The term “genetic code” is associated with a measure of ambiguity.  For molecular biologists, “genetic code” has historically referred to a table that provides for each messenger RNA ribonucleotide triplet the corresponding amino acid that is incorporated into the growing end of a nascent polypeptide chain, i.e. the translation from RNA sequence to protein sequence.  In colloquial parlance, “genetic code” is frequently used to refer to all or part of the deoxribonucleotide sequence of a genome.  A recent paper, published online ahead of print in Proceedings of the National Academy of Sciences (Bezerra et al., PNAS, 2013) demonstrates that this semantic ambiguity can have a counterpart in the ribosomal interpretation of the genetic code, using the technical molecular biological meaning of the latter term.

Bezerra et al. exploit the documented tendency of a strain (SN148) of Candida albicans to incorporate either serine (Ser) or leucine (Leu) at polypeptide sites encoded by the CUG triplet, which in most other cells directs incorporation only of leucine.  The authors note that fungal cells from three genera (Candida, Debaryomyces, and Lodderomyces) have previously been shown to exhibit this interesting variation on the normal correspondence between codon and amino acid.  This deviation from the typical codon-amino acid relationship is not entirely unprecedented.  The authors cite other examples of documented alterations in the standard genetic code in organisms or organelles such as Mycoplasma, Micrococci, ciliates, and eukaryotic mitochondria.

Specifically, the authors manipulate the frequency of Leu incorporation associated with CUG codons in SN148 strain of cells of Candida albicans, which unlike most cells usually incorporate Ser in response to CUG codons.  The specific modifications including inserting one or two copies of a yeast tRNA gene (tDNACAGLeu) associated with incorporation of Leu and, in some strains, deleting one or two copies of the gene encoding the tRNACAGSer (i.e., the gene encoding the tRNA that favors incorporation of Ser over Leu (97% to 3% in 60% of the protein-coding genes).  Consequently, the investigators created a series of C. albicans strains that incorporated progressively higher percentages of Leu instead of Ser.  The effects of these engineered differences in the fungal proteome were dramatic and included:

1) new colony morphologies observed for genetically-modified strains,

2) altered responses under a variety of growth conditions such that the strains  incorporating increased percentages of Leu at CUG codons were superior to control strains in some conditions and inferior to control strains in others,

3) increased stomach-associated inflammatory responses, assessed by measuring inflammation-associated cytokines (e.g., TNF-alpha, IL-17A, IL-10) elicited by intragastric infection of mice with C. albicans strains incorporating more Leu than by infection with control strains,

4) more severe stomach pathology caused by modified as compared to control strains in infected mice, presumably related to the alterations in cytokine production cited in the preceding point,

5) and increased tolerance for some clinical antifungal agents, such as fluconazole and itraconazole, and less tolerance for another antifungal, caspofungin .

When the authors analyzed the genomes of the control strains and the genetically-manipulated strains, they uncovered more interesting findings.  There were many additional genetic variants found in the strains with modified codon interpreration due to knocked out and knocked in tRNA genes.  These additional genetic variants were found at loci encoding proteins related to filamentous growth, cell adhesion, metabolism, transcriptional control, and other unknown functions.  The authors interpreted these results to imply that new mutations were selected for as compensation for the altered functions of the many gene products that had Leu residues where Ser residues normally resided (in the wild-type C. albicans strain).

This study by itself does not definitively determine the role of codon ambiguity in the evolution of this human pathogen, but it demonstrates unequivocally the potential for phenotypic impact of such translational variability.  Furthermore, the data are highly suggestive that this form of variation can promote further genetic diversification and in ways that could affect pathogen virulence and drug sensitivity.  It appears that some mutations in genes encoding tRNAs and translation release factors can result in altered codon translation.  Once again, we see that evolution by natural selection is a process ‘willing’ to try and if successful exploit any conceivable genetic mechanism that does not violate the laws of chemistry and physics.

Acknowledgments.  I wish to thank Peter Harte, a geneticist and Case colleague, for bringing this paper to my attention.



Ana R. Bezerra, João Simões, Wanseon Lee, Johan Rung, Tobias Weil, Ivo G. Gut, Marta Gut, Mónica Bayés, Lisa Rizzetto, Duccio Cavalieri, Gloria Giovannini, Silvia Bozza, Luigina Romani, Misha Kapushesky, Gabriela R. Moura, and Manuel A. S. Santos. Reversion of a fungal genetic code alteration links proteome instability with genomic and phenotypic diversification PNAS 2013 ; published ahead of print June 17, 2013, doi:10.1073/pnas.1302094110