Raymond De Paulo
Department of Psychiatry; Johns Hopkins School of Medicine
Bipolar disorder and related depressive illnesses are among the most prevalent and costly of medical conditions. The absence of a pathology or pathophysiology for the illness have limited diagnostic procedures to the clinical evaluation. Treatments are empirical as the mode of action is unknown for the medications employed to treat it. The advent of molecular maps of the genome a decade ago focused attention on the potential of genetic studies to remedy this state of affairs especially given the clear evidence from twin and adoption studies for a powerful genetic contribution to the illness. However, finding genes for bipolar disorder has proved a much more daunting task than was anticipated. Several "false starts" in our field led to a more stringent realignment of the statistical requirements for claiming linkage. And even assuming that the 5 loci with "suggestive evidence for linkage" and support from independent samples all harbor genes which influence risk, the chromosomal regions are so large as to make substantial narrowing of them an uncertain proposition.
Our approach to this problem has been to undertake clinical analyses to identify genetic mechanisms which might tell us what type of gene to look for or to divide the phenotype into genetically meaningful subgroups to increase the homogeneity of our samples. Our group reported an apparent parent of origin effect with excess maternal transmission in families selected for our linkage studies. When this observation was factored into a linkage analysis of bipolar disorder and loci on chromosome 18, we found that the maternal chromosomes (and/or maternal families) provided no evidence for linkage while the paternal chromosomes provided significant evidence supporting linkage. This has been partially confirmed by our analysis of a second sample and has been supported as well by some but not all clinical and linkage studies from other groups. Whether this phenomenon is an artifact of ascertainment, an indication of genetic heterogeneity, or evidence for imprinting remains unknown to date. Resolution of this question could help in locating and identifying the genes relevant to bipolar disorder.