Over-expression of Grhl2 causes spina bifida in the Axial defects mutant mouse.

Cranial neural tube defects (NTDs) occur in mice carrying mutant alleles of many different genes, whereas isolated spinal NTDs (spina bifida) occur in fewer models, despite being common human birth defects. Spina bifida occurs at high frequency in the Axial defects (Axd) mouse mutant but the causative gene is not known. In the current study, the Axd mutation was mapped by linkage analysis. Within the critical genomic region, sequencing did not reveal a coding mutation whereas expression analysis demonstrated significant up-regulation of grainyhead-like 2 (Grhl2) in Axd mutant embryos. Expression of other candidate genes did not differ between genotypes. In order to test the hypothesis that over-expression of Grhl2 causes Axd NTDs, we performed a genetic cross to reduce Grhl2 function in Axd heterozygotes. Grhl2 loss of function mutant mice were generated and displayed both cranial and spinal NTDs. Compound heterozygotes carrying both loss (Grhl2 null) and putative gain of function (Axd) alleles exhibited normalization of spinal neural tube closure compared with Axd/+ littermates, which exhibit delayed closure. Grhl2 is expressed in the surface ectoderm and hindgut endoderm in the spinal region, overlapping with grainyhead-like 3 (Grhl3). Axd mutants display delayed eyelid closure, as reported in Grhl3 null embryos. Moreover, Axd mutant embryos exhibited increased ventral curvature of the spinal region and reduced proliferation in the hindgut, reminiscent of curly tail embryos, which carry a hypomorphic allele of Grhl3. Overall, our data suggest that defects in Axd mutant embryos result from over-expression of Grhl2.

Toward positional cloning of the curly tail gene.

BACKGROUND: The curly tail (ct) mutant mouse is one of the best-studied mouse models of spina bifida. The ct mutation has been localized to distal chromosome 4 in two independent studies and was recently postulated to be in the Grhl-3 gene. METHODS: A recombinant BALB/c-ct strain was generated and used to precisely map the ct gene. RESULTS: We report the absence of gross chromosomal abnormalities and the precise mapping of the ct gene to a 3-Mb region at 135 Mb (66 cM) from the centromere, closely linked to the polymorphic microsatellite marker D4Mit148. Candidate genes, Idb3, Wnt4, Cdc42, and perlecan, all localized in the critical region, were studied by sequence and expression analyses. Our data indicate that these genes in all probability do not account for the ct phenotype. In addition, our expression data do not provide strong evidence that Grhl-3 is indeed the ct gene. CONCLUSIONS: The ct gene has not yet been identified. A total of 29 candidate genes remain present in the critical region. Refined mapping studies need to be performed to further narrow the region and additional candidate genes need to be examined. Supplementary material for this article can be found on the Birth Defects Research (Part A) website (http://www.mrw.interscience.wiley.com/suppmat/1542-0752/suppmat/2005/73/tables_S3-S6.doc).