Successful therapeutic intervention in new mouse models of frizzled 2-associated congenital malformations.

Frizzled 2 (FZD2) is a transmembrane Wnt receptor. We previously identified a pathogenic human FZD2 variant in individuals with FZD2-associated autosomal dominant Robinow syndrome. The variant encoded a protein with a premature stop and loss of 17 amino acids, including a region of the consensus dishevelled-binding sequence. To model this variant, we used zygote microinjection and i-GONAD-based CRISPR/Cas9-mediated genome editing to generate a mouse allelic series. Embryos mosaic for humanized Fzd2W553* knock-in exhibited cleft palate and shortened limbs, consistent with patient phenotypes. We also generated two germline mouse alleles with small deletions: Fzd2D3 and Fzd2D4. Homozygotes for each allele exhibit a highly penetrant cleft palate phenotype, shortened limbs compared with wild type and perinatal lethality. Fzd2D4 craniofacial tissues indicated decreased canonical Wnt signaling. In utero treatment with IIIC3a (a DKK inhibitor) normalized the limb lengths in Fzd2D4 homozygotes. The in vivo replication represents an approach for further investigating the mechanism of FZD2 phenotypes and demonstrates the utility of CRISPR knock-in mice as a tool for investigating the pathogenicity of human genetic variants. We also present evidence for a potential therapeutic intervention.

Differential requirements of tubulin genes in mammalian forebrain development.

Tubulin genes encode a series of homologous proteins used to construct microtubules which are essential for multiple cellular processes. Neural development is particularly reliant on functional microtubule structures. Tubulin genes comprise a large family of genes with very high sequence similarity between multiple family members. Human genetics has demonstrated that a large spectrum of cortical malformations are associated with de novo heterozygous mutations in tubulin genes. However, the absolute requirement for many of these genes in development and disease has not been previously tested in genetic loss of function models. Here we directly test the requirement for Tuba1a, Tubb2a and Tubb2b in the mouse by deleting each gene individually using CRISPR-Cas9 genome editing. We show that loss of Tubb2a or Tubb2b does not impair survival but does lead to relatively mild cortical malformation phenotypes. In contrast, loss of Tuba1a is perinatal lethal and leads to significant forebrain dysmorphology. We also present a novel mouse ENU allele of Tuba1a with phenotypes similar to the null allele. This demonstrates the requirements for each of the tubulin genes and levels of functional redundancy are quite different throughout the gene family. The ability of the mouse to survive in the absence of some tubulin genes known to cause disease in humans suggests future intervention strategies for these devastating tubulinopathy diseases.