Polymorphism in the 5′-promoter region of serine racemase gene in schizophrenia

Abstract

There are several lines of evidence suggesting that the regulation of D-aminoacids may play a role in development or modulation of schizophrenia. First, endogenous D-aspartate and D-serine have been recently found in mammalian nervous system. The D-serine is enriched in brain, particularly in prefrontal and parietal cortex, regions implicated in the pathogenesis of schizophrenia.1 Second, D-serine has potency as an endogenous coagonist at the glycine site of NMDA receptors. The hypofunction of NMDA receptor-mediated neurotransmission may contribute to development of schizophrenia.2, 3 Third, the genetic role of the G72 gene for the putative protein termed D-amino-acid oxidase activator (DAOA) and the gene for D-amino-acid oxidase (DAAO) were recently implicated in schizophrenia.4 The original genetic association between SNPs and haplotypes in DAOA and schizophrenia was supported by at least several subsequent studies of other case–control groups including schizophrenia and other psychiatric illnesses.5, 6 The data implicate that genes for proteins involved in metabolism of D-amino-acids, modulators of NMDA receptors, might potentially contribute to schizophrenia pathogenesis. The gene for mammalian serine racemase that catalyses the synthesis of D-serine from L-serine was recently described.7 We searched for variations in the 5'-promoter region of gene for human serine racemase located at 17p13 and characterized a common SNP in this region that may affect putative transcriptional regulatory sites. We examined the frequency of this SNP in schizophrenics and control individuals collected in Moscow. The original case–control sample was composed of 234 schizophrenic patients and 251 control individuals from Russia. The patients were recruited from psychiatric hospitals in Moscow. The control samples were from blood donors from Moscow with no psychiatric illnesses according to the medical records. The diagnosis of schizophrenia was made in accordance with International Classification of Disorders 10 (ICD-10) criteria. All patients in this study met criteria for paranoid schizophrenia (F20.0 in ICD-10). DNA was extracted from peripheral venous blood specimens using the phenol–chloroform method. We analyzed the polymorphism found in 5'-promoter region of serine racemase gene (G/C SNP at the 12 bp upstream of the most 5'-site of the mRNA SRR transcript, SNP at the position 2124325 NT_086857.1, dbSNP:408067). Screening of the 5'-region of the SRR (3100 bp) gene for putative regulatory sites using the MatInspector program from the Genomatix Software server (http://www.genomatix.de) identified that the short sequence stretch (gcgcagcG/Ccccgcccgccccaccc) harboring the G/C polymorphism similar to regulatory sequence elements for early growth factor response 1 (EGR1), activator protein 2 and other putative transcriptional factors. Direct and reverse primer oligonucleotides were designed for the polymorphic region to generate a PCR product of 329 bp and for the genotyping by DNA restriction analysis (HaeII endonuclease). Association analysis between the SNP allele or genotypes and schizophrenia were performed by exact tests and 2-test. Observed genotype frequencies showed no deviation from Hardy–Weinberg equilibrium in control and case samples (Table 1). The frequency of the genotypes showed that although the 5'-G/C SRR polymorphism is obviously not a major risk factor for schizophrenia, the modest contribution of the genotypes as a protective or vice versa disease risk factors have to be considered further. The statistically significant prevalence of the G allele was found in the total group of paranoid schizophrenia from the Moscow population group (Ms cohort) (P=0.024, OR=1.361, C.I.=1.042–1.777). The GG genotype was a more frequent than CC genotype in schizophrenia versus control (P=0.047, OR=1.774, C.I.=1.005–3.132). However, when several individuals with self-proxy determined non-Russian ethnicity (6% in the total Moscow schizophrenia sample) were excluded from the analysis (Ms-r cohort) the association between G allele and schizophrenia failed to reach the statistically significant value (P=0.068, OR=1.288, C.I.=0.981–1.690). The data demonstrate that minor ethnic heterogeneity or reduced power may affect the genetic association data, which are at the borderline significance. The stratification of schizophrenia group by age and gender had no significant effect on the results (Table 1). When the data were in preparation for publication other group reported no significant association between polymorphisms in SRR gene and schizophrenia in Japanese population.8 Although the 5'-SRR polymorphism alone is unlikely to be a strong risk factor for paranoid schizophrenia the etiology of schizophrenia may depend on a combination of variations in genes for proteins involved in metabolism of D-aminoacids or D-amino-acid regulated NMDA receptor signaling. The study of synergistic effects of biologically meaningful variations in genes involved in this pathway may be a promising strategy to define genetic risk factors for schizophrenia.