Faculty Bibliography

This report summarizes the deliberations of a panel with representation from diverse disciplines of relevance to the genetics of mood disorders. The major charge to the panel was to develop a strategic plan to employ the tools of genetics to advance the understanding, treatment, and outcomes for mood disorders. A comprehensive review of the evidence for the role of genetic factors in the etiology of mood disorders was conducted, and the chief impediments for progress in gene identification were identified. The National Institute of Mental Health (NIMH) portfolios in the Genetics Research Branch and the Division of Mental Disorders, Behavioral Sciences, AIDS, and all genetics training activities were reviewed. Despite some promising leads, there are still no confirmed linkage findings for mood disorders. Impediments to gene finding include the lack of phenotypic validity, variation in ascertainment sources and methodology across studies, and genetic complexity. With respect to linkage, the committee recommended that a large-scale, integrated effort be undertaken to examine existing data from linkage and association studies of bipolar disorders using identical phenotypes and statistical methods across studies to determine whether the suggestive linkage findings at some loci can be confirmed. Confirmation would justify more intensive approaches to gene finding. The committee recommended that the NIMH support continued efforts to identify the most heritable subtypes and endophenotypes of major depression using the tools of genetic epidemiology, neuroscience, and behavioral science. The field of genetic epidemiology was identified as an important future direction because population-based, epidemiologic studies of families and unrelated affected individuals assume increasing importance for common chronic diseases. To prepare for shifts to more complex genetic models, the committee recommended that the NIMH develop new interdisciplinary training strategies to prepare for the next generation of genetics research.

The human 13q32-q33 region has been linked to both bipolar disorder and schizophrenia. Before completion of the draft sequences, we developed an approximately 15-Mb comprehensive map for the region extending from D13S1300 to ATA35H12. This map was assembled using publicly available mapping data and sequence-tagged site (STS)-based PCR confirmation. We then compared this map with the NCBI, Celera Genomics, and UCSC Golden Path data in February, June, and September 2001. All data sets showed gaps, misassignment of STSs, and errors in orientation and marker order. Surprisingly, the completed sequences of many bacterial artificial chromosomes (BACs) had been truncated. Of 21 gaps that were detected, 4 were present in both the NCBI and Celera databases. All gaps could be filled using 1-2 BAC clones. A total of 39 loci mapped to additional sites within the human genome, providing evidence of segmental duplications. Additionally, 61 unique cDNA clones were sequenced to increase available transcribed sequence, and 11,353 reference single-nucleotide polymorphisms (SNPs) with an average density of 1 SNP/3720 bases were identified. Overall, integration of the data from multiple sources is still needed for complete assembly of the 13q32-q33 region. (c)

Hirschsprung disease (HSCR), the most common hereditary cause of intestinal obstruction, shows considerable variation and complex inheritance. Coding sequence mutations in RET, GDNF, EDNRB, EDN3 and SOX10 lead to long-segment (L-HSCR) and syndromic HSCR but fail to explain the transmission of the much more common short-segment form (S-HSCR). We conducted a genome scan in families with S-HSCR and identified susceptibility loci at 3p21, 10q11 and 19q12 that seem to be necessary and sufficient to explain recurrence risk and population incidence. The gene at 10q11 is probably RET, supporting its crucial role in all forms of HSCR; however, coding sequence mutations are present in only 40% of linked families, suggesting the importance of noncoding variation. Here we show oligogenic inheritance of S-HSCR, the 3p21 and 19q12 loci as RET-dependent modifiers, and a parent-of-origin effect at RET. This study demonstrates by a complete genetic dissection why the inheritance pattern of S-HSCR is nonmendelian.

Five members of a family are described, all of whom suffered from chronic constipation and megacolon. Detailed clinical and histologic evaluation of each member revealed that two individuals have histologic evidence of desmosis coli and three have Hirschsprung's disease, one of whom also has desmosis coli. The latter combination has never been described before, either in a family or in a single patient. Genetic studies of the family did not reveal an increase in the number of shared markers for the RET proto-oncogene, suggesting that this previously undescribed familial association is likely not caused by a mutation in the RET gene, but by other genetic abnormalities.

Thymic stromal derived lymphopoietin receptor (TSLPR) is a novel receptor subunit that is related in sequence to the interleukin (IL)-2 receptor common gamma chain. TSLPR forms a heterodimeric complex with the IL-7 receptor alpha chain to form the receptor for thymic stromal derived lymphopoietin, a cytokine involved in B- and T-cell function. We have cloned the TSLP receptor from rat and find that the WSXWX motif commonly found in extracellular domains of cytokine receptors is conserved as a W(T/S)XV(T/A) motif among TSLP receptors from mouse, rat and human. As in the mouse, TSLP receptor is widely expressed in rats suggesting that TSLPR may have roles in signaling outside the hematopoietic system. A zooblot analysis revealed that TSLPR is expressed in all vertebrate species examined. The absence of TSLPR in Saccharomyces cerevisiae, Drosophila melanogaster and Caenorhabditis elegans genomes is similar to the expression of several other cytokine receptors that have been characterized thus far. We have also characterized the genomic structure of the murine Tslpr gene which shows that in addition to primary sequence homology, it shares a common genomic organization of coding exons with the murine IL-2 receptor common gamma chain (Il2rg). Use of an alternative splice acceptor site leads to two alternatively spliced transcript variants of murine TSLPR, both of which are functional receptors. Finally, using linkage analysis, we mapped the murine Tslpr gene to mouse chromosome 5 between the Ecm2 and Pxn genes.

Idiopathic congenital central hypoventilation syndrome (CCHS) is an unique disorder of respiratory control, occurring in association with Hirschsprung disease (HSCR), tumors of neural crest origin, and symptoms of autonomic nervous system dysfunction (ANSD). CCHS is thought to be genetic in origin based upon 1) affected sib pairs, 2) genetic analysis, and 3) identification of genetic mutations in both HSCR and CCHS patients. Because these mutations have been found in but a few cases of CCHS, exploration of other candidate genes has continued. Brain-derived neurotrophic factor (BDNF) represents a potential candidate gene to consider because of altered respiratory control in the BDNF knock-out mouse model and localization to the enteric nervous system in human tissue. The objective of this study was to determine the frequency of BDNF mutations among 19 children with CCHS (five with HSCR) compared to 40 unaffected unrelated controls. Using the known genomic DNA sequence for BDNF, polymerase chain reaction (PCR)-amplified genomic DNA was analyzed by standard sequencing methods. A discrete mutation was identified in one of 19 children with isolated CCHS and the unaffected father. Specifically, an isoleucine was substituted for a threonine or serine in the amino acid sequence. Absence of this mutation in 40 controls confirmed that this mutation was likely not a common polymorphism. These data further support a genetic basis for CCHS, though mutations of BDNF are not consistent in this disorder.

To assess evidence for a gene with large effect on systolic blood pressure (SBP), diastolic blood pressure (DBP), and body mass index (BMI), we conducted segregation analyses on 261 nuclear families collected from a rural Caucasian community in Michigan. The families were ascertained through a hypertensive proband. Each phenotype was adjusted for significant covariate effects (e.g., gender and age). We used class D regressive models to conduct the segregation analyses. Our analysis results support the segregation of a major gene for BMI, but not for SBP or DBP. A recessive locus effect provided the best explanation for BMI where approximately 43% of the variance of BMI was due to this gene.

The human genome project is producing an enormous amount of sequence data, based on which single base changes between individuals can be identified. Unfortunately, computer tools that were adequate for sequence assembly are less than ideal for the characterization of polymorphism data [single nucleotide (snp) or insertion/deletion (indel)] and other sequence features, and their relationship to each other. We have developed viewGene as a flexible tool that takes input from a number of sequence formats and analysis programs (Genbank, FASTA, RepeatMasker, Cross match, BLAST, user-defined data) to construct a sequence reference scaffold that can be viewed through a simple graphical interface. polymorphisms generated from many sources can be added to this scaffold through the same sequence formats, with a variety of options to control what is displayed. Large amounts of polymorphism data can be organized so that patterns and haplotypes can be readily discerned. In our laboratory, viewGene has been used to view annotated genbank records, find nonrepetitive sequence fragments for polymorphism detection, and visualize similarity search results. Manipulation, cross-referencing, and haplotype viewing of snp data are essential for quality assessment and identification of variants associated with genetic disease, and viewGene provides all three of these important functions.

A genome-wide scan using 387 short tandem repeat markers was conducted for obesity among 618 black individuals from 202 families residing in a suburb of Chicago. Evidence for linkage was evaluated with BMI and percent body fat (PBF) using a variance component analysis approach. Suggestive evidence for linkage was found for BMI on chromosome 5 (logarithm of odds [LOD] score = 1.9) and PBF on chromosome 6 (LOD score = 2.7). One additional region on chromosome 3 was linked to these phenotypes at a lower level of significance (LOD score = 1.8 and 0.95 for BMI and PBF, respectively); the linked marker on this chromosome lies in the same region implicated as harboring obesity genes in a previous study of a white population. The replication of linkage evidence using different ethnic groups reinforces the potential significance of this latter candidate region.