Animal genomes riddled with the 'skeletons' of ancient viruses

It’s time for animals - including humans - to admit that the bacteria, viruses and other microbes have won. Our bodies are home to many times more bacterial cells than animal cells and countless trillions of viruses. Ancient retroviruses make up a good size chunk of our genome. Now, scientists have discovered that most any virus can set up shop in an animal's genomes and lay dormant for millions of years.


A scan of 44 mammal genomes, plus those of several mosquito and tick vectors and two birds that could serve as reservoirs, has uncovered DNA sequences that can be traced to 10 different families of viruses, including some related to viruses that cause hepatitis B, Ebola, rabies and dengue. Most of the viral sequences are riddled with enough mutations to be considered junk, but some appear to encoding working genes co-opted by their host. The work is published online today in the journal PLoS Genetics.

It’s not obvious how all these viruses got into animal genomes. The researchers, Aris Katzourakis at the University of Oxford, UK, and Robert Gifford at Rockefeller University in New York, searched specifically for viruses that aren’t retroviruses, which are obligated to copy their DNA into hosts. Many but not all of the viruses infect their hosts persistently or replicate inside of the nucleus, however, offering ample opportunity to take up residence in the genomes of germ cells.

The work is just a first look at all the non-retroviruses in the animal genome, but Katzourakis and Gifford turned up a few interesting findings. For instance, their scan identified sequences from filoviruses, the family Ebola belongs to, in the genomes of bats, tarsiers, several rodents, opossums and even wallabies. This hints that filoviruses have a much wider host range than the primate and bat species which these viruses are known to infect.

The paper also hints at unknown ancient transmissions of viruses between hosts. The bottlenose dolphin genome, it turns out, is home to sequences of a kind of parvovirus similar to one found in birds, suggesting that the viruses may have jumped between mammals and birds in the past.

Most of these sequences are junk, so filled with mutations that they can’t make working proteins. But some of the viral sequences might do something inside their hosts. One example is a bornavirus gene called EBLN-1 that took up residence in ancient primate genomes some 50 million years ago and survives intact in many modern primates, including humans. A similar protein latches onto RNA in bornaviruses, so it might do the same in primates as part of a viral defence mechanism, Gifford speculates.

Like the ancient retroviruses locked inside animal genomes, these viruses offer a window into infections that occurred millions of years ago.

“People who are looking at the ecology of those diseases, they very much work in recent time and they have no assumptions that it’s an old system that might have evolved over billions of years,” says Gifford. "The data that we’re finding is really contradicting that and providing the first evidence that these are really old relationships between hosts and viruses, and I think it’s really critical to how we underestand them to get that context right."