CONSIDER THE CONTEXT: CHARACTERIZING STIMULUS DEPENDENT GENETIC EFFECTS IN HUMAN NEURAL PROGENITOR CELLS

GWAS-implicated common genetic variants associated with complex brain traits and neuropsychiatric disorders primarily reside in non-coding regions. Functionally characterizing these variants is challenging because their regulatory effects are often context-dependent, becoming apparent only in specific cell types, developmental windows, or in response to stimulus. Studying variants in static, bulk post-mortem tissues obscure this crucial context-specificity. To address this, we utilized a "Gene-by-Environment (GxE) in a dish" approach, leveraging a large, genetically diverse cohort of primary human neural progenitor cells (hNPCs), a relevant fetal-stage context for cortical development. We performed multi-omic profiling, RNA-sequencing (RNA-seq) for gene expression and ATAC-sequencing (ATAC-seq) for chromatin accessibility, across disease-relevant contexts. These included stimulation of the fundamental neurodevelopmental Wnt signaling cascade and exposure to psychiatric medications Valproic Acid (VPA) and Lithium Chloride (LiCl). This interrogation successfully revealed thousands of genetic detected only under stimulated conditions. Many of the Wnt-stimulated eGenes were novel when compared to aggregated bulk-tissue cohorts, demonstrating that context is a crucial consideration in understanding genetic architecture. Furthermore, the Wnt-responsive regulatory quantitative trait loci (rQTLs) demonstrated significant colocalization with GWAS risk loci for schizophrenia and autism spectrum disorder (ASD), providing mechanistic insights. The analysis of Gene-by-Treatment (GxT) interactions with VPA identified genetically mediated alterations in key pathways, such as folate metabolism, associated with inter-individual variability in cognitive traits. In parallel, analysis of LiCl identified genetic variants influencing hNPC proliferation that colocalized with risk for bipolar disorder, establishing a molecular mechanism for differences in drug response. These findings provide a functional basis for developing targeted therapeutics and pharmacogenomic strategies to mitigate side effects or improve clinical outcomes in patient populations. We also generated a pilot genetic resource exploring the cellular responses to additional contexts, including maternal immune activation (MIA), signaling pathways, and estradiol. This further demonstrates the utility of context-dependent genetic mapping and provides targets for future, focused exploration. This research establishes a robust methodology and a comprehensive multi-omic resource to bridge the critical gap between genetic association and complex phenotypes. Our findings unequivocally reinforce that context is an essential layer of information required to fully understand genetic contributions to neurodevelopmental and neuropsychiatric disorder risk.