Epistasis and Evolution

The concept of epistasis was introduced into evolutionary theory more than a hundred years ago. Its history is marked by controversies regarding its importance for the evolutionary process, as exemplified by the debate between Ronald Fisher and Sewall Wright in the wake of the modern synthesis. In this case the disagreement was about the shape of the adaptive landscape, which is determined by epistasis. Wright believed that epistasis causes the adaptive landscape to be rugged with many local peaks, whereas Fisher viewed evolution as a smooth, steady progression toward a unique optimum. Even today, the different meanings attributed to epistasis continue to spawn confusion. Nevertheless, a consensus is emerging, according to which the term should be used to designate interactions between genetic effects on phenotypes in the broadest sense. Stated differently, in the presence of epistasis the phenotypic effects of a gene depend on its genetic context. In evolutionary theory the phenotype of primary interest is organismal fitness, but principally the concept applies to any genotype-phenotype map. Reflecting the Fisherian view, throughout the 20th century epistasis was often considered to be a residual perturbation on the main effects of individual genes. Following the advent of sequencing techniques providing insights into the molecular basis of genotype-phenotype maps, over the past two decades it has become clear, however, that epistasis is the rule rather than an exception. This has motivated a large number of empirical studies exploring the patterns and evolutionary consequences of epistasis across a wide range of scales of organismal and genomic complexity. Correspondingly, mathematical and computational tools have been developed for the analysis of experimental data, and models have been constructed to elucidate the mechanistic and statistical origins of genetic interactions. Despite a certain inherent vagueness, the concept takes center stage in modern evolutionary thought as a framework for organizing the accumulating understanding of the relationship among genotype, phenotype, and organism.