Schizophrenia is a debilitating mental illness and disturbance in Gamma-aminobutyric acid (GABA) transmission systems have been previously observed in schizophrenics. To test the potential involvement of the GABA_A receptors, eighteen single nucleotide polymorphisms (SNPs) in the α₁, β₂, γ₂, ε and π subunit genes of the receptor were assessed for association with schizophrenia. Five of the SNPs in intron 8 of the β₂ subunit gene showed significant associations in Han Chinese (schizophrenia 170; control subjects 120). This initial finding was subjected to cross-validation with Japanese (JP) and German Caucasian (GE) populations....[ Read more ]

Schizophrenia is a debilitating mental illness and disturbance in Gamma-aminobutyric acid (GABA) transmission systems have been previously observed in schizophrenics. To test the potential involvement of the GABA_A receptors, eighteen single nucleotide polymorphisms (SNPs) in the α₁, β₂, γ₂, ε and π subunit genes of the receptor were assessed for association with schizophrenia. Five of the SNPs in intron 8 of the β₂ subunit gene showed significant associations in Han Chinese (schizophrenia 170; control subjects 120). This initial finding was subjected to cross-validation with Japanese (JP) and German Caucasian (GE) populations.

Un-related individuals of JP (schizophrenia, 304; control subjects, 207) and GE (schizophrenia, 301, including 155 of unsystematic schizophrenia and 146 of systematic schizophrenia; control subjects, 190) origin were recruited. SNP discovery and genotyping were carried out through resequencing of a 1,839 base pair (bp) region in GABRB2. Tagging-SNPs (tSNPs) were selected based on linkage disequilibrium (LD), combinations of which were analyzed with Bonferroni-correction and permutation for disease association. Random re-sampling was applied to generate size- and gender-balanced case and control groups to minimize the potential effects of unequal sample size and imbalanced gender ratio. Seventeen SNPs were revealed (9.2 SNPs/ 1,000 bp), including eleven known and six newly-discovered. Out of the seventeen, six were population-specific. Within this apparently high LD region, at least two low LD points were identified in both populations. However, population variations in LD were observable. Although disease association at single SNP level were only shown in GE, strong disease associations of tSNP haplotypes were demonstrated in both JP (P = 0.0002 - 0.0191) and GE (P = 0.0033 - 0.0410) subjects, centering on haplotypes bearing rs1816072 and rs1816071. Among different clinical subtypes, the most significant association with genotypes was exhibited by systematic schizophrenia. Cross-population validation of GABRB2 association with schizophrenia has been obtained with JP and GE subjects, the genotype-disease correlations being strongest in systematic schizophrenia, the most severe subtype of the disease.

To investigate the potential biological significance of these disease associated variants in GABRB2, signatures of selection were examined in the region. A 3,551-bp region harboring the disease associated variants was completely re-sequenced in 638 human subjects of four populations, African (AF), GE, Caucasians from United States (US) and JP. Comparative genome analysis of human and non-human primate sequences revealed a human-specific Alu insertion in intron 8. Among the human populations, four schizophrenia-associated SNPs, rs6556547, rs187269, rs1816071, and rs1816072 showed high frequency of derived allele in at least three of the four populations. The contemporary major human haplotype bearing the derived alleles of the four SNPs accounts for > 50 % of total chromosomes, while the other 74 inferred haplotypes share the remaining ~ 50 %. This sharp contrast of haplotype diversity suggested positive selection on the four derived alleles and the human major haplotype. Positive selection on the major haplotype was tested against the effects of genetic drift and demographic history. Significant positive selection was essentially obtained for GE, US and JP, but not in AF. This population variation proposed the selection is a rather recent event. The findings by Zhao et al. (2006), shed light on the biological implications of the observed selection.