* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
Download Centromeres: An Integrated Protein/DNA Complex
Survey
Document related concepts
Extracellular matrix wikipedia , lookup
Histone acetylation and deacetylation wikipedia , lookup
Organ-on-a-chip wikipedia , lookup
Cellular differentiation wikipedia , lookup
Protein moonlighting wikipedia , lookup
Endomembrane system wikipedia , lookup
Cell nucleus wikipedia , lookup
Signal transduction wikipedia , lookup
Biochemical switches in the cell cycle wikipedia , lookup
Cell growth wikipedia , lookup
Cytokinesis wikipedia , lookup
Spindle checkpoint wikipedia , lookup
Transcript
Annual Reviews www.annualreviews.org/aronline Annu. Rev. Cell. Biol. 1991.7:311-336. Downloaded from arjournals.annualreviews.org by University of North Carolina - Chapel Hill on 11/14/07. For personal use only. Annu. Rev. Cell Biol. 1991.7 ." 311 36 Copyright © 1991 by Annual Reviews Inc. All rights reserved CENTROMERESAN INTEGRATED PROTEIN/DNA COMPLEX REQUIRED FOR CHROMOSOME MOVEMENT I. Schulmanand K. S. Bloom Department of Biology, University of North Carolina, Chapel Hill, North Carolina 27599-3280 KEY WORDS: kinetochore structure, satellite DNA,nucleosomalarrays, mitosis spindle attachments CONTENTS INTRODUCTION .............................................................................................................. CENT~,OMEmC ONA ........................................................................................................ Budding Yeast .......................................................................................................... Fission Yeast ............................................................................................................ Mammalian Centromeres ......................................................................................... 311 312 312 313 314 STRUCTURAL ORGANIZATION OFDNA IN THE CELL .......................................................... ChromatinStructure of FungalCentromeres............................................................ Structural Features of Mammalian Centromeres...................................................... CENTROMERIC PROTEINS ................................................................................................ Centromere-Associated Antigens.............................................................................. Histones ................................................................................................................... Centromere DNA-Binding Factors............................................................................ Satellite DNA-Binding Proteins................................................................................ 316 316 317 REGULATION ANDASSEMBLY OFCENTROMERIC COMPONENTS ........................................... Chromosome CopyControlMutants........................................................................ Chromosome Fidelity Mutants.................................................................................. 326 327 329 EVOLUTION OFTHE KINETOCHORE .................................................................................. 329 318 318 319 320 324 INTRODUCTION Centromeres of eukaryotic chromosomes are specific regions along the chromatin fiber that play a fundamental role in chromosome movement 311 0743-4634/91 / 1115-0311 $02.00 Annual Reviews www.annualreviews.org/aronline Annu. Rev. Cell. Biol. 1991.7:311-336. Downloaded from arjournals.annualreviews.org by University of North Carolina - Chapel Hill on 11/14/07. For personal use only. 312 SCHULMAN & BLOOM during mitosis and meiosis. The centromere is a distinct DNAdomain that appears as the primary constriction along metaphasechromosomes. The centromeremediates sister chromatidinteraction and nucleates the assemblyof a kinetochore.Thekinetochoreis a region of structural differentiation that is composedof centromeric DNA and protein components and that can be visualized as a trilaminar plaqueor disc on the surface of the chromosome. Thecentromere,therefore, is responsible for delineating a chromosomal domainthat is accessible to the segregation apparatus as well as providing a scaffold for assemblyof kinetochorecomponents.The apparatusthat regulates the distribution of chromosomes to daughtercells is broughtabout by the assemblyof protein subunits into spindle tubules. Thefunctionof the kinetochore,at least in part, is to capturethese spindle microtubulesand generate polewardforces (Mitchison &Kirschner 1985). Several reports indicate that kinetochoresfacilitate microtubulebinding while allowing the microtubules to freely exchangetubulin subunits at their attached ends (Mitchison et al 1986; Salmon1989). Kinetochores mayalso be the domain where mechanochemicalmotors required for chromosome movements are localized (Rieder & Alexander1990; Nicklas 1989). Recent studies demonstrate that dynein, a force-producing molecule, is concentrated along the primary constriction of the chromosome (Pfarr et al 1990; Steuer et al 1990). A completeunderstanding of the molecularmechanisms by whichcentromericregions becomeassociated with the microtubules requires a knowledgeof the organization of centromeric DNAsequences, centromeric proteins, and other chromatin componentsspecifically associated with the centromere. This review focuses on the primaryfolding ofcentromericDNA sequencesinto nucleoprotein complexes. CENTROMERIC DNA Buddin 9 Yeast Functional centromere DNA sequences have been isolated from the budding yeasts, Saccharomyces cerevisiae (Fitzgerald-Hayeset al 1982;Hieter et al 1985), S. uvarum(Huberman et al 1986), and K. lactis (Heuset al 1990). TheS. cerevisiae centromerecan be localized to approximately125 base pairs of DNA.The sequence elements required for function (CDE centromereDNA element I, 8 bp; II, 78-86 bp of AT-rich DNA;and III, 25 bp) are conservedin their spatial and sequencearrangementin 13 yeast centromeres sequencedto date. The spacing between CDEI and III is maintained by CDEII, which is comprised of greater than 90%A+T nucleotides. Replacementof virtually every nucleotide has confirmedthe essential role for specific basesin CDEIII and for the general requirement Annual Reviews www.annualreviews.org/aronline Annu. Rev. Cell. Biol. 1991.7:311-336. Downloaded from arjournals.annualreviews.org by University of North Carolina - Chapel Hill on 11/14/07. For personal use only. STRUCTURE AND FUNCTION OF CENTROMERES 313 of AT-rich sequences in CDEII (reviewed by Gaudet& Fitzgerald-Hayes 1990). Chromosomes bearing mutations, in CDEI and II are still maintained at frequencies of 99 per 100 to 9999per 10,000 cell divisions. However,inversion of CDEIII relative to CDEI and II is deleterious to centromerefunction. Thus the requirements for proper mitotic function reside in the primarysequenceof CDEIII and in its spatial orientation relative to the AT-richCDEII. The requirement for ATsequencesjuxtaposedto a putative recognitiondomainis significant whenone considers that oligo A base pairs deformDNA with specific spatial parameters. The migrationof centromereDNA in polyacrylamidegels is noticeably altered in comparisonto bulk DNA sequences. CDEII has been projected in three dimensions using the wedgeprogramdescribed by Eckdhal & Anderson (1987).Thepredictedstructure consists of a rigid backbone that is deflected backand forth along its helical axis in a zig-zag fashion, quite distinct from other structural sequences such as bent DNA,or the bent DNA locus in kinetoplast DNA and autonomouslyreplicating sequences. This structural motif is conservedin 12 CDEII regions that exhibit minimal sequence homology.Whetherthis sequenceis deformedin accord with a subjacent protein core, or has a role in specifyingthe exquisite sequence specificity characteristic of CDE Ill is not known,but suggestsa role for structural DNA in this unique protein/DNAcomplex. CentromereDNAfrom a related species of Saccharomyes,S. uvarum, can stabilize minichromosomes in S. cerevisiae, and it contains the conserved sequenceelementsCDEI-III. It is unlikely that centromerefunction extends past the species barrier, however. CentromereDNA from anothergenusof Ascomycetes, K. lactis, does not function in S. cerevisiae (Heuset al 1990).K. lactis is a buddingyeast used primarilyin industrial fermentations with a chromosomenumber(6) intermediate between cerevisiae (16) and S. pombe(3). Five of the six centromereshavebeen isolated fromK. lactis bytheir ability to confermitoticstability to plasmids containing an autonomouslyreplicating sequence. Thefunctional domains have been localized on DNA fragments ranging from two to five kilobase pairs in length. Sequenceanalysis of these centromereswill reveal whether the K. lactis centromereis intermediatein size betweenthe S. cerevisiae and S. pombecentromeres, or contains the CDEelements characteristic of S. cerevisiaein a slightly altered sequenceand/orspatial configuration, Fission Yeast Thecentromereof the fission yeast Schizosaccharomycespombe is approximately two to three orders of magnitudelarger than the S. cerevisiae centromere (50-150 kb vs 150 bp) (Nakasekoet al 1986; Fishel et 1988; Niwaet al 1989; Clarke & Baum1990; Matsumotoet al 1990). Annual Reviews www.annualreviews.org/aronline Annu. Rev. Cell. Biol. 1991.7:311-336. Downloaded from arjournals.annualreviews.org by University of North Carolina - Chapel Hill on 11/14/07. For personal use only. 314 SCHULMAN & BLOOM Distinguishing features are blocks of repeated DNAelements that comprise this centromeric domain. A schematic representation of the sequence organization of the centromere region from S. pombe is illustrated in Figure 1. At least three middle repetitive elements K (6.4 kb), L (4.5 kb), and B (1-2 kb) flank a nonconservedsingle copy central core, cc2 (7 kb chromosome2). Independent investigations have also shown the S. pombe centromere to be made up of a unique central domain flanked by repeated elements dg (3.8 kb), dh (4.0 kb), and yn/tm (less than 1 kb) (Nakaseko al 1986; Niwa et al 1989; Chikashige et al 1989; Matsumotoet al 1990). The dh region corresponds to a portion of K-L; dg to K, and yn/tm to B, respectively. Eachof these elements, or portions thereof, occur at all three S. pombecentromeres. These repeats are in turn tandemly reiterated to generate an inverted repeating motif that flanks the central core. The detailed analysis of such large DNAfragments has only recently been madepossible through the use of yeast artificial chromosomevectors (YACs, Burke et al 1987). The large repeated sequences present in pombecentromeres are unstable in E. coli and are lost or rearranged when maintained in a bacterial host. However, when repeated domains are introduced into vectors containing S. cerevisiae centromeres and origins for DNAreplication, these sequences can be subcloned and propagated in S. cerevisiae or S. pombe without deletion or rearrangement. Upon introduction into S. pombehost cells, minichromosomescontaining entire centromere regions are faithfully segregated both during mitosis and meiosis and are stably maintained in the absence of genetic selection (Hahnenberger et al 1989; Clarke & Baum1990). Deletion analysis of functional minichromosomesindicates that the central core flanked by the short inverted repeats (B) and at least one copy of the K (dg) repeat are required for mitotic function (Matsumoto et al 1990; Hahnenberger et al 1991). Deletions of large regions of the repeats (K-L-B) have a disproportional effect on meiotic function, as witnessed by a high degree of precocious sister chromatidseparation in meiosis I. In no instance has the central core or the tandem array of repeats been shown to function independently in mitotic or meiotic cell divisions. Mammalian Centromeres A functional analysis of mammaliancentromeres has not progressed to the extent described for the yeast systems. Our knowledgeis limited to a molecular description of the abundant sequences located in or near the constricted region of mitotic chromosomes. Centromeric DNAis predominantly heterochromatic and rich in highly repetitive satellite DNA. The primary sequence motif in primates is alpha satellite or alphoid DNA (Arn et al 1989; Willard et al 1989). Alpha satellite DNAis composed Annual Reviews www.annualreviews.org/aronline 315 STRUCTURE AND FUNCTION OF CENTROMERES Annu. Rev. Cell. Biol. 1991.7:311-336. Downloaded from arjournals.annualreviews.org by University of North Carolina - Chapel Hill on 11/14/07. For personal use only. ~ 100bp S. cerevlslae CEN3 2kb S. pombe cen2 ~ 170bp ’,~ 0.68 - 6kb H. sapiens alphoid DNA ~ Fi#ure 1 Schematic representation of the centromere regions from S. cerevisiae (top), pombe(middle), and H. sapiens (bottom). (Top) The elements of centromere DNAhomology CDEI and III arc indicated. A~-rowscorrespond to nuclease cutting sites delineating the centromere core. The central oval represents the protected core, flanked by ordered nucleosomal subunits. The minimal fragment with full segregation function is drawn immediately below. (Middle) The repeated elements B, L, K are shownrelative to the single copy central core (cc2). The darkened boxes represent 1.5 kb of inverted repeat sequences that occur only in this region of the genome(adapted from Clarke 1990). The filled circles along the DNA represent nuclcosomal subunits. Nucleosomesmaybe masked or absent within the cc2. Their position with respect to the DNAsequence has not been determined. The minimal active segment is drawn immediately below. (Bottom) The alphoid repeats are arranged in dimers or pentamers, which in turn are organized into 0.68-6.0 kb sequences. The larger units are repeated 100-1000 times along the centromere (Arn et al 1989). Nucleosomesoccupy one major and multiple minor phases with respect to the repeat sequence. These are simply indicated by ovals. Note the scale differences in the various organisms. Annual Reviews www.annualreviews.org/aronline Annu. Rev. Cell. Biol. 1991.7:311-336. Downloaded from arjournals.annualreviews.org by University of North Carolina - Chapel Hill on 11/14/07. For personal use only. 316 SCHULMAN & BLOOM tandemarrays of 170 bp polymeric units that exhibit 60-95%homology betweenindividual units. Dimeror pentamericunits of the alphoid DNA, as shownin Figure l, are in turn organizedinto larger clusters ranging from 0.68-6.0 kilobase pairs (Willard &Waye1987). Hierarchical arrays of these clusters comprise up to millions of base pairs of DNA.DNA sequencingof cloned repeat units, along with conventionalrestriction mapping, have revealed numerouspolymorphismsamongdifferent chromosomes as well as within chromosome-specific subsets. There is a tendency for monomerson one chromosome to be more conserved than those fromdifferent chromosomes. In addition, the organizationof large blocks of repeats within a chromosome are specific to that chromosome (Wervick & Willard 1989; Grieg et al 1989; Mahtani & Willard 1990). With the advent of techniques to manipulate large fragments of DNA and to construct large mammalian artificial chromosomes, the identification of mammalian centromeres is soon likely to follow. The paridigms for DNA manipulationwithin the’ centromereregion of S. pombewill certainly be invaluable in subsequent functional analysis of their mammalian counterparts. STRUCTURAL ORGANIZATION OF DNA IN THE CELL ChromatinStructure of Fungal Centromeres s. CEREVISIAE Molecularmappingtechniques have been used to visualize the chromatinstructure of four S. cerevisiae centromeres(CEN3,CEN11, Bloom& Carbon 1982; CEN4,Bloomet al 1984a; CEN14,Funk et al 1989). Approximately150 to 200 bp of DNA are folded into a highly nuclease-resistantcore structure. Thesingle base changesin the central C nucleotide of CDEIII, whicheradicate segregation function, completely disrupt the characteristic protected structure. TheCDE I and II mutations that retain partial centromerefunction exhibit chromatinconformations of altered dimension(Saunderset al 1988). Deletions of CDEI decrease the fidelity of chromosome transmissionfive- to tenfold, andthe protected structures of these altered centromeresare shortenedby 10-20 base pairs. Mutations in which the length between CDEI and CDEIII is increased or decreasedexhibit lengthenedor shortenedprotected structures, respectively, and have intermediate effects on chromosome segregation. The chromatinparticles associated with these defective centromerescan range from 145 to greater than 300 bp. Interestingly, mutants with less than 145 bp of centromere DNAincorporate flanking DNAinto protected structures that span 145 bp of chromatinDNA. Theplasticity in structural organization indicates that the chromatincore is likely to represent a Annual Reviews www.annualreviews.org/aronline STRUCTURE AND FLrNCTION OF CENTROMERES 317 Annu. Rev. Cell. Biol. 1991.7:311-336. Downloaded from arjournals.annualreviews.org by University of North Carolina - Chapel Hill on 11/14/07. For personal use only. complexinteraction betweencentromere-specific DNA-binding proteins, histones, and centromere DNA. s. POtaBE The structural organization of the centromererepeats and the unique central core have recently been determinedby nuclease digestion studies (Polizzi & Clarke 1991). The large repeated domains,K and exhibit nucleosomearrays that typify the bulk of the chromatin DNA. These nucleosomalrepeats are 155 bp, with an average spacer length of 10 base pairs. Theprecise position of nucleasecutting sites with respect to the DNAsequence has not been determined. The nucleosomalarrays are not as ordered as in S. cerevisiae, whichindicates the availability of alternative registers of nucleosomal arrays in a givencell. Onesuchordered array of nucleosome packagingis illustrated in Figure1. Themoststriking feature of the chromatinstructure is the absenceof ordered nucleosomes in the central core. It is unlikely that this regionis completelydevoidof nucleosomesbecause monomersand dimers can be visualized following micrococcalnuclease digestion. However,the spacing is randomizedor maskedto an extent that precludes the appearanceofnucleosomeladders. Thejunction betweenthe central core and the flanking repeats reveals a third pattern that is intermediatebetweenthe typical nucleosome ladders of K and L and the randomizedcentral core. Thenucleosomeswithin the B repeat and the core-associated repeats (darkenedblocks, Figure 1) are less distinct andexhibit higher levels of background hybridization. While the central domainlacks an organized nucleosomearray, it appears that in the junction region a moreorganizedstructure begins to formand by repeat L is arrayedin multipleregisters distally fromthe centromere.The lack of nucleosomestructure in the central core does not reflect any inherent nucleotide sequencethat precludes nucleosome assembly.Polizzi & Clarke (1991) demonstratedthe assemblyof nucleosomesubunits within the central core whenthese sequenceswereintroducedinto S. cerevisiae. Thusthe functional properties of the centromerein fission yeasts are intertwinedwith this apparentlyatypical subunit structure. It is possible that histone subunits are predominantly absent, modified, or occupy limited sites within this region. Alternatively,the bindingof centromerespecific nonhistoneproteins mayprevent exogenousnucleolytic digestion of the central core. Thechromatinorganization of the centromereregion in S. pombeis certainly a complexstructure of several functional domains and mayrepresent a compound version of that found in S. cerevisiae. Structural Features of Mammalian Centromeres Thechromatinstructure of the alphoid satellite has been studied extensively in the African green monkey.AlphoidDNA is packagedinto highly Annual Reviews www.annualreviews.org/aronline Annu. Rev. Cell. Biol. 1991.7:311-336. Downloaded from arjournals.annualreviews.org by University of North Carolina - Chapel Hill on 11/14/07. For personal use only. 318 SCHULMAN & BLOOM organized nucleosomearrays (Zhanget al 1983; van Holde1989), which are positionedin one of eight phases (one major andseven minorphases). The observation of multiple phases for nucleosomepositioning is also apparentin the rat satellite I and mousesatellite chromatin.Suchpatterns of multiplephasingare at least consistent with the data obtainedin the K and L repeats in the S. pombecentromere. Nucleosomesare present on the KandL repeats, but mayoccupyalternate positions in different cells. The binding of nonhistone chromosomalproteins to specific sequences maybe an importantfactor in the delineation of various phases in individual cells and the subsequentcompactionof heterochromatin. Several nonhistone proteins, one of whichis a kinetochore component,bind to specific regions in the alpha satellite sequence.CENP-B (Masumoto et al 1989), HMG-I (Solomonet al 1986; Disney et al 1989), and D1(Levinger & Varshavsky1982) bind to A-T rich regions of chromosomes,and are likely candidates in the compactionof centromericheterochromatin(see below). CENTROMERIC PROTEINS Centromere-Associated Antigens The identification of centromeric proteins in mammalian cells has been facilitated by the discoveryof kinetochore-specificautoantibodiesin the serumof patients with the calcinosis/Raynaud’s phenomenon/esophageal dysmotility/sclerodactyly/telangiectasia variant (CREST) of scleroderma. Thedisease is a progressivesystemicsclerosis that can affect manyorgans includingskin, subcutaneous tissue, gastrointestinal tract, heart, lung, and kidney. Moroiet al (1980) demonstratedthat patients with this disorder contain autoantibodies that react with the kinetochore of chromosomes. The characterization of CREST antigens [centromere proteins (CENP)-A, B, C, D, and E], as well as inner centromereproteins (INCENPs) and chromatid-linkingproteins (CLiPs), has been instrumental in delineating distinct domainswithin the primary constriction of metaphasechromosomes [Brinkley 1990; Cookeet al 1987; Earnshaw& Rattner 1989 (see Figure1 therein); Rattner et al 1988;Rattner 1991;Yenet al 1991;Willard 1991]. Theseproteins partition into three distinct regions within the primary constriction, or centromere domain. The kinetochore domainis assembledon the outer surface of the centromereand is comprisedof five subdomains:a fibrous corona, an outer, middle, and inner plate, and subjacent chromatinfibrils. The central domainof the centromerelies belowthe kinetochore plates and is followed by the innermost pairing domain,whichis responsiblefor tethering sister chromatidstogether prior to their separation. Annual Reviews www.annualreviews.org/aronline Annu. Rev. Cell. Biol. 1991.7:311-336. Downloaded from arjournals.annualreviews.org by University of North Carolina - Chapel Hill on 11/14/07. For personal use only. STRUCTURE AND FUNCTION OF CENTROMERES 319 CENPsA and C are located in, but not restricted to, the inner kinetochore plate. CENP-Ais a histone H3 variant and may be involved in the subunit organization of the centromeric DNA(Palmer et al 1990; see below). CENP-Cis restricted to active centromeres (Earnshaw et al 1989) and is equally distributed between different chromosomes. The microtubule-based mechanochemicalmotor protein, dynein, has also been localized to the kinetochore domain(Pfarr et al 1990; Steuer et al 1990). The ability ofdynein to translocate along a microtubule lattice has considerable impact with respect to chromosomesegregation mechanisms. The accessibility of dynein to microtubules is most likely maintained in the kinetochore structure, thus implicating dynein as a primary candidate for the proteinaceous substance that comprises the outermost fibrous corona of the kinetochore. CENP-Bis the most extensively characterized CRESTantigen to date and binds the s-satellite DNA(see below). It is localized throughout centromeric heterochromatin in the central domain beneath the kinetochore (Cooke et al 1990). The distribution of CENP-Bvaries between different chromosomesand correlates well with the amountof s-satellite DNAper chromosome. CENP-B-~-satellite DNAcomplexes may comprise the bulk of the central domainand could well provide an internal frameworkrequisite for the positioning of the kinetochore domain. INCENPs and CLiPs are localized to the inner surface of the centromere and along the arms where contacts between sister chromatids occur. INCENPs and CLiPsare likely to represent at least two classes of proteins, based from their behavior in anaphase. INCENPs migrate to the equatorial zone of the spindle and are destined for the midbody, whereas the CLiPs are no longer detected following chromatid separation. Additional centromere-associated proteins have recently been identified by raising antibodies against mitotic chromosomescaffolds. Comptonet al (1991) generated antibodies to four classes of centromere-associated proteins distinguished by their staining characteristics throughout the chromosomecycle. Several of these antigens localize to centrosomes, the microtubule organizing centers of the cell, as well as metaphase chromosomes. The organization of centromere and kinetochore domains into visible structural entities will involve diverse classes of proteins that contribute both to the fidelity of kinetochore assembly as well as to active chromosome segregation. Histones S. CEREVISIAE Histones are the major structural proteins required for chromosomepackaging into nucleosomes, and their stoichiometric production is a contributing factor to the fidelity of chromosome segregation (Meeks-Wagner & Hartwell 1986). Analysis of specific roles for histone Annual Reviews www.annualreviews.org/aronline Annu. Rev. Cell. Biol. 1991.7:311-336. Downloaded from arjournals.annualreviews.org by University of North Carolina - Chapel Hill on 11/14/07. For personal use only. 320 SCHULMAN & BLOOM chromosome organizationhas beenfacilitated by the introduction of single copies of histone H2Bor H4, whoseexpression can be modulateddepending on the carbonsource(Hanet al 1987). Deletionof the aminoterminus of histone H4is dispensablefor cellular growth,but essential for transcriptional repressionof the silent matingtype loci (Kayneet al 1988).Thus histone-nonhistoneprotein interactions are likely to be as intricate and essential as histone-DNAinteractions for chromosomestructure and function. Repressionof either H2Bor H4histones also leads to the increased nucleaseaccessibility of the centromere(Saunderset al 1990). Structural perturbation of the centromere complexfollowing histone repression reflects either the presenceof histones at the centromere,or a role for flanking nucleosomal chromatin in centromere assembly and/or maintenance. Strains harboring an aminoterminal deletion of histone H4(428 aa) as the only source of H4are able to grow; however,bulk chromatin stru6ture is significantly disrupted(Kayneet al 1988). B. Frediani and Bloom(unpublished results) demonstratedthat centromere structure these strains is unaltered.Thuscentromerestructure is not simplyreflective of bulk chromatin perturbation caused by histone deletions. The dissociation pattern of centromereproteins following salt elution of the chromatincomplexis consistent with a modelin whichhistones are present at the centromere (Bloom& Carbon1982). At NaCIconcentrations that dissociate histones (0.75-1.25M), there is a concomitantdisruption of the centromere core, while at lower salt concentrations, the centromere remainsunaffected. In addition, the linking numberof centromereplasmidsisolated fromlogarithmically growingcells is indistinguishable from plasmids of identical size, whichcontain exclusively nucleosomalDNA (Bloomet al 1984b). Thesedata led us to consider a histone structural base for the centromereitself. MAMMALIAN CENP-A,the centromere-associated antigen localized to the inner plate of the kinetochore,is tightly associated with mononucleosome particles following digestion of chromatin with micrococcal nuclease (Palmeret al 1989). Peptide sequenceanalysis has identified CENP-A be a histone H3 variant (Palmer et al 1990). Thushistone proteins, modifiedhistones are also localized within mammalian kinetochores. Centromere DNA-Bindin9 Factors CENTROMERE-BINDING FACTOR 1 (CBF1) The isolation of a centromere DNA-bindingprotein that recognizes the centromere DNAelement I (CDEI) octamerof S. cerevisiae has beenreported by a numberof groups (CBF1,Cai & Davis 1989, 1990; centromereprotein 1, CP1,Baker et al Annual Reviews www.annualreviews.org/aronline Annu. Rev. Cell. Biol. 1991.7:311-336. Downloaded from arjournals.annualreviews.org by University of North Carolina - Chapel Hill on 11/14/07. For personal use only. STRUCTURE AND FUNCTION OF CENTROMERES 321 1989; Baker & Masison 1990; centromere and promoter factor 1, CPFI, Mellor et al 1990). CBF1is an abundant cellular protein (600 copies per cell, Bakeret al 1989) that is not restricted to the centromericregion (Bram & Kornberg 1987). CBF1 binds both centromeres and promoters and exhibits the helix-loop-helix dimerization motif characteristic of other DNA-binding proteins involved in transcriptional control. Complete deletion of the coding information for CBF1 from the yeast genomeresults in a five- to tenfold increase in chromosome malsegregation, commensurate with the increase in chromosomeloss upon deletion of the CDEIoctanucleotide. The pleiotropic nature of the binding interactions is manifested in a genetic sense as well. CBF1deletion strains are methionine auxotrophs. Thus both genetic and biochemical criteria indicate that this protein participates in diverse cellular functions. In addition to the helixloop-helix domain, 20%of CBF1 is comprised of acidic residues, aspartate and glutamate. The charged nature of CBF1exhibits similarity to the HMGproteins and the highly acidic domains in CENP-B.Domain mapping of CBF1is in progress, and deletion of two clusters of negative charges (amino acids 80-88 and 184-194) only increases the chromosome loss rate by a factor of 1.5 comparedto the wild-type protein (Mellor et al 1990). The proximity of negatively charged domainsto the centromere mayreflect a role in facilitating the binding of key nonhistoneproteins, or other componentsthat must access an otherwise highly compacted region. CENTROMERE-BINDINGFACTOR 3 (CBF3) The difficulty in demonstrating the binding of a factor to centromere DNAelement CDEIII by similar methodological approaches may reflect the low copy number of such molecules, or the lack of faithfully reproducing the in vivo assembly pathway and/or chromatin requirements. Lechner & Carbon (1991) recently succeeded in identifying centromere DNA-bindingfactors that specifically recognize the CDEIII element. CBF3is a 240-kd phosphoprotein polymeric complex of at least three monomericspecies (110, 64, and 58 kd; CBF3A,B, C, respectively) that are capable of discriminating wild-type CDEIII from an inactivating point mutation in CDEIII. The proteins are present in vanishingly small amountsin the cell. Conservative estimates from the purification procedures show that there are only 20 copies per yeast cell in a logarithmically growingpopulation. If these proteins are restricted to a particular stage of the cell cycle, however, they maybe present in higher quantities. In any case, it remains that the CBF3complex is at least an order of magnitude less abundant than CBF1. CBF3protects a 56 bp region from DNAaseI digestion when complexed to CENDNA.The complex is asymmetrically distributed over CDEIII; Annual Reviews www.annualreviews.org/aronline Annu. Rev. Cell. Biol. 1991.7:311-336. Downloaded from arjournals.annualreviews.org by University of North Carolina - Chapel Hill on 11/14/07. For personal use only. 322 SCHULMAN & BLOOM 6 base pairs of CDEII to the left, and 24 base pairs of flanking DNAto the right are protected in the binding reaction. Perhaps the most striking feature of the reaction is the requirement for assembly factors. Extracts from either budding yeast, E. coli, or purified casein from bovine milk, are necessary but not sufficient to support assembly. Ancillary protein factors required in assembly of macromolecular structures, such as the centromere DNA-protein complex, may prevent nonproductive proteinprotein interactions, while promoting more favorable interactions. The role of proteins, such as nucleoplasmin, in facilitating nucleosome assembly sets precedence for invoking a similar requirement in centromere assembly. The ability of the acidic protein casein to substitute for an endogenousassembly factor(s) suggests a role for acidic domains in promoting centromere DNA-protein assembly. The structure of the centromere DNAmay contribute to the binding specificity as well. Proteins that stabilize the natural curvature of DNA,such as histones and high mobility group (HMG)proteins, may play a role in positioning the centromere DNAelements in a conformation that enhances and/or stabilizes the binding of critical nonhistone proteins. In this regard it is noteworthy that while only 6 bp of CDEII are complexed in the CBF3footprint, at least 20-30 bp of flanking A + T DNAon the left are required for in vivo function. The additional 15-25 runs of A+Tnucleotide may position CDEIII around the subjacent protein core, but may not be in direct contact with CBF3itself. AFFINITY CHROMATOGRAPHY OF CENTROMERIC CHROMATIN The complexity of form and function of yeast and mammaliankinetochores has directed efforts in our ownlaboratory towards the isolation of an intact chromatin complex from yeast cells. Restriction enzymelinkers encoding the BamHIrecognition site have been ligated to a 289 bp CEN3DNA segment encompassingthe 220-250 bp protected structure. An entire functional segment of centromeric chromatin can be excised from the chromosomeof S. cerevisiae by BamHIdigestion (Kenna et al 1988). The ability to excise an intact functional core segment, which retains centromeric proteins, offers the opportunity to analyze this autonomous DNA-protein complex. Release of this complex from any torsional constraints exerted by the flanking chromosomalarms does not perturb protein binding to centromere DNA.Differential sedimentation in linear glycerol gradients reveals biochemical differences between naked DNA,a functional centromere, and a nonfunctional altered centromere. However, the methodis not sufficiently selective to visualize centromere proteins. An affinity method was developed for protein-DNA complexes that makes use of the interaction between the lac repressor and its operator (Figure Annual Reviews www.annualreviews.org/aronline STRUCTURE AND FUNCTION OF CENTROMERES CEN3 289bp I EcoRI I lacO CEN3 Annu. Rev. Cell. Biol. 1991.7:311-336. Downloaded from arjournals.annualreviews.org by University of North Carolina - Chapel Hill on 11/14/07. For personal use only. I EcoRI lacO I BamHI I I EcoRI I BamHI BamHI 323 I II EcoRI BamHI lacl-lacZ Anti-lacZ Sepharose Figure 2 Affinity chromatography of centromeric chromatin. A strategy for affinity purification of centromeric chromatin has been devised that exploits the interaction between the E. coli lac operator (lacO) and its repressor (lacI) (Saunders 1989; Deanet al 1989). has been cloned adjacent to centromere DNAand introduced into yeast cells. The IacOCENcassette is excised from yeast nuclei by digestion with EcoRI and subsequently incubated with lac repressor-fl-galactosidase fusion protein (lacIlacZ). The complexis applied to sepharose column containing immobilized antibodies directed against the fl-galacatoside portion of the fusion protein. Treatment of the immobilized complex with IPTGresults in release of centromeric chromatin. 2). A symmetrical lac operator DNAsequence was synthesized and cloned adjacent to the centromere DNA.This lac operator-CEN cassette was introduced into yeast by transformation, and centromeres assembled on this construct in vivo. The intact cassette was excised from the cell nucleus with restriction enzymes and incubated with a fl-galactosidase-lac repressor fusion protein that specifically binds lac operator. The tripartite complex of centromere proteins, centromere-lac operator DNA,and fusion protein was purified by passage through anti-fl-galactosidase Annual Reviews www.annualreviews.org/aronline 324 SCHULMAN & BLOOM sepharose. The complex was specifically eluted with isopropyl-B-D-thiogalactopyranoside (IPTG), a competitive inhibitor of the lac repressor. Wehave constructed a similar cassette containing a lac operator-mutant centromere that is completely deficient in its segregation function and intend to study its intracellular interactions. Comparisonsbetween the proteins associated with wild-type and mutant centromeres should allow definitive identification of in vivo centromere DNA-bindingproteins. Annu. Rev. Cell. Biol. 1991.7:311-336. Downloaded from arjournals.annualreviews.org by University of North Carolina - Chapel Hill on 11/14/07. For personal use only. Satellite DNA-Bindin# Proteins HIGHMOBILITY GROUP PROTEINS Several proteins have been isolated that appear to interact specifically with highly repetitive A-I-T-rich satellite DNA.Two of these proteins, D1 from Drosophila melano#aster and HMG-Ifrom mammaliancells, were originally characterized as members of the high mobility group (HMG)family of nonhistone chromosomal proteins (Alfagemeet al 1980; Lurid et al 1983). HMGs are characterized by their solubility in low salt (0.35 MNaC1)and 5%perehloric acid, and an abundancy of both basic and acidic amino acids. The basic and acidic amino acids are distributed in a polar fashion with the amino-terminal region predominantly basic and the carboxy-terminus acidic (see Johns 1982). Immunofluorescent staining of HMG-Iand D 1 reveals their localization to predominantly satellite heterochromatin, including centromeres (Alfagemeet al 1980; Disney et al 1989). Morerecently another Drosophila protein, HP-1, was identified; it is localized to centric heterochromatin and is responsible in part for position-effect variegation. HP-1is a low molecular weight protein (18,000 K) that exhibits high levels of acidic and basic residues typical to HMGs and has one stretch of six glutamic acids (James &Elgin 1986; James et al 1989). Both D 1 and HMG-Ihave been shown to interact with A + T-rich DNA in vitro (Levinger & Varshavsky 1982; Solomon et al 1986; Reeves Nissen 1990). Analysis of the amino acid sequences of D1 and HMG-I has identified a potential DNA-bindingmotif repeated ten times in DI and three times in HMG-I. The motif consists of Gly-Arg-Pro (GRP) located within a cluster of basic aminoacids (Ashely et al 1989; Reeves Nissen 1990). A synthetic eleven amino acid peptide of the consensusbinding domain is able to bind the minor groove of A+T-rich DNA (Reeves & Nissen 1990). The predicted secondary structure of this peptide is similar to the structure of A + T-rich DNA-bindingdrugs, netropsin and distamycin, as well as Hoechst 33258. These compoundsdisplace HMGI from naked DNAand disrupt the structure of kinetochores in whole cells (Lica et al 1986). Structural studies of HMG-nucleosome complexes place HMG-likeproteins along the DNAduplex as it enters and exits the histone octomer of a single nucleosome. Considering that the specificity Annual Reviews www.annualreviews.org/aronline STRUCTURE AND FUNCTION OF CENTROMERES 325 Annu. Rev. Cell. Biol. 1991.7:311-336. Downloaded from arjournals.annualreviews.org by University of North Carolina - Chapel Hill on 11/14/07. For personal use only. for nucleosomepositioning can be reconstituted in purified systems with alphoid DNAand purified histone octamers (Linxweiler & Horz 1985), the HMG proteins are likely to confer additional structural information. Mutations in HP-1 that directly impinge upon heterochromatin physiology indicate that HMG-likeproteins play important roles in higher levels of heterochromatin condensation and subsequent formation of the kinetochore. CENP-BCENP-Bis an acidic protein that binds to alphoid satellite DNA (Masumotoet al 1989; Sullivan & Glass 1991). The amino acid sequence of CENP-Breveals the presence of two highly acidic domains located in the carboxy-terminalhalf of the protein (Earnshawet al 1987). In addition, CENP-Bcontains a high percentage of basic amino acids (approximately 12%). Almost half of these amino acids are located in the amino-terminal quarter of the protein. The overall charge of this region is + 17. The last 50 amino acids of the protein are also highly basic (22%, overall charge + 19). Thus CENP-Bhas an HMG-like composition. In addition, CENPB contains one copy of the tripeptide Gly-Arg-Pro (GRP); however, unlike HMG-Iand D1 DNA-bindingdomains this tripeptide is not imbedded in a cluster of basic aminoacids. The ability to extract CENP-B from nuclei with relatively low salt (0.5 MNaCI; Masumotoet al 1989) is also consistent with an HMG-likecharacterization. It has been reported that CENP-Bis an abundant component of the insoluble mitotic scaffold (Earnshawet al 1984). CENP-B maybe subject to cell-cycle modifications, and/or its interphase properties may differ from CENP-B found in mitotic preparations. Alternatively, differences in extraction protocols may account for the enhanced dissociation of these antigens from interphase cells compared to their mitotic counterparts. CENP-Bis an abundant protein (20,000-50,000 copies per cell) that binds individual chromosomes according to their alphoid DNAcontent (Earnshaw et al 1987). Thus CENP-Bbinds alphoid regions in a stoichiometric fashion in vivo. Injection of antibodies against CENP-Band other kinetochore antigens into living cells directly implicates their functional role in chromosome movementas well (Simerly et al 1990; Bernat et al 1990; Yen et al 1991). Interestingly, injection of the antibodies against predominantly CENP-B only blocks chromosome movements prior to the metaphase/anaphase transition. Once the cells traverse this point, the kinetochore antigens cannot be neutralized by the antibodies. As discussed below, these data indicate at least a two-step process in the assembly of a kinetochore complex: protein DNArecognition and maturation into an active segregation complex. Annual Reviews www.annualreviews.org/aronline 326 SCHULMAN Annu. Rev. Cell. Biol. 1991.7:311-336. Downloaded from arjournals.annualreviews.org by University of North Carolina - Chapel Hill on 11/14/07. For personal use only. REGULATION COMPONENTS & BLOOM AND ASSEMBLY OF CENTROMERIC The dynamicsof centromerereplication, assembly, and maturation have been studied in greatest detail at the molecularlevel in the yeast, S. cerevisiae. CentromericDNA is replicated early in the S-phaseof the cell cycle (McCarroll & Fangman1988) and is correlated temporally with replication of the spindle pole bodies. Theuniquechromatinstructure is present at all stages of the cell cycle (Yeh1985; E. Yeh& K. Bloom, unpublishedresults). OnlywhenDNA replication is allowed to proceed in the absenceof histone synthesis does the centromerebecomeexposed to nucleolytic digestion (Saunders et al 1990). Thus the assembly centromeric chromatin is likely to be coordinated with centromere DNA replication. Thestructural assemblyof a limited numberof critical proteins such as CBF3requires their nonrandomassembly. These data are therefore consistent with the view that centromeric componentsprovide a template for formation of newstructures as the replication fork proceeds. Amaturationstep can be defined as the transition from protein binding to centromereDNA to a centromerethat actively segregates chromosomes. Protein bindingitself is not sufficient fromchromosome segregation. This is mostevident fromstudies in whichthe protein-bindingcapacity of the centromereDNA has been uncoupledfrom its partitioning function. The centromere can be conditionally inactivated by transcription from an adjacent promoter (Hill & Bloom1987). The conditionally inactivated centromereis organizedinto a protected structure characteristic of wildtype centromere.Determination of the initiation site and length of adjacent RNAsupontranscriptional activation reveal that the transcripts do not penetrate the centromere core (Hill 1988; Russo & Sherman1988). The terminationsite of the transcripts mapsto the boundariesof the protected core structure and confirms that a portion of the centromeric proteins remain bound to the DNA under conditions of centromere inactivation. In contrast, particular mutations in the centromere DNAthat disrupt function, disrupt structure as well (Saunderset al 1988). Thusthe conditional centromereand selected point mutations in the centromereDNA are indistinguishablein their loss of function, but differ in the mechanism of inactivation. Theseobservationsprovide a framework for distinguishing mutations in protein binding at centromere, from mutations in proteinprotein interactions, or post-translational modifications, whichare required for the maturation of this complexinto an active partitioning element. Annual Reviews www.annualreviews.org/aronline STRUCTURE AND FUNCTION OF CENTROMERES Annu. Rev. Cell. Biol. 1991.7:311-336. Downloaded from arjournals.annualreviews.org by University of North Carolina - Chapel Hill on 11/14/07. For personal use only. Chromosome Copy Control 327 Mutants Competitionfor the binding of a limited pool of centromereproteins or other spindle components is manifestedin the severe limitations for extra centromeresin yeast, first demonstratedby Futcher & Carbon(1986). Wild-typeyeast cells contain 16 chromosomes, and therefore 16 centromeres. Thecells can tolerate 1-2 extrachromosomal centromereswithout any measurablegrowthor segregation defect. The application of genetic pressure to maintain an increased numberof centromeres(greater than five to ten) results in cell death. Theuse of geneticmarkersthat complement a deficiency in a gene dosage-dependentfashion provides a meansfor screening heterogeneity in plasmidcopy numberin a population of cells in the absenceof geneticselection pressures. Resnicket al (1990)used the CUP1gene, whichprovides resistance to exogenouscopper, to demonstrate limitations in centromere copy numberof approximatelyten per cell. Thusboth genetic selection for increased copy numberand indirect assessmentof copy numberby gene dosage reveal limited tolerance for centromere DNA. Relaxation of stringent copy control mechanisms mayreveal mutations in critical components that are otherwiselimiting in the cell. Theuse of gene dosage markers provides a meansfor determining plasmid copy number, as well as providing selection for mutations in copy number control. The drug-resistant marker, G418(Tschumper& Carbon 1987), and particular mutationsin the leu2 geneexemplifysuch dosage-dependent markers. The leu2-d allele contains a truncated promoterthat is only partially functional (Runge&Zakian1989), and therefore is required high copy number(100-200copies per cell) to efficiently complement leu2 auxotrophs. Fewer than 10 copies of leu2-d results in poor complementation, or a complete failure to grow, dependingon the particular strain of yeast. Theleu2-d allele has beenclonedinto plasmidsharboring a variety of wild-type and mutantcentromeres.Results from quantitative hybridization to yeast colonies containing these plasmidsreveal a highly elevated plasmidcopynumberwithout the centromere,low levels of wildtype centromere plasmids, and intermediate levels of plasmids with a mutant centromere(central C residue to CDEIII changedto A) (Figure 3). This eis mutation in centromere DNA is clearly not subject to the stringent copycontrol of wild-typecentromeres,but is regulated compared to vector sequenceswith no centromere.This quantitative assay for plasmid copy numberreveals measurable protein-binding capacity of centromeremutantsthat are otherwisedefective in their segregationcapacity. Whena conditional centromerewas inactivated by growthon galactose, Annual Reviews www.annualreviews.org/aronline Annu. Rev. Cell. Biol. 1991.7:311-336. Downloaded from arjournals.annualreviews.org by University of North Carolina - Chapel Hill on 11/14/07. For personal use only. 328 SCHULMAN & BLOOM Figure 3 Quantitative analysis of minichromosome copy number by colony hybridization. S. cereuisiue cells contain minichromosomes carrying the leu2-dallele (see text) and no centromere (top); mutant centromeres containing A in place of the central C residue of CDE 111 that inactives segregation function (middle); or wild-type centromcrc (bottom). Cells were grown directly onto nitrocellulose filters in the absence of leucine. The colonies were lysed on the filters and hybridized to "P-labeled pBR322 DNA to visualize exclusively plasmid sequences. After hybridization, filters were washed and exposed to X-ray film. Annual Reviews www.annualreviews.org/aronline STRUCTURE AND FUNCTION OF CENTROMERES 329 only 10-20 copies accumulated per cell (I. Schulman& K. Bloom, unpublished results). These data are consistent with the observation that conditionally inactive centromeres bind limiting componentsin the cell, but are uncoupled from their segregation function. The ability to dissect the cell’s tolerance for centromere DNAfrom its segregation function should allow us to separate mutations in protein binding from the more complex partitioning functions characteristic of the centromere. Annu. Rev. Cell. Biol. 1991.7:311-336. Downloaded from arjournals.annualreviews.org by University of North Carolina - Chapel Hill on 11/14/07. For personal use only. ChromosomeFidelity Mutants A number of genetic screens have been developed to search for genes whose products are involved in chromosometransmission. These screens typically exploit a genetic trait that can be monitored in a visual colony assay. A numberof genes have been identified and will certainly comprise proteins involved in replication, assembly, and segregation (McGrewet al 1989; Meeks-Wagneret al 1986; Spencer et al 1988, 1990; Gerring et al 1991). Identification of critical componentsin the segregation machinery per se will require secondary screens to distinguish regulatory and structural components. Characterization of the gene product of one gene required for the fidelity of chromosome transmission reveals protein kinase domains (Shero & Hieter 1991). Considering the role of phosphorylation in the binding specificity of CBF3and the activation of cell cycle cascades, the identification and characterization of protein kinases are important steps towards understanding processes of assembly and maturation of centromeric components. EVOLUTION OF THE KINETOCHORE Centromeres from both S. cerevisiae and S. pombe contain regions of single-copy DNAsequence that are required for proper centromere function. At this time there is no evidence for the presence of unique sequence DNAas a major component of human centromeres (Willard et al 1989). The current observations suggest that the entire region of humancentromeres maybe permissive for kinetochore assembly. The exact placement of the kinetochore could be randomwithin the larger centromeric domain (Radic et al 1987) and guided by the initial deposition of critical kinetochore components. The assembly of the segregation machinery is likely to represent the compilation of repeating structural units. A threshold of critical mass wouldbe necessary to ensure the capture and stabilization of kinetochore mictotubules that emanate from the centrosome. Annual Reviews www.annualreviews.org/aronline Annu. Rev. Cell. Biol. 1991.7:311-336. Downloaded from arjournals.annualreviews.org by University of North Carolina - Chapel Hill on 11/14/07. For personal use only. 330 SCHULMAN & BLOOM Tandemly repeated arrangements of kinetochore units in mammalian systems are an attractive mechanismto explain an apparent disparity in centromere DNAorganization throughout phylogeny (see Brinkley et al 1989). The coordination of multiple centromeres in mitotically stable arrangements can be addressed in the artificial chromosomesystems. Many dicentric chromosomes are unstable during mitosis (McClintock 1939; Mann& Davis 1983; Oertel & Mayer 1984; Vig & Paweletz 1988; Koshland et al 1987; Hill & Bloom1987, 1989). However, Koshland et al (1987) demonstratedthat mitotic instability of small circular dicentric molecules was a function of the distance between the two centromeres. The closer the two centromeres were in proximity; the more stable the artificial chromosome. E. Amaya& K. Bloom (unpublished results) juxtaposed two centromere core structures on a chromosomeand found these molecules to be very stable. The structural organization of adjacent centromeres revealed two distinct centromeric cores. Thus it is unlikely that one centromere precludes the other from forming. Rather, the two seem to be coordinately controlled and attached to the same pole, tandemly arranged in a physically stable structure. Stable dicentric chromosomesin mammaliancell are typified by one functional centromere (primary constriction) and one nonfunctional centromere, recognized by C-band staining and, in some cases, by the CREST antisera (Merry et al 1985; Earnshaw & Migeon 1985; Zinkowski et al 1986; Peretti et al 1986; Cherry & Wang1988; Vig & Paweletz 1988). Cytological examination of dicentric chromosomesindicates that in most cases the inactive centromere appears smaller. Examination ofcentromere antigens by Earnshaw et al (1989) revealed the presence of CENP-B both functional as well as nonfunctional centromeres, consistent with its binding to alphoid DNA.However, another centromere antigen, CENPC, seems to be restricted to the active centromere. The inactive centromere maybe devoid of key binding sites for segregation function, or altered in such a way that DNApackaging and/or size precludes visualization of antibody staining. A mechanismfor the loss of centromeric DNAis suggested from studies in S. cerevisiae that indicate monocentricchromosomes can be derived from dicentrics, by excision of a small region of DNA containing one of the centromeres (Hill & Bloom 1989; J. Brock & K. Bloom, unpublished observations). The breakage-fusion-bridge cycle described by McClintock(1939) would also result in centromeric deletions until segregation function ceases. These remnant centromeres maycontain a sub-threshold level of repeats that are competent to bind a set of kinetochore proteins, but unable to attain the critical mass necessary for capturing microtubules. Microtubule attachments that persist on individual units would support Annual Reviews www.annualreviews.org/aronline Annu. Rev. Cell. Biol. 1991.7:311-336. Downloaded from arjournals.annualreviews.org by University of North Carolina - Chapel Hill on 11/14/07. For personal use only. STRUCTURE AND FUNCTION OF CENTROMERES 331 the view of a mammalian centromere comprised of tandem arrays of smaller functional units. A recent report has demonstrated that kinetochores are comprised of such autonomous elements. Brinkley and his co-workers (Brinkley et al 1988) utilized caffeine to induce cells blocked with hydroxyurea to segregate their unreplicated chromosomes[mitosis with unreplicated genomes (MUGs)]. The kinetochores were fragmented and physically detached from the bulk of the chromosome in MUGed cells. Kinetochores retained the characteristic trilaminar ultrastructure and short fragments of centromeric DNA.Surprisingly, the microtubule attachments were stable, and the detached kinetochores continued their migration to the spindle pole. These data extend the observations of Kenna et al (1988) that excision of yeast centromeres does not perturb protein binding. Experimental detachment of the centromere region from mammalian chromosomesdoes not perturb their functional properties. The kinetochore region is certainly autonomous from the chromosomal arms and is likely to be redundant in its composition. The ability of single centromeric units to function in budding yeasts may reflect their continual attachment to the spindle apparatus. Systems in which kinetochores disassemble following mitosis require a mechanism to establish kinetochore-microtubule connections. The dynamic nature of the mitotic spindle ensures that an excess of microtubule ends are available for kinetochore capture. In fact, microtubule ends will be continually available until such time that every chromosome is tethered to the spindle (see Nicklas 1988). An effective kinetochore must be distinguished from the surrounding chromatin, however, and perhaps a critical size favors these chance encounters with microtubules. The large repeating motifs in S. pombe and mammalianspecies represent a region of differentiation along the chromosomal DNAthat may serve such a purpose. Wesuggest that the centromere DNA-protein-microtubule connections are fundamentally similar throughout phylogeny, and vast differences in kinetochore size reflect elaborations and variation in modesof microtubule capture. ACKNOWLEDGMENTS Wethank Elaine Yeh and JoAnn Brock for critical review of the manuscript. Workfrom our laboratory was supported by National Institutes of Health grant RO1GM32238.K.B. was supported in part by a Research Career Development Award from the National Cancer Institute of the National Institutes of Health. I.S. was supported by a postdoctoral fellowship from the DamonRunyon-Walter Winchell Cancer Foundation. Annual Reviews www.annualreviews.org/aronline 332 SCHULMAN & BLOOM Annu. Rev. Cell. Biol. 1991.7:311-336. Downloaded from arjournals.annualreviews.org by University of North Carolina - Chapel Hill on 11/14/07. For personal use only. Literature Cited Alfageme, C. R., Rudkin, G. T., Cohen, L. H. 1980. Isolation, properties and cellular distribution of DI, a chromosomal protein of Drosophila. Chromosoma78:1-31 Am, P. H., Ketabgian, A. A., Smith, S., Schwartz, D. C., Jobs, E. W. 1989. The macromolecular organization of human centromeric regions. In Mechanisms of chromosomedistribution and aneuploidy, ed. M. A. Resnick, I:l. K. Vig, pp. 9-18. NewYork: Liss. 400 pp. Ashely, C. T., Pendleton, C. G., Jennings, W. W., Saxena, A., Glover, C. V. C. 1989. Isolation and sequencing of cDNAclones encoding Drosophila chromosomal protein DI. J. Biol. Chem.264: 8394-840l Baker, R. E., Fitzgerald-Hayes, M., O’Brien, T. C. 1989. Purification of the yeast centromere-binding protein CP1 and a mutational analysis of its binding site. J. BioL Chem. 264:10843-50 Baker, R. E., Masison, D. C. 1990. Isolation of the gene encoding the Saccharomyces cerevisiae centromere-binding protein CP1. MoLCell Biol. 10:1863-72 Bernat, R. L., Borisy, G. G., Rothfield, N. F., Earnshaw, W. C. 1990. Injection of anticentromere antibodies in interphase disrupts events required for chromosome movementsat mitosis. J. Cell Biol. 111: 1519-33 Bloom~K. S., Amaya,E., Carbon, J., Clarke, L., Hill, A., Yeh, E. 1984a. Chromatin conformation of yeast centromeres. J. Cell Biol. 99:1559~8 Bloom, K., Amaya, E., Yeh, E. 1984b. Centromeric DNAstructure in yeast chromatin. In Molecular BiolotTy of the Cytoskeleton, ed. G. G. Borisy, D. W. Cleveland, D. B. Murphy, pp. 175-84. NewYork: Cold Spring Harbor. 512 pp. Bloom, K. S., Carbon, J. 1982. Yeast centromere DNAis in a unique and highly ordered structure in chromosomes and small circular minichromosomes.Cell 29: 305-17 Bram,R. J., Kornberg, R. D. 1987. Isolation of a Saccharomyces cerevisiae centromere DNA-binding protein, its human homolog, and its possible role as a transcription factor. MoLCell Biol. 7:403-9 Brinkley, B. R. 1990. Centromeres and kinetochores: integrated domains of eukaryotic chromosomcs.Curr. Op. Cell Biol. 2: 446-52 Brinkley, B. R., Valdiva, M. M., Tousson, A., Balczon, R. D. 1989. The kinetochore: structure and molecular organization. In Mitosis Molecules and Mechanisms, ed. J. S. Hyams,B. R. Brinkley, pp. 77-118. San Diego: Academic. 350 pp. Brinkley, B. R., Zinkowski, W. L., Mollon, F. M., Davis, F. M., Pisegna, M., et al. 1988. Movement and segregation of kinetochores experimentally detached from mammalian chromosomes. Nature 336:251-54 Burke, D. T., Carle, G. F., Olson, M. V. 1987. Cloning of large segmentsof exogenous DNAinto yeast by means of artificial chromosomes. Science 236:806-12 Cai, M., Davis, R. W. 1989. Purification of a yeast centromere-binding protein that is able to distinguish single-base pair mutationsin its recognition site. Mol. Cell. Biol. 9:2544-50 Cai, M., Davis, R. W. 1990. Yeast centromere binding protein CBF1, of the helix-loop-helix family, is required for chromosome stability and methionine prototrophy. Cell 61:437-46 Cherry, L. M., Wang, R.-Y. 1988. Kinetochore staining in multicentric chromosomesof murine cancer lines. Cytobios 53: 173-83 Chikashige, Y., Kinoshita, N., Nakaseko, Y., Matsumoto, T., Murakami, S., et al. 1989. Composite motifs and repeat symmetry in S. pombe centromeres: direct analysis by integration of Notl restriction sites. Cell 57: 739-5l Clarke, L. 1990. Centromeres of budding and fission yeasts. Trends Genet. 6:150-54 Clarke, L., Baum, M. P. 1990. Functional analysis ofa centromerefrom fission yeast: a role for centromere-specific repeated DNAsequences. Mol. Cell Biol. I0:186372 Cooke, C. A., Heck, M. M. S., Earnshaw, W. C. 1987. The INCENPantigens: movement from the inner centromere to the midbodyduring tnitosis. J. Cell Biol. 105: 2053q57 Compton,13. A., Yen, T. J., Cleveland, D. W. 1991. Identification of novel centromere/kinetochore associated proteins using monoclonal antibodies generated against human mitotic chromosome scaffolds. J. Cell Biol. 122:1083-98 Dean, A., Pederson, D. S., Simpson, R. T. 1989. Isolation of yeast plasmid chromatin. In Methods in Enzymology, ed. P. M. Wasserman, R. D. Kornberg, pp. 26-41. " NewYork: Academic. 683 pp. Disney, J. E., Johnson, K. R., Magnuson, N. S., Sylvester, S. R., Reeves, R. 1989. High-mobility group protein HMG-I localizes to G/Q- and C-bands of human and mouse chromosomes. J. Cell Biol. 109:1975-82 Earnshaw, W. C., Halligan, N., Cooke, C. A., Rothfield, N. 1984. The kinetochore is Annual Reviews www.annualreviews.org/aronline Annu. Rev. Cell. Biol. 1991.7:311-336. Downloaded from arjournals.annualreviews.org by University of North Carolina - Chapel Hill on 11/14/07. For personal use only. STRUCTURE AND FUNCTION OF CENTROMERES 333 part of the metaphase chromosome Construction of functional artificial scaffold. J. Cell Biol. 98:352-57 minichromosomes in the fission yeast Earnshaw, W. C., Migeon, B. R. 1985. Three Schizosaccharomyces pombe. Proc. Natl. related centromere proteins are absent Acad. Sci. USA 86:577-81 Hahnenberger, K. M., Carbon, J., Clarke, from the inactive centromere of a stable isodicentric chromosome. Chromosoma L. 1991. Identification of DNAregions 92:290-96 required for mitotic and meiotic functions Earnshaw, W. C., Ratrie, H., Stetten, G. within the centromere of Schizosac1989. Visualization ofcentromere proteins charomyces pombe chromosome I. Mol. CENP-Band CENP-Con a stable dicenCell. Biol. In press tric chromosomein cytological spreads. Han, M., Chang, M., Kim, U., Grunstein, Chromosoma 98:1-12 M. 1987. Histone H2Brepression causes Earnshaw, W. C., Rattner, J. B. 1989. A map cell-cycle-specific arrest in yeast: Effects of the centromere (primary constriction) on chromosomalsegregation, replication, in vertebrate chromosomesat metaphase. and transcription. Cell 48:589 See Arn et al 1989, pp. 33~42 Heus, J. J., Zonneveld, B. J. M., Steensma, Earnshaw, W. C., Sullivan, K. F., Machlin, H. Y., Van den Berg, J. A. 1990. Centromeric DNAof Kluyceromyces lactis. P. S., Cooke,C. A., Kaiser, D. A., et al. 1987. Molecular cloning of cDNAfrom Curr. Genet. 18:517-22 CENP-B, the major human centromere Hieter, P., Pridmore, D., Hegemann,J. H., autoantigen. J. Cell Biol. 104:817-29 Thomas, M., Davis, R. W., Philippsen, P. Eckdahl, T. T., Anderson, J. A. 1987. Com1985. Functional selection and analysis of puter modelling of DNAstructures yeast centromeric DNA.Cell 42:913-21 involved in chromosome maintenance. Hill, A. 1988. Genetic manipulation of cenNucleic Acids Res. 15:853145 tromere function in Saccharomycescerevisiae. PhDthesis. Univ. N. Carolina. 133 Fishel, B., Amstutz, H., Baum,M., Carbon, J., Clarke, L. 1988. Structural organppization and functional analysis of cenHill, A., Bloom, K. 1987. Genetic maniputromeric DNAin the fission yeast Schizolation of centromere function. Mol. Cell. saccharomycespombe. Mol. Cell. Biol. 8: Biol. 7:2397-2405 754~63 Hill, A., Bloom,K. S. 1989. The acquisition Fitzgerald-Hayes, M., Clarke, L., Carbon, and processing of dicentric chromosomes J. 1982. Nucleotide sequence comparisons in yeast. Mol. Cell. Biol. 9:1368-70 and functional analysis of centromeric Huberman, J. A., Pridmore, R. D., Jager, DNAof yeast centromere DNAs. Cell 29: D., Zonneveld, B., Philippsen, P. 1986. 235-44 Centromeric DNAfrom Saccharomyces Funk, M., Hegemann,J. H., Philippsen, P. uvarum is functional in Saccharomyces 1989. Chromatindigestion with restriction cerevisiae. Chromosoma94:162~58 endonucleases reveals 150-160 bp of proJames, T. C., Eissenberg, J. C., Craig, C., tected DNAin the centromere region of Dietrich, V., Hobson,A., Elgin, S. C. R. chromosome XIV in Saccharomyces cere1989. Distribution patterns of HP1, a visiae. Monogr.Gen. Genet. 219:153 60 heterochromatin-associated nonhistone Futcher, B., Carbon, J. 1986. Toxic effects chromosomal of Drosophila. Eur. J. Cell Biol. protein 50:170-80 of excess cloned centromeres. Mol. Cell. Biol. 6:2213-22 James, T. C., Elgin, S. C. R. 1986. IdentiGaudet, A. M., Fitzgerald-Hayes, M. 1990. fication of a nonhistone chromosomal The function of centromeres in chroprotein associated with heterochromatin mosomesegregation. In The Eukaryotic in Drosophila melanogaster and its gene. Nucleus, ed. P. R. Strauss, S. H. Wilson, Mol. Cell. Biol. 6:3862-72 2: 845-81. NewJersey: Telford. 881 pp. Kayne, P. S., Kim, U.-J., Han, M., Mullen, Gerring, S. L., Spenser, F., Hieter, P. 1991. J. R., Yoshizaki, F., Grunstein, M. 1988. The CHL1 (CTF1) gene product of SacExtremely conserved histone H4 terminus charomyces cerevisiae is important for is dispensable for growth but essential for chromosometransmission and normal cell repressing the silent mating type loci in cycle progression in G2/M. EMBOJ. 9: yeast. Cell 55:27-39 4347-58 Johns, E. W. 1982. History, definitions and Greig, G. M., England, S. B., Bedford, H. problems. In The HMGChromosomal M., Willard, H. F. 1989. ChromosomeProteins, ed. E. W. Johns, pp. 1 8. Lonspecific alpha satellite DNAfrom the cendon: Academic. 251 pp. tromere of chromosome 16. Am. J. Hum. Kenna, M., Amaya, E., Bloom, K. S. 1988. Genel. 45:862-72 Selective excision of the centromere Hahnenberger, K. M., Baum,M. P., Polizzi, chromatin complex from Saccharomyces C. M., Carbon, J., Clarke, L. 1989. cerevisiae. J. Cell Biol. 107:9-15 Annual Reviews www.annualreviews.org/aronline Annu. Rev. Cell. Biol. 1991.7:311-336. Downloaded from arjournals.annualreviews.org by University of North Carolina - Chapel Hill on 11/14/07. For personal use only. 334 SCHULMAN& BLOOM Koshland,D. L., Rutledge,L., Fitzgeraldgenes that affect mitotic chromosome Hayes, M., Hartwell, L. 1987. A genetic transmissionin S. cerevisiae. Cell 44: 53analysis of dicentric minichromosomes in 63 Saccharomyces cerevisiae. Cell 48:801-12 Mellor,J., Jiang, W.,Rathjen,J., Barnes,C. Lechner, J., Carbon, J. 1991. A 240 kD A., Hinz, T., et al. 1990. CPF1,a yeast multisubunit protein complex(CBF3)is protein whichfunctions in centromeres major componentof the budding yeast and promoters. EMBO J. 9:4017-26 centromere.Cell 64:717-25 Merry, D. E., Pathak, S., Hsu, T. C., Levinger,L., Varshavsky,A. 1982. Protein Brinkley, B. R. 1985. Anti-kinetochore D1 preferentially binds to A+T-rich antibodies: use as probesfor inactive cenDNA in vitro and is a componentof Drotromeres. Am. J. Hum.Genet. 37:425-30 sophila melanogaster nucleosomescon- Mitchison,T. J., Evans,L., Schulze,E., Kirtaining A+T-richsatellite DNA.Proc. schner, M.W.1986. Sites of microtubule Natl. Acad. Sci. USA79:7152-56 assemblyand disassemblyin the mitotic Lica,L. M., Narayanswami,S., Hamakalo, spindle. Cel145:515-21 B. A. 1986. Mousesatellite DNA,cenMitchison, T. J., Kirschner, M. W.1985. tromere structure, and sister chromatid Properties of the kinetochorein vitro. I. pairing. J. Cell BioL103:1145-51 Microtubulenucleation and tubulin bindLinxweiler, W., Horz, W. 1985. Reconing. J. Cell Biol. 101:755-65 stitution experimentsshowthat sequence- Moroi,Y., Peebles,C., Fritzler, M.J., Steispecific histone-DNA interactions are the gerwald, J., Tan, E. M.1980. Autoantibody to centromere (kinetochore) basis for nucleosomephasing on mouse satellite DNA.Cell 42:281-90 scherodermasera. Proc. Natl. Acad.Sci. Lund, T., Holtlund, J., Fredrickson, M., USA 77:1627-31 Laland, S. G. 1983. On the presence of Nakaseko,Y., Adachi, Y., Funahashi, S., two newmobility group-like proteins in Niwa, O., Yanagida, M. 1986. ChroHeLa$3 cells. FEBSLett. 152:163-67 mosomewalking shows a highly homMahtani, M. M., Willard, H. F. 1990. ologousrepetitive sequencepresent in all Pulsed-field gel analysis of or-satellite the centromereregions of fission yeast. DNAat the humanX chromosomecentroEMBOJ.5:i011 21 mere: high frequency polymorphismsand Nicklas, R. B. 1988. Chanceencountersand array size estimates. Genomics7:607-13 precisionin mitosis. J. Cell. Sci. 89: 283Mann,C., Davis, R. W.1983. Instability of 85 dicentric plasmidsin yeast. Proc. Natl. Nicklas, R. B. 1989. Themotorfor poleward chromosomemovementin anaphase is in Acad. Sci. USA80:228-32 Masumoto,H., Masukata, H., Muro, Y., or near the kinetochore.J. Cell Biol. 109: Nozaki, N., Okazaki, T. 1989. A human 2245-55 centromere antigen (CENP-B)interacts Niwa, O., Matsumoto,T., Chikashige, Y., with a short specific sequencein alphoid Yanagida, M. 1989. Characterization of Schizosaccharomyces pombeminichroDNA,a humancentromeric satellite. J. Cell Biol. 109:1963-73 mosome deletion derivatives and a functional allocation of their centromere. Matsumoto,T., Murakami,S., Niwa, O., Yanagida, M. 1990. Construction and EMBOJ. 8:3045-52 characterization of centric circular and Oertel, W., Mayer,M.1984. Structure and mitotic stability of minichromosomes acentric linear chromosomes in fission yeast. Curr. Genet.18:323-30 originating in yeast cells transformedwith McCarroll, R. M., Fangman,W. L. 1988. tandem dimers of CENll plasmids. Mol. Timeof replication of yeast centromeres Gen. Genet. 198:300-7 and telomeres. Cell 54:505-13 Palmer, D. K., O’Day,K., Margolis, R. L. 1989. Biochemicalanalysis of CENP-A, a McClintock,B. 1939. The behavior of successive nuclear divisions of a chromosome centromeric protein with histone-like properties. SeeAmet al 1989,pp. 61-72 brokenat meiosis. Proc.Natl. Acad.Sci. Palmer, D. K., O’Day,K., Trong, H. L., USA 25:405 16 McGrew, J. T., Xiao, Z., Fitzgerald-Hayes, Charbonneau,H., Margolis, R. L. 1990. Partial sequence analysis of the cenM. 1989. Saccharomyces cerevisiae tromere specific protein CENP-A reveals mutants defective in chromosomesegextensive homology to histone H3. J. Cell regation. Yeast 5:271-84 Meeks-Wagner, D., Hartwell, L. H. 1986. Biol. 111(5):281(Abstr.) Normalstoichiometry of histone dimer Peretti, D., Maraschio,P., Lambiase,S., sets is necessaryfor highfidelity of mitotic Curto, F. L., Zuffardi, O. 1986.Indirect immunofluorescence of inactive cenchromosome transmission. Cell 44:43-52 tromeresas indicator of centromerefuncMccks-Wagner, D., Wood,J. S., Garvik, B., tion. Hum.Genet. 73:12-16 Hartwell, L. H. 1986. Isolation of two Annual Reviews www.annualreviews.org/aronline Annu. Rev. Cell. Biol. 1991.7:311-336. Downloaded from arjournals.annualreviews.org by University of North Carolina - Chapel Hill on 11/14/07. For personal use only. STRUCTUREAND FUNCTION OF CENTROMERES 335 putative protein kinase. GenesDev.5: 549Pfarr, C. M., Coue, M., Grissom, P. M., Hays,T., Porter, M. E., Mclntosh,J. R. 60 1990. Cytoplasmicdyneinis localized to Simerly, C., Balczon,R., Brinkley, B. R., Schatten, G. 1990. Microinjectedkinetokinetochoresduring mitosis. Nature345: chore antibodies interfere with chro263-65 Polizzi, C., Clarke, L. 1991. Thechromatin mosome movementin meiotic and mitotic structure of centromeres from fission mouseoocytes. J. Cell Biol. 111: 1491yeast: differentiation of the central core 1504 that correlateswithfunction.J. Cell Biol. Solomon,M.J., Strauss, F., Varshavsky,A. 112:191-201 1986. A mammalianhigh mobility group protein recognizesanystretch of six A-T Radic, M. Z., Lundgren,K., Hamakalo,B. A. 1987. Curvature of mousesatellite base pairs of duplex DNA.Proc. Natl. DNAand condensation of heteroAcad. Sci. USA83:1276-80 chromatin. Cell 50:1101-8 Spencer,F., Connelly,C., Lee, S., Hieter, P. 1988. Isolation and cloning of conRattner, J. B. 1991. Thestructure of the mammalian centromere. Bioessays 13:51ditionally lethal chromosometrans56 mission fidelity genes in Saccharomyces Rattner, J. B., Kingwell,B. G., Fritzler, M. cerevisiae. In CancerCells Eukaryotic J. 1988.Detectionof distinct structural DNAReplication, ed. T. Kelly, B. Stilldomainswithin the primary constriction man, 6: 441-52. NewYork: Cold Spring using autoantibodies. Chromosoma 96: Harbor Lab. 360-67 Spencer, F., Gerring, S. L., Connelly, C,, Reeves, R., Nissen, M. S. 1990. The A-THieter, P. 1990. Mitotic chromosome DNA-binding domain of mammalian transmission fidelity mutants in Sachigh mobility group I chromosomal procharomyees cerevisiae. Genetics 124:237 teins. J. Biol. Chem.265:8573-82 49 Rieder, C. L., Alexander,A. P. 1990. Kine- Steuer, E. R., Wordeman, L., Schroer, T. tochores are transported polewardalong A., Sheetz, M. P. 1990. Localization of a single astral microtubuleduring chrocytoplasmicdyneinto mitotic spindles and mosome attachmentto the spindle in newt kinetochores. Nature345:266-68 lung cells. J. Cell Biol. 110:8145 Tschumper, G., Carbon, J. 1987. SaeResnick, M. A, Westmoreland,J., Bloom, charomycescerevisiae mutants that tolK. 1990. Heterogeneity and maintenance erate centromere plasmids at high copy of centromere plasmid copy number in number.Proc. Natl. Acad. Sci. USA84: Saccharomyeescerevisiae. Chromosoma 7203-7 99:281-88 Sullivan, K. F., Glass, C. A. 1991. CENP-B Runge,K. W., Zakian, V. A. 1989. Introis a highly conserved mammaliancenductionof extra telomericsequencesinto tromere protein with homologyto the Saccharomyces eerevisiae results in telohelix-loop-helixfamilyof proteins. Chromereelongation. Mol.Cell. Biol. 9: 1488mosoma.In press 97 van Holde, K. E. 1989. Chromatin. New Russo, P., Sherman,F. 1988. Transcription York:Springer-Verlag.497 pp. terminates near the poly(A) site in the Vig, B. K., Paweletz,N. 1988. Sequenceof CYC1gene of the yeast Saecharomyces chromosomeseparation: generation of cerevisiae. Proc.Natl. Acad.Sci. USA86: unstable multicentric chromosomes in a 8348-52 rat cell line. Chromosoma 96:275-82 Salmon,E. D. 1989. Microtubule dynamics Wevrick, R., Willard, H. F. 1989. Longand chromosomemovement.See Brinkley rangeorganizationof tandemarrays of ~tet al 1989,pp. 119-82 satellite DNAat the centromeres of Saunders, M. 1989. The molecular archihuman chromosomes: High-frequency tecture of yeast centromeres.PhDthesis. array-length polymorphismand meiotic Univ. N. Carolina. 144 pp. stability. Proc. Natl. Acad.Sci. USA86: Saunders, M. A., Fitzgerald-Hayes, M., 9394-98 Bloom,K. S. 1988. Chromatinstructure Willard, H. F. 1990. Centromeres of of altered yeast centromeres.Proc. Natl. mammaliancentromeres. Trends Genet. Acad. Sei. USA85:175-79 6:410-16 Saunders, M. A., Yeh, E., Grunstein, M., Willard,H. F., Waye,J. S. 1987. Hieratchal Bloom,K. S. 1990. Nucleosome depletion order in chromosome-specific human alters the chromatin structure of Sacalpha satellite DNA. TrendsGenet.3: 192charomycescerevisiae centromeres.Mol. 98 Cell. Biol. 10:5721-27 Willard, H. F., Wervick,R., Warburton,P. Shero, J. H., Hieter, P. 1991. A suppressor E. 1989. Humancentromere structure: ofa centromere DNAmutation encodes a organization and potential role of alpha Annual Reviews www.annualreviews.org/aronline 336 SCHULMAN & BLOOM Annu. Rev. Cell. Biol. 1991.7:311-336. Downloaded from arjournals.annualreviews.org by University of North Carolina - Chapel Hill on 11/14/07. For personal use only. satellite DNA.In Mechanisms of Chroreleased from chro~nosomes at the onset mosomeDistribution and Aneuploidy, ed. ofanaphase. EMBOJ. 10:1245-54 M. A. Resnick, B. K. Vig, pp. 9-18. New Zhang, X.-Y., Fittler, F., Horz, W. 1983. York: Liss. 400 pp. Eight different highly specific nucleosome Yeh, E. 1985. PhD thesis. Studies of the phases and e-satellite DNA in the African RNAtranscripts within the centromere green monkey. Nucleic Acids Res. I1: region of yeast chromosomeIII and XI. 4287-4306 Univ. Calif., Santa Barbara. 160 pp. Zinkowski, R. P., Vig, B. K., Broccoli, D. Yen, T. J., Compton, D. A., Earnshaw, W. 1986. Characterization of kinetochores in multicentric chromosomes. Chromosoma C., Cleveland, D. W. 1991. CENP-E, a human centromere associated protein 94:243-48 Annu. Rev. Cell. Biol. 1991.7:311-336. Downloaded from arjournals.annualreviews.org by University of North Carolina - Chapel Hill on 11/14/07. For personal use only. Annu. Rev. Cell. Biol. 1991.7:311-336. Downloaded from arjournals.annualreviews.org by University of North Carolina - Chapel Hill on 11/14/07. For personal use only.