The outbreak of poliomyelitis that hit the Caribbean island of Hispaniola last summer took everyone by surprise. The news that weakened, live viruses in the oral polio vaccine had reverted to a form that can initiate a disease outbreak is forcing the World Health Organization to re-examine the final stages of its global campaign to eradicate polio (see page 278).

Thankfully, the outbreak amounted to no more than eight cases. But it has exposed the limits of our knowledge about the polio virus. The genetic determinants of its transmissibility, for instance, remain obscure. And they seem likely to remain so, with polio now a low priority for the world's biomedical research agencies, which view it as a 'conquered' disease.

This attitude is short-sighted, because viruses have a habit of producing unpleasant surprises. In 1996 and 1997, for instance, as civil war raged in what is now the Democratic Republic of the Congo, the country was hit by a substantial outbreak of monkeypox. In Central Africa, contact with primates means that people do catch this disease from time to time. But the severity of the outbreak sparked fears that the virus had mutated into a more pathogenic form, or had acquired an enhanced ability to spread from person to person.

In the event, this appeared not to have happened. But some virologists argue that the threat is real, as the global eradication of smallpox — completed more than two decades ago — has left the door open for the related monkeypox virus to evolve and fill its vacant niche. Once more, however, testing this theory does not seem to be high on the list of biomedical research priorities.

The last official samples of smallpox reside in secure facilities in the United States and Russia. But other stocks may exist in undeclared bioweapons laboratories — and the threat of biowarfare or bioterrorism provides another reason to maintain a broad view of the potential threat posed by viral diseases.

This was underlined last week by the revelation that a team in Australia had inadvertently created an unusually lethal form of mousepox. The researchers were trying to create a contraceptive vaccine, using a benign mousepox virus to express genes for proteins carried on mouse eggs. They also added the gene for interleukin-4 (IL-4), in an attempt to boost the antibody response against these proteins. But a paper in the latest issue of the Journal of Virology (75, 1205–1210; 2001) describes how the IL-4 gene also shut down the cellular arm of the animals' immune systems. As a result, the virus killed the vaccinated mice within days. Such findings show that it may not take much for a malevolent bioengineer to create a hideously effective weapon.

Advances in fields from immunology to epidemiological modelling mean that we are in a better position than ever to respond to the threats posed by viral evolution and bioterrorism. High-throughput gene sequencing, in particular, means that a virus's genetic secrets can be rapidly laid bare. But deploying these tools effectively may require a strategy of expecting the unexpected. Apparently conquered or benign viruses cannot simply be ignored.