Genome-wide association study of alcohol dependence:significant findings in African- and European-Americans including novel risk loci

Abstract

We report a GWAS of alcohol dependence (AD) in European-American (EA) and African-American (AA) populations, with replication in independent samples of EAs, AAs and Germans. Our sample for discovery and replication was 16 087 subjects, the largest sample for AD GWAS to date. Numerous genome-wide significant (GWS) associations were identified, many novel. Most associations were population specific, but in several cases were GWS in EAs and AAs for different SNPs at the same locus,showing biological convergence across populations. We confirmed well-known risk loci mapped to alcohol-metabolizing enzyme genes, notably ADH1B (EAs: Arg48His, P=1.17 × 10(-31); AAs: Arg369Cys, P=6.33 × 10(-17)) and ADH1C in AAs (Thr151Thr, P=4.94 × 10(-10)), and identified novel risk loci mapping to the ADH gene cluster on chromosome 4 and extending centromerically beyond it to include GWS associations at LOC100507053 in AAs (P=2.63 × 10(-11)), PDLIM5 in EAs (P=2.01 × 10(-8)), and METAP in AAs (P=3.35 × 10(-8)). We also identified a novel GWS association (1.17 × 10(-10)) mapped to chromosome 2 at rs1437396, between MTIF2 and CCDC88A, across all of the EA and AA cohorts, with supportive gene expression evidence, and population-specific GWS for markers on chromosomes 5, 9 and 19. Several of the novel associations implicate direct involvement of, or interaction with, genes previously identified as schizophrenia risk loci. Confirmation of known AD risk loci supports the overall validity of the study; the novel loci are worthy of genetic and biological follow-up. The findings support a convergence of risk genes (but not necessarily risk alleles) between populations, and, to a lesser extent, between psychiatric traits.

Figure 1

Chromosome 4 regional Manhattan plots. (a) European-Americans: regional Manhattan plot for the ADH gene cluster and adjacent regions on chromosome 4 showing the adjusted symptom count meta-analysis P-values from the discovery sample EAs, as well as points for the meta-analysis results after including our replication sample (diamonds). The points are colored based on degree of LD with rs1229984. Square points indicate genotyped SNPs and round points indicate imputed SNPs. All SNPs in the replication sample were genotyped. Data include the GCD sample plus SAGE, and GCD and German replication samples (diamonds). (b) African-Americans: regional Manhattan plot for the ADH gene cluster and adjacent regions on chromosome 4 showing the adjusted symptom count meta-analysis P-values from AAs in the discovery sample, as well as points for the meta-analysis result after inclusion of our replication sample (diamonds). The points are colored based on degree of LD with rs2066702. Square points indicate genotyped SNPs and round points indicate imputed SNPs. All SNPs in the replication sample were genotyped. Data include the GCD sample plus SAGE and the GCD replication sample (diamonds). AA, African-American; EA, European-American; LD, linkage disequilibrium; SAGE, study of addiction: genetics and environment.

Figure 2

(a) Chromosome 2 regional Manhattan plot in European-Americans: regional Manhattan plot for the 136.70–136.90 MB region on chromosome 2 showing the adjusted symptom count meta-analysis P-values from EAs in the discovery sample. The points are colored based on the LD with rs925966. Square points indicate genotyped SNPs and round points indicate imputed SNPs. Data include the GCD sample plus SAGE. (b) Chromosome 4 PDLIM5 region in EAs: regional Manhattan plot for the 95.27–95.69 MB region containing PDLIM5 on chromosome 4 showing the adjusted symptom count meta-analysis P-values from EAs in the discovery sample. The points are colored based on the LD with rs10031423. Square points indicate genotyped SNPs and round points indicate imputed SNPs. Data include the GCD sample plus SAGE. (c) Chromosome 5 regional Manhattan plot in African-Americans: regional Manhattan plot for the 3.98–4.18 MB region on chromosome 5 showing the adjusted symptom count meta-analysis P-values from AAs in the discovery sample. The points are colored based on degree of LD with rs1493464. Square points indicate genotyped SNPs and round points indicate imputed SNPs. Data include the GCD sample plus SAGE. AA, African-American; EA, European-American; LD, linkage disequilibrium; SAGE, study of addiction: genetics and environment.