Figure 3.

Excess expression of Grhl3 results in spinal NTDs owing to failure of PNP closure. (A) Among litters from Grhl3^{ct/ct;TgGrhl3/0} matings analysed at E8.5–10.5, the PNP was significantly enlarged among Grhl3^{ct/ct;TgGrhl3/TgGrhl3} embryos (n = 64) from the 12–15 somite stage onwards (#, significant difference from both other genotypes, P < 0.001). PNP closure was normalized in Grhl3^{ct/ct;TgGrhl3/0} embryos (n = 123) compared with Grhl3^ct/ct littermates (n = 176) at 28–31 somites (^*, significantly different from Grhl3^ct/ct, P < 0.001). Mean ± SEM values; n = 4–77 embryos/genotype/stage (see Supplementary, Fig. S2 for plot of individual data). (B) Abundance of Grhl3 mRNA varies significantly with genotype at E9.5 and E10.5 (^*P < 0.0001 ANOVA; Holm-Sidak pairwise analysis). For context, we previously found that Grhl3 expression in ct/ct embryos at E10.5 was ∼50% of that in partially congenic wild-type (+^ct) embryos (11). (C) Analysis of individual embryos at E10.5 (28–29 somites) shows that increased abundance of Grhl3 genomic DNA (in transgenic embryos) results in gene dosage-dependent increase in Grhl3 mRNA expression. qG-PCR signal corresponds to 2, 3 and 4 copies of Grhl3 in ct/ct, hemizygous and homozygous transgenics, respectively. (D) Moderate overexpression of Grhl3 normalizes PNP closure in individual Grhl3^{ct/ct;TgGrhl3/0} embryos, compared with Grhl3^ct/ct, whereas excess expression prevents PNP closure, in Grhl3^{ct/ct;TgGrhl3/TgGrhl3}. (E–H) Scanning electron micrographs at E9 (13 somite stage) show typical appearance of the closing PNP. Neural folds are elevated and apposed in all three genotypes (E–G) but failure of closure progression is already apparent in Grhl3^{ct/ct;TgGrhl3/TgGrhl3} embryos (G,H), leaving a narrow unclosed region (compare region adjacent to white arrow E–H). Scale bars represent 0.1 mm.