NEW YORK — Using a massively parallel reporter assay (MPRA) and other techniques, researchers have functionally validated thousands of schizophrenia risk variants from previous genome-wide association studies (GWAS), finding more than 400 with allelic regulatory effects.
As reported in Cell Genomics on Thursday, a team led by researchers from the University of North Carolina at Chapel Hill used MRPA on 5,173 fine-mapped schizophrenia GWAS variants in primary human neural progenitor cells to determine their function.
While previous GWAS have identified hundreds of risk loci, the authors wrote, their function has been largely unknown as many reside in noncoding DNA regions and because each locus comprises many different variants.
For their study, the researchers built an adeno-associated virus-based MPRA vector (AAV-MPRA) comprising a 150 bp target sequence with the variant in the center, a minimal promoter, green fluorescent protein, and a 20 bp unique barcode.
Next, they generated an AAV-MPRA library from a computationally predicted set of schizophrenia risk variants, including 5,173 variants in their final design that covered 143 GWAS loci.
Subsequently, they packaged the schizophrenia MPRA library into the AAV, administered it to human neural progenitor cells, and quantified the expression of green fluorescent protein using RNA sequencing two weeks later.
The reason behind using human neural progenitors was that schizophrenia genetic risk factors are enriched in regulatory elements of the developing cortex, and these cells model human neural development, the authors noted.
The experiments identified 439 MPRA-positive variants that showed allelic regulatory activity. Among them were only a few index variants at a given GWAS locus — those variants with the most significant GWAS association — the authors noted.
Highlighting the benefits of the MPRA, they said that the approach refined the number of regulatory variants and reduced the number of variants per locus. While, on average, 36.2 variants per locus were identified via computational fine-mapping approaches, MPRA trimmed those to 4.3 variants per locus.
Many of the MPRA-tested variants were located in intergenic and intronic regions, with only a few present in exons and promoters.
Further analysis of the variants revealed that many do not operate in the conventional way of promoting strong enhancer activity. In fact, 77 percent of the MPRA-positive variants were annotated by chromatin states, suggesting that they can exert their regulatory effects through various epigenetic mechanisms, the researchers explained.
Further, by combining their results with those from brain expression quantitative trait loci studies, they found that 64 percent of MPRA-positive variants did not overlap with eQTL variants. The characterization of GWAS variants via MPRA could therefore identify functional regulatory variants under selective pressure that eQTLs may be unable to detect, they wrote.
They also combined allelic activity, chromatin accessibility data, and chromatin contact frequency, using what they referred to as the accessibility-by-contact (ABC) model. Using this model, they could link variants to their putative gene targets and predict their expression. The findings also suggested that the ABC model was superior to an additive model in which allelic activity of variants is added to predict gene expression.
Overall, the findings shed light on the complex mechanisms underlying the molecular pathology of schizophrenia, the authors concluded.