SEARCH

SEARCH BY CITATION

Keywords:

  • GWAS ;
  • Alcohol and Nicotine Codependence;
  • cis-eQTL;
  • IPO11 ;
  • HTR1A

Background

Alcohol and nicotine codependence can be considered as a more severe subtype of alcohol dependence. A portion of its risk may be attributable to genetic factors.

Methods

We searched for significant risk genomic regions specific for this disorder using a genome-wide association study. A total of 8,847 subjects underwent gene–disease association analysis, including (i) a discovery cohort of 818 European American cases with alcohol and nicotine codependence and 1,396 European American controls, (ii) a replication cohort of 5,704 Australian family subjects with 907 affected offspring, and (iii) a replication cohort of 449 African American cases and 480 African American controls. Additionally, a total of 38,714 subjects of European or African descent in 18 independent cohorts with 10 other nonalcoholism neuropsychiatric disorders were analyzed as contrast. Furthermore, 90 unrelated HapMap CEU individuals, 93 European brain tissue samples, and 80 European peripheral blood mononuclear cell samples underwent cis-acting expression quantitative locus (cis-eQTL) analysis.

Results

We identified a significant risk region for alcohol and nicotine codependence between IPO11 and HTR1A on chromosome 5q that was reported to be suggestively associated with alcohol dependence previously. In the European American discovery cohort, 381 single nucleotide polymorphisms (SNPs) in this region were nominally associated with alcohol and nicotine codependence (p < 0.05); 57 associations of them survived region- and cohort-wide correction (α = 3.6 × 10−6); and the top SNP (rs7445832) was significantly associated with alcohol and nicotine codependence at the genome-wide significance level (p = 6.2 × 10−9). Furthermore, associations for 34 and 11 SNPs were replicated in the Australian and African American replication cohorts, respectively. Among these replicable associations, 4 reached genome-wide significance level in the meta-analysis of European Americans and European Australians: rs7445832 (p = 9.6 × 10−10), rs13361996 (p = 8.2 × 10−9), rs62380518 (p = 2.3 × 10−8), and rs7714850 (p = 3.4 × 10−8). Cis-eQTL analysis showed that many risk SNPs in this region had nominally significant cis-acting regulatory effects on HTR1A or IPO11 mRNA expression. Finally, no markers were significantly associated with any other neuropsychiatric disorder examined.

Conclusions

We speculate that this IPO11-HTR1A region might harbor a causal variant for alcohol and nicotine codependence.