Genetic Variation of the Ghrelin Signalling System in Individuals with Amphetamine Dependence

The development of amphetamine dependence largely depends on the effects of amphetamine in the brain reward systems. Ghrelin, an orexigenic peptide, activates the reward systems and is required for reward induced by alcohol, nicotine, cocaine and amphetamine in mice. Human genetic studies have shown that polymorphisms in the pre-proghrelin (GHRL) as well as GHS-R1A (GHSR) genes are associated with high alcohol consumption, increased weight and smoking in males. Since the heritability factor underlying drug dependence is shared between different drugs of abuse, we here examine the association between single nucleotide polymorphisms (SNPs) and haplotypes in the GHRL and GHSR, and amphetamine dependence. GHRL and GHSR SNPs were genotyped in Swedish amphetamine dependent individuals (n = 104) and controls from the general population (n = 310). A case-control analysis was performed and SNPs and haplotypes were additionally tested for association against Addiction Severity Interview (ASI) composite score of drug use. The minor G-allele of the GHSR SNP rs2948694, was more common among amphetamine dependent individuals when compared to controls (pc = 0.02). A significant association between the GHRL SNP rs4684677 and ASI composite score of drug use was also reported (pc = 0.03). The haplotype analysis did not add to the information given by the individual polymorphisms. Although genetic variability of the ghrelin signalling system is not a diagnostic marker for amphetamine dependence and problem severity of drug use, the present results strengthen the notion that ghrelin and its receptor may be involved in the development of addictive behaviours and may thus serve as suitable targets for new treatments of such disorders.


Introduction
Amphetamine and its derivatives are widely abused drugs and amphetamine dependence is of great concern in today's society. The development of drug dependence, such as for amphetamine, largely depends on the effects of amphetamine in the brain reward systems including the mesolimbic dopamine system (for review see [1]). These systems are important in mediating the rewarding properties of natural incentives, e.g. food, as well as of drugs, such as alcohol and amphetamine [2][3][4]. The drug response to amphetamine is heterogeneous and depends on factors such as expectancies of the drug, psychiatric disorders or personality traits, gender differences as well as genetics (for review see [5]). The heritability factor underlying drug dependence is shared between different drugs of abuse [6][7][8], and we hypothesize that ghrelin and its receptor, growth hormone secretagogue receptor (GHS-R1A), may be common denominators.
In addition, central ghrelin signalling is required for the reinforcing properties of drugs of abuse as supported by preclinical findings (for review see [25] or [26]). Specifically, a GHS-R1A antagonist attenuates the rewarding properties of alcohol, nicotine, cocaine and amphetamine as measured by locomotor stimulation, accumbal dopamine release and condition place preference [27][28][29]. In rodents, ghrelin has been shown to enhance cocaineinduced locomotor stimulation, to condition a place preference for cocaine and to induce cocaine-seeking behaviours [30][31][32]. Moreover, ghrelin increases, and a GHS-R1A antagonist decreases the consumption and operant self-administration of alcohol in rodents [28,33,34].
We have obtained support for these preclinical data in human genetic studies in which single nucleotide polymorphisms (SNPs) and haplotypes in the pre-proghrelin (GHRL) as well as GHS-R1A (GHSR) genes were associated with high alcohol consumption, increased weight and smoking in males as well as with personality traits related to alcohol use disorder in males with an alcohol use disorder [35][36][37]. Additionally, associations between a GHRL haplotype and paternal heredity of alcohol use disorder in females as well as increased weight in alcohol dependent males has been reported [37]. Haplotypes of the ghrelin signalling system have also been linked to increased sucrose consumption [23]. Given the shared genetic contribution to drug abuse, that polymorphisms in ghrelin related genes are associated with alcohol intake and smoking and that GHS-R1A antagonists blocks amphetamineinduced reward in mice, we hypothesise that genetic variations in the GHRL and GHSR may be associated with amphetamine dependence in humans. Thus, in the present association study SNPs and haplotypes of the GHRL and GHSR were investigated in Swedish individuals with a Diagnostic and Statistical Manual of Mental Disorders, 4 th edition, (DSM-IV) diagnosis of amphetamine dependence and who had additionally been assessed using the Addiction Severity Index (ASI).

Subjects
All individuals with a history of amphetamine dependence (26 women, 78 men) were recruited at the Stockholm Centre for Dependency Disorders (Beroendecentrum Stockholm), Sweden. The individuals had a DSM-IV diagnosis of amphetamine dependence and reported amphetamine as their primary drug of choice. They did not have a current diagnosis of any other major psychiatric disorders (such as bipolar, schizophrenia) or major somatic disorder (e.g. heart disease). Subjects testing positive for any other illicit substance were excluded. Data on nicotine use was not systematically collected on all patients, however, a majority (85%) were dependent on nicotine. The mean age of the patients was 40.3 years (SD 10.5, range 20-60). All patients gave their written informed consent to the study and the protocol was approved by the regional ethics review board at Karolinska Institutet, Stockholm, Sweden.
The control subjects consist of individuals (214 women, 96 men) randomly selected among a larger sample of low alcohol consuming individuals (i.e. self-reported alcohol consumption of 0.3-2.3 g alcohol/day) recruited for the INTERGENE study (see [36] for detailed description of this selection). These controls are Caucasians and did not self-report any drug use. However, 54 (out of 308 i.e. 17.8%) reported that they were current smokers. INTERGENE is a population based research program with the main objective of assessing the INTERplay between GENEtic susceptibility and environmental factors in the risk of developing chronic diseases in western Sweden [38][39][40]. The mean age of the selected control subjects was 56.6 years (SD 13.5, range 20-76). All control subjects gave their written informed consent to the study and the protocol was approved by the regional ethics review board at University of Gothenburg, Gothenburg, Sweden.

Clinical Assessment
The individuals in the present study are Caucasians and fulfilled the diagnosis of amphetamine dependence based on the DSM-IV criteria. In addition to the DSM-IV interview, patient alcohol and drug usage patterns were measured via the Addiction Severity Interview (ASI), a structured interview which maps the severity and duration of dependence across seven sub-scales [41]. 27 subjects were not interviewed with the ASI. All individuals included had the DSM-IV diagnosis of current amphetamine dependence. They did not have co-diagnosis of any other substance dependence ( Table 1). 51 of 62 individuals met ASI criteria for previous heavy alcohol consumption (more than 4 drinks for women and 5 for men, per occasion) in the past year. The urine toxicology of the patients confirmed that all patients were current D/L-amphetamine users. The average length of use was 13 years of regular usage (.4 days per week). Consumption ranged between 1-2 grams per occasion (average consumption of 6 grams per week). These amounts are typical for chronic amphetamine dependent individuals treated at the clinic. See table 1 for data on patient substance abuse characteristics.

SNP selection and genotyping
The Tag SNP selection procedure has been described in detail previously [37]. Briefly, genotyping data for the pro-ghrelin and GHS-R1A genes for the Caucasian CEPH population was downloaded from the International Haplotype Mapping Project web site (http://www.hapmap.org). Six Tag SNPs in the proghrelin gene (rs4684677, rs42451, rs35680, rs3491141, rs696217, and rs26802) and four Tag SNPs in the GHS-R1A gene (rs2948694, rs572169, rs2232165, and rs495225) were selected by using the Tagger function in the Haploview software with the r 2 set to a minimum of 0.80 (for their location and the SNPs which they tag). Out of these SNPs rs4684677 rs3491141, rs696217, rs572169, and rs495225 were force-included due to previous studies (for refs, see [37]).
The genotyping procedure has been described in detail previously [37]. Human genomic DNA was extracted from blood samples from both patients and controls using the QIAamp 96 DNA Blood Blood Kit (Qiagen, Hilden, Germany). The studied SNPs were genotyped from the obtained genomic DNA using TaqMan Pre-Designed SNP Genotyping AssaysH (Applied Biosystems, Foster City, CA) on the ABI PRISM 7900HT Sequence Detection System (Applied Biosystems) using the TaqMan Allelic Discrimination technology [42]. In the TaqMan Pre-Designed SNP Genotyping AssaysH a polymerase chain reaction amplifies the region around the SNPs. This polymerase chain reaction also includes two oligonucleotide probes, specific for one SNP allele each. These probes are labelled with one fluorescent reporter dye each (VIC and FAM) at one end and a quencher that absorbs fluorescence at the other end. When the DNA polymerase extends the newly synthesized DNA strand, it cleaves any probe that is tightly bound to the DNA, causing an increase in fluorescence. After polymerase chain reaction amplification the signal strength of each reporter molecule is measured and displayed in a scatter plot with the fluorescence of the probes are plotted on the X-and Y-axis. Given that each point represents one sample the genotype of each sample can be determined. For details on the studied Tag  SNPs and the TaqMan assay ID, see table 2.

Statistical analyses
Deviation from Hardy-Weinberg equilibrium (HWE) was assessed for all genotyped SNPs in cases and controls separately. Allele frequencies were compared between amphetamine dependent patients and controls using Chi2-test. Homogeneity of odds ratios was tested with a Breslow-Day test. The possible influence of the studied SNPs on the ASI composite score of drug use was investigated using linear regression with sex and age as covariates.
For each gene, haplotype analyses were performed including all SNPs using stepwise logistic regression in the case-control analysis and linear regression including sex and age as covariates in the ASI composite score of drug use analysis. Haplotype frequencies were estimated using the expectation-maximisation algorithm [43]. A significance level of 0.05 was used. Due to LD between SNPs correction for multiple testing was carried out using 10 000 permutations in each gene (i.e. six in GHRL and four in GHSR1).

Results
All studied SNPs in both the patient and control group had a HWE p-value.0.01. As seen in table 3, the minor allele of the rs2948694 SNP located in intron 1 of the GHSR gene was associated with amphetamine dependence (p c = 0.02, allelic OR 1.84). When men and women were analysed separately for rs2948694 the ORs were 1.48 and 2.11 respectively, but not significantly different (p = 0.5). When the control group was split into non-smokers and current smokers the minor allele frequencies were 10.0% and 13.0% respectively. Pairwise comparisons of the ASI, addiction severity index. *Variation in n due to variation in number of individuals responding to the ASI items. 1 The composite score ranges from 0 (no problems) to 1 (severe problems). 2 Percentage users out of valid responses. 3 Data expressed as mean (SD) of number of years in life that the drug has been abused (subjects who had not taken the given drug were excluded). 4 The ASI does not specify which drug was used intravenously. doi:10.1371/journal.pone.0061242.t001  Single marker analysis also showed association between the minor allele (90Leu) of the GHRL SNP rs4684677 and increased ASI composite scores for drug use (table 4).
Although associations were found both between a haplotype in the GHSR gene and diagnosis as well as a haplotype in the GHRL gene and ASI composite scores for drug use these analyses did not contribute to additional information next to the single marker analysis.

Discussion
The present study is, to our knowledge, the first investigating the GHRL and GHSR genes, encoding ghrelin and its receptor, in relation to amphetamine dependence and problem severity of drug use. We found that the minor allele of a SNP located in GHSR (rs2948694) is associated with amphetamine dependence. This finding was strengthened by the fact that the control smokers had an allele frequency intermediate between non-smokers and patients. We also found a significantly higher ASI drug composite score in patients carrying the minor allele of a GHRL SNP (rs4684677).
Common neurobiological mechanisms for the rewarding properties of both drugs of abuse and food have been suggested in both animal and human studies [4,44,45], and the ghrelin signalling system could be one of the mediators of this effect. This may be one reason for the high co-morbidity between eating disorders and drug dependence [46]. Both preclinical and human genetic findings have previously suggested an important role for GHS-R1A in drug-induced reward (vide infra). Specifically, the ability of alcohol, nicotine, cocaine as well as amphetamine to induce locomotor stimulation, increase accumbal dopamine release and to condition a place preference, are attenuated by GHS-R1A antagonists in mice [28,29,47]. Previous human genetic studies have reported association between SNPs and haplotypes of the GHSR gene and heavy alcohol consumption as well as body mass index in heavy alcohol consuming individuals [37]; smoking and type II alcohol dependence in women [36] as well as with self-directedness in alcohol dependent individuals [35]. Collectively, these data suggest that GHS-R1A may be involved in drug reward, drug intake as well as characteristic traits of addictive behaviours and the receptor may therefore be a potential target for treatment of various addictions. It has been suggested that either the constitutive activity of the GHS-R1A [48] or the ability of the GHS-R1A to form a heterodimer with dopamine D1-like or D2 receptors [49,50] may affect the set point of the mesolimbic dopamine system. The functional role of the studied SNP is not known and the significance of these results cannot be determined. However, we hypothesize that the SNP in the GHS-R1A gene could, tentatively, influence the activity of tegmental dopamine neurons and thereby influence the ability of these dopamine neurons to be activated by addictive drugs. However, this needs to be further elucidated.
We also found that patients with the uncommon variant (90Leu) of the GHRL SNP (rs4684677) have higher ASI composite scores for drug use. Our previous human genetic studies have reported associations between SNPs and haplotypes of the GHRL gene and body mass index in heavy alcohol consuming individuals [37]; paternal alcohol dependence and withdrawal symptoms in female alcohol-dependent individuals [36] and with self-transcendence in alcohol dependent individuals [35], and ghrelin may thus be implicated in mechanisms underlying drug use. Preclinical data support an important role of ghrelin in drug intake and drugreward. In mice, ghrelin administration into brain ventricles or locally into reward nodes (ventral tegmental area or laterodorsal tegmental area) increases the intake of alcohol [28] and alcoholinduced reward is attenuated in ghrelin knockout mice [47]. Moreover, the ability of cocaine to increase locomotor activity and  condition a place preference is enhanced by ghrelin in mice and high plasma levels of ghrelin is associated with cocaine-seeking behaviours in rats [30][31][32]. Although hypothetical, it could be speculated that this SNP, if resulting in higher ghrelin levels, could lead to drug seeking or craving, as ghrelin plasma levels have been previously shown to be higher during alcohol abstinence [51] and craving for alcohol [52]. Interestingly ghrelin and its receptor appear to be important for reward in general. Specifically, suppression of the GHS-R1A reduces reward induced by, the intake of and motivation to consume palatable foods [21,22]. It has also been shown that ghrelin and its receptor regulates the intake of and motivation to consume sucrose and saccharin [24,53,54], at the level of the mesolimbic dopamine system [55]. Human functional magnetic resonance imaging studies show that ghrelin administration increases the natural response to food in reward related areas, including the mesolimbic dopamine system [56]. In addition, human genetic findings show an association between a haplotype in the GHRL gene and high sucrose consumption [23]. Taken together with the data regarding ghrelin, the GHS-R1A and drugs of abuse it may be suggested that ghrelin and its receptor may have an important role in regulating reward in general and that GHS-R1A may be a potential target for new treatments of addictive behaviours such as compulsive overeating.
The current study is faced with limitations, the main being the small sample size, the possibility of a type 1 error and population stratifications. Therefore our study should be considered as a pilot study and additional studies replicating our data are warranted. Another confounding factor may be that the controls self-reported nicotine and alcohol usage via a questioner. Unreported usage of addictive drugs could therefore tentatively affect the obtained results. Additionally, the modest sample size may limit the power to detect effects. However, the consistency in our findings where a SNP in GHSR and GHRL were associated with amphetamine dependence and ASI drug composite score, may suggests an important role for ghrelin and its receptor in drug dependence. Given previous reports showing associations between the studied SNPs and heavy alcohol-use as well as smoking [37] and the comorbidity between amphetamine, alcohol and nicotine dependence, we cannot rule out the possibility of an interaction between the studied SNPs and alcohol and/or nicotine dependence. Additionally, the functional value of the studied SNPs is yet to be determined. Although our research is highly hypothesis driven, the findings reported here must be regarded as preliminary until replicated.
In conclusion, even though genetic variability of ghrelin and its receptor is not diagnostic marker for amphetamine dependence and problem severity of drug use, the present results still strengthen the notion that ghrelin and the GHS-R1A are involved in some of the mechanisms underlying the development of addictive behaviours.