Variant-to-gene-mapping followed by cross-species genetic screening identifies GPI-anchor biosynthesis as novel regulator of sleep

Abstract

Sleep is nearly ubiquitous throughout the animal kingdom, with deficiencies in sleep having been linked to a wide range of human disorders and diseases. While genome wide association studies (GWAS) in humans have been identified loci robustly associated with several heritable diseases or traits, little is known about the functional roles of the underlying causal variants in regulating sleep duration or quality. We applied an ATAC-seq/promoter focused Capture C methodology in iPSC-derived neural progenitors to carry out a variant-to-gene mapping campaign that identified 88 candidate sleep effector genes connected to relevant GWAS signals. To functionally validate the role of the implicated effector genes in sleep regulation, we performed a neuron-specific RNAi screen in the fruit fly, Drosophila melanogaster. This approach identified a number of genes that regulated sleep, including phosphatidylinositol N-acetylglucosaminyltransferase subunit Q (PIG-Q). This gene encodes an enzyme involved in the first step of glycosylphosphatidylinositol (GPI)-anchor biosynthesis. We show that flies deficient for PIG-Q have longer sleep during both the day and night due to an increase in the total number of sleep bouts. Subsequent systematic investigation of other PIG-family genes identified increased sleep in flies for multiple different genes within the PIG pathway. We then mutated the PIG-Q locus in zebrafish and identified similar increases in sleep to those observed in Drosophila, confirming deep homology of PIG-Q mediated sleep regulation. These results provide the first physical variant-to-gene mapping of human sleep genes, and reveals a novel and conserved role for GPI-anchor biosynthesis in sleep regulation.

Zebrafish screen of high-confidence effector genes at insomnia GWAS loci implicates conserved regulators of sleep-wake behaviors

“…1B). CRISPR mutagenesis at the single-cell stage produced highly-efficient disruption of the target gene in greater than 90 percent of cells as previously demonstrated 23,25 . Single guide RNAs (sgRNAs) were designed to target the exonic sequence with highest conservation between zebrafish and human.…”

“…Genes assessed in this study were previously implicated through our established variant-togene mapping approach [15][16][17][18][19]23 , which integrates chromatin conformation and accessibility datasets in relevant cell types to infer causal effector genes at GWAS loci. Promoters of the candidate effector genes used in this study reside in open chromatin and display high resolution chromatin contacts with putative insomnia causal variants associated with significant GWAS loci (Fig.…”

Validation of Candidate Sleep Disorder Risk Genes Using Zebrafish

“…The use of rodent models is limited by the challenges of generating, maintaining, and testing large numbers of mutant lines, while zebrafish provide a useful alternative vertebrate model to validate human candidate disease risk genes efficiently and cost-effectively, and to then perform mechanistic studies to understand the involvement of each gene in disease. Indeed, zebrafish have been used to validate human candidate genes associated with insomnia (Palermo et al, 2021 ), schizophrenia (Thyme et al, 2019 ), diabetes (Adeyemo et al, 2019 ), abnormal liver function (Liu et al, 2013 ), idiopathic scoliosis (Mathieu et al, 2021 ), attention deficit hyperactivity disorder (ADHD; Dark et al, 2020 ), and other psychiatric disorders (Tang et al, 2020 ). The strategies described in this review are not limited to GWAS, but can also be applied to other disease gene discovery strategies.…”

“…Strategies such as heuristic linkage disequilibrium, penalized regression, and Bayesian fine-mapping have been used to first narrow down causal variants and then genes using variant-to-gene mapping (Schaid et al, 2018 ). More recently, GWAS data and chromatin conformation capture were used for variant-to-gene mapping to identify 88 candidate insomnia risk genes (Palermo et al, 2021 ). The orthologs of many of these genes were then functionally tested using RNAi knockdown in Drosophila , which identified a small number of genes whose loss resulted in a sleep phenotype.…”

“…This targeted loss-of-function approach has also been used to characterize the swimming behavior of zebrafish containing mutations in the orthologs of 90 genes that have been associated with psychiatric disorders (Tang et al, 2020 ). Most recently, zebrafish were used to validate a candidate insomnia risk gene identified from a targeted Drosophila screen (Palermo et al, 2021 ) based on a human insomnia GWAS (Jansen et al, 2019 ). Automated annotation of the zebrafish genome found that over 70% of human genes have at least one zebrafish ortholog (Howe et al, 2013 ), and manual annotation suggests that this number is closer to 90% (Tran and Prober, unpublished observation).…”