Molecular Psychiatry (2004) 9, 494–499 & 2004 Nature Publishing Group All rights reserved 1359-4184/04 $25.00 www.nature.com/mp ORIGINAL RESEARCH ARTICLE , ionotropic, N-methyl D-aspartate 2A (GRIN2A) as a positional candidate for attention- deficit/hyperactivity disorder in the 16p13 region J Adams1, J Crosbie2, K Wigg1, A Ickowicz2, T Pathare2, W Roberts3, M Malone3, R Schachar2, R Tannock2, JL Kennedy4 and CL Barr1,2 1Cell and Molecular Biology Division, Toronto Western Research Institute, University Health Network, Toronto, ON, Canada; 2Department of Psychiatry, Brain and Behaviour Programme, The Hospital for Sick Children, Toronto, ON, Canada; 3Division of Neurology, Brain and Behaviour Programme, The Hospital for Sick Children, Toronto, ON, Canada; 4Neurogenetics Section, Centre for Addiction and Mental Health, Department of Psychiatry, University of Toronto, Toronto, ON, Canada

Keywords: ADHD; genetics; glutamate; GRIN2A; NMDA; A second executive control process, working memory, linkage the active maintenance and manipulation of informa- tion necessary to guide decision-making and beha- The glutamate system may be involved in the develop- 3,8 ment of attention-deficit/hyperactivity disorder (ADHD) vioral responses, may also be related to ADHD. based on animal models and the role of N-methyl-D- On the basis of animal models, pharmacological aspartate receptors (NMDAR) in cognition and motor interventions and neuroimaging studies, ADHD and processes. A follow-up study of the first genome scan its associated cognitive deficits are believed to result for ADHD identified significant evidence for linkage to from dysregulation of neurotransmitter systems. The the 16p13 region.1 The glutamate receptor, ionotropic, dopamine system has been the primary focus in N-methyl D-aspartate 2A (GRIN2A) gene that encodes studies of ADHD to date; however, given that the the 2A subunit of the NMDA receptor, resides in this glutamatergic system is involved in both cognitive region and a recent study has reported an association 2 and motor function, and modulates the dopamine and between this gene and ADHD. We tested for linkage 9 between the alleles and haplotypes of four polymorph- serotonin systems, dysregulation in this system isms at the GRIN2A and ADHD in our sample of might also underlie the development of ADHD. 183 nuclear families with 229 affected children. In The action of glutamate is mediated by several contrast to previous findings, we did not identify any families of receptors, one of which is the N-methyl-D- evidence for a relationship of these markers and ADHD. aspartate receptor (NMDAR) family. NMDARs are Owing to the role of GRIN2A in aspects of cognition, we heteromeric complexes differentially expressed investigated the relationship of this gene to the throughout the central nervous system and develop- cognitive phenotypes of inhibitory control, verbal ment. Several distinct subtypes of NMDAR have been short-term memory and verbal working memory. There identified. Each NMDAR subtype is composed of at was no significant evidence of linkage between GRIN2A least one obligatory subunit, termed NR1, and one or and these phenotypes. While the results were not significant in our sample, the previous association more of the four NR2 subunits (2A, 2B, 2C, 2D). While finding suggests that further study of this gene is the specific biological and pharmacological proper- warranted. ties differ for each NMDAR subtype, it has been well Molecular Psychiatry (2004) 9, 494–499. established that NMDARs have critical roles in doi:10.1038/sj.mp.4001455 excitatory synaptic transmission, neuronal survival Published online 23 December 2004 and plasticity, and in cognitive processes and motor functioning. Attention-deficit/hyperactivity disorder (ADHD) is a Numerous studies have investigated the role of prevalent psychiatric disorder among children, char- NMDARs in cognition. Studies of long-term potentia- acterized by inappropriate levels of inattention, over- tion, the phenomenon thought to underlie long-term activity and impulsivity. Substantial evidence memory, have shown that NMDARs play an essential suggests that, to a large degree, cognitive deficits, role in both the summation of cellular responses and particularly deficits in executive control, underlie the induction of neural plasticity. Also, studies in rats etiology of ADHD.3,4 Executive control functions in have identified hippocampal NMDARs as being the preparation, maintenance, delaying and switching involved in multiple aspects of spatial short-term of action and thoughts. Although the precise cogni- memory.10 Antagonism of NMDARs in non-human tive characteristics of ADHD are debated, the execu- primates causes impairments in short-term memory tive control process inhibition has been identified as a performance.11 Finally, in humans, administration of potential core deficit in ADHD.5,6 Inhibition is ketamine, an NMDAR antagonist, decreases declara- defined as the ability to suddenly and completely tive memory performance,12 as well as both verbal stop a planned or ongoing thought and action.7 and nonverbal working memory performance.13,14 GRIN2A as a positional candidate for ADHD in the 16p13 region J Adams et al 495 These observations demonstrate that NMDARs are involved in multiple cognitive processes, and suggest that NMDAR functioning is related to both short-term and working memory. Collective evidence from a number of studies also supports a role for the glutamate system in the Anneal. temp. regulation of motor function. Hyperactivity can be induced by pharmacological disruption of NMDAR glutamatergic transmission.15,16 Also, infusion of c cccag 58 t cttgattt 57 tagatcca 60 NMDA into the ventral hippocampus to stimulate g NMDARs was shown to cause dose-dependant short- g

17 sequence term hyperactivity. Several strains of transgenic Probe FAM mice carrying mutations in, or knockouts of encoding NMDAR subunits exhibit changes in motor and cognitive function. Mice lacking the NMDAR 2A subunit gene GluRe1, the mouse homolog of the c cccag tgccaggtt c tagatcca ctcagg GRIN2A gene in humans, show increased sponta- cttgatttc aagtgac a neous locomotor activity in a novel environment, a impaired latent learning associated with selective sequence attention9 and deficits in spatial learning.18 Recently, Fisher et al19 reported their findings from the first genomewide scan for loci linked to ADHD. Although no regions reached genomewide signifi- cance in this initial study, regions on 5p12, 10q26, 12q23 and 16p13 had LOD scores greater than 1.5. The significance of the linkage finding in the 16p13 region was increased by genotyping additional families and markers. Results from this analysis

yielded a multi-point maximum LOD score of 4.2 Reverse primer Probe VIC near marker D16S3114.1 The GRIN2A gene maps within 2 MB of the 1-LOD support interval surround- ing this maximum. Support for GRIN2A as the susceptibility gene in this region was obtained recently in a family-based study in which a signifi- cant association between ADHD and the Grin2a-5 polymorphism was observed.2 b ct catattgttactatcgatgatccat N/A N/A 58

In the present study, we investigated the relation- g ship of the GRIN2A gene to ADHD and cognitive measures using the transmission disequilibrium test (TDT) to examine the inheritance patterns of four GRIN2A polymorphisms in a sample of 183 small nuclear families ascertained through an ADHD pro- Forward primer band. The markers analysed in this study include the Grin2a-5 polymorphism, as well as three additional single-nucleotide polymorphisms (SNPs) dispersed over the length of the gene (Table 1). The allele frequencies for the Grin2a-5 polymorph- a ism reported for the Turic sample (allele 1 ¼ 0.69, allele 2 ¼ 0.31) and this sample (allele 1 ¼ 0.72, allele 2 ¼ 0.28) are similar (Table 1). TDT results for the four markers individually are shown in Table 2. No evidence for biased transmission of alleles at any of the markers was observed. Strong linkage disequili- brium (LD) was observed among polymorphisms Grin2a-605, Grin2a-5 and Grin2a-531, located in intron 5, exon 6 and exon 14, respectively. LD was not observed between the intron 3 polymorphism Grin2a-503, and any of the three downstream mar- Genotyping conditions kers, indicating that LD does not extend across the entire length of the gene. TDT analysis of haplotypes, Bolded nucleotides show the mutagenic site in the primer. Based on NCBI genomic contig NT_010393. Bolded nucleotides are the polymorphic nucleotides examined. Grin2a-531 1014531 Exon 14, UTR aattgcttgtctcactaaact ctcaaatcaaaagccttcaa ctcagg Table 1 Probe name rs# Location a b c Grin2a-605Grin2a-5 727605 N/A Intron 5 Exon 6 gcatagaaacttgtttgtgcatacc cgtggaagacatagaccc ctgatctaaagagtatcccagaaaacc agtgac using the TRANSMIT program, showed no significant Grin2a-503 1070503 Intron 3 ctgaaaatagagattctttaaacaaatacc tgccaaggattaataactcagtgatag tgccaggtt

Molecular Psychiatry GRIN2A as a positional candidate for ADHD in the 16p13 region J Adams et al 496 Table 2 ETDT analysis of the GRIN2A markers

Polymorphism Allele Allele freq. Transmissions Nontransmissions w2 P-valuea

Grin2a-503 C(1) 0.86 44 39 0.30 0.58 T(2) 0.14 39 44 Grin2a-605 G(1) 0.70 77 78 0.01 0.94 A(2) 0.30 78 77 Grin2a-5 G(1) 0.72 72 68 0.11 0.74 A(2) 0.28 68 72 Grin2a-531 G(1) 0.62 81 93 0.83 0.36 A(2) 0.38 93 81

aOne degree of freedom.

Table 3 GRIN2A haplotype frequencies and TRANSMIT analysis

Marker alleles Haplotype frequencya Transmissions

Haplotype 503 605 5 531 Obs.b Exp.c w2 P-valued

1 1 1 1 1 0.47 195.22 202.53 1.15 0.28 2 1 1 1 2 0.14 64.13 57.41 2.07 0.15 3 1 2 2 2 0.19 81.09 79.29 0.11 0.74

Global w2 test for haplotypes with frequencies greater than 0.10 ¼ 2.57, 3 df, P ¼ 0.46. aOnly haplotypes with a frequency greater than 0.10 were used in the analysis. bTest statistic representing the observed number of transmissions. cExpected value of the test statistic under the null hypothesis of no linkage or association. dOne degree of freedom.

evidence for biased transmission of haplotypes (Table the chance of a spurious finding, and their relatively 3). There was no significant evidence for a relation- large sample of 238 nuclear families should be fairly ship between the transmission of alleles and scores of robust to false positives. Notably, we also used a the three cognitive phenotypes inhibitory control, family-based control design, and a similar sample size verbal short-term memory and verbal working mem- of 183 families with a total of 229 affected children, ory (Table 4). and therefore should have sufficient power to detect The role of NMDARs in cognition and motor linkage if the polymorphism is of similar effect in activity, and the location of GRIN2A in the 16p13 both samples. linkage region identified in an ADHD genome scan, ADHD is a clinically heterogeneous disorder made GRIN2A a strong candidate gene for considera- that has many etiologies, genetic and nongenetic, tion as an ADHD susceptibility locus. Support for this which may be differentially represented in different hypothesis has been shown by the previous associa- samples and in turn influence the strength of a tion of alleles of the Grin2a-5 polymorphism and specific genetic effect. Two possible sources could ADHD.2 In contrast to those results, we found no cause this divergence: (i) the population each sample significant evidence of linkage in our sample. is drawn from differs in these representations and/or It has been observed that strong associations (ii) the manner in which the sample is collected apparent in initial genetic association studies often leads to different representations. The ethnic compo- become less prominent as more data are collected. sition of the population and/or sample is included in Three possible reasons could explain this trend: these sources. In contrast to the British Caucasian (i) the results of the first study could be a spurious sample used by Turic et al,21 most of the families in finding (ie false positive), (ii) the results of a our study are of mixed European Caucasian descent. subsequent study could be false negative, (iii) the In addition to issues related to ethnicity, differences genetic effect of the polymorphism(s) tested is between the samples due to participant ascertainment stronger in some samples than in others.20 Turic and strategies, selection criteria such as age, gender or colleagues’ finding of an association between GRIN2A absence of comorbidities, and/or the source of and ADHD could be spurious. However, their use of a diagnostic information may have influenced the family-based study design would reduce the possibi- sample composition and the strength of a specific lity of population stratification, and thereby reduce genetic effect.

Molecular Psychiatry GRIN2A as a positional candidate for ADHD in the 16p13 region J Adams et al 497 Table 4 FBAT analysis of GRIN2A alleles transmission in relation to inhibitory control, verbal short-term memory and verbal working memory

Task Marker Allele Na Sb Ec ZP-valued

Stop-task Grin2a-503 1 27 11981.06 11497.28 0.63 0.53 2 4357.38 4841.17 À0.63 Grin2a-605 1 43 15576.74 16298.11 À0.72 0.47 2 10659.58 9938.21 0.72 Grin2a-5 1 40 14528.03 14958.82 À0.45 0.65 2 10153.67 9722.89 0.45 Grin2a-531 1 50 16915.16 17018.89 À0.10 0.92 2 14033.98 13930.26 0.10

Digit span forwards Grin2a-503 1 36 577.00 568.00 0.27 0.79 2 221.00 230.00 À0.27 Grin2a-605 1 59 804.00 800.50 0.08 0.94 2 484.00 487.50 À0.08 Grin2a-5 1 56 850.00 824.50 0.57 0.57 2 440.00 465.50 À0.57 Grin2a-531 1 69 766.00 838.00 À1.47 0.14 2 754.00 682.00 1.47

Digit span backwards Grin2a-503 1 36 577.00 576.00 0.03 0.98 2 231.00 232.00 À0.03 Grin2a-605 1 59 800.00 819.50 À0.42 0.68 2 524.00 504.50 0.42 Grin2a-5 1 56 824.00 811.50 0.28 0.78 2 462.00 474.50 À0.28 Grin2a-531 1 69 792.00 845.50 À1.06 0.29 2 766.00 712.50 1.06 aNumber of informative families. bTest statistic representing the observed number of transmissions, weighted to the phenotype measure. cExpected value of S under the null hypothesis of no linkage or association. dTwo-tailed.

To date, a functional consequence of the Grin2a-5 Materials and methods polymorphism has not been detected and, therefore, Participants we must consider that the association finding from Participant assessment and diagnostic criteria have Turic et al is likely to be due to LD between the 22 Grin2a-5 polymorphism and a functional DNA var- been described previously. Probands and affected iant(s). Although the frequency of the Grin2a-5 siblings between the ages of 7 and 16 years were polymorphism is similar in the two samples, the recruited from the Child Development and Neurop- frequency of the functional polymorphism may not sychiatry Clinics at the Hospital for Sick Children, be. Therefore, an association of GRIN2A and ADHD and met the Diagnostic and Statistical Manual of may be differentially detected, depending on sample Mental Disorder, 4th Edition (DSM-IV) criteria for composition and the amount of LD between the ADHD. All children were free of medication for a marker and the functional variant. minimum of 24 h before assessment and cognitive In addition to testing for linkage between GRIN2A testing. This protocol was approved by the Hospital and ADHD, we also tested for relationships between for Sick Children’s Research Ethics Board and GRIN2A and three cognitive phenotypes, inhibitory informed written consent was obtained for all control, verbal short-term memory and verbal working participants. In total, 183 nuclear families including memory. We found that variation in GRIN2A was not 46 siblings of the probands were genotyped for this related to measures of these cognitive phenotypes. study. Although NMDARs are thought to play pivotal roles in multiple memory processes including working Cognitive phenotype measures memory, their involvement in specific aspects and/ Inhibitory control was measured using the stop task or tasks remains unclear. Therefore, our analysis of paradigm described in greater detail in Schachar these specific cognitive phenotypes does not rule out et al.6 The stop task is a laboratory paradigm that the genetic contribution of GRIN2A to other cognitive provides a direct measure of the speed with which phenotypes. one can execute and voluntarily inhibit a speeded

Molecular Psychiatry GRIN2A as a positional candidate for ADHD in the 16p13 region J Adams et al 498 motor response. The stop task involves two concur- Genotypes for the Grin2a-5 polymorphism were rent tasks. The go task is a simple choice reaction time determined by restriction enzyme digestion of PCR task that individuals are asked to perform as quickly products. A product of 154 bp was amplified using and as accurately as possible. The stop task involves a the primer sequences listed in Table 1. The forward tone emitted from the computer, which signals primer is a mutagenic primer which converts the ‘C’ participants to withhold their response on that trial. located three nucleotides upstream of the SNP to a ‘G’ Tones occur randomly and on 25% of trials. We used resulting in an MspA1I site in the PCR product a ‘tracking’ procedure that converges on the delay derived from the ‘G’ variant (Allele 1) of this SNP. The between the go and stop signals at which individuals PCR reactions (20 ml volume) contained 50 ng of

are able to stop 50% of the time. We can estimate the genomic DNA and 1.5 mM of MgCl2. The thermal latency of the unobserved stop process, stop signal cycling conditions were as follows: 941C for 1 min, reaction time (SSRT), by subtracting the mean delay and then 35 cycles of 941C for 30 s, 581C for 30 s and at which the subject inhibits 50% of the time from 721C for 30 s, and then 10 min at 721C. Amplification mean go reaction time (see Logan7 and Logan and product (8 ml) was digested with 5 U of MspA1I (New Cowan23). SSRT was corrected for age based on England Biolabs, Beverly, MA, USA) at 371C for 1.5– population-based norms.24 2 h. Restriction fragments of 154 bp (site absent: Allele Verbal short-term and working memory was as- 2) and/or 131 plus 23 bp (site present: Allele 1) were sessed using the digit span subtest of the Wechsler resolved on agarose gels consisting of 2% agarose plus Intelligence Scale for Children – Third Edition 2% NuSieve 3 : 1 agarose (Mandel Scientific Com- (WISC-III). This test provides an overall memory pany, Inc., Guelph, Ontario, Canada) and visualized score, which is then divided into two parts, digits with ethidium bromide staining. forward (DF) and digits backward (DB). DF provides a measure of verbal short-term memory span. The Statistical analysis experimenter speaks aloud a sequence of digits (at For the categorical analysis of ADHD, the extended the rate of one per second) and the child is asked to 28 repeat them in the order in which they were TDT program (ETDT) was used to test for biased presented. The dependant measure is the number of transmission of individual marker alleles. Transmis- sion of haplotypes was analyzed using the program trials in which the lists are recalled in the correct 29 order. The DB task is a measure of verbal working TRANSMIT. Only haplotypes with a frequency memory. The presenter reads a series of numbers to above 0.10 were included in the analysis. The two- the child. The child is then required to repeat the locus linkage disequilibrium program (2LD; http:// numbers in a reverse order. For both the forwards and www.iop.kcl.ac.uk/loP/Departments/PsychMed/GE- piBSt/software.shtml) was used to calculate the the backwards components, age-normed standardized 0 scores were used.25 coefficient of LD, D between marker alleles in the parental . The FBAT program30 was Genotyping used for the analysis of quantitative measures. Four SNPs were genotyped for this study: Grin2a-503, Grin2a-605, Grin2a-5 and Grin2a-531 located in intron 3, intron 5, exon 6 and exon 14, respectively. Acknowledgements Previously, the Grin2a-5 and Grin2a-531 polymorph- This work was supported by grants from the Natural isms were reported to be located in exon 5 and exon 2,26 0 Sciences and Engineering Research Council (JHA), 13, respectively ; however, the recent finding of a 5 The Hospital for Sick Children Psychiatric Endow- untranslated exon (see NCBI genomic contig 27 ment Fund (CLB), the Canadian Institutes of Health NT_010393 and Itokawa et al ) increases the pre- Research MT14336 (CLB) and MOP-14336 (CLB). vious numbering of all exons and introns by one. Also, we thank Drs Anita Thapar and Michael The SNPs designated Grin2a-503, Grin2a-605 and O’Donovan for providing information on the Grin2a- Grin2a-531 were genotyped with the ABI 7900-HT 5 polymorphism prior to publication. Sequence Detection System (Applied Biosystems, Foster City, CA, USA) using the TaqMan 50 nuclease assay for allelic discrimination. Primer and probe References sequences, and annealing temperatures used, are listed in Table 1. The PCR reactions (10 ml volume), 1 Smalley SL, Kustanovich V, Minassian SL, Stone JL, Ogdie MN, contained 30 ng of genomic DNA, 10 mmol of Taq- McGough JJ et al. Genetic linkage of attention-deficit/hyperactivity Mans Universal PCR Master Mix (Applied Biosys- disorder on 16p13, in a region implicated in autism. Am J Hum Genet 2002; 71: 959–963. tems) and 0.25 ml of allelic discrimination mix 2 Turic D, Langley K, Mills S, Stephens M, Lawson D, Govan C et al. (Applied Biosystems) containing 36 mM of each primer Follow-up of genetic linkage findings on chromosome 16p13: and 8 mM of each probe. The thermal cycling condi- evidence of association of N-methyl-Daspartate glutamate receptor tions were 501C for 2 min, 951C for 10 min and then 40 2A gene polymorphism with ADHD. Mol Psychiatry, (in press). 1 3 Pennington BF, Ozonoff S. Executive functions and developmental cycles of 92 C for 15 s, and the annealing temperature psychopathology. J Child Psychol Psychiatry 1996; 37: 51–87. for 1 min. Each 96-well plate contained two negative 4 Quay HC. Inhibition and attention deficit hyperactivity disorder. controls. J Abnorm Child Psychol 1997; 25: 7–13.

Molecular Psychiatry GRIN2A as a positional candidate for ADHD in the 16p13 region J Adams et al 499 5 Schachar R, Logan GD. Impulsivity and inhibitory control in 19 Fisher SE, Francks C, McCracken JT, McGough JJ, Marlow AJ, normal development and childhood psychopathology. Dev Psy- MacPhie IL et al. A genomewide scan for loci involved in chol 1990; 26: 710–720. attention-deficit/hyperactivity disorder. Am J Hum Genet 2002; 6 Schachar R, Mota VL, Logan GD, Tannock R, Klim P. Confirmation 70: 1183–1196. of an inhibitory control deficit in attention-deficit/hyperactivity 20 Ioannidis JP, Ntzani EE, Trikalinos TA, Contopoulos-Ioannidis DG. disorder. J Abnorm Child Psychol 2000; 28: 227–235. Replication validity of genetic association studies. Nat Genet 7 Logan GD. On the ability to inhibit thought and action: a users’ 2001; 29: 306–309. guide to the stop signal paradigm. In: Carr DD, Carr TH (eds) 21 Holmes J, Payton A, Barrett JH, Hever T, Fitzpatrick H, Trumper Inhibitory Processes in Attention, Memory, and Language. Aca- AL et al. A family-based and case–control association study of the demic Press: San Diego, CA, 1994 pp 189–239. dopamine D4 receptor gene and dopamine transporter gene in 8 Barnett R, Maruff P, Vance A, Luk ES, Costin J, Wood C et al. attention deficit hyperactivity disorder. Mol Psychiatry 2000; 5: Abnormal executive function in attention deficit hyperactivity 523–530. disorder: the effect of stimulant medication and age on spatial 22 Barr CL, Wigg K, Malone M, Schachar R, Tannock R, Roberts W working memory. Psychol Med 2001; 31: 1107–1115. et al. Linkage study of catechol-O-methyltransferase and attention- 9 Miyamoto Y, Yamada K, Noda Y, Mori H, Mishina M, Nabeshima deficit hyperactivity disorder. Am J Med Genet 1999; 88: T. Hyperfunction of dopaminergic and serotonergic neuronal 710–713. systems in mice lacking the NMDA receptor epsilon1 subunit. 23 Logan GD, Cowan WB. On the ability to inhibit thought and J Neurosci 2001; 21: 750–757. action: a theory of an act of control. Psychol Rev 1984; 91: 10 Lee I, Kesner RP. Differential contribution of NMDA receptors in 295–327. hippocampal subregions to spatial working memory. Nat Neurosci 24 Williams BR, Ponesse JS, Schachar RJ, Logan GD, Tannock R. 2002; 5: 162–168. Development of inhibitory control across the life span. Dev 11 Taffe MA, Davis SA, Gutierrez T, Gold LH. Ketamine impairs Psychol 1999; 35: 205–213. multiple cognitive domains in rhesus monkeys. Drug Alcohol 25 Kaplan E, Fein D, Kramer J, Delis D, Morris R. WISC-III PI Manual Depend 2002; 68: 175–187. The Psychological Corporation: San Antonio, TX, 1999. 12 Newcomer JW, Farber NB, Jevtovic-Todorovic V, Selke G, Melson 26 Williams NM, Bowen T, Spurlock G, Norton N, Williams HJ, AK, Hershey T et al. Ketamine-induced NMDA receptor hypofunc- Hoogendoorn B et al. Determination of the genomic structure and tion as a model of memory impairment and psychosis. Neurop- mutation screening in schizophrenic individuals for five subunits sychopharmacology 1999; 20: 106–118. of the N-methyl-D-aspartate glutamate receptor. Mol Psychiatry 13 Grunwald T, Beck H, Lehnertz K, Blumcke I, Pezer N, Kurthen M 2002; 7: 508–514. et al. Evidence relating human verbal memory to hippocampal 27 Itokawa M, Yamada K, Yoshitsugu K, Toyota T, Suga T, Ohba H N-methyl-D-aspartate receptors. Proc Natl Acad Sci USA 1999; 96: et al. A microsatellite repeat in the promoter of the N-methyl-D- 12085–12089. aspartate receptor 2A subunit (GRIN2A) gene suppresses tran- 14 Adler CM, Goldberg TE, Malhotra AK, Pickar D, Breier A. Effects scriptional activity and correlates with chronic outcome in of ketamine on thought disorder, working memory, and semantic schizophrenia. Pharmacogenetics 2003; 13: 271–278. memory in healthy volunteers. Biol Psychiatry 1998; 43: 811–816. 28 Sham PC, Curtis D. An extended transmission/disequilibrium test 15 Carlsson M, Carlsson A. The NMDA antagonist MK-801 causes (TDT) for multi-allele marker loci. Ann Hum Genet 1995; 59(Part marked locomotor stimulation in monoamine-depleted mice. 3): 323–336. J Neural Transm 1989; 75: 221–226. 29 Clayton D. A generalization of the transmission/disequilibrium 16 Corbett R, Camacho F, Woods AT, Kerman LL, Fishkin RJ, Brooks K test for uncertain-haplotype transmission. Am J Hum Genet 1999; et al. Antipsychotic agents antagonize non-competitive N-methyl- 65: 1170–1177. D-aspartate antagonist-induced behaviors. Psychopharmacology 30 Horvath S, Xu X, Laird NM. The family based association test (Berl) 1995; 120: 67–74. method: strategies for studying general genotype–phenotype 17 Zhang WN, Bast T, Feldon J. Effects of hippocampal N-methyl-D- associations. Eur J Hum Genet 2001; 9: 301–306. aspartate infusion on locomotor activity and prepulse inhibition: differences between the dorsal and ventral hippocampus. Behav Neurosci 2002; 116: 72–84. Correspondence: CL Barr, Toronto Western Hospital, Main 18 Sakimura K, Kutsuwada T, Ito I, Manabe T, Takayama C, Kushiya E Pavilion, Rm. 14-302, 399 Bathurst St., Toronto, Ontario, Canada et al. Reduced hippocampal LTP and spatial learning in mice M5T 2S8. E-mail: [email protected] lacking NMDA receptor epsilon 1 subunit. Nature 1995; 373: Received 22 August 2003; revised 15 September 2003; accepted 151–155. 14 October 2003

Molecular Psychiatry