Skip to main content
PMC home page
Alcohol and Alcoholism (Oxford, Oxfordshire) logoLink to Alcohol and Alcoholism (Oxford, Oxfordshire)
. 2012 Sep 27;48(1):9–14. doi: 10.1093/alcalc/ags104

Rare ADH Variant Constellations are Specific for Alcohol Dependence

Lingjun Zuo 1, Heping Zhang 2, Robert T Malison 1, Chiang-Shan R Li 1, Xiang-Yang Zhang 3, Fei Wang 1, Lingeng Lu 2, Lin Lu 4, Xiaoping Wang 5, John H Krystal 1,6,7, Fengyu Zhang 8, Hong-Wen Deng 9, Xingguang Luo 1,*
PMCID: PMC3523382  PMID: 23019235

Abstract

Aims: Some of the well-known functional alcohol dehydrogenase (ADH) gene variants (e.g. ADH1B*2, ADH1B*3 and ADH1C*2) that significantly affect the risk of alcohol dependence are rare variants in most populations. In the present study, we comprehensively examined the associations between rare ADH variants [minor allele frequency (MAF) <0.05] and alcohol dependence, with several other neuropsychiatric and neurological disorders as reference. Methods: A total of 49,358 subjects in 22 independent cohorts with 11 different neuropsychiatric and neurological disorders were analyzed, including 3 cohorts with alcohol dependence. The entire ADH gene cluster (ADH7–ADH1C–ADH1B–ADH1A–ADH6–ADH4–ADH5 at Chr4) was imputed in all samples using the same reference panels that included whole-genome sequencing data. We stringently cleaned the phenotype and genotype data to obtain a total of 870 single nucleotide polymorphisms with 0< MAF <0.05 for association analysis. Results: We found that a rare variant constellation across the entire ADH gene cluster was significantly associated with alcohol dependence in European-Americans (Fp1: simulated global P = 0.045), European-Australians (Fp5: global P = 0.027; collapsing: P = 0.038) and African-Americans (Fp5: global P = 0.050; collapsing: P = 0.038), but not with any other neuropsychiatric disease. Association signals in this region came principally from ADH6, ADH7, ADH1B and ADH1C. In particular, a rare ADH6 variant constellation showed a replicable association with alcohol dependence across these three independent cohorts. No individual rare variants were statistically significantly associated with any disease examined after group- and region-wide correction for multiple comparisons. Conclusion: We conclude that rare ADH variants are specific for alcohol dependence. The ADH gene cluster may harbor a causal variant(s) for alcohol dependence.

INTRODUCTION

Alcohol dehydrogenases (ADHs) are largely distributed in the liver (e.g. ββADH encoded by ADH1B, γγADH encoded by ADH1C and ADH6 enzyme encoded by ADH6) and the upper digestive tract (e.g. σσADH encoded by ADH7) and partly in the central nervous system (e.g. σσADH and ADH6) (Shmueli et al., 2003; Yanai et al., 2005). They possess high activity in converting ethanol to toxic acetaldehyde (Yasunami et al., 1991). Alterations of this activity may influence human drinking behavior and thus the risk of alcohol dependence. Additionally, these enzymes are also efficient in the oxidization of retinol, a vitamin A precursor (summarized in Satre et al., 1994; Luo et al., 2008). For example, σσADH, also called retinol dehydrogenase, is the most efficient enzyme among ADHs in catalyzing retinol formation (Satre et al., 1994); the ADH6 enzyme is efficient in the oxidization of retinol as well (Km:15–40 µM) (Satre et al., 1994). Specifically, they convert retinol to retinal, which in turn is synthesized to retinoic acid (RA), the active form of vitamin A. RA is a pleiotrophic regulator of gene expression in vertebrates and plays a role in regulating embryonic development (including development of the brain). Dopamine neurons contain all necessary enzymatic components for these regulations. Proper development and maintenance of a functional dopaminergic system may depend strongly upon the supply of RA. Functional alterations of these enzymes can thus influence the development and maintenance of physiological dopaminergic system functioning (Luo et al., 2008). In addition to ethanol and retinol, ADH enzymes are also implicated in the metabolism of various dopamine-related neurotransmitters. These support the hypothesis that, in addition to alcohol dependence, there could be associations between ADH gene variants and more neuropsychiatric and neurological disorders, given that the dopaminergic system is well known to play an important role in the etiology of those disorders. Furthermore, alcohol dependence has high rates of co-morbidity with numerous psychiatric disorders including anxiety disorders, major depression, bipolar disorders, schizophrenia, post-traumatic stress disorder, etc. (Regier et al., 1990; Kessler et al., 1996; Grant et al., 2004), which also supports the hypothesis that alcohol dependence and other neuropsychiatric disorders could have common susceptibility genes including ADH genes. So far, numerous studies have reported associations between ADH variants and alcohol dependence; ADH variants have also been associated with Parkinson's disease (ADH1C and ADH7) (Buervenich et al., 2000, 2005), cerebral infarction and lacunae (ADH1B) (Suzuki et al., 2004).

It is well known that at least four functional ADH gene variants including rs1229984 (ADH1B*2; Arg48His), rs2066702 (ADH1B*3; Arg370Cys), rs1693482 (ADH1C*2; Arg272Gln) and rs698 (ADH1C*2; Ile350Va) that significantly affect the risk of alcohol dependence are rare variants in most populations, e.g. in Asians [minor allele frequency (MAF) frs2066702 = 0.000; frs1693482 = 0.023; frs698 = 0.025], Europeans (frs2066702 = 0.000; frs1229984 = 0.008) and/or Africans (frs1229984 = 0.000; frs1693482 = 0.052; frs698 = 0.042) (Luo et al., 2006). A recent genome-wide association study identified a common variant (rs1789891; f = 0.192) that was significantly associated with alcohol dependence in people of German descent [P = 1.3 × 10−8; odds ratio (OR) = 1.46] (Frank et al., 2012). Notably, this significant risk variant is located between the four functional ADH rare variants. These suggest to us that rare ADH variants may play important roles in human diseases.

The role of rare genetic variants in human diseases has not been well studied until recently. An important hypothesis in medical genetics research is that many genetically influenced human diseases may not result from a single common variant, but rather, from a constellation of more rare, regionally concentrated, disease-causing variants. The signals of association credited to common genetic variants may be synthetic associations resulting from the contributions of multiple rare variants within a given gene region (Dickson et al., 2010). With the emergence of sequencing technology, it is now feasible to test this hypothesis by thoroughly investigating the rare variants across the genome (e.g. capitalizing on the vast array of rare variant data deposited in databases such as the 1000 Genome Project).

In this study, we aimed to comprehensively examine the associations between rare ADH variants (MAF < 0.05) and 11 different neuropsychiatric and neurological disorders in subjects of European or African descent, which included three independent cohorts with alcohol dependence in European-Americans, European-Australians and African-Americans. In these three cohorts, no significant common ADH variants for risk of alcohol dependence have been found before (Bierut et al., 2010; Edenberg et al., 2010; Heath et al., 2011). This study would help us to know whether the rare ADH variants are specific for alcohol dependence or shared by susceptibility to other disorders.

MATERIALS AND METHODS

Subjects

A total of 49,358 subjects in 22 independent cohorts with 11 different neuropsychiatric and neurological disorders were analyzed (Tables 1 and 2). These 22 cohorts included case–control and family-based samples, genotyped on different microarray platforms. These 11 disorders included alcohol dependence, major depression, bipolar disorder, schizophrenia, autism, attention deficit hyperactivity disorder (ADHD), Alzheimer's disease, amyotrophic lateral sclerosis (ALS), early onset stroke, ischemic stroke and Parkinson's disease. These data were all of those with neuropsychiatric and neurological disorders available for our analysis from the database of Genotypes and Phenotypes (dbGaP). Detailed demographics data are shown in Table 1.

Table 1.

Demographic data of all cohorts

Pedigrees Subjects Affected subjects
 Unaffected subjects
Total Total Male
 Female
 Male
 Female
Human disease Ethnicity Data set name n n n Age (yrs) n Age (yrs) n Age (yrs) n Age (yrs)
Alcohol dependence EA (CC) SAGE+COGA 2927 2927 883 39.0 ± 10.4 526 36.7 ± 8.8 445 37.9 ± 10.1 1073 39.0 ± 9.1
Alcohol dependence AA (CC) SAGE+COGA 1189 1189 428 41.0 ± 8.3 253 39.8 ± 6.8 169 40.2 ± 8.4 339 39.6 ± 6.8
Alcohol dependence EAu (Fam) OZ-ALC 1856 6410 1011 42.0 ± 8.4 622 39.2 ± 7.3 1709 46.3 ± 9.8 2213 45.6 ± 9.5
Major depression CA (CC) PRSC 3625 3625 548 44.2 ± 12.0 1257 41.2 ± 12.8 694 47.1 ± 14.4 1126 43.8 ± 13.7
Bipolar disorder EA (CC) BDO+GRU 1402 1402 190 43.1 ± 8.0 178 45.4 ± 10.0 532 54.7 ± 17.3 502 50.1 ± 17.6
Bipolar disorder EA (CC) BARD+GRU 1687 1687 322 42.1 ± 8.3 331 44.4 ± 9.7 532 54.7 ± 17.3 502 50.1 ± 17.6
Bipolar disorder AA (CC) BARD+GRU 812 812 39 42.4 ± 7.9 102 42.0 ± 7.8 272 46.0 ± 14.0 399 45.7 ± 13.5
Schizophrenia AA (CC) GAIN 2149 2149 746 41.9 ± 10.8 449 43.0 ± 9.8 362 46.2 ± 13.7 592 45.0 ± 12.9
Schizophrenia EA (CC) GAIN 2729 2729 947 42.5 ± 11.3 404 45.1 ± 11.2 634 53.5 ± 17.0 744 49.2 ± 16.7
Schizophrenia EA (CC) nonGAIN 2784 2784 996 42.3 ± 11.8 441 44.2 ± 12.4 669 51.8 ± 15.3 678 47.9 ± 16.1
Schizophrenia AA (CC) nonGAIN 118 118 60 41.5 ± 11.3 38 42.9 ± 10.5 20 49.7 ± 9.2 0 —
Autism EA (Fam) AGP 1366 4075 1121 7.2 ± 3.2 209 7.1 ± 3.0 0 — 0 —
ADHD CA (Fam) IMAGE 922 2757 802 10.9 ± 2.8 122 10.8 ± 3.0 0 — 0 —
Alzheimer's disease CA (Fam) LOAD × 4 2243 5219 788 84.1 ± 8.1 1510 86.3 ± 8.7 486 66.7 ± 10.7 773 66.1 ± 10.5
Alzheimer's disease EA (CC) GenADA 1588 1588 340 77.6 ± 8.3 466 78.3 ± 8.8 279 74.4 ± 7.7 503 72.8 ± 8.1
ALS CA (CC) GRU 507 507 138 56.5 ± 11.9 123 59.2 ± 11.6 136 69.6 ± 8.6 110 69.8 ± 8.9
Early onset stroke EA (CC) GEOS × 3 802 802 198 42.7 ± 6.1 174 38.7 ± 7.6 208 40.6 ± 6.4 222 37.8 ± 7.3
Early onset stroke AA (CC) GEOS × 3 599 599 144 42.3 ± 6.2 165 41.0 ± 7.3 129 40.9 ± 6.5 161 38.4 ± 7.5
Ischemic stroke CA (CC) ISGS 485 485 119 71.8 ± 8.3 100 71.6 ± 8.1 128 69.7 ± 8.6 138 69.7 ± 8.9
Parkinson's disease CA (CC) NGRC 3986 3986 1346 67.2 ± 10.5 654 67.3 ± 11.0 769 70.7 ± 13.9 1217 70.1 ± 14.2
Parkinson's disease CA (CC) PDRD+GRU 1767 1767 537 70.5 ± 9.3 363 70.2 ± 10.0 346 53.5 ± 15.7 521 55.2 ± 11.2
Parkinson's disease CA (CC) lng_coriell_pd 1741 1741 560 66.3 ± 10.9 380 65.9 ± 11.2 336 62.3 ± 14.4 465 56.0 ± 17.2
Open in a new tab

In the family data, only the affected and unaffected offspring are listed. Data set names refer to dbGaP.

n, sample size; yrs, years; CC, case–control sample; Fam, family sample. EA, European-American; AA, African-American; EAu, European-Australian; CA, Caucasian; ADHD, attention deficit hyperactivity disorder; ALS, amyotrophic lateral sclerosis. [GenADA: Li et al. Arch Neurol. 2008; 65(1):45–53; Filippini et al. Neuroimage. 2009;44(3):724–728. AGP: The AGP Consortium. Nature. 2010;466(7204):368–372; Human Molecular Genetics. 2010;19(20):4072–4082; Nature Genetics. 2007;39(3):319–328].

Table 2.

Associations between ADH gene cluster and different neuropsychiatric or neurological disorders

Most sig. Affected
 Unaffected
 Minimal SNP # SNP #
Human disease Ethnicity Data set name dbGaP# SNP Gene n MAF n MAF P-value (total) (P < 0.05)
Alcohol dependence EA (CC) SAGE+COGA phs000092.v1.p1 rs1596180 ADH7 1409 0.018 1518 0.007 0.0009 343 9
Alcohol dependence AA (CC) SAGE+COGA phs000092.v1.p1 rs114618736 ADH1C 681 0.018 508 0.006 0.0108 486 6
Alcohol dependence EAu (Fam) OZ-ALC phs000181.v1.p1 rs11733695 ADH6 1633 0.042 1633 0.105 0.0353 385 2
Major depression CA (CC) PRSC phs000020.v2.p1 rs7690269 ADH7 1805 0.019 1820 0.009 0.0004 341 16
Bipolar disorder EA (CC) BDO+GRU phs000017.v3.p1 rs6532797 ADH4 368 0.043 1034 0.021 0.0072 215 10
Bipolar disorder EA (CC) BARD+GRU phs000017.v3.p1 rs1391088 ADH1C 653 0.060 1034 0.038 0.0224 250 7
Bipolar disorder AA (CC) BARD+GRU phs000017.v3.p1 rs283417 ADH1C 141 0.029 671 0.008 0.0094 193 2
Schizophrenia AA (CC) GAIN phs000021.v3.p2 rs4699743 ADH1C 1195 0.013 954 0.030 0.0086 276 11
Schizophrenia EA (CC) GAIN phs000021.v3.p2 rs60652198 ADH4 1351 0.041 1378 0.027 0.0165 277 4
Schizophrenia EA (CC) nonGAIN phs000167.v1.p1 rs71612689 ADH7 1437 0.004 1347 0.000 0.0220 354 5
Schizophrenia AA (CC) nonGAIN phs000167.v1.p1 rs76919634 ADH6 98 0.008 20 0.045 0.2905 35 0
Autism EA (Fam) AGP phs000267.v1.p1 rs62325239 ADH5 1330 0.003 1330 0.002 0.0141 361 44
ADHD CA (Fam) IMAGE phs000016.v2.p2 rs1442483 ADH7 924 0.017 924 0.059 0.0039 356 9
Alzheimer's Disease CA (Fam) LOAD × 4 phs000168.v1.p1 rs116192122 ADH4 2298 0.012 2298 0.002 0.0030 356 10
Alzheimer's disease EA (CC) GenADA phs000219.v1.p1 rs35361391 ADH4 806 0.034 782 0.009 0.0162 267 18
ALS CA (CC) GRU phs000101.v3.p1 rs115081066 ADH4 261 0.006 246 0.033 0.0017 334 7
Early Onset Stroke EA (CC) GEOS × 3 phs000292.v1.p1 rs1596180 ADH1B 372 0.019 430 0.003 0.0048 301 8
Early onset stroke AA (CC) GEOS × 3 phs000292.v1.p1 rs114188790 ADH1C 309 0.081 290 0.036 0.0026 451 48
Ischemic stroke CA (CC) ISGS phs000102.v1.p1 rs72681936 ADH7 219 0.078 266 0.034 0.0202 348 6
Parkinson's disease CA (CC) NGRC phs000196.v2.p1 rs78304974 ADH1B 2000 0.007 1986 0.002 0.0089 341 7
Parkinson's disease CA (CC) PDRD+GRU phs000126.v1.p1 rs1693457 ADH7 900 0.029 867 0.048 0.0129 360 25
Parkinson's disease CA (CC) lng_coriell_pd phs000089.v3.p2 rs28472487 ADH7 940 0.026 801 0.043 0.0260 354 11
Open in a new tab

Only the most significant risk markers are listed; in family-based cohorts, N = sample size of affected offspring; “affected MAF” = “transmitted MAF”, “unaffected MAF” = “untransmitted MAF” in offspring. Data set names refer to dbGaP. COGA data set access number is phs000125.v1.p1. MAF, minor allele frequency; n, CC, Fam, EA, AA, EAu, CA, ADHD and ALS: also see Table 1. The corrected α was 2.1 × 10−5 (European) and 1.5 × 10−5 (African), respectively.

These subjects contained three cohorts with alcohol dependence, including 1409 European-American cases, 1518 European-American controls, 6410 European-Australian family subjects with 1633 alcohol-dependent probands, 681 African-American cases and 508 African-American controls. All subjects in these three cohorts were interviewed using the Semi-Structured Assessment for the Genetics of Alcoholism (Bucholz et al., 1994). Affected subjects met DSM-IV criteria for alcohol dependence (American Psychiatric Association, 1994). Additionally, 65.9% of patients with major depression had alcohol-drinking behavior (data not shown), i.e. at least 12 alcoholic drinks in the past 12 months. The samples with alcohol dependence and major depression were identical to those used in the published work (Boomsma et al., 2008; Zuo et al., 2011a,b).

Imputation

To make the genetic marker sets consistent across the different cohorts, we imputed the missing single nucleotide polymorphisms (SNPs) across the entire ADH gene cluster (ADH7–ADH1C–ADH1B–ADH1A–ADH6–ADH4–ADH5 at Chr4: 100,204,900–100,631,900) in all samples using the same reference panels that included whole-genome sequencing data. To maximize the success rate and accuracy of imputation, we (a) used both 1000 Genome Project and HapMap 3 panels as the reference, and separated the European (CEU) and African (YRI) ethnicities during imputation; (b) used a Markov Chain Monte Carlo algorithm implemented in the program IMPUTE2 (Howie et al., 2009) to derive full posterior probabilities, not the ‘best-guess’, of the genotypes of each SNP; (c) set the imputation parameters at burnin = 10,000, iteration = 10,000, k = 100, Ne = 11,500 and confidence level = 0.99 (Howie et al., 2009); (d) merged, within the same ethnicity, the data sets as much as possible to increase sample sizes and marker density for imputation, being subject to the following criteria: cases and controls that were paired within the same study; different panels of array data in the same subjects; and separate samples that had the same phenotype and were genotyped on the same microarray platform and (e) stringently cleaned the imputed data before association analysis (see below). Additionally, because the imputation process did not incorporate the family relationship information, Mendelian errors might occur in the imputed data. Thus, the families with at least one individual who had >0.5% Mendel errors (considering all SNPs tested) and the SNPs with >0.5% Mendel errors (considering all individuals tested) were excluded too. Finally, for SNPs that were directly genotyped, we used the direct genotypes rather than the imputed data.

Data cleaning

We stringently cleaned the phenotype data and the genotype data before association analysis (detailed previously; Zuo et al., 2011a). Subjects with poor genotypic data and questionable diagnostic information, allele discordance, duplicated IDs, potential sample misidentification, sample relatedness, sample misspecification, gender anomalies, missing race, non-European and non-African ethnicity, population group outliers, a mismatch between self-identified and genetically inferred ethnicity, a missing genotype call rate ≥2% across all SNPs and subjects overlapped between two data sets [e.g. the Study of Addiction: Genetics and Environment (SAGE) data set and the Collaborative Study on the Genetics of Alcoholism (COGA) data set] were excluded (one copy). Furthermore, we excluded monomorphic SNPs and SNPs with allele discordance, Mendelian errors (in family samples) and an overall missing genotype call rate ≥2%. For those data sets merged from the separate samples (e.g. SAGE and COGA) that had the same phenotype and were genotyped on the same microarray platform, SNPs with allele frequency differences >2% between the original separate samples were excluded. For all merged data sets, SNPs with missing rate differences >2% between the original separate samples were also excluded. The SNPs with MAF = 0 in either cases or controls were excluded, because it could not be determined if they were missed during the imputation process or truly non-polymorphic in nature in some phenotype groups. Finally, only a total of 870 SNPs with 0<MAF<0.05 in either cases or controls were extracted for association analysis (Supplementary data, Table A1). The cleaned sample sizes, cleaned SNP numbers, ethnicity, diagnosis, dbGaP access numbers and data set name abbreviations of these samples are shown in Tables 1 and 2.

Association tests for region-wide rare variant constellations

Synthetic effects of region-wide rare variant constellations may be more significant than individual rare variants in some specific gene regions on disease phenotypes. These effects were tested using a score-type program, SCORE-Seq (Lin and Tang, 2011). The mutation information was aggregated by virtue of a weighted linear combination across all rare variants of the entire ADH gene cluster or across each ADH gene region, and then related to disease phenotypes using appropriate regression models. Sex, age, alcohol drinking and the first 10 principal components served as the covariates in the regression models. Principal component scores for each individual were estimated using the program EIGENSTRAT (Price et al., 2006). The first 10 principal components explained >95% of variance in our samples. Two fixed MAF threshold with flexible weight tests (Fp1: MAF <0.01; Fp5: MAF <0.05) and one variable threshold with fixed weight test (VT test: MAF <0.05) were performed to derive the global P-values from these regression models (Table 3). In Fp tests, the weight was 1/sqrt(p(1-p)) where P was the estimated MAF with pseudo counts in the pooled sample. In VT test, the weight was 1 when MAF <threshold and 0 otherwise, where the threshold varied between 0 and 0.05. Statistical significance was assessed by resampling 1 million times (Lin and Tang, 2011). Additionally, we used ARIEL (Asimit et al., 2012), a regression-based collapsing approach that incorporates variant quality scores, to confirm the tests by SCORE-Seq. All association analyses were performed within the same ethnicity.

Table 3.

P-values for associations between rare variant constellations and diseases

Diseases Ethnicity Data set name Tests MAF upper bound ADH cluster ADH1A ADH1B ADH1C ADH4 ADH5 ADH6 ADH7
Alcohol dependence EA SAGE+COGA Fp 0.01 0.045 0.231 0.979 0.591 0.085 0.725 0.008 0.076
Alcohol dependence EA SAGE+COGA Fp 0.05 0.707 0.929 0.965 0.210 0.893 0.965 0.655 0.650
Alcohol dependence EA SAGE+COGA VT Variable 0.089 0.239 0.887 0.877 0.329 0.986 0.010 0.138
Alcohol dependence EA SAGE+COGA Collapsing 0.05 0.542 0.827 0.943 0.197 0.693 0.847 0.479 0.568
Alcohol dependence EAu OZ-ALC Fp 0.01 0.206 0.092 0.509 0.056 0.645 0.765 0.397 0.195
Alcohol dependence EAu OZ-ALC Fp 0.05 0.027 0.635 0.025 0.034 0.315 0.480 0.449 0.009
Alcohol dependence EAu OZ-ALC VT Variable 0.055 0.396 0.103 0.023 0.344 0.797 0.030 0.047
Alcohol dependence EAu OZ-ALC Collapsing 0.05 0.038 0.851 0.016 0.038 0.335 0.482 0.472 0.005
Alcohol dependence AA SAGE+COGA Fp 0.01 0.543 0.941 0.795 0.452 0.492 0.447 0.496 0.889
Alcohol dependence AA SAGE+COGA Fp 0.05 0.050 0.784 0.723 0.151 0.077 0.121 0.051 0.491
Alcohol dependence AA SAGE+COGA VT Variable 0.226 0.822 0.854 0.563 0.273 0.173 0.099 0.981
Alcohol dependence AA SAGE+COGA Collapsing 0.05 0.038 0.555 0.861 0.139 0.051 0.069 0.056 0.581
Major depression CA PRSC Fp 0.01 0.307 0.643 0.282 0.339 0.765 0.646 0.040 0.697
Major depression CA PRSC Fp 0.05 0.107 0.349 0.319 0.856 0.123 0.071 0.294 0.675
Major depression CA PRSC VT Variable 0.557 0.198 0.678 0.496 0.607 0.314 0.302 0.133
Major depression CA PRSC Collapsing 0.05 0.108 0.336 0.359 0.770 0.106 0.072 0.361 0.621
Open in a new tab

MAF, minor allele frequency; Fp1, Fp5 and VT, association tests using SCORE-Seq; Collapsing, association test using ARIEL; EA, EAu, AA, CA and data set names refer to Table 1. Significant P-values are bold.

Association tests for individual rare variants

For case–control samples, the allele frequencies of each SNP were compared between cases and controls using logistic regression analysis as implemented in PLINK (Purcell et al., 2007). Diagnosis served as the dependent variable, alleles served as the independent variables and sex, age, alcohol drinking and the first 10 principal components served as the covariates. For family samples, we tested the allele-disease associations using the program Family-Based Association Test (Horvath et al., 2001). The MAFs and P-values of the most significant risk SNPs and the numbers of the nominally significant risk SNPs (P < 0.05) in all samples are shown in Table 2.

Correction for multiple testing in single-point association tests

The experiment-wide significance levels (α) were corrected for the numbers of cohorts (i.e. 22) and the numbers of effective markers that were calculated by the program SNPSpD (Li and Ji, 2005), which is an adjusted Bonferroni correction taking the linkage disequilibrium structure into account. Approximately, 110 and 150 effective SNPs captured most of the information of all rare variants across the entire ADH gene cluster in cohorts of European and African descent, respectively. Thus, the corrected significance levels (α) for single-point association tests were set at 2.1 × 10−5 in cohorts of European descent and 1.5 × 10−5 in cohorts of African descent, respectively.

RESULTS

The rare variant constellation across the entire ADH gene cluster was specifically associated with alcohol dependence in European-Americans [Fp1: global P = 0.045; 108 variants (SNPs) with 2067 minor alleles], European-Australians (Fp5: global P = 0.027; Collapsing P = 0.038; 388 variants with 92,429 minor alleles) and African-Americans (Fp5: global P = 0.050; Collapsing P = 0.038; 486 variants with 20,513 minor alleles), but not with any other neuropsychiatric disease (P > 0.10). In testing the rare variant constellations within each individual gene region, several results were obtained. First, the ADH6 variant constellation was significantly associated with alcohol dependence in European-Americans (Fp1: P = 0.008; VT: P = 0.010; 10 variants with 155 minor alleles), European-Australians (VT: P = 0.030; 49 variants with 10,546 minor alleles) and African-Americans (Fp5: P = 0.051; Collapsing P = 0.056; 85 variants with 4529 minor alleles). Second, the ADH7 variant constellation was significantly associated with alcohol dependence in European-Australians (Fp5: P = 0.009; VT: P = 0.047; Collapsing P = 0.005; 98 variants with 20,280 minor alleles), and suggestively in European-Americans (Fp1: P = 0.076; 22 variants with 348 minor alleles). Third, the ADH1B and ADH1C variant constellations were modestly associated with alcohol dependence in European-Australians (for ADH1B: Fp5: P = 0.025 and collapsing: P = 0.016; for ADH1C: Fp1: P = 0.056, Fp5: P = 0.034, VT: P = 0.023 and collapsing: P = 0.038), but not in European-Americans and African-Americans (P > 0.10; Table 3). Additionally, single-point association analysis showed that, of a total of 343 individual rare variants in European-Americans, 9 SNPs were nominally associated with alcohol dependence (P < 0.05), the most significant of which (rs1596180, at 5′ of ADH7) was suggestively associated with alcohol dependence (P = 0.0009; Table 2).

The rare variant constellation across the ADH6 gene region was also modestly associated with major depression in Caucasians (Fp1: P = 0.040; 10 variants with 307 minor alleles). This association turned out to be non-significant after correction for multiple testing. Furthermore, among a total of 341 individual rare variants in Caucasians, 16 SNPs were nominally associated with major depression (P < 0.05), the most significant of which (rs7690269, at 5′ of ADH7) was suggestively associated with major depression (OR = 2.16; P = 0.0004). This rs7690269 was also the most significant one among all 22 cohorts. Finally, no individual variants were statistically significantly associated with any disease examined after group- and region-wide correction (P >α), including alcohol dependence and major depression (Table 2).

DISCUSSION

We found that rare ADH variant constellations were specific for alcohol dependence. In particular, a rare ADH6 variant constellation showed replicable association with alcohol dependence across three independent cohorts of European or African descent. Additionally, ADH7, ADH1B and ADH1C variant constellations might also be implicated in the risk for alcohol dependence. We speculate that the ADH gene cluster may harbor a causal variant(s) for alcohol dependence.

Searching the entire ADH cluster, we found no individual rare variants which were statistically significantly associated with any disease examined (including alcohol dependence) after group- and region-wide correction for multiple comparisons. Our study provides an additional example to support the hypothesis that the synthetic effects of region-wide rare variant constellations may be more significant than individual rare variants on disease phenotypes. Using multiple cohorts with large sample sizes, we found that rare ADH variant constellations were specific for alcohol dependence, but not associated with any other disease, which was consistent with previous reports (Luo et al., 2006) and with the fact that the ADH enzymes are mainly distributed in the liver, but only partly distributed in the central nervous system. Although the synthetic effects of rare ADH variants on alcohol dependence seemed to be modest in the present study, these effects appeared to be highly significant when compared with those on other ‘non-alcohol dependence’ neuropsychiatric disorders.

When testing each gene region, we detected modest associations between rare ADH1B and ADH1C variant constellations and alcohol dependence in European-Australians. The variants in these two genes may influence the risk of alcohol dependence via ethanol metabolism pathways, which is well-known by numerous studies. However, these associations were not strong and not replicated in other populations in the present study. They remained to be confirmed in the future.

More robust associations were detected between ADH6 variants and alcohol dependence, which was replicable in three cohorts. Alteration of ADH6 enzyme activity caused by ADH6 variants may influence the ethanol metabolism as introduced above, and thus may influence the human drinking behavior and the risk for alcohol dependence. Alternatively, the retinol metabolism pathway or other non-ethanol metabolism pathways introduced above may be other possible mechanisms underlying the associations between ADH6 variant constellation and alcohol dependence, and possibly the suggestive association between ADH6 variant constellation and major depression as well. Similarly, these mechanisms might also underlie the suggestive associations between the rare ADH7 variant constellation and alcohol dependence and between individual ADH7 variants and major depression.

SUPPLEMENTARY DATA

Supplementary data are available at Alcohol and Alcoholism online.

Funding

This work was supported in part by National Institute on Drug Abuse (NIDA) grants K01DA029643, K24DA017899 and R01DA016750, National Institute on Alcohol Abuse and Alcoholism (NIAAA) grants R01 AA016015 and R21 AA020319 and the National Alliance for Research on Schizophrenia and Depression (NARSAD) Award 17616 (L.Z.). We thank NIH GWAS Data Repository, the Contributing Investigator(s) who contributed the phenotype and genotype data from his/her original study (e.g. Drs Bierut, Edenberg, Heath, Singleton, Hardy, Foroud, Myers, Gejman, Faraone, Sonuga-Barke, Sullivan, Nurnberger, Devlin, Monaco, etc.), and the primary funding organization that supported the contributing study. Funding and other supports for phenotype and genotype data were provided through the National Institutes of Health (NIH) Genes, Environment and Health Initiative (GEI) (U01HG004422, U01HG004436 and U01HG004438); the GENEVA Coordinating Center (U01HG004446); the NIAAA (U10AA008401, R01AA013320, P60AA011998); the NIDA (R01DA013423); the National Cancer Institute (P01 CA089392); the Division of Neuroscience, the NIA National Institute of Neurological Disorders and Stroke (NINDS); the NINDS Human Genetics Resource Center DNA and Cell Line Repository; the NIH contract ‘High throughput genotyping for studying the genetic contributions to human disease’ (HHSN268200782096C); the Center for Inherited Disease Research (CIDR); a Cooperative Agreement with the Division of Adult and Community Health, Centers for Disease Control and Prevention; the NIH Office of Research on Women's Health (ORWH) (R01NS45012); the Department of Veterans Affairs; the University of Maryland General Clinical Research Center (M01RR165001), the National Center for Research Resources, NIH; the National Institute of Mental Health (R01MH059160, R01MH59565, R01MH59566, R01MH59571, R01MH59586, R01MH59587, R01MH59588, R01MH60870, R01MH60879, R01MH61675, R01MH62873, R01MH081803, R01MH67257, R01MH81800, U01MH46276, U01MH46282, U01MH46289, U01MH46318, U01MH79469, U01MH79470 and R01MH67257); the NIMH Genetics Initiative for Bipolar Disorder; the Genetic Association Information Network (GAIN); the Genetic Consortium for Late Onset Alzheimer's Disease; the Autism Genome Project, the MARC: Risk Mechanisms in Alcoholism and Co-morbidity; the Molecular Genetics of Schizophrenia Collaboration; the Medical Research Council (G0601030) and the Wellcome Trust (075491/Z/04), University of Oxford; the Netherlands Scientific Organization (904-61-090, 904-61-193, 480-04-004, 400-05-717, NWO Genomics, SPI 56-464-1419) the Centre for Neurogenomics and Cognitive Research (CNCR-VU); Netherlands Study of Depression and Anxiety (NESDA) and the Netherlands Twin Register (NTR); and the European Union (EU/WLRT-2001-01254), ZonMW (geestkracht program, 10-000-1002). Assistance with phenotype harmonization and genotype cleaning, as well as with general study coordination, was provided by the Genetic Consortium for Late Onset Alzheimer's Disease, the GENEVA Coordinating Center (U01 HG004446), and the National Center for Biotechnology Information. Genotyping was performed at the Johns Hopkins University Center for Inherited Disease Research, and GlaxoSmithKline, R&D Limited. The data sets used for the analyses described in this manuscript were obtained from dbGaP at http://www.ncbi.nlm.nih.gov/sites/entrez?Db=gap through dbGaP accession numbers listed in Table 2.

Conflict of interest statement. J.H.K. has been a paid consultant for Aisling Capital, LLC, AstraZeneca Pharmaceuticals, Brintnall & Nicolini, Inc., Easton Associates, Gilead Sciences, Inc., GlaxoSmithKline, Janssen Pharmaceuticals, Lundbeck Research USA, Medivation, Inc., Merz Pharmaceuticals, MK Medical Communications, F. Hoffmann-La Roche Ltd, SK Holdings Co., Ltd, Sunovion Pharmaceuticals, Inc., Takeda Industries and Teva Pharmaceutical Industries, Ltd. He serves as a member of Scientific Advisory Boards for Abbott Laboratories, Bristol-Myers Squibb, Eisai, Inc., Eli Lilly and Co., Forest Laboratories, Inc., Lohocla Research Corporation, Mnemosyne Pharmaceuticals, Inc., Naurex, Inc., Pfizer Pharmaceuticals and Shire Pharmaceuticals. He is the Editor for Biological Psychiatry, a member of Board of Directors of Coalition for Translational Research in Alcohol and Substance Use Disorders, and the President Elect for American College of Neuropsychopharmacology. He also gets support from Tetragenex Pharmaceuticals. Additionally, H.Z. has been a paid consultant for Eisai, Inc. Other authors have no conflict of interest.

See Letters to the Editor (p.129) for a response to this article.

Supplementary Material

Supplementary Data
supp_48_1_9__index.html (862B, html)

References

  1. American Psychiatric Association. Diagnostic and Statistical Manual of Mental Disorders. Washington, DC: American Psychiatric Press; 1994. [Google Scholar]
  2. Asimit JL, Day-Williams AG, Morris AP, et al. ARIEL and AMELIA: testing for an accumulation of rare variants using next-generation sequencing data. Hum Hered. 2012;73:84–94. doi: 10.1159/000336982. doi:10.1159/000336982. [DOI] [PMC free article] [PubMed] [Google Scholar]
  3. Bierut LJ, Agrawal A, Bucholz KK, et al. A genome-wide association study of alcohol dependence. Proc Natl Acad Sci USA. 2010;107:5082–7. doi: 10.1073/pnas.0911109107. doi:10.1073/pnas.0911109107. [DOI] [PMC free article] [PubMed] [Google Scholar]
  4. Boomsma DI, Willemsen G, Sullivan PF, et al. Genome-wide association of major depression: description of samples for the GAIN Major Depressive Disorder Study: NTR and NESDA biobank projects. Eur J Hum Genet. 2008;16:335–42. doi: 10.1038/sj.ejhg.5201979. doi:10.1038/sj.ejhg.5201979. [DOI] [PubMed] [Google Scholar]
  5. Bucholz KK, Cadoret R, Cloninger CR, et al. A new, semi-structured psychiatric interview for use in genetic linkage studies: a report on the reliability of the SSAGA. J Stud Alcohol. 1994;55:149–58. doi: 10.15288/jsa.1994.55.149. [DOI] [PubMed] [Google Scholar]
  6. Buervenich S, Sydow O, Carmine A, et al. Alcohol dehydrogenase alleles in Parkinson's disease. Mov Disord. 2000;15:813–8. doi: 10.1002/1531-8257(200009)15:5<813::aid-mds1008>3.0.co;2-y. doi:10.1002/1531-8257(200009)15:5<813::AID-MDS1008>3.0.CO;2-Y. [DOI] [PubMed] [Google Scholar]
  7. Buervenich S, Carmine A, Galter D, et al. A rare truncating mutation in ADH1C (G78Stop) shows significant association with Parkinson disease in a large international sample. Arch Neurol. 2005;62:74–8. doi: 10.1001/archneur.62.1.74. doi:10.1001/archneur.62.1.74. [DOI] [PubMed] [Google Scholar]
  8. Dickson SP, Wang K, Krantz I, et al. Rare variants create synthetic genome-wide associations. PLoS Biol. 2010;8:e1000294. doi: 10.1371/journal.pbio.1000294. doi:10.1371/journal.pbio.1000294. [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. Edenberg HJ, Koller DL, Xuei X, et al. Genome-wide association study of alcohol dependence implicates a region on chromosome 11. Alcohol Clin Exp Res. 2010;34:840–52. doi: 10.1111/j.1530-0277.2010.01156.x. doi:10.1111/j.1530-0277.2010.01156.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
  10. Frank J, Cichon S, Treutlein J, et al. Genome-wide significant association between alcohol dependence and a variant in the ADH gene cluster. Addict Biol. 2012;17:171–80. doi: 10.1111/j.1369-1600.2011.00395.x. doi:10.1111/j.1369-1600.2011.00395.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
  11. Grant BF, Stinson FS, Dawson DA, et al. Prevalence and co-occurrence of substance use disorders and independent mood and anxiety disorders: results from the National Epidemiologic Survey on Alcohol and Related Conditions. Arch Gen Psychiatry. 2004;61:807–16. doi: 10.1001/archpsyc.61.8.807. doi:10.1001/archpsyc.61.8.807. [DOI] [PubMed] [Google Scholar]
  12. Heath AC, Whitfield JB, Martin NG, et al. A quantitative-trait genome-wide association study of alcoholism risk in the community: findings and implications. Biol Psychiatry. 2011;70:513–8. doi: 10.1016/j.biopsych.2011.02.028. doi:10.1016/j.biopsych.2011.02.028. [DOI] [PMC free article] [PubMed] [Google Scholar]
  13. Horvath S, Xu X, Laird NM. The family based association test method: strategies for studying general genotype–phenotype associations. Eur J Hum Genet. 2001;9:301–6. doi: 10.1038/sj.ejhg.5200625. doi:10.1038/sj.ejhg.5200625. [DOI] [PubMed] [Google Scholar]
  14. Howie BN, Donnelly P, Marchini J. A flexible and accurate genotype imputation method for the next generation of genome-wide association studies. PLoS Genet. 2009;5:e1000529. doi: 10.1371/journal.pgen.1000529. doi:10.1371/journal.pgen.1000529. [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Kessler RC, Nelson CB, McGonagle KA, et al. The epidemiology of co-occurring addictive and mental disorders: implications for prevention and service utilization. Am J Orthopsychiatry. 1996;66:17–31. doi: 10.1037/h0080151. doi:10.1037/h0080151. [DOI] [PubMed] [Google Scholar]
  16. Li J, Ji L. Adjusting multiple testing in multilocus analyses using the eigenvalues of a correlation matrix. Heredity (Edinb) 2005;95:221–7. doi: 10.1038/sj.hdy.6800717. doi:10.1038/sj.hdy.6800717. [DOI] [PubMed] [Google Scholar]
  17. Lin DY, Tang ZZ. A general framework for detecting disease associations with rare variants in sequencing studies. Am J Hum Genet. 2011;89:354–67. doi: 10.1016/j.ajhg.2011.07.015. doi:10.1016/j.ajhg.2011.07.015. [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. Luo X, Kranzler HR, Zuo L, et al. Diplotype trend regression analysis of the ADH gene cluster and the ALDH2 gene: multiple significant associations with alcohol dependence. Am J Hum Genet. 2006;78:973–87. doi: 10.1086/504113. doi:10.1086/504113. [DOI] [PMC free article] [PubMed] [Google Scholar]
  19. Luo X, Kranzler HR, Zuo L, et al. ADH7 variation modulates extraversion and conscientiousness in substance-dependent subjects. Am J Med Genet B Neuropsychiatr Genet. 2008;147B:179–86. doi: 10.1002/ajmg.b.30589. doi:10.1002/ajmg.b.30589. [DOI] [PMC free article] [PubMed] [Google Scholar]
  20. Price AL, Patterson NJ, Plenge RM, et al. Principal components analysis corrects for stratification in genome-wide association studies. Nat Genet. 2006;38:904–9. doi: 10.1038/ng1847. doi:10.1038/ng1847. [DOI] [PubMed] [Google Scholar]
  21. Purcell S, Neale B, Todd-Brown K, et al. PLINK: a tool set for whole-genome association and population-based linkage analyses. Am J Hum Genet. 2007;81:559–75. doi: 10.1086/519795. doi:10.1086/519795. [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. Regier DA, Farmer ME, Rae DS, et al. Comorbidity of mental disorders with alcohol and other drug abuse. Results from the Epidemiologic Catchment Area (ECA) Study. J Am Med Assoc. 1990;264:2511–18. doi:10.1001/jama.1990.03450190043026. [PubMed] [Google Scholar]
  23. Satre MA, Zgombic-Knight M, Duester G. The complete structure of human class IV alcohol dehydrogenase (retinol dehydrogenase) determined from the ADH7 gene. J Biol Chem. 1994;269:15606–12. [PubMed] [Google Scholar]
  24. Shmueli O, Horn-Saban S, Chalifa-Caspi V, et al. GeneNote: whole genome expression profiles in normal human tissues. C R Biol. 2003;326:1067–72. doi: 10.1016/j.crvi.2003.09.012. doi:10.1016/j.crvi.2003.09.012. [DOI] [PubMed] [Google Scholar]
  25. Suzuki Y, Fujisawa M, Ando F, et al. Alcohol dehydrogenase 2 variant is associated with cerebral infarction and lacunae. Neurology. 2004;63:1711–13. doi: 10.1212/01.wnl.0000142971.08275.db. doi:10.1212/01.WNL.0000142971.08275.DB. [DOI] [PubMed] [Google Scholar]
  26. Yanai I, Benjamin H, Shmoish M, et al. Genome-wide midrange transcription profiles reveal expression level relationships in human tissue specification. Bioinformatics. 2005;21:650–9. doi: 10.1093/bioinformatics/bti042. doi:10.1093/bioinformatics/bti042. [DOI] [PubMed] [Google Scholar]
  27. Yasunami M, Chen CS, Yoshida A. A human alcohol dehydrogenase gene (ADH6) encoding an additional class of isozyme. Proc Natl Acad Sci USA. 1991;88:7610–4. doi: 10.1073/pnas.88.17.7610. doi:10.1073/pnas.88.17.7610. [DOI] [PMC free article] [PubMed] [Google Scholar]
  28. Zuo L, Gelernter J, Zhang CK, et al. Genome-wide association study of alcohol dependence implicates KIAA0040 on chromosome 1q. Neuropsychopharmacology. 2011a;37:557–66. doi: 10.1038/npp.2011.229. doi:10.1038/npp.2011.229. [DOI] [PMC free article] [PubMed] [Google Scholar]
  29. Zuo L, Zhang CK, Wang F, et al. A novel, functional and replicable risk gene region for alcohol dependence identified by genome-wide association study. PLoS ONE. 2011b;6:e26726. doi: 10.1371/journal.pone.0026726. doi:10.1371/journal.pone.0026726. [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supplementary Data
supp_48_1_9__index.html (862B, html)
supp_ags104_ags104supp.doc (36KB, doc)

Articles from Alcohol and Alcoholism (Oxford, Oxfordshire) are provided here courtesy of Oxford University Press

RESOURCES