Codon utilisation in the pathogenic yeast, Candida albicans.

The imperfect fungus, Candida albicans is the major yeast pathogen in humans (1). We have analysed the codon usage in eleven C. albicans gene sequences reported in the EMBL and GenBank databases (Table). The codons are divided into three groups (low, medium and high) based upon their codon usage (Table), and these groups are compared with analogous groups for the highly expressed genes from S. cerevisiae and tine divergent yeast, Schizosaccharomyces pombe (12). The subsets of strongly preferred codons in each organism overlap (for example, with the codons GGT, GTT, TTG and CAA). However, significant differences exist between these subsets, and for C.albicans these differences represent a strong preference for an A or T in the third, or 'wobble' position. For example, the preferred codon for lysine is AAA in C.albicans, but is AAG in 5. cerevisiae and S.pombe, and the leucine codon TTA is frequently used in C.albicans, but not in the other two yeasts (Table). This should be considered when using amino acid sequence data for the design of synthetic oligonucleotide probes to isolate C.albicans genes. The bias towards an A or T in the wobble position may reflect the high A:T content of C.albicans DNA (65.3% in the genes analysed here). However, the preference for the termination codon TAA is probably due to the efficiency of natural suppression of termination by TGA and TAG in C.albicans (13). The CTG codon, which normally encodes leucine, has been shown to encode serine in C.cylindracea (14). Since to our knowledge there has been no analogous report for C.albicans, no reference to this is made in the Table. Such a change in tRNA specificity would prevent the use of heterologous reporter genes in C.albicans.