Despite major advances in genomic technologies, the integration of genetic data with clinical management remains challenging. This problem is particularly poignant for copy number variants (CNVs), lesions that typically perturb the dosage of dozens of genes and, as recent data have revealed, are major contributors to genetic disorders, most notably neurodevelopmental traits. Recent advances that afford superior resolution in the detection of genomic lesions have accelerated the discovery of CNVs in patients; however, with the exception of rare cases with point mutations of discrete genes within a CNV, the transition from CNV detection to assigning phenotypic contribution of specific genes remains largely intractable. We have taken an orthogonal approach to the problem, grounded on two key observations: a) that some CNVs associated with neurodevelopmental traits exhibit quantitative anatomical phenotypic surrogates; and b) that some CNVs manifest in reciprocal relationships (deletions and duplications of the same or overlapping genomic segment) with either similar or mirroring clinical phenotypes. Grounded on these observations, we have performed systematic overexpression and suppression studies in zebrafish embryos and have identified KCTD13 and CHD1L as the primary drivers of the neuroanatomical phenotypes associated with the 16p11.2 and 1q21.1 CNVs respectively, results that have been substantiated by the subsequent identification of rare atypical de novo deletions in these two genes in patients with autism spectrum disorders (ASD). Here we propose to develop further our tools to systematically dissect other reciprocal CNVs associated with neurocognitive traits. Under a series of filters that select for number of protein encoding genes, and mandate tractable anatomical features that can be recapitulated in zebrafish embryos, we will focus initially on 12 CNVs and assay in vivo the possible contribution of each gene contained within these lesions. Additionally, we will ask whether concomitant dosage misregulation of the primary CNV driver locus and each additional gene within the CNV might influence the expressivity of relevant anatomical, quantitative phenotypes. For the resulting group of candidate or contributing CNV driver genes, we will ask whether these genes a) are also involved in atypical, rare deletions identified by high resolution array comparative genomic hybridization (aCGH) datasets generated from cohorts with an enrichment of neurocognitive phenotypes; and b) bear point mutations in ASD exomes that will be functionally tested using a zebrafish model to assay pathogenicity and direction of effect. Finally, we will ask whether testing of all alleles detected for each transcript in cases and controls might reveal a mutational burden at these loci that is invisible by statistical means alone. Together, these data will identify a number of genes that drive neurocognitive effects in humans and provide an efficient platform for the systematic dissection of CNVs discovered in patient exomes and genomes.
The interpretation of genetic and genomic lesions discovered in patients is critical to diagnosis and clinical management and yet represents a fundamental bottleneck in human and medical genetics. In this proposal, we will develop further the tools to systematically dissect the primary genes that drive neuroanatomical phenotypes seen in patients who bear rare copy number variants (CNVs). This work will leverage existing genetic and genomic data to provide a platform for functional interpretation, but it is also directly relevant to the future dissection of CNVs relevant to a broad range of human genetic disorders.
