Genome-wide DNA methylation patterns in discordant sib pairs with alcohol dependence

Authors

  • Rongrong Zhao MD,

    1. Mental Health Institute, The Second Xiangya Hospital, Central South University, Changsha, Hunan, China
    Search for more papers by this author
  • Ruiling Zhang MD,

    1. Department of Psychiatry, The Second Affiliated Hospital of Xinxiang Medical University, Xinxiang, Henan, China
    2. Henan Key Lab of Biological Psychiatry, Xinxiang Medical University, Xinxiang, Henan, China
    Search for more papers by this author
  • Wenqiang Li MS,

    1. Department of Psychiatry, The Second Affiliated Hospital of Xinxiang Medical University, Xinxiang, Henan, China
    2. Henan Key Lab of Biological Psychiatry, Xinxiang Medical University, Xinxiang, Henan, China
    Search for more papers by this author
  • Yanhui Liao MD,

    1. Mental Health Institute, The Second Xiangya Hospital, Central South University, Changsha, Hunan, China
    Search for more papers by this author
  • Jinsong Tang MD,

    1. Mental Health Institute, The Second Xiangya Hospital, Central South University, Changsha, Hunan, China
    Search for more papers by this author
  • Qin Miao MD,

    1. Department of Psychiatry, The Second Affiliated Hospital of Xinxiang Medical University, Xinxiang, Henan, China
    2. Henan Key Lab of Biological Psychiatry, Xinxiang Medical University, Xinxiang, Henan, China
    Search for more papers by this author
  • Wei Hao MD

    Corresponding author
    • Mental Health Institute, The Second Xiangya Hospital, Central South University, Changsha, Hunan, China
    Search for more papers by this author

Correspondence

Wei Hao MD, Mental Health Institute, The Second Xiangya Hospital, Central South University, 139 Renmin (M) Road, Changsha, Hunan 410011, China.

Tel: +86,731 85292156

Fax: +86,731 85360160

Email: weihaochangsha@gmail.com

Abstract

Introduction

Alcohol dependence is a complex disease caused by a confluence of environmental and genetic factors. Epigenetic mechanisms have been shown to play an important role in the pathogenesis of alcohol dependence.

Methods

To determine if alterations in gene-specific methylation were associated with alcohol dependence, a genome-wide DNA methylation analysis was performed on peripheral blood mononuclear cells from alcohol-dependent patients and siblings without alcohol dependence as controls. The Illumina Infinium Human Methylation450 BeadChip was used and gene-specific methylation of DNA isolated from peripheral blood mononuclear cells was assessed. Genes ALDH1L2, GAD1, DBH and GABRP were selected to validate beadchip results by pyrosequencing.

Results

Compared to normal controls, 865 hypomethylated and 716 hypermethylated CG sites in peripheral blood mononuclear cell DNA in alcohol-dependent patients were identified. The most hypomethylated CG site is located in the promoter of SSTR4 (somatostatin receptor 4) and the most hypermethylated CG site is GABRP (gamma-aminobutyric acid A receptor). The results from beadchip analysis were consistent with that of pyrosequencing.

Discussion

DNA methylation might be associated with alcohol dependence. Genes SSTR4, ALDH1L2, GAD1, DBH and GABRP may participate in the biological process of alcohol dependence.

Introduction

Alcohol dependence (AD) is a common disorder that causes physical, psychological, and social problems. Over 300 million people are dependent on alcohol (Rehm et al., 2004). Alcohol use disorders have been the second cause of the global burden of this disease (Collins et al., 2011). The pathogenesis of AD is multifactorial, including genetic factors, environmental factors and social culture. Twin, family, and adoption studies have consistently shown that genetic factors play an important role in the development of alcohol dependence (Kimura and Higuchi, 2011). Twin studies have estimated the heritability of alcohol dependence to be about 50–65% in both sexes (Kendler et al., 1992, 1997; Heath et al., 1997). Environmental influences predominated in remission from alcohol use disorder and which were from individual experiences (McCutcheon et al., 2012). In addition, epigenetic mechanisms may be an important factor in AD. In a related report, the epigenetic mechanisms of AD were demonstrated (Bönsch et al., 2005; Bleich et al., 2006).

Epigenetic mechanisms that regulate gene activity without altering the DNA code have been shown to produce long-lasting changes in gene expression essential to development and adaptation to environmental changes (Ducci and Goldman, 2008). DNA methylation is one of the epigenetic mechanisms. DNA methylation affects gene transcription by binding a methyl group to a CpG island in the genomic sequence. Most of CpG sequences are methylated, but CpG islands in the promoter regions of genes are generally less methylated. In most cases, higher methylation of the genomic sequence leads to an inactivation of the involved gene, while less methylation leads to activation of the involved gene (Doerfler, 1983; Egger et al., 2004).

There were some previous researches on genome DNA methylation pertaining to alcoholism. Manzardo et al. used the Roche NimbleGen Human DNA Methylation 2.1M Deluxe Promoter Array to interrogate methylation loci in the brain of alcoholics relative to matched controls in order to identify global methylation disturbances and affected candidate genes for alcoholism (Manzardo et al., 2012). Thapar et al. used the Illumina GoldenGate Methylation Cancer Panel (Illumina, San Diego, CA, USA), which probes the methylation profile at 1505 CpG sites from 807 cancer-related genes, to assess whether DNA methylation patterns in chronic alcoholics are different from non-alcoholic sibling controls (Thapar et al., 2012). Thapar et al. (2012) did not reveal any significant differences in the average methylation score between alcoholic and non-alcoholic siblings associated with 743 genes implicated in carcinogenesis. This result contradicts our study, so epigenetics in the context of alcoholism is likely, but is not clearly demonstrated to date. Therefore, further investigations should try to solve these problems.

In addition, some studies have shown changes in DNA methylation in the promoter regions of specific genes in AD. Muschler et al. (2010) showed that the DNA methylation status in the gene sequence of polypeptide pro-opiomelanocortin (POMC) differ between patients with alcohol dependence and healthy controls and identified a specific cluster of CpG islands showing a significant association with alcohol craving. A higher level of DNA methylation in the alpha synuclein (SNCA) gene promoter and decreased expression of SNCA gene mRNA have been found in patients with AD compared with normal controls (Bönsch et al., 2004, 2005). SNCA is known to be involved in dopaminergic neurotransmission and it catalyzes the intake of dopamine (Perez et al., 2002), which has been indicated as a main mechanism associated with AD. DAT (dopamine transporter) is responsible for the reuptake of dopamine from the synaptic gap and plays a crucial role in dopaminergic neurotransmission. The DAT gene promoter was shown to have significant hypermethylation in AD patients compared with normal controls (Hillemacher et al., 2009). Other genes also showed methylation related to AD, such as the homocysteine-induced endoplasmatic reticulum protein (HERP) gene (Bleich et al., 2006) and N-methyl-D-aspartate2b receptor subtype (NR2B) gene (Marutha Ravindran and Ticku, 2005; Biermann et al., 2009).

At present, DNA methylation related to AD was almost wholly confined to a single gene, which did not provide a comprehensive analysis. Epigenetics in the context of alcoholism is likely but is not clearly demonstrated to date. Therefore, in this study genome-wide DNA methylation analysis was performed to assess the relationship between DNA methylation in peripheral blood mononuclear cells and AD in discordant sib pairs.

Methods

This study was approved by the Ethics Committee of the Second Affiliated Hospital of Xinxiang Medical University, and written informed consent was obtained from all participants. Twenty participants were Han Chinese living in the north Henan province and their biological grandparents were of Han Chinese ancestry. Individuals with a history of severe medical complications, organic brain disease, concomitant major psychiatric disorders, substance dependence apart from alcohol, neoplastic disorders and severe immune disease were excluded. The consensus diagnoses for AD were made by at least two psychiatrists according to the Diagnostic and Statistical Manual of Mental Disorders-Fourth Edition IV (DSM-IV). Alcohol craving was assessed at admission using the Alcohol Use Disorders Identification Test (AUDIT, the value is no less than 8, which indicated AD tendency).

All blood samples were stored at −80°C immediately after collection. Total leukocyte DNA was extracted from frozen EDTA-blood using a QIAmp DNA Blood Mini Kit (Qiagen, Hilden, Germany). Bisulfite conversion of DNA was performed using the Zymo EZ DNA Methylation Kit (Zymo Research, Orange, CA, USA). This technique involves treating methylated DNA with bisulfite, which converts unmethylated cytosines into uracil. Methylated cytosines remain unchanged during the treatment. Once converted, the methylation profile of the DNA can be determined by PCR amplification followed by DNA sequencing. Occasionally, it should be noted that high input levels of DNA may result in incomplete bisulfite conversion for some GC-rich regions.

Bisulfite-converted patient and sib pair DNA samples were prepared and quantified using a NanoDrop 2000 scanning spectrophotometer (Thermo, Wilmington, DE, USA). The NanoDrop 2000/2000c pedestal option can measure higher protein concentrations than traditional cuvette-based spectrophotometers. The sample retention system employs surface tension to hold the sample in place between two optical fibers. This enables the measurement of very highly concentrated samples without the need for dilutions. Using this technology, the full spectrum (190–840 nm) NanoDrop 2000/2000c Spectrophotometers have the capability to measure sample concentration up to 200 times more concentrated than samples measured using the standard cuvette.

For each sample, 500ng of bisulite-converted DNA was used for hybridization on the Infinium HumanMethylation450 BeadChip for whole genome analysis, following the Illumina Infinium HD Methylation protocol. Infinium HumanMethylation450 BeadChip includes 485,764 cytosine positions of the human genome covering more than 14,000 genes. The incorporated CpG sites are distributed among all 22 autosomal and one sex chromosome pairs of humans. The beadchip was scanned with an Illumina iScan. Raw data were exported for further analysis using Illumina BeadStudio software and the Methylation Module add-in (Illumina, San Diego, CA, USA). The relative level of methylation for each CG site was evaluated as the ratio of methylated-probe signal to the total locus signal intensity and was defined within a 0 to 1 range average beta value (AVB), exported from the Illumina BeadStudio software package. The methylation status of specific cytosines is indicated by an AVB where 1 corresponds to complete methylation and 0 to no methylation. Signals of probes with P ≥ 0.05 were excluded from the analysis. Loci were scored as hypermethylated if the AVB was greater than or equal to 0.8. As changes in AVB greater than 0.2 can be detected with 95% statistical confidence, this value was the threshold to identify significant methylation changes in our analyses. The expression difference score, DiffScore, takes into account background noise and sample variability (Chudin et al., 2006). Differential CG sites were selected according to the absolute value of the DiffScore (no less than 20).

The Infinium Methylation system includes a variety of both sample-independent and sample-dependent controls, which were evaluated in each chip. Sample-independent controls included staining controls consisting of high- and background-intensity dinitrophenyl (DNP) and biotin; hybridization controls consisting of low, medium and high levels; target removal controls; and extension controls of both DNP (A/T) and biotin (C/G) fluorescence channels. Sample-dependent controls included both high- and background-level bisulfite conversion controls for each of the four basic oligomer types, specificity controls for mismatches and perfectly matched sequences in DNP and biotin channels, 600 negative controls and non-polymorphic controls per chip.

CG sites were chosen from the Illumina assay repertoire for pyrosequencing to validate beadchip results. Analysis of the methylation status in this manner exploits the quantitative nature of pyrosequencing by reporting the ratio of cytosine to thymine at each analyzed CpG site, which reflects the proportion of methylated DNA. These sites were located within promoter regions of hypomethylated ALDH1L2, GAD1, DBH and hypermethylated GABRP. Because genes ALDH1L2, GAD1, DBH and GABRP validation methods were built in our laboratory and they were identified as related to AD by DAVID (Database for Annotation, Visualization and Integrated Discovery 6.7), they were chosen to validate.

Primers were designed using Pyrosequencing Assay Design Software (Biotage AB, Uppsala, Sweden). Sequences of the primers are listed in Table 1. The PCR was carried out with 10 ng of bisulfite treated DNA using TaqGold DNA polymerase (Applied Biosystems) under the following conditions: 10 min at 95°C, followed by 50 cycles of 35 sec at 95°C, 35 sec at 57.5°C, and 1 min at 72°C. Pyrosequencing reactions were performed with a Biotage PyroMark MD System (Biotage) according to the manufacturer's protocols by the sequential addition of single nucleotides in a predefined order. Raw data were analyzed using Pyro Q-CpG 1.0.9 analysis software (Biotage). The CpG methylation level (ranging from 0 to 1) was represented by the percentage of methylated C among the sum of methylated and unmethylated C.

Table 1. Primers for pyrosequencing methylation assay
GenesPrimerNotesSequences (from 5′ to 3′)
GABRPGABRP1fPCR-forwardTTGGAGGTAGTAGTTATAGTAGGAGTT
GABRP1rPCR-reverse, biotin-labeledCCCCATCCTTAATAAAAACACTAACTTATC
GABRP1sSequencing-forwardAGTAGTTATAGTAGGAGTTG
GAD1GAD11fPCR-forwardTTGAGGAGAAATTGTTTTGGGTTAAAATGT
GAD11rPCR-reverse, biotin-labeledACCTAATAAAAAACCTCCTAACCTATA
GAD11ssequencing-forwardTGTTTTGGGTTAAAATGTT
ALDH1L2ALDH1L21fPCR-forwardGGAAGGTGAAGTAGAAAAGTAAAG
ALDH1L21rPCR-reverse, biotin-labeledAAATATAATTTAACCTCAAAAATTTTTACA
ALDH1L21ssequencing-forwardTTAAGTTTTTTTTAAAGTTAGATTG
DBHDBH1fPCR-forwardTTATTGGGTTGTGGTTAGGAGGTTA
DBH1rPCR-reverse, biotin-labeledACTCCCAAATACTTCTAAAAAAAATCTTC
DBH1sSequencing-forwardGGGTTTATGTAGAGTTAGT

Linear regression analysis was used to compare Illumina DNA methylation microarrays with pyrosequencing.

To identify the function of genes, including differential CG sites, the Database for Annotation, Visualization and Integrated Discovery 6.7 (DAVID; http://david.abcc.ncifcrf.gov) was used, which provides a set of data mining tools that systematically combine functionally descriptive data with intuitive graphical displays. DAVID provides exploratory visualization tools that promote discovery through functional classification, biochemical pathway maps, and conserved protein domain architectures, while simultaneously remaining linked to rich sources of biological annotation (Dennis et al., 2003). Analysis of biological pathways, including Gene Ontology (GO; http://www.geneontology.org/) and the Kyoto Encyclopedia of Genes and Genomes (KEGG; http://www.genome.jp/kegg/), were used to search functionally related genes to AD through DAVID web-accessible programs. Gene Ontology is one of four web-based analysis modules of DAVID, which is a major bioinformatics initiative with the aim of standardizing the representation of gene and gene product attributes across species and databases. The project provides a controlled vocabulary of terms for describing gene product characteristics and gene product annotation data from GO Consortium members, as well as tools to access and process this data. While KEGG is another one of four web-based analysis modules of DAVID, which is a database resource for understanding high-level functions and utilities of the biological system, such as the cell, the organism and the ecosystem, from molecular-level information, especially large-scale molecular datasets generated by genome sequencing and other high-throughput experimental technologies.

Results

Using a matching design, there was no difference in age and sex between the patients with AD and controls. The mean age for patients was 44.90 ± 5.90 years and the mean age for controls was 44.20 ± 7.19 years, showing no difference (P = 0.701; Table 2).

Table 2. Demographic characteristic of alcohol dependence patients and their normal siblings
Case IDCase ageAuditControl IDControl ageAudit
 143291457
 249202536
 345303535
 442154377
 542125403
 654286475
 734307305
 853248466
 942349495
10453410424

There were, according to the absolute value of the DiffScore, 865 hypomethylated and 716 hypermethylated CG sites in the peripheral blood mononuclear cell DNA in AD patients. These sites were involved in 827 known genes. 98 CG sites were located in the 5′-untranslated region and 62 CG sites were located in the 3′-untranslated region. The top five hypomethylated and hypermethylated CG sites are listed in Table 3, the top 30 hypomethylated and hypermethylated CG sites are listed in supporting Table S1 and the heatmap of these differential CG sites between AD patients and their siblings are displayed in Figure 1. The most hypomethylated CG site is located in the promoter of SSTR4 and the most hypermethylated CG site is located in GABRP.

Figure 1.

Heatmap of differentially methylated genes between alcohol dependence (AD) patients and their siblings. Methylation values were standardized across rows and clustered with a Euclidean distance metric and average linkage. The red/green gradient represents a standardized level of hypermethylation/hypomethylation in AD patients compared to controls. The order of clustered genes is represented in supporting Table S1. Case, AD patients; controls, their siblings. Cg, the location of a CpG island.

Table 3. Five top-ranked hypomethylated genes and five top-ranked hypermethylated genes
SymbolGeneTarget IDDiffScoreDelta BetaChrRelation
SSTR4somatostatin receptor 4cg01471923−334.109−0.2370920N_Shore
TAGLN3transgelin3cg08522473−334.109−0.31173unknown
CCDC85Ccoiled-coil domain containing 85Ccg09214175−334.109−0.2429714unknown
RFPL3ret finger protein-like 3cg16142906−334.109−0.1874922unknown
RPH3ALrabphilin 3A-like (without C2 domains)cg27575622−334.109−0.2152817N_Shelf
GABRPgamma-aminobutyric acid A receptorcg27129755336.67210.18026625unknown
SPDEFSAM pointed domain containing ets, transcription factorcg01395541336.67210.1935356unknown
FLJ31306uncharacterized LOC379025cg01801090336.67210.15985814N_Shelf
PLK5Ppolo-like kinase 5cg02872767336.67210.165919Island
CDKN2Dcyclin-dependent kinase inhibitor 2Dcg03007623336.67210.14309319Island

Through analysis of biological pathways, the differential methylation genes are part of 13 biological processes, 19 molecular functions and 20 cellular components in GO analysis (Figure 2). Biological regulation accounted for the largest proportion of biological processes. Thirty-three genes of 130 differential methylation genes used in KEGG analysis belonged to metabolic pathways (Table 4). Two hypomethylated genes and two hypermethylated genes were identified as related to AD by DAVID (Table 5) and they were analyzed by pyrosequencing to validate the data from Illumina DNA methylation microarrays in all samples. Linear regression analysis revealed the microarray methylation and pyrosequencing were well correlated (Table 6).

Figure 2.

Through GO analysis, 827 known genes, including 1581 differential CG sites, are part of 13 biological processes, 19 molecular functions and 20 cellular components.

Table 4. Differentially methylated genes KEGG pathway analysis
KEGG pathwayNo. genes
Type I diabetes mellitus11
Cell adhesion molecules (CAMs)16
Allograft rejection10
Graft-versus-host disease10
Autoimmune thyroid disease10
Viral myocarditis11
Intestinal immune network for IgA production8
Antigen processing and presentation10
Tight junction11
Metabolic pathways33
Table 5. Differentially methylated genes (beadchip)
GeneDiffScoreDelta betaCase AVG betaControl AVG beta
  1. DiffScore: expression difference score. Delta beta: the difference between case AVG beta and control AVG beta. AVG Beta were evaluated as the ratio of methylated-probe signal to total locus signal intensity.
GABRP336.67210.18026620.81853320.638267
ALDH1L2−176.0189−0.1195070.8130590.9325656
GAD1−43.36211−0.11855040.50820740.6267577
DBH−60.23603−0.13719820.42022680.557425
Table 6. Correlation (R2) between Illumina DNA methylation microarray and pyrosequencing (validation)
GeneMicroarrayPyrosequencingR2P-value
Mean AVG betaSDMean AVG betaSD
GABRP0.65390.39990.60460.41040.8700.000
ALDH1L20.87490.14270.85200.13560.9770.000
GAD10.56630.19890.56750.19800.9930.000
DBH0.48770.32800.48750.32500.9980.000

Discussion

Before this study, there was a shortage of reports on genome-wide DNA methylation studies in discordant sib pairs with AD. To our knowledge, this is the first report of a genome-wide methylation analysis in AD. Therefore, the results of the study are potentially of great interest for the field, especially as knowledge of epigenetic alterations in AD is currently very limited. In this study, a genome-wide DNA methylation study was performed to provide evidence for an association between gene-specific DNA methylation and AD. The current study could provide new insight into the mechanisms underlying AD. A discordant sib pair study was performed to provide hereditary background similarity, and environmental effect on sib pairs had small difference.

According to the results, there are 865 upregulated and 716 downregulated CG sites in peripheral blood mononuclear cell DNA in AD patients. This result revealed DNA methylation differences between AD patients and their siblings at the level of the whole genome. The most hypomethylated CG site is located in the promoter of SSTR4 and the most hypermethylated CG site is located in GABRP.

Somatostatin (SST) is a regulatory peptide that activates G protein-coupled receptors comprised of five members (somatostatin receptors [SSTRs] 1–5) and triggers multiple transmembrane signaling pathways (Reisine and Bell, 1995; Csaba and Dournaud, 2001). Regionally selective modifications in somatostatin levels caused various neurological disorders such as Alzheimer's, Parkinson's disease, depression, and epilepsy (Bissette and Myers, 1992; Epelbaum et al., 1994; Vecsei and Klivenyi, 1995; Eve et al., 1997; Vezzani and Hoyer, 1999). Somatostatin receptors (SSTRs) are widely distributed throughout many tissues and show different functions in various cell and tissue types (Weckbecker et al., 2003; Lahlou et al., 2004). One study represented that SSTR4 was thought to have a specific function in hepatic oval cells (Jung et al., 2006). Gastambide et al. (2009) found that the impairment of place learning and memory induced by hippocampal SST is mainly mediated by SSTR, hippocampal SSTR4 are functionally involved in a switch from hippocampus based memory to dorsal striatum-based memory. At present, it has not yet reports about SSTR4 related to AD. In this study, the CG sites of SSTR4 were hypomethylated, which inferred SSTR4 inactivity. Somatostatin (SST) may activate G protein-coupled receptors, which is corresponding receptor of neurotransmitter controlling behavior and mood including serotonin, dopamine, gamma-aminobutyric and glutamic acid related to AD. It may be a new gene related to AD; further research will go on.

Through pathway analysis, AD is involved in 13 biological processes and biological regulation accounted for the largest proportion of biological processes of AD. Alcoholism may lead to upregulation of systems that mediate the production of cytokines and other proinflammatory molecules (e.g. protein kinase C and activator protein 1) (Pandol and Raraty, 2007), which results in diseases such as pancreatitis (da Costa et al., 2011). Metabolic processes were associated with 215 of 827 differential methylation genes and it was also a main biological process in AD. Some complications of AD, such as type 2 diabetes mellitus, were involved in metabolic abnormalities (Ju et al., 2011).

Long-term alcoholism may result in changes in gene transcription by epigenetic modifications. Specifically, altered epigenetic modifications probably induced behavioral changes that are likely interactions between genetic and environmental factors for the genesis and maintenance of alcohol-seeking behavior. Further analysis of the differentially methylated genes revealed functionally relevant enrichment in biological processes and pathways involved in related genes.

The genes that included differential CG sites related to AD are listed in Table 5. Several significant associations with AD have been reported for subsets of the alcohol metabolism gene (e.g. aldehyde dehydrogenase [ALDH]), primarily in East Asian populations (Chen et al., 1999; Cheng et al., 2004; Higuchi et al., 2004). The ALDH gene, coding for the alcohol metabolizing enzyme ALDH, is clearly relevant and the most studied. ALDH transforms acetaldehyde into acetic acid, which can then be easily excreted. Mitochondrial ALDH2 is considered to play a major role in eliminating acetaldehyde (Vasiliou et al., 2004). The significant impact on acetaldehyde elimination of inactive ALDH2 causes high blood acetaldehyde concentrations after drinking and results in a painful adverse effect called the flushing response, including facial flushing, nausea, headache, and rapid heartbeat after drinking (Yoshida, 1992). This response generally leads to the avoidance of excessive alcohol consumption. But in this study, the ALDH1L2 gene is a differential gene, which has been rarely reported. A recent study reported the ALDH1L2 gene converted 10-formyltetrahydrofolate to tetrahydrofolate and CO2 in a NADP+-dependent reaction, which may be a likely source of CO2 production from 10- formyltetrahydrofolate in mitochondria, and played an essential role in the distribution of one-carbon groups (Krupenko et al., 2010). It also may be involved in acetaldehyde metabolism, which was not found. The hypomethylation of the referring gene in AD indicated ALDH1L2 activation and accelerated acetaldehyde elimination, and not high blood acetaldehyde concentrations, with increased alcohol consumption. Thus individuals had enhanced tolerance of alcohol.

Another gene family that has been studied is the gamma-aminobutyric acid type a receptors (GABRP) family, which codes for a family of chloride ion channels that mostly mediate rapid inhibitory neurotransmission throughout the CNS. The neurotransmitter GABA binds to these receptors, changing their conformation state and opening the channel to allow chloride ions to pass down an electrochemical gradient. The flux of chloride ions hyperpolarizes the membrane leading to neuronal inhibition (Kumar et al., 2009). In this study, the methylation degree of GABRP was the highest, with inactivity accompanied by a GABRP downregulation function, and decreased binding to GABA, increased GABA release, inhibitory weakening and then excitability enhancement (e.g. seizure severity). This result is consistent with some former studies. One of the studies reported that chronic ethanol exposure downregulates GABA receptor functions (Crews et al., 1996). In addition, ethanol-induced increases in GABA release likely contribute to ethanol-induced changes in GABAA receptor surface expression, which appear to be involved in ethanol tolerance, dependence, and withdrawal (Ziskind-Conhaim et al., 2003; Ariwodola and Weiner, 2004; Criswell and Breese, 2005; Criswell et al., 2008). This study suggested that long-term alcohol consumption led to GABAP gene methylation changes and then behavioral effects (e.g. seizure severity and withdrawal), and finally a higher alcohol intake.

In addition, according to the analysis of biological pathways related to AD using DAVID, some other genes were selected including dopamine beta-hydroxylase (DBH) and glutamate decarboxylase1 (GAD1).

DBH, a constituent of the catecholamine biosynthetic pathway, catalyzes the conversion of dopamine to noradrenaline or norepinephrine (Tang et al., 2007). Dopamine is considered to mediate the reward system in mesolimbic neurons related to AD and produce the experience of pleasure in activities. Several association studies on AD and DBH have been reported (Kato et al., 1979; Schuckit et al., 1981; Bagdy and Arató, 1987; Köhnke et al., 2006). If DBH activity increased, the conversion of dopamine increased and there was a reduction in dopamine. In this study, the CG sites of DBH were hypomethylated, which inferred DBH activity and dopamine reduction. These results are what are called “tolerance”. Once the feelings of pleasure have dissipated, induced by dopamine reduction, it requires more alcohol to achieve the same results. The more often you use alcohol, the more sensitized your receptors become and the more alcohol you require to achieve the effect. Alcohol addiction develops over time. Alcohol intake largely affects DBH methylation and DBH methylation further affects behavior.

GAD, the rate-limiting enzyme in the biosynthesis of GABA, may be involved in the development of alcoholism. Two isoforms of GAD have been identified: GAD1 and GAD2. One former study indicated a possible significant role for the GAD1 gene in the development of alcohol dependence and/or the course of alcohol withdrawal and outcome of alcoholism (Loh el et al., 2006). Another report presented that acute response to ethanol resulted in an increase of GABA release in the brain and required more GAD1 to transfer glutamic acid to produce GABA in animal studies (Kuo et al., 2009), which was consistent with this study. In this study, the GAD1 gene was hypomethylated and active, so it transferred more glutamic acid to produce GABA and an increase in GABA release. Repeated intake of alcohol resulted in upregulated GABA and downregulated glutamic acid.

Methylation levels generally are highly tissue- and region-specific. The use of blood mononuclear cells instead of brain cells to study genome-wide DNA methylation status may be regarded as a limitation of the present study. However, brain cells are difficult to obtain. Some studies showed that peripheral methylation and expression in peripheral blood cells may reflect changes in the central nervous system (Czermak et al., 2004; Hillemacher et al., 2009). For example, Hillemacher showed an association with a psychological parameter (craving), which can not be determined by peripheral regulation. Further research on the pathophysiological function of DAT and of its epigenetic regulation elucidated the role of epigenetic mechanisms in the neurobiology of alcohol craving (Hillemacher et al., 2009). Further investigations should try to replicate these findings using cerebral cells (e.g. from animal experiments). In addition, the identification of methylation loci did not necessitate a physiological effect and the prediction of gene targets on a global scale is hypothetical at best. Thus, the relevance of noted methylation difference must be validate on an individual bases.

In summary, genome-wide DNA methylation differs between AD patients and their siblings. This work provides novel insights that alcohol intake (environment) likely results in epigenetic modifications (e.g. DNA methylation) and epigenetic modifications (e.g. DNA methylation) may result in behavioral changes (e.g. alcohol intake largely). It may be helpful to better understand gene-environment interactions in the neurobiology of AD and more effective therapeutic options may be possible in the future. Further studies on AD and gene methylation are warranted.

Acknowledgments

This work was supported by Major Program of National Natural Science Foundation of China (20111483) and National Key Basic Research and Development Program (973) (2009CB522007), Natural Science Foundation of China (81130020) Henan Science Technology Committee (094200510005) to Wei Hao, National Natural Science Foundation of China (81171261) to Ruiling Zhang, and Fund for Talents with Innovation in Medical Science and Technology of Henan Province (3052) to Ruiling Zhang; Natural Science Foundation of China (81100996), Central Colleges basic scientific research operating expenses (2011QNZT170) and Specialized Research Fund for the Doctoral Program of Higher Education (20110162120013) to Yanhui Laio; Natural Science Foundation of China (30900486) to JinsongTang.

We thank the study participants, our research staffs, and Henan Mental Hospital. We acknowledge Henan Key Lab of Biological Psychiatry of Xinxiang Medical University, which provided all equipment for experiment. We wish to express our gratitude to the staff of Hunan Xinkaipu Compulsory Drug Rehabilitation Center, Hunan Baimalong Compulsory Drug Rehabilitation Center, and the Voluntary Drug Rehabilitation Center of Hunan Brain Hospital for their assistance throughout the course of this study.

Conflict of interest

The authors have declared that no competing interests exist.

Ancillary