Faculty Bibliography

The first demonstration of an autosomal dominant human disease caused by segmental trisomy came in 1991 for Charcot-Marie-Tooth disease type 1A (CMT1A). For this disorder, the segmental trisomy is due to a large tandem duplication of 1.5 Mb of DNA located on chromosome 17p11.2-p12. The search for the CMT1A disease gene was misdirected and impeded because some chromosome 17 genetic markers that are linked to CMT1A lie within this duplication. To better understand how such a duplication might affect genetic analyses in the context of disease gene mapping, we studied the effects of marker duplication on transmission probabilities of marker alleles, on linkage analysis of an autosomal dominant disease, and on tests of linkage homogeneity. We demonstrate that the undetected presence of a duplication distorts transmission ratios, hampers fine localization of the disease gene, and increases false evidence of linkage heterogeneity. In addition, we devised a likelihood-based method for detecting the presence of a tandemly duplicated marker when one is suspected. We tested our methods through computer simulations and on CMT1A pedigrees genotyped at several chromosome 17 markers. On the simulated data, our method detected 96% of duplicated markers (with a false-positive rate of 5%). On the CMT1A data our method successfully identified two of three loci that are duplicated (with no false positives). This method could be used to identify duplicated markers in other regions of the genome and could be used to delineate the extent of duplications similar to that involved in CMT1A.

With the advent of the Human Genome Project, the ability to rapidly construct comprehensive and accurate linkage maps based on genetic marker data from family studies is an absolute necessity. In addition to their usefulness in localizing genes for both simple and complex disorders, linkage maps are invaluable tools for genetic counseling using linked marker genes. Several computer program packages are publicly available which aid in the construction of linkage maps by computing multipoint likelihoods for specified locus orders. However, these programs work in a step-by-step fashion, requiring intensive user-intervention and analysis at each step. Such a repetitive process is amenable to computerized automation. We have developed and tested an expert system computer program, MultiMap, for automated genetic linkage mapping. This program greatly reduces the amount of user-computer interaction, increasing the accuracy and speed with which a map can be constructed. In addition, because the total mapping time is greatly reduced through automation, it is now feasible to explore and compare various mapping heuristics and mapping criteria in order to develop the most appropriate approach, or set of approaches, for genetic linkage mapping. MultiMap need not be restricted to the construction of genetic maps, but could be adapted to aid in the automated construction of physical maps as well.

Hirschsprung disease (HSCR) is characterized by a congenital absence of enteric ganglia along a variable length of the intestine. Although long considered to be a multifactorial disease, we have identified linkage in a subset of five HSCR families to the pericentromeric region of chromosome 10, thereby providing monogenic inheritance in some families. A maximum two-point lod score of 3.37 (theta = 0.045) was observed between HSCR and D10S176, under an incompletely penetrant dominant model. Multipoint, affecteds-only and non-parametric analyses supported this finding and localize this gene to a region of approximately 7 centiMorgans, in close proximity to the locus for multiple endocrine neoplasia type 2 (MEN2). The co-occurrence of these two entities in some families might be attributable to shared pathogenetic origins.

Using a case-control design, a reported association of schizophrenia with homozygosity at the dopamine D3 receptor gene locus was investigated in a group of patients (n = 53), with schizophrenia (DSM-III-R), and psychiatrically normal controls (n = 61), matched for ethnicity and area of residence. No significant differences in the distribution of alleles or genotypes between the two groups could be detected. However, among patients with a family history of schizophrenia, as compared to controls without such family history, an association with allele 1 at this locus was noted (Odds ratio 12.4, C.I. 1.61, 96.35).

A physical map of 4q35 was constructed through radiation hybrid analysis of 134 clones generated from the cell line HHW416, a chromosome 4-only human-hamster somatic cell hybrid. This subtelomeric region contains the as-yet-unidentified gene responsible for facioscapulohumeral muscular dystrophy. The most likely order of 15 loci within 4q35 was determined. The loci ordered on this radiation hybrid map include both genes and polymorphic loci, as well as monomorphic loci which cannot be placed on a genetic linkage map. The physical distance spanning these loci was estimated to be approximately 4.5 Mb, by using a kilobase/centiray conversion factor derived from 4p16.3 marker analysis through the same set of radiation hybrids. The comparison of this physical map to establish genetic maps suggests that this region is smaller than initially estimated and that recombination rates are increased near the telomere.

This is the first report of an individual with several congenital abnormalities, including those suggestive of Treacher Collins syndrome, and an atypical schizophrenic illness. Cytogenetic studies have failed to detect any recognizable chromosomal abnormality.

We used single-strand conformation polymorphism (SSCP) to detect DNA polymorphisms in the 3' untranslated (3'UT) region of the gene for cystathionine beta-synthase (CBS). A polymorphism due to a T-to-C substitution at nucleotide 549 of the 3'UT region with heterozygosity of 46% has been identified. Genotypes for this polymorphism have been obtained in all of the informative CEPH families, and CBS has been placed in the linkage map of human chromosome 21.

DNA polymorphisms can be used to place loci and phenotypes on the linkage maps of human chromosomes. In an effort to localize genes on the linkage map of human chromosome 21 better, we examined their 3' untranslated (3'UT) regions for the presence of polymorphisms. We amplified the 3'UT region of 17 genes of chromosome 21 by the polymerase chain reaction and subjected the product to single-stranded conformation analysis (SSCA). We have found eight polymorphisms in the 3'UT region of genes. The total area examined was 8144 nucleotides and therefore the variability detected by this method was 1 in 1018 nucleotides. This is not different from the estimated variability of DNA sequences based on restriction analysis. Sequence analysis revealed that all polymorphisms found are due to single nucleotide substitutions. Additional polymorphisms were identified in the last intron of BCEI gene and in the 3'-flanking region of the S100B gene. We conclude that the 3'UT region of genes is a relatively rich source of polymorphisms and that SSCA is an effective method of detecting the normal sequence variation in the human genome.