Structure, organization, and sequence of alpha satellite DNA from human chromosome 17: evidence for evolution by unequal crossing-over and an ancestral pentamer repeat shared with the human X chromosome.

The centromeric regions of all human chromosomes are characterized by distinct subsets of a diverse tandemly repeated DNA family, alpha satellite. On human chromosome 17, the predominant form of alpha satellite is a 2.7-kilobase-pair higher-order repeat unit consisting of 16 alphoid monomers. We present the complete nucleotide sequence of the 16-monomer repeat, which is present in 500 to 1,000 copies per chromosome 17, as well as that of a less abundant 15-monomer repeat, also from chromosome 17. These repeat units were approximately 98% identical in sequence, differing by the exclusion of precisely 1 monomer from the 15-monomer repeat. Homologous unequal crossing-over is suggested as a probable mechanism by which the different repeat lengths on chromosome 17 were generated, and the putative site of such a recombination event is identified. The monomer organization of the chromosome 17 higher-order repeat unit is based, in part, on tandemly repeated pentamers. A similar pentameric suborganization has been previously demonstrated for alpha satellite of the human X chromosome. Despite the organizational similarities, substantial sequence divergence distinguishes these subsets. Hybridization experiments indicate that the chromosome 17 and X subsets are more similar to each other than to the subsets found on several other human chromosomes. We suggest that the chromosome 17 and X alpha satellite subsets may be related components of a larger alphoid subfamily which have evolved from a common ancestral repeat into the contemporary chromosome-specific subsets.     

Interhomologue sequence variation of alpha satellite DNA from human chromosome 17: Evidence for concerted evolution along haplotypic lineages

Alpha satellite DNA is a family of tandemly repeated DNA found at the centromeres of all primate chromosomes. Different human chromosomes 17 in the population are characterized by distinct alpha satellite haplotypes, distinguished by the presence of variant repeat forms that have precise monomeric deletions. Pairwise comparisons of sequence diversity between variant repeat units from each haplotype show that they are closely related in sequence. Direct sequencing of PCR-amplified alpha satellite reveals heterogeneous positions between the repeat units on a chromosome as two bands at the same position on a sequencing ladder. No variation was detected in the sequence and location of these heterogeneous positions between chromosomes 17 from the same haplotype, but distinct patterns of variation were detected between chromosomes from different haplotypes. Subsequent sequence analysis of individual repeats from each haplotype confirmed the presence of extensive haplotype-specific sequence variation. Phylogenetic inference yielded a tree that suggests these chromosome 17 repeat units evolve principally along haplotypic lineages. These studies allow insight into the relative rates and/or timing of genetic turnover processes that lead to the homogenization of tandem DNA families. Correspondence to: H.F. Willard     

Structure of DNA near long tandem arrays of alpha satellite DNA at the centromere of human chromosome 7.

The centromeric regions of human chromosomes contain long tracts of tandemly repeated DNA, of which the most extensively characterized is alpha satellite. In a screen for additional centromeric DNA sequences, four phage clones were obtained which contain alpha satellite as well as other sequences not usually found associated with tandemly repeated alpha satellite DNA, including L1 repetitive elements, an Alu element, and a novel AT-rich repeated sequence. The alpha satellite DNA contained within these clones does not demonstrate the higher-order repeat structure typical of tandemly repeated alpha satellite. Two of the clones contain inversions; instead of the usual head-to-tail arrangement of alpha satellite monomers, the direction of the monomers changes partway through each clone. The presence of both inversions was confirmed in human genomic DNA by polymerase chain reaction amplification of the inverted regions. One phage clone contains a junction between alpha satellite DNA and a novel low-copy repeated sequence. The junction between the two types of DNA is abrupt and the junction sequence is characterized by the presence of runs of A's and T's, yielding an overall base composition of 65% AT with local areas > 80% AT. The AT-rich sequence is found in multiple copies on chromosome 7 and homologous sequences are found in (peri)centromeric locations on other human chromosomes, including chromosomes 1, 2, and 16. As such, the AT-rich sequence adjacent to alpha satellite DNA provides a tool for the further study of the DNA from this region of the chromosome. The phage clones examined are located within the same 3.3-Mb SstII restriction fragment on chromosome 7 as the two previously described alpha satellite arrays, D7Z1 and D7Z2. These new clones demonstrate that centromeric repetitive DNA, at least on chromosome 7, may be more heterogeneous in composition and organization than had previously been thought.     

Nonrandom localization of recombination events in human alpha satellite repeat unit variants: implications for higher-order structural characteristics within centromeric heterochromatin.

Tandemly repeated DNA families appear to undergo concerted evolution, such that repeat units within a species have a higher degree of sequence similarity than repeat units from even closely related species. While intraspecies homogenization of repeat units can be explained satisfactorily by repeated rounds of genetic exchange processes such as unequal crossing over and/or gene conversion, the parameters controlling these processes remain largely unknown. Alpha satellite DNA is a noncoding tandemly repeated DNA family found at the centromeres of all human and primate chromosomes. We have used sequence analysis to investigate the molecular basis of 13 variant alpha satellite repeat units, allowing comparison of multiple independent recombination events in closely related DNA sequences. The distribution of these events within the 171-bp monomer is nonrandom and clusters in a distinct 20- to 25-bp region, suggesting possible effects of primary sequence and/or chromatin structure. The position of these recombination events may be associated with the location within the higher-order repeat unit of the binding site for the centromere-specific protein CENP-B. These studies have implications for the molecular nature of genetic recombination, mechanisms of concerted evolution, and higher-order structure of centromeric heterochromatin.     

Patterns of intra- and interarray sequence variation in alpha satellite from the human X chromosome: evidence for short-range homogenization of tandemly repeated DNA sequences

A number of processes, such as sequence conversion, unequal crossingover, and molecular drive, have been postulated to explain the homogenization of tandemly repeated DNA families. To investigate the nature and extent of such processes in the alpha satellite family of centromeric DNA, we determined the nucleotide sequence of approximately 700 bp from each of 40 representative alpha satellite repeats from six sources of human X chromosomes, obtaining a total of approximately 28 kb of sequence data. Sequence divergence among the repeats examined was low, with an average pairwise difference of approximately 1%. Pairwise comparisons of all repeats indicate that the degree of similarity for those repeats in physical proximity (within approximately 15 kb) of each other is significantly greater than that for randomly located repeats, from either the same or different X chromosomes, suggesting that the mechanisms predicted to homogenize these arrays are effectively short-range in action. Analysis of individual patterns of sequence variation allows the assignment of haplotypes for five high-copy-number diagnostic positions and reveals distinct positions of equilibrium and disequilibrium within the repeat. These analyses address hypotheses about the origin of the observed patterns of variation throughout alpha satellite evolution.     

Concerted evolution of primate alpha satellite DNA. Evidence for an ancestral sequence shared by gorilla and human X chromosome alpha satellite

To understand evolutionary events in the formation of higher-order repeat units in alpha satellite DNA, we have examined gorilla sequences homologous to human X chromosome alpha satellite. In humans, alpha satellite on the X chromosome is organized as a tandemly repeated, 2.0 x 10(3) base-pairs (bp) higher-order repeat unit, operationally defined by the restriction enzyme BamHI. Each higher-order repeat unit is composed of 12 tandem approximately 171 base-pair monomer units that have been classified into five distinct sequence homology groups. BamHI-digested gorilla genomic DNA hybridized with the cloned human 2 x 10(3) bp X alpha satellite repeat reveals three bands of sizes approximately 3.2 x 10(3), 2.7 x 10(3) and 2 x 10(3) bp. Multiple copies of all three repeat lengths have been isolated and mapped to the centromeric region of the gorilla X chromosome by fluorescence in situ hybridization. Long-range restriction mapping using pulsed-field gel electrophoresis shows that the 2.7 x 10(3) and 3.2 x 10(3) bp repeat arrays exist as separate but likely neighboring arrays on the gorilla X, each ranging in size from approximately 200 x 10(3) to 500 x 10(3) bp, considerably smaller than the approximately 2000 x 10(3) to 4000 x 10(3) bp array found on human X chromosomes. Nucleotide sequence analysis has revealed that monomers within all three gorilla repeat units can be classified into the same five sequence homology groups as monomers located within the higher-order repeat unit on the human X chromosome, suggesting that the formation of the five distinct monomer types predates the divergence of the lineages of contemporary humans and gorillas. The order of 12 monomers within the 2 x 10(3) and 2.7 x 10(3) bp repeat units from the gorilla X chromosome is identical with that of the 2 x 10(3) bp repeat unit from the human X chromosome, suggesting an ancestral linear arrangement and supporting hypotheses about events largely restricted to single chromosome types in the formation of alpha satellite higher-order repeat units.     

Chromosome-specific alpha satellite DNA: nucleotide sequence analysis of the 2.0 kilobasepair repeat from the human X chromosome.

The pericentromeric region of the human X chromosome is characterized by a tandemly repeated family of 2.0 kilobasepair (kb) DNA fragments, initially revealed by cleavage of human DNA with the restriction enzyme BamHI. We report here the complete nucleotide sequence of a cloned member of the repeat family and establish that this X-linked DNA family consists entirely of alpha satellite DNA. Our data indicate that the 2.0 kb repeat consists of twelve alpha satellite monomers arranged in imperfect, direct repeats. Each of the alpha X monomers is approximately 171 basepairs (bp) in length and is 60-75% identical in sequence to previously described primate alpha satellite DNAs. The twelve alpha X monomers are 65-85% identical in sequence to each other and are organized as two adjacent, related blocks of five monomers, plus an additional two monomers also related to monomers within the pentamer blocks. Partial nucleotide sequence of a second, independent copy of the 2.0 kb BamHI fragment established that the 2.0 kb repeat is, in fact, the unit of amplification on the X. Comparison of the sequences of the twelve alpha X monomers allowed derivation of a 171 bp consensus sequence for alpha satellite DNA on the human X chromosome. These sequence data, combined with the results of filter hybridization experiments of total human DNA and X chromosome DNA, using subregions within the 2.0 kb repeat as probes, provide strong support for the hypothesis that individual human chromosomes are characterized by different alpha satellite families, defined both by restriction enzyme periodicity and by chromosome-specific primary sequence.     

Chromosome-specific organization of human alpha satellite DNA

Restriction endonuclease analysis of human genomic DNA has previously revealed several prominent repeated DNA families defined by regularly spaced enzyme recognition sites. One of these families, termed alpha satellite DNA, was originally identified as tandemly repeated 340- or 680-base pair (bp) EcoRI fragments that hybridize to the centromeric regions of human chromosomes. We have investigated the molecular organization of alpha satellite DNA on individual human chromosomes by filter hybridization and in situ hybridization analysis of human DNA and DNA from rodent/human somatic cell hybrids, each containing only a single human chromosome. We used as probes a cloned 340-bp EcoRI alpha satellite fragment and a cloned alpha satellite-containing 2.0-kilobase pair (kbp) BamHI fragment from the pericentromeric region of the human X chromosome. In each somatic cell hybrid DNA, the two probes hybridized to a distinct subset of DNA fragments detected in total human genomic DNA. Thus, alpha satellite DNA on each of the human chromosomes examined--the X and Y chromosomes and autosomes 3, 4, and 21--is organized in a specific and limited number of molecular domains. The data indicate that subsets of alpha satellite DNA on individual chromosomes differ from one another, both with respect to restriction enzyme periodicities and with respect to their degree of sequence relatedness. The results suggest that some, and perhaps many, human chromosomes are characterized by a specific organization of alpha satellite DNA at their centromeres and that, under appropriate experimental conditions, cloned representatives of alpha satellite subfamilies may serve as a new class of chromosome-specific DNA markers.     

Molecular analysis of a polymorphic domain of alpha satellite from the human X chromosome.

Alpha satellite DNA, a diverse family of tandemly repeated DNA sequences located at the centromeric region of each human chromosome, is organized in a highly chromosome-specific manner and is characterized by a high frequency of restriction-fragment-length polymorphism. To examine events underlying the formation and spread of these polymorphisms within a tandem array, we have cloned and sequenced a representative copy of a polymorphic array from the X chromosome and compared this polymorphic copy with the predominant higher-order repeat form of X-linked alpha satellite. Sequence data indicate that the polymorphism arose by a single base mutation that created a new restriction site (for HindIII) in the sequence of the predominant repeat unit. This variant repeat unit, marked by the new HindIII site, was subsequently amplified in copy number to create a polymorphic domain consisting of approximately 500 copies of the variant repeat unit within the X-linked array of alpha satellite. We propose that a series of intrachromosomal recombination events between misaligned tandem arrays, involving multiple rounds of either unequal crossing-over or sequence conversion, facilitated the spread and fixation of this variant HindIII repeat unit.     

Genomic organization of alpha satellite DNA on human chromosome 7: evidence for two distinct alphoid domains on a single chromosome.

A complete understanding of chromosomal disjunction during mitosis and meiosis in complex genomes such as the human genome awaits detailed characterization of both the molecular structure and genetic behavior of the centromeric regions of chromosomes. Such analyses in turn require knowledge of the organization and nature of DNA sequences associated with centromeres. The most prominent class of centromeric DNA sequences in the human genome is the alpha satellite family of tandemly repeated DNA, which is organized as distinct chromosomal subsets. Each subset is characterized by a particular multimeric higher-order repeat unit consisting of tandemly reiterated, diverged alpha satellite monomers of approximately 171 base pairs. The higher-order repeat units are themselves tandemly reiterated and represent the most recently amplified or fixed alphoid sequences. We present evidence that there are at least two independent domains of alpha satellite DNA on chromosome 7, each characterized by their own distinct higher-order repeat structure. We determined the complete nucleotide sequences of a 6-monomer higher-order repeat unit, which is present in approximately 500 copies per chromosome 7, as well as those of a less-abundant (approximately 10 copies) 16-monomer higher-order repeat unit. Sequence analysis indicated that these repeats are evolutionarily distinct. Genomic hybridization experiments established that each is maintained in relatively homogeneous tandem arrays with no detectable interspersion. We propose mechanisms by which multiple unrelated higher-order repeat domains may be formed and maintained within a single chromosomal subset.     

Polymorphisms and genomic organization of repetitive DNA from centromeric regions of Arabidopsis chromosomes.

A highly abundant repetitive DNA sequence family of Arabidopsis, AtCon, is composed of 178-bp tandemly repeated units and is located at the centromeres of all five chromosome pairs. Analysis of multiple copies of AtCon showed 95% conservation of nucleotides, with some alternative bases, and revealed two boxes, 30 and 24 bp long, that are 99% conserved. Sequences at the 3' end of these boxes showed similarity to yeast CDEI and human CENP-B DNA-protein binding motifs. When oligonucleotides from less conserved regions of AtCon were hybridized in situ and visualized by using primer extension, they were detected on specific chromosomes. When used for polymerase chain reaction with genomic DNA, single primers or primer pairs oriented in the same direction showed negligible amplification, indicating a head-to-tail repeat unit organization. Most primer pairs facing in opposite directions gave several strong bands corresponding to their positions within AtCon. However, consistent with the primer extension results, some primer pairs showed no amplification, indicating that there are chromosome-specific variants of AtCon. The results are significant because they elucidate the organization, mode of amplification, dispersion, and evolution of one of the major repeated sequence families of Arabidopsis. The evidence presented here suggests that AtCon, like human alpha satellites, plays a role in Arabidopsis centromere organization and function.     

Evolutionary dynamics of a satellite DNA in the tiger beetle species pair Cicindela campestris and C. maroccana.

Satellite repeat elements are an abundant component of eukaryotic genomes, but not enough is known about their evolutionary dynamics and their involvement in karyotype and species differentiation. We report the nucleotide sequence, chromosomal localization, and evolutionary dynamics of a repetitive DNA element of the tiger beetle species pair Cicindela maroccana and Cicindela campestris. The element was detected after restriction digest of C. maroccana total genomic DNA with EcoRI as a single band and its multimers on agarose gels. Cloning and sequencing of several isolates revealed a consensus sequence of 383 bp with no internal repeat structure and no detectable similarity to any entry in GenBank. Hybridization of the satellite unit to C. maroccana mitotic and meiotic chromosomes revealed the presence of this repetitive DNA in the centromeres of all chromosomes except the Y chromosome, which exhibited only a very weak signal in its short arm. PCR-based tests for this satellite in related species revealed its presence in the sister species C. campestris, but not in other closely related species. Phylogenetic analysis of PCR products revealed well-supported clades that generally separate copies from each species. Because both species exhibit the multiple X chromosome karyotypic system common to Cicindela, but differ in their X chromosome numbers (four in C. maroccana vs. three in C. campestris), structural differences could also be investigated with regard to the position of satellites in a newly arisen X chromosome. We find the satellite in a centromeric position in all X chromosomes of C. maroccana, suggesting that the origin of the additional X chromosome involves multiple karyotypic rearrangements.     

Organization and evolution of alpha satellite DNA from human chromosome 11

The human alpha satellite repetitive DNA family is organized as distinct chromosomal subsets located at the centromeric regions of each human chromosome. Here, we describe a subset of the alpha satellite which is localized to human chromosome 11. The principal unit of repetition of this alpha satellite subset is an 850 bp XbaI fragment composed of five tandem diverged alphoid monomers, each 171 bp in length. The pentamer repeat units are themselves tandemly reiterated, present in 500 copies per chromosome 11. In filter hybridization experiments, the Alpha 11 probes are specific for the centromeric alpha satellite sequences of human chromosome 11. The complete nucleotide sequences of two independent copies of the XbaI pentamer reveal a pentameric configuration shared with the alphoid repeats of chromosomes 17 and X, consistent with the existence of an ancestral pentameric repeat common to the centromeric arrays of at least these three human chromosomes.     

Chromosome-specific alpha satellite DNA from human chromosome 1: hierarchical structure and genomic organization of a polymorphic domain spanning several hundred kilobase pairs of centromeric DNA

The human alpha satellite repetitive DNA family is organized as distinct chromosome-specific subsets localized to the centromeric region of each chromosome. Here, we report he isolation and characterization of cloned repeat units which define a hierarchical subset of alpha satellite on human chromosome 1. This subset is characterized by a 1.9-kb higher-order repeat unit which consists of 11 tandem approximately 171-bp alpha satellite monomer repeat units. The higher-order repeat unit is itself tandemly repeated, present in at least 100 copies at the centromeric region of chromosome 1. Using pulsed-field gel electrophoresis we estimate the total array length of these tandem sequences at the centromere of chromosome 1 to be several hundred kilobase pairs. Under conditions of high stringency, the higher-order repeat probe hybridizes specifically to chromosome 1 and can be used to detect several associated restriction fragment length DNA polymorphisms. As such, this probe may be useful for molecular and genetic analyses of the centromeric region of human chromosome 1.     

Distribution of gamma satellite DNA on the human X and Y chromosomes suggests that it is not required for mitotic centromere function

The bulk of the DNA found at human centromeres is composed of tandemly arranged repeats, the most abundant of which is alpha satellite. Other human centromeric repetitive families have been identified, one of the more recent being gamma satellite. To date, gamma satellite DNAs have been reported at the centromeres of human chromosomes 8 and X. Here, we show that gamma-X satellite DNA is not interspersed with the major DZX1 alpha-X block, but rather is organised as a single array of approximately 40-50 kb on the short-arm side of the alpha satellite domain. This repeat array is absent on two mitotically stable Xq isochromosomes. Furthermore, a related repeat DNA has been identified on the human Y chromosome. Fluorescence in situ hybridisation has localised this satellite DNA to the long arm side of the major DYZ3 alpha-Y domain, outside the region previously defined as that required for mitotic centromere function. Together, these data suggest that while blocks of highly related gamma satellite DNAs are present in the pericentromeric regions of both human sex chromosomes, this repeated DNA is not required for mitotic centromere function.     

Chromosome-specific subsets of human alpha satellite DNA: Analysis of sequence divergence within and between chromosomal subsets and evidence for an ancestral pentameric repeat

Summary The centromeric regions of human chromosomes are characterized by diverged chromosome-specific subsets of a tandemly repeated DNA family, alpha satellite, which is based on a fundamental monomer repeat unit 171 bp in length. We have compared the nucleotide sequences of 44 alphoid monomers derived from cloned representatives of the multimeric higher-order repeat units of human chromosomes 1, 11, 17, and X. The 44 monomers exhibit an average 16% divergence from a consensus alphoid sequence, and can be assigned to five distinct homology groups based on patterns of sequence substitutions and gaps relative to the consensus. Approximately half of the overall sequence divergence can be accounted for by sequence changes specific to a particular homology group; the remaining divergence appears to be independent of the five groups and is randomly distributed, both within and between chromosomal subsets. The data are consistent with the proposal that the contemporary tandem arrays on chromosomes 1, 11, 17, and X derive from a common multimeric repeat, consisting of one monomer each from the five homology groups. The sequence comparisons suggest that this pentameric repeat must have spread to these four chromosomal locations many millions of years ago, since which time evolution of the four, now chromosome-specific, alpha satellite subsets has been essentially independent.     

PCR amplification of chromosome-specific alpha satellite DNA: definition of centromeric STS markers and polymorphic analysis.

Alpha satellite DNA is a tandemly repetitive DNA family found at the centromere of every human chromosome. Chromosome-specific subsets have been isolated for over half the chromosomes and have prove useful as markers for both genetic and physical mapping. We have developed specific oligonucleotide primer sets for polymerase chain reaction (PCR) amplification of alpha satellite DNA from chromosomes 3, 7, 13/21, 17, X, and Y. For each set of primers, PCR products amplified from human genomic DNA are specific for the centromere of the target chromosome(s), as shown by somatic cell hybrid mapping and by fluorescence in situ hybridization. These six subsets represent several evolutionarily related alpha satellite subfamilies, suggesting that specific primer pairs can be designed for most or all chromosomal subsets in the genome. The PCR products from chromosome 17 directly reveal the polymorphic nature of this subset, and a new DraI polymorphism is described. The PCR products from chromosome 13 are also polymorphic, allowing in informative cases genetic analysis of this centromeric subset distinguished from the highly homologous chromosome 21 subset. These primer sets should allow placement of individual centromeres on the proposed STS map of the human genome and may be useful for somatic cell hybrid characterization and for making in situ probes. In addition, the ability to amplify chromosome-specific repetitive DNA families directly will contribute to the structural and functional analysis of these abundant classes of DNA.     

A homologous subfamily of satellite III DNA on human chromosomes 14 and 22.

We describe a new subfamily of human satellite III DNA that is represented on two different acrocentric chromosomes. This DNA is composed of a tandemly repeated array of diverged 5-base-pair monomer units of the sequence GGAAT or GGAGT. These monomers are organised into a 1.37-kilobase higher-order structure that is itself tandemly reiterated. Using a panel of somatic cell hybrids containing specific human chromosomes, this higher-order structure is demonstrated on chromosomes 14 and 22, but not on the remaining acrocentric chromosomes. In situ hybridisation studies have localised the sequence to the proximal p-arm region of these chromosomes. Analysis by pulsed-field gel electrophoresis (PFGE) reveals that 70-110 copies of the higher-order structure are tandemly organised on a chromosome into a major domain which appears to be flanked on both sides by non-tandemly repeated genomic DNA. In addition, some of the satellite III sequences are interspersed over a number of other PFGE fragments. This study provides fundamental knowledge on the structure and evolution of the acrocentric chromosomes, and should extend our understanding of the complex process of interchromosomal interaction which may be responsible for Robertsonian translocation and meiotic nondisjunction involving these chromosomes.     

The organisation,nucleotide sequence,and chromosomal distribution of a satellite DNA from Allium cepa

We have investigated the organisation, nucleotide sequence, and chromosomal distribution of a tandemly repeated, satellite DNA from Allium cepa (Liliaceae). The satellite, which constitutes about 4% of the A. cepa genome, may be resolved from main-band DNA in antibiotic-CsCl density gradients, and has a repeat length of about 375 base pairs (bp). A cloned member of the repeat family hybridises exclusively to chromosome telomeres and has a non-random distribution in interphase nuclei. We present the nucleotide sequences of three repeats, which differ at a large number of positions. In addition to arrays made up of 375-bp repeats, homologous sequences are found in units with a greater repeat length. This divergence between repeats reflects the heterogeneity of the satellite determined using other criteria. Possible constraints on the interchromosomal exchange of repeated sequences are discussed.     

Sequence and sequence variation within the 1.688 g/cm3 satellite DNA of Drosophila melanogaster.

We have determined the complete nucleotide sequence of the monomer repeating unit of the 1.688 g/cm³ satellite DNA from Drosophila melanogaster. This satellite DNA, which makes up 4% of the Drosophila genome and is located primarily on the sex chromosomes, has a repeat unit 359 base-pairs in length. This complex sequence is unrelated to the other three major satellite DNAs present in this species, each of which contains a very short repeated sequence only 5 to 10 base-pairs long. The repeated sequence is more similar to the complex repeating units found in satellites of mammalian origin in that it contains runs of adenylate and thymidylate residues. We have determined the nature of the sequence variations in this DNA by restriction nuclease cleavage and by direct sequence determination of (1) individual monomer units cloned in hybrid plasmids, (2) mixtures of adjacent monomers from a cloned segment of this satellite DNA, (3) mixtures of monomer units isolated by restriction nuclease cleavage of total 1.688 g/cm³ satellite DNA. Both direct sequence determination and restriction nuclease cleavage indicate that certain positions in the repeat can be highly variable with up to 50% of certain restriction sites having altered recognition sequences. Despite the high degree of variation at certain sites, most positions in the sequence are highly conserved. Sequence analysis of a mixture of 15 adjacent monomer units detected only nine variable positions out of 359 base-pairs. Total satellite DNA showed only four additional positions. While some variability would have been missed due to the sequencing methods used, we conclude that the variation from one repeat to the next is not random and that most of the satellite repeat is conserved. This conservation may reflect functional aspects of the repeated DNA, since we have shown earlier that part of this sequence serves as a binding site for a sequence-specific DNA binding protein isolated from Drosophila embryos (Hsieh &; Brutlag, 1979).