Killer findings: scientists piece together 1918-flu virus.

The "Spanish" flu killed an estimated 20 million to 50 million people worldwide between 1918 and 1919. Hoping to prevent such a deadly outbreak from recurring, scientists have long strived to figure out what characteristics differentiate that strain from other, more-benign varieties. Because researchers have lacked live samples of the killer virus, however, they couldn't answer this pivotal question.

Two new studies now shed unprecedented light on the 1918 strain.

The first study caps a 9-year effort led by Jeffrey Taubenberger of the Armed Forces Institute of Pathology in Rockville, Md., to attain a complete genome sequence for the 1918 strain (SN: 3/22/97, p. 172). Taubenberger and his colleagues collected virus particles from samples that had been preserved after autopsies of 1918-flu casualties and from a single additional victim interred in the Alaskan permafrost.

The virus had long since degraded in these samples, but it left behind tiny bits of RNA that encode the virus' eight major gene segments.

Previously, Taubenberger and his colleagues had used these RNA fragments to sequence five of the virus' gene segments. In the Oct. 6 Nature, the team reveals the final three sequences. The genes in these blocks code for the 1918 flu's polymerases, which are proteins crucial for a virus' replication in animal hosts.

Taubenberger's team found a striking resemblance between the 1918 virus and modern bird-flu strains, including the deadly HSN1 strain currently circulating in Southeast Asia (SN: 9/10/05, p. 171). These results add to evidence suggesting that the 1918 flu originated as an avian strain that acquired the capacity to infect people.

Taubenberger notes that figuring out how the virus adapted to human hosts could aid researchers in preventing modern bird-flu outbreaks from becoming pandemics in people. "If we could identify which [parts of the genome] are important in adaptation, we could provide a checklist for surveillance of strains just beginning to show adaptation to humans," he says.

Using the just-completed sequence, a team led by Terrence Tumpey of the Centers for Disease Control and Prevention in Atlanta has partially reconstructed the 1918 virus.

The researchers synthesized the eight major pieces of RNA according to the virus' genetic code. They then combined them with bits of RNA from related flu viruses. The latter genetic material enabled animal cells to read the viral genes.

Working under biosafety level 3, the second-highest level of protection against biohazards, Tumpey's team found that the reconstructed virus killed otherwise-healthy mite in 3 to 5 days. It was also lethal to chicken embryos developing inside eggs, supporting its likely origin as an avian-flu virus. When the scientists infected samples of human-lung cells with the virus, it replicated readily.

By mixing and matching the 1918 virus' genes with genes from contemporary-flu strains, the team found that the 1918 virus' polymerase genes and its hemagglutinin gene, which sneaks the virus into cells, seem to play pivotal roles in virulence. The group published its results in the Oct. 7 Science.

Flu researcher Yoshihiro Kawaoka of the University of Wisconsin-Madison suggests is that he and other researchers can now determine how the virus' genes made the 1918 flu so deadly. With that information, scientists could craft new vaccines and drugs to combat future outbreaks.

"By understanding why this [strain] was so pathogenic, we tan prepare for the potential of future viruses that may appear," Kawaoka says.