GATA4 variants were not associated with AUD in either the European ancestry or African ancestry groups after correcting for multiple comparisons. Rs10112596 demonstrated a significant relationship with an anxiety measure among the African ancestry group with AUD.
Abstract
Aims
Previous genetic association studies have shown that variation in the GATA4 gene encoding the GATA binding protein 4, a binding protein that binds to the ANA sequence GATA, increase susceptibility for alcohol use disorder (AUD). In this study, we aimed to replicate those findings in an independent sample and analyze their association with anxiety.
Methods
Overall, 1044 individuals with AUD [534 European American (EA), 510 African Americans (AA)] and 645 controls [413 EA, 232 AA] were genotyped using 34 markers. Genotype and allele frequencies were compared between cases and controls using chi-square analysis. Other phenotype data were analyzed for possible associations with GATA4 single-nucleotide polymorphisms (SNPs) in individuals with AUD.
Results
Rs6601604 was nominally significantly associated with AUD in EA, and 3 SNPs (rs6990313, rs11250159 and rs17153694) showed trend-level significance (P < 0.10) in AA. However, none of the SNPs were significant after correcting for multiple testing. Haplotype analysis of the 34 SNPs did not find a significant association between haplotype blocks and AUD diagnosis after correcting for multiple testing. From the phenotype analysis, anxiety was associated with GATA4 SNP rs10112596 among the AA group with AUD after a correction for multiple testing.
Conclusions
Although previous studies have shown a relationship between variants of the GATA4 gene and a diagnosis of AUD, we did not replicate these findings in our sample. Additional studies of variation in this gene are needed to elucidate whether polymorphisms of the GATA4 gene are associated with AUD and other alcohol-related phenotypes.
Short Summary
GATA4 variants were not associated with AUD in either the European ancestry or African ancestry groups after correcting for multiple comparisons. Rs10112596 demonstrated a significant relationship with an anxiety measure among the African ancestry group with AUD.
INTRODUCTION
The prevalence, individual health risks and societal costs of excessive alcohol use demonstrate the importance of identifying underlying causes of pathological alcohol consumption and developing novel treatment approaches (Bouchery et al., 2011; Stahre et al., 2014; Grant et al., 2015). Genetic factors account for ~40–60% of the variance in risk of developing alcohol use disorder (AUD) (Rietschel and Treutlein, 2013; Tawa et al., 2016); however, AUD is a complex disorder, with many factors contributing to its onset and maintenance. Identification of the underlying genetic risk variants will further our understanding of the disorder’s neurobiology and may direct the development of novel individualized (i.e. pharmacogenetic) treatment options for AUD.
Recent case-control genome-wide association studies (GWAS) implicate GATA binding protein 4 (GATA4), a gene located on chromosome 8, in associations with vulnerability to AUD diagnosis (Treutlein et al., 2009; Edenberg et al., 2010). While a previous candidate gene study found a significant association between GATA4 and AUD using gene-level testing (Karpyak et al., 2014), a more recent study by Degenhardt et al. (2016) failed to show an association between rare GATA4 variants and AUD. However, it should be noted that Degenhardt et al. (2016) attempted to identify only rare risk-associated variants, which precluded them from identifying likely protective variants in GATA4.
The GATA4 gene encodes transcription factor GATA4, which regulates the expression of the atrial natriuretic peptide (ANP), among other proteins (McBride and Nemer, 2001). Importantly, GATA4 protein is expressed throughout cells in the central nervous system (CNS). Reduced ANP expression in the CNS is associated with the dysregulation of stress and anxiety mechanisms in the brain, suggesting a possible link between ANP and AUD (Jorde et al., 2014). ANP also influences hypothalamic–pituitary–adrenal (HPA) axis functioning, as well as amygdala activation, further supporting the relationship between ANP and AUD-related phenotypes (McBride and Nemer, 2001). Clinically, post-detoxification patients with AUD and decreased ANP plasma levels report increased craving and anxiety levels compared to both detoxified patients with AUD and higher ANP plasma levels, as well as controls (Kiefer et al., 2002). GATA4 also acts as a transcription factor for brain natriuretic peptide (BNP), a peptide involved in the regulation of the cardiovascular system. Interestingly, BNP is also involved in stress responses when found in the CNS (Amir et al., 2010). One study demonstrated a relationship between GATA4 binding site methylation and BNP expression among alcohol-dependent patients experiencing alcohol withdrawal (Glahn et al., 2016).
Previous GWAS highlight an association between the specific intronic single-nucleotide polymorphism (SNP) rs13273672 in the GATA4 gene and AUD-related phenotypes (Kiefer et al., 2011; Jorde et al., 2014). In a randomized, double-blind, placebo-controlled study, Kiefer et al. (2011) showed that alcohol-dependent individuals with the rs13273672 G allele had a decreased time to relapse following Acamprosate treatment compared to A allele carriers with AUD. Furthermore, the G allele of this SNP was predictive of a significant decrease in variance in ANP plasma concentration compared to the A allele. Acamprosate is an FDA-approved pharmacological intervention for AUD that decreases cravings to reduce relapse risk; it is thought to primarily work through the glutamatergic system, although the exact mechanism of action remains unclear (Kiefer et al., 2011). Further research identified the AA genotype as associated with stronger alcohol cue-induced amygdala activation, and this association was predictive of a lower relapse risk (Jorde et al., 2014). Zois et al. (2016) expanded on this work by identifying an interaction between GATA4 genotype and gray matter volume on relapse risk, such that the AA genotype group showed an association between greater gray matter and a reduced relapse risk. This provides further support for the possible protective nature of the AA genotype.
Furthermore, AUD has been shown to be associated with mood and anxiety disorders, neuroticism and alcohol withdrawal (Regier et al., 1990; Malouff et al., 2007; Becker and Mulholland, 2014), all of which may play a role in genetic predisposition to AUD (Tawa et al., 2016). In particular, much research has determined the significant co-occurrence of AUD and anxiety disorders, and it is likely that genetic variation influences this comorbidity (Poikolainen, 2000; Smith and Randall, 2012). According to the common factor model proposed by Smith , a third variable (e.g. genetic susceptibility) explains the presence of both AUD and anxiety disorders. In line with this model, Merikangas et al. (1996) found that relatives of patients with anxiety disorders had an increased risk for alcohol dependence. This result could be partly explained by shared genetic factors influencing susceptibility to AUD and anxiety disorders. Moreover, a review by Kenna et al. (2012) highlights more recent research that has found an association between 5-HTTLPR, a 5-HT transporter polymorphism, and both alcohol dependence and anxiety symptoms. Given the high degree of heritability of AUD and frequent comorbid occurrence of anxiety symptoms, identifying genetic risk factors that contribute to their shared pathophysiology may improve our understanding of comorbid AUD and anxiety, as well as inform the development of pharmacological treatments. Therefore, the present study aimed to explore associations between genetic variation in GATA4 and anxiety, as assessed by the Brief Scale for Anxiety.
In summary, previous studies indicate a possible association between variation in the GATA4 gene and AUD. However, the underlying mechanisms of this relationship are still relatively unknown and poorly understood. Therefore, additional studies of SNPs within GATA4 and their association with AUD-related phenotypes are needed. In this case-control study, we sought to (a) replicate findings associating variants in GATA4 with increased susceptibility to AUD and (b) determine associations between GATA4 variants and alcohol-related clinical phenotypes, specifically anxiety, as assessed by the Brief Scale for Anxiety. Identifying genetic variants associated with AUD and related clinical phenotypes could be used to identify individuals at risk of developing AUD. Ultimately, this could inform the development of more targeted pharmacological prevention and treatment approaches for AUD.
MATERIALS AND METHODS
Participants
This study was approved by the Institutional Review Board at the National Institutes of Health (NIH). All participants provided written informed consent and permission to use collected samples. Out of 1778 individuals with collected samples, 1044 individuals with AUD [534 European Americans (EA), 510 African Americans (AA)] and 645 controls [413 EA and 232 AA] took part in this study. The 89 missing participants were excluded because they did not have a completed SCID diagnosis. Study participants were recruited between 2005 and 2016 from the inpatient unit and outpatient clinic of the Laboratory of Clinical and Translational Studies at the National Institute on Alcohol Abuse and Alcoholism (NIAAA), NIH (Bethesda, MD). Participants were recruited from three screening protocols, all of which excluded those under 18 years of age. Two of the screening protocols included only those in good health without major medical problems, and excluded individuals that were under court-mandated or involuntary treatment. The third protocol excluded prisoners, as well as pregnant women. Alcohol-dependent subjects were diagnosed with alcohol dependence according to the Diagnostic and Statistical Manuel for Mental Disorders, 4th edn, Text-revised (DSM-IV-TR) (American Psychiatric Association, 2000). Participants were diagnosed using the Structured Clinical Interview (SCID-I) for DSM-IV-TR (First, et al., 2002). Given the overlap between the DSM-IV alcohol dependence criteria and the Diagnostic and Statistical Manual of Mental Disorders, 5th edn (DSM-5) (American Psychiatric Association, 2013) AUD criteria, all participants also met criteria for AUD; however, a separate clinical interview was not conducted. Informed consent was obtained from all subjects who participated in accordance with the Declaration of Helsinki.
Genotyping and SNP selection
Large-scale genotyping was performed at the NIAAA Laboratory of Neurogenetics using the Illumina OmniExpress BeadChip (Illumina, San Diego, CA). Data for all SNPs located within the GATA4 gene that were genotyped on the array were extracted using PLINK version 1.07 (Purcell et al., 2007) (http://pngu.mgh.harvard.edu/purcell/plink/), based on start and end base pair positions for the gene located on chromosome 8 (11561716, 1161750; GRCh37/hg19 assembly). This procedure resulted in genotype data for 34 SNPs. Ancestry informative markers (AIMs; n = 2500) were also extracted from the Illumina array to calculate ancestral proportions for all study participants. Using methods described previously for an AIM panel including 186 markers (Hodgkinson et al., 2008), which were not available for the current data set, the ancestry assessment identified six ethnic factors (Africa, Europe, Asia, Far East Asia, Oceania and Americas). An analysis of the 34 SNPs among the full sample (n = 1778) found that all were in Hardy–Weinberg equilibrium (HWE) except rs12550668 (P < 0.005) and rs3729856 (P < 0.033) in the EA group. In the AA group, all SNPs were in HWE except rs6601604 (P < 0.03), rs804280 (P < 0.004) and rs867858 (P < 0.003).The same analysis found that rs10105409 in the EA group and rs13275657, rs17153747, rs3729856, rs804290 and rs11785481 in the AA group all had minor allele frequencies (MAF) < 5%. All other SNPs had a MAF > 5%.
Analysis
Allele frequencies for each SNP were determined using PLINK for the sample as a whole, and then separately for subjects of European and African ancestry (based on self-report). Due to multiple differences in allele frequency across the 34 SNPs, subsequent analyses were conducted separately in each group. Single marker association analyses were conducted using frequency comparison by chi-square test, which is the standard case-control method in PLINK, with adjustment for multiple comparisons performed using the Benjamini–Hochberg method for false discovery rate (FDR) (Benjamini and Hochberg, 1995). The threshold for FDR was set at q = 0.05. Haplotype blocks were determined using HaploView software (Barrett et al., 2005), with haplotype blocks defined using the default D’/LOD method. Haplotype association tests using these defined blocks were conducted in PLINK, and were corrected for multiple comparisons using permutation tests (5000 permutations).
Participants also completed a variety of clinical assessments, including the Alcohol Dependence Scale (ADS; Skinner and Allen, 1982), Montgomery Asberg Depression Rating Scale (MADRS; Montgomery and Asberg, 1979), Brief Scale for Anxiety (BSA; Tyrer et al., 1984), State-Trait Anxiety Inventory (STAI; Spielberger et al., 1970), Clinical Institute Withdrawal Assessment for Alcohol (CIWA; Sullivan et al., 1989) and NEO-PI-R (Costa and McCrae, 2002). Sample sizes for these assessments are inconsistent due to missing data, particularly among the control group who were not administered these assessments until later in the study’s recruitment. Single marker association and haplotype analyses were conducted for these continuous outcomes using linear regression models in PLINK. These analyses controlled for age, gender, and African and European ancestry via the AIMS scores for Africa and Europe, based on research showing age, gender and ethnicity differences in alcohol consumption and its related consequences (Delker et al. 2016).
RESULTS
Table 1 shows the demographic information of participants, as well as differences in the clinical assessments between groups. As expected, AUD participants had significantly greater scores in all alcohol-related phenotype measures when compared to controls in both the EA and AA group. To analyze the association between the 34 GATA4 SNPs and AUD diagnosis, single marker association analyses were conducted using frequency comparison by chi-square test with adjustment for multiple comparisons using the Benjamini–Hochberg method for FDR. Results revealed that one SNP (rs6601604) was nominally significantly associated with AUD in the EA group (P = 0.036). However, this SNP was not significant after adjusting for multiple comparisons (Table 2). Three SNPs (rs6990313, rs11250159 and rs17153694) trended towards a significant association with AUD in the AA group (Ps ≤ 0.065) (Table 3), but these SNPs failed to reach trend-level significance after adjusting for multiple comparisons. The a priori SNP of interest, rs13273672, was not significantly associated with AUD for either ancestry group in our sample.
Table 1.
Demographic and clinical assessment information
| European ancestry | African ancestry | |||||
|---|---|---|---|---|---|---|
| AUD (n = 534) | Controls (n = 413) | P-valuea | AUD (n = 510) | Controls (n = 232) | P-valuea | |
| Gender count (female) | 166 (31.1%) | 175 (42.4%) | 0.0003 | 140 (27.5%) | 107 (46.1%) | <0.0001 |
| Mean age (SD) | 42.5 (11.4) | 32.0 (12.1) | <0.0001 | 43.2 (10.1) | 35.6 (11.0) | <0.0001 |
| Mean ADS score (SD) | 21.8 (8.3) [n = 451] | 2.1 (4.1) [n = 95] | <0.0001 | 17.7 (8.6) [n = 345] | 1.3 (3.9) [n = 94] | <0.0001 |
| Mean MADRS score (SD) | 15.3 (9.7) [n = 444] | 1.5 (3.5) [n = 166] | <0.0001 | 11.2 (9.4) [n = 441] | 1.2 (3.1) [n = 155] | <0.0001 |
| Mean BSA score (SD) | 11.1 (7.0) [n = 446] | 1.3 (2.6) [n = 166] | <0.0001 | 8.7 (7.2) [n = 441] | 1.1 (2.6) [n = 155] | <0.0001 |
| Mean STAI score (SD) | 43.2 (13.6) [n = 199] | 33.6 (11.5) [n = 185] | <0.0001 | 41.1 (12.4) [n = 235] | 32.4 (11.9) [n = 145] | <0.0001 |
| Mean neuroticism score (SD) | 56.3 (11.5) [n = 483] | 44.8 (10.0) [n = 377] | <0.0001 | 54.6 (9.7) [n = 444] | 44.9 (8.4) [n = 202] | <0.0001 |
aChi-square test for gender; t-test for all remaining continuous variables.
Table 2.
Associations between GATA4 gene SNPs and AUD in EA samplea
| SNP | A1/A2b | MAF casesc | MAF controls | Chi square | Odds ratio | P-value | FDR |
|---|---|---|---|---|---|---|---|
| rs6990313 | A/C | 0.10 | 0.09 | 1.37 | 1.21 | 0.243 | 0.836 |
| rs10105409 | G/A | 0.01 | 0.00 | 0.05 | 1.16 | 0.818 | 0.904 |
| rs6601604 | A/G | 0.29 | 0.34 | 4.42 | 0.81 | 0.036* | 0.836 |
| rs10112596 | A/G | 0.17 | 0.19 | 1.33 | 0.87 | 0.248 | 0.836 |
| rs12550668 | A/G | 0.40 | 0.43 | 1.74 | 0.88 | 0.188 | 0.836 |
| rs2898292 | G/A | 0.10 | 0.10 | 0.00 | 1.00 | 0.995 | 0.995 |
| rs4840579 | G/A | 0.39 | 0.40 | 0.28 | 0.95 | 0.597 | 0.836 |
| rs11250159 | A/C | 0.08 | 0.09 | 0.65 | 0.87 | 0.421 | 0.836 |
| rs17153694 | A/G | 0.07 | 0.08 | 0.18 | 0.93 | 0.670 | 0.836 |
| rs17153698 | A/G | 0.16 | 0.15 | 0.24 | 1.06 | 0.625 | 0.836 |
| rs6983129 | C/A | 0.47 | 0.48 | 0.59 | 0.93 | 0.444 | 0.836 |
| rs2898295 | A/G | 0.50 | 0.47 | 1.06 | 1.10 | 0.304 | 0.836 |
| rs11250163 | C/A | 0.47 | 0.45 | 0.69 | 1.08 | 0.407 | 0.836 |
| rs13275657 | A/G | 0.19 | 0.20 | 0.64 | 0.91 | 0.423 | 0.836 |
| rs2029969 | G/A | 0.38 | 0.36 | 0.53 | 1.07 | 0.465 | 0.836 |
| rs2173117 | A/C | 0.33 | 0.30 | 1.78 | 1.14 | 0.182 | 0.836 |
| rs3779664 | A/G | 0.14 | 0.16 | 1.08 | 0.87 | 0.299 | 0.836 |
| rs3735814 | A/G | 0.48 | 0.49 | 0.18 | 0.96 | 0.671 | 0.836 |
| rs2740434 | A/G | 0.33 | 0.34 | 0.13 | 0.97 | 0.720 | 0.844 |
| rs2645399 | A/G | 0.34 | 0.36 | 0.49 | 0.93 | 0.486 | 0.836 |
| rs11784693 | A/G | 0.29 | 0.32 | 1.44 | 0.89 | 0.231 | 0.836 |
| rs804283 | G/A | 0.29 | 0.31 | 1.24 | 0.89 | 0.265 | 0.836 |
| rs17153747 | G/A | 0.13 | 0.12 | 0.75 | 1.13 | 0.386 | 0.836 |
| rs804282 | C/A | 0.45 | 0.46 | 0.40 | 0.94 | 0.529 | 0.836 |
| rs13264774 | A/G | 0.15 | 0.15 | 0.02 | 1.02 | 0.880 | 0.935 |
| rs13273672 | G/A | 0.30 | 0.29 | 0.26 | 1.05 | 0.610 | 0.836 |
| rs804280 | C/A | 0.43 | 0.44 | 0.32 | 0.95 | 0.574 | 0.836 |
| rs3729856 | G/A | 0.14 | 0.14 | 0.16 | 1.06 | 0.689 | 0.836 |
| rs867858 | C/A | 0.31 | 0.31 | 0.05 | 1.02 | 0.824 | 0.904 |
| rs1062219 | A/G | 0.45 | 0.46 | 0.20 | 0.96 | 0.655 | 0.836 |
| rs804290 | A/G | 0.24 | 0.22 | 1.31 | 1.13 | 0.253 | 0.836 |
| rs11785481 | A/G | 0.14 | 0.15 | 0.26 | 0.93 | 0.607 | 0.836 |
| rs12458 | T/A | 0.33 | 0.31 | 0.36 | 1.06 | 0.548 | 0.836 |
| rs3203358 | C/G | 0.33 | 0.32 | 0.01 | 1.01 | 0.917 | 0.944 |
aN = 534 cases, 413 controls.
bAlleles 1 and 2 refer to minor and major allele, respectively.
cMAF = minor allele frequency.
*P < 0.05.
Table 3.
Associations between GATA4 gene SNPs and AUD in AA samplea
| SNP | A1/A2b | MAF casesc | MAF controls | Chi square | Odds ratio | P-value | FDR |
|---|---|---|---|---|---|---|---|
| rs6990313 | A/C | 0.28 | 0.32 | 3.39 | 0.80 | 0.065 | 0.742 |
| rs10105409 | G/A | 0.17 | 0.19 | 0.96 | 0.87 | 0.327 | 0.930 |
| rs6601604 | A/G | 0.43 | 0.41 | 0.62 | 1.09 | 0.431 | 0.930 |
| rs10112596 | A/G | 0.16 | 0.14 | 0.86 | 1.16 | 0.353 | 0.930 |
| rs12550668 | G/A | 0.11 | 0.10 | 0.66 | 1.16 | 0.417 | 0.930 |
| rs2898292 | G/A | 0.30 | 0.27 | 2.06 | 1.20 | 0.151 | 0.921 |
| rs4840579 | A/G | 0.47 | 0.46 | 0.40 | 1.07 | 0.527 | 0.930 |
| rs11250159 | A/C | 0.21 | 0.17 | 3.41 | 1.31 | 0.065 | 0.742 |
| rs17153694 | A/G | 0.09 | 0.06 | 3.79 | 1.54 | 0.051 | 0.742 |
| rs17153698 | A/G | 0.32 | 0.31 | 0.29 | 1.07 | 0.588 | 0.930 |
| rs6983129 | C/A | 0.38 | 0.35 | 1.46 | 1.15 | 0.226 | 0.930 |
| rs2898295 | A/G | 0.39 | 0.38 | 0.12 | 1.04 | 0.728 | 0.930 |
| rs11250163 | C/A | 0.10 | 0.10 | 0.02 | 0.98 | 0.898 | 0.930 |
| rs13275657 | A/G | 0.04 | 0.03 | 0.64 | 1.27 | 0.425 | 0.930 |
| rs2029969 | G/A | 0.24 | 0.24 | 0.01 | 0.99 | 0.927 | 0.930 |
| rs2173117 | A/C | 0.24 | 0.23 | 0.34 | 1.08 | 0.559 | 0.930 |
| rs3779664 | A/G | 0.06 | 0.06 | 0.13 | 1.09 | 0.721 | 0.930 |
| rs3735814 | A/G | 0.49 | 0.49 | 0.06 | 1.03 | 0.802 | 0.930 |
| rs2740434 | A/G | 0.31 | 0.31 | 0.01 | 1.01 | 0.930 | 0.930 |
| rs2645399 | A/G | 0.48 | 0.51 | 0.69 | 0.91 | 0.408 | 0.930 |
| rs11784693 | A/G | 0.22 | 0.21 | 0.12 | 1.05 | 0.726 | 0.930 |
| rs804283 | G/A | 0.13 | 0.11 | 1.72 | 1.26 | 0.190 | 0.921 |
| rs17153747 | G/A | 0.05 | 0.04 | 0.27 | 1.15 | 0.605 | 0.930 |
| rs804282 | C/A | 0.46 | 0.48 | 0.18 | 0.95 | 0.671 | 0.930 |
| rs13264774 | A/G | 0.22 | 0.26 | 1.82 | 0.84 | 0.177 | 0.921 |
| rs13273672 | G/A | 0.37 | 0.38 | 0.20 | 0.95 | 0.657 | 0.930 |
| rs804280 | C/A | 0.42 | 0.41 | 0.20 | 1.05 | 0.655 | 0.930 |
| rs3729856 | G/A | 0.02 | 0.02 | 0.63 | 0.74 | 0.428 | 0.930 |
| rs867858 | C/A | 0.23 | 0.23 | 0.06 | 0.97 | 0.810 | 0.930 |
| rs1062219 | A/G | 0.17 | 0.17 | 0.03 | 1.03 | 0.867 | 0.930 |
| rs804290 | A/G | 0.05 | 0.03 | 2.20 | 1.54 | 0.138 | 0.921 |
| rs11785481 | A/G | 0.03 | 0.03 | 0.03 | 0.94 | 0.852 | 0.930 |
| rs12458 | T/A | 0.39 | 0.39 | 0.02 | 0.98 | 0.892 | 0.930 |
| rs3203358 | C/G | 0.06 | 0.06 | 0.19 | 1.11 | 0.667 | 0.930 |
aN = 510 cases, 232 controls.
bAlleles 1 and 2 refer to minor and major allele, respectively.
cMAF = minor allele frequency.
In addition to tests of single SNP associations, we ran haplotype analyses of the 34 SNPs. Interestingly, the haplotype structure differed in the EA and AA populations. There were nine haplotype blocks in the EA group (Supplementary Fig. S1), and seven haplotype bocks in the AA group (Supplementary Fig. S2).
Two haplotype blocks (Blocks 1 and 5) were nominally significantly related to AUD in the EA group (P = 0.037, P = 0.015, respectively). Block 1 included rs6601604 (Supplementary Table S1), which was nominally significantly associated with AUD in the single SNP association (Table 2). One block (Block 9) reached trend-level significance in this group (P = 0.087) (Supplementary Table S1). Two haplotype blocks (Block 2 and Block 3) trended towards a significant relationship with AUD in the AA group (Ps ≤ 0.068). Block 3 contained both rs11250159 and rs17153694 (Supplementary Table S2), both of which trended towards significance in the single SNP association (Table 3). There was no significant association between haplotype blocks and AUD diagnosis after correction for multiple testing.
Further analyses of continuous phenotype outcomes co-varied for age, gender, and African and European ancestry. These analyses showed no significant relationships that survived multiple comparisons between the GATA4 SNPs and scores on the ADS, MADRS, STAI, CIWA or NEO Neuroticism (data not shown). When analyzing only those with current AUD, one phenotype, anxiety, as measured by the BSA was significantly associated with GATA4 SNP rs10112596 when adjusting for multiple comparisons (P = 0.032) in the AA ancestry only (Table 4). This SNP was not in a haplotype block.
Table 4.
Associations between GATA4 gene SNPs and Brief Scale for Anxiety (BSA) scores in AA sample with AUD
| SNP | A1a | N | BETA | STAT | P-value | FDR |
|---|---|---|---|---|---|---|
| rs6990313 | A | 414 | 1.14 | 2.09 | 0.037 | 0.418 |
| rs10105409 | G | 415 | 1.55 | 2.43 | 0.015 | 0.262 |
| rs6601604 | A | 415 | 0.24 | 0.47 | 0.642 | 0.845 |
| rs10112596 | A | 415 | −2.24 | −3.33 | 0.001* | 0.032* |
| rs12550668 | G | 415 | −0.76 | −0.95 | 0.344 | 0.828 |
| rs2898292 | G | 415 | 0.94 | 1.72 | 0.086 | 0.418 |
| rs4840579 | A | 415 | 0.30 | 0.58 | 0.560 | 0.828 |
| rs11250159 | A | 415 | 1.07 | 1.73 | 0.085 | 0.418 |
| rs17153694 | A | 412 | 1.03 | 1.17 | 0.241 | 0.746 |
| rs17153698 | A | 415 | 0.83 | 1.58 | 0.115 | 0.489 |
| rs6983129 | C | 414 | 0.34 | 0.65 | 0.516 | 0.828 |
| rs2898295 | A | 415 | 0.22 | 0.43 | 0.671 | 0.845 |
| rs11250163 | C | 415 | −0.27 | −0.32 | 0.753 | 0.883 |
| rs13275657 | A | 411 | −1.33 | −0.94 | 0.346 | 0.828 |
| rs2029969 | G | 415 | −0.38 | −0.63 | 0.532 | 0.828 |
| rs2173117 | A | 415 | −0.44 | −0.74 | 0.461 | 0.828 |
| rs3779664 | A | 415 | −0.44 | −0.43 | 0.665 | 0.845 |
| rs3735814 | A | 415 | 0.92 | 1.86 | 0.064 | 0.418 |
| rs2740434 | A | 415 | 0.42 | 0.77 | 0.440 | 0.828 |
| rs2645399 | A | 414 | 0.42 | 0.83 | 0.405 | 0.828 |
| rs11784693 | A | 414 | 0.33 | 0.53 | 0.598 | 0.845 |
| rs804283 | G | 415 | 0.06 | 0.08 | 0.940 | 0.940 |
| rs17153747 | G | 415 | −0.72 | −0.61 | 0.545 | 0.828 |
| rs804282 | C | 414 | −0.04 | −0.09 | 0.927 | 0.940 |
| rs13264774 | A | 415 | 0.42 | 0.67 | 0.506 | 0.828 |
| rs13273672 | G | 415 | 0.20 | 0.36 | 0.717 | 0.870 |
| rs804280 | C | 415 | 0.46 | 0.87 | 0.384 | 0.828 |
| rs3729856 | G | 415 | 0.17 | 0.10 | 0.924 | 0.940 |
| rs867858 | C | 415 | −0.10 | −0.15 | 0.877 | 0.940 |
| rs1062219 | A | 412 | −1.15 | −1.78 | 0.076 | 0.418 |
| rs804290 | A | 415 | −1.60 | −1.40 | 0.161 | 0.548 |
| rs11785481 | A | 415 | 0.33 | 0.22 | 0.825 | 0.935 |
| rs12458 | T | 413 | −0.32 | −0.61 | 0.541 | 0.828 |
| rs3203358 | C | 415 | −1.57 | −1.43 | 0.153 | 0.548 |
aAllele 1 refers to minor allele.
bAnalysis controls for age, gender, African ancestry informative markers and European ancestry informative markers.
*P < 0.05.
DISCUSSION
Previous studies have found evidence that implicates the GATA4 gene in susceptibility to alcohol dependence (Treutlein et al., 2009; Edenberg et al., 2010; Karpyak et al., 2014). In particular, the SNP rs13273672 has been found to be related to variance in ANP expression, alcohol-induced cue reactivity and relapse risk (Kiefer et al., 2011; Jorde et al., 2014; Zois et al., 2016).
This case-control study aimed to replicate previous GWAS and candidate gene studies relating GATA4 and SNP rs13273672 with AUD and alcohol-related phenotypes. Although previous studies have shown a relationship between variants of this gene and a diagnosis of AUD, we did not replicate these findings in our sample.
There are several explanations for these discrepant results. First, as our study consisted of a relatively small sample size, we may have lacked adequate power to detect small effects, which is a limitation of the current study. This limitation might be particularly relevant given the number of SNPs that did not have a MAF > 5%. While Karpyak et al. (2014) used a sample of over 800 AD cases, we were limited to 534 and 510 AUD cases in the EA and AA subgroups, respectively. Clinical heterogeneity, such as differences in anxiety or participant status, may also account for our inability to replicate past studies. Edenberg et al. (2010) and Treutlein et al. (2009) used a sample of participants receiving treatment for their alcohol use, while our cohort included both treatment-seeking and non-treatment-seeking individuals. Although all AUD patients in the present study received a diagnosis based on the DSM-IV, it is possible that the two cohorts represent different phenotypes, which may have confounded the analysis. Furthermore, it is likely that multiple genes are involved in AUD, with only their interaction accumulating to account for a significant proportion of the variance. Therefore, additional studies of genetic variation are needed to elucidate whether polymorphisms of the GATA4 gene interact with other genes to contribute to the genetic risk for AUD and other alcohol-related phenotypes. Given that one SNP in the EA group (rs6601604) and three SNPs in the AA group (rs6990313, rs11250159 and rs17153694) did not survive correction for multiple comparisons, our data indicate a need for further replication studies with larger sample sizes. Karpyak et al. (2014) used gene-level testing to identify an association between AUD diagnoses and GATA4 variation at the gene-level. Future studies should use this additional analysis to replicate these findings and identify any gene-level association between GATA4 variants and alcohol-related clinical phenotypes. Confirming a gene-level association between GATA4 and AUD would provide a target for identifying and treating maladaptive alcohol use.
Our finding of a relationship between SNP rs10112596 and an anxiety measure in the AA group with AUD is novel, as this marker has not yet been associated with any alcohol-related phenotype. ANP levels might underlie this correlation, as decreased ANP levels have been associated with increased anxiety among individuals with AUD during detoxification (Kiefer et al., 2002). This result also supports GATA4 as a potential new target for research on comorbid AUD and anxiety, particularly among those of African ancestry. Consistent with a prior study that found a moderating effect of ethnicity on the association between alcohol abuse and an anxiety-related neuroendocrine biomarker, we found this association between GATA4 and anxiety in only the AA group with AUD (Ransome et al. 2017). Our finding suggests that the minor allele in rs10112596 might provide a protective effect in lowering anxiety susceptibility among African Americans with AUD. However, given our sample size, this association should be further examined in future studies. We did not find any association that survived correction for multiple testing between GATA4 SNPs and several other alcohol-related phenotypes. Further studies should utilize gene-level testing to determine whether there is a gene-level association between GATA4 and comorbid AUD and alcohol-related phenotypes.
Our study endeavored to extend the knowledge of the genetic components relating to AUD. We did not replicate previous studies that found an association between GATA4 variants and risk for AUD; however, further studies with larger samples and gene-level testing techniques are needed. We did find an association between rs10112596 and anxiety in the AA group, suggesting that this SNP may contribute to risk for AUD and anxiety in individuals of AA but not EA. This finding also implicates GATA4 in the relationship between AUD and anxiety, indicating a possible protective effect of the rs10112596 A minor allele. This investigation contributes meaningfully to the field because it extends the discovery of certain genotypes that may be associated with a higher risk of developing and maintaining AUD, as well as those genotypes that constitute part of a genetic ‘protective’ factor. As medicine and treatment plans are becoming more personalized and patient-specific, it becomes crucial to elucidate the mechanisms behind the genetic contribution to complex disorders. Ideally, genetic information will provide us with the tools to better diagnose and prevent psychiatric disorders, including AUD. Genetics can also provide meaningful information regarding the underlying biological basis of a disease when designing treatment strategies. An understanding of genetic susceptibility to AUD may inform the development of individualized pharmacological interventions that may provide patient-specific drug efficacy.
Supplementary Material
SUPPLEMENTARY MATERIAL
Supplementary data are available at Alcohol And Alcoholism online.
CONFLICT OF INTEREST STATEMENT
The authors declare no conflict of interest.
FUNDING
This work was supported by the National Institutes of Health (NIH) intramural funding [ZIA-AA000242; Section on Clinical Genomics and Experimental Therapeutics; to F.W.L.; Division of Intramural Clinical and Biological Research of the National Institute on Alcohol Abuse and Alcoholism (NIAAA)].
REFERENCES
- American Psychiatric Association (2000) Diagnostic and Statistical Manual of Mental Disorders, 4th edn, Text-revised. Washington, DC: American Psychiatric Association. [Google Scholar]
- American Psychiatric Assocaition (2013) Diagnostic and Statistical Manual of Mental Disorders, 5th edn Washington, DC: American Psychiatric Association. [Google Scholar]
- Amir O, Sagiv M, Eynon N, et al. (2010) The response of circulating brain natriuretic peptide to academic stress in college students. Stress 13:83–90. [DOI] [PubMed] [Google Scholar]
- Barrett JC, Fry B, Maller J, et al. (2005) Haploview: analysis and visualization of LD and haplotype maps. Bioinformatics 21:263–5. [DOI] [PubMed] [Google Scholar]
- Becker HC, Mulholland PJ (2014) Neurochemical mechanisms of alcohol withdrawal. Handb Clin Neurol 125:133–56. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Benjamini Y, Hochberg Y (1995) Controlling the false discovery rate—a practical and powerful approach to multiple testing. J R Stat Soc 57:289–300. [Google Scholar]
- Bouchery EE, Harwood HJ, Sacks JJ, et al. (2011) Economic costs of excessive alcohol consumption in the U.S., 2006. Am J Prev Med 41:516–24. [DOI] [PubMed] [Google Scholar]
- Costa PT, McCrae RR (2002) NEO Personality Inventory-Revised (NEO-PI-R). Washington, DC: American Psychological Association. [Google Scholar]
- Degenhardt F, Kramer L, Frank J, et al. (2016) Analysis of rare variants in the alcohol dependence candidate gene GATA4. Alcohol Clin Exp Res 40:1627–32. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Delker E, Brown Q, Hasin DS (2016) Alcohol consumption in demographic subpopulations: an epidemiologic overview. Alcohol Res 38:7–15. [PMC free article] [PubMed] [Google Scholar]
- Edenberg HJ, Koller DL, Xuei X, et al. (2010) Genome-wide association study of alcohol dependence implicates a region on chromosome 11. Alcohol Clin Exp Res 34:840–52. [DOI] [PMC free article] [PubMed] [Google Scholar]
- First MB, Gibbon M, Spitzer RL, et al. (2002) User’s Guide for the Structured Clinical Interview for DSM-IV-TR Axis I Disorders—Research Version. New York, NY: Biometrics Research Department, New York State Psychiatric Institute. [Google Scholar]
- Glahn A, Rhein M, Heberlein A, et al. (2016) The epigenetic regulation of GATA4-dependent brain natriuretic peptide expression during alcohol withdrawal. Neuropsychobiology 74:131–38. [DOI] [PubMed] [Google Scholar]
- Grant BF, Goldstein RB, Saha TD, et al. (2015) Epidemiology of DSM-5 alcohol use disorder results from the national epidemiologic survey on alcohol and related conditions III. JAMA Psychiatry 72:757–66. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Hodgkinson CA, Yuan Q, Xu K, et al. (2008) Addictions biology: haplotype-based analysis for 130 candidate genes on a single array. Alcohol Alcohol 43:505–15. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Jorde A, Bach P, Witt SH, et al. (2014) Genetic variation in the atrial natriuretic peptide transcription factor GATA4 modulates amygdala responsiveness in alcohol dependence. Biol Psychiatry 75:790–7. [DOI] [PubMed] [Google Scholar]
- Karpyak VM, Winham SJ, Biernacka JM, et al. (2014) Association of GATA4 sequence variation with alcohol dependence. Addict Biol 19:312–5. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Kenna GA, Roder-Hanna N, Leggio L, et al. (2012) Association of the 5-HTT gene-linked promoter region (5-HTTLPR) polymorphism with psychiatric disorders: review of psychopathology and pharmacotherapy. Pharmgenomics Pers Med 5:19–35. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Kiefer F, Andersohn F, Jahn H, et al. (2002) Involvement of plasma atrial natriuretic peptide in protracted alcohol withdrawal. Acta Psychiatr Scand 105:65–70. [DOI] [PubMed] [Google Scholar]
- Kiefer F, Witt SH, Frank J, et al. (2011) Involvement of the atrial natriuretic peptide transcription factor GATA4 in alcohol dependence, relapse risk and treatment response to acamprosate. Pharmacogenomics J 11:368–74. [DOI] [PubMed] [Google Scholar]
- Malouff JM, Thorsteinsson EB, Rooke SE, et al. (2007) Alcohol involvement and the five-factor model of personality: a meta-analysis. J Drug Educ 37:277–94. [DOI] [PubMed] [Google Scholar]
- McBride K, Nemer M (2001) Regulation of the ANF and BNP promoters by GATA factors: lessons learned for cardiac transcription. Can J Physiol Pharmacol 79:673–81. [PubMed] [Google Scholar]
- Merikangas KR, Stevens D, Fenton B (1996) Comorbidity of alcohol and anxiety disorders: the role of family studies. Alcohol Health Res World 20:100–06. [PMC free article] [PubMed] [Google Scholar]
- Montgomery SA, Asberg M (1979) A new depression scale designed to be sensitive to change. Br J Psychiatry 134:382–9. [DOI] [PubMed] [Google Scholar]
- Poikolainen K. (2000) Risk factors for alcohol dependence: a case-control study. Alcohol Alcohol 35:190–6. [DOI] [PubMed] [Google Scholar]
- Purcell S, Neale B, Todd-Brown K, et al. (2007) PLINK: a tool set for whole-genome association and population-based linkage analyses. Am J Hum Genet 81:559–75. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Ransome Y, Slopen N, Karlsson O, et al. (2017) The association between alcohol abuse and neuroendocrine system dysregulation: race differences in a National sample. Brain Behav Immun 66:313–21. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Regier DA, Farmer ME, Rae DS, et al. (1990) Comorbidity of mental disorders with alcohol and other drug abuse. Results from the Epidemiologic Catchment Area (ECA) Study. J Am Med Assoc 264:2511–8. [PubMed] [Google Scholar]
- Rietschel M, Treutlein J (2013) The genetics of alcohol dependence. Ann NY Acad Sci 1282:39–70. [DOI] [PubMed] [Google Scholar]
- Skinner HA, Allen BA (1982) Alcohol dependence syndrome: measurement and validation. J Abnorm Psychol 91:199–209. [DOI] [PubMed] [Google Scholar]
- Smith JP, Randall CL (2012) Anxiety and alcohol use disorders: comorbidity and treatment considerations. Alcohol Res 34:414–31. [PMC free article] [PubMed] [Google Scholar]
- Spielberger CD, Gorsuch RL, Lushene RE (1970) Manual for the State-Trait Anxiety Inventory. Palo Alto, CA: Consulting Psychologists Press. [Google Scholar]
- Stahre M, Roeber J, Kanny D, et al. (2014) Contribution of excessive alcohol consumption to deaths and years of potential life lost in the United States. Prev Chronic Dis 11:E109. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Sullivan JT, Sykora K, Schneiderman J, et al. (1989) Assessment of alcohol withdrawal: the revised clinical institute withdrawal assessment for alcohol scale (CIWA-Ar). Br J Addict 84:1353–7. [DOI] [PubMed] [Google Scholar]
- Tawa EA, Hall SD, Lohoff FW (2016) Overview of the genetics of alcohol use disorder. Alcohol Alcohol 51:507–14. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Treutlein J, Cichon S, Ridinger M, et al. (2009) Genome-wide association study of alcohol dependence. Arch Gen Psychiatry 66:773–84. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Tyrer P, Owen RT, Cicchetti DV (1984) The brief scale for anxiety: a subdivision of the comprehensive psychopathological rating scale. J Neurol Neurosurg Psychiatry 47:970–5. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Zois E, Vollstadt-Klein S, Hoffmann S, et al. (2016) GATA4 variant interaction with brain limbic structure and relapse risk: a voxel-based morphometry study. Eur Neuropsychopharmacol 26:1431–7. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.