The centromeric regions of human chromosomes are dominated by a class of tandemly repeated DNA, alpha satellite, which consists of an extensive group of related, highly diverged repeats, based on a monomer repeat length of ~171 basepairs. Long tandem arrays of alpha satellite, estimated to be hundreds to thousands of kilobasepairs long, are located at the centromeric region of each chromosome. In some cases, centromeric regions are characterized by multiple, independent alpha satellite subsets or by subsets of different classes of satellite DNA (beta or classical), in addition to alpha satellite. Because conventional strategies for mapping human chromosomes cannot easily accommodate long, megabase stretches of repetitive DNA, a directed approach is required for high-resolution physical and genetic mapping of the centromeric region of each human chromosome, for full integration into the complete chromosome maps being generated in other laboratories. The current proposal is an extension of our centromere mapping project, which has been ongoing for the past two years. We propose: (i) to isolate representative DNA probes for the alpha, beta, and classical satellite classes located at or near the centromere of all 24 types of human chromosome and to develop specific sequence tagged sites (STS's) for these centromeric loci; (ii) to develop high-frequency polymorphisms for each centromeric locus and to genotype the 40 CEPH families to contribute to centromere-based genetic linkage maps of each chromosome; (iii) to measure the size and amount of variation of centromeric DNA for each chromosome in a representative collection of individuals to provide a physical estimate to close the "centromere gap" in conventional maps of each chromosome; and (iv) to characterize in detail the complete arrays of four centromeres selected as models (chromosomes 4,7,17, and X), using one- and two-dimensional long-range pulsed-field mapping and short- and long-range cloning in lambda phage and in yeast artificial chromosomes, and strategies designed to identify and clone the junctions between the edges of the centromere array and the chromosome arms. The proposed studies address directly several of the Five-Year Goals of the Human Genome Project: to complete a fully connected human genetic map by providing a polymorphic marker, identified by an STS, at the centromere of each chromosome; to contribute to a fully assembled STS physical map of each human chromosome by providing a well-characterized STS at the centromere of each chromosome and by generating the pulsed-field restriction mapping data needed to fully integrate the centromere into complete chromosome maps; and to generate overlapping sets of closely spaced, ordered markers that span the centromeres of selected chromosomes, with continuity over several million basepairs.