adaptive evolution

Halligan DL, Oliver F, Eyre-Walker A, Harr B & Keightley PD 2010 Evidence for pervasive adaptive protein evolution in wild mice. PLoS Genet 6:e1000825.

  • estimates of the proportion of adaptive amino acid changes (α) vary greatly between taxa
  • being 50% or more in bacteria and fruit flies
  • at most 13% in hominids
  • we infer the frequencies of polymorphism in protein-coding genes of 15 Mus musculus castaneus individuals
  • by combining the polymorphisms data with nucleotide divergence to the related murid species M. famulus and the rat, we obtain an estimate for α of 57%
  • this represents the first estimate of α for a mammal other than humans
  • hominids may be somewhat unusual in having low rates of adaptive protein evolution
  • one possible cause of this is the low effective population size in humans
  • this is consistent with the higher frequency of nearly neutral deleterious amino acid mutations in hominids than murids
  • higher FST within genic than nongenic regions of the human genome suggests that genic regions are subject to local adaptation
  • an approach that tests for an excess of substitutions at selected sites is the McDonald-Kreitman test
  • which contrasts levels of polymorphism with divergence at selected sites (e.g., nonsynonymous sites) and linked putatively neutral sites (e.g., synonymous sites)
  • a recent extension of this test quantifies molecular adaptation as the fraction of substitutions driven to fixation by positive selection (α)
  • some estimates of α are compromised by slightly deleterious mutations
  • these contribute proportionately more than neutral polymorphisms to diversity than divergence
  • if slightly deleterious mutations are prevalent and not properly accounted for then α could be substantially underestimated
  • this may partially explain low estimates of α obtained using methods that do not incorporate explicit population genetics models
  • recently, improved methods to estimate α have been developed that model the contribution of slightly deleterious mutations to polymorphism and divergence
  • the estimates obtained for yeast and Arabidopsis are not based on an explicit population genetics model
  • even though attempts have been made to reduce the impact of slightly deleterious mutations, the estimates may still be downwardly biased
  • the low estimated proportion of adaptive substitutions in hominids may reflect their low Ne
  • this will increase the proportion of effectively neutral advantageous and deleterious mutations
  • low Ne will also reduce the rate of adaptive evolution if the rate is limited by the supply of mutations
  • this is consistent with the low recent Ne estimates for humans [23], chimpanzees [39] and gorillas [40]
  • it is also possible that most adaptive evolution occurs in noncoding regions in primates [41]
  • alternatively, changes in effective population size can lead to bias in the estimate of α