A shared pathogenic mechanism for valproic acid and SHROOM3 knockout in a brain organoid model of neural tube defects

Neural tube defects (NTDs) including anencephaly and spina bifida are common major malformations of fetal development resulting from incomplete closure of the neural tube. These conditions lead to either universal death (anencephaly) or life-long severe complications (spina bifida). Despite hundreds of genetic mouse models having neural tube defect phenotypes, the genetics of human NTDs are poorly understood. Furthermore, pharmaceuticals such as antiseizure medications have been found clinically to increase the risk of NTDs when administered during pregnancy. Therefore, a model that recapitulates human neurodevelopment would be of immense benefit to understand the genetics underlying NTDs and identify teratogenic mechanisms. Using our self-organizing single rosette spheroid (SOSRS) brain organoid system, we have developed a high-throughput image analysis pipeline for evaluating SOSRS structure for NTD-like phenotypes. Similar to small molecule inhibition of apical constriction, the antiseizure medication valproic acid (VPA), a known cause of NTDs, increases the apical lumen size and apical cell surface area in a dose-responsive manner. This expansion was mimicked by GSK3-β and HDAC inhibitors; however, RNA sequencing suggests VPA does not inhibit GSK3-β at these concentrations. Knockout of SHROOM3, a well-known NTD-related gene, also caused expansion of the lumen as well as reduced f-actin polarization. The increased lumen sizes were caused by reduced cell apical constriction suggesting that impingement of this process is a shared mechanism for VPA treatment and SHROOM3-KO, two well-known causes of NTDs. Our system allows the rapid identification of NTD-like phenotypes for both compounds and genetic variants and should prove useful for understanding specific NTD mechanisms and predicting drug teratogenicity.