Figure 5.

ddm1 is a zygotic modifier of genomic imprinting. (A) Crossing scheme to test the effect of ddm1 on mea seed abortion. Progeny from the cross fell into two classes, characterized by a different frequency of aborted seeds (Ab) and the presence or absence of enlarged seeds (E). (# seeds) Numer of seeds scored. Data of six plants from each class were pooled because pairwise χ² tests showed no significant differences in the frequency of aborted and viable [normal (N) plus enlarged seeds] produced by plants in either class. Among the plants carrying a mutant ddm1-2 allele one-eighth of the seeds that inherit a maternal mea-1^m allele are expected to also carry a wild-type paternal MEA^p allele and to be homozygous for ddm1-2 (0.5 × ¼ × ½ = 0.0625; X² = 2.50 < χ²_0.05[1] = 3.84; without correction for the Col/Ler background effect). Given that 3.6% of the seeds that inherit a mutant maternal mea-1^m allele survive in the Col/Ler hybrid background (see text), a frequency of 5.7% is in very good agreement with the expected 5.8% [(0.5 − 0.036) ×  = 0.058; X² = 0.86 < χ²_0.05[1] = 3.84; with correction for the Col/Ler background effect]. (B) Rescued mea-1^m/MEA^p; ddm1-2/ddm1-2 seeds (top) are larger than phenotypically wild-type siblings (bottom). (C) Embryos from dissected, rescued seeds (top) are larger than wild-type embryos (bottom). Bar, 400 μm.