Raymond De Paulo
Family, twin, and adoption studies provide strong evidence for a genetic etiology in bipolar disorder (BPD). Early studies seeking to locate genes reported very large statistical odds favoring linkage of BPD to genetic markers at two chromosomal sites. These reports were not replicated, and several full-genome scans have produced largely negative findings, leading to more than a few discouraging words. The genome scans done with larger sample sizes have made it clear that single-gene forms of BPD, if they exist, must be very uncommon. Nonetheless, positive linkage findings in multiple independent samples on chromosomes 4p, 4q, l0p, 12q, 13q, l8p, 18q, 21q, and 22q have emerged in recent years as attractive candidate regions for further study.
Clinical strategies for increasing the power of linkage and association studies include ascertaining very large family samples, and dividing the phenotypes and families into genetically meaningful subgroups to decrease heterogeneity. The publication of the draft sequence of the genome and over 300,000 single-nucleotide polymorphism markers provides a substantial new resource for these studies. The Human Genome Project will provide the completed sequences for all genes within a few years. In doing so, it will make the identification of candidate genes for BPD a routine step in future studies. I will summarize the results of the current studies at Johns Hopkins University using families, all ascertained through BPI probands but with differing patterns of clinical subtypes within the family members (i.e., containing BPII-BPII sib pairs versus multiple family members with psychotic affective disorders). We found that preferential transmission of paternal alleles on 18q21-22 is associated with the BPII sib pair families, while a different pattern may be seen in the families with multiple psychotic relatives. Delineation of families with differing genes could be predictive of differences in treatment response, as I will discuss.