Cross-species variant-to-function analyses implicate MEIS1 in conferring sleep abnormalities and impaired cerebellar development

Abstract

Genome wide association study (GWAS) reports substantially outpace subsequent functional characterization. Pinpointing the causal effector gene(s) at GWAS loci remains challenging given the non-coding genomic residency of >98% of these signals. We previously implicated effector genes at GWAS loci for the complex and polygenic disorder of human insomnia using a high-resolution cell-specific, chromatin capture-based variant-to-gene mapping protocol, paired with sleep phenotyping in Drosophila. In this study, we leveraged a diurnal vertebrate model with higher genomic conservation, namely zebrafish, to screen our six highest confidence candidate genes and identify those whose loss-of-function impaired sleep characteristics related to human insomnia-like behaviors. Of these genes, we observed that CRISPR-mediated deletion of the zebrafish ortholog of MEIS1 produced nighttime specific sleep fragmentation and increased latency to sleep, pointing to a conserved role for MEIS1 in sleep maintenance. Comparing our human cell-based chromatin accessibility and contact maps with publicly available zebrafish spatial genomic data revealed highly conserved genomic architecture harboring the insomnia GWAS variant of interest. Notably, this genomic conservation was selective for the zebrafish ortholog which contributed to the sleep phenotype, meis1b, while the duplicated ohnolog meis1a proved dispensable. Motivated by this, we characterized the spatio-temporal expression of meis1b in zebrafish, showing it is comparable to human with respect to cerebellar granule progenitors. Ultimately, we found that loss of meis1b impairs cerebellar development. Together, our work provides a powerful model for screening human disorder risk genes for sleep fragmentation using a tractable vertebrate and supports a conserved cerebellar role for MEIS1 in sleep disturbance.