Schizophrenia (SCZ) is a common and debilitating psychiatric disorder that imposes tremendous personal and societal burdens, and studies have demonstrated substantial heritability reflecting common and rare alleles at many loci. Most genetic or mechanistic studies of SCZ still focus predominantly on protein-coding genes; however, the majority of SCZ risk variants reside in noncoding regions of the genome. Long noncoding RNAs (lncRNAs) account for a significant fraction of functional noncoding elements and (like enhancers) are enriched for SCZ risk variants, but so far remain largely uncharacterized. Though the functions of most lncRNAs are unknown, many have now been implicated in the regulation of gene expression and chromatin architecture, and there is emerging evidence that lncRNAs are important during neurodevelopment. As such, there is an urgent need to understand their role in SCZ. Comprehensive profiling of lncRNAs has remained challenging because they are typically expressed at low levels compared to other transcripts. We therefore propose here to leverage new RNA Capture technologies, which we have developed and applied to control and autism brains, to deeply profile the lncRNA transcriptome in a uniquely large resource of diverse data types from post-mortem dorsolateral prefrontal cortex (DLPFC) brain samples of 350 SCZ cases and 350 matched controls. By deep short-read sequencing of samples enriched for lncRNAs using specific capture probes, we will identify lncRNAs that are dysregulated in SCZ cases (Aim 1). We will integrate our noncoding expression data with existing standard RNA-Seq data generated for the same samples by the CommonMinds Consortium (CMC) to construct coding/noncoding co-expression networks to identify key regulatory lncRNAs whose dysregulation may contribute to SCZ risk. Network analyses will be supported by the availability of high-resolution chromosome confirmation capture maps (Hi-C) to identify direct interactions between lncRNAs and their targets (Aim 2). Finally, in silico regulatory lncRNA predictions will be validated in-situ and in-vitro by mapping their complete genomic loci using full-length transcript sequencing technology, and analyzing the effect of lncRNA perturbations on target gene expression and regulatory interactions in neural cells derived from human induced pluripotent stem cells (hiPSCs) (Aim 3). As part of all these analyses we will also identify and integrate lncRNA gene expression quantitative trait loci (lncQTL) with existing PsychENCODE epigenetic histone modifications and open-chromatin QTLs for the same samples, to assess the mechanistic impacts of SCZ risk variants. Together these results will not only improve our understanding of the role of lncRNAs in SCZ etiology, potentially providing therapeutic targets, but also provide a robust framework for future noncoding RNA studies in any disease context.
Most genetic risk loci reported for schizophrenia (SCZ) to date reside in noncoding genomic regions. Here we plan to systematically assess the role of a large new class of genes encoding long noncoding RNAs (lncRNAs) in SCZ risk by deeply profiling lncRNA expression in postmortem brain tissue, identifying dysregulated lncRNAs in SCZ cases vs controls in coding/noncoding gene networks, and assessing how perturbations of regulatory lncRNAs affect their target genes by taking advantage of the latest genomics technologies, along with newly developed algorithms and experimental approaches.
