Coding translational rates: the hidden genetic code

Abstract

In this paper we propose that translational rate is modulated by pairs of consecutive codons or bicodons. By a statistical analysis of coding sequences, associated with low or with high abundant proteins, we found some bicodons with significant preference usage for either of these sets. These usage preferences cannot be explained by the frequency usage of the single codons. We compute a pause propensity measure of all bicodons in nine organisms, which reveals that in many cases bicodon preference is shared between related organisms. We found that bicodons associated with sequences encoding low abundant proteins are involved in translational attenuation reported in SufIprotein in E. coli. Furthermore, we observe that the misfolding in the drug-transport protein, encoded by MDR1 gene, is better explained by a big change in the pause propensity due to the synonymous bicodon variant, rather than by a relatively small change in the codon usage. These findings suggest that bicodon usage can be a more powerful framework to understand translational speed, protein folding efficiency, and to improve protocols to optimize heterologous gene expression.