Ionotropic Glutamate Receptor Gene GRIK3 SER310ALA Functional Polymorphism Is Related to Delirium Tremens in Alcoholics

Ionotropic glutamate receptor gene GRIK3 SER310ALA functional polymorphism is related to delirium tremens in alcoholics

UW Preuss1,2, P Zill1, G Koller1, Upregulation of glutamatergic neurotransmission resulting from chronic 1 3 ethanol intoxication may cause a hyperexcitable state during alcohol B Bondy , V Hesselbrock and withdrawal, which may lead to seizures and delirium tremens. The aim of 1 M Soyka our study was to evaluate the association between a history of alcohol withdrawal-induced seizures and delirium tremens, and a functional 1Department of Psychiatry, Ludwig-Maximilians Universita¨t Mu¨nchen, Mu¨nchen, Germany; polymorphism (Ser310Ala) of the GRIK3 gene coding for the glutamatergic 2Johanna-Odebrecht-Stiftung, Greifswald, kainate receptor subunit GlurR7 in a sample of well-characterized alcoholics Greifswald, Germany and 3Department of compared to controls. In total, 233 patients meeting DSM-IV alcohol Psychiatry, University of Connecticut School of dependence criteria and 309 controls, all of German descent, were Medicine, Farmington, CT, USA investigated. GRIK3 functional polymorphism was determined using PCR Correspondence: (polymerase chain reaction) of lymphocyte DNA. History of alcohol PD Dr med. U.W. Preuss, Johanna-Odebrecht- withdrawal-induced delirium tremens and seizures were obtained using the Foundation, Gu¨tzkower Landstrasse 69, 17489 SSAGA (Semi-Structured Assessment for the Genetics of Alcoholism). Data Greifswald, Germany. were cross-checked with in-patients’ clinical files. While a significant E-mail: [email protected] relationship between history of delirium tremens and the Ser310 allele was detected, no significant results were obtained for alcohol withdrawal-related seizures. Although this result is suggestive for a significant role of this polymorphism in the pathogenesis of delirium tremens in alcohol-dependent individuals, further investigation and confirmation are warranted. The Pharmacogenomics Journal (2006) 6, 34–41. doi:10.1038/sj.tpj.6500343; published online 29 November 2005

Keywords: GRIK3; glutamate system; alcohol dependence; withdrawal; delirium tremens

Introduction

Ethanol withdrawal following chronic alcohol consumption is reported to produce dramatic changes in the content of brain excitatory neurotransmitters as well as their associated receptor sensitivity and regulation.1 Upregulation of glutamate neurotransmission as a consequence of chronic alcohol intake may contribute not only to neurotoxicity but also to neuronal death observed in brains of alcohol-dependent individuals (e.g.Davis and Wu,2 and Bleich3). This upregulation is related to the neuroadaptation, which results from chronic ethanol intoxication, causing a hyperexitable state that may also lead to seizures and delirium tremens during alcohol withdrawal.1,4,5 However, while most research focuses on the effects and consequences of acute and chronic alcohol Received 16 November 2004; revised 31 August 2005; accepted 1 September 2005; consumption on glutamatergic ionotropic NMDA (N-methyl-D-aspartate) recep- published online 29 November 2005 tors, several other ligand-gated channels activated by glutamate might be Kainate receptor functional polymorphism and delirium tremens UW Preuss et al 35

involved in the pathophysiology of alcoholism, including confirmed the functional relevance of this polymorphism. alcohol withdrawal-related complications such as delirium First, a lower KA-receptor expression of T-allele carriers tremens and seizures. (Ser310) was detected. In seven of nine human brain tissue Besides NMDA and AMPA (alpha-amino-3-hydroxy- samples under investigation, a 1.4- to 4.7-fold reduction of 5-methylisoxazole-4-propionate) receptors, kainate (KA) this allele expression was found. Second, while the Ser310 receptors are the third class in members of the glutamate- allele expression levels in the occipital cortex, mesencepha- gated family of ionotropic receptors. It is assumed that lon, cerebellum basal ganglia and thalamus was significantly KA-receptors channels are formed by tetramers or pentamers lower, the frontal brain had a two-fold higher expression of GluR5-GluR7, KA1, KA2 subunits. While GluR5-7 repre- level of this allele.28 sent low-affinity subunits and KA1-2 are considered high- The aim of this analysis is to investigate the influence of a affinity subunits,6 it is not clear as to which combination GluR7 functional polymorphism on a history of epileptic actually exist in native KA receptors in mammalian brains. seizures or delirium tremens during alcohol withdrawal in Some pathways involved in physiological activation of KA an in patient sample of alcohol-dependent subjects. Delir- receptors are still unknown,7 although physiological studies ium tremens and epileptic seizure may have different and have identified both post- and presynaptic roles for KA independent neurobiological mechanisms, they often co- receptors. This class of receptors has also been suggested to occur during withdrawal. For a subsequent analysis, there- play an important role in short- and long-term synaptic fore, three groups of alcohol-dependent subjects with a plasticity.8 KA receptors have been reported to generate a lifetime history of delirium tremens only, epileptic seizures slow excitatory activity-dependent synaptic current, which only and co-occuring delirium tremens and seizures are greatly augments excitatory currents of hippocampal CA3 associated with this GluR7 functional polymorphism. pyramidal cells.9 While located both pre- and postsynap- Furthermore, the potential influence of severity of alcohol ticlly, KA receptors play significant roles in the regulation of dependence, reflected by illness duration, severity of alcohol hippocampal excitability. They inhibit glutamate release dependence according to number of DSM IV criteria presynaptially10–12 and also regulate action potential-depen- endorsed, poor somatic condition, early age of onset and dent GABA release from interneurons. KA receptor mediated daily alcohol intake as well as subtypes of alcohol-depen- increase of interneuronal GABA release may function as an dent subjects (Babor’s Type A and B) are evaluated. All these important homeostatic mechanism that prevents overexci- characteristics of alcohol dependence might increase the tation of hippocampal CA1 pyramidal neurons.13,14 risk for both epileptic seizures and delirium tremens during While little was known about adaptations in KA receptors alcohol withdrawal and mediate the influence of genetic after chronic ethanol intake, in recent years KA receptors polymorphism on alcohol withdrawal complications. have emerged as important targets of alcohol’s action in the central nervous system.15 Exposure to ethanol for 1–5 days selectively decreased the sensitivity of KA receptor responses Results to low KA concentrations.16 Furthermore, an upregulation of these receptors under chronic administration of alcohol Two hundred and thirty-three alcohol-dependent indivi- in hippocampal regions has been reported. This upregula- duals (189 males and 44 females, aged from 18 to 74 years, 7 tion was suggested to eventually result in an increased mean age 42.02 9.8 years, see Table 1), all meeting ICD10 inhibition of GABAergic interneurons in the hippocampus and DSM-IV criteria for alcohol dependence were recruited. during alcohol withdrawal.17 This makes KA receptors As controls, 309 persons (161 males, 148 females, aged from 7 prime candidates in hippocampal kindling, which is 19 to 76 years; mean 46.8 12.1 years) were enrolled into suggested to be an important pathophysiological factor the study. History of alcohol withdrawal-induced delirium generating alcohol withdrawal-induced seizures and tremens was noted in 45 (19.3%), while a history of alcohol delirium tremens.18 withdrawal-induced epileptic seizures was reported from 54 The GluR7 subunit of KA receptors has been mapped to (23.2%) of the 233 alcohol-dependent subjects. Also pre- chromosome 1 at 1p34–p33,19 close to a region reported to sented in Table 1 are the duration of alcohol dependence, be associated with alcoholism and alcoholism-related phe- number of DSMIV alcohol-dependence criteria endorsed and notypes from linkage studies.20–22 GluR7 receptor subunits the number of subjects meeting Babor’s type A and B criteria. form a functional homomeric receptor channel with low sensitivity to glutamate.23 The GluR7 contains an exonic Genotype results thymine/guanine nucleotide variation that determines a The genotype frequencies of alcohol dependent subjects and serine or alanine at position 310 in the N-terminal controls are presented in Table 2. Data for both alcohol- extracellular domain of the receptor.24 This domain has dependent individuals and controls followed the Hardy– been reported to affect receptor desensitization25 and to Weinberg equilibrium (w2: 2.18, df: 1, P: 0.34 and w2: 0.17, participate in ligand binding.26,27 Prior electrophysiological df: 1, P: 0.92). investigations of both alleles in HEK 293 cells did not To assess the influence of GRIK3 polymorphism on indicate that the replacement of serine 310 by alanine alcohol dependence, GRIK3 genotype and allele frequencies affects receptor responses to glutamate.28 However, expres- were compared between alcoholic subjects and controls. No sion studies and characterization of putative consequences significant differences were detected.

The Pharmacogenomics Journal Kainate receptor functional polymorphism and delirium tremens UW Preuss et al 36

Table 1 Sample characteristics

Alcohol-dependent subjects Controls

Gender (m/f) 189/44 (n ¼ 233) 161/148 (n ¼ 309) Age (years) 42.0279.8 40.03715.0 Age of onset (years) 29.5179.0 Years of dependence (years) 12.5779.1 Daily alcohol intake (g/day) 331.657191.3 Number of DSM IV alc. dependence criteria 5.6071.3 Number of alcohol-related somatic problems 2.0272.6 Alcohol-withdrawal induced seizures 45/9 (23.2%) Delirium tremens 41/6 (20.2%) Babor type A/B 158 (67.8%)/75 (32.2%) m: male; f: female; g/d: gram per day.

Table 2 GRIK3 receptor functional polymorphism genotype distribution in alcoholics and controls

GRIK3 (GluR7) genotype w2 value df Sig

Ser/Ser Ser/Ala Ala/Ala Ser Ala

Alcohol-dependent subjects Sex (m/f); n ¼ 233 106/20 (54.1%) 66/15 (34.8%) 17/9 (11.2%) 71.5% 28.5% 0.66 2 0.72

Control sample Sex (m/f) n ¼ 309 89/76 (53.4%) 56/64 (38.8%) 16/8 (7.8%) 72.8% 27.2%

Alcoholic subjects with Controls Alcoholics without delirium/seizures

w2 df Sig w2 df Sig

Delirium tremens Ser310 3.91 1 0.030 7.67 1 0.002 (0.008 corrected) Seizures Ser310 1.86 1 0.177 0.10 1 0.75

Delirium tremens 310Ala 3.56 1 0.041 4.79 1 0.021 (0.084 corrected) Seizures 310Ala 1.34 1 0.173 1.19 1 0.17

GRIK3: ionotropic glutamate receptor kainate 3; m: male; f: female. Bold represents values which are statistically significant.

Alcohol withdrawal-induced seizures, delirium tremens and GRIK3 1.39). Comparing alcoholics with a history of delirium functional polymorphisms tremens to controls, again a significant influence of the Alcohol-dependent individuals with a history of alcohol Ser310 allele was detected (w2: 3.91, df: 1, P: 0.03, OR: 1.17, withdrawal-induced seizures showed no significant relation- CI 95%: 1.12–1.22). ship to either the GRIK3 Ser310 or 310Ala allele compared Of the 29 patients without Ser310 allele, none had even to healthy controls and alcoholics without seizures one episode of delirium tremens, compared with 12.5% (11 (see Table 2). of 88) of patients with one Ser310 allele and 21.5% (15 of Comparing alcohol-dependent individuals with a history 116) patients homozygous for this allele (w2: 6.81, df 1, of alcohol withdrawal-induced delirium tremens with con- P: 0.009, Cochran-Armitage linear test). trols, a weak but significant relationship to the GRIK3 While not reaching statistical significance, alcoholic polymorphism was detected (w2: 3.91, df: 2, P: 0.03). A more subjects who are homozygous carriers of the Ser310 allele pronounced result was detected comparing alcohol-depen- reported a slightly higher rate and an earlier age of onset of dent individuals with delirium tremens to individuals delirium tremens, compared to alcoholic subjects hetero- without (w2: 7.67, df: 1, P: 0.002; Table 2). zygous for this allele (rate of delirium tremens: 2.5572.2 vs Alcohol-dependent subjects who reported a history of 1.8471.40, t-value: À0.98, P: 0.35; delirium tremens age of delirium tremens were significantly more often carriers of onset: 30.36711.0 vs 33.32712.8, t-value: 0.65, P: 0.51). the Ser310 allele as compared to alcoholics without delirium Finally, the effects of the GRIK3 polymorphism on history tremens (w2: 7.67, df: 1, P: 0.002, OR 1.29, CI 95%: 1.20– of delirium tremens and epileptic seizures occurring

The Pharmacogenomics Journal Kainate receptor functional polymorphism and delirium tremens UW Preuss et al 37

together were assessed. In total, 26 (11.2%) subjects had a COGA study.20,21 This association was recently supported history of delirium tremens only, 29 (12.4%) of epileptic by a family-based study using genetic fine-mapping of the seizures without delirium tremens while 28 (12.0%) subjects 1p region of chromosome 1.29 reported epileptic seizures and delirium occurring together Comparing gentoype frequencies in our study with at least once. previous research (Begni et al.,30 n ¼ 116 and Schiffer et The GRIK3 Ser310 allele was found significantly more al.,28 n ¼ 35), control subjects had a similar distribution of often in the delirium tremens only group (w2: 3.76, df: 1, alleles and genotypes. However, the allele frequencies in our P: 0.035) while no significant relationship was detected alcoholic subjects were different from those obtained by between the GRIK3 polymorphism and history of epileptic Begni et al.,30 who reported lower genotype frequencies of seizures only (w2: 0.40, df: 1, P: 0.33) and for epileptic the GRIK3 polymorphism Ser allele in his sample of seizures and delirium tremens occurring together (w2: 1.24, schizophrenic patients. These patients had a very different df: 1, P: 0.23). psychiatric diagnosis of schizophrenia while the in-patient sample enrolled into this study all had a diagnosis of Alcoholism-related clinical characteristics and GRIK3 genotypes alcohol-dependence. Moreover, subjects with a comorbid Comparing alcoholism-related clinical characteristics, schizophrenic psychosis were excluded from our study. which may have a significant influence on the risk of Alcoholism-related traits are reported to have a complex delirium tremens such as previous alcoholism-related treat- inheritance with genetic heterogeneity of the human ment, duration of alcohol dependence, number of alcohol- susceptibility to delirium tremens and other alcoholism- related physical problems and age of onset, no significant related phenotypes. Thus, these phenotypes including association with the GRIK3 polymorphism was found. delirium tremens might be caused by a number of genes While a statistical tend was found for more DSM IV with small effects. Furthermore, little is known about the criteria endorsed in GRIK3 Ser310 allele carriers (ANOVA interplay and control of these genes on the genetic and F-value: 2,73, df: 2, P: 0.067), no relationship was detected meta-genetic level. GRIK3 might be one of them involved in for Babor’s A/B subtypes. the pathophysiology of delirium tremens. This is the first study conducted on KA receptor GRIK3 Discussion Ser310Ala functional polymorphism and alcohol withdrawal-related phenotypes. A growing number of previous In this analysis of the Munich Gene Data Bank for studies have reported an association between putative alcoholism (MGBA), associations between the history of functionally relevant dopamine transporter polymorphisms delirium tremens and alcohol withdrawal-induced seizures and withdrawal seizures or delirium tremens in alcohol- with a functional polymorphisms of the KA receptor subunit dependent subjects,31–34 while some of these positive gene GRIK3 coding for GluR7 were investigated. The findings are not unanimously confirmed by other research significant association between delirium tremens history groups using a more specific but less powerful family-based with the Ser310 allele supports the hypothesis of a association approach.35 significant role of this GRIK3 functional polymorphism in In mice, an association between quantitative trait loci in the pathogenesis of severe complications in alcohol with- proximity to genes that directly or indirectly affect gamma- drawal. The result is strengthened by the two comparisons aminobutyric acid (GABA)-A receptor-mediated transmis- conducted, which were both significant: alcoholics with and sion and withdrawal intensity have been reported.36 without delirium tremens, and alcoholics with delirium Reduced coupling between the GABA recognition site on tremens and controls. However, the latter association was the GABA-A receptor and the chloride channel has been weaker and the odds ratio decreased from 1.29 to 1.17. An noted.37 The interaction between glutamatergic and GA- explanation could be that this GRIK3 functional poly- BAergic neurotransmission may be of high significance in morphism is modifying the alcoholism phenotype rather the development of delirium tremens. Decreased GABA-A- than being a risk factor for alcohol dependence itself. This mediated inhibition during alcohol withdrawal has been hypothesis is supported by the analysis of the relationship related to increased glutamate-mediated excitability and between GRIK3 polymorphism, delirium tremens and both may be critical mechanisms of alcohol withdrawal- epileptic seizures occurring alone or in combination. This induced seizures38 and the development of delirium tre- GRIK3 polymorphism might exert a differential effect on the mens, which often occur simultaneously since many liability for delirium tremens in alcohol-dependent subjects alcoholics with delirium tremens also experience epileptic only, since no significant effects was found for the epileptic seizures.39 Glutamate enhances inhibitory GABA neuro- seizure or the combined epileptic seizure and delirium transmission.1 During chronic alcohol consumption, de- tremens group. Furthermore, a gene–dose effect of the creased glutamatergic neurotransmission may eventually GRIK3 Ser310 allele was found where the rate of delirium cause GABA to be low. During withdrawal, increased tremens increased significantly in those persons who carried excitatory neurotransmission may contribute to even lower one or two copies of this allele. GABAergic neurotransmission. The combined effect of the The GRIK3 gene is localized on chromosome 1 (1p34– inferred increased excitatory neurotransmission and de- p3319) in vicinity of a genomic area related to alcoholism creased GABAergic neurotransmission would lead to sub- and alcoholism-related phenotypes as reported by the stantial amplification of the overall excitatory

The Pharmacogenomics Journal Kainate receptor functional polymorphism and delirium tremens UW Preuss et al 38

neurotransmission.1 These effects eventually may result in Limitations of this study include a retrospective design. more severe withdrawal and neuronal death since even Our data were obtained from in-patient alcoholics after selective neuronal degeneration has been reported resulting alcohol withdrawal using a semistructured interview and from the activation of glutamatergic ionotropic receptors.40 later crosschecked with in-patients’ clinical files. The While the exact role of ionotropic KA receptors in alcohol number of in-patients enrolled into the study who were dependence is unknown, previous research reported that in affected by delirium tremens and alcohol-withdrawal related the presence of low concentrations of EtOH (5–10 mM), KA seizures was only a small subgroup of the alcoholic patients receptors in hippocampal interneurons are inhibited, which investigated. Subsequently, the number of patients with this reduces the excitatory effect to these neurons, subsequently phenotype compared to the control group was small decreases GABAergic transmission and increases the excit- (delirium tremens: n ¼ 45, 19% of the patients). Thus, the ability of pyramidal neurons. In contrast, higher concentra- results reported from this analysis need replication in larger tions of ethanol reduce the inhibitory effect of KA receptors samples to exclude a spurious association. on GABAergic transmission,41 inhibit KA-receptor-mediated Due to the exploratory character of this study, it should be synaptic currents in hippocampal CA3 pyramidal neurons42 mentioned that in a phenotype with a complex inheritance, and modulate the function of other neuronal proteins, a number of genes and polymorphisms might be involved. including GABA-A and NMDA-receptors.15 The end results Each of these genes may contribute in a minor manner to of the combined action of ethanol on all these effects is the phenotype and might be only detected in very large believed to cause neuronal depression,43 which may also be samples with sufficient power if a very low significance level a result of chronic alcohol intake. During withdrawal, how- is applied. ever, upregulated NMDA-receptors are no longer blocked by Another limitation of this study is the lack of a genomic chronic alcohol intake, and the decreased number or control technique to confirm the results of the study. While impaired function of GABA-A receptors may be too low to all the subjects enrolled into the study were carefully compensate for excessive glutamatergic excitation.44 examined regarding their ethnic background, stratification Concerning the potential role of hippocampal KA recep- bias cannot be excluded. A family-based association method tors during withdrawal, while previous research reported using case–parent triad design to avoid ethnic stratification significant acute effects of alcohol intake on KA receptor bias, however, was not used in this analysis and might be expression and function,15,16 chronic and high-dose alcohol necessary in subsequent studies. consumption may lead to upregulation of KA receptors, as well as NMDA receptors, in particular in hippocampal regions and subsequently resulting in stronger GABAergic Materials and methods transmission and neuronal depression. During withdrawal from chronic alcohol intake, higher expression of KA Patients receptors may eventually amplify low GABAergic transmis- Seven years ago, the Munich Gene Data Bank for alcoholism sion, beside the direct effects of chronic alcohol ingestion (MGBA) project was initiated.48–50 All participants were on GABA-receptors. These effects may contribute to the recruited as in patients from a ward for treatment of alcohol generation and intensity of alcohol withdrawal by facilitat- dependence. All patients were older than 18 years and met ing kindling in hippocampal regions. Alcohol withdrawal- ICD10 and DSM-IV criteria for alcohol dependence. Previous related kindling is a phenomenon in which a weak electrical and current epileptic seizures, delirious symptoms such as or chemical stimulus, which initially causes no overt hallucinations or severe disturbance of consciousness were behavioral response, results in the appearance of behavioral assessed with a structured interview (SSAGA: Semi-Struc- effects, such as more severe withdrawal or seizures, when it tured Assessment for the Genetics of Alcoholism51,52). is administered repeatedly.45 Repeated withdrawal from Personality diagnosis, including antisocial personality dis- chronic alcohol consumption is reported frequently from order, was assessed using the SCID II interview (German alcohol-dependent subjects and over time might increase Version53). Furthermore, a comprehensive psychiatric ex- the risk of developing an alcohol withdrawal-related amination was performed by two experienced psychiatrists seizures46 or delirium tremens.47 (UWP or MS). The patients’ reports were cross-checked with Genetic polymorphisms of the GRIK3 gene may explain the original patient files. Using ICD10 criteria, the history, inter-individual differences in the density and affinity of KA number of alcohol withdrawal-induced seizures (F10.31 or receptors. Ser310 allele carriers have lower expression levels F10.41) and delirium tremens (F10.4) were identified. In- of this receptor during alcohol withdrawal. This may patients were investigated 2 weeks after admission and after destabilize the interaction between KA receptors and alcohol withdrawal when free of any psychopharmacologi- GABAergic neurotransmission and may eventually be a risk cal treatment. factor for more severe withdrawal and delirium tremens. Age of onset of alcohol dependence was assessed comput- However, the exact effects of chronic alcohol consump- ing the mean of retrospectively obtained first alcohol tion on expression levels of KA receptors depending on dependence ages of onset criteria as mentioned in DSM-IV Ser310Ala alleles are not known and deserve further research by the SSAGA: higher consumption of alcohol than to reveal its role in alcohol withdrawal kindling and the intended, attempts to stop or control alcohol consumption, pathogenesis of delirium tremens. significant time spent to consume alcohol or recover from

The Pharmacogenomics Journal Kainate receptor functional polymorphism and delirium tremens UW Preuss et al 39

alcohol intake, regular withdrawal symptoms during im- 0.5 mM of the forward and reverse primer and 0.75 U Ampli portant daily obligations like school or work, reduction of Taq Gold DNA polymerase (Applied Biosystems, Foster City, important occupational or private activities because of USA). After an initial denaturation step of 951C for 10 min, alcohol intake, continued alcohol consumption despite there were 35 cycles of denaturation at 951C for 30 s, the occurrence of psychological or physical harm and annealing at 621C for 30 s and extension at 721C for 1 min. A occurrence of 50% higher tolerance to alcohol effects. Daily final step was performed at 721C for 7 min. The PCR product alcohol intake was obtained using the typical daily average was digested with 5 U SmaI over night at 251C. An uncut alcohol consumption of the last 8 days before admission. fragment of 180 bp represents the A-allele, whereas two Pure alcohol intake was computed in g/day. Duration of fragments of 116 and 64 bp indicate the C-allele. alcohol dependence was computed as the difference between current age and age of onset. Statistics Alcohol-dependent subjects were subgrouped according All continuous data were tested for normal distribution. 54 to Babor types A and B by using five characteristics of Hardy–Weinberg equilibrium was analyzed for each sample. 55 alcohol dependence mentioned by a subsequent study, The relationships between continuous parameters like age which were assessed by the SSAGA. These characteristics and alcohol dependence criteria such as age of onset, included daily alcohol intake, relief drinking, medical duration of alcohol dependence and daily alcohol intake conditions due to alcohol, physical and social consequences. across GRIK3 genotype were computed using one-way Composite scales were calculated for the five characteristics ANOVA. Post hoc tests (Scheffe´’s test) were applied when and dichotomized using median split. Subjects were subtyped necessary. according to type A or B, if they met three or more of the The relationships between other variables like frequencies dichotomized scores criteria (upper half: type B, lower half of seizures and delirium tremens with genotypes were tested type A). Furthermore, taking the SSAGA data, number of DSM using w2 tests. To permit comparisons of the different IV criteria endorsed and number of six alcoholism-related associations found, independently of the sample size, odds physical problems were computed. Finally, to compare the ratios (OR) were computed with 95% confidence intervals influence of the GRIK3 polymorphism on in-patient’s history (CI) comparing allele frequency and disease-related traits. of delirium tremens and epileptic seizures, alcohol-depen- Linear trend in variations of frequencies of the SER310ALA dent in-patients subjects were subgrouped into subjects with polymorphism alleles (presence of 0, 1 or 2 alleles) with a history of delirium tremens only or epileptic seizures only, respect to the presence of clinical phenotypes (delirium or or both together. seizures) was subsequently evaluated with the Cochran– Control persons from the general population were Armitage test.57 Clinical variables (such as duration of recruited at different locations (e.g. libraries, road-works, alcoholism and age at interview) were compared in different big stores), representing different social groups from un- groups (with vs without the studied allele) using an analysis skilled worker to university graduates. In all control persons, of variance. a comprehensive medical and psychiatric assessment was A two-tailed a-significance level of Po0.05 was defined to carried out together with routine laboratory screening to be statistically significant. Due to the experimental char- exclude severe physical or psychiatric axis I or axis II disorders acter of this analysis, correction of significance levels was such as schizophrenia, depression, personality disorders and not used. substance use disorders including alcohol dependence. In order to assess psychological problems, personality Ethical standards traits and possible present psychopathological symptoms A signed written informed consent was obtained from were evaluated using the MMPI (Minnesota Multiphasic patients and controls after complete and extensive descrip- Personality Inventory56), one of the most commonly used tion of the study. The Ethics Committee of the Ludwig- personality questionnaires in clinical practice. All persons Maximilians-University of Munich approved of the study. with first-degree relatives suffering from any axis I disorders or alcohol dependence were not included. All patients and controls were Caucasian and of German descent to avoid Duality of Interest ethnic stratification bias. None declared. Genotyping Genomic DNA was isolated from whole blood according to Acknowledgments standard procedures. The GRIK3 ser310ala (T928G) polymorphism (rs6691840) was detected by PCR amplification, We thank Dr WM Wong for English language review. restriction enzyme digestion, electrophoresis on 2% agarose gels and visualization by ethidium bromide staining under References the following conditions: forward primer: 50-GAA GCC CCT GGA ATT CAC CT-30; reverse primer: 50-CCG GAT TCT CAA 1 Tsai G, Gastfriend DR, Coyle JT. The glutamatergic basis of human alcoholism. Am J Psychiatry 1995; 152: 332–340. 0 TGT GGA CAA-3 . PCR was carried out in a final volume of 2 Davis KM, Wu JY. Role of glutamatergic and GABAergic systems in 25 ml containing 50 ng genomic DNA, 200 mM of each dNTP, alcoholism. J Biomed Sci 2001; 8: 7–19.

The Pharmacogenomics Journal Kainate receptor functional polymorphism and delirium tremens UW Preuss et al 40

3 Bleich S, Degner D, Bandelow B, von Ahsen N, Ruther E, Kornhuber J. domains structurally related to bacterial amino acid-binding proteins. Plasma homocysteine is a predictor of alcohol withdrawal seizures. Neuron 1994; 13: 1345–1357. Neuroreport 2000; 11: 2749–2752. 27 Armstrong N, Sun Y, Chen GQ, Gouaux E. Structure of a glutamate 4 Tsai GE, Ragan P, Chang R, Chen S, Linnoila VM, Coyle JT. Increased receptor ligand-binding core in complex with kainate. Nature 1998; glutamatergic neurotransmission and oxidative stress after alcohol 395: 913–917. withdrawal. Am J Psychiatry 1998; 155: 726–732. 28 Schiffer HH, Swanson GT, Masliah E, Heinemann SF. Unequal 5 Spanagel R, Pendyala G, Abarca C, Zghoul T, Sanchis-Segura C, expression of allelic kainate receptor GluR7 mRNAs in human brains. J Magnone MC et al. The clock gene Period2 influences the glutamatergic Neurosci 2000; 20: 9025–9033. system and thereby modulates alcohol consumption. Nat Med 2005; 29 Lappalainen J, Kranzler HR, Petrakis I, Somberg LK, Page G, Krystal JH et 11: 35–42. al. Confirmation and fine mapping of the chromosome 1 alcohol 6 Hollmann M, Heinemann S. Cloned glutamate receptors. Annu Rev dependence risk locus. Mol Psychiatry 2004; 9: 312–319. Neurosci 1994; 17: 31–108. 30 Begni S, Popoli M, Moraschi S, Bignotti S, Tura GB, Gennarelli M. 7 Liu QS, Xu Q, Arcuino G, Kang J, Nedergaard M. Astrocyte-mediated Association between the ionotropic glutamate receptor kainate 3 activation of neuronal kainate receptors. Proc Natl Acad Sci 2004; 101: (GRIK3) ser310ala polymorphism and schizophrenia. Mol Psychiatry 3172–3177. 2002; 7: 416–418. 8 Huettner JE. Kainate receptors and synaptic transmission. Prog Neurobiol 31 Schmidt LG, Harms H, Kuhn S, Rommelspacher H, Sander T. 2003; 70: 387–407. Modification of alcohol withdrawal by the A9 allele of the dopamine 9 Castillo PE, Malenka RC, Nicoll RA. Kainate receptors mediate a slow transporter gene. Am J Psychiatry 1998; 155: 474–478. postsynaptic current in hippocampal CA3 neurons. Nature 1997; 388: 32 Sander T, Harms H, Podschus J, Finckh U, Nickel B, Rolfs A et al. Allelic 182–186. association of a dopamine transporter gene polymorphism in alcohol 10 Kamiya H, Ozawa S. Kainate receptor-mediated inhibition of presynap- dependence with withdrawal seizures or delirium. Biol Psychiatry 1997; tic Ca2+ influx and EPSP in area CA1 of the rat hippocampus. J Physiol 41: 299–304. (London) 1998; 509: 833–845. 33 Gorwood P, Limosin F, Batel P, Hamon M, Ades J, Boni C. The A9 allele 11 Vignes M, Clarke VR, Parry MJ, Bleakman D, Lodge D, Ornstein Pl et al. of the dopamine transporter gene is associated with delirium tremens The GluR5 subtype of kainate receptor regulates excitatory synaptic and alcohol-withdrawal seizure. Biol Psychiatry 2003; 53: 85–92. transmission in areas CA1 and CA3 of the rat hippocampus. 34 Wernicke C, Smolka M, Gallinat J, Winterer G, Schmidt LG, Rommel- Neuropharmacology 1998; 37: 1269–1277. spacher H. Evidence for the importance of the human dopamine 12 Frerking M, Schmitz D, Zhou Q, Johansen J, Nicoll RA. Kainate receptors transporter gene for withdrawal symptomatology of alcoholics in a depress excitatory synaptic transmission at CA3-CA1 synapses in the German population. Neurosci Lett 2002; 333: 45–48. hippocampus via a direct presynaptic action. J Neurosci 2001; 21: 2958– 35 Franke P, Schwab SG, Knapp M, Gansicke M, Delmo C, Zill P et al. DAT1 2966. gene polymorphism in alcoholism: a family-based association study. Biol 13 Kullmann DM. Presynaptic kainate receptors in the hippocampus: Psychiatry 1999; 45: 652–654. slowly emerging from obscurity. Neuron 2001; 32: 561–564. 36 Buck KJ, Finn D. Genetic factors in addiction: QTL mapping and 14 Jiang L, Xu J, Nedergaard M, Kang J. A kainate receptor increases the candidate gene studies implicate GABAergic genes in alcohol and efficacy of GABAergic synapses. Neuron 2001; 30: 503–513. barbiturate withdrawal in mice. Addiction 2001; 96: 139–149. 15 Carta M, Ariwodola OJ, Weiner JL, Valenzuela CF. Alcohol potently 37 Klein RL, Mascia MP, Whiting PJ, Harris RA. GABAA receptor function inhibits the kainate receptor-dependent excitatory drive of hippocam- and binding in stably transfected cells: chronic ethanol treatment. pal interneurons. Proc Natl Acad Sci 2003; 100: 6813–6818. Alcohol Clin Exp Res 1995; 19: 1338–1344. 16 Dildy-Mayfield JE, Harris RA. Acute and chronic ethanol exposure alters 38 Faingold C, Li Y, Evans MS. Decreased GABA and increased glutamate the function of hippocampal kainate receptors expressed in Xenopus receptor-mediated activity on inferior colliculus neurons in vitro are oocytes. J Neurochem 1992; 58: 1569–1572. associated with susceptibility to ethanol withdrawal seizures. Brain Res 17 Carta M, Olivera DS, Dettmer TS, Valenzuela CF. Ethanol withdrawal 2000; 868: 287–295. upregulates kainate receptors in cultured rat hippocampal neurons. 39 Wojnar M, Bizon Z, Wasilewski D. Assessment of the role of kindling in Neurosci Lett 2002; 327: 128–132. the pathogenesis of alcohol withdrawal seizures and delirium tremens. 18 Dietrich D, Beck H, Kral T, Clusmann H, Elger CE, Schramm J. Alcohol Clin Exp Res 1999; 23: 204–208. Metabotropic glutamate receptors modulate synaptic transmission in 40 Olney JW. Excitotoxic amino acids and neuropsychiatric disorders. Annu the perforant path: pharmacology and localization of two distinct Rev Pharmacol Toxicol 1990; 30: 47–71. receptors. Brain Res 1997; 767: 220–227. 41 Crowder TL, Ariwodola OJ, Weiner JL. Ethanol antagonizes kainate 19 Puranam RS, Eubanks JH, Heinemann SF, McNamara JO. Chromosomal receptor-mediated inhibition of evoked GABAA inhibitory postsynaptic localization of gene for human glutamate receptor subunit-7. Somat Cell currents in the rat hippocampal CA1 region. J Pharmacol Exp Ther 2002; Mol Genet 1993; 19: 581–588. 303: 937–944. 20 Reich T, Edenberg HJ, Goate A, Williams JT, Rice JP, Van Eerdewegh P et 42 Weiner JL, Dunwiddie TV, Valenzuela CF. Ethanol inhibition of al. Genome-wide search for genes affecting the risk for alcohol synaptically evoked kainate responses in rat hippocampal CA3 dependence. Am J Med Genet 1998; 81: 207–215. pyramidal neurons. Mol Pharmacol 1999; 56: 85–90. 21 Foroud T, Edenberg HJ, Goate A, Rice J, Flury L, Koller DL et al. 43 Little HJ. The contribution of electrophysiology to knowledge of the Alcoholism susceptibility loci: confirmation studies in a replicate sample acute and chronic effects of ethanol. Pharmacol Ther 1999; 84: 333– and further mapping. Alcohol Clin Exp Res 2000; 24: 933–945. 353. 22 Dick DM, Nurnberger Jr J, Edenberg HJ, Goate A, Crowe R, Rice J et al. 44 Heinz A, Schafer M, Higley JD, Krystal JH, Goldman D. Neurobiological Suggestive linkage on chromosome 1 for a quantitative alcohol-related correlates of the disposition and maintenance of alcoholism. Pharma- phenotype. Alcohol Clin Exp Res 2002; 10: 1453–1460. copsychiatry 2003; 36(Suppl 3): 255–258. 23 Schiffer HH, Swanson GT, Heinemann SF. Rat GluR7 and a carboxy- 45 Becker HC. Kindling in alcohol withdrawal. Alcohol Res Health 1998; 22: terminal splice variant, GluR7b, are functional kainate receptor subunits 25–33. with a low sensitivity to glutamate. Neuron 1997; 19: 1141–1146. 46 Brown ME, Anton RF, Malcolm R, Ballenger JC. Alcohol detoxification 24 Nutt SL, Hoo KH, Rampersad V, Deverill RM, Elliott CE, Fletcher EJ et al. and withdrawal seizures: clinical support for a kindling hypothesis. Biol Molecular characterization of the human EAA5 (GluR7) receptor: a Psychiatry 1988; 23: 507–514. high-affinity kainate receptor with novel potential RNA editing sites. 47 Booth BM, Blow FC. The kindling hypothesis: further evidence from a Receptors Channels 1994; 2: 315–326. US national study of alcoholic men. Alcohol Alcohol 1993; 28: 593–598. 25 Krupp JJ, Vissel B, Heinemann SF, Westbrook GL. N-terminal domains in 48 Preuss UW, Soyka M, Bahlmann M, Wenzel K, Behrens S, de Jonge S et the NR2 subunit control desensitization of NMDA receptors. Neuron al. Serotonin transporter gene regulatory region polymorphism (5- 1998; 20: 317–327. HTTLPR), [3H]paroxetine binding in healthy control subjects and 26 Stern-Bach Y, Bettler B, Hartley M, Sheppard PO, O’Hara PJ, Heinemann alcohol-dependent patients and their relationships to impulsivity. SF. Agonist selectivity of glutamate receptors is specified by two Psychiatry Res 2000; 96: 51–61.

The Pharmacogenomics Journal Kainate receptor functional polymorphism and delirium tremens UW Preuss et al 41

49 Preuss UW, Koller G, Soyka M, Bondy B. Association between suicide 53 Wittchen HU, Zaudig M, Fydrich T. SKID. Strukturiertes Klinisches attempts and 5-HTTLPR-S-allele in alcohol-dependent and control Interview fu¨r DSM-IV. Achse I und II. Hogrefe: Go¨ttingen, 1997. subjects: further evidence from a German alcohol-dependent inpatient 54 Babor TF, Hofmann M, DelBoca FK, Hesselbrock V, Meyer RE, Dolinsky sample. Biol Psychiatry 2001; 50: 636–639. ZS et al. Types of alcoholics, I. Evidence for an empirically derived 50 Soyka M, Preuss UW, Koller G, Zill P, Bondy B. Dopamine D 4 receptor typology based on indicators of vulnerability and severity. Arch Gen gene polymorphism and extraversion revisited: results from the Munich Psychiatry 1992; 49: 599–608. gene bank project for alcoholism. J Psychiatr Res 2002; 36: 429–435. 55 Schuckit MA, Tipp JE, Smith TL, Shapiro E, Hesselbrock VM, Bucholz KK 51 Bucholz KK, Cadoret R, Cloninger CR, Dinwiddie SH, Hesselbrock VM, et al. An evaluation of type A and B alcoholics. Addiction 1995; 90: Nurnberger Jr JI et al. A new, semi-structured psychiatric interview for 1189–1203. use in genetic linkage studies: a report on the reliability of the SSAGA. J 56 Endler NS, Parker JD, Butcher JN. A factor analytic study of coping Stud Alcohol 1994; 55: 149–158. styles and the MMPI-2 content scales. J Clin Psychol 1989; 49: 52 Hesselbrock M, Easton C, Bucholz KK, Schuckit M, Hesselbrock V. A 523–527. validity study of the SSAGA – a comparison with the SCAN. Addiction 57 Armitage P. Tests for linear trends in proportions and frequencies. 1999; 94: 1361–1370. Biometrics 1955; 11: 375–386.

The Pharmacogenomics Journal