Patterns of genome-wide codon usage bias in tobacco, tomato and potato

Abstract

Understanding codon usage patterns and their shaping factors in members of the Solanaceae family may reflect important features of their evolutionary history. Here, we investigated codon usage bias (CUB) over the whole genome of tobacco, tomato and potato. Analyzing the effective number of codons (ENc) revealed a weak CUB over the whole genome of the three genomes. The correlation analysis between the ENc and the nucleotide content at the third codon position indicated that genes with high ENc have high A/T and low G/C contents. Analyzing nucleotide composition and the relative synonymous codon usage confirmed that the three genomes are AT-rich and favor codons ending with T and/or A. Comparing CUB in each of the three genomes revealed that approximately 50% and 45% of the genes were dominated by mutational and natural selection, respectively. The remaining 5% of the genes were affected by combined factors of mutation and translation selection. The first four axes of the correspondence analysis (CA) were the major contributors to the variation within the codon sequences (CDS) of the three genomes. Results of a multi-regression analysis between the 4 axes and the CUB indices verified the fact that natural selection and directional mutation for base composition along with other factors could describe the variations in the CUB within the CDS of the three genomes. In conclusion, the CUB of the three species showed the same weak trend and that mutation and natural selection were the main factors influencing the nuclear genes of the three genomes.