Phillips PC. Epistasis-the essential role of gene interactions in the structure and evolution of genetic systems. Nat Rev Genet 9: 855-867

Center for Ecology and Evolution, University of Oregon, Eugene, Oregon 97403 USA.
Nature Reviews Genetics (Impact Factor: 36.98). 11/2008; 9(11):855-67. DOI: 10.1038/nrg2452
Source: PubMed


Epistasis, or interactions between genes, has long been recognized as fundamentally important to understanding the structure and function of genetic pathways and the evolutionary dynamics of complex genetic systems. With the advent of high-throughput functional genomics and the emergence of systems approaches to biology, as well as a new-found ability to pursue the genetic basis of evolution down to specific molecular changes, there is a renewed appreciation both for the importance of studying gene interactions and for addressing these questions in a unified, quantitative manner.

  • Source
    • "If the double mutant expression is lower than predicted, then mutations are considered to be in negative epistasis. If the individual mutation causes an increase in the expression but the double mutant containing that mutation leads to its reduction (or vice versa), the mutation is deemed to be in sign epistasis (Phillips 2008). Epistatic interactions between mutations in CREs define the robustness as well as the evolvability of regulatory elements—not only how transcription levels can be modulated but also how new functional CREs could evolve (Payne and Wagner 2014). "
    [Show abstract] [Hide abstract]
    ABSTRACT: Changes in gene expression are an important mode of evolution, however the proximate mechanism of these changes is poorly understood. In particular, little is known about effects of mutations within cis binding sites for transcription factors, or the nature of epistatic interactions between these mutations. Here, we tested the effects of single and double mutants in two cis binding sites involved in the transcriptional regulation of the Escherichia coli araBAD operon, a component of arabinose metabolism, using a synthetic system. This system decouples transcriptional control from any post-translational effects on fitness, allowing a precise estimate of the effect of single and double mutations, and hence epistasis, on gene expression. We found that epistatic interactions between mutations in the araBAD cis regulatory element are common, and that the predominant form of epistasis is negative. The magnitude of the interactions depended on whether the mutations are located in the same or in different operator sites. Importantly, these epistatic interactions were dependent on the presence of arabinose, a native inducer of the araBAD operon in vivo, with some interactions changing in sign (e.g., from negative to positive) in its presence. This study thus reveals that mutations in even relatively simple cis regulatory elements interact in complex ways such that selection on the level of gene expression in one environment might perturb regulation in the other environment in an unpredictable and uncorrelated manner.
    Preview · Article · Nov 2015 · Molecular Biology and Evolution
  • Source
    • "Furthermore, parallel changes at the nucleotide level appear more probable if there is little epistasis, i.e. when the fitness effects of individual mutations is independent of the particular order in which they appear (Lobkovsky & Koonin 2012). However, epistasis is pervasive in all genetic systems and consequently also in complex eukaryotes (Phillips 2008). Therefore, it appears more likely that early beneficial mutations have a particularly large influence on the further evolutionary trajectory via pleiotropic interactions, priority effects etc. (Lenormand et al. 2008). "
    [Show abstract] [Hide abstract]
    ABSTRACT: Replicated ecological gradients are prime systems to study processes of molecular evolution underlying ecological divergence. Here, we investigated the repeated adaptation of the neotropical fish Poecilia mexicana to habitats containing toxic hydrogen sulphide (H2S) and compared two population pairs of sulphide-adapted and ancestral fish by sequencing population pools of >200 individuals (Pool-Seq). We inferred the evolutionary processes shaping divergence and tested the hypothesis of increase of parallelism from SNPs to molecular pathways. Coalescence analyses showed that the divergence occurred in the face of substantial bidirectional gene flow. Population divergence involved many short, widely dispersed regions across the genome. Analyses of allele fre- quency spectra suggest that differentiation at most loci was driven by divergent selection, followed by a selection-mediated reduction of gene flow. Reconstructing allelic state changes suggested that selection acted mainly upon de novo mutations in the sulphide-adapted populations. Using a corrected Jaccard index to quantify parallel evolution, we found a negligible proportion of statistically significant parallel evolution of Jcorr = 0.0032 at the level of SNPs, divergent genome regions (Jcorr = 0.0061) and genes therein (Jcorr = 0.0091). At the level of metabolic pathways, the overlap was Jcorr = 0.2545, indicating increasing parallelism with increasing level of biological integration. The majority of pathways contained positively selected genes in both sulphide populations. Hence, adaptation to sulphidic habitats necessitated adjustments throughout the gen- ome. The largely unique evolutionary trajectories may be explained by a high proportion of de novo mutations driving the divergence. Our findings favour Gould’s view that evolution is often the unrepeatable result of stochastic events with highly contingent effects.
    Full-text · Article · Oct 2015 · Molecular Ecology
    • "KEYWORDS epistasis; genomic selection; genomic best linear unbiased prediction (G-BLUP); extended G-BLUP (EG-BLUP); reproducing kernel Hilbert space regression (RKHS); GenPred; shared data resource E PISTASIS has long been recognized as an important component in dissecting genetic pathways and understanding the evolution of complex genetic systems (Phillips 2008). It is hence a biologically influential component contributing to the genetic architecture of quantitative traits (Mackay 2014). "
    [Show abstract] [Hide abstract]
    ABSTRACT: Modelling epistasis in genomic selection is impeded by a high computational load. The extended genomic best linear unbiased prediction (EG-BLUP) with an epistatic relationship matrix and the reproducing kernel Hilbert space regression (RKHS) are two attractive approaches reducing the computational load. In this study, we proved the equivalence of EG-BLUP and genomic selection approaches explicitly modelling epistatic effects. Moreover, we have shown why the RKHS model based on a Gaussian kernel captures epistatic effects among markers. Using experimental data sets in wheat and maize, we compared different genomic selection approaches and concluded that prediction accuracy can be improved by modelling epistasis for selfing species but may not for out-crossing species. Copyright © 2015, The Genetics Society of America.
    No preview · Article · Jul 2015 · Genetics
Show more