Many millions of years ago, two yeast cells became unable to have productive sex with each other. These eventually gave rise to separate species, or what we now know as Saccharomyces cerevisiae and Saccharomyces mikatae respectively. But how did the original barrier to mating arise in these yeasts, and how has it been maintained for all this time? In studying such speciation events, geneticists have been limited to retrospective studies that infer what might have happened. Now that has changed — in the 6 March issue of Nature, Delneri et al. actually 'do the experiment' to test the effects of chromosomal translocations on speciation.

The genomes of S. cerevisiae and S. mikatae are known to vary by at least two reciprocal chromosomal translocations, which disturb the collinearity of the two genomes. If these yeast species attempt to mate, sterile progeny result — presumably from the inability of the two rearranged genomes to complement each other to produce viable spores. We do not know what initiates the speciation process, but it has been speculated that genome rearrangements between proto-species reinforce their reproductive isolation.

Delneri et al. effectively backtracked in evolution by engineering laboratory strains of S. cerevisiae to the S. mikatae state at the translocation breakpoint. The popular Cre/loxP system was used to create large reciprocal translocations, resulting in a new strain with a genome that is more collinear with that of S. mikatae. When these engineered strains were mated to S. mikatae, viable progeny resulted. Even so, the matings were not 100% fertile, which indicated that the translocation is not the only important genomic difference between the two species. Further important variations might exist at the single-gene scale, which would only be discovered by sequencing both genomes — projects to sequence multiple yeast species are well underway.

Interestingly, the viable hybrid spores that were recovered were often extensively aneuploid, retaining chromosomes from one parent more often than should be the case and having two copies of many chromosomes. The likely explanation is the duplication of one parental genome followed by some chromosome loss, but future experiments will be needed to discover the details. With this work, Delneri et al. have provided a new approach for further exploring these evolutionary mysteries. Although we still do not know what led to the divorce of these two yeasts, we now know what keeps them from reconciling.