Viruses are resilient. They mutate and adapt to survive, fueling the fear of uncontrollable epidemics. Scientists recently found, however, that with rotavirus, evolution may be slower than previously thought. It's an insight that potentially could inform care and prevention efforts.

Globally, rotavirus is responsible for nearly half a million deaths in children each year.

Sarah McDonald, an assistant professor at the Virginia Tech Carilion Research Institute, led the study, recently published in the Journal of Virology, which found that the rotavirus genome does not change as often as expected.

McDonald's team - including researchers Shu Zhang and Paul McDonald from the Virginia Tech Carilion Research Institute and Travis Thompson, then a student at the Virginia Tech Carilion School of Medicine - collaborated with scientists at the National Institutes of Health and the J. Craig Venter Institute to examine rotaviruses in clinical samples that were collected from a single hospital in Washington, D.C., over the course of 18 years, from 1974 to 1991.

"We needed access to a big dataset of specimens from which we could extract the viral genome and sequence it," McDonald said. By looking at how each of the 11 viral genes change individually and together, scientists can begin to understand how rotavirus evolves as it spreads in the human population.

When children become infected with rotavirus, they are often hit with the double-whammy of a co-infection. Rotaviruses have a segmented genome, so they can swap genes, resulting in a new, blended rotavirus. The daughter rotavirus has a combination of genes, some of which may be beneficial for its survival.

"The leading thought in the field was that rotaviruses reassorted their genes frequently in nature," McDonald said. "We looked at that for the Washington, D.C., rotaviruses, but then took it a step further to examine the mechanism that may have either promoted or limited gene reassortment."

The results were intriguing.

"We noticed that certain sets of genes, called gene constellations, tended to stick around longer than others," McDonald said. "One group of viruses, collected over a span of 13 years, didn't change at all."

That's not quite the blink-of-an-eye evolution anticipated with viruses.

"You think of a virus and you think of mutation," McDonald said. "They change so rapidly."

Their finding compelled McDonald's team to find out why rotaviruses behaved so unexpectedly.

One hypothesis was that rotaviruses retain certain gene constellations because their encoded proteins work best when they are kept together.

McDonald's team used amino acid covariance analysis to test their theory. If two proteins co-evolved to operate best when kept together, a signature would appear in their amino acid sequences.

"You could imagine that two proteins interact and evolve to have a nice fit, like a lock and key," McDonald said. "All of a sudden, one of the partners gets swapped out for something that looks very different and now, at the protein level, it's no longer quite right. The virus with the new protein combination won't emerge in the population because it doesn't work as well."

The researchers found what they expected - amino acid covariances in connections among physically interacting proteins. They also found something puzzling - there were more connections between proteins not known to interact at all. While the results were initially startling, McDonald said it's not that surprising in an evolutionary sense; separate changes might synergize to affect fitness outcomes.

The next step is for McDonald's team to sequence contemporary strains from a variety of locations. The results of these studies are expected to provide an important scientific foundation for future vaccine design.

"Rotavirus has evolved to be exquisitely good at what it does," McDonald said. "It has a lot more to teach us."