A recent PLoS One article “A Genetic Code Alteration Is a Phenotype Diversity Generator in the Human Pathogen Candida albicans” finds some pretty dramatic changes in gene expression and phenotypes by replacing the tRNAs for CUG back to Leucine (Leu; in the standard genetic code) from their meaning of Serine (Ser) in these Candida species. The CUG codon transition in some Candida spp has been of interest since it is an example of a recent change in the genetic code and provides a comparative system to study the mechanism and genome changes of how a genetic code shift is manifested.
In this paper the authors introduce Saccharomyces cerevisiae tRNAs on a plasmid that code CUG as Leu and allowed them to compete with the native C. albicans wild-type tRNA which codes for Ser. Several interesting phenotypes emerged from this synthetic construct. There are a variety of morphological changes known to occur in Candida albicans including white-opaque switch and yeast to mycelia transitions (reviewed here [not free]). These are induced and shown in Figure 4 (above). The authors also report increased cell adhesion or flocculation (this is important in pathogens who need to stick to the host cells and important in making beer with Saccharomyces yeasts so that the cells stick together and drop to the bottom of the fermentation tank. Flocculation is controlled by the expression of genes on the cell surface – these genes in Saccharomyces (such as FLO1) are known to have high intrastrain variability to present diverse surface architectures. In pathogens this variability may also help prevent the host immune system from properly recognizing the cells as invaders.
As has been seen in Candida and other yeasts before, ploidy changes can be a simple mechanism to change gene expression. In this constructed environment of ambiguous CUG tRNAs there was also an increase in the number of ploidy changing events. It would seem this ambiguity is might be a stress condition, perhaps due to limited proper translation of essential genes, that the cells are able to escape through increased copy number of several chromosomes. Figure 8 (above) shows the variation in ploidy in the strains with the introduced plasmids (pUA13 , pUA14, pUA15).
So what is going on with all these changes that normally are only seen conditionally under particular cultivated environments (or hosts)? How is competition among the native tRNA (Ser) and the introduced (Leu) affecting the gene expression, ploidy? It was difficult to get a stable phenotype from the cells so the authors report they were unable to complete a detailed gene expression study – (I am not sure if this is due to the plasmid transcribed nature of the tRNAs or what). The authors point to increased expression of genes WOR1 and HWP1, which play a role in White-Opaque switching and hyphal specific growth respectively, may be controlling a large part of the gene expression and phenotypes. So irregular expression of these genes may be having an effect on the rest of the downstream pathways which are linked to these. The roles of induced virulence related genes (phospholipaes, cell adhesion genes, etc) may be a standard stress response as by-product of improper expression. But I think the actual cause is still not pinned down, but it certainly make for a very interesting attempt to reconstruct the heterogenous tRNA environment which may have been present in the ancestor at some point.
In terms of the evolution of CUG capture event I point to a couple of papers that have explored the evolution of the CUG change. I include a snippet of a tree from a paper I worked on (Fitzpatrick et al, BMC Evol Biology) to make the point that not all Candida species are really part of the same monophyletic clade (C. glabrata being mentioned in the paper but really belonging to a different genus).
I also want to point to a paper from former labmate Stephanie Diezmann (Diezmann et al, JCM) which describes the relationship of the Candida species and a nice study of correlated character evolution showing that presence of coenzyme Q9 (“a mitochondrial electroncarrier with various numbers of isoprene units”) is highly correlated with CUG capture event.
Finally it is interesting to note a statement made by the authors about other possible reasons for the varying degree of phenotypes that are observed. The authors state:
“However, one should not exclude the hypothesis that genome destabilization contributes to exposure of hidden phenotypes through ambiguous CUG decoding”
Evolutionarily I think this is interesting given what Leah Cowen and colleagues found when studying the evolution of drug resistance under different regiments of Hsp90 (chaperone protein) inhibition. Perhaps some of the genes which are poorly expressed or improperly translated when using the wrong tRNA and so there are hidden phenotypes only exposed under extreme conditions. I’m not sure there are plausible conditions when Leu-CUG tRNAs would be present in a wild-type cell, but it maybe possible that there is hidden variability that is only exposed under certain stressed conditions. It would be most useful to get a better handle on the genes which are playing a role in these hidden phenotypes to see if there is any particular avoidance (or lack of) in using CUG codons in these genes. A really nice followup would be protein sequencing to see if translated peptides are mixed or preferentially using Leu or Ser in the CUG place and how this correlates with the phenotype changes.