Functional dissection of the recurrent reciprocal 1q21.1 autism-associated CNV. N. Katsanis1,2, I. Blumenthal3,4, A. Ragavendran3,4, M. E. Talkowski3,4, C. Golzio1 1) Center for Human Disease Modeling and Department of Cell biology, Duke University, Durham, NC; 2) Department of Pediatrics, Duke University, Durham, NC; 3) Molecular Neurogenetics Unit, Center for Human Genetic Research, Massachusetts General Hospital, Boston, MA; 4) Departments of Neurology and Genetics, Harvard Medical School, Boston, MA.

   Copy number variants (CNVs) are frequent lesions involved in both rare and complex human traits. This has raised the challenge of identifying which genes within the CNV drive observed clinical traits. We have previously shown how the combinatorial use of neuroanatomical surrogate phenotypes in zebrafish embryos and genomic studies can dissect a single contributory locus to these phenotypes within the 600kb CNV on 16p11.2 associated with multiple developmental and psychiatric disorders. These findings prompted us to ask whether this approach could be useful for the systematic dissection of other CNVs that manifest similar defects. The 1q21.1 CNV is associated with congenital heart defects, head size defects, and represents the second most common lesion in autism. We noted that the minimal nine-gene deletion of this CNV is associated with microcephaly, while the reciprocal duplication is associated with macrocephaly, rendering it tractable by our methods. Systematic overexpression and suppression of all genes in the CNV provided a striking result that only one gene, the chromodomain-helicase-DNA-binding protein CHD1L, gave significant head size changes; overexpression of human CHD1L mRNA led to macrocephaly, while suppression of chd1l lead to a significant decrease in head size. These phenotypes were likely driven by defects in neurogenesis; chd1l morphants exhibited reduced rates of neuronal proliferation while, conversely, overexpressants showed increased rates of neuronal proliferation in vivo. Given the role of chromodomain genes in transcriptional regulation, and their emerging roles in neuropsychiatric disorders, we next asked which genes are regulated by CHD1L during brain development and whether there are any common transcriptional networks relevant to autism. We thus performed a transcriptomic evaluation of RNA-seq data from heads of control, CHD1L RNA-, and chd1l MO-injected embryos at 5 days post-fertilization. Using a linear regression model, we identified 257 loci with significant reciprocal differential expression correlated with gene dosage. An intriguing subset these loci have been associated with autism or connected to known loci by a first-order interaction. We are evaluating this dysregulated network in vivo with regard to neurogenesis and head size regulation and we will present data on the ability of this gene set to contribute to neurodevelopment and neurodevelopmental disorders.

You may contact the first author (during and after the meeting) at