The formation of two separate species from a single ancestor, and the subsequent maintenance of these new species without hybridization, is a central problem in evolutionary biology. The availability of entire genomes in a range of organisms could help address the problem, as E. J. Louis of the University of Oxford and colleagues now explore in Nature1.

To test one possible speciation mechanism -- chromosomal rearrangement -- Louis and colleagues studied the relationships of brewers' yeast, Saccharomyces cerevisiae, with several of its close relatives. Chromosomes are the bodies along which genes are arranged in a linear order. Occasionally, chromosomes may get themselves in a muddle: a segment of one chromosome is switched around, or moved to a different chromosome. This interferes with 'meiosis', which produces sex cells -- sperm and eggs.

Because of chromosomal rearrangements, sperm and egg cells may have too many chromosomes, or too few. This can prevent the successful development of fertile or viable offspring. Chromosomal rearrangement, therefore, could be a useful way of enforcing the differences between two otherwise very similar species, and has indeed been observed in several cases.

The availability of the entire genome of Saccharomyces cerevisiae -- all 16 chromosomes of it -- now makes it possible systematically to test the involvement of chromosomal rearrangement in speciation. The researchers mapped the pattern of chromosomal rearrangement in brewers' yeast and its relatives.

If chromosomal rearrangement has been a constant, consistent factor in speciation, then we would expect a greater amount of rearrangement with evolutionary time. In other word, closest relatives would have only a small amount of rearrangement, but more distant relatives would have more divergent chromosomes.

The researchers now find that the converse is true: the most extensive rearrangements are seen between close relatives, and distant relatives share the same chromosomal order. This suggests that chromosome rearrangement is not a prerequisite for yeast speciation, though it might contribute to it.

If not speciation, what is the consequence of chromosome rearrangement? At this point the researchers found something very interesting: chromosomal rearrangements are not random, but tend to form clusters of particular chromosomal rearrangements in certain yeasts. Why could this be? An intriguing suggestion is that chromosomal rearrangement is a response to environmental challenges.

Chromosomal rearrangements of the kind seen by Louis and colleagues have been observed in industrial strains of yeast, brought in from the wild to the very different regime of the culture vat. In response to such a drastic environmental shift, natural selection might relax its normal stranglehold on DNA housekeeping, allowing a mild amount of chromosomal shuffling to juggle the genes into new combinations, some of which might be suited to the new circumstances.