Academic journal article Genetics

The Evolution of Epistasis and Its Links with Genetic Robustness, Complexity and Drift in a Phenotypic Model of Adaptation

Academic journal article Genetics

The Evolution of Epistasis and Its Links with Genetic Robustness, Complexity and Drift in a Phenotypic Model of Adaptation

Article excerpt


The epistatic interactions among mutations have a large effect on the evolution of populations. In this article we provide a formalism under which epistatic interactions among pairs of mutations have a distribution whose mean can be modulated. We find that the mean epistasis is correlated to the effect of mutations or genetic robustness, which suggests that such formalism is in good agreement with most in silico models of evolution where the same pattern is observed. We further show that the evolution of epistasis is highly dependant on the intensity of drift and of how complex the organisms are, and that either positive or negative epistasis could be selected for, depending on the balance between the efficiency of selection and the intensity of drift.

(ProQuest: ... denotes formulae omitted.)

THE long-term evolutionary fate of a population relies on the shape of its adaptive landscape. The adaptive landscape is the precise map that associates a fitness value to any possible genotype or phenotype. While a perfect knowledge of the adaptive landscape is out of reach, even for the simplest organisms, understanding its statistical properties can be very valuable to decipher the different selective pressures acting on an organism and its genetic system (e.g., its mutation rate or its recombination rate). Among such statistical properties, the mean epistasis among mutations is one that has been broadly studied in population genetics (Wolf et al. 2000). Epistasis refers to the existence of interactions between mutations: the effect of a mutation depends on the genetic background in which it appears. The spread of a mutation in a population depends on its effect on fitness and as epistasis affects fitness it can potentially influence the evolution of a population. This is why we focus here on the effect of mutations and their interactions on fitness.

Epistasis has been studied widely because it was demonstrated that depending on its sign selection could favor or not the evolution of recombination (Kondrashov 1993). Since that study was published, much more convincingmodels have been developed on the evolution of sex (Otto and Barton 2001; Poon and Chao 2004; Barton and Otto 2005; de Visser and Elena 2007). Nevertheless, the early model generated a strong interest for the experimental study of epistasis among the population genetics community (de Visser et al. 1996, 1997; Elena and Lenski 1997; Bonhoeffer et al. 2004; Burch and Chao 2004; Sanjuan et al. 2005; Beerenwinkel et al. 2007; Jasnos and Korona 2007)

There are several closely related ways to calculate epistasis (Wolf et al. 2000); in this article epistasis between two mutations is defined as

... (1)

(Martin et al. 2007).

In this definition e measures the deviation of the log- fitness for a double mutant, log(f(x12)/f(x0)), from that expected if the fitness effects of each individual mutation were multiplicative. (x0 is the ancestral genotype, x1 and x2 are the genotypes having mutation 1 or mutation 2, and x12 is the double mutant). With this definition epistasis is a relative deviation from single-mutant effects and is therefore not directly connected to the amplitude of the effect of single mutations on fitness. A positive epistasis refers to the case in which a combination of mutations has a higher fitness than the one expected from independent mutations. Negative epistasis refers to the opposite: double mutants have lower fitness than expected.

Epistasis is a key component of genomic architecture. Two main approaches, which can be traced back to the Wright vs. Fisher debate, have been used to study epistasis. In a Wrightian approach epistasis dominates the adaptive landscape such that the landscape is defined only by the epistatic interactions among mutations. Depending on the amount of epistasis and its (a priori imposed) distribution, smooth or rugged fitness landscapes can emerge (Goodnight 1988; Kauffman 1993; Hansen and Wagner 2001). …

Search by... Author
Show... All Results Primary Sources Peer-reviewed


An unknown error has occurred. Please click the button below to reload the page. If the problem persists, please try again in a little while.