Ultra-rare de novo damaging coding variants are enriched in attention-deficit/hyperactivity disorder and identify risk genes

Attention-deficit/hyperactivity disorder (ADHD) is a common and impairing neurodevelopmental disorder in which genetic factors play an important role. DNA sequencing of parent-child trios provides a powerful approach for identifying de novo (spontaneous) variants, which has led to the discovery of hundreds of clinically informative risk genes for other neurodevelopmental disorders but has yet to be extensively leveraged in studying ADHD. Here, we conducted whole-exome DNA sequencing in 152 parent-child trios with ADHD and demonstrate for the first time a significant enrichment of rare and ultra-rare de novo protein-truncating variants and missense variants predicted to be damaging in ADHD cases compared to unaffected controls. Combining these results with a large independent case-control DNA sequencing cohort (3,206 ADHD cases and 5,002 controls), we identify lysine demethylase 5B (KDM5B) as a high-confidence risk gene for ADHD as well as two likely risk genes. We estimate that 862 genes contribute to ADHD risk. Finally, using our list of genes harboring ultra-rare de novo damaging variants, we show that these genes overlap with previously reported risk genes for other neuropsychiatric conditions in both DNA sequencing and genome-wide association studies. We also show that these genes are enriched for several canonical biological pathways, suggesting early neurodevelopmental underpinnings of ADHD. Overall, this work provides critical new insight into the biology of ADHD and demonstrates the discovery potential of DNA sequencing in larger parent-child trio cohorts. Significance statementGiven the important role of genetic factors in the development of attention-deficit/hyperactivity disorder (ADHD), research aimed at identifying risk genes can provide critical insight into underlying biological processes. We conducted whole-exome DNA sequencing in parent-child trios with ADHD, showing that these children have a significantly greater rate of rare and ultra-rare de novo gene-damaging mutations compared to unaffected controls, expanding our understanding of the genetic landscape of ADHD. We then use this information to identify KDM5B as a high-confidence risk gene for ADHD and highlight several enriched biological pathways. This work advances our etiologic understanding of ADHD and illustrates a previously unexplored path for risk gene discovery in this common neurodevelopmental disorder.