Scientific Correspondence 448 Molecular Psychiatry genotype distributions of this SNP do not deviate from Hardy–Weinberg equilibrium in both patients and con- trols. The genotype and allele frequencies of this SNP in both patients and controls were listed in Table 1. No significant differences of genotype or allele fre- quency were detected between patients and controls, suggesting that this SNP is unlikely involved in the pathogenesis of schizophrenia. Reeler mouse is an autosomal recessive mutant mouse with defective reelin gene. Heterozygous reeler mouse showed 50% decrease of reelin mRNA and pro- tein in the brain, deficits in prepulse inhibition of startle, 7 indicating a possible link of genetic defects of reelin gene and pathogenesis of schizophrenia. In humans, mutations in the reelin gene were found to be associated with lissencephaly, a rare autosomal recess- ive neurological disease with cerebellar hypoplasia. 8 Surprisingly, no evidence of increased risk to schizo- phrenia was found in individuals with heterozygous or homozygous human reelin mutations. 8 In this study, we report no association of a novel putative functional SNP at the 5'-UTR of human reelin gene with schizophrenia. Our result should be inter- preted with caution. First, SSCP is not a perfect mutation screening method, there might be other func- tional polymorphisms missed in this study. Second, human reelin gene is a very large gene, there might be other functional mutations in the protein-coding sequences and exon-intron splicing sites that may down regulate the reelin expression in schizophrenic patients. Third, in addition to heterozygous reeler mutant mouse, there are several other animal models with reduced expression of reelin, such as prenatal human influenza virus infection, 9 experimental hypo- thyroidism, and chemical lesions. 10 Hence, in addition to defective reelin gene, environmental factors as men- tioned above should be also taken into consideration when investigating the etiology of the reduced reelin expression in schizophrenic patients. Finally, although the allelic association of this SNP with schizophrenia does not reach statistic significance (P = 0.08) in our population, it is warranted to replicate this association Table 1 Genotype and allele frequencies of the single nucle- otide polymorphism g.-888GC of the reelin gene in schizo- phrenic patients and controls Polymorphism Schizophrenia Control  2 df P n = 279 n = 255 g.-888GC Genotype 2.99 2 0.22 GG 243 (87.1%) 209 (81.9%) GC 34 (12.2%) 42 (16.5%) CC 2 (0.7%) 4 (1.6%) Allele 3.16 1 0.08 G 520 (93.2%) 460 (90.2%) C 38 (6.8%) 50 (9.8%) study in other population and subgroups of schizo- phrenia. M-L Chen 1 , S-Y Chen 2 , C-H Huang 3 and C-H Chen 3 1 Institute of Medical Sciences, Tzu-Chi University, Hualien City, Taiwan; 2 Department of Life Sciences, Tzu-Chi University, Hualien City, Taiwan; 3 Depart- ment of Psychiatry, Tzu-Chi General Hospital, Hualien City, Taiwan Correspondence should be addressed to: C-H Chen. E-mail: cchen mail.tcu.edu.tw 1 Impagnatiello F et al. Proc Natl Acad Sci USA 1998; 95: 15718– 15723. 2 Guidotti A et al. Arch Gen Psychiatry 2000; 57: 1061–1069. 3 Fatamil SH et al. Mol Psychiatry 2000; 5: 654–663. 4 Persico AM et al. Mol Psychiatry 2001; 6: 150–159. 5 Curran T, D’Arcangelo G. Brain Res Rev 1998; 26: 285–294. 6 Prestridge D. CABIOS 1991; 7: 203–206. 7 Tueting P et al. Neuroreport 1999; 10: 1329–1334. 8 Hong SE et al. Nat Genet 2000; 26: 93–96. 9 Fatemi SH et al. Mol Psychiatry 1999; 4: 145–154. 10 Fatemi SH. Mol Psychiatry 2001; 6: 129–133. Epistatic effect of genes from the dopamine and serotonin systems on the temperament traits of Novelty Seeking and Harm Avoidance Molecular Psychiatry (2002) 7, 448–450. doi:10.1038/ sj.mp.4001005 SIR – Extremely high and low scoring individuals on the temperament dimensions of novelty seeking and harm avoidance were genotyped for three candidate genes involved in the dopaminergic and serotonergic systems. Associations were found between the dopa- mine transporter gene and novelty seeking, and between the serotonin transporter gene and harm avoidance. The associations were the strongest in females. In addition, the associations differed although not significantly, in groups of individuals with or with- out the 7-repeat allele of the dopamine D4 receptor gene, suggestive of an epistatic effect. Genetic studies on temperament have been focused mainly on the dimensions of novelty seeking (NS) and harm avoidance (HA). High NS was found to be asso- ciated with the 7-repeat allele of a 48-bp polymorphic repeat in the dopamine D4 receptor gene (DRD4), 1,2 and high HA with the short allele of a 44-bp insertion/deletion polymorphism in the upstream regulatory region of the serotonin transporter gene (5- HTT). 3 In addition interaction has been reported between the 5-HTT and DRD4 polymorphisms in NS and HA. 4–6 Since the DRD4 and 5-HTT associations were not found in cross cultural and ethnic groups,