We undertook a genomewide linkage study in a total of 353 affected sib pairs (ASPs) with schizophrenia. Our sample consisted of 179 ASPs from the United Kingdom, 134 from Sweden, and 40 from the United States. We typed 372 microsatellite markers at approximately 10-cM intervals. Our strongest finding was a LOD score of 3.87 on chromosome 10q25.3-q26.3, with positive results being contributed by all three samples and a LOD-1 interval of 15 cM. This finding achieved genomewide significance (P<.05), on the basis of simulation studies. We also found two regions, 17p11.2-q25.1 (maximum LOD score [MLS] = 3.35) and 22q11 (MLS = 2.29), in which the evidence for linkage was highly suggestive. Linkage to all of these regions has been supported by other studies. Moreover, we found strong evidence for linkage (genomewide P<.02) to 17p11.2-q25.1 in a single pedigree with schizophrenia. In our view, the evidence is now sufficiently compelling to undertake detailed mapping studies of these three regions.

The high-activity homozygotes showed significantly higher scores of aggression, whereas the heterozygotes showed significantly lower scores. The odds ratio for aggression for the high-activity homozygotes was 2.07 (95% Cl=1.03-4.15), whereas that for the heterozygotes was 0.54 (95% Cl=0.30-1.00). CONCLUSIONS; The high-activity COMT homozygote confers a higher risk of recorded aggression in schizophrenia. Heterozygotes had a significantly lower risk, which may represent an example of heterosis/heterozygote advantage.

Some studies have reported associations between COMT and MAO genotypes and aggression, though results have been inconsistent. We examined the relationship between Overt aggression scale (OAS) scores, and both MAOA and MAOB polymorphisms in a well-powered sample of 346 subjects with schizophrenia. We also examined COMT in a Stage II replication sample of 150 individuals, and combined these results with our previously reported (Stage I) findings for COMT. We found no evidence of any associations between OAS ratings and any of the polymorphisms investigated under different genetic models. There was no evidence of epistatic interaction between MAOA and COMT on OAS scores. These results fail to support the theory that functional polymorphisms within the MAOA, MAOB, or COMT genes, as determinants of catecholamine enzymatic activity, are risk factors for aggressive behavior.

Homer proteins are a group of proteins that regulate group 1 metabotropic glutamate receptor function. As altered glutamate function has been implicated in many neuro psychiatric disorders, particularly schizophrenia, we have screened all three known Homer genes for sequence variation for use under the candidate gene association paradigm. We found seven SNPs, including three in exons. Of these, none was non-synonymous. Allele frequencies of all the detected SNPs were estimated in DNA pools of 368 schizophrenics and 368 controls. Only one (Homer 1 IVS4 + 18A > G) was associated with schizophrenia in this sample, a finding confirmed by individual genotyping (P = 0.01). However, in our extended sample of 680 cases and 671 controls, the evidence for association diminished (P = 0.05). Our results suggest it is unlikely that sequence variants in the Homer genes contribute to the aetiology of schizophrenia, but the variants we identified are plausible candidates for other neuropsychiatric phenotypes.

Several lines of evidence suggest that psychosis is associated with altered dopaminergic neurotransmission. Dopamine is catabolized by monoamine oxidase (MAO) and catechol-O-methyl transferase (COMT). We hypothesized that the genes encoding MAOA and COMT might contain genetic variation conferring increased risk to schizophrenia. In order to test this hypothesis, we genotyped the 941T > G and the promoter VNTR polymorphisms in the MAOA gene and the V158M COMT polymorphism in 346 DSMIV schizophrenics and 334 controls. We also genotyped the-287A > G COMT promoter polymorphism in 177 schizophrenics and 173 controls. No significant differences were found in allele or genotype frequencies between affecteds and controls for any of the polymorphisms. As both genes are involved in degrading catecholamines, we also sought evidence for additive and epistatic effects but none was observed. Our data, therefore, do not support the hypothesis that genetic variation in MAOA and COMT is involved individually or in combination in the etiology of schizophrenia.