An analysis of the selection, usage and bias of optimal codons in the excavate protozoan parasite Trichomonas vaginalis

Introduction

Trichomonas vaginalis (T. vaginalis) is a single-celled eukaryotic parasite and excavate protist. The genome of T. vaginalis was fully sequenced in 2007 and this has presented an opportunity to study various features of its genome such as introns/exons, transposable elements etc. However, while the structure of its genome has been studied by a number of research groups, the codon usage and genomic bias of this organism has not had a significant amount of research dedicated to it. This study attempted to look in more detail at the codons of the organism with regards to selection, bias and usage as well as the structure and evolutionary factors that may be affecting these preferences. As well as this, the organism’s genome was studied using current methods and resources to determine if the existing understanding of T. vaginalis’s genome structure are accurate.

Methodology

The original genome dataset for T. vaginalis was analysed in order to verify its gene content as well as identify which genes were truly coding, after which the optimal codons were determined and their bias analysed using codonW in the coding genes and transposable elements. An analysis of mutation pressure was also performed by determining the ratio of coding GC3s to non-coding GC’s in genes with high and weak expression levels, as well as a phylogenetic analysis on the candidate genes for adenosine deaminase acting on tRNA (ADAT) to determine if these genes are present and actively transcribed.

Results and Conclusions

Many of the genes may potentially be false positives due to their lack of functional domains and failure to fall into a recognised KOG category. As well as this, the overall GC3 distribution was low, likely due to the AT bias of the organism’s genome. The candidate optimal codons were also identified and matched most of the corresponding tRNAs in both the coding and full gene datasets. The few exceptions may potentially be the result of the deamination of adenosine in their corresponding tRNAs. This is evidence that natural selection is having an influence on the selection of OCs in T. vaginalis and is consistent with mutation/drift overriding the selection of many of the codons in the dataset, skewing them towards AT, and contributing to the AT bias of the T. vaginalis genome. The optimal codon frequency (Fop) in the coding genes was higher in their domain regions, suggesting that selection operates at the level of translational accuracy. In the phylogenetic analysis, ADAT1 gene phylogeny and tree topology matched the expected species phylogeny, while the ADAT2 and 3 phylogenies only did so after adding an outgroup of prokaryote tRNA TadA genes. This suggests that these genes are the true ADAT genes and implies that the deamination of adenosine may indeed be occurring in the tRNAs and influencing the selection of OCs. The analysis of mutation pressure also indicated that mutation pressure may have an influence on codon usage as there was a statistically significant difference between coding GC3s and non-coding GC content. As well as this, the Fop analysis of the domain and non-domain regions, also suggests that selection operates at the level of translational accuracy. With regards to the transposable elements, while overall, they showed no relationship between GC3s and Fop, when separated into their respective families, the LTR retrotransposons alone displayed a positive relationship. This indicates that the retrotransposons do select for optimal codons, but the other transposons do not.