Risk Locus Identification Ties Alcohol Withdrawal Symptoms to SORCS2

• Published in: Alcoholism, clinical and experimental research clinical and experimental research, 2018-12, Vol.42 (12), p.2337-2348
• Main Authors: Smith, Andrew H., Ovesen, Peter L., Skeldal, Sune, Yeo, Seungeun, Jensen, Kevin P., Olsen, Ditte, Diazgranados, Nancy, Zhao, Hongyu, Farrer, Lindsay A., Goldman, David, Glerup, Simon, Kranzler, Henry R., Nykjær, Anders, Gelernter, Joel
• Format: Article
• Language: English
• Subjects: Adult; Alcohol; Alcohol use; Alcohol Withdrawal; Alcohol Withdrawal Seizures - genetics; Alcoholic beverages; Black People; Cell culture; Cell Line; Chromosome 4; Computational Biology; Dexamethasone - pharmacology; Ethanol; Female; Genome-wide association studies; Genome-Wide Association Study; Genomes; Genotyping; Haplotypes; Hippocampus - metabolism; Human Genetics; Humans; Male; Meta-analysis; Middle Aged; Minority & ethnic groups; Nervous system; Neurotrophic factors; Receptors, Cell Surface - genetics; Risk Factors; Seizures; Stress; Stress Hormones; White People
• ISSN: 0145-6008; 1530-0277; 1530-0277
• DOI: 10.1111/acer.13890

Background Efforts to promote the cessation of harmful alcohol use are hindered by the affective and physiological components of alcohol withdrawal (AW), which can include life‐threatening seizures. Although previous studies of AW and relapse have highlighted the detrimental role of stress, little is known about genetic risk factors. Methods We conducted a genome‐wide association study of AW symptom count in uniformly assessed subjects with histories of serious AW, followed by additional genotyping in independent AW subjects. Results The top association signal for AW severity was in sortilin family neurotrophin receptor gene SORCS2 on chromosome 4 (European American meta‐analysis n = 1,478, p = 4.3 × 10−9). There were no genome‐wide significant findings in African Americans (n = 1,231). Bioinformatic analyses were conducted using publicly available high‐throughput transcriptomic and epigenomic data sets, showing that in humans SORCS2 is most highly expressed in the nervous system. The identified SORCS2 risk haplotype is predicted to disrupt a stress hormone‐modulated regulatory element that has tissue‐specific activity in human hippocampus. We used human neural lineage cells to demonstrate in vitro a causal relationship between stress hormone levels and SORCS2 expression, and show that SORCS2 levels in culture are increased upon ethanol exposure and withdrawal. Conclusions Taken together, these findings indicate that the pathophysiology of withdrawal may involve the effects of stress hormones on neurotrophic factor signaling. Further investigation of these pathways could produce new approaches to managing the aversive consequences of abrupt alcohol cessation. This study identifies genome‐wide significant risk variants in SORCS2 associated with increased alcohol withdrawal severity. The SORCS2 (black) risk haplotype (red) is predicted to disrupt a stress hormone‐modulated regulatory element (green) that has tissue‐specific activity in human hippocampus (dark blue). We demonstrate in vitro a causal relationship between stress hormones and SORCS2 expression, and show that SORCS2 levels increase upon exposure to and withdrawal from ethanol. The present work thus moves from genetic association signals to insight into regulatory processes.