Faculty Bibliography

The evolutionary histories and relationships among African, Eurasian, and Pacific Island populations are investigated by using observations on five polymorphic restriction sites in the beta-globin gene cluster. We present new data on 222 chromosomes from a global sample and combine these with previously published observations on 591 chromosomes. It is shown that the data are rich in rare haplotypes and that rare variants are not helpful for standard methods of population structure analysis. Consequently, a new approach is developed. We first consider the phylogeny of beta-globin haplotypes. The roles of mutation, gene conversion, and recombination in the generation of haplotype diversity are specifically focused upon. The relationships among human populations are then inferred from the phylogenetic relationships among the haplotypes, their presence or absence, and frequencies within populations. Questions regarding whether or not a phyletic process can account for relationships among the major geographical populations and whether or not an extant human population exhibits the qualities that would be expected of an ancestral group are addressed. The results of this analysis support an African origin for modern Homo sapiens and a phyletic structuring of the major geographical regions. However, it is shown that divergence times for the various populations cannot be determined from these data.

Approximately 70 percent of the mutations in cystic fibrosis patients correspond to a specific deletion of three base pairs, which results in the loss of a phenylalanine residue at amino acid position 508 of the putative product of the cystic fibrosis gene. Extended haplotype data based on DNA markers closely linked to the putative disease gene locus suggest that the remainder of the cystic fibrosis mutant gene pool consists of multiple, different mutations. A small set of these latter mutant alleles (about 8 percent) may confer residual pancreatic exocrine function in a subgroup of patients who are pancreatic sufficient. The ability to detect mutations in the cystic fibrosis gene at the DNA level has important implications for genetic diagnosis.

The fluorescence anisotropy of 1,6-diphenyl-1,3,5-hexatriene in labeled platelet membranes, an index of membrane fluidity, is a stable, familial trait that is associated with a clinically distinct subtype of Alzheimer disease. Complex segregation analysis of this continuous variable was performed on 95 members of 14 pedigrees identified through probands who had autopsy-confirmed or clinically diagnosed Alzheimer disease. The results suggest that platelet membrane fluidity is controlled by a single genetic locus, PMF, with two alleles that have additive effects. The PMF locus appears to explain approximately 80% of the total variation in platelet membrane fluidity within the families of patients with Alzheimer disease.

For studying the biology of autosomal trisomies it is necessary to establish the parental origin and meiotic stage of nondisjunction by using genetic markers. Theoretical formulas are obtained for calculating the probability of establishing (1) parental origin and meiotic stage of nondisjunction by using a centromeric marker, (2) parental origin of nondisjunction by using a noncentromeric marker, and (3) meiotic stage, given parental origin of nondisjunction. These theoretical calculations demonstrate that parental origin of nondisjunction can be identified with virtual certainty by utilizing multiple genetic markers along a chromosome arm. Centromeric markers are by themselves inefficient for determining meiotic stage of the error, but the efficiency can be considerably increased if parental origin is known with certainty. Even then, multiple centromeric markers may be necessary.

We have constructed a genetic linkage map of 17 markers on the long arm of human chromosome 21, including six genes and two anonymous loci with a variable number of tandem repeats. The estimated length of the map is 103 cM in males and 140 cM in females, assuming Kosambi interference. Recombination in females was approximately twice that in males between proximal markers. However, over half of the recombination events in either sex occur distally, in 21q22.3, although this region accounts for only about 15% of the physical length of chromosome 21.

Mutations in the low-density lipoprotein (LDL) receptor gene result in the autosomal dominant disorder familial hypercholesterolemia (FH). Many different LDL receptor mutations have been identified and characterized, demonstrating a high degree of allelic heterogeneity at this locus. The ability to identify mutant LDL receptor genes for prenatal diagnosis of homozygous FH or to study the role of the LDL receptor gene in polygenic hypercholesterolemia requires the use of closely linked RFLPs. In the present study we used 10 different RFLPs, including three newly described polymorphisms, to construct 123 independent haplotypes from 20 Caucasian American pedigrees. Our sample contained 31 different haplotypes varying in frequency from 0.8% to 29.3%; the five most common haplotypes account for 67.5% of the sample. The heterozygosity and PIC of each site were determined, and these values disclosed that eight of the RFLPs were substantially polymorphic. Linkage-disequilibrium analysis of the haplotype data revealed strong nonrandom associations among all 10 RFLPs, especially among those sites clustered in the 3' region of the gene. Evolutionary analysis suggests the occurrence of both mutational and recombinational events in the generation of the observed haplotypes. A strategy for haplotype analysis of the LDL receptor gene in individuals of Caucasian American descent is presented.

The gene for von Recklinghausen neurofibromatosis type 1 (NF1) has recently been mapped to the pericentromeric region of human chromosome 17. To further localize the NF1 gene, linkage analysis using chromosome 17 DNA markers was performed on 11 multigeneration families with 175 individuals, 57 of whom were affected. The markers used were D17Z1 (p17H8), D17S58 (EW301), D17S54 (EW203), D17S57 (EW206), D17S73 (EW207), CRI-L946, HOX-2, and growth hormone. Tight linkage was found between NF1 and D17Z1, D17S58, and D17S57 with a recombination fraction of zero. One recombinant was detected between NF1 and D17S73, showing linkage with a 10% recombination fraction. No linkage was detected between NF1 and CRI-L946 or between HOX-2 and growth hormone. Our data are consistent with the proposed gene order pter D17S58-D17Z1-NF1-D17S57-D17S73 qter.