Man's best virus

ﾩ H.C. ROBINSON / PHOTO RESEARCHERS, INC.

It might be considered the cat's revenge on the dog that chased it around the house and yard: Sometime in the late 1960s or 1970s, deadly feline parvovirus jumped from cats to dogs, becoming canine parvovirus. Then, in 1978, it started killing puppies at an alarming rate. "Most viruses go into a new host and just die out," says Laura Shackelton, a postdoc at Pennsylvania State University, who has studied the evolution of both subgroups. "This one took off."

In both species, parvovirus causes serious problems for newborns and can be fatal: Kittens develop hypoplasia, and puppies sicken with myocarditis. Older animals can suffer some effects, but they are mild and temporary. The virus can stay active in soil for long periods of time, so dogs must be vaccinated close to birth, immediately after maternal antibodies wane and before they start digging for whatever they dig for.

The carnivore parvovirus subgroup, as it is called, also includes variants that can infect the blue fox, raccoon, and mink. The epidemiology and small size of the viral genome has made it an excellent model for looking at cross-species convergence, says Shackelton. What happened to the virus was particularly compelling: CPV2, the first known canine strain, infected feline cells in culture, but didn't infect cats. As the subgroup population changed, however, with CPV2a replacing CPV2, the virus regained the ability to infect both cats and dogs. The residues that changed between CPV2 and CPV2a, however, were not the same as those that had changed when the virus jumped from cats to dogs. "This wasn't a reversion," she notes.

Shackelton and Andrew Rambaut, her advisor at Oxford where she did a postdoc, set out to better characterize those mutations. Perhaps somewhat fittingly, they used the Bayesian Evolutionary Analysis Sampling Trees (BEAST) program to analyze sequences of the canine and feline parvoviruses. Using a Markov Chain Monte Carlo approach to look at samples of feline parvovirus dating back to the 1960s, they found that the rate of evolution was closer to that of RNA viruses than that of DNA viruses. "Everyone assumed the underlying error rate during replication was similar to that of their host cells because the virus replicates with the host polymerase," Shackelton says of the small eukaryotic DNA viruses, but that doesn't seem to be the case. She and colleagues published their results in the Proceedings of the National Academy of Sciences in January 2005 (102:379-84).

Building on the knowledge that parvovirus mutation rates were not as slow as that of larger DNA viruses such as the herpesviruses, as everyone had thought, Shackelford has also turned her attention to human parvoviruses. She naturally wondered first whether the carnivore parvovirus mutation rate was similar to the human rate. "Was the average rate just due to huge selection pressures during emergence, or is there a consistently elevated rate of mutation in all parvoviruses, all single-stranded DNA viruses, or even all small DNA viruses?"

A large amount of parvovirus B19 was available. In children, that virus causes Fifth disease, a mild rash on the body and typically described by doctors as a "slapped-cheek" pattern on the face. They found that B19 had the same rate of mutation as the canine parvovirus, suggesting the rate wasn't specific to the events that had led to the canine virus' emergence from cats, and confirming that parvoviruses have a faster mutation rate than other DNA viruses (J Virol, 80:3666-9, 2006).

Shackelton says there's still much to be learned about human parvoviruses: "Is this a genome-size thing? A genome-characteristic thing?" Jan Drake, of the US National Institute of Environmental Health Sciences, has suggested that mutation rates vary by size.

Meanwhile, Shackelton and her collaborators, Karin Hoelzer and Colin Parrish, at Cornell University, are working to characterize the molecular basis of the elevated rate and the emergence event as well as what the pattern says about viral evolution and emergence. "It's a great model virus that has really been underutilized," she says.