Cytogenetic and genetic evidence supports a role for the kainate-type glutamate receptor gene, GRIK4, in schizophrenia and bipolar disorder

Article metrics


In the search for the biological causes of schizophrenia and bipolar disorder, glutamate neurotransmission has emerged as one of a number of candidate processes and pathways where underlying gene deficits may be present. The analysis of chromosomal rearrangements in individuals diagnosed with neuropsychiatric disorders is an established route to candidate gene identification in both Mendelian and complex disorders. Here we describe a set of genes disrupted by, or proximal to, chromosomal breakpoints (2p12, 2q31.3, 2q21.2, 11q23.3 and 11q24.2) in a patient where chronic schizophrenia coexists with mild learning disability (US: mental retardation). Of these disrupted genes, the most promising candidate is a member of the kainate-type ionotropic glutamate receptor family, GRIK4 (KA1). A subsequent systematic case–control association study on GRIK4 assessed its contribution to psychiatric illness in the karyotypically normal population. This identified two discrete regions of disease risk within the GRIK4 locus: three single single nucleotide polymorphism (SNP) markers with a corresponding underlying haplotype associated with susceptibility to schizophrenia (P=0.0005, odds ratio (OR) of 1.453, 95% CI 1.182–1.787) and two single SNP markers and a haplotype associated with a protective effect against bipolar disorder (P=0.0002, OR of 0.624, 95% CI 0.485–0.802). After permutation analysis to correct for multiple testing, schizophrenia and bipolar disorder haplotypes remained significant (P=0.0430, s.e. 0.0064 and P=0.0190, s.e. 0.0043, respectively). We propose that these convergent cytogenetic and genetic findings provide molecular evidence for common aetiologies for different psychiatric conditions and further support the ‘glutamate hypothesis’ of psychotic illness.


Both de novo and familial chromosome abnormalities have been described in patients with psychiatric or neurological conditions.1, 2 A small proportion of these have been studied at the molecular level.3 We have previously described a large Scottish family in which there is a highly significant co-segregation (LOD>7)4 between major mental illness and the presence of a balanced translocation between chromosomes 1 and 11.5 The DISC1 gene is directly disrupted by the chromosome 1 breakpoint6, 7 and supportive evidence for its involvement with schizophrenia and bipolar disorder has come from linkage and case–control association studies on karyotypically normal populations8, 9, 10, 11, 12 and from our recent demonstration of its activity-dependent interaction with the protein encoded by an independently disrupted gene, PDE4B.13 Through the analysis of further cytogenetic abnormalities in schizophrenic patients, we and others have identified NPAS3, GRIA3 and DIBD1 as potential candidate genes.14, 15, 16, 17

Here we describe a patient with chronic schizophrenia and mild learning disability (US: mental retardation). This co-morbid state has been previously shown to be highly familial, associated with changes in brain structure on imaging typical of schizophrenia alone, and also associated with increased prevalences of chromosomal abnormalities.18, 19 The patient possesses a complex chromosomal rearrangement involving several chromosomes. We hypothesized that the patient's diagnosis was as a direct result of gene deficits caused by the cytogenetic abnormality. Therefore, we have positioned the constituent chromosomal breakpoints and identified a number of potential candidate genes in their vicinity. One of these, GRIK4, is directly disrupted at a breakpoint and encodes a member of the ionotropic glutamate receptor family – an established focus for candidate disease gene hunting because of the hypothesis that defects in glutamate neurotransmission may underlie schizophrenia.

As the most promising candidate, we wished to determine the relevance of GRIK4 to psychiatric illness beyond this unique patient. Therefore, a large case–control association study was undertaken, which demonstrated that single nucleotide polymorphism (SNP) markers and haplotypes within the GRIK4 locus are strongly associated with both schizophrenia and bipolar disorder.

Materials and methods

Clinical findings

A female patient was identified as part of a survey of patients with mild mental retardation (UK term – learning disability) and coexisting schizophrenia.18 There was no evidence for any previous karyotype investigation before this ascertainment. Full, informed consent for the current study was obtained from the subject.

In her mid 60s when examined, the patient gave a near 30-year history of severe chronic schizophrenia requiring repeated admissions to a hospital specializing in the care of people with mental retardation and concurrent psychiatric or behavioural disorders. Initially in mainstream schooling, from the age of 8 she received education in an establishment for those with special educational needs – she had a mild degree of intellectual impairment (IQ 65–70). Her early psychiatric history is unclear and, although psychosis was only formally documented from her early 40s, there is evidence of earlier onset. The schizophrenic illness showed the same features during all episodes with clear and persistent auditory hallucinations and persecutory delusions. Treatment was initially with standard typical anti-psychotic medications and, latterly, with atypical anti-psychotics owing to the onset of persistent extrapyramidal side effects. Between episodes resolution was good and she regained full insight. The diagnosis of chronic schizophrenia was confirmed using SADS-L structured interview,20, 21 to generate DSM-IV and ICD-10 criteria, by a psychiatrist experienced in both general psychiatry and the psychiatry of mental retardation (WM). Consensus diagnosis was reached on review by two psychiatrists (WM and DB). IQ scores were generated from WAIS-R and their stability shown by similar levels detected by psychological examination at different times throughout her life. No other family member agreed to participate in the study or to be karyotyped. Clear family history information is not available. However, as far as can be ascertained, there were no first-degree relatives known with schizophrenia or mental retardation.

The patient was not dysmorphic and was of average stature. During childhood, repeated mastoid infections and corrective surgery led to conduction deafness in one ear and partial conduction deafness in the other. Hearing-aid correction was excellent and meant that her communication skills were fully intact. Although there is an older literature suggesting a link between some cases of schizophrenia (especially late-onset paraphrenia) and chronic deafness, more recent analysis suggests that there is little evidence of any true association.22 During her 60s, the patient developed sero-negative rheumatoid arthritis, principally affecting her distal joints.

Cell culture

The patient was karyotyped using conventional Giemsa, G-band, staining. Lymphocytes were extracted from 7 ml of patient venous blood (for storage and generation of Epstein–Barr virus-transformed cell lines) using density gradient separation (Histopaque-1077, Sigma, Gillingham, UK). The resulting lymphoblastoid cell line failed to thrive, perhaps owing to the patient age and medication, such that limited cell material was produced. However, metaphase-arrested chromosomes for cytogenetic analysis were successfully prepared from 0.8 ml of patient blood cultured for 71 h in medium containing phytohaemagglutinin (Peripheral Blood Medium, Sigma). These short-term cultures were treated with colcemid for 1 h followed by a conventional cell membrane lysis and methanol/acetic acid fixing procedure. Fixed chromosomes were dropped onto microscope slides and stored for 1 week before use in fluorescence in situ hybridization (FISH) experiments.

Selection of YAC/BAC and cosmid clones for FISH probe synthesis

Clones were initially selected from the Whitehead/MIT map of the relevant chromosome in the cytogenetic intervals to which the G-band defined breakpoints were mapped. Yeast artificial chromosomes (YACs) were obtained from the HGMP Resource Centre, Babraham Bioincubator, Babraham, Cambridge, UK ( Clone DNA was prepared by standard methods and PCR amplified using primers designed against consensus sequence elements within the archetypal Alu repeat.23 This ‘Alu-PCR’ gives a representative spread of non-repetitive sequence over the full length of the YAC and generates a better FISH probe than native YAC DNA. Alu-PCR was performed using the Expand Long Template PCR kit (Roche, Lewes, UK). Cycling conditions were as follows: 94°C – 45 s, 55°C – 30 s, 68°C – 8 min, 35 cycles; 68°C −10 min final extension. Bacterial artificial chromosome (BAC) clones corresponding to positive YAC regions were arranged into contigs by consulting the Washington University FPC ( and UCSC Human Genome Browser ( databases. BAC clones were supplied by BACPAC Resources, Oakland, CA, USA ( Clone selection was biased to gene-containing BACs. Once a breakpoint-spanning BAC was identified, the precise position of the breakpoint in relation to candidate gene exons was determined by FISH probes generated from cosmids isolated from gridded chromosome-specific libraries (HGMP Resource Centre, see below for primers used to amplify GRIK4 exon regions for use as radioactive probes in cosmid library screens) or precisely positioned, repeat element-free, long-range PCR products (Expand long range PCR kit, Roche; see below for primer sequences). Cycling conditions were as follows: 94°C – 45 s, 52°C – 30 s, 68°C – 11 min, 35 cycles; 68°C – 15 min final extension.

FISH protocol

Probe template DNA (pooled YAC Alu-PCR products, BAC clone DNA, cosmid clone DNA or long-range PCR products) was labelled by nick translation and hybridized to patient metaphase spreads using standard FISH methods. Slides were counterstained with 4,6-diamidino-2-phenylindole in Vectashield anti-fade solution (Vector Laboratories, Peterborough, UK). A Zeiss Axioskop2 fluorescence microscope with a chroma number 81000 multi-spectral filter set was used in conjunction with CoolSnap HQ camera (Photometrics). Images were captured using SmartCapture2 (Digital Scientific, UK). FISH signals observed on derived chromosomes dictated the selection of further clones required to ‘walk’ towards, and eventually cross, the breakpoint.

PCR primers

Long-range PCR for FISH probe templates:

5′ ITGA4a::


5′ ITGA4b::


3′ ITGA4a::


3′ ITGA4b::


GRIK4 exon region-specific PCR: screening of chromosome 11 cosmid libraries:

Ex1 a::


Ex1 b::


Ex2/3 a::


Ex2/3 b::


Cycling conditions were as follows: 94°C – 2 min initial denaturation; 94°C – 1 min, 52°C – 1 min, 72°C – 75 s, 33 cycles; 72°C – 15 min final extension.

Association study

SNP marker selection was achieved by analysing the public International HAPMAP Project ( genotyping data (release #7) from the CEU population (Utah residents with ancestry from northern and western Europe) in the Haploview application (v.2.5) ( LD blocks across the GRIK4 locus (including regions 5′ and 3′ of the transcribed gene) were defined using Haploview's ‘solid spine of LD’ method with D′ values of greater than 0.8. Tagging SNPs were then selected both by their ability to represent haplotype diversity within these blocks down to the 10% haplotype frequency level and also their predicted ability to form reliable assays on the Illumina Inc., San Diego, CA, USA genotyping platform. Twenty-seven such SNPs were chosen over a span of 344 kb to cover the GRIK4 gene (326 kb). These SNPs were typed at Illumina Inc., San Diego, CA, USA using their proprietorial bead-array technology on 368 bipolar disorder cases and 386 schizophrenia cases comprised of patients of hospitals in southeast and south central Scotland, and 458 controls of matching geographical distribution from the Scottish National Blood Transfusion Service: a cohort described in more detail elsewhere.25 Analyses by our own and other groups on the Scottish population suggest that it is generally stable and homogeneous, with no obvious evidence for substructure.

Statistical analysis

Formatted genotype data were analysed using Cocaphase26 ( in order to derive P-values for both single marker associations and also two-, three- and four-marker global and individual haplotype tests – a surrogate for linkage disequilibrium (LD) blocks in our population.

Cocaphase uses the EM algorithm to estimate the haplotype frequencies of unphased genotype data and standard unconditional logistic regression analysis, applying the likelihood ratio test under a log-linear model to compare haplotype frequencies between cases and controls. In order to avoid misleading results caused by rare haplotypes, all haplotypes with a frequency less than or equal to 1% in both cases and controls were declared as rare and clumped together for the test of the null hypothesis using the command line option ‘-rare 0.01’. P-values for both global and individual tests of haplotype frequencies were determined. The global test P-value assesses the significance of the overall difference in the distribution of haplotype frequencies between cases and controls. The P-value from the individual test represents the significance of the difference in frequency of an individual haplotype between cases and controls.

In order to account for the multiple SNPs and haplotypes tested, permutation analysis (1000 permutations) was performed also using Cocaphase. Cocaphase randomly reassigns the diagnosis labels (case vs control) of the individuals. All single markers or haplotypes of a specified window size are then tested and the most significant P-value from each permutation is stored. Based on the distribution of these stored P-values, an ‘experimentwise’ significance level27 is provided for the most significant P-value observed for the particular sliding-window size (i.e. one to four). Odds ratios (ORs) were calculated online (


Characterization of loci affected by the karyotype rearrangement

The chromosomal rearrangement was discovered on fine analysis to be more complex than the description in the original study18 (see legend to Figure 1). It is likely that the cytogenetic abnormality originates from a pericentric inversion of chromosome 2 coupled with a modifed translocation event between chromosomes 2 and 11 (Figure 1). In addition, a segment of 11q adjacent to the translocation has been transferred onto chromosome 8. A representative image from a BAC FISH probe crossing the breakpoint on chromosomes 11 is shown in Figure 2. FISH analysis positioned five breakpoints. Table 1 details the YAC and BAC probes that define the breakpoint positions and Figure 3 shows the gene content in their vicinity. The relatively recent annotation of sequences from the Human Genome Project has meant that our understanding of the precise relationship between breakpoint and transcriptional units has changed over time: the data presented here are a combination of the May 2004 freeze of the UCSC Human Genome Browser ( and in-house annotation through the collation of expressed sequence tags (ESTs) and incomplete cDNA sequences from multiple species. We were not able to characterize in detail the region of 8q13 where the segment of chromosome 11q23.3–q24.2 has inserted owing to a limited supply of patient chromosome material. The breakpoint at 2p12 lies within a gene-poor region (3 Mb in size) and was not pursued beyond the resolution of YAC clones. The breakpoint at 2q21.2 disrupts a potentially aberrant fusion transcript, which appears to be a non-genic LINE1 transcript intergenically spliced to a genuine, CpG island-associated gene, NAP5 (NCK-Associated Protein). The functional NAP5 gene has been described in its full-length form only in the mouse (NM_172484), but the cognate human orthologue is not disrupted, lying centromeric to the breakpoint.

Figure 1

Representation of complex chromosomal rearrangement in the translocation patient. The pericentric chromosome 2 inversion is coupled with a reciprocal translocation with chromosome 11. The chromosome 11 region between the 11q23.3 and 11q24.3 breakpoints has inserted on chromosome 8q13. The full karyotype of this rearrangement is (46, XX, ins(8;11)(q13;q23.3q24.2) inv(2)(p12q32.1) t(2;11)(q21.3;q24.2) der(2)(2qter → 2q32.1::2p12 → 2q21.3::11q24.2 → 11qter) der(11)(11pter → 11q23.3::2q21.3 → 2q32.1::2p12 → 2pter) der(8)(8pter → 8q13::11q23.3 → 11q24.2::8q13 → 8qter).

Figure 2

FISH hybridization. The red signals (biotinylated probe) correspond to whole chromosome 8 paint. The green signals (digoxigeninylated probe) indicate hybridization of a probe derived from CEPH YAC 936_d_9. This YAC probe hybridizes to normal 11q23.3, the derived chromosome 11 and the derived chromosome 8. This split signal indicates that the YAC spans the breakpoint in the region of the GRIK4 gene. A BAC RP11-89p5-derived probe and cosmid-derived probes (data not shown) gave FISH results indicating that the breakpoint lies within the body of the GRIK4 gene (Figures 3 and 4).

Table 1 FISH probes used in the definition of the breakpoints in the translocation patient
Figure 3

Genomic arrangement of the genes within 500 kb of each chromosomal breakpoint in the translocation patient. To the right of the figure is a block listing the cytogenetic bands at which breakpoints were localized. The horizontal lines to the left of the block represent 1 Mb stretches of genomic DNA centred on the position of the breakpoints. The resolution accuracy of each breakpoint position, as determined by FISH, is indicated by the length of the black bars superimposed on the chromosome stretches. Rectangular boxes indicate the location, genomic extent and transcriptional orientation of each gene: those situated above the chromosome line are transcribed from left to right, and those below from right to left. Shading within the rectangles indicates a direct gene disruption.

The 11q23.3 breakpoint lies within the high-affinity kainate ionotropic glutamate receptor, GRIK4, locus (acc. S67803 and NM_014619, alternative nomenclature KA1/EAA1). The function of this gene and the neurotransmitter receptor it encodes are described in more detail in the Discussion section. FISH with cosmid-derived probes (Figure 4) positioned the breakpoint between exons 2 and 3 of GRIK4. The expected outcome of the position of the breakpoint is the truncation of all putative transcript forms such that a functional receptor cannot be encoded from the derived chromosome 11. PKNOX2/PREP2, encoding a transcription factor, is located at the 11q24.2 breakpoint.28, 29, 30 The reference sequence (NM_022062) for this gene is probably incomplete; further 5′ exons extending to a CpG island are present in several ESTs (e.g. BC045626). The putative complete gene sequence is disrupted by the breakpoint resulting in the loss of one functional allele in this patient. PKNOX2/PREP2 encodes a member of the MEINOX-TALE class of homeodomain-containing proteins that heterodimerizes with PBX1A to form a functional transcriptional regulator. Little is known about the function of this gene product other than that it is expressed widely with the highest levels observed in the brain, pancreas and lung.31 The integrin A4 cell adhesion molecule (ITGA4, acc. L12002 and X16983) and the candidate retinitis pigmentosa gene, RP26/CERKL (acc. NM_201548), are both located at the 2q31.3 breakpoint. The breakpoint-spanning BAC RP11-358m9 contains both genes but long-range PCR-based FISH probes corresponding to the 5′ and 3′ ends of ITGA4 (see Materials and methods for primer sequences) indicated that the breakpoint must lie within a 50 kb window downstream of its coding sequence. This region consists entirely of RP26/CERKL gene sequence, indicating that this gene was also directly disrupted. A recessive nonsense mutation in the RP26/CERKL gene has been described in two unrelated Spanish families with retinitis pigmentosa,32 indicating that disruption and the resulting haploinsufficiency of this gene is unlikely to explain the clinical phenotype of the patient. The direct gene disruptions of the PKNOX2/PREP2 and RP26/CERKL genes are not discussed further in this paper.

Figure 4

The GRIK4 gene: annotation pertaining to the cytogenetics and association studies. Exons are represented by numbered vertical lines. The breakpoint position is marked by a longer vertical line. The large rectangle at the top signifies the breakpoint-spanning BAC clone as determined by FISH. Labelled arrows indicate the location of cosmid clones used in the positioning of the breakpoint. The direction of the arrows represents the direction towards the breakpoint as interpreted from FISH images using probes derived from these clones. The 27 SNPs typed in the association study are shown together with the extended haplotype regions associated with schizophrenia and bipolar disorder. Filled triangles indicate significant single marker association with disease. Core (filled rectangles) and extended (bracketed lines) haplotypes associated with both schizophrenia and bipolar disorder are shown beneath the individual markers. The three rows at the bottom of the figure show LD blocks calculated from HapMap genotyping data processed in the Haploview application set at three defined stringency levels of multiallelic D′ before adjacent blocks were joined. A scale bar is shown.


Of the several genes disrupted or potentially deregulated in this subject, GRIK4 is the outstanding candidate to explain the observed symptomatology. This is owing to its neuronal expression pattern, a previous report of a disrupted glutamate receptor in a psychiatric condition16 and the hypothesized involvement of aberrant glutamate neurotransmission in susceptibility to psychiatric illness (see Discussion).

In order to assess the contribution of the GRIK4 gene to psychiatric illness in the karyotypically normal population, we undertook a case–control association study employing 27 SNP markers selected across the gene for their ability to ‘tag’ common haplotype variation (see Materials and methods). Analysis of genotyping data from schizophrenia cases, bipolar disorder cases and controls was carried out using Cocaphase. At the level of individual markers (Table 2), two principal regions of the gene demonstrated allele frequency changes in the case groups: SNPs 15, 16 and 17 showed significant association with schizophrenia (P=0.001, 0.004 and 0.015, respectively) and SNPs 23 and 26 were associated with bipolar disorder (P=0.003 and 0.020, respectively). Of these, SNP 15 remained significantly associated with schizophrenia (P=0.030, s.e. 0.005; Table 2) after permutation analysis to counter multiple testing errors (see Materials and methods). These positionally bi-partite findings (Table 2) were further supported by a sliding window global analysis of haplotypes of lengths two to four, such that distinct clusters of significant P-values were observed in both schizophrenia (best global haplotype, SNPs 15 and 16; P=0.005) and bipolar disorder (best global haplotype, SNPs 26 and 27; P=0.002). The latter global haplotype approached significance after permutation analysis (P=0.049, s.e. 0.007; Table 2). Additional significant global P-values were observed outside of the two regions, but without clear clustering between overlapping windows. Table 2 also shows that tests of the case group, consisting of combined diagnoses, resulted in less significant P-values at both regions when compared to the tests of the individual diagnoses, suggesting that different regions within the GRIK4 locus confer disease-specific risk/protective effects for the two disorders.

Table 2 Single SNP marker and sliding window global haplotype (2–4 SNP markers)

We examined the associations in further detail by assessing the contribution of individual haplotypes to the global associations across the entire gene (Table 3). An individual haplotype consisting of three SNPs (CTT, SNPs 15–17) was responsible for the most significant association with schizophrenia (P=0.0005, est. freq. cases: 0.38, est. freq. controls: 0.29, OR of 1.453 (95% CI 1.182–1.787)). Although this three-SNP haplotype just failed to remain significant after permutation correction, the component two-SNP haplotype (TT, SNPs 16 and 17; P=0.0006) did remain significant (P=0.0430, s.e. 0.0064; Table 3) after such correction. An underlying extended haplotype consisting of nine SNPs (TCCTTGTGA, SNPs 13–21) appeared upon alignment of the local individual haplotypes: this haplotype conferred susceptibility to schizophrenia. For bipolar disorder, a two-SNP haplotype (GC, SNPs 26 and 27) showed the greatest significance (P=0.0002, est. freq. cases: 0.16, est. freq. controls: 0.23, OR of 0.624 (95% CI 0.485–0.802)). This haplotype remained significant after permutation correction (P=0.0190, s.e. 0.0043; Table 3). Again, overlapping individual haplotype results in this region suggested that this was a component of an underlying seven-SNP haplotype (ATCATGC, SNPs 21–27), in this case conferring a protective effect against bipolar disorder.

Table 3 Key individual haplotypes associated with SCZ (top) and BPD (bottom)

In addition, it has been shown previously that a genuinely significant individual haplotype can exist in the absence of a significant global test.33 We identified several consistent individual haplotypes between markers 4 and 8, which were significantly associated with a protective effect against bipolar disorder despite this region not appearing in the global analysis.


Five chromosomal breakpoints have been positioned at the gene resolution level in a patient with schizophrenia and mental retardation. We have described the disruption of the gene, GRIK4 (KA1/EAA1), encoding the high-affinity subunit of the kainate-type ionotropic glutamate receptor.34, 35 In this patient, we believe that haploinsufficiency of this gene is most likely responsible for the psychiatric component of the patient's diagnosis although the existence of several other disrupted genes in this patient, the potential for position effects on genes at a distance from the breakpoints36, 37 and the possibility of cryptic deletions associated with the chromosomal rearrangement38 mean that we cannot definitively rule out other mechanisms. We have not yet examined these genes at the molecular or genotyping level and so these alternative explanations remain uninvestigated in this patient.

GRIK4 lies at the edge of a schizophrenia linkage region described in a recent publication39 with one microsatellite marker, D11S925, located within an intron at the 3′ end of the gene. In contrast, a recent meta-analysis of several bipolar disorder linkage studies failed to observe any significant findings on chromosome 11.40 However, a conventional linkage study approach would be unlikely to discover loci with protective effects. Recently, an association study was carried out on GRIK4 in Japanese schizophrenia patients.41 The study failed to identify any statistically significant single SNP or haplotype associations despite good marker coverage. Explanations for this might be the modest size of the study (100 cases, 100 controls) and the different population studied.

Beyond the bona fide gene mutation in the translocation patient, the most compelling evidence we describe here in support of a role for GRIK4 in schizophrenia and bipolar disorder comes from our findings from a comprehensive case–control association study in Scottish population. Interestingly, two physically separated haplotype regions within the gene showed independent and significant association with schizophrenia and bipolar disorder, respectively. This finding mirrors the repeated observation of overlapping linkage ‘hotspots’ for these two conditions.42 The OR values detailed above for the schizophrenia and bipolar disorder core individual haplotypes reflect their high frequency in the population (0.38 and 0.23, respectively) and are of a magnitude entirely consistent with previously described complex disease-susceptibility variants.43 Although the corrected P-values are only moderately significant, we believe they should be taken in the context of the prior probability conferred on the GRIK4 gene through the cytogenetic study. Replicating these findings in additional European sample sets should be an important target for future genetic studies.

The schizophrenia risk region lies primarily over exons encoding the N-terminal extracellular domain, whereas the bipolar disorder protective region lies over exons encoding the cytoplasmic C-terminus of the protein, suggesting that the, as yet undiscovered, underlying causative mutations for these two regions could act through different aspects of GRIK4 function or regulation. Further analysis (data not shown) failed to reveal any effect of the protective haplotype on diagnostic/phenotypic subclasses of the bipolar disorder cohort (e.g. BP1 vs BP2, hallucinations vs delusions). This suggests that the haplotype confers a protective effect on all forms of bipolar disorder.

To our knowledge, no previous example has been documented where a single gene contains distinct and independent risk and protective haplotypes for two disorders. However, there are examples of complex genetic disorders where all the individual features of the GRIK4 association findings have been described. Independent CARD15/NOD2 gene mutations can give rise to Blau syndrome44 or Crohn's disease45, 46 – distinct disorders with a common inflammatory basis. There are also examples where both risk and protective alleles have been identified in genes such as PPARG in insulin resistance disorders.47 Moreover, in this example, like GRIK4 and DISC1, the PPARG mutations exist in both rare, severe familial forms and common, low-penetrance population-level forms.

With regard to GRIK4 function, abnormalities in glutamatergic neurotransmission have previously been proposed as a mechanism for psychiatric illness. The ‘Glutamate Hypothesis’ of schizophrenia was devised to explain the psychotic symptoms exhibited by individuals following administration of ionotropic glutamate receptor antagonists such as phencyclidine (PCP; ‘Angel Dust’) and ketamine.48, 49 Several studies also point to changes, predominantly decreases, in glutamate receptor subunit expression in the post-mortem brains of patients with schizophrenia.50, 51, 52 Four classes of ionotropic glutamate receptors have been identified on the basis of their pharmacological profiles and sequence homologies: NMDA receptors, AMPA receptors, kainate receptors and delta receptors. Intriguingly, with the inclusion of this study, examples from each class have now been associated with the pathophysiology of psychiatric illness. NMDAR1/Grin1 knockdown mice display components of schizophrenia-like behaviour,53 the AMPA receptor GRIA3 has been shown to be disrupted in a patient diagnosed with bipolar disorder co-morbid with mental retardation16 and a breakpoint has been discovered adjacent to a delta-type receptor (Pickard et al., manuscript in preparation).

GRIK4 is expressed in the amygdala, hippocampal formation (CA3 pyramidal and dentate granule cells) and entorhinal cortex,54, 55 a pattern matching brain regions implicated in the psychoses.56 In addition, the potential role of long-term potentiation (LTP; the molecular model now generally accepted to underlie observed changes in activity-dependent synaptic plasticity57, 58, 59) in psychiatric illness has also been widely hypothesized. This is as a result of the well-documented deficits in particular cognitive tasks involving learning and memory in individuals with schizophrenia, bipolar disorder and unipolar major depression.60, 61, 62 Presynaptic changes in kainate receptor-dependent plasticity have been described at hipocampal mossy fibre synapses,63, 64 suggesting that the expected reduction in GRIK4 protein levels in the karyotypically abnormal patient might be expected to modify kainate receptor channel properties, and their contribution to LTP, by altering subunit stoichiometry.

The findings described in this report provide supportive evidence for a glutamatergic contribution to psychosis. The susceptibility and protective haplotypes described above present future opportunities for replication studies and for re-sequencing coding and non-coding portions of GRIK4, bounded by the relevant LD blocks, to identify causative mutations. The findings also suggest that the study of rare chromosomal rearrangements can direct research towards pertinent biological pathways and processes that may be targets for further genetic investigation and potential therapeutic strategies.

Accession codes




  1. 1

    MacIntyre DJ, Blackwood DH, Porteous DJ, Pickard BS, Muir WJ . Chromosomal abnormalities and mental illness. Mol Psychiatry 2003; 8: 275–287.

  2. 2

    Nielsen J, Wohlert M . Chromosome abnormalities found among 34,910 newborn children: results from a 13-year incidence study in Arhus, Denmark. Hum Genet 1991; 87: 81–83.

  3. 3

    Pickard BS, Millar JK, Porteous DJ, Muir WJ, Blackwood DH . Cytogenetics and gene discovery in psychiatric disorders. Pharmacogenomics J 2005; 5: 81–88.

  4. 4

    Blackwood DH, Fordyce A, Walker MT, St Clair DM, Porteous DJ, Muir WJ . Schizophrenia and affective disorders – cosegregation with a translocation at chromosome 1q42 that directly disrupts brain-expressed genes: clinical and P300 findings in a family. Am J Hum Genet 2001; 69: 428–433.

  5. 5

    St Clair D, Blackwood D, Muir W, Carothers A, Walker M, Spowart G et al. Association within a family of a balanced autosomal translocation with major mental illness. Lancet 1990; 336: 13–16.

  6. 6

    Millar JK, Wilson-Annan JC, Anderson S, Christie S, Taylor MS, Semple CA et al. Disruption of two novel genes by a translocation co-segregating with schizophrenia. Hum Mol Genet 2000; 9: 1415–1423.

  7. 7

    Millar JK, Christie S, Anderson S, Lawson D, Hsiao-Wei Loh D, Devon RS et al. Genomic structure and localisation within a linkage hotspot of Disrupted In Schizophrenia 1, a gene disrupted by a translocation segregating with schizophrenia. Mol Psychiatry 2001; 6: 173–178.

  8. 8

    Ekelund J, Hovatta I, Parker A, Paunio T, Varilo T, Martin R et al. Chromosome 1 loci in Finnish schizophrenia families. Hum Mol Genet 2001; 10: 1611–1617.

  9. 9

    Hennah W, Varilo T, Kestila M, Paunio T, Arajarvi R, Haukka J et al. Haplotype transmission analysis provides evidence of association for DISC1 to schizophrenia and suggests sex-dependent effects. Hum Mol Genet 2003; 12: 3151–3159.

  10. 10

    Gejman PV, Martinez M, Cao Q, Friedman E, Berrettini WH, Goldin LR et al. Linkage analysis of fifty-seven microsatellite loci to bipolar disorder. Neuropsychopharmacology 1993; 9: 31–40.

  11. 11

    Hwu HG, Liu CM, Fann CS, Ou-Yang WC, Lee SF . Linkage of schizophrenia with chromosome 1q loci in Taiwanese families. Mol Psychiatry 2003; 8: 445–452.

  12. 12

    Curtis D, Kalsi G, Brynjolfsson J, McInnis M, O'Neill J, Smyth C et al. Genome scan of pedigrees multiply affected with bipolar disorder provides further support for the presence of a susceptibility locus on chromosome 12q23–q24, and suggests the presence of additional loci on 1p and 1q. Psychiatr Genet 2003; 13: 77–84.

  13. 13

    Millar JK, Pickard BS, Mackie S, James R, Christie S, Buchanan SR et al. DISC1 and PDE4B are interacting genetic factors in schizophrenia that regulate cAMP signalling. Science 2005; 310: 1187–1191.

  14. 14

    Kamnasaran D, Muir WJ, Ferguson-Smith MA, Cox DW . Disruption of the neuronal PAS3 gene in a family affected with schizophrenia. J Med Genet 2003; 40: 325–332.

  15. 15

    Pickard BS, Malloy MP, Porteous DJ, Blackwood DH, Muir WJ . Disruption of a brain transcription factor, NPAS3, is associated with schizophrenia and learning disability. Am J Med Genet B Neuropsychiatr Genet 2005; 136: 26–32.

  16. 16

    Gecz J, Barnett S, Liu J, Hollway G, Donnelly A, Eyre H et al. Characterization of the human glutamate receptor subunit 3 gene (GRIA3), a candidate for bipolar disorder and nonspecific X-linked mental retardation. Genomics 1999; 62: 356–368.

  17. 17

    Baysal BE, Willett-Brozick JE, Badner JA, Corona W, Ferrell RE, Nimgaonkar VL et al. A mannosyltransferase gene at 11q23 is disrupted by a translocation breakpoint that co-segregates with bipolar affective disorder in a small family. Neurogenetics 2002; 4: 43–53.

  18. 18

    Doody GA, Johnstone EC, Sanderson TL, Owens DG, Muir WJ . ‘Pfropfschizophrenie’ revisited: schizophrenia in people with mild learning disability. Br J Psychiat 1998; 173: 145–153.

  19. 19

    Sanderson TL, Best JJ, Doody GA, Owens DG, Johnstone EC . Neuroanatomy of comorbid schizophrenia and learning disability: a controlled study. Lancet 1999; 354: 1867–1871.

  20. 20

    Wing JK, Ustun TB, Sartorius N (eds) Diagnosis and Clinical Measurement in Psychiatry: A Reference Manual for Scan/PSE-10. Cambridge University Press: Cambridge, 1998.

  21. 21

    Endicott J, Spitzer RL . A diagnostic interview: the schedule for affective disorders and schizophrenia. Arch Gen Psychiat 1978; 35: 837–844.

  22. 22

    Carvill S . Sensory impairments, intellectual disability and psychiatry. J Intellect Disabil Res 2001; 45: 467–483.

  23. 23

    Breen M, Arveiler B, Murray I, Gosden JR, Porteous DJ . YAC mapping by FISH using Alu-PCR-generated probes. Genomics 1992; 13: 726–730.

  24. 24

    Barrett JC, Fry B, Maller J, Daly MJ . Haploview: analysis and visualization of LD and haplotype maps. Bioinformatics 2005; 21: 263–265.

  25. 25

    Thomson PA, Wray NR, Millar JK, Evans KL, Le Hellard S, Condie A et al. Association between the TRAX/DISC locus and both bipolar disorder and schizophrenia in the Scottish population. Mol Psychiatry 2005; 10: 657–668.

  26. 26

    Dudbridge F . Pedigree disequilibrium tests for multilocus haplotypes. Genet Epidemiol 2003; 25: 115–121.

  27. 27

    Churchill GA, Doerge RW . Empirical threshold values for quantitative trait mapping. Genetics 1994; 138: 963–971.

  28. 28

    Imoto I, Sonoda I, Yuki Y, Inazawa J . Identification and characterization of human PKNOX2, a novel homeobox-containing gene. Biochem Biophys Res Commun 2001; 287: 270–276.

  29. 29

    Haller K, Rambaldi I, Kovacs EN, Daniels E, Featherstone M . Prep2: cloning and expression of a new prep family member. Dev Dyn 2002; 225: 358–364.

  30. 30

    Fognani C, Kilstrup-Nielsen C, Berthelsen J, Ferretti E, Zappavigna V, Blasi F . Characterization of PREP2, a paralog of PREP1, which defines a novel sub-family of the MEINOX TALE homeodomain transcription factors. Nucleic Acids Res 2002; 30: 2043–2051.

  31. 31

    Villaescusa JC, Verrotti AC, Ferretti E, Farookhi R, Blasi F . Expression of Hox cofactor genes during mouse ovarian follicular development and oocyte maturation. Gene 2004; 330: 1–7.

  32. 32

    Tuson M, Marfany G, Gonzalez-Duarte R . Mutation of CERKL, a novel human ceramide kinase gene, causes autosomal recessive retinitis pigmentosa RP26. Am J Hum Genet 2004; 74: 128–138.

  33. 33

    Fallin D, Cohen A, Essioux L, Chumakov I, Blumenfeld M, Cohen D et al. Genetic analysis of case/control data using estimated haplotype frequencies: application to APOE locus variation and Alzheimer's disease. Genome Res 2001; 11: 143–151.

  34. 34

    Werner P, Voigt M, Keinanen K, Wisden W, Seeburg PH . Cloning of a putative high-affinity kainate receptor expressed predominantly in hippocampal CA3 cells. Nature 1991; 351: 742–744.

  35. 35

    Kamboj RK, Schoepp DD, Nutt S, Shekter L, Korczak B, True RA et al. Molecular cloning, expression, and pharmacological characterization of humEAA1, a human kainate receptor subunit. J Neurochem 1994; 62: 1–9.

  36. 36

    Velagaleti GV, Bien-Willner GA, Northup JK, Lockhart LH, Hawkins JC, Jalal SM et al. Position effects due to chromosome breakpoints that map approximately 900 Kb upstream and approximately 1.3 Mb downstream of SOX9 in two patients with campomelic dysplasia. Am J Hum Genet 2005; 76: 652–662.

  37. 37

    Kleinjan DA, van Heyningen V . Long-range control of gene expression: emerging mechanisms and disruption in disease. Am J Hum Genet 2005; 76: 8–32.

  38. 38

    Gribble SM, Prigmore E, Burford DC, Porter KM, Ng BL, Douglas EJ et al. The complex nature of constitutional de novo apparently balanced translocations in patients presenting with abnormal phenotypes. J Med Genet 2005; 42: 8–16.

  39. 39

    Gurling HM, Kalsi G, Brynjolfson J, Sigmundsson T, Sherrington R, Mankoo BS et al. Genomewide genetic linkage analysis confirms the presence of susceptibility loci for schizophrenia, on chromosomes 1q32.2, 5q33.2, and 8p21–22 and provides support for linkage to schizophrenia, on chromosomes 11q23.3–24 and 20q12.1–11.23. Am J Hum Genet 2001; 68: 661–673.

  40. 40

    McQueen MB, Devlin B, Faraone SV, Nimgaonkar VL, Sklar P, Smoller JW et al. Combined analysis from eleven linkage studies of bipolar disorder provides strong evidence of susceptibility Loci on chromosomes 6q and 8q. Am J Hum Genet 2005; 77: 582–595.

  41. 41

    Shibata H, Aramaki T, Sakai M, Ninomiya H, Tashiro N, Iwata N et al. Association study of polymorphisms in the GluR7, KA1 and KA2 kainate receptor genes (GRIK3, GRIK4, GRIK5) with schizophrenia. Psychiat Res 2006; 141: 39–51.

  42. 42

    Kohn Y, Lerer B . Excitement and confusion on chromosome 6q: the challenges of neuropsychiatric genetics in microcosm. Mol Psychiatry 2005; 10: 1062–1073.

  43. 43

    Wang WY, Barratt BJ, Clayton DG, Todd JA . Genome-wide association studies: theoretical and practical concerns. Nat Rev Genet 2005; 6: 109–118.

  44. 44

    Miceli-Richard C, Lesage S, Rybojad M, Prieur AM, Manouvrier-Hanu S, Hafner R et al. CARD15 mutations in Blau syndrome. Nat Genet 2001; 29: 19–20.

  45. 45

    Hugot JP, Chamaillard M, Zouali H, Lesage S, Cezard JP, Belaiche J et al. Association of NOD2 leucine-rich repeat variants with susceptibility to Crohn's disease. Nature 2001; 411: 599–603.

  46. 46

    Ogura Y, Bonen DK, Inohara N, Nicolae DL, Chen FF, Ramos R et al. A frameshift mutation in NOD2 associated with susceptibility to Crohn's disease. Nature 2001; 411: 603–606.

  47. 47

    Meirhaeghe A, Amouyel P . Impact of genetic variation of PPARgamma in humans. Mol Genet Metab 2004; 83: 93–102.

  48. 48

    Goff DC, Wine L . Glutamate in schizophrenia: clinical and research implications. Schizophr Res 1997; 27: 157–168.

  49. 49

    Javitt DC . Glutamate as a therapeutic target in psychiatric disorders. Mol Psychiatry 2004; 9: 984–997.

  50. 50

    Ibrahim HM, Hogg AJ, Healy Jr DJ, Haroutunian V, Davis KL, Meador-Woodruff JH . Ionotropic glutamate receptor binding and subunit mRNA expression in thalamic nuclei in schizophrenia. Am J Psychiat 2000; 157: 1811–1823.

  51. 51

    Meador-Woodruff JH, Davis KL, Haroutunian V . Abnormal kainate receptor expression in prefrontal cortex in schizophrenia. Neuropsychopharmacology 2001; 24: 545–552.

  52. 52

    Sokolov BP . Expression of NMDAR1, GluR1, GluR7, and KA1 glutamate receptor mRNAs is decreased in frontal cortex of ‘neuroleptic-free’ schizophrenics: evidence on reversible up-regulation by typical neuroleptics. J Neurochem 1998; 71: 2454–2464.

  53. 53

    Mohn AR, Gainetdinov RR, Caron MG, Koller BH . Mice with reduced NMDA receptor expression display behaviors related to schizophrenia. Cell 1999; 98: 427–436.

  54. 54

    Chittajallu R, Braithwaite SP, Clarke VR, Henley JM . Kainate receptors: subunits, synaptic localization and function. Trends Pharmacol Sci 1999; 20: 26–35.

  55. 55

    Lerma J, Paternain AV, Rodriguez-Moreno A, Lopez-Garcia JC . Molecular physiology of kainate receptors. Physiol Rev 2001; 81: 971–998.

  56. 56

    Antonova E, Sharma T, Morris R, Kumari V . The relationship between brain structure and neurocognition in schizophrenia: a selective review. Schizophr Res 2004; 70: 117–145.

  57. 57

    Miller S, Mayford M . Cellular and molecular mechanisms of memory: the LTP connection. Curr Opin Genet Dev 1999; 9: 333–337.

  58. 58

    Martin SJ, Grimwood PD, Morris RG . Synaptic plasticity and memory: an evaluation of the hypothesis. Annu Rev Neurosci 2000; 23: 649–711.

  59. 59

    Abel T, Lattal KM . Molecular mechanisms of memory acquisition, consolidation and retrieval. Curr Opin Neurobiol 2001; 11: 180–187.

  60. 60

    Kuperberg G, Heckers S . Schizophrenia and cognitive function. Curr Opin Neurobiol 2000; 10: 205–210.

  61. 61

    Elvevag B, Goldberg TE . Cognitive impairment in schizophrenia is the core of the disorder. Crit Rev Neurobiol 2000; 14: 1–21.

  62. 62

    Blackwood DH, Glabus MF, Dunan J, O'Carroll RE, Muir WJ, Ebmeier KP . Altered cerebral perfusion measured by SPECT in relatives of patients with schizophrenia. Correlations with memory and P300. Br J Psychiat 1999; 175: 357–366.

  63. 63

    Contractor A, Swanson G, Heinemann SF . Kainate receptors are involved in short- and long-term plasticity at mossy fiber synapses in the hippocampus. Neuron 2001; 29: 209–216.

  64. 64

    Lauri SE, Bortolotto ZA, Bleakman D, Ornstein PL, Lodge D, Isaac JT et al. A critical role of a facilitatory presynaptic kainate receptor in mossy fiber LTP. Neuron 2001; 32: 697–709.

Download references


We thank Judy Fantes, Maura Walker, Margaret van Beck, Paul Perry, Veronica van Heyningen and Western General Hospital Cytogenetics service for expert assistance and advice. We also wish to acknowledge the HGMP Resource Centre for the provision of YAC, Cosmid and IMAGE clones and gridded DNA libraries. This work was supported by a collaborative research agreement with Merck, Sharp and Dohme (The Neuroscience Research Centre, Terlings Park, UK), SHERT Grant RG45/01, Wellcome Trust ‘Genes to Cognition’ grant and CSO Grant K/MRS/50/C2789. Part of the patient sample collection was supported by a grant from the State Hospital for Scotland at Carstairs.

The authors declare that there are no conflicts of interest.

Author information

Correspondence to B S Pickard.

Rights and permissions

Reprints and Permissions

About this article


  • association study
  • chromosome abnormality
  • glutamate receptor
  • haplotype
  • psychiatry

Further reading