Epistasis

In Mendelian genetics, epistasis refers to the situation in which the phenotype caused by a mutation in one gene (A) masks the phenotype resulting from a mutation in another, nonallelic gene (B). Thus the individual carrying mutations in both genes A and B exhibits the same phenotype as the individual carrying a mutation in gene A alone. In this case, gene A is said to be epistatic over gene B, and gene B hypostatic to gene A. A classic example of epistasis is found among genes that determine mouse coat color. In the presence of albino mutant alleles, the mutant phenotypes of other coat-color determinants are not revealed, due to lack of pigment. Albino mutations, therefore, are epistatic over all other mutations that affect coat-color (1).

An epistatic relationship between two gene functions is often used to infer their relative order within a genetic pathway. Interpretation of such analysis depends on the type of genetic pathways. In biosynthetic pathways in which a series of genes modify a common intermediate substrate, epistatic genes are placed upstream of hypostatic ones. An example of such pathways is the morphogenetic pathway of T4 phage for head, tail, and tail-fiber assembly (2). In regulatory gene pathways, in contrast, where a member gene regulates the expression or activity of a downstream gene or gene product, epistatic genes are generally located downstream. A good example is found in the regulatory gene pathway that determines the dorsoventral body axis of the embryo in the fruit fly, Drosophila melanogaster (3). Most loss-of-function mutations that affect this pathway dorsalize the embryo. In contrast, a gain-of-function mutant allele of Toll, known as Toll^D, has been found that ventralizes the embryo. The presence of the opposite phenotypes allowed the investigators to combine a dorsalizing allele in a tester gene with the ventralizing allele of Toll and ask which of the two alleles is epistatic in double-mutant embryos. If the double mutant shows a dorsalized phenotype ) ie, the tester allele is epistative over Toll^D), the tester gene is placed downstream of Toll. If the double mutant shows a ventralized phenotype, the tester gene is placed upstream. In this way, the genes in the pathway that affect the embryonic dorsoventral axis have been placed upstream or downstream relative to the Toll gene.

 The logic of this type of analysis is straightforward if the pathway is a simple (ie, linear) series of genes or gene products. In such a pathway, the output depends solely on the activity of the terminal member of the pathway, and other members can be thought of as a series of on-off switches that determine the active or inactive state of the output gene. Regulatory gene pathways are not always simple, however, and there are situations in which epistatic genes act upstream. For excellent discussions of more complicated situations, as well as on assumptions and “rules” used in epistasis analysis (4).

 In population and quantitative genetics, epistasis refers to the genotype value for fitness contributed by gene interaction between two loci. The aggregate genotypic value of an individual (G) may deviate from the sum of the genotypic values attributable to the first locus (G_A) and the second locus) G_B). If it does, the deviation or nonadditive component (I_AB) is ascribed to epistasis, or interactions, between the two loci (5). Thus, G = G_A + G_B + I_AB. An example of epistasis can be seen when quantitative trait loci (QTL) from lines selected for high numbers of abdominal bristles in Drosophila are tested in pairwise combinations. In such tests, several pairs of QTLs for abdominal bristle number showed strong epistasis (interaction), resulting in abdominal bristle numbers that were beyond the additive effects of individual QTL (6).

References

1. W. K. Silvers (1979) The coat Colors of Mice, Springer-Verlag, New York. 

2. M. Levine (1969) Annu. Rev. Genet. 3, 323–342. 

3. K. V. Anderson et al. (1985) Cell 42, 779–789. 

4. L. Avery and S. Wasserman (1992) Trends Genet. 8, 312–316. 

5. D. S. Falconer and T. F. C. Mackay (1996) Introduction to Quantitative Genetics, 4th ed, Longman, Essex, England. 

6. A. D. Long et al. (1995) Genetics 139, 1273–1291.