Viruses can be thought of as hyperspeed shape-shifters, organisms that can adapt quickly to overcome barriers to infection. But recent research has been finding ancient traces of many viruses in animal genomes, DNA insertions that have likely been there for much longer than the viruses were previously thought to have existed at all.

A new study describes evidence of a hepadnavirus (a virus group that includes hepatitis B, which infects humans as well as other mammals and ducks) hiding in the genomes of modern songbirds. By tracing back to these bird species' common ancestors, the researchers behind the new work estimate that this family of viruses has been around for at least 19 million years—and possibly as long as 40 million years—rather than the several thousand years researchers had estimated.

Such a primordial start is difficult to square with how similar these ancient snippets' DNA looks to currently circulating versions of the virus and what we know about viruses' ability to change so rapidly. "It's just something we don't quite grasp in the evolution of viruses," says Cédric Feschotte, an associate professor at the University of Texas Arlington's Department of Biology and co-author of the new study. "I think that's pretty exciting." By complicating the understanding of viral evolution, the new findings also promise to help inform transmission dynamics and the ways in which viruses move among different host species.

The new estimate would slow the average rate of hepadnavirus mutation some 1,000-fold, wrote the researchers of the new the study, published online September 28 in PLoS Biology.

Given the bold new picture of virus evolution that many of this and other recent genomic insertion discoveries imply, however, Feschotte is braced for blowback from the field. But the study's numbers look solid, says Harmit Malik, an evolutionary geneticist at the Fred Hutchinson Cancer Research Center who was not involved in the new research. "There was a fairly high burden of proof" in the study, he says; "I think the authors have done a really good job."

Beyond retro
These so-called viral fossils are not mineralized relics but rather bits of genetic code held over in the genome of a modern-day host organism. Until recently, most known viral genomic insertions came from retroviruses, which use host cells' nuclei to replicate—thereby making accidental incorporation into the host's genome more likely than from viruses that replicate outside of the nucleus. The human genome, for example, is thought to owe some 8 percent of its code to endogenous retroviruses. And a July study, published online in PLoS Pathogens, described dozens of examples of viral code in vertebrate genomes—much of which had likely been there for some 40 million years.

As viruses go, hepadnaviruses are a likely candidate to get mixed up with the host's own genetic material. Hepatitis B, a disease that kills some 600,000 people worldwide each year, replicates using RNA, and bits of the virus's genetic material are frequently found inside a host's liver cells.

In order to have become integrated in bird genomes and be passed along for millions of years, however, these hepadnaviruses would have needed to insert genetic material into the sperm or egg cells of their avian hosts. The authors propose that given the viral fossils' time span and diversity, the insertions likely happened on multiple occasions—possibly over the course of millions of years. Feschotte suggests that it is likely having a bit of the virus's code integrated in a bird's genome might have conferred immune protection against similar viruses.

Different speeds
Modern-day viruses are an extremely diverse bunch. Because of the observed quick mutation rates, however, "you could explain away all the diversity as arising from a common ancestor maybe 10,000 years ago," Malik says.

In studying the molecular changes in these viruses, "you would never predict to see something that is 20 million years old that resembles so closely the circulating ones," Feschotte notes. The location of the insertions, however, suggests that these integrations are not modern flukes.

In the birds the viral fragments were located in "the exact same genomic position," a finding that would be "extremely unlikely" to occur at different times across different species more recently, the researchers noted in their paper. Thus, the hepadnaviruses instead worked their way into common ancestors tens of millions of years ago and has been passed down in the genetic code to new species that evolved in the intervening millennia.

Widening window

As the field of paleovirology has taken off in recent years—thanks in large part to quicker genetic sequencing technology—more and more surprises are emerging from the dark, heretofore little-explored reaches of vertebrate genomes. And Feschotte predicts an increasing number of viral findings in the near future.

"Any kind of viruses can potentially do this as long as they can infect the germ cells," Feschotte says. "This is not going to stop. We're picking the low-hanging fruit right now," he says of viral fossil study. And new research profiling more endogenous viruses is slated to be published soon.

Genetic scans can also provide a new, more even-keeled perspective on viral history, Feschotte says. He notes that the study of virus evolution has likely been subjected to a general selection bias in what strains undergo examination. Most viral subjects have been the cause of modern illness—whether in humans or other organisms. But looking into the genetic past "opens an interesting window, a much less biased window, I would argue, to study viruses," he says.

Feschotte proposes that most viruses that have come under study show such a fast mutation rate because they have recently—in the past thousands of years—crossed over into a new species and "are in their adaptation phase, so it kind of makes sense that we see them evolving much faster." He points to the presence of the Ebola virus, which is relatively new to—and so deadly in—humans, that seems to have existed for far longer in bats and hardly makes them sick. But taking a broader viral history survey can offer a more representative picture of how viruses evolved in and with their hosts.

Malik is not entirely convinced by this interpretation, noting that the unexpectedly slow average mutation rate might have more to do with host genetics than those of the viruses. But many of these uncertainties might be answered with further study in the lab, assessing and even reassembling these extinct viruses.

"These are DNA fossils, and we can always put back together and resurrect extinct species," Feschotte says. "I know it sounds very scary, like science fiction," he adds, "but that can really answer questions in terms of the biology of the virus."

Feschotte explains that other researchers might have to puzzle over fossilized bones and argue indefinitely about whether a hominid walked upright. But with the genetic material of an extinct virus in front of them, researchers can recreate it in the lab. The process (which Malik hastens to note is done "in a highly protected setting") could eventually "reveal to us why these hosts were able to survive and thrive with these insertions," Malik says.