The Genom of Homo sapiens.pdf

386 PORTEOUS ET AL.Figure 2. The t(1:11) translocation and segregation with psychosis.Diagram of the t(1;11), which breaks at ch1q42 andch11q14. The diagnosis of patients (ascertained blinded to karyotypestatus) is tabulated. The LOD score for SCZ alone is 3.4and for SCZ plus BPAD and recurrent major depression is 7.1(Blackwood et al. 2001).The most parsimonious explanation for the correlationbetween the t(1;11) and psychosis is as a consequence ofDISC1/DISC2 gene disruption (Millar et al. 2003b), butbrief mention should be made of an alternative explanationproposed by Klar (2002), who seized on the observationthat about half of the t(1:11) subjects developed amajor psychotic diagnosis, and about half did not. Klar(2002) proposes an elegant, if elaborate, hypothesis thatrequires three hypothetical genes, DOH1, SEG, andRGHT. DOH1 is required for brain laterality specificationand is active on one chromatid, but switched off on theother by imprinting; SEG exists on the same chromosome,but is unlinked to DISC1; and a trans-acting factor,RGHT (right hander), utilizes SEG to govern nonrandomsegregation of sister chromatids at the critical period ofdevelopment when brain hemisphere asymmetry is determined.It is then hypothesized that the translocation separatesDOH1 from SEG. The net consequence of thet(1;11) according to this scheme is that 50% of individualscarrying the translocation will have abnormal brainlaterality and develop psychiatric illness, and the other50% will be normal. The clinical picture is not so simple.This five-generation family has been under clinical observationand care for over 30 years (Blackwood et al.2001). Overall, 62% of t(1;11) carriers are affected, but ifthe youngest generation is set aside (many of the individualsnot yet having reached the average age of onset at thetime of ascertainment), the figure rises to 70% of translocationcarriers being affected. When the P300 event-relatedpotential (P300 ERP), a trait marker of risk, is takeninto account, the proportion of affected translocation carriersincreases still further (Blackwood et al. 2001). TheP300 ERP is considered to be a measure of the pace andefficiency of information processing in the brain and isabnormal in translocation carriers, with or without a majorpsychiatric diagnosis. This indicates the presence ofcentral nervous system abnormalities in all translocationcarriers, including those who are clinically unaffected,and is thus inconsistent with the strand segregation hypothesis.In contrast, the proportion of clinically affectedtranslocation carriers and the P300 ERP data are consistentwith our model of dominant inheritance coupled withreduced penetrance and variable expressivity dependenton the action of modifiers (genetic and/or environmental).Indeed, we propose that disruption of DISC1 andDISC2, plus the actions of genetic and environmentalmodifying factors, are sufficient to explain the psychiatricillness arising from inheritance of the t(1;11) chromosome(Millar et al. 2003b). In support of this argument,the relative risk of the t(1;11) is equivalent to thatof the monozygotic co-twin of an affected individual.Replication of the Genetic Evidence for the DISC1Locus as a Risk FactorThe most convincing independent genetic evidence sofar for the involvement of the TRAX/DISC1 region inmental illness has come from the Finnish population. Initiallinkage evidence was found in an internal isolate ofFinland (LOD = 3.7) and a common haplotype spanning6.6 cM reported in 3 of the 20 families multiply affectedwith SCZ and schizoaffective disorder (Hovatta et al.1999). A later analysis of 134 sib pairs collected fromthroughout Finland gave additional support for the same1q32.2-q41 region on Chromosome 1 with a maximumLOD of 2.6 for a diagnostic class that included SCZ,schizophrenia spectrum disorders, and bipolar and unipolardisorders (Ekelund et al. 2000). Fine mapping of bothFinnish populations resulted in maximum LOD scores forSCZ and schizoaffective disorder at D1S2709, a microsatellitelocated in intron 9 of DISC1, in both the combinedsample (Z max = 2.71) and outside the internal isolate(Z max = 3.21) (Ekelund et al. 2001).Other data implicating this region in psychosis havecome from linkage analysis in families of diverse ethnicorigin. Detera-Wadleigh et al. (1999) reported a maximumLOD of 2.67 in a 30-cM region spanning 1q25-1q42 in bipolar families, some of whose members wereaffected by SCZ or schizoaffective disorder. Gejman etal. (1993) reported a maximum LOD of 2.39 at D1S103(1q42.2) in one North American family with bipolar disorder.Most recently, a suggestive nonparametric linkagescore (NPL = 2.18), equating approximately to a LODscore of 1.2, has been reported in Taiwanese families withdiagnoses of SCZ, schizoaffective disorder, and othernon-affective psychotic disorders (Hwu et al. 2003). Ourown studies have also found evidence for linkage in asubset of BPAD families in Scotland (S. MacGregor etal., unpubl.).Finally, as discussed earlier in the context of the t(1;11)clinical follow-up study, the use of endophenotypes inmolecular studies may overcome the effect of reducedpenetrance on the ability to detect linkage and associationand help to resolve the affected status of family members.In this context, it is worth mentioning that Gasperoni etal. (2003) undertook a QTL analysis of spatial workingmemory, an endophenotype for SCZ (Cannon et al. 2000;Glahn et al. 2003). Gasperoni et al. (2003) provide evi-