Patient-specific iPSC models of neural tube defects identify underlying deficiencies in neuroepithelial cell shape regulation and differentiation

Abstract

Spina bifida and anencephaly are neural tube defects caused by failure of embryonic neural tube closure. Successful closure requires apicobasal elongation and apical constriction of neuroepithelial cells. These critical behaviours have not been studied in human, patient-specific models. We characterise a human iPSC-derived neuroepithelial morphogenesis model that is highly reproducible across three parental iPSC lines of diverse origin and reprogramming technologies. Differentiated neuroepithelial cells actively undergo ROCK-dependent apical constriction. ROCK acts downstream of planar cell polarity/VANGL2 in other species. We find that VANGL2 is up-regulated and phosphorylated during iPSCs-to-neuroepithelial differentiation. The patient-identified VANGL2-R353C mutation does not alter VANGL2 expression, localisation or phosphorylation, but reduces myosin-II phosphorylation and apical constriction relative to congenic control iPSCs. Non-congenic comparisons and forward genetic testing are also informative in this reproducible model. We compare lines reprogrammed from amniocytes of two patients with spina bifida, versus two controls. One patient-derived line forms a morphologically normal neuroepithelium, but fails to differentiate into neurons. The second fails to undergo apicobasal elongation, and harbours compound heterozygous mutations in the MED24 gene previously implicated in neuroepithelial elongation in mice. Thus, iPSC-derived neuroepithelial modelling reveals mechanistic insights into conserved cell behaviours, including apical constriction, apicobasal elongation and neural differentiation, links genetically-impaired apical constriction to human disease, and establishes patient-specific models which recapitulate failures across these heterogenous neuroepithelial functions.