Reservoirs for Retrovirus Evolution

Retroviruses are a family of viruses that use RNA instead of DNA as their genetic material. The 'retro' designation comes from the fact that retroviruses copy their RNA back into DNA once they get inside a cell, and that DNA gets integrated into the host's genome and copied along with it. If a retrovirus is lucky enough to find itself in a germ cell, it will get passed on into the gametes and so become incorporated in the organism's offsprings' genome, too. Retroviruses that pull this trick off are called endoretroviruses (ERVs). Once a retrovirus becomes an ERV, it sets off on a separate evolutionary course from the retroviruses that spawned it. The ERV will evolve in (and with) its host's genome, while the free retroviruses will evolve in their own right, infecting other organisms — possibly even other species — and spawning other ERVs along the way. ERVs can also become active again and produce viruses that go on to infect other creatures. It turns the grand, stately evolutionary tree into a beautiful, fascinating, skein of entangled threads.

Gammaretorviruses are a great example of this kind of mess. They're a genus of retroviruses that are linked with a variety of diseases in vertebrates, and researchers also suspect that they're pretty good at switching host species. In a paper published in PNAS last year, a team from Uppsala University investigated the evolution of gammaretroviruses and their hosts. They used software to comb through the published genomes of 60 vertebrates and identify ERVs and then developed a method to construct phylogenies from such a large dataset (they recovered nearly 90,000 ERV sequences!) and to account for the different selective pressures acting on ERVs and free retroviruses. Their analysis showed clear evidence of host switching, with related ERVs being spread across widely divergent hosts instead of neatly clustered within a host clade. Their ancestors, free-living gammaretroviruses, were evidently happy to hop between different species. The difference between the ERV phylogeny and that of their hosts is most strikingly evident in this tanglegram, which shows hows the two trees line up:

The researchers discovered another interesting pattern: ERVs from rats and mice were spread throughout the phylogeny. The ERVs in many other species often had closely related ERVs in rodents, suggesting that these animals may have served as a retrovirus reservoir, spreading retroviruses (and thus ERVs) to a variety of mammals over the course of evolution. The team also noticed that ERVs derived from gammaretroviruses were much less common in large herbivores like cattle, elephants, and sloths, while they were over-represented in the true carnivores. Data from the panda reinforced the notion; although a member of the order Carnivora, it is a strict vegetarian, and, in keeping with that, had few gammaretrovirus-derived ERVs. The possible link between diet and retroviral infection is intriguing, especially considering the potential evolutionary impact of ERVs. What an amazingly roundabout link between diet and evolutionary fate, and what a gloriously tangled mess the whole story is!

Ref
Hayward A, Grabherr M, Jern P. Broad-scale phylogenomics provides insights into retrovirus-host evolution. PNAS 110(50):20146-20151. (2013) doi:10.1073/pnas.1315419110

Image credits
The tanglegram is Figure 4 in the paper.