The contribution of epistatic pleiotropy to the genetic architecture of covariation among polygenic traits in mice

The contribution that pleiotropic effects of individual loci make to covariation among traits is well understood theoretically and is becoming well documented empirically. However, little is known about the role of epistasis in determining patterns of covariation among traits. To address this problem we combine a quantitative trait locus (QTL) analysis with a two-locus model to assess the contribution of epistasis to the genetic architecture of variation and covariation of organ weights and limb bone lengths in a backcross population of mice created from the M16i and CAST/Ei strains. Significant epistasis was exhibited by 14 pairwise combinations of QTL for organ weights and 10 combinations of QTL for limb bone lengths, which contributed, on average, about 5% of the variation in organ weights and 8% in limb bone lengths beyond that of single-locus QTL effects. Epistatic pleiotropy was much more common in the limb bones (seven of 10 epistatic combinations affecting limb bone lengths were pleiotropic) than the organs (three of the 14 epistatic combinations affecting organ weights were pleiotropic). In both cases, epistatic pleiotropy was less common than single-locus pleiotropy. Epistatic pleiotropy accounted for an average of 6% of covariation among organ weights and 21% of covariation among limb bone lengths, which represented an average of one-fifth (for organ weights) and one-third (for limb bone lengths) of the total genetic covariance between traits. Thus, although epistatic pleiotropy made a smaller contribution than single-locus pleiotropy, it clearly made a significant contribution to the genetic architecture of variation/covariation.