We describe a class of human repetitive DNA, called (3 satellite, that, at a most fundamental level, exists as tandem arrays of diverged -68-base-pair monomer repeat units. The monomer units are organized as distinct subsets, each characterized by a multimeric higher-order repeat unit that is tandemly reiterated and represents a recent unit of amplification. We have cloned, characterized, and determined the sequence of two (3 satellite higher-order repeat units: one located on chromosome 9, the other on the acrocentric chromosomes (13, 14, 15, 21, and 22) and perhaps other sites in the genome. Analysis by pulsed-field gel electrophoresis reveals that these tandem arrays are localized in large domains (50-300 kilobase pairs) that are marked by restriction fragment length polymorphisms. In total, .8 satellite sequences comprise several million base pairs of DNA in the human genome. Analysis of this DNA family should permit insights into the nature of chromosome-specific and nonspecific modes of satellite DNA evolution and provide useful tools for probing the molecular organization and concerted evolution of the acrocentric chromosomes.It has been recognized for nearly 20 years that a considerable proportion of the human genome consists of constitutive heterochromatin, regions known to be rich in highly repetitive "satellite" DNA (1, 2). Recently, considerable progress has been made toward the systematic classification of satellite DNAs based on the size and composition, as well as the genomic distribution and subchromosomal localization, of the tandem repeat unit. Studies employing cloned satellite probes have revealed the existence of two broad classes of human satellite DNA. One class of sequences, comprising the major components of several classical satellite DNAs originally isolated by altered buoyant density (base composition), consists of relatively short oligonucleotide tandem repeat units organized in long chromosome-specific arrays. These sequences are located primarily in the heterochromatic long arm regions of human chromosomes 1, 9, 16, and Y (3-7). A second class of sequences, a satellite, consists of =171-base-pair (bp) tandem repeat units localized to the centromeric region of each human chromosome (8). Although both of these major classes of human satellite DNA are organized in a largely chromosome-specific manner, a notable exception involves sequences at the centromere and on the short arms of the five pairs of acrocentric chromosomes (9-12). Repetitive DNA sequences and the ribosomal RNA (rRNA) genes localized to these chromosomes do not exhibit chromosome specificity but rather demonstrate complex patterns of interchromosomal homology, reflecting partial or complete sequence homogenization and concerted evolution (13-18).In this report, we describe the isolation and characterization of a human satellite DNA family, f satellitet which is unrelated in structure or sequence to any previously described human satellite DNAs. These sequences comprise a minimum of several million base pairs o...