Abstract
Different synonymous codons are favored by natural selection for translation efficiency and accuracy in different organisms. The rules governing the identities of favored codons in different organisms remain obscure. In fact, it is not known whether such rules exist or whether favored codons are chosen randomly in evolution in a process akin to a series of frozen accidents. Here, we study this question by identifying for the first time the favored codons in 675 bacteria, 52 archea, and 10 fungi. We use a number of tests to show that the identified codons are indeed likely to be favored and find that across all studied organisms the identity of favored codons tracks the GC content of the genomes. Once the effect of the genomic GC content on selectively favored codon choice is taken into account, additional universal amino acid specific rules governing the identity of favored codons become apparent. Our results provide for the first time a clear set of rules governing the evolution of selectively favored codon usage. Based on these results, we describe a putative scenario for how evolutionary shifts in the identity of selectively favored codons can occur without even temporary weakening of natural selection for codon bias.
Highlights
The genetic code is redundant with most amino acids encoded by several synonymous codons
We further use the data in bacteria to demonstrate that once nucleotide substitution patterns are taken into account additional amino-acid specific rules determining the identity of favored codons become apparent
For each amino acid we identified the most favored codon defined as the codon that showed both the strongest and statistically significant positive Spearman correlation with the overall level of codon bias (P#0.05/n, where n is the number of codons encoding the amino acid in question, Materials and Methods)
Summary
The genetic code is redundant with most amino acids encoded by several synonymous codons. The identity of the codons used by many genes that are not highly expressed may be determined to a large extent by the nucleotide substitution patterns of the genome that are unrelated to natural selection at the level of translation. Unlike in Drosophila, the identities of favored codons in bacteria, archea, and fungi correspond to the nucleotide content of the intergenic regions of each genome. We further use the data in bacteria to demonstrate that once nucleotide substitution patterns are taken into account additional amino-acid specific rules determining the identity of favored codons become apparent.
Sign-in/Register to view highlights & summary 
Published Version (
Free)
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call