IPSC-derived neurons as a model for studying the role of RELN in autism

Abstract

RELN is strongly associated with Autism Spectrum Disorder (ASD). Homozygous loss of the encoded protein REELIN is associated with severe neurodevelopmental phenotypes characterized by lissencephaly and cerebellar hypoplasia, yet the ASD linked variants are typically heterozygous and appear to require additional genetic risk to cause ASD. To functionally characterize a RELN variant in a patient with ASD, we used induced pluripotent stem cells (iPSCs) from a family of non-autistic parents and their son who had ASD (the proband). The proband has a maternally-inherited missense variant (R2457C) in the RXR motif of the REELIN protein. We differentiated the iPSCs into two types of neurons, inhibitory neurons which model the inhibitory forebrain neurons that secrete REELIN, and excitatory neurons which model the cortical pyramidal neurons that respond to REELIN. Immunoblotting revealed that the proband inhibitory neurons had a lower ratio of extracellular/intracellular REELIN compared to that of the parental neurons, suggesting a decreased REELIN secretion. Sholl analysis on the proband excitatory neurons showed reduced dendritic complexity and reduced total length compared to the parental neurons. REELIN treatment increased the dendritic length and complexity in proband neurons up to the level of parental neurons. CRISPR/Cas9-mediated RELN KO did not change the dendritic phenotype in the excitatory neurons, ruling out a cell autonomous role for REELIN in these neurons. The proband excitatory neurons also had lower mRNA expression of WNT target genes in response to WNT3a, suggesting an underactive WNT signaling, as well as higher total GSK3β protein and lower phosphorylation at the inhibitory S9 site, indicating an overactive GSK3β signaling. Inhibition of GSK3β improved the proband neurons dendritic complexity in the proximal parts of the dendritic arbor. However, inhibition of mTOR signaling, which has shown to regulate REELIN signaling, did not change the dendritic morphology. In conclusion, the pathophysiology of ASD in the proband likely consists of a reduced REELIN secretion from the inhibitory neurons and an additional vulnerability in the REELIN-responding excitatory neurons, the latter likely being an overactive GSK3β and an underactive WNT signaling, all of which result in reduced dendritic complexity.