Fig. 2. Maternal Dkk1 inhibits canonical Wnt signaling. (A) Embryos derived from sibling control (uninj.) and Dkk1-depleted oocytes (5, 7.5 and 10 ng oligo injected) at the early tailbud stage. This phenotype was seen in seven experiments in a total of 85% of cases (124/165). (B) The phenotype of Dkk1-depleted (Dkk^-) embryos was partially rescued by the reintroduction of 20 pg human Dkk1 mRNA (Dkk^-+mRNA) before fertilization. Here, Dkk1-depleted embryos (14/17) had the elongated phenotype shown compared with 4/14 for Dkk1^-+mRNA and 0/24 uninjected; 20 pg human Dkk1 mRNA alone (Dkk mRNA) caused enlargement of head structures (16/18). The experiment was repeated with a similar result. (C) The relative expression levels of Xnr3 and Xnr5 in control (uninj.), in Dkk1 depleted (Dkk^-), in 20 pg human Dkk1 mRNA (Dkk mRNA) and in Dkk1 depleted+20 pg human Dkk1 mRNA injected (Dkk^-+mRNA) embryos assayed by real-time RT-PCR at the late blastula stage; siblings of those shown in B. (D) TOPflash reporter activation after injection into two dorsal cells of four-cell stage control embryos compared with sibling Dkk1-depleted embryos frozen at the eight-cell, mid-(stage 8), late blastula (stages 9, 9.5) and early gastrula stages (stages 10, 10.5). (E) Western blot of total β-catenin protein in control and Dkk1-depleted sibling early blastulae (stage 7), using -tubulin as a loading control. Quantitation is shown on the right. (F,G) In situ hybridization of sibling control and Dkk1-depleted early gastrulae (Xnr5, F; chordin, G). (H) Western blot of phospho-Smad2 and -Smad1 proteins in control and Dkk1-depleted sibling at late blastulae (stage 9.5) and early gastrulae (stage 10.5). Before freezing, embryos were hemisected into batches of four dorsal (dor) and four ventral (ven) halves.