A novel meiotic drive locus almost completely distorts segregation in mimulus (monkeyflower) hybrids

Abstract

We report the discovery, mapping, and characterization of a meiotic drive locus (D) exhibiting nearly 100% nonrandom transmission in hybrids between two species of yellow monkeyflowers, outcrossing Mimulus guttatus and selfing M. nasutus. Only 1% of F(2) hybrids were M. nasutus homozygotes at the marker most tightly linked to D. We used a set of reciprocal backcrosses to distinguish among male-specific, female-specific, and zygote-specific sources of transmission ratio distortion. Transmission was severely distorted only when the heterozygous F(1) acted as the female parent in crosses to either parental species, ruling out pollen competition and zygote mortality as potential sources of drive. After four generations of backcrossing to M. nasutus, nearly isogenic lines were still >90% heterozygous at markers linked to D, suggesting that heterozygosity at the drive locus alone is sufficient for nonrandom transmission. A lack of dramatic female fitness costs in these lines rules out alternatives involving ovule or seed mortality and points to a truly meiotic mechanism of drive. The strength and direction of drive in this system is consistent with population genetic theory of selfish element evolution under different mating systems. These results are the first empirical demonstration of the strong female-specific drive predicted by new models of selfish centromere turnover.

Figure 1.—

Map position and transmission ratio distortion of markers on LG11 in the M. nasutus × M. guttatus F₂ mapping population (N = 287). The horizontal line indicates the Mendelian expectation for either parental homozygote class (0.25). The dotted segment between bc374 and MgSTS87 continues the symmetry of homozygote frequencies shown by the other markers, but we do not have data on the frequency of M. guttatus homozygotes at bc374 because it is a dominant AFLP. Distances are in cM Kosambi.

Figure 2.—

Transmission ratio distortion at LG11 markers in backcross populations. Data for codominant markers MgSTS87-MgSTS26 are shown. The solid line is the Mendelian expectation of 0.50. Values above the dotted line deviated significantly (P < 0.05) from the Mendelian expectation by one-tailed χ² with 1 d.f., but each value was tested against a slightly different threshold because of differences in sample size. Because linked markers are not independent and the pattern in F₁ female backcrosses is clear, we did not correct for multiple tests. The most distorted marker in the F₁ male × M. guttatus backcross (MgSTS87; 61% M. guttatus transmission through F₁) deviates from the Mendelian expectation with P = 0.0027, suggesting that milder distortion in this cross also has a real biological basis.

Figure 3.—

LG11 heterozygosity in NILs formed by four generations of backcrossing to the M. nasutus parental line (BN4, N = 187). Data for codominant markers MgSTS87-MgSTS26 are shown. The dashed line indicates the expected (Mendelian) heterozygote frequency (0.0625). The F₂, BC₁, and NIL data suggest that the D locus is within 2 cM of MgSTS87.