British Journal of Psychiatry (1991), 158, 605—610 Molecular Biology Symposium Collaborative Strategies in the Molecular Genetics of the Major Psychoses MARION LEBOYER and PETER McGUFFIN Most authorities now agree that a genetic contribution is the best established aetiological factor in manic depressive illness (McGuffin & Katz, 1989) and schizophrenia (Gottesman & Shields, 1982). In both cases the size of the genetic contribution is sub stantial, as reflected in estimates of broad heritability (the proportion of total phenotypic variance accounted for bygenes) of 60—80%.However, the evidence in favour of a genetic contribution from family, twin and adoption studies tells us that we are dealing with complex conditions which do not have simple Mendelian modes of transmission and in which environmental influences almost certainly play a part. The main debate about mode of transmission has centred upon whether it is entirely polygenic (i.e. dependent upon the action of many different genes, each of small effect) or whether for each of the major psychoses there is a gene of major effect. Many proponents of the single major effect believe that heterogeneity is probable, so that schizophrenia and manic—depressiveillness will each turn out to be a group of conditions, at least some of which show inheritance of the major-effect gene, even if other forms are purely polygenic or even predominantly environmental. The attraction of the hypotheses of a single major effect is that, given the recent advances in recombinant DNA technology, they provide a starting point for a proper understanding of the aetiology of the psychoses. It is now theoretically feasible for any illness which shows inheritance of a major-effect gene but has an otherwise obscure aetiology to detect linked genetic markers, to move from the markers to the gene itself and thence to discover the abnormal (or missing) gene products responsible for the disorder. This ‘¿ reverse genetic' approach holds great promise and has already produced strikingly successful results in monogenic conditions such as Duchenne's muscular dystrophy (Hoffmann & Kunkel, 1989) and cystic fibrosis (Riordan et al, 1989). Until fairly recently the search for genes of major effect in complex conditions such as schizophrenia and manic—depressive illness has relied upon studies of the segregation of the disorders in families and the application of statistical methods which compare the theoretical expectations under a variety of models with what is actually observed. The results have largely been disappointing. Complex segregation analysis has provided limited support for inheritance of such a gene in manic—depression(O'Rourke et al, 1983; Rice et al, 1987) but has provided no clear evidence either for or against this in schizophrenia (Carter & Chung, 1980; Vogler et al, 1990). While it seems certain that a ‘¿ major' gene on its own cannot explain the transmission of schizophrenia, the possibility of such a gene operating together with polygenic or familial environmental effects cannot be excluded (McGue et al, 1985). Thus there is no very convincing evidence in favour of genes of major effect in the inheritance of manic—depression or schizophrenia, but neither is there strong evidence against the existence of such genes. With this background, and given the very rapid advances which have taken place in mapping the human genome (i.e. in discovering and locating linkage markers throughout the length of the 22 pairs of autosomes and the sex chromosomes), there has been considerable optimism about the potential of genetic linkage studies in complex conditions such as the psychoses. In linkage studies, the usual aim is to study the cosegregation of genetic markers and the disease within families in which multiple members are affected. The tendency for a particular marker allele and the disease to be inherited together within a pedigree can suggest that the disease and the marker locus are close together on the same chromosome. Unfortunately, so far in respect of the psychoses, the results of linkage marker studies have been contra dictory and confusing. The promising finding of linkage between manic—depressionand markers on chromosome 11 (Egeland et al, 1987) was not replicated elsewhere (Detera-Wadleigh et al, 1987; Hodgkinson et al, 1987; Gill et al, 1988) and further studies on the large pedigree in which chromosome-i 1 linkage was first reported cast doubt on the original findings (Kelsoe et al, 1989). There have been similarly conflicting results regarding the claims of X linked genes for manic—depression(e.g. Mendlewicz et al, 1987; Berettini et al, 1990). In respect of schizophrenia, the most provocative finding has concerned a report of a linkage marker on chromosome 5 (Sherrington et al, 1988). Again, failures of replication followed (Kennedy et al, 1988; St Clair et al, 1989)and recently reported data, combined 605