Genetics of human methylation

A cloned DNA segment 1.25 kilobases upstream from the joining segments of the human heavy chain immunoglobulin gene revealed extensive polymorphic variation at this locus. Sequencing of this segment cloned from twp different individuals showed that the polymorphism is caused by the insertion and deletion of 50-base-pair tandem repeats. I tested the hypothesis that CMI37, a VNTR locus, is within a hot spot for meiotic recombination by comparing physical distances with genetic distances between markers flanking this locus. I found no evidence of enhanced recombination within the intervals measured, since I detected no recombinants between the markers studied. I also characterized the physical structure of the CMI37 locus. I have developed a strategy to distinguish between the methylation patterns of homologous chromosomes in tissues, and to follow these patterns in human pedigrees. This genetic approach uncovered evidence of variation in the methylation of allelic sites on homologous chromosomes. This variation was tissue-specific and reproducible after transmission through the germ line, demonstrating that homologous chromosomes have distinct blue-prints for the tissue-specific determination of methylation. Furthermore, this approach can be used to study the relationship between parental imprinting and methylation in native mammalian loci. Several human tissues tested were shown to be mosaic for methylation, including clonal tumors of the uterine wall. These tumors allowed us to compare the methylation of phenotypically and genotypically identical cell lineages generated in vivo. Surprisingly, I found that their mosaic methylation-patterns were not random, because independent leiomyomas from the same individual often shared identical methylation patterns. Moreover, independent samples from the same leiomyoma always revealed the same mosaic methylation-pattern. Hence, I have demonstrated that single cells can generate independent lineages with similar methylation faiths," and that postembryonic cells retain the blue-prints for the determination of methylation patterns. I have also established that the methylation of neighboring sites can be independent and probabilistic. These results suggest that a probabilistic machinery coul