Abstract
Biased usage of synonymous codons has been elucidated under the perspective of cellular tRNA abundance and more recently by the mRNA secondary structure folding stability of the corresponding genes. Taking advantage of publicly available gene expression data for Escherichia coli, a comprehensive investigation of the three classes of genes having different codon usage biases was performed from the standpoint of tRNA abundance, mRNA secondary structure folding stability, and translational error minimization procedure. We detected the different evolutionary forces for translational and/or transcriptional regulation of highly expressed genes depending upon their codon bias. Additionally, the possible role of mRNA folding stability in maintaining the overall high expressivity of the set of lowly biased genes has been articulated. These novel findings certainly strengthen the understanding of the codon usage bias in the Escherichia coli genome.