The reconstructed version of the flu virus that caused the 1918 world pandemic will be mailed to registered labs that ask for it, despite previous assurances to the contrary, Nature has learned.

The DNA sequence of the virus, which killed some 50 million people when it swept the globe in 1918–19, was reported by US researchers last month (J. K. Taubenberger et al. Nature 437, 889–893; 2005). At the same time, another group described how it used that sequence to reconstruct the complete virus for the first time (T. M. Tumpey et al. Science 310, 77–80; 2005).

Critics such as virologist Jens Kuhn of Harvard Medical School say that the virus should never have been recreated in the first place, as it could escape and cause another pandemic. Giving the virus to more labs, not to mention sending it in the mail, adds to that risk, they say. Others argue that such studies are safe, and can help to discover what made the virus so deadly. “We have to be careful,” says US health secretary Michael Leavitt. “But on the other hand we have to have the ability to study it.”

Scientists at the Centers for Disease Control and Prevention (CDC) in Atlanta, Georgia, where the virus is held, initially answered concerns that it might escape by emphasizing that only one person, microbiologist Terrence Tumpey, had access to it (see Nature 437, 794–795; 2005). Tumpey said that he had undergone extensive background security checks, and had to pass through tough safety procedures every time he entered the lab. The CDC also said that the virus would not be sent out to any other labs.

The agency now seems to have changed its mind. Spokesman Von Roebuck told Nature last week that labs that are registered to work with select agents — in particular, dangerous pathogens that are subject to specific handling rules — will be able to request the virus. The reconstructed flu virus was added to the CDC's select-agents list on 20 October.

The designation means that labs that operate under enhanced biosafety level-3 conditions or higher will be able to work with the virus. The highest biosafety level is 4, which requires full body suits. But enhanced level-3 conditions still require lab workers to wear respirators and to shower before leaving the lab. According to the select-agent rule, it is also “strongly recommended” that lab workers are vaccinated with the annual flu vaccine, which may offer partial protection.

No lab has yet formally requested the virus, Roebuck says. But some scientists say they would like to work with it. Michael Katze, a virologist at the University of Washington in Seattle, says there are plans to renovate his institution's primate centre so that his group can at least work with forms of the virus that contain some of the 1918 genes. Katze says he will study the virulence of the 1918 strain by infecting macaques with a mutated version of the virus. At some point, he also wants to work with the fully reconstructed virus, he says.

Scientists in Canada are planning to work with the virus, although they will not request it from the CDC. Constructs containing the virus's DNA will be made at the University of Wisconsin-Madison and will be sent to the National Microbiology Laboratory in Winnipeg, Canada, according to Frank Plummer, the laboratory's scientific director. Reconstructing the live virus from its DNA would then take just a few days, he says.

Making tracks: the 1918 flu virus could be transported between labs by commercial carriers. Credit: T. GLASGOW/AP

The researchers at the Winnipeg lab, which has level-4 certification, will infect mice and macaques to identify which parts of the virus make it so virulent. Such knowledge may enable researchers to identify potentially dangerous flu strains and develop vaccines against them. The work has been given extra urgency by the avian flu strains currently being carried by birds migrating from Asia. Researchers analysing the 1918 sequence have concluded that the virus was one that had jumped to humans from birds.

If it is sent out, the 1918 virus could be shipped by commercial carrier services that allow it to be tracked, says Clarence Peters, a virologist at the University of Texas Medical Branch in Galveston. Frozen samples of select agents are shipped in a plastic vial wrapped in absorbent material that would soak up any leaks. That vial is placed inside at least one other plastic container. This is held inside a polystyrene box containing dry ice, which itself lies inside a heavy cardboard box. Some of the packaging must pass tests such as free-fall drops and has survived air crashes, says Plummer: “It's very, very safe.”

Critics note that there is still a risk that the package could be lost or misdelivered, but Plummer says that is unlikely because the sender and recipient would know that the package is in transit. He adds that other select agents, including Ebola virus, are frequently sent in this way. If the virus did escape, it is unclear how serious the consequences would be (see ‘How bad would it be if the virus escaped?’).

Kuhn argues that the more places and people work with the virus, the greater the chance that it will escape or be stolen. He says that there should be an international agreement that restricts work on the virus to a few laboratories worldwide.

See Editorial, Flu in circulation