Background: Meta-analyses have suggested an association between schizophrenia (SZ) and a coding polymorphism (rs6280/Ser9Gly) at the dopamine D3 receptor gene (DRD3), but results have been inconsistent. Because most studies have evaluated only rs6280, the inconsistencies might reflect associations with other variants.
Methods: We analyzed polymorphisms spanning 109kb in two independent samples (United States: 13 single nucleotide polymorphisms (SNPs), 331 cases, 151 trios, 274 control subjects; India: 11 SNPs, 141 trios).
Results: In the U.S. samples, significant associations were detected with eight SNPs, including rs6280 (p = .001, odds ratio: 1.5). Consistent associations in the case-control and family-based analyses were detected with a common haplotype spanning intron 1 to the 3' region of the gene (rs324029-rs7625282-rs324030-rs2134655-rs10934254; case-control, p = .002; transmission disequilibrium test [TDT], p = .0009; global p-values = .002 and .007, respectively). In the Indian sample, one SNP was associated (rs10934254, p = .03). Moreover, over-transmission of the same common haplotype as the U.S. sample was observed in this cohort (TDT, p = .005; global test, p = .009). Ser9Gly (rs6280) was associated with SZ against this haplotype background but not other haplotypes.
Conclusions: These data suggest previous inconsistencies might have resulted from associations with other DRD3 variants. A liability locus might be in linkage disequilibrium (LD) with or carried against, an associated haplotype 3' to rs6280. Comprehensive SNP evaluation in larger samples is needed.