Fruitfly studies suggest that there is a common genetic basis to reproductive isolation. Credit: janeff/iStockphoto

The genomes of both plants and animals seem to act as engines for speciation, as they are 'hard-wired' to encourage the development of new species, according to findings on two very different groups of organisms.

Two papers published in this week's Science1,2 provide the first empirical evidence that when two diverging populations of organisms become unable to interbreed — a phenomenon known as hybrid sterility or incompatibility — the number of genes that prevent them from producing viable offspring starts to 'snowball'. This process encourages the evolution of new species.

Incompatibility genes can have a variety of effects in offspring, such as interfering with sperm production.

The Dobzhansky-Muller model of hybrid incompatibility, named for the two researchers who in the 1930s and '40s laid the theoretical groundwork for the snowball effect, has, however, remained only a theory till now.

Previous studies looked at levels of infertility rather than the number of genes involved and failed to find the snowball effect. But now two papers produced independently and published in Science provide the first empirical evidence backing the theory, and between them they span both plants and animals. Dmitry Filatov, an evolutionary geneticist at the University of Oxford, UK, says this is surprising because plants are much more prone to interspecies hybridisation than animals.

In the first of the Science studies, Daniel Matute and his colleagues at the University of Chicago in Illinois counted the number of genes involved in species incompatibility between fruitflies. Fruitflies lend themselves to this work because three closely related species of fruitfly — Drosophila melanogaster, D. simulans and D. santomea — can still interbreed, although any offspring are sterile.

The researchers created two hybrid populations — one by crossing D. melanogaster with D. simulans and the other by crossing D. melanogaster with D. santomea — and counted the number of incompatibilities in the hybrid chromosomes.

The species the parent flies belonged to diverged at different times: D. simulans diverged from D. melanogaster about 5.4 million years ago, and D. santomea split from D. melanogaster around 12.8 million years ago. This gave the team a time difference against which to compare the accumulation of incompatibility genes to see whether there was snowballing.

Because divergence dates are always approximate, the team instead used the number of changes in the DNA codes between species as a proxy for time.

Runaway train

They found 65 incompatibilities in the hybrids between D. melanogaster and D. santomea, while there were just ten in the hybrids between D. melanogaster and D. simulans. Taking into account the relative number of genetic changes between species, this suggests that the number of incompatibilities does not increase in a linear way, but is instead accelerating.

In a separate study of the plant genus, Solanum, which includes potatoes and tomatoes, Leonie Moyle of Indiana University in Bloomington and Takuya Nakazato, now at the University of Memphis in Tennessee, looked at the number of genes coding for traits involved in seed and pollen sterility. The genomes of three wild Solanum species — S. pennellii, S. habrochaites and S. lycopersicoides — were introduced into domesticated tomato plants, S. lycopersicum, by repeated breeding, and the number of incompatibility genes were counted. The researchers also looked at the number of genes involved in some traits unrelated to fertility, for the sake of comparison. Again, changes in the DNA code between the species under investigation were used as a proxy for time.

The researchers saw results similar to those in Matute's fruitflies: numbers of genes contributing to seed sterility were swelling much faster than the linear rate, whereas genes that had nothing to do with seed sterility were not.

However, the snowball effect was not seen in genes involved in pollen sterility, suggesting that further work could be needed to refine the theory.

Moyle remains optimistic. "This is exciting because we're at the point of being able to generate enough data to test the theory and because Daniel and I have evaluated it in two really different groups," she says, adding that both sets of results suggest a "deep-seated common genetic basis to reproductive isolation among very different organisms".

Filatov says that this evidence for the snowball theory is "interesting and useful". "It's suggestive", he says, "but there are only a few data points in these papers. I wouldn't put all my money on this effect just yet."