IRB Protocol 80-M-0083 NCT00001174 Starting in 1993, in collaboration with 10 academic centers across the United States, we recruited a large sample of over 3,000 individuals with bipolar or related mood disorders. All participants did a diagnostic interview and provided a blood sample for DNA analysis. DNA and clinical data are available through the NIMH Center for Genetics. Genetic linkage studies suggested several chromosomal regions may contain genes that contribute to mood disorders in this sample. To identify individual causal genes, we conducted the first genome-wide association study of bipolar disorder in 2007. The results implicated several genes, each of small effect, suggesting that bipolar disorder is a polygenic disease. Our 2011 meta-analysis of independent case-control studies of bipolar disorder supported association with a cluster of genes on chromosome 3p21, markers near TRANK1, LMAN2L, and PTGFR, and 2 independent regions of ANK3 implicated in previous studies. Many of these findings have now been replicated in independent samples. Ongoing work, supported by NARSAD, is using genome-wide association methods in large samples of cases and controls to detect additional risk genes on the X-chromosome, which has so far been little studied in bipolar disorder by GWAS methods. We have shown that models based on large numbers of markers can distinguish between cases and controls in independent datasets with high significance, but only modest predictive value. These studies also suggest that common alleles predisposing to bipolar disorder also predispose to major depression and schizophrenia, but not to neurological diseases such as Parkinson Disease. Collaborative work we did with the Cross-Disorder Group of the Psychiatric Genomics Consortium has confirmed that there are substantial genetic correlations between schizophrenia and bipolar disorder, schizophrenia and major depressive disorder, bipolar and major depressive disorder, and between ADHD and major depressive disorder. These results point to shared genetic risk factors among major psychiatric disorders that may reflect shared pathophysiologies. To identify genetic variants that may have a larger impact on individual risk for major mood disorders, we have undertaken genome sequencing studies in selected populations with reduced genetic diversity and large families. Large families increase the opportunities for ascertaining distant relatives with a mood disorder, and individuals belonging to different nuclear families are often related, forming an extended kindred ideal for genome sequencing studies. So far we have collected close to 1400 individuals from Amish and Mennonite communities whose unique genetic history makes them especially good candidates for this kind of study. All blood samples are processed by the Rutgers Cell and DNA Repository who also establish lymphoblastoid cell lines and distribute DNA as a resource for the general scientific community. We will also continue to collect families as they are identified. Our goal is to collect at least 100 additional cases, along with their parents and offspring. Offspring are studied in collaboration with Dr Leibenluft's group at NIMH IRP, which performs additional cognitive and emotional testing on these high-risk children. Genome sequencing began in 2012 using technology that captured only the exome, or the expressed portions of the genome. So far we have completed about 80 exome sequences in individuals of Amish or Mennonite ancestry. Although no rare, damaging mutations have been found that are shared by the majority of cases in our study, we have identified over 1000 promising variants shared by distantly-related cases. These have been submitted to the Bipolar Sequencing Consortium (BSC) where they will become part of a large meta-analysis of sequenced cases, controls, and families. Skin biopsies are obtained on sequenced individuals and converted to fibroblasts. Several fibroblast lines have been reprogrammed into induced pluripotent stem cells (iPSc) for functional genomic studies. Last year, with support and collaboration from the Institute of Systems Biology (ISB), we completed whole genome sequence on 105 individuals from the collection. Ongoing analyses are aimed at detecting genetic variants that influence BD risk by virtue of their impact on gene regulation. We are also examining small deletions and insertions that can be readily detected in whole genome sequence. Those that overlap known genes and tend to segregate with mood disorders in families may be particularly important. Also in collaboration with ISB, and the Bipolar Disorder Genome Study (BiGS), we have analyzed whole-genome sequences performed on 200 members of 41 families multiply affected with bipolar disorder. Several classes of coding and non-coding (regulatory) variants segregating in these pedigrees were enriched for neuronal excitability genes. Most affected individuals inherited several of these variants, suggesting that genetic risk is generally oligogenic. Targeted sequencing of 26 neuronal excitability genes in an additional 3457 cases and 2384 controls confirmed rare-variant burdens in ANK3, CACNA1B, CACNA1C, CACNA1D, CACNG2, and CAMK2A. Variants in promoters and 3 and 5 untranslated regions contributed more strongly than coding variants to risk for bipolar disorder, both in pedigrees and in the case-control cohort. Ongoing work is aimed at confirming these findings in additional samples and characterizing the functional impact of the implicated genetic variants. If confirmed, these results could have important implications for our understanding of the causes of bipolar disorder and provide clues for better treatment. We are also searching for genetic markers that might help predict an individual's response to lithium, one of the most effective known treatments for bipolar disorder. To this end, we organized a large international collaboration, known as the Consortium on Lithium Genetics (ConLiGen), which aims to characterize lithium response in a large group of patients using reliable instruments, and then perform a genome-wide association study. In collaboration with the University of Bonn, we completed genome-wide SNP array genotyping on over 3000 cases. Association analysis was performed using definitions of lithium response that showed high inter-site reliability. We identified a genome-wide significant signal in a region on chromosome 21 that appears to encode a long-noncoding RNA molecule. Individuals with the response-associated markers were 50% more likely to respond to lithium. Ongoing work will test this finding for replication in additional samples. If replicated, these results would point to a novel pathway whereby lithium exerts its therapeutic effects on bipolar disorder.
Showing the most recent 10 out of 56 publications
