MINI REVIEW A gene’s eye view of epistasis, selection and speciation

The genetic architecture of a phenotype consists of the genes, the interactions among them (epistasis), and the interactions among genes and environments (G · E) that affect the phenotype’s expression (Wade et al., 2001). For a phenotype with a ‘complex’ genetic architecture, epistasis and G · E play significant roles as opposed to a phenotype with a ‘simple’ architecture, in which interactions of any sort are relatively unimportant. Epistasis contributes to inbreeding depression, developmental homeostasis, plasticity, evolution of sex and recombination, mating system evolution, speciation and interdemic selection, because all of these topics involve phenotypes with a complex genetic architecture. For example, in speciation, epistasis contributes to reproductive isolation because genes that function well in the genetic background of conspecifics function poorly in the genetic background of interspecific hybrids. Such a change in the sign of a gene’s effect from positive to negative can only be achieved by interactions (Wade, 1992; Johnson & Wade, 1996). In the same way, gene interactions are essential to understanding developmental homeostasis (or canalization), in which alleles at one locus reciprocally diminish allelic effects at other loci (Wagner et al., 1998). Wright (1931, 1969) considered epistasis to be ‘ubiquitous’ and stated that ‘The inadequacy of any evolutionary theory that treats genes as if they had constant effects, favourable or unfavourable, irrespective of the rest of the genome, seems clear’ (Wright, 1969, p. 88). He emphasized that the ‘...existence [of epistasis] must be taken as a major premise in any serious discussion of population genetics and evolution’ (Wright, 1969, p. 105). Wright (1931) developed his ‘adaptive landscape’ as a way of illustrating nonlinearities in the map between two-locus gene combinations and fitness, although he found it inadequate for representing gene interactions of higher dimensionality. Nevertheless, Wright’s heuristic and the genetic ideas on which it is based have been criticized (Coyne et al., 1997, 2000; Phillips, 1998). The role of epistasis in microevolutionary processes within a single population often can be minimized by assumption or by transformation (Brodie, 2000). Speciation theory is