 influenza A viruses. Science. 2001;293: 1840-1842.
2. Webster RG. Virology: a molecular whodunit. Science. 2001;293:1773-1775.
3. Laver G, Garman E. Virology: the origin and control of pandemic influenza. Science. 2001;293: 1776-1777.

Return to table of contents