The role of Lrp6-mediated Wnt/β-catenin signaling in the development and intervention of spinal neural tube defects in mice

Abstract

Neural tube defects (NTDs) are among the common and severe birth defects with poorly understood etiology. Mutations in the Wnt co-receptor LRP6 are associated with NTDs in humans. Either gain-of-function (GOF) or loss-of-function (LOF) mutations of Lrp6 can cause NTDs in mice. NTDs in Lrp6-GOF mutants may be attributed to altered β-catenin-independent noncanonical Wnt signaling. However, the mechanisms underlying NTDs in Lrp6-LOF mutants and the role of Lrp6-mediated canonical Wnt/β-catenin signaling in neural tube closure remain unresolved. We previously demonstrated that β-catenin signaling is required for posterior neuropore (PNP) closure. In the current study, conditional ablation of Lrp6 in dorsal PNP caused spinal NTDs with diminished activities of Wnt/β-catenin signaling and its downstream target gene Pax3, which is required for PNP closure. β-catenin-GOF rescued NTDs in Lrp6-LOF mutants. Moreover, maternal supplementation of a Wnt/β-catenin signaling agonist reduced the frequency and severity of spinal NTDs in Lrp6-LOF mutants by restoring Pax3 expression. Together, these results demonstrate the essential role of Lrp6-mediated Wnt/β-catenin signaling in PNP closure, which could also provide a therapeutic target for NTD intervention through manipulation of canonical Wnt/β-catenin signaling activities.

Keywords: Genetic rescue; Lrp6; Pharmacological intervention; Spinal neural tube defects; Wnt/β-catenin signaling.

Fig. 1.

Spinal bifida aperta and diminished canonical Wnt signaling by conditional ablation of Lrp6 in Pax3-expressing dorsal neural folds. (A) Dorsal–posterior view of an X-gal-stained (blue) E8.5 embryo for genetic fate mapping of Pax3^Cre/+;Rosa26-lacZ demonstrates the Cre recombination pattern in the dorsal region of the recently closed and pending-closing posterior neuropore (PNP; indicated by dashed line). (B-D) The conditional mutants of Pax3^Cre/+;Lrp6-cKO embryos exhibit open spinal neural tube defects (NTDs), as shown at E12.5 and E18.5. Dashed line brackets indicate the open lesion regions. (E-H) Sagittal caudal bodies of X-gal-stained Wnt/β-catenin signaling reporters BATgal or TOPgal show higher activities in the littermate control embryos (E,G) and diminished activities in the Pax3^Cre/+;Lrp6-cKO embryos (F,H) at E9.5. Arrows indicate recently closed dorsal neural tube regions. Arrowheads indicate the closing or pending-closing regions. nc, notochord. (I,J) Transverse sections show in situ hybridization signal of a Wnt/β-catenin target and feedback gene Axin2, which is high in the dorsal PNP of a littermate control (dashed line oval in I) and low in the mutant PNP (dashed line oval in J) at E9.5.

Fig. 2.

Wholemount in situ hybridization results show diminished gene expression of NTD-associated transcription factors Pax3, Cdx2 and Cdx4 in the dorsal PNPs of Pax3-Cre;Lrp6-cKOs at E9.5. (A-D) Pax3 expression is strong at the PNP closure site, as shown in a littermate control embryo (brackets in A, sagittal view and in B, dorsal view), whereas it is diminished specifically at the defective closure site of the mutant PNP (dashed line brackets in C,D). (E-H) Cdx2 is widely expressed in the caudal body of the control embryo, including dorsal PNP (bracket in E, sagittal view; arrowheads in F, transverse section from the region of the dashed line in E), and it is specifically diminished in the dorsal PNP of the mutant embryo (dashed line bracket in G and arrowheads in H). (I-L) Cdx4 is expressed in the dorsal PNP of the control embryo (bracket in I and arrowheads in J), and it is specifically diminished in the dorsal PNP of the mutant embryo (dashed line bracket in K and arrowheads in L). Asterisks indicate the dorsolateral hinge points.

Fig. 3.

Wholemount in situ hybridization results on Wnt genes and additionally relevant Wnt signaling downstream target genes around the PNP regions of littermate controls and Pax3-Cre;Lrp6-cKOs at E9.5. (A,B) Msx1 is restrictively expressed in the dorsal PNP of the normal control embryo (bracket in A), and its expression is significantly diminished in the mutant PNP (bracket in B). (C-J) No obvious changes in T (C,D), Tbx6 (E,F), Wnt1 (G,H) and Wnt5a (I,J) expression patterns around the mutant PNP regions were observed compared to respective expression patterns in the littermate controls.

Fig. 4.

Wholemount in situ hybridization results on Fgf genes and related Mesp2 expression around PNP regions of the littermate controls and Pax3-Cre;Lrp6-cKOs at E9.5. (A-F) No obvious differences in Fgf8 (A,B), Fgf17 (C,D) and Fgf18 (E,F) expression around PNP regions (brackets) between the control and mutant embryos were observed. (G,H) No obvious differences in Fgf-regulated Mesp2 expression in the presomites between the littermate control and mutant embryos were observed.

Fig. 5.

Noncanonical Wnt/PCP signaling activities in littermate controls and Pax3-Cre;Lrp6-cKOs at E9.5. (A-D) Transverse sections after wholemount in situ hybridization show no obvious changes in Vangl2 (A,B) and Ptk7 (C,D) expression patterns at the PNP closure sites in normal control and mutant embryos. (E) Immunoblots show no differences in phosphorylated (higher band that is linked with PCP signaling) and nonphosphorylated (lower band) Dvl2 proteins between the control and mutant PNP samples.

Fig. 6.

Proliferation and apoptosis at the PNP closure sites of the littermate controls and Pax3-Cre;Lrp6-cKOs at E9.5. (A-C) BrdU incorporation and detection experiments show no significant differences in proliferating cells in the dorsal PNPs above the dorsolateral hinge points (dashed lines in A,B, transverse PNP sections) between the control and mutant embryos. (D-F) TUNEL assays demonstrate no significant differences in apoptotic cells (green in D,E) in the dorsal PNPs between the control and mutant embryos. n.s., no statistical significance (P>0.05; unpaired, two-tailed Student's t-test).

Fig. 7.

Genetic rescue of PNP closure defects in the Pax3-Cre;Lrp6-cKOs by β-catenin gain-of-function (GOF). (A,B) Failed PNP closure as shown in the dorsal–posterior view of an Lrp6-cKO embryo (A) and in a transverse PNP section (arrows in B, cut through the dashed line in A) at E10.5. (C,D) Rescued PNP closure as shown in the dorsal–posterior view of an Lrp6-cKO;β-catenin-GOF embryo (A) and in a transverse PNP section that shows abnormally widened but closed dorsal PNP (arrowhead in D, cut through the dashed line in C) at E10.5. (E) A transverse PNP section of a littermate control embryo shows normally closed PNP at the dorsal midline (arrowhead in E) at E10.5.

Fig. 8.

Pharmaceutical intervention of spinal NTDs in Pax3-Cre;Lrp6-cKOs by maternal supplementation of a Wnt/β-catenin signaling agonist. (A) The embryo numbers and Lrp6-cKO ratios are not significantly different between the control and lithium-treated groups at E18.5 as compared with the expected Mendelian ratio (25% cKOs) (P>0.05, chi-square test). (B) Dorsal–caudal body views of a double heterozygous (Het) Pax3^Cre;Lrp6^flox/+ embryo that shows no NTD and with normal tail (B1), an Lrp6-cKO embryo in the control group that shows the severest lumbosacral NTD (with ∼8 mm lesion length, bracket in B2), an Lrp6-cKO embryo treated with lithium that shows a fully closed or rescued spinal cord (0 mm lesion length in B3, asterisk shows partially rescued tail growth), and a mutant embryo treated with lithium that shows milder NTD (with 3 mm lesion length in B4) at E18.5. (C) Rescue effects of spinal NTDs in Lrp6-cKOs examined at E18.5 after maternal supplementation of lithium chloride (LiCl) from E7.5 to E9.5. The lesion lengths (mm) were measured under a microscope. *P=0.02 (Fisher exact test); after all samples combined and averaged in each group, P=0.01 (unpaired, two-tailed Student's t-test). (D) RT-qPCR results demonstrate significant restoration of Pax3 mRNA in the lithium-treated Lrp6-cKO PNPs at E9.5. n.s., no statistical significance (P>0.05); **P<0.01 (unpaired, two-tailed Student's t-test). (E) Illustrative summary of Lrp6-mediated β-catenin–Pax3/Cdx2/Cdx4 signaling underlying PNP closure/elongation and spinal NTD; the latter can be rescued by either genetic activation of β-catenin or maternal supplementation of lithium ion, which stabilizes intracellular β-catenin by inhibiting Gsk3 in the canonical Wnt signaling pathway, thus restoring a key downstream transcription factor Pax3 in Lrp6-deficient PNPs. Red font, mutants and phenotypes; green font and arrows, genetic or pharmacological rescues demonstrated in the current study.