ULK4 Genetic Variants Have Pleiotropic Effect on Risk of Autism, Associated with Brain mRNA Expression and Antipsychotic Treatment Response

Abstract

Background: ULK4 genetic variants have been implicated for adult-onset psychiatric disorders, and common variants are associated with hematologic and cardiologic disorders at genome-wide significance. This study aimed to examine the pleiotropic effect of ULK4 on the risk of autism, cis-association with mRNA and impact on antipsychotic treatment response in humans. Methods: The clinical genetic data comprised one cohort of autism case-parent triad sample in the Han Chinese and three cohorts of family-based samples in the European ancestry, from Autism Genetic Research Exchange, the Autism Genome Project and the Simons Foundation for Autism Research Initiative; mRNA expression in postmortem human prefrontal cortex across the lifespan and different brain regions of postmortem human brain and other tissues from two independent datasets were used for examining the cis-association with ULK4 variants. Antipsychotic treatment response data were from the Clinical Antipsychotic Trials in Intervention Effectiveness in patients with chronic schizophrenia. Transmission disequilibrium test was used to examine the genetic association with autism. General linear regression analysis was performed for cis-association with mRNA expression. The Cox proportion hazard model was used to analyze the primary outcome, the time to discontinued use of antipsychotics. Results: Multiple functional SNPs including rs2272007 in strong linkage disequilibrium at ULK4 were associated with autism in the Han Chinese sample (minimum p < 0.00071) which survived the Bonferroni correction for multiple testing. SNP rs2272007 and other SNPs were significantly associated with ULK4 expression in postmortem human prefrontal cortex in subjects across the lifespan and multiple brain areas in two independent datasets. In addition, two SNPs rs7651623 (Hazard Ratio, HR = 16.33; p = 5.00 × 10−4) and rs2030431 (HR = 17.25; p = 3.00 × 10−4) in strong LD were associated with the risk of discontinuing use of antipsychotic medications in the patients with schizophrenia. SNP rs2272007, perfect LD with rs7651623, was associated with treatment response in olanzapine only (HR = 4.22; p = 0.0034). Conclusion: We provide evidence at multiple layers for ULK4 common genetic variants associated with the risk of autism. This may have clinical implication for translational research and precision psychiatry