This collaboration has developed a bipolar disorder family resource that has proven to be uniquely valuable in testing clinically-based hypotheses about the genetic heterogeneity of the illness. We have ascertained and assessed about 150 nuclear families with a bipolar I proband and two or more siblings with a major affective disorder. The analysis of clinical features in the ascertained families has revealed important patterns, including bipolar II, psychotic bipolar disorder, and panic disorder comorbidity. One important goal is to derive and test genetically meaningful phenotypic subtypes. Tests of familial aggregation of variables and factors are also being performed, and positive results will be followed by covariate analyses of genetic linkage and association data. We previously reported results of an association study of close to 5000 SNP markers on chromosomes 13, 18, and 22, as well as in selected candidate genes. Preliminary results implicated several genes in bipolar disorder that are good candidates for replication in independent samples. We have found evidence that risk of bipolar disorder in these families is associated with genetic variation in the gene encoding FKBP5, genes encoding proteins involved in the Wnt signaling pathway, and genes that influence circadian rhythms. We have also performed association studies through SNP genotyping (using the Illumina platform) in 4 chromosomal regions: 6q21-23, 13q31-33, 18q21-22 and 22q12. Analyses of the SNP data have revealed several signals that will be followed up with replication in an independent sample. We have also generated a database of clinical variables, The Bipolar Disorder Phenome Database, based in part on this sample. The database contains several hundred clinical variables, harmonized with those collected by the NIMH Genetics Initiative, describing the course, symptoms, and clinical picture of bipolar disorder. We have made the database available to the scientific community to support research into the genetics of bipolar disorder and related conditions. For example, we used these data to demonstrate that postpartum mood symptoms aggregate in bipolar disorder pedigrees, suggesting that these symptoms may help define a familial subtype of the disorder. In recent years, in collaboration with scientists at Cold Spring Harbor Laboratory, we have used cutting-edge genetic tools to search for relatively small structural variations in chromosomes, known as copy number variants(CNVs), that may play a role in bipolar disorder. Using these tools, we found evidence that some CNVs strongly over-represented among people with schizophrenia may also play a role in bipolar disorder. Work over the past year was focussed on detecting de novo CNVs that may give rise to bipolar disorder, especially in an early-onset form, among the offspring of parents who do not suffer from the condition. We postulated that singleton probands with unaffected parents may be enriched for rare, de novo CNVs, but these are scarce, so we pulled together 4 other groups to help us with sample collection and diagnostic assessment. CNV screening was done in Jonathan Sebats lab. We found a slight excess transmission of the large chromosome 16p11.2 microduplication in 161 trio families, but the difference became statistically significant only after adding over 3,000 additional cases in a meta-analysis. It is possible that an excess of smaller, de novo CNVs may be detectable in bipolar disorder, especially if we increase the sample size for study. This is the focus of our ongoing collaboration with the Sebat lab, which is using higher resolution oligo arrays that have better sensitivity for smaller CNVs, down to 10 kb in size, in a larger sample of trios.
Showing the most recent 10 out of 56 publications
