Studying Patterns of Recent Evolution at Synonymous Sites and Intronic Sites in Drosophila melanogaster
- 250 Downloads
Most previous studies of the evolution of codon usage bias (CUB) and intronic GC content (iGC) in Drosophila melanogaster were based on between-species comparisons, reflecting long-term evolutionary events. However, a complete picture of the evolution of CUB and iGC cannot be drawn without knowledge of their more recent evolutionary history. Here, we used a polymorphism dataset collected from Zimbabwe to study patterns of the recent evolution of CUB and iGC. Analyzing coding and intronic data jointly with a model which can simultaneously estimate selection, mutational, and demographic parameters, we have found that: (1) natural selection is probably acting on synonymous codons; (2) a constant population size model seems to be sufficient to explain most of the observed synonymous polymorphism patterns; (3) GC is favored over AT in introns. In agreement with the long-term evolutionary patterns, ongoing selection acting on X-linked synonymous codons is stronger than that acting on autosomal codons. The selective differences between preferred and unpreferred codons tend to be greater than the differences between GC and AT in introns, suggesting that natural selection, not just biased gene conversion, may have influenced the evolution of CUB. Interestingly, evidence for non-equilibrium evolution comes exclusively from the intronic data. However, three different models, an equilibrium model with two classes of selected sites and two non-equilibrium models with changes in either population size or mutational parameters, fit the intronic data equally well. These results show that using inadequate selection (or demographic) models can result in incorrect estimates of demographic (or selection) parameters.
KeywordsDrosophila melanogaster Codon usage Mutational bias Natural selection GC content
We thank Dr. Andrea Betancourt for helpful discussions, and Dr. David Turissini for providing the Drosophila dataset. This work made use of the resources provided by the Edinburgh Compute and Data Facility (ECDF; http://www.ecdf.ed.ac.uk/). The ECDF is partially supported by the e-Science Data, Information and Knowledge Transformation (eDIKT) initiative (http://www.edikt.org.uk). K.Z. acknowledges support from the School of Biological Sciences, University of Edinburgh, and a Biomedical Personal Research Fellowship given by the Royal Society of Edinburgh and the Caledonian Research Foundation.