An integrated view of protein evolution

Top Cited Papers
Publisher Website
Google Scholar
Abstract
Variations in the rate of protein evolution are determined by biases in the mutation rate and fixation rate (which are either protein specific or linked to genomic location). By drawing on accumulating genomic data, evolutionary studies have moved from studying individual proteins to characterizing global cellular factors. Protein-specific biases in fixation rate are due to differences in both purifying and positive selection across genes. Although theoretical considerations that are based on purifying selection suggest that the importance of a gene (or its dispensability) is a key determinant of protein evolution, experimental data confirm at best a moderate influence. An important concept in thinking about protein evolution is fitness density, that is, measuring the weighted fraction of sites at which mutations result in phenotypes with modified fitness. Selection on protein structure and stability is presumably responsible for the largest contribution to fitness density. The position of a protein in biological networks seems to be only of minor importance, despite much recent excitement. Broadly expressed and highly expressed proteins evolve slowly; expression level is by far the strongest predictor of evolutionary rate in yeast (possibly because of selection for robust folding in highly expressed proteins). Some recent studies suggest that a large fraction (∼30%) of amino-acid changes might be driven by positive selection, contrary to expectations that are based on the (nearly) neutral theory. Positive selection often reflects compensatory mutations or arms races rather than adaptation. Further research is needed to understand the relative importance of the different factors that affect protein evolution; future studies will be most effective if combined with the development of a coherent theory that is based on population genetics models.