FOXC1 Is Required for Normal Cerebellar Development and Is a Major Contributor to Chromosome 6p25.3 Dandy-Walker Malformation

Background: Cerebellar malformations (CbM) are common human structural birth defects of the brain that affect an estimated 1/5,000 births and typically cause mental retardation, developmental delay, motor and visual impairments, and have been observed in a subset of patients with autism spectrum disorder (ASD). Dandy-Walker malformation (DWM) is the most common CbM and is linked to at least four loci in humans. DWM gene identification remains a challenge due to its complex inheritance pattern and frequent confusion with other malformations of the cerebellum and posterior fossa. However, identifying the molecular basis for malformations of the cerebellum and posterior fossa provides a method to clarify whether these abnormalities represent distinct entities or examples of phenotypes along a spectrum of the same molecular pathogenesis. Objectives: To classify CbM among patients with 6p25.3 copy number variants (CNVs), identify the major DWM-causative gene on chromosome 6p25.3, and test the hypothesis that molecularly defined CbM represent an ASD subgroup. Methods: CNV mapping and MRI evaluation were performed for patients with chromosome 6p25.3 CNVs to determine the minimal region of overlap on 6p25.3 associated with CbM. Mouse in situ hybridization (ISH) expression analyses and examination of mouse mutants was performed for the eight 6p25.3 DWM candidate genes. Extensive analysis of Foxc1 mutant mice was also performed. Results: Human CNV and mouse ISH expression analyses identified FOXC1 as the major DWM-causative gene on 6p25.3. In parallel, examination of Foxc1 mouse mutants revealed striking deficits in early cerebellar development, including an early expansion of roof plate and choroid plexus and premature loss of Math1 in cerebellar rhombic lip progenitors. Mice with a hypomorphic Foxc1 allele additionally displayed abnormalities in midline fusion, foliation, and Purkinje cell morphology consistent with a mouse DWM-like phenotype. The expression of Foxc1 in the cerebellum-adjacent mesenchyme regulates secretion of trophic factors required for cerebellar development and implicates skull developmental abnormalities as an integral component of DWM pathogenesis. Additional MRI analysis indicated that mutation of the human FOXC1 gene alone is sufficient to cause cerebellar abnormalities, but not the full DWM phenotype. Three DWM deletion 6p25.3 patients were also diagnosed with ASD and an additional deletion 6p25.3 ASD patients were reported in the literature without MRIs to evaluate for CbM. Conclusions: FOXC1 regulates cerebellar development in both humans and mice, but must interact with other 6p25.3 genes to cause the complete DWM phenotype. Together, these studies provide novel mechanistic insight into the pathogenesis of both DWM and ASD and highlight the importance of reciprocal diagnostic evaluation for patients with these disorders identified in the clinic.