Department of Medical Genetics, Institute of Basic Medical Sciences, University of Tsukuba, , Ibaraki, Japan; 2

Transcription

1 ORIGINAL RESEARCH ARTICLE (2000) 5, Macmillan Publishers Ltd All rights reserved /00 $ Mutation and association analysis of the 5 region of the dopamine D3 receptor gene in schizophrenia patients: identification of the Ala38Thr polymorphism and suggested association between DRD3 haplotypes and schizophrenia H Ishiguro 1,2, Y Okuyama 1, M Toru 2 and T Arinami 1 1 Department of Medical Genetics, Institute of Basic Medical Sciences, University of Tsukuba, , Ibaraki, Japan; 2 Department of Neuropsychiatry, Tokyo Medical and Dental University School of Medicine, 113, Tokyo, Japan Keywords: population; genome; linkage; variation; association; schizophrenia Although the association between the Ser9Gly polymorphism of the dopamine D3 receptor gene (DRD3) and schizophrenia has been investigated by many research groups, it is not known whether the Ser9Gly polymorphism alone or a variation in linkage disequilibrium may effect susceptibility to schizophrenia. We searched the 5 region of the DRD3 gene and found three novel polymorphisms: 712G/C, 205A/G, and Ala38Thr. The Ala38Thr polymorphism is located in the first transmembrane region and is conserved in the monkey, mouse, and rat. Case-control comparisons in 153 Japanese schizophrenia patients and 122 Japanese controls did not suggest an association between Ala38Thr and schizophrenia. However, there was a marginally significant association between the Ser9 allele of the Ser9Gly polymorphisms and schizophrenia (P = 0.02). Furthermore, there was a highly significant association between haplotypes of the 712G/C, 205A/G, and Ser9Gly polymorphisms and schizophrenia (P = , corrected P = 0.007). These positive findings were replicated in an additional 99 Japanese schizophrenia patients and 132 controls (P = 0.04 and , respectively). The most allelic differences of the Ser9Gly polymorphism between patient and control groups arose from the chromosome carrying specific alleles of the other three polymorphisms. This study indicates unknown variant(s) in linkage disequilibrium with the DRD3 haplotypes associated with schizophrenia. Molecular Psychiatry (2000) 5, Several lines of evidence suggest that the dopamine D3 receptor is involved in the pathophysiology of schizophrenia. The D3 receptor has a restricted pattern of expression in brain limbic areas associated with cognitive function and motivated behavior. 1 It may mediate the pharmacological effects of antipsychotic drugs, since typical and atypical antipsychotic drugs have a relatively high affinity for the dopamine D3 receptor. 2 Postmortem studies suggest a selective loss of DRD3 mrna in parietal and motor cortical regions of patients with chronic schizophrenia. 3 D3 selective agonists and antagonists reveal an inhibitory role on motor behavior for the D3 receptor, opposite to that of the D2 receptor. Tolerance to the motor but not to the therapeutic effects of neuroleptics is observed after prolonged treatment, which up-regulate D2 but not D3 receptor levels in animals. 4 A genetic association between the Ser9Gly polymorphism of the dopamine D3 receptor gene (DRD3) and schizophrenia has been investigated by many researchers. Groups in the UK and France found an association between homozygosity of this polymorphism and schizophrenia. 5 Some studies imply that a disproportionate frequency of homozygosity may be pronounced in patients with a family history of schizophrenia 6 and in patients with a good response to neuroleptics. 7 Although many researchers reported negative findings in unselected samples, 7 15 a meta-analysis of over 30 case-control studies resulted in a higher frequency of both forms of the homozygote in schizophrenia patients than in controls, 16 at least among Caucasians. 17 However, it is not known whether the Ser9Gly polymorphism alone or a variation in linkage disequilibrium has any effect upon susceptibility to schizophrenia. Receptor binding analysis using the Ser9Gly mutant human dopamine D3 receptor expressed in the Semliki Forest Virus system showed a significantly higher dopamine binding affinity for the Gly9 homozygote, but the binding of the heterozygote did not differ from that of the Ser9 homozygote. The D3-selective ligand GR99841 showed significantly lower binding activity for the Ser9 homozygote than both the homo- and heterozygote for Gly9. 18 However, the differences in the binding affinity are small, and the relevance of this finding in relation to disease is unclear. To find DNA variant(s) in linkage disequilibrium with the Ser9Gly polymorphism, we searched for mutation(s) in the 5 region and exon 1 of the DRD3 gene in 48 randomly-selected schizophrenia patients using PCR single-strand conformation polymorphism (SSCP) method. The PCR primers and conditions are listed in Table 1. We detected three novel polymorphisms: 712G/C, 205A/G, and Ala38Thr (Figure 1).

2 434 Table 1 PCR primers used to search for nucleotide variants in the DRD3 gene Region Primers (F: Forward; R: Reverse) 3 end of primer sequence Polymerase a Annealing temp 5 region No. 5 (F) 5 -TTTTGCTTGGGTTTCAGGGTTACTC 849 bp upstream to exon 1 HiFi 60 C 5 region No. 5 (R) 5 -GAGGGTGTGAGGTAGACAGATTGTG 555 bp upstream to exon 1 5 region No. 4 (F) 5 -AGCTCAGAGAGGTAAAGTAATATGC 658 bp upstream to exon 1 HiFi 60 C 5 region No. 4 (R) 5 -GAGAACAACACGGGCCGCAGGAAGG 462 bp upstream to exon 1 5 region No. 3 (F) 5 -CTTTCAATAGCCCTGCACAATCTGT 546 bp upstream to exon 1 HiFi 60 C 5 region No. 3 (R) 5 -TGTGGTGTGTTCAATAATGATACGT 261 bp upstream to exon 1 5 region No. 2 (F) 5 -ATCTCCTCCAGGTCAAGACTCAATT 365 bp upstream to exon 1 HiFi 58 C 5 region No. 2 (R) 5 -CCTGTGAGGAGACAGAAAACAATAT 1 bp from transcription site in exon 1 5 region No. 1 (F) 5 -CCAACTCTCACACAGTGCCTACTA 114 bp upstream to exon 1 LT 60 C 5 region No. 1 (R) 5 -GTGTAGTTCAGGTGGCTACTCAGCT 51 bp from transcription site in exon 1 5 region No. 0 (F) 5 -GCCCCAAAGAGTCTGATTTTATTAA 24 bp upstream to exon 1 LT 60 C 5 region No. 0 (R) 5 -TCCTTCAGCACAGCCATGCACAGCA 180 bp from transcription site in exon 1 5 region No. 5 mp (F) 5 -TTACATGGGAAGAATCTGGAGTGCA 677 bp upstream to exon 1 Taq 60 C a LT: Expand Long Template PCR System; HiFi: Expand High Fidelity PCR System (Roche Diagnostics, Mannheim, Germany). The nucleotides underlined in 5 region No. 6 mp (mismatch) primer were artificially replaced from CT to TG. Figure 1 Direct sequencing of the identified variants in the 5 region and exon 1 of the DRD3 gene and genotyping using PCR- RFLP. Restriction enzymes used are shown above gel figures. M: 100-bp DNA ladder marker (New England BioLabs, Beverly, MA, USA). PCR primers used to genotype the polymorphisms are No. 5mp(F) and No. 5(R) (Table 1) for 712G/C, No. 2(F) and No. 2(R) for 205A/G, and No. 0(F) and No. 0(R) for Ala38Thr, respectively.

3 Table 2 Pairwise linkage disequilibria between four polymorphisms of DRD3. (Values above the diagonal are nonstandardized D- and standardized D/D max -values in the controls and values below the diagonal are those in the schizophrenia patients) 712G/C 205A/G S9G A38T 712G/C 0.20, , , A/G 0.19, , , 0.82 S9G 0.13, , , 1 A38T 0.03, , , 0.64 The four polymorphisms were in strong linkage disequilibrium with each other (Table 2). Specifically, the 712G/C and 205A/G polymorphisms were in almost complete linkage disequilibrium in both control and schizophrenic subjects. Genotype and allele distributions of these novel polymorphisms and the Ser9Gly polymorphism in 153 schizophrenia patients and in 122 controls are shown in Table 3. The genotype distributions of these polymorphisms did not deviate significantly from Hardy Weinberg expectations in the control or patient groups. The frequency of the Ser9 allele was significantly higher (P = 0.02) and that of the Ser9/Ser9 genotype tended to be higher (P = 0.09) in the schizophrenia patients than in the controls. Estimated haplotype frequencies are shown in Table 4. Estimated haplotype frequencies of all four polymorphisms were significantly different between the schizophrenia patients and controls (P = 0.015). The most marked difference in haplotype frequencies occurs for the C-G-Gly-Ala haplotype, which was more common in controls than schizophrenia patients. In addition, we compared estimated haplotype frequencies of 10 possible combinations of the four polymorphisms between patient and control groups (Table 4). Differences in estimated haplotype distributions of the 712G/C, 205A/G, and Ser9Gly polymorphisms were also significant (P = , corrected P = 0.007). Different allele frequencies of the Ser9Gly polymorphism between the schizophrenia patients and the controls largely resulted from the chromosomes carrying the C allele of the 712G/C polymorphism. Case-control genetic association studies are susceptible to the ability of population stratification to produce type 1 error. To confirm the findings suggesting the allelic and haplotype associations described above, we analyzed an additional 99 Japanese schizophrenia patients and 132 controls. The Ser9 allelic (P = 0.008) and genotypic (P = 0.04) association of the Ser9Gly polymorphism and haplotype association of the 712G/C, 205A/G, and Ser9Gly polymorphisms (P = ) were again suggested (Tables 5 and 6). In this study, we detected one novel missense polymorphism resulting in the substitution of threonine, a hydrophilic amino acid, for alanine, a hydrophobic amino acid, at codon 38 in exon 1 of the DRD3 gene. This polymorphism is located in the first transmembrane region of DRD3 and Ala38 is conserved in the green monkey, mouse, and rat. Although the Ala38Thr variant may have an altered receptor function, our subjects did not show any association between this polymorphism and schizophrenia. No homozygote for the Thr38 allele was found among our subjects. Shaikh and colleagues have reported a significant association between the Ser9 allele and the Ser9/Ser9 genotype of the Ser9Gly polymorphism with schizophrenia in 133 Caucasians. 19 In this and other metaanalyses, Ser9 and the Ser9/Ser9 genotype were found to be significantly associated with schizophrenia. 17,19 In this present study, we also observed an increased frequency of the Ser/Ser genotype and the Ser allele in 435 Table 3 Observed genotypic and allelic counts (frequencies) of the polymorphisms in the DRD3 gene in controls and schizophrenics Population Sample Genotype number (frequency) Heterogeneity Allele number Heterogeneity size test (frequency) test (n) The 712/G polymorphism G/G G/C C/C G G Controls (0.53) 43 (0.35) 14 (0.11) P (0.71) 71 (0.29) P 0.1 Schizophrenia patients (0.52) 67 (0.44) 7 (0.05) 225 (0.74) 81 (0.26) The 205A/G polymorphism A/A A/G G/G A G Controls (0.53) 40 (0.33) 17 (0.14) P (0.70) 74 (0.30) P 0.1 Schizophrenia patients (0.48) 70 (0.46) 9 (0.06) 218 (0.71) 88 (0.29) The S9G polymorphism S/S S/G G/G S G Controls (0.46) 51 (0.42) 15 (0.12) P = (0.67) 81 (0.33) P = 0.02 Schizophrenia patients (0.55) 61 (0.40) 8 (0.05) 229 (0.75) 77 (0.25) The A38T polymorphism A/A A/T T/T A T Controls (0.93) 8 (0.07) 0 (0.00) P (0.97) 8 (0.03) P 0.1 Schizophrenia patients (0.93) 11 (0.07) 0 (0.00) 295 (0.96) 11 (0.04)

4 436 Table 4 Estimated haplotype frequencies ± SD in the controls and schizophrenics a Controls Schizophrenics P (n = 244) (n = 306) (Corrected P) Haplotypes of four polymorphisms ( 712G/C, 20A/G, S9G, A38T) G-A-Ser-Ala ± ± C-G-Gly-Ala ± ± G-A-Gly-Ala ± ± C-G-Gly-Thr ± ± G-G-Ser-Ala ± ± C-G-Ser-Ala ± ± C-A-Gly-Ala ± ± G-A-Gly-Thr ± ± G-G-Gly-Ala ± ± C-G-Ser-Thr ± ± Haplotypes of the 712G/C, 205A/G, and S9G polymorphisms ( 712G/C, 205A/G, S9G) (0.007) C-A-Gly ± ± C-G-Ser ± ± C-G-Gly ± ± G-A-Ser ± ± G-A-Gly ± ± G-G-Ser ± ± G-G-Gly ± ± Haplotypes of the 712G/C and S9G polymorphisms ( 712G/C, S9G) (0.02) G-Ser ± ± G-Gly ± ± C-Gly ± ± C-Ser ± ± Haplotypes of the 250A/G and S9G polymorphisms 0.02 ( 712G/C, S9G) (0.2) A-Ser ± ± A-Gly ± ± G-Ser ± ± G-Gly ± ± a Among 11 possible combinations of the four polymorphisms, haplotypes with uncorrected P 0.05 were listed. Table 5 Replication of associations between the Ser9Gly polymorphism and schizophrenia in additional 132 controls and 99 schizophrenia patients Sample Genotype number (frequency) Heterogeneity Allele number Heterogeneity size test (frequency) test (n) S/S S/G G/G S G Controls (0.45) 58 (0.44) 15 (0.11) P = (0.69) 88 (0.31) P = Schizophrenia patients (0.62) 31 (0.31) 7 (0.07) 153 (0.82) 45 (0.18) schizophrenia patients as compared with controls. It is possible that the haplotype association observed in this study reflects an association of the Ser9Gly polymorphism with schizophrenia. In fact, haplotype frequencies that were significantly different between the subject groups were those containing the Ser9Gly polymorphism (Table 4), indicating that the Ser9Gly polymorphism is the greatest contributor of the four polymorphisms to the haplotype association. However, because most allelic differences of the Ser9Gly poly-

5 Table 6 Replication of the associations between the 712G/C, 205A/G, and S9G polymorphism haplotype and schizophrenia in an additional 132 controls and 99 schizophrenia patients Controls Schizophrenics P (n = 264) (n =198) ( 712G/C, 205A/G, (Estimated haplotype S9G) frequencies ± SD) C-A-Gly ± ± C-G-Ser ± ± C-G-Gly ± ± G-A-Ser ± ± G-A-Gly ± ± G-G-Ser ± ± G-G-Gly ± ± morphism between patient and control groups arose from the chromosome carrying the C allele of the 712G/C and the G allele of the 204A/G polymorphism, we conclude that the haplotypes of the 712G/C, 204A/G, and the Ser9Gly polymorphisms, rather than the Ser9Gly polymorphism alone, are associated with schizophrenia. We believe that unknown variant(s) in linkage disequilibrium with the haplotypes have an effect on susceptibility to schizophrenia. The haplotype association observed in this study is just a hint for further studies on the relation between the DRD3 gene and schizophrenia. The haplotype association needs to be replicated in independent samples. Since haplotype frequencies and linkage disequilibrium are often different between racial/ethnic groups, independent studies in various populations are important. Studies using large sample sizes are needed to detect effects of the numerous rare haplotypes. The structure of the promoter and regulatory regions of the human DRD3 gene have to be elucidated. Without the structural information of the gene, the functional significance of the 712G/C and 204A/G polymorphisms can not be addressed experimentally. Further studies will elucidate the role of the DRD3 gene in schizophrenia. Methods Subjects The schizophrenia patients analyzed in this study were 153 unrelated Japanese patients (88 men and 65 women with a mean ± SD age of 45.3 ± 12.1 years). They were recruited from a group of patients who were being treated at six hospitals near Tokyo in Japan. All patients satisfied the ICD-10 or DSM-III-R criteria for schizophrenia. The control subjects consisted of 122 unrelated Japanese (59 men and 63 women with a mean ± SD age of 47.2 ± 10.5 years). They were hospital staff members. An additional 99 Japanese schizophrenia patients (52 men and 47 women with a mean ± SD age of 47.8 ± 9.8 years) and 132 controls (73 men and 59 women with a mean ± SD age of 48.5 ± 11.3 years) were analyzed to confirm the positive findings observed in the first samples. This study was approved by the Ethics Committees of Tsukuba University and Tokyo Medical and Dental University. All subjects agreed to participate in this study and gave written informed consent. DNA analysis Genomic DNAs were prepared from peripheral whole blood cells by the standard phenol extraction method. Polymerase chain reaction (PCR) was used to amplify the 5 -flanking region ranging from 800 bp to the transcription site and exon 1 of the DRD3 gene. The genomic nucleotide sequence was obtained from GenBank accession numbers AF and HSU The PCR amplification conditions and primers are shown in Table 1. Genetic variants were screened in 48 randomlyselected schizophrenia patients by the SSCP method using the PhastSystem (Pharmacia, Piscataway, NJ, USA). Direct sequencing was carried out using a Dye Terminator Cycle Sequencing Kit and an ABI PRISM 310 DNA Sequencer (Perkin-Elmer, Oak Brook, IL, USA). The nucleotide variants detected in this study were genotyped by restriction fragment length polymorphism (RFLP) method after PCR amplification (Figure 1). Genotyping of the Ser9Gly polymorphism was carried out according to the method described by Lannfelt et al. 20 The fragments were separated on either 2% agarose gels or 3% agarose and 1.5% Nu-Sieve gels, and visualized by ethidium bromide staining and ultraviolet transillumination. Statistical procedures The following statistical procedures were carried out using Arlequin ver 1.1 ( unige.ch/arlequin). Deviation of genotype frequencies from Hardy Weinberg values was tested. Maximumlikelihood haplotype frequencies were estimated from the observed data using an expectation-maximization (EM) algorithm. Linkage disequilibrium values (D), standardized disequilibrium values (D = D/D max ), and significance of linkage disequilibrium were computed from estimated haplotype frequencies. Differences in observed allele and genotype frequencies and estimated haplotype frequencies between groups were tested for significance using the tests analogous to Fisher s exact test on a 2 2 contingency table or a 2 (No. of genotypes or haplotype) contingency table. P-values were obtained after more than Markov chain steps. A P-value 0.05 was considered significant in tests for Hardy Weinberg equilibrium. When genotypic and allelic distribution of each polymorphism was compared between the patient and control groups, no correction for multiple testing was made because the four polymorphisms examined in this study are in strong linkage disequilibrium each other. For comparisons of estimated haplotype frequencies, a P-value 0.05 was considered significant when haplotypes of all four polymorphisms were evaluated. When the haplotypes of two or three polymorphisms were selec- 437

6 438 ted from four polymorphisms for comparison, ten polymorphism combinations are possible and a P-value (0.05/10) was considered significant. References 1 Sokoloff P, Giros B, Martres MP, Bouthenet ML, Schwartz JC. Molecular cloning and characterization of a novel dopamine receptor (D3) as a target for neuroleptics. Nature 1990; 347: Schwartz JC, Levesque D, Martres MP, Sokoloff P. Dopamine D3 receptor: basic and clinical aspects. Clin Neuropharmacol 1993; 16: Schmauss C, Haroutunian V, Davis KL, Davidson M. Selective loss of dopamine D3-type receptor mrna expression in parietal and motor cortices of patients with chronic schizophrenia. Proc Natl Acad Sci USA 1993; 90: Griffon N, Sokoloff P, Diaz J, Levesque D, Sautel F, Schwartz JC et al. The dopamine D3 receptor and schizophrenia: pharmacological, anatomical and genetic approaches. Eur Neuropsychopharmacol 1995; 5 Suppl: Crocq MA, Mant R, Asherson P, Williams J, Hode Y, Mayerova A et al. Association between schizophrenia and homozygosity at the dopamine D3 receptor gene. J Med Genet 1992; 29: Nimgaonkar VL, Zhang XR, Caldwell JG, Ganguli R, Chakravarti A. Association study of schizophrenia with dopamine D3 receptor gene polymorphisms: probable effects of family history of schizophrenia? Am J Med Genet 1993; 48: Jonsson E, Lannfelt L, Sokoloff P, Schwartz JC, Sedvall G. Lack of association between schizophrenia and alleles in the dopamine D3 receptor gene. Acta Psychiatr Scand 1993; 87: Nothen MM, Cichon S, Propping P, Fimmers R, Schwab SG, Wildenauer DB. Excess of homozygosity at the dopamine D3 receptor gene in schizophrenia not confirmed. J Med Genet 1993; 30: Nanko S, Sasaki T, Fukuda R, Hattori M, Dai XY, Kazamatsuri H et al. A study of the association between schizophrenia and the dopamine D3 receptor gene. Hum Genet 1993; 92: Yang L, Li T, Wiese C, Lannfelt L, Sokoloff P, Xu CT et al. No association between schizophrenia and homozygosity at the D3 dopamine receptor gene. Am J Med Genet 1993; 48: Laurent C, Savoye C, Samolyk D, Meloni R, Mallet J, Campion D et al. Homozygosity at the dopamine D3 receptor locus is not associated with schizophrenia. J Med Genet 1994; 31: Chen CH, Liu MY, Wei FC, Koong FJ, Hwu HG, Hsiao KJ. Further evidence of no association between Ser9Gly polymorphism of dopamine D3 receptor gene and schizophrenia. Am J Med Genet 1997; 74: Hawi Z, McCabe U, Straub RE, O Neill A, Kendler KS, Walsh D et al. Examination of new and reported data of the DRD3/MscI polymorphism: no support for the proposed association with schizophrenia. Mol Psychiatry 1998; 3: Gaitonde EJ, Morris A, Sivagnanasundaram S, McKenna PJ, Hunt DM, Mollon JD. Assessment of association of D3 dopamine receptor MscI polymorphism with schizophrenia: analysis of symptom ratings, family history, age at onset, and movement disorders. Am J Med Genet 1996; 67: Di Bella D, Catalano M, Strukel A, Nobile M, Novelli E, Smeraldi E. Distribution of the MscI polymorphism of the dopamine D3 receptor in an Italian psychotic population. Psychiatr Genet 1994; 4: Williams J, Spurlock G, Holmans P et al. A meta-analysis and transmission disequilibrium study of association between the dopamine D3 receptor gene and schizophrenia [published erratum appears in Mol Psychiatry 1998; 3: 458]. Mol Psychiatry 1998; 3: Dubertret C, Gorwood P, Ades J, Feingold J, Schwartz JC, Sokoloff P. Meta-analysis of DRD3 gene and schizophrenia: ethnic heterogeneity and significant association in Caucasians. Am J Med Genet 1998; 81: Lundstrom K, Turpin MP. Proposed schizophrenia-related gene polymorphism: expression of the Ser9Gly mutant human dopamine D3 receptor with the Semliki Forest virus system. Biochem Biophys Res Commun 1996; 225: Shaikh S, Collier DA, Sham PC, Ball D, Aitchison K, Vallada H et al. Allelic association between a Ser-9-Gly polymorphism in the dopamine D3 receptor gene and schizophrenia. Hum Genet 1996; 97: Lannfelt L, Sokoloff P, Martres MP, Pilon C, Giros B, Pilon C et al. Amino acid substitution in the dopamine D3 receptor as a useful polymorphism for investigating psychiatric disorders. Psychiatr Genet 1992; 2: Correspondence: T Arinami, MD, Department of Medical Genetics, Institute of Basic Medical Science, University of Tsukuba, Tsukuba, Ibaraki 305, Japan. tarinami md.tsukuba.ac.jp Received 13 October 1999; revised and accepted 10 January 2000