Two recent meta-analyses supported possible linkage at 1q, 2q, 3p, 5q, 6p, 8p, 11q, 13q, 14p, 20q and 22q, though they only agreed with respect to 8p and 22q (Badner and Gershon, 2002; Lewis et al., 2003), exemplifying the problems of linkage analysis for schizophrenia. Linkage studies have also not yet identified specific risk genes, though candidates such as neuregulin 1 (NRG1) and catechol O-methyltransferase (COMT) are present within the 8p and 22q candidate loci. It remains unclear, however, if this reflects a true association.
Family studies suggest that there may be many genetic loci involved in the genesis of schizophrenia and that no one gene is likely to confer greater than a threefold increased risk for schizophrenia (Risch, 1990). However, despite the relatively small effect of any one gene, there is a significant collective genetic risk for schizophrenia, with estimates of heritability as high as 80% (Cardno and Gottesman, 2000). Twin studies have provided the most significant evidence for a genetic component, with concordance between monozygotic twins in the range of 50%, even when raised from birth in different homes (Kendler et al., 1994; Tienari et al., 2004). This, however, leaves 50% discordance between monozygotic twins, which appears to be considerably more discordance than might be predicted from the calculated heritability of 80%.
Epigenetic influences, which alter gene expression through gene methylation, histone de-actylation and other regulatory mechanisms, may account for some of the difference. Evidence supporting the role of epigenetic effects is presented in a study that found increased methylation in a male without schizophrenia for the dopamine(Drug information on dopamine) receptor gene DRD2, as compared to his discordant monozygotic twin brother with schizophrenia and another sibling with schizophrenia (Petronis et al., 2003).
Another possibility is that the difference between monozygotic twins reflects probabilistic, rather than mechanistic, outcomes. This hypothesis, reminiscent of quantum theory, suggests that as the number of risk genes increases, so too does the probability of developing schizophrenia, but there is no specific threshold that must be crossed before disease is expressed (Procopio, 2005).
With respect to actual risk genes for schizophrenia, numerous genes have been implicated but the evidence remains spartan. The strongest evidence currently available is for the genes dysbyndin and neuregulin 1 (NRG1) with somewhat less support available for COMT, DISC1, mGluR3, proline dehydrogenase (PRODH), G72, DAO and RGS4 (Harrison and Weinberger, 2005) (Table). The most likely candidates at this time include dysbyndin. This gene, found on chromosome 6p22.3, has widespread distribution throughout the brain. Reduced levels of dysbyndin protein and mRNA expression have been found in the dorsal prefrontal cortex of postmortem brains of patients with schizophrenia (Weickert et al., 2004). The function of this gene is speculative, but it has been suggested that it may be involved in synaptic glutamate release (Numakawa et al., 2004). However, despite studies showing an association, no mutation has yet been conclusively associated with schizophrenia.
Neuregulin 1 has also been associated with schizophrenia in several populations (Harrison and Weinberger, 2005) and appears to express multiple proteins with many functions relevant to neurodevelopment, plasticity and transmitter activity (Corfas et al., 2004). However, it remains unclear which are of relevance.
Catechol O-methyltransferase appears plausible because of mapping and deletion studies (particularly associated with velocardiofacial syndrome, which has a very high risk for psychotic disorders including schizophrenia) (Schosser and Aschauer, 2004). In addition, COMT is one of the two enzymes degrading catecholamines such as dopamine. Therefore, it plays a large role in cortical dopamine metabolism; dopamine being a central component of current theories of schizophrenia genesis (Keshavan, 1999). However, evidence regarding allelic variants and protein expression/activity differences in schizophrenia remain conflicting.
Indeed, this appears to be the case for other candidate genes (Craddock et al., 2005; Harrison and Weinberger, 2005). Their potential role as neurotransmitter modulators is biologically plausible because neurotransmitters, particularly dopamine, γ-aminobutyric acid (GABA) and glutamate, have all been strongly implicated in schizophrenia (Coyle, 2004; Kalkman and Loetscher, 2003).
