Interrogating the Links between Genetics and Psychiatric Disorders with High Throughput Assays

What It Covers

Joe Dougherty, associate professor at Washington University in St. Louis, explains how massively parallel reporter assays (MPRAs) help researchers identify which noncoding genetic variants among thousands of candidates actually affect gene expression, with applications to psychiatric disorders including autism, depression, and schizophrenia.

Key Questions Answered

• •Noncoding variant problem: Human genetics has identified 200+ schizophrenia loci and 200+ autism-associated genes, but genome sequencing reveals 10 times more noncoding mutations than coding ones. Unlike protein-coding variants where a triplet lookup table predicts consequences, noncoding variants require functional assays to determine which ones actually alter gene expression versus background noise.
• •MPRA scaling mechanism: Convert fluorescent reporter assays into sequencing experiments by attaching unique 6-8 nucleotide barcodes to each variant construct's 3' UTR. Using oligonucleotide synthesis providers like Twist Bioscience, researchers can order 10,000-20,000 oligos simultaneously, clone them in a single reaction, and measure all variants in one sequencing run rather than individual transfection wells.
• •Multi-SNP causality model: When Bernie Mulvey screened 30 depression-associated loci using MPRA, most loci contained multiple functional SNPs rather than a single causal variant. This finding shifts the research model away from identifying one causal SNP per locus toward understanding how several variants within a linked block act in concert to modulate disease risk.
• •Pharmacogenomics application: Retinoid transcription factor binding sites appeared repeatedly across depression-associated loci in Mulvey's MPRA data, connecting to Accutane's FDA black-box warning for increased suicidality. Re-running the MPRA experiment with retinoids added to cell culture revealed increased activity across multiple loci, demonstrating MPRAs can map drug-gene interactions at scale to identify at-risk patient populations.
• •Experimental quality controls: To achieve reliable MPRA results with small-effect common variants, use 6-10 barcodes per allele to prevent barcode sequence effects from overwhelming the SNP signal, maximize replicate transfection wells, select biologically relevant cell types where target transcription factors are expressed, and prioritize high-uniformity oligo pools to avoid deep sequencing costs and unrepresented library members.