A Functional Polymorphism in the Promoter Region of the Dopamine D2 Receptor Gene Is Associated with Schizophrenia

Abstract

An excess dopaminergic activity may be implicated in the etiology of schizophrenia. Our objective was to identify nucleotide variants in the 5′ region of the dopamine D2 receptor gene (DRD2) and to clarify their effects on schizophrenia. We identified two polymorphisms, the A-241G and −141C Ins/Del, by examination of 259 bp in the 5′-flanking region and 249 bp of exon 1 of DRD2. Reporter constructs containing the −141C Delallele cloned into a luciferase reporter plasmid drove 21% (Y-79 cells) and 43% (293 cells) expression compared with the −141C Ins allele. In a case-control study, the −141C Delallele frequency was significantly lower in 260 schizophrenic patients than in 312 controls (OR = 0.60, 95%CI 0.44–0.81, P < 0.001). No significant association was found between the A-241G polymorphism and in vitro luciferase activity, or in allele frequency between the patients versus controls. These findings show that the −141C Ins/Delmay be a functional polymorphism in the 5′-promoter region of DRD2 and may affect the susceptibility to schizophrenia.

Introduction

The concept that an excess dopaminergic activity leads to the psychotic symptoms of schizophrenia is based on pharmacological observations (1–3). Dopamine agonists can cause or exacerbate psychotic symptoms (4). The antipsychotic potency of a wide range of neuroleptic drugs is correlated with the ability to block dopamine D2 receptors (2,3). Also, the density of dopamine D2 and/or other D2-like receptors (D3 and D4) is elevated in the brain of schizophrenics obtained at post-mortem who had been neuroleptic-free for many years before death (5). However, the issue of whether the schizophrenic process itself is associated with an increase in dopamine receptor is controversial (6) and it is not yet clear which D2-like receptors are raised in post-mortem schizophrenic brains (7,8).

An increase in the density of D2 receptors in schizophrenia could result from abnormal processing of genetically abnormal receptors. However, an intensive search for nucleotide variants causing alterations in the amino acid sequence has identified only a few molecular variants. The identified variants, Ser311Cys (9,10), Val96A1a (10), and Pro310Ser (10), are relatively rare. An association has been reported between the Cys311 variant and type I schizophrenia (11,12) or with alcoholism (13). However, other investigators failed to find an association with schizophrenia (10,14–19) or alcoholism (20). No other variant has been reported to be associated with schizophrenia or alcoholism (10).

Variation in the genomic sequence of the promoter region of the D2 receptor gene (DRD2) could affect the expression or regulation of the gene. The first exon of DRD2 is ∼250 kb apart from the second exon where the translation start codon exists (21). Since the possibility that promoter variants are associated with schizophrenia cannot be excluded, we investigated the sequences of the first exon and of the 5′-flanking region to find out functional variants associated with the expression of DRD2.

Results

Identification of polymorphisms in the 5′-region of DRD2

Direct sequencing of selected individual samples in which sequence variations were suggested by SSCP analysis carried out in randomly selected 20 schizophrenic and 20 control subjects revealed variants at two positions. A-241 was substituted with G and an insertion/deletion (Ins/Del) variant was found at position −141 where one cytosine was deleted from a run of two cytosines when compared with the published sequence (22; Fig.1). In addition, base differences at two positions from the published sequence (not polymorphic) were found by PCR direct sequencing in five individuals (Fig. 1).

No evidence for linkage disequilibrium between the A-241G and −141C Ins/Del polymorphisms was obtained (delta value = 0.012, χ² = 1.66, df = 1, P = 0.20) (23). The −141C Ins/Del polymorphism was not in linkage disequilibrium with the S311C polymorphism in exon 7 of DRD2 (delta value = −0.001, P = 0.79, n = 273), or with the TaqIA polymorphism located 3′ to DRD2 (delta value = −0.01, P = 0.31, n = 179).

Figure 1.

Polymorphisms in the 5′-region of the human DRD2. (A) Nucleotide sequence of the 5′ flanking region and exon 1 of the human D2 receptor gene. Upper case letters represent exon 1 (33) and lower case letters, the 5′-flanking sequence. The location and orientation of oligonucleotides used for PCR are shown by arrows. The nucleotide sequence is numbered from the 5′ end of the D2a cDNA (33) as indicated at the right of each line. Bases different from the published sequence (22) are shown by asterisks. The position of the −141C Ins/Del polymorphism and the A-241G polymorphism are shown (box). (B1) Direct sequencing from individuals homozygous for the A-241G polymorphism; (B2) genotyping for the A-241G polymorphism. The fragments amplified by PCR with primers D2–676 and -677 were digested withMaelII. Lane M represents pGEM marker (Promega); lane AA is a homozygote for the A-241 allele; lane AG is a heterozygote for the A-241 and G-241 alleles; lane GG is a homozygote for the G-241 allele. (C1) Direct sequencing from individuals homozygous for the −141C Ins/Del polymorphism; (C2) genotyping for the -141C Ins/Del polymorphism. The fragments amplified by PCR with primers D2–676 and -677 were digested with BstNI. Lane M represents pGEM marker (Promega); lane II is a homozygote for the −141C Ins allele; lane DI is a heterozygote for the −141C Ins and −141C Del alleles; lane DD is a homozygote for the −141C Del allele.

Figure 2.

Transient expression of luciferase enzymatic activity driven by the DRD2 5′-flanking 304 bp containing the A-241 and −141C Del alleles, the A-241 and −141C Ins alleles, and the G-241 and −141C Ins alleles in Y-79 cells (A) and in 293 cells (B.) All data are normalized to β-galactosidase expression driven by co-transfected pSV-β-galactosidase control plasmid (Promega). Activities are expressed as percentages of the activity of the positive control plasmid, pGL3-promoter (Promega). Each value is the mean ± SEM for three independent experiments each performed in duplicate. P values are from a t-test (two-tailed).

Table 1.

Distribution of the A-241G and the −141C Ins/Del polymorphisms of DRD2

Transient transfection and luciferase assay

To test the effect of a polymorphic sequence on gene expression, fragments containing the three haplotypes (the −141C Ins and A-241 alleles, the −141C Ins and G-241 alleles, the −141 Del and A-241 alleles) were fused to luciferase reporter constructs and transiently transfected into D2-expressing human retinoblastoma Y79 cells (24) and D2-non-expressing human kidney 293 cells. Stable cultured tumor cells of human CNS origin constitutively expressing D2-receptors have rarely been reported (24,25). Y-79 cells express both neural- and glial-specific cellular protein markers (26) and D2 receptors expressed in Y-79 cells are functional (24). It has been confirmed that 293 cells do not express D2 receptors (27). The haplotype containing the A-241 and −141C Ins alleles is the most common and may be the wild haplotype. The wild-type fragment with the A-241/−141C Ins alleles directed luciferase synthesis to a level of 180% and 48% of the SV40 promoter-reporter plasmid, or of 38- and 16-fold greater than the promoter-less basal plasmid in Y79 and 293 cells, respectively (Fig. 2). By contrast, the fragment bearing the A-241/−141C Del alleles directed synthesis with significantly less promoter strength than the fragment with the wild haplotype (39 and 21% of the SV-40 promoter-reporter plasmid, or 8- and 7-fold greater than the promoter-less basal plasmid in Y79 cells and 293 cells, respectively). Reporter constructs containing the A-241/−141C Del allele cloned into a luciferase reporter plasmid drove 21% (Y-79 cells) and 43% (293 cells) expression compared with the wild A-241/−141C Ins haplotype. The promoter strength of the fragment with the G-241/−141C Ins alleles did not significantly differ from that of the fragment with the wild haplotype in Y79 and 293 cells.

Association of DRD2 5′-promoter polymorphism with schizophrenia

No significant difference in genotype or allele frequency of the A-241G polymorphism was observed between the schizophrenic and control groups (Table 1). The frequency of the −141C Del allele was significantly decreased in the schizophrenic subjects compared with the controls. The odds ratio for schizophrenia associated with the −141C Ins allele was 0.60 (P < 0.001).

Discussion

The DRD2 5′-promoter fragments drove the transcription of heterologous luciferase constructs in Y79 cell line expressing DRD2 as well as in DRD2 non-expressing 293 cells. Although the three D2 5′-fragment constructs exhibited from 1.9- to 3.7-fold higher expression in Y79 cells than those in 293 cells, the promoter-less basic plasmid also showed 1.6-fold higher expression in Y79 cells than that in 293 cells. It is likely that the 304 bp sequence of DRD2 analyzed in the present study did not contain the elements required to confer a tissue-specific expression of the gene. It has been reported that the 1 kb rat D2 promoter and the 450 bp human D5 promoter do not show cell-specific expression for D2 or D5 expressing cells (28,29). Also, there may be repressor elements between position −1140 and −352 of the rat m4 cholinergic muscarinic receptor gene that repress transcription in non-expressing cells (30).

The fragment that contained the −141C Del allele showed a decrease in promoter strength as compared with the fragment that contained the −141CIns allele in Y-79 and 293 cells. Although the transcription factors involved in this allelic difference have not been identified, the position of the polymorphism is part of a putative binding site for Sp-1, 5′-CCAGGCCGGGGATCGCC.

Whether the −141C Ins/Del polymorphism is actually related to DRD2 gene expression in human brains is not yet clear. Our preliminary data showed that the number of spiperone binding sites (B_max) in the putamen of the post-mortem brains tended to be decreased in four non-schizophrenics who carried the −141C Del allele compared with six non-schizophrenics who did not: the mean ± SD (fmol/mg protein) of the former and latter groups were 216.8 ± 56.8 and 144.1 ± 58.7, respectively, P < 0.09, Student's /-test, two-tailed. This trend was in the expected direction from the results of the in vitro luciferase assay experiments. However, the binding of spiperone to D4 receptor could interfere with measurements of B_max.

A negative association betweeen the −141C allele associated with lower luciferase activity and schizophrenia was suggested by the results of the present study. Previous post-mortem and neuroimaging studies indicate that striatal D2 receptor density is slightly increased in drug-free schizophrenics as compared with controls (5). A decreased frequency of the −141C Del allele in schizophrenics may contribute to elevation of D2 receptor density in schizophrenics. If this association is confirmed by family samples, it will be evidence supporting the dopamine hypothesis for schizophrenia.

While it is generally accepted that genetic heterogeneity is a less likely bias in Japan, various biases including undetected population stratification may affect case-control comparisons. Furthermore, data that suggest a polymorphism in DRD2 associated with schizophrenia appear to be incompatible with linkage studies, since no study has reported positive linkage between schizophrenia and DRD2 (31,32). However, if the association suggested by the present study is true, the effect is small. Linkage tests using large numbers of sib pairs and/or transmission disequilibrium test are more suitable for confirming or excluding the allelic susceptiblity to schizophrenia.

In conclusion, a possible functional polymorphism in the 5′-promoter region of DRD2 was identified and the polymorphism may be associated with schizophrenia. However, both case-control comparisons and promoter studies are not without potential pitfalls; our conclusions must be considered tentative until they have been subjected to the test of independent replication.

Materials and Methods

Population

To evaluate the association between the identified polymorphisms and schizophrenia, we examined 260 unrelated Japanese patients [151 men and 119 women, aged 19–81 years (mean 44.4); age at disease onset 12–39 years (mean 22.4)] who met the criteria for schizophrenia of the American Psychiatric Association's Diagnostic and Statistical Manual of Mental Disorders (DSM-III-R) (1987). The patients were receiving treatment at eight hospitals within 200 km of Tokyo. The control subjects were 312 unrelated Japanese [173 males and 139 females, aged 29–75 years (mean 48.7)]. Among the controls, 135 were hospital staff members documented to be free of psychosis. The remainder were corporate employees who had requested annual physical examinations but had not been evaluated for psychiatric disorders by a psychiatrist. The cases and controls were resident in the same area of Japan. No minorities in Japan were included in this study. Samples of venous blood were collected after written informed consent had been obtained. The present study was approved by the ethics committees of Tokyo Medical and Dental University and University of Tsukuba.

Amplification of exon 1 and the 5′-flanking region of DRD2

DNA was prepared from blood using standard techniques. The genomic sequence of 284 bp of the 5′-flanking region and 274 bp of exon 1 of DRD2 was amplified by PCR with the primer pair D2–677 (5′-ACTGGCGAGCAGACGGTGAGGACCC; nn −284 ∼ −260) and D2–676 (5′-TGCGCGCGTGAGGCTGCCGGTTCGG; nn −5 ∼ +20), and the pair of D2–1073 (5′-CGCCGAGGAGGTACAGCTCCTTTGGTG; nn −354∼ −325) or D2–977 (5′-GCCGAACCGGCAGCCTCACGCGCGCA; nn −6 ∼ +21) and D2–976 (5′-GGGGCAGAGACGGCGCCGGCTGCTT; nn +250 ∼ +274). Nucleotides are numbered from the 5′ end of the human D2A cDNA (33), modified as described in the paper by Gandelman et al. (22). Native Pfu polymerase (Stratagene) was used to amplify the fragments by incubation at 98°C for 1 min followed by 35 cycles of 98°C for 20 s and 74°C for 5 min for primers of D2–676 and -677. When using primers D2–1073 or D2–977 and D2–976, the annealing/extension temperature was 77°C. PCR reaction buffers were supplemented with formamide at 4% final concentration.

Single strand conformation polymorphism (SSCP) and direct sequencing

The SSCP method was used to screen polymorphisms using PhastSystem (Pharmacia). Direct sequencing was carried out using a Sequenace kit (US Biochemical Corp.) or cycle sequencing using Taq polymerase and dye terminators on an Applied Biosystems automated sequencer (Perkin-Elmer).

Genotyping

PCR-restriction fragment length polymorphism (RFLP) analysis was performed on amplified fragments digested with BstN1 (for the −141C Ins/Del polymorphism) or with MaeIII (for the A-241G polymorphism). The digested fragments were electrophoresed in 2% agarose gel and were visualized by ethidium bromide staining.

Construction of D2-promoter-luciferase plasmids

Fragments consisting of 284 bp of the 5′-flanking sequence and 20 bp of the first exon were obtained from individuals with identified haplotypes. The fragments were amplified by PCR with the MluI linker added D2–677 and the BglII linker added D2–676. The fragments were cloned into MluI/BglII-cut pGL3-basic plasmid (Promega). The clones of the promoter segments used were sequenced in full to rule out any sequence alterations.

Cell culture, plasmid transfections and luciferase assays

Human retinoblastoma Y-79 cells purchased from The American Type Culture Collection (ATCC) were grown in suspension in RPMI 1640 supplemented with 15% fetal bovine serum. Human embryonal kidney 293 cells purchased from ATCC were grown in Eagle's minimal essential medium (MEM) supplemented with 10% horse serum.

The DNA purified by a Qiagen column was transiently transfected into cells using Tfx-50 (Promega). An aliquot of 3 µg of promoter-luciferase fusion plasmid, pGL3-basic and pGL3-promoter plasmid DNAs was co-transfected with pSV-β-galactosidase control plasmid (Promega). After 5 h, cells were washed in phosphate-buffered saline and maintained in the appropriate media supplemented with sera. After an additional 24 h, cells were harvested, lysed, and assayed for luciferase and β-galactosidase enzymatic activity according to the manufacturer's recommendations (Promega).

Statistical procedures

Luciferase activity was normalized to β-galactosidase activity and compared by Student's t-test. Group difference in allele frequency was evaluated by the χ² test in a case-control study. A two-tailed a criterion of <0.05 was considered statistically significant.

Acknowledgements

This study was supported by a scientific research grant (No. 08672594) from the Ministry of Education, Science and Culture of Japan, and a scientific research grant from the National Center of Neurology and Psychiatry of the Ministry of Health and Welfare of Japan.