Base Composition and Translational Selection are Insufficient to Explain Codon Usage Bias in Plant Viruses

Daniel Cardinale,Kate Derosa,Siobain Duffy

doi:10.3390/v5010162

Abstract

Viral codon usage bias may be the product of a number of synergistic or antagonistic factors, including genomic nucleotide composition, translational selection, genomic architecture, and mutational or repair biases. Most studies of viral codon bias evaluate only the relative importance of genomic base composition and translational selection, ignoring other possible factors. We analyzed the codon preferences of ssRNA (luteoviruses and potyviruses) and ssDNA (geminiviruses) plant viruses that infect translationally distinct monocot and dicot hosts. We found that neither genomic base composition nor translational selection satisfactorily explains their codon usage biases. Furthermore, we observed a strong relationship between the codon preferences of viruses in the same family or genus, regardless of host or genomic nucleotide content. Our results suggest that analyzing codon bias as either due to base composition or translational selection is a false dichotomy that obscures the role of other factors. Constraints such as genomic architecture and secondary structure can and do influence codon usage in plant viruses, and likely in viruses of other hosts.

Highlights

All organisms exhibit some degree of codon usage bias (CUB), the unequal usage of synonymous codons [1,2]
Codon bias may vary among genes of the same organism, which is associated with factors like asymmetrical mutation pressures or tissue-specific gene expression, but is relatively uniform within the most highly expressed genes [3,4,5,6,7,8,9,10]
When CUB diverges from the null hypothesis of genomic nucleotide content, translational selection—selection for optimal speed and accuracy of translation—is routinely invoked

Summary

Introduction

All organisms exhibit some degree of codon usage bias (CUB), the unequal usage of synonymous codons [1,2]. Codon bias may vary among genes of the same organism, which is associated with factors like asymmetrical mutation pressures or tissue-specific gene expression, but is relatively uniform within the most highly expressed genes [3,4,5,6,7,8,9,10]. CUB is often explained as the product of two potentially competing factors: genomic base composition and translational selection (Table 1). Translational selection should exert an influence on CUB because preferred codons tend to correlate with the most common tRNAs [11,12], allowing for faster, yet accurate, codon recognition and translation of highly expressed genes [13,14]. Genomic composition and translational selection need not be acting antagonistically on CUB, and sequences can show CUB distinct from that predicted by either force

Methods

Results

Discussion

Conclusion