The genetic code is fixed in most organisms, but sometimes microbes pull off a swap. How that swap works is hard to fathom, because you can screw up nearly all of your genes by doing it. A paper out today in Nature finds (among many other things) looks at how the yeast genus Candida pulled off such a swap.

Candida yeast are responsible for most human yeast infections. To get a better handle on how these yeast reproduce and become virulent, a group sequenced the genomes of 6 different Candida species, bringing the total number of Candida genomes to 8. By comparing these genomes, researchers can learn things about the evolutionary history of these yeast species, and how they've adapted to their lifestyle as pathogens.

About the swap: in Candida, the translate the DNA sequence CTG as the amino acid serine, instead of the standard one, leucine. That means that at some point in evolutionary history, leucine would be swapped for serine in a lot of Candida proteins, with some significant negative consequences. The biochemical details of this kind of a swap (involving things like wobble bases and inosine-containing tRNAs) will have to be saved for another day.

What's interesting about this genome paper is that the researchers found that the ancestral CTG codons have been completely replaced. A leucine-to-serine switch was almost never tolerated in Candida proteins; in most places where a protein requires leucine, the ancestral CTG codon has been changed to one of the other leucine-coding codons (CTT, CTC, or CTA). In other yeast species (ones that didn't make the genetic code swap), the CTG codon is still in place.

What you seen then here is a bit of the evolutionary history: we knew (based on genetics and biochemistry) that Candida used CTG to code for serine, but now we can see, by comparing species, what effects that has had on the Candida genome.