ddg330.pdf - Oxford Academic

Analysis of mammalian proteins involved inchromatin modification reveals new metaphasecentromeric proteins and distinct chromosomaldistribution patterns

Jeffrey M. Craig, Elizabeth Earle, Paul Canham, Lee H. Wong,

Melissa Anderson and K.H. Andy Choo*

Murdoch Childrens Research Institute, Royal Children’s Hospital, Flemington Road, Melbourne,

Victoria 3052, Australia

Received June 26, 2003; Revised September 15, 2003; Accepted September 23, 2003

We have examined the metaphase chromosomal localization of 15 proteins that have previously beendescribed as involved in mammalian chromatin modification and/or transcriptional modulation.Immunofluorescence data indicate that all the proteins localize to human and mouse centromeres, aneocentromere, and the active centromere of a dicentric chromosome, with six of these proteins (Sin3A,PCAF, MYST, MBD2, ORC2, P300/CBP) being demonstrated at mammalian centromeres for the first time.Most of these proteins fall into two distinct chromosomal distribution patterns: (a) kinetochore-associatedproteins (Sin3A, PCAF, MYST and BAF180), which colocalize with metaphase kinetochores, but not any of thepericentric and other major heterochromatic regions; and (b) heterochromatin-associated proteins (MeCP2,MBD1, MBD2, ATRX, HP1a, HDAC1, HDAC2, DNMT1 and DNMT3b), which colocalize with centromeric/pericentric heterochromatin and all other major heterochromatic sites. A heterogeneous third group (c)consists of the origin recognition complex subunit ORC2 and the histone acetyltransferase P300/CBP, whichassociate generally with kinetochores in humans and centromeric/pericentric heterochromatin in mouse,with some minor differences in localization. These observations indicate an extensive sharing of proteincomponents involved in chromatin modification at gene loci, centromeres and various chromosomalheterochromatic landmarks. The definition of distinct patterns of chromosomal distribution for theseproteins provides a useful basis for the further investigation of the broad-ranging roles of these proteins.

INTRODUCTION

Centromeres are responsible for the attachment of chromo-somes to the mitotic and meiotic spindle, and are essential forchromosome segregation. They perform most of these func-tions via the kinetochore, which contains a variety of proteinswith roles ranging from establishing a centromeric nucleosomestructure to checkpoint control (1). Neocentromeres are fullyfunctional centromeres that occur at ectopic genomic locations(2). Human neocentromeres contain no detectable a-satelliteDNA that resides at typical human centromeres, and have beenshown to bind all essential centromere proteins (3). Of note,despite the lack of long tracts of tandemly repeated DNA,human neocentromeres are heterochromatic in nature, as shownby their association with known heterochromatin proteins (3,4).

These properties make neocentromeres a useful system tofacilitate the study of centromere structure and function (5,6).

Most eukaryotic centromeres, including those of fissionyeast, Drosophila and mammals, contain regions of hetero-chromatin; that is, chromatin packaged in a condensed statethroughout the cell cycle (7). On a cytological scale, thisheterochromatin may appear to overlap with the kinetochore(8–10), and in some organisms flank the kinetochore along thechromosome axis (11,12). Pericentric heterochromatin containsfew transcribed genes and can form domains that are involvedin gene silencing in trans within the interphase nucleus (13,14).

The traditional view of heterochromatin has been developedprincipally around Drosophila and mammalian pericentricheterochromatin that usually contain large tracts of tandemrepeats (7). Human centromeres also contain pericentric

*To whom correspondence should be addressed. Tel: þ61 383416306; Fax: þ61 393481391; Email: [email protected]

Human Molecular Genetics, 2003, Vol. 12, No. 23 3109–3121DOI: 10.1093/hmg/ddg330

Human Molecular Genetics, Vol. 12, No. 23 # Oxford University Press 2003; all rights reserved

Dow

nloaded from https://academ

ic.oup.com/hm

g/article/12/23/3109/692455 by guest on 17 August 2022

heterochromatin, with several chromosomes exhibiting largerheterochromatic bands recognizable at the cytological level,at 1q12, 9q12, 16q11.2, Yq12 and the p-arms of all theacrocentric chromosomes. More recently, heterochromatinproteins have been found at silenced gene promoters (15) andhuman neocentromeres (3,4), indicating the need to broadenthe definition of the heterochromatic status of a genomic regionto reflect the presence of heterochromatin proteins (16).

Investigation of the protein components of the silent matingtype loci and centromeres of fission yeast, and gene promotersand pericentric heterochromatin of higher eukaryotes, have ledto the ‘histone code hypothesis’, whereby dynamic changes inthe covalent modification of histones, such as acetylation andmethylation, provide a code for the correct regulation of open(transcriptionally competent) or closed (transcriptionally silent)chromatin, mostly by affecting chromatin structure and intera-ctions of non-histone proteins with chromatin (17). The twomain types of epigenetic modifier proteins are those involved inthe covalent modification of histones, such as histone deace-tylases (HDACs) and histone acetyltransferases (HATs), andATP-dependent chromatin remodelling factors that enablehistone-DNA shuffling (7,18).

HDACs are central to establishing and maintaining silentchromatin. CpG-methylation ( meCpG) is another integral compo-nent of the silent chromatin. DNA methyltransferases (DNMTs),which maintain meCpG, and meCpG-binding proteins (MeCPs),both act as corepressors by interacting with HDACs (18,19). DNA-binding transcription factors such as those of the Ikaros family,can recruit HDACs directly to promoters or via corepressors suchas CtBP or Sin3A (18,20). These factors form macromolecularcomplexes that cooperate with ATP-dependent chromatin remo-delling factors to modify chromatin (18).

Some of the proteins involved with chromatin modification atgene promoters have been localized cytologically to centro-meric regions (reviewed in 21). However, most studies haveconcentrated on interphase chromosomes, and many of theseproteins remain untested for their interaction with the centro-mere. In this study, we have used antibodies to 15 differentproteins, most of which are known to play major roles in theepigenetic control of gene expression, to determine theirrelationship with human and mouse kinetochores and peri-centric heterochromatin.

RESULTS

Antibodies to 15 different proteins were used in immunofluo-rescence study. These proteins included the histone deacetylases(HDAC1 and HDAC2), DNA methyltransferases (DNMT1 andDNMT3b), MeCPs (MeCP2, MBD1 and MBD2), the transcrip-tional corepressor Sin3A, chromatin remodellers (ATRX andBAF180), origin recognition complex subunit ORC2, chromo-domain protein HP1a, and histone acetyltransferases (HATs—PCAF, MYST and P300/CBP). All of these proteins, exceptmethyl binding domain protein 1 (MBD1) and the chromatinremodellers, have been shown to be subunits of, or interactdirectly with, macromolecular chromatin protein complexes, andall proteins except the chromatin remodellers have previouslybeen found at gene promoters (18,22,23).

It is noteworthy that, with a few exceptions, humanchromosomes in general exhibit less heterochromatin thantypical mouse chromosomes (24,25). The exceptions are thelarge bands of heterochromatin seen in the pericentric q-armregions of human chromosomes 1, 9 and 16, the q12 regionof Y, and the pericentric p-arm regions of the acrocentricchromosomes. For the purpose of illustration for our immuno-fluorescence results, we have used randomly selected mousechromosomes, but for humans, we chose to use chromosome 1(and acrocentric chromosomes in some cases) for consistencyof displaying the protein distribution patterns for boththe kinetochore and pericentric heterochromatin on the samechromosome. In addition, we also showed partial metaphasespreads containing examples of at least 10 random human ormouse chromosomes stained with anti-centromere antiserum(ACA) CREST3 or anti-CENP-A antibody. The results aresummarized in Table 1 and described below.

A number of different mouse and human cell lines were usedfor immunofluorescence analysis (see Materials and Methods).These included a human cell line containing a dicentricchromosome with two well-separated repetitive centromericheterochromatic regions, only one of which forms an active,kinetochore-bearing centromere (26). We also examined thelocalization of these antibodies on our well-characterized mardel(10) neocentromere (5,27–29). Antibodies were co-detectedwith antibodies against previously cytologically mappedcentromere proteins: inner-kinetochore protein CENP-A (30)or ACA CREST3 (see Materials and Methods). In addition,we co-detected proteins with antibodies to outer-kinetochoremitotic checkpoint protein BubR1 (31), and heterochromatinproteins HP1a (32) or MBD1 (33). Analysis was initially basedon cytological observations, which enabled us to categorizeproteins as kinetochore- or heterochromatin-specific. We furthercharacterized protein-distribution patterns in two ways. The firstwas to apply fluorescence intensity line scans (graphs of pixelintensity versus distance) laterally through sister kinetochoresand longitudinally through the centromere along the chromo-some axis (see below). In the second method, applicable only tokinetochore-associated proteins where distinct doublet kineto-chore-signals were discernible [see Table 1 subgroups (a) and(c); and described below], we directly measured the inter-kinetochore distance for each signal-doublet of at least40 chromosomes, and compared this distance with that of theco-detected CENP-A/CREST3 or the outer-kinetochore proteinBubR1 (see Materials and Methods). This allowed us todetermine whether a protein was localised significantly furtherapart than CENP-A (i.e. in an outer-kinetochore location), andto confirm such a conclusion using BubR1 colocalizationmeasurements.

Kinetochore-associated proteins

Antibodies to Sin3A, PCAF, MYST and BAF180 all showedpreferential staining at kinetochores on mouse and humanmetaphase spreads, but not to any significant extent at thepericentric or other heterochromatin (Table 1). Figures 1 and 2show typical examples of immunofluorescence staining fromtwo members of this group, PCAF and Sin3A, respectively.

Antibodies to the kinetochore-associated proteins all loca-lized at the lateral outside edges of the CREST3 signals on

3110 Human Molecular Genetics, 2003, Vol. 12, No. 23

Dow


ic.oup.com/hm


human centromeres (Figs 1A and I, 2A and I, and 3).Localization to the outer kinetochores was confirmed bycolocalization with a known outer-kinetochore mitotic check-point protein BubR1 (31) (Figs 1B, 2B and 3). On mousechromosomes, colocalization of these proteins to the outerkinetochore regions is more difficult to discern, due presum-ably to the tighter centromeric chromatin structures of thesechromosomes (Figs 1 and 2C, D, and J). Using antibodies toheterochromatin proteins HP1a or MBD1, Sin3A, PCAF,MYST and BAF180 all consistently occupied a significantlysmaller region especially axially, than HP1a or MBD1, at themouse and human pericentric heterochromatin (Figs 1 and 2Eand F).

Immunofluorescence analysis of the mardel(10) chromosomeshowed positive signals at the neocentromere for each of theantibodies to this subgroup of proteins (Figs 1G and 2G;Table 1). Testing of all members of this group on the humandicentric chromosome further demonstrated binding only to theactive centromere (Figs 1H and 2H; Table 1), except for MYST,where the determination of whether this protein was present orabsent at the inactive centromere was prevented by relativelyhigh background signals on the chromosome arms.

Heterochromatin-associated proteins

Antibodies to this subgroup of proteins, consisting of MeCP2,MBD1, MBD2, ATRX, HP1a, HDAC1, HDAC2, DNMT1 andDNMT3b, all exhibited a similar staining pattern in thedifferent cell types investigated, as exemplified by DNMT1 andMBD2 in Figures 4 and 5, respectively. The results indicated

that, unlike the discreet doublet signals of a kinetochore-associated protein, immunofluorescence staining for theseproteins generally mirrored that of other known heterochroma-tin proteins, such as HP1a and MBD1, showing strong signalsat both the centromeres and the large bands of pericentricheterochromatin in human and mouse chromosomes, includingthe short-arm heterochromatin of the human acrocentricchromosomes (Figs 4 and 5A–E), as described for MBD1,HP1a and HP1b and ATRX (9,10,33,34). In addition, MBD2,HDAC1, HDAC2, DNMT1 and DNMT3b showed significantstaining over whole chromosome arms, as was described forMBD1 (33).

When the binding properties of this subgroup of proteins wereexamined on the mardel(10) chromosome, an elevated doubletsignal was observed at the neocentromere. The intensity of theneocentromere signal was consistently lower than at all othercentromeres, but higher than that observed at the 10q25 regionon the normal human chromosome 10 (Figs 4F and 5F; Table 1,and data not shown). We further determined the binding patternof these proteins on the human dicentric chromosome, andshowed a strong localization at both the active and inactivecentromeres (Figs 4G and 5G); exceptions were the antibodiesto HDAC1 and HDAC2, where high chromosome-arm back-ground signals have prevented a clear visualization of theseproteins at the inactive centromere (Table 1).

Kinetochore- and heterochromatin-associated proteins

In human cells, ORC2 and P300/CBP both localized to thekinetochores, but not to the pericentric heterochromatin

Table 1. Localization of proteins on normal metaphase, dicentric and neocentric chromosomes

Protein Functional class Metaphase localization Dicentric Neocen

Active Inactive

Subgroup (a): kinetochore-associated proteinsSin3Aa–c Transcriptional corepressor Outer kine, no peric or other major het þ � þPCAFa,b HAT Outer kine, no peric or other major het þ � þMYSTa,b HAT Outer kine, no peric or other major het nd nd þBAF180 Chromatin remodeller Outer kine, no peric or other major het þ � þ

Subgroup (b): heterochromatin-associated proteinsMeCP2 MeCP Centric/Peric het, all major het þ þ þMBD1 MeCP Centric/Peric het, all major het þ þ þMBD2a,b MeCP Centric/Peric het, all major het þ þ þATRX Chromatin remodeller Centric/Peric het, all major het þ þ þHP1a Chromodomain Centric/Peric het, all major het þ þ þHDAC1b HDAC Centric/Peric het, all major het nd nd þHDAC2b HDAC Centric/Peric het, all major het nd nd þDNMT1b DNMT Centric/Peric het, all major het þ þ þDNMT3bb DNMT Centric/Peric het, all major het þ þ þ

Subgroup (c): kinetochore- or heterochromatin-associated proteinsORC2a,b Origin recognition complex Inner kine and some acrocentric short arms in human,

all centric/peric het in mouseþ þ þ

P300/CBPa,b HAT Outer kine in human, all centric/peric het in mouse þ � þ

Localizations to inner or outer kinetochore (kine) were established from cytological observations and confirmed by inter-kinetochore distance measurements (seeFig. 3). Peric het, pericentric heterochromatin; major het, major regions of cytologically defined heterochromatin; dicentric, human dicentric chromosome with anactive and an inactive centromere; neocen, neocentromere; nd, not determinable.aLocalization of these proteins to mammalian centromeres has not been previously reported.bLocalization of these proteins to mammalian metaphase centromeres has not been previously reported.cA homologue of this protein has been found at fission yeast centromeres (53).

Human Molecular Genetics, 2003, Vol. 12, No. 23 3111

Dow


ic.oup.com/hm


Figure 1. Immunofluorescence results showing distribution of PCAF on human or mouse chromosomes. For human chromosomes, examples of chromosome 1 (A,B, E) and a partial metaphase spread (I) are shown for the colocalization with known centromere proteins. For mouse, a randomly chosen chromosome (C, D, F)and a partial metaphase spread (J) are shown. Known centromere proteins were represented by ACA CREST3 (equivalent to inner-centromere protein CENP-A);the outer-kinetochore protein BubR1; heterochromatin protein MBD1 (or HP1a for some other proteins); and CREST6, an autoimmune serum containing anti-CENP-A and anti-CENP-B antibodies, and localizing to both the active (a) and inactive (i) centromeres of a dicentric chromosome. FISH indicates the signal froma mardel(10) neocentromere-specific BAC probe. Fluorescence intensity line-scans taken laterally (three pixels wide) and axially (13 pixels wide) through the kineto-chore (as shown by arrows in K and L, respectively), are also presented. For each chromosome, multi-colour and split-colour images are shown: blue (DAPI) for thechromosome; red (Texas Red) for PCAF; and green (FITC) for known centromere proteins or FISH signal. PCAF is shown to localize on the lateral outside edgesof CENP-A (A, C, I) and fully with BubR1 (B, D); to identify only a subset of pericentric heterochromatin (E, F); to localize to the mardel(10) neocentromere (G),and to the active centromere only of a human dicentric chromosome (H).


Dow


ic.oup.com/hm


(Figs 6A, F and J, and 7A, E and I). ORC2, however, wasfrequently observed at the satellites at the ends of the sort armof the acrocentric chromosomes (Fig. 6B). In addition, whenORC2 was analysed against CENP-A and the outer-kinetochore protein BubR1, the protein was seen to colocalizewith CENP-A at the inner kinetochores (Figs 6A and 3). Incontrast, a similar analysis for P300/CBP against CREST3 andBubR1 in human cells indicated that this protein colocalizedmore with BubR1 (Figs 7A, B and 3).

The immunofluorescence distribution patterns on mousechromosomes was considerably different to those seen on thehuman chromosomes, where ORC2 and P300/CBP bindingwas detected at both the centromere and throughout thepericentric heterochromatin, reflecting both centromere andheterochromatin association (Figs 6D, E, G and K and Figs 7C,D, F and J). Immunofluorescence analysis also showed thatboth the proteins were enhanced at the mardel(10) neocen-tromere (Figs 6H and 7G) and the active centromere of a human

Figure 2. Immunofluorescence results showing distribution of Sin3A on human or mouse chromosomes. Detailed explanations are as for Figure 1.


Dow


ic.oup.com/hm


dicentric chromosome (Fig. 6I and 7H). However, ORC2was also present at the inactive centromere of the dicentricchromosome (Fig. 6I).

DISCUSSION

Recent studies have identified an increasing number of proteinsresponsible for the epigenetic modification of chromatin, inparticular in establishing and maintaining active or silentchromatin found at gene promoters (reviewed in 7,17,18). Thepossible involvements of these proteins at the centromere andcytologically defined heterochromatin have been less wellstudied. In order to shed light on such possible involvements,we have investigated the detailed distribution of 15 of theseproteins on human and mouse metaphase chromosomes.These proteins include those involved in covalent modifi-cation of histones (HDACs and HATs), chromatin remodelling(BAF180 and ATRX), meCpG-maintenance or binding(DNMTs, MeCPs), transcriptional corepression (Sin3A), het-erochromatin formation (Hp1a) and origin binding (ORC2).Our results have identified all 15 proteins to be enriched at thehuman and mouse centromeres, with most of these proteinsfalling into two general chromosomal distribution subgroups(Table 1).

The first chromosomal subgroup, consisting of Sin3A, PCAF,MYST and BAF180, localize to active kinetochores only. Allmembers bind to outer kinetochores except BAF180, whichcolocalizes with CENP-A. Apart from BAF180, localizationof the other three members to mammalian centromeres

has not previously been described. The similar chromosomaldistribution pattern of the histone H3 methyltransferaseSUV39H1 (35) suggests that this protein also belongs to thissubgroup. These results therefore define a substantial subgroupof chromatin modifier proteins whose binding is linked tofunctional kinetochores.

The second subgroup of proteins, comprising MeCP2,MBD1, MBD2, ATRX, HP1a, HDAC1, HDAC2, DNMT1and DNMT3b localize, at the current level of resolution, tocentromeres as well as all major mouse and human regions ofpericentric and non-pericentric heterochromatin, and theinactive dicentric centromere. These results confirm previouslocalization of MeCP2, MBD1, ATRX and HP1a on thepericentric heterochromatin of mammalian metaphase chromo-somes (10,33,34,36), and extend the previously observedinterphase localization of HDAC1, HDAC2, DNMT1, andDNMT3b (14,37–39) to metaphase chromosomes. In addition,the results reveal for the first time the localization of MBD2 atmammalian pericentric and kinetochore regions.

The apparently indiscriminate binding of this subgroup ofproteins to all major sites of cytological heterochromatin is instark contrast to the behaviour of proteins of subgroup (a).These heterochromatic sites are known to contain large arraysof a variety of tandem repeats, such as centromeric mouseminor and pericentric major satellites, and human a- and b-satellites and classical satellites I–III (25). This apparentsequence-independent mode of chromosomal association maybe related to the unique tertiary structures presented bytandemly repeated DNA arrays, and/or chromatin-relatedmacromolecular complexing of these proteins at these sites.

Figure 3. Measurement of inter-signal distances at human sister kinetochores. Distances between centres of fluorescent signals for all proteins localizing to humankinetochores were measured for at least 40 kinetochores, averaged, and compared to co-detected inner kinetochore CENP-A/ACA CREST3 or outer kinetochoreBubR1. Distances were then then converted to distances relative to CENP-A or BubR1. Error bars of �1 SD and P-values for each antibody/CENP-A pair are alsoshown on each graph. Sin3A, PCAF, MYST, BAF180 and P300/CBP were all found to localize significantly outer to CENP-A/CREST3 and to colocalize withBubR1, whereas ORC2 co-localized with CENP-A/CREST3 and was significantly inner to BubR1.


Dow


ic.oup.com/hm


Localization of ORC2 and P300/CBP to mammaliancentromeres has not been previously described. These proteinsappear to exhibit some of the kinetochore-binding properties ofsubgroup (a), especially with human chromosomes, and someof the heterochromatin-binding properties of subgroup (b),especially mouse chromosomes.

What are the roles of these proteins at the centromere, and thesignificance of the different patterns of chromosomal distribu-tion? The detection of all proteins at the mardel(10) neocen-tromere clearly indicates that the presence of large arrays of

heterochromatic satellite DNA is not a prerequisite for thebinding of any of these proteins. Indeed, this finding establishesa direct functional link between the establishment of centromereactivity and the recruitment and assembly of these proteins.Since the roles of these proteins have previously been studiedmostly in relation to transcriptional regulation, their specificroles at the centromere are largely unclear at present. It has beensuggested that one of the members of this subgroup, SUV39H1,in addition to performing a role in setting up silent chromatin atcentromeres, could also be involved in centromere-directed

Figure 4. Immunofluorescence results showing distribution of DNMT1 on human and mouse chromosomes. For human chromosomes, examples of chromosome1, an acrocentric chromosome, and a partial metaphase spread, are shown for colocalization with known centromere proteins (A, C, D, H); for mouse, a randomlychosen chromosome or a partial metaphase spread is shown (B, E, I). Known centromere proteins and FISH probe, and fluorescence intensity line-scans, are asdetailed in Figure 1 except that anti CENP-A was used instead of CREST3. For each chromosome, multi-colour and split-colour images are shown: blue (DAPI) forthe chromosome; red (Texas Red) for DNMT1; and green (FITC) for known centromere proteins or FISH signal. DNMT1 is shown to localize to a larger regionthan CENP-A (A, B, H, I); to colocalize with pericentric heterochromatin at human chromosome 1 (C), human acrocentric chromosomes (D), and mouse chromo-somes (E); to localize to the mardel(10) neocentromere (F), and to both the active (a) and inactive (i) centromeres of a human dicentric chromosome (G).


Dow


ic.oup.com/hm


alignments of chromosomes as they move to the metaphaseplate and/or play a role in regulating higher-order structure ofthe kinetochore (4). It is therefore possible that the other proteinsin subgroup (a) may perform similar or related functions.

Amongst the six new mammalian centromere-bindingproteins we have described, we have identified a functionallyrelated class that has not previously been localized tocentromeres. These are the histone acetyltransferases PCAF,MYST and P300/CBP. Why would proteins usually involved informing active chromatin be present at centromeres? Oneexplanation is that acetylation levels are the result of a dynamicequilibrium between HDACs and HATs and therefore bothgroups of proteins are expected to show overlapping localiza-tion (40). In addition, it has been shown that PCAF andHDAC1 interact in vivo and that HATs are integrated into alarge multiprotein HDAC complex (41).

The loss of DNMT3b has been shown to result in ICFsyndrome (42), which manifests at the cytological level asdecondensation and instability of the major regions of humanpericentric heterochromatin (reviewed in 43). The loss ofheterochromatin proteins has also been shown to result in anumber of other genetic diseases: e.g. Rett syndrome due toMeCP2 (44), ATRX syndrome due to ATRX (34), andTownes–Brocks syndrome due to SALL1 (45). The possibilitythat disruption in pericentric heterochromatin organization mayplay a significant part in the aetiology of these diseases shouldbe a useful area of further investigation.

In summary, we have demonstrated an association withmammalian metaphase centromeres of 15 proteins previouslyreported to participate in a wide range of roles in chromatin andtranscriptional modulation. Of these proteins, six have notpreviously been described at mammalian centromeres, while

Figure 5. Immunofluorescence results showing distribution of MBD2 on human and mouse chromosomes. Detailed explanations are as for Figure 4.


Dow


ic.oup.com/hm


Figure 6. Immunofluorescence results showing chromosomal distribution of ORC2. For human chromosomes, examples of chromosome 1, an acrocentric chro-mosome, and a partial metaphase spread are shown (A, B, C, F, J); for mouse, a randomly chosen chromosome and a partial metaphase spread are shown (D, E, G,K). In human chromosomes, ORC2 is shown to colocalize with CENP-A (A, B, J) and slightly on the lateral inside edges of BubR1 (C), and identified a smallerregion than heterochromatin protein MBD1 (F) and frequently to human acrocentric short arm satellites (B). In mouse chromosomes, however, ORC2 identified alarger region than CENP-A and BubR1 (D, E, K) and colocalized with pericentric heterochromatin protein MBD1 (G). ORC2 also localized to the mardel(10)neocentromere (H) and to the active (a) and inactive (i) centromeres of a human dicentric chromosome (I).


Dow


ic.oup.com/hm


the remaining ones have been reported as present on eithermetaphase or interphase centromeres (10,21,33,34,36,46).Other chromatin-associated proteins that have been reportedto bind mammalian pericentric heterochromatin, and which arenot included in the present study, include the HP1 isoformHP1b (9,10); its binding partner SUV39H1 (35), and chromatintarget, MeH3K9 (47); the histone variant H2A.Z (48);transcriptional corepressors TIF1b, KRAZ1, KRAZ2, SALL1;and the chromatin remodellers ACF1 and WSTF (reviewed

in 21). Together, these observations point to the extensivesharing of protein components, many likely to be acting co-operatively through macromolecular complexes, at gene loci,centromeres and various chromosomal heterochromatic land-marks. The classification of these protein components intothree distinct chromosomal distribution subgroups shouldprovide a useful basis for the further delineation of thebroad-ranging and largely unknown roles of these proteins atthe different chromosomal locations.

Figure 7. Immunofluorescence results showing distribution of P300/CBP on representative human and mouse chromosomes. Detailed explanations are as forFigure 6, except that P300/CBP did not normally localize to human acrocentric short arm satellites (data not shown) or to the inactive centromere (i) of the dicentricchromosome (H).


Dow


ic.oup.com/hm


MATERIALS AND METHODS

Cell lines

We used a lymphoblastoid cell line established from a patientwith the mardel(10) neocentromere (27), maintained in RPMI1640 medium (Trace Biosciences, Australia) supplementedwith 20% fetal calf serum (FCS); a somatic cell hybrid [ES-mardel(10)-1] made by transferring the mardel(10) chromo-some into the mouse ES cell line ES129-1, maintained in EScell medium (Trace) supplemented with b-mercaptoethanol andLIF; a dicentric lymphoblastoid cell line from a patient with at(X;15) translocation (26) maintained in RPMI 1640 medium(Trace) supplemented with 20% FCS; and a transformed mouseT cell line VL3-3M2, maintained in RPMI 1640 medium(Trace) supplemented with 5% FCS. Cells were arrested inmetaphase using standard Colcemid treatment.

Antibodies and immunofluorescence

The specificity of all antibodies used in this study had beenprevious tested by one or more of the following techniques:western blotting, immunoprecipitation, immunohistochemistryand ELISA. Human anti-centromere antisera (ACA) wereprovided and characterized by S. Wittingham and T. Kaye,Royal Melbourne Hospital. ACA CREST3 recognized primarilyCENP-A, and ACA CREST6 recognized primarily CENP-Aand CENP-B (27). Rabbit anti-mouse CENP-A, and rabbit anti-human CENP-A antibodies have been described elsewhere(27,49,50). CREST3 was used in place of anti CENP-A whenco-detected with antibodies raised in rabbits and for consis-tency when comparing inner or outer centromere localization;CREST6 was used to identify both centromeres of dicentricchromosomes. Other antibodies used were as follows (withdilutions and preferred buffers, see below): mouse anti humanATRX (23c or 39f), from Doug Higgs, MRC Molecular

Haematology Unit, IMM, Oxford, UK [1:20, conditions asMcDowell et al. (34)]; rabbit anti human BAF180 fromWeidong Wang, Laboratory of Genetics, NIH, Baltimore, MD,USA (1:100, KCM); goat anti human DNMT1, from SantaCruz (1:25, TEENþ) or mouse anti human DNMT1 fromImgenex, San Diego, CA, USA (1:25, TEENþ); goat antihuman DNMT3b, from Santa Cruz (1:25, TEENþ) or mouseanti mouse DNMT3b, from Imgenex (25, TEENþ); rabbitanti human HDAC1, from Calbiochem, San Diego, CA,USA (1:50, TEENþ) or goat anti human HDAC1, fromSanta Cruz (1:100; TEENþ); rabbit anti human HDAC2, fromCalbiochem (1:50, TEENþ) or goat anti human HDAC2,from Santa Cruz (1:100; TEENþ); mouse anti human HP1a(clone 2HP1H5), from Pierre Chambon, CNRS, INSERM,Strasbourg, France [1:250; conditions as Minc et al. (54)];sheep anti human MBD1, from Brian Hendrich, ICMB,University of Edinburgh, UK (1:200, TEENþ); sheep antihuman MBD2, from Brian Hendrich (1:100, TEENþ) ormouse anti human MBD2 from Imgenex (1:50, TEENþ);rabbit anti rat MeCP2, from Brian Hendrich (1:200, TEENþ)or rabbit anti mouse MeCP2, from Upstate Biotechnology,Waltham, MA, USA (1:200, TEENþ); rabbit anti yeast MYSTfamily, from Upstate Biotechnology (1:100, TEENþ); mouseanti human ORC2 (Ab-2), from Calbiochem (1:20, TEEN) ormouse anti human ORC2, from NeoMarkers, Fremont, CA,USA (1:50, TEENþ); mouse anti human P300/CBP (mixedepitope), from Upstate (1 :25, TEENþ) or mouse anti humanP300/CBP (RW128), from Upstate (1:50, TEENþ); mouseanti human PCAF, from Santa Cruz (1:25, KCM) or rabbit antihuman PCAF, from Upstate (1:25, TEENþ); rabbit anti mousemSin3A (AK-11), from Santa Cruz (1:100, TEENþ) or rabbitanti mouse mSin3A, from Upstate (1 :25, TEENþ). Table 2summarizes the quality control information (i.e. whetheraffinity purified, and/or analysed by western blotting, immu-nohistochemistry or ELISA) for each of the proteins that havenot previously been observed at metaphase centromeres

Table 2. Quality control data for proteins not previously localized to mammalian metaphase centromeres (see Table 1)

Protein Source(s) of antibody Immunogen species Raised in Affinity purified Quality controla

DNMT1 Santa Cruz Human Goat Yes WB, IHCImgenex Human Mouse No WB

DNMT3b Santa Cruz Human Goat Yes WBImgenex Mouse Mouse No WB

HDAC1 Santa Cruz Human Goat Yes WBCalbiochem Human Rabbit Yes WB

HDAC2 Santa Cruz Human Goat Yes WBCalbiochem Human Rabbit Yes WB

MBD2 Brian Hendrich Human Sheep No WBImgenex Human Mouse ND WB, ELISA

MYST Upstate Yeast Rabbit ND WBORC2 Calbiochem Human Mouse Yes WB

NeoMarkers Human Mouse Yes WBP300/CBP Upstate (mix epitope) Human Mouse ND WB

Upstate (RW128) Human Mouse ND WB, IPPCAF Santa Cruz Human Mouse No WB, IP

Upstate Human Rabbit ND WB, IPSin3A Santa Cruz Mouse Rabbit Yes WB

Upstate Mouse Rabbit No WB

aWB, western blotting; IHC, immunohistochemistry; ELISA, enzyme-linked immunosorbent assay. ND, not determinable. Note that, with the exception of MYST,immunofluorescence studies were performed using antibodies from two different sources.


Dow


ic.oup.com/hm


(see Results and Table 1), and indicates that, with the exceptionof MYST, our results were confirmed with antibodies from twodifferent sources.

Immunofluorescence was carried out on unfixed, permea-blized cells according to published methods (26,51,52) withminor modifications. Antibodies were diluted either in TEENþ[1 mM triethanolamine–HCl (pH 8.5), 0.2 mM NaEDTA (pH9.0), 25 mM NaCl, 0.1% Triton X-100, 0.1% BSA] or KCM[20 mM KCl; 20 mM NaCl; 10 mM Tris–HCl; 0.5 mM NaEDTA;0.1% (v/v) Triton X-100]. An exception was the goat primaries,where 5% donkey serum was substituted for BSA in theTEENþ. All washes were performed using KCM andantibodies were detected using fluorescently labelled second-aries (Jackson Immunoresearch Laboratories, West Grove, PA,USA) and in the case of ATRX, HP1 (on human chromosomes)and ORC2, a fluorescently labelled tertiary antibody followingan unlabelled secondary. Antibodies were detected with TexasRed-conjugated secondaries and were co-localized with FITC-detected ACA CREST3 or anti-CENP-A, or ACA CREST6 fordicentric chromosomes and antibodies against HP1a (or MBD1for primaries raised in mouse) to stain pericentric hetero-chromatin. Immuno-FISH was carried out using the mardel(10)neocentromere-specific BAC E8 as previously described (27)and detected using FITC-labelled antibodies. Images werecaptured using IPLab software (Scanalytics, Fairfax, VA, USA).Graphs of total pixel intensity versus distance (line scans) weregenerated using the IPLab extension Line Measure with a pixelwidth of 3 for lateral scans across kinetochores and 13 for axialscans through centromeres (as illustrated in Fig. 1K and L). Thedistances between kinetochore doublet-signals for all thechromosomes within a single metaphase spread were measuredusing IPLab and Adobe Photoshop. Student’s t-test was appliedto co-localization data-sets and a p-value of 0.05 was used as acut-off above which two proteins were concluded to coincide.

ACKNOWLEDGEMENTS

We thank L. Shaffer for MES10 and dicentric cell lines,respectively; S. Smale for cell line VL3-3M2; P. Kalitsis for theCENP-A antibody; D. Higgs and T. McDowell for ATRXantibodies; W. Wang and D. Murray for BAF180 antibody; P.Chambon for HP1a antibody; and B. Hendrich for MBD1,MBD2 and MeCP2 antibodies. J.M.C. and K.H.A.C. areassociates of the Department of Paediatrics, University ofMelbourne. K.H.A.C. is a Senior Principal Research Fellow ofNH&MRC of Australia. This work was supported by fundingfrom NH&MRC.

REFERENCES

1. Cleveland, D.W., Mao, Y. and Sullivan, K.F. (2003) Centromeres andkinetochores: from epigenetics to mitotic checkpoint signaling. Cell,112, 407–421.

2. Amor, D.J. and Choo, K.H.A. (2002) Neocentromeres: role in humandisease, evolution, and centromere study. Am. J. Hum. Genet., 71, 695–714.

3. Saffery, R., Irvine, D.V., Griffiths, B., Kalitsis, P., Wordeman, L. andChoo, K.H.A. (2000) Human centromeres and neocentromeres showidentical distribution patterns of >20 functionally importantkinetochore-associated proteins. Hum. Mol. Genet., 9, 175–185.

4. Aagaard, L., Schmid, M., Warburton, P. and Jenuwein, T. (2000) Mitoticphosphorylation of SUV39H1, a novel component of active centromeres,coincides with transient accumulation at mammalian centromeres. J. CellSci., 113 (Pt 5), 817–829.

5. Lo, A.W., Craig, J.M., Saffery, R., Kalitsis, P., Irvine, D.V., Earle, E.,Magliano, D.J. and Choo, K.H.A. (2001) A 330 kb CENP-A bindingdomain and altered replication timing at a human neocentromere. EMBO J.,20, 2087–2096.

6. Craig, J.M., Wong, L.H., Lo, A.W.I., Earle, E., Choo, K. H. A. (2003)Centromeric chromatin pliability and memory at a human neocentromere.EMBO J., 22, 2495–2504.

7. Dillon, N. and Festenstein, R. (2002) Unravelling heterochromatin:competition between positive and negative factors regulates accessibility.Trends Genet., 18, 252–258.

8. Nicol, L. and Jeppesen, P. (1994) Human autoimmune sera recognize aconserved 26 kD protein associated with mammalian heterochromatin thatis homologous to heterochromatin protein 1 of Drosophila. ChromosomeRes., 2, 245–253.

9. Wreggett, K.A., Hill, F., James, P.S., Hutchings, A., Butcher, G.W. andSingh, P.B. (1994) A mammalian homologue of Drosophilaheterochromatin protein 1 (HP1) is a component of constitutiveheterochromatin. Cytogenet. Cell Genet., 66, 99–103.

10. Minc, E., Allory, Y., Courvalin, J.C. and Buendia, B. (2001)Immunolocalization of HP1 proteins in metaphasic mammalianchromosomes. Meth. Cell Sci., 23, 171–174.

11. Partridge, J.F., Borgstrom, B. and Allshire, R.C. (2000) Distinct proteininteraction domains and protein spreading in a complex centromere.Genes Dev., 14, 783–791.

12. Blower, M.D. and Karpen, G.H. (2001) The role of Drosophila CID inkinetochore formation, cell-cycle progression and heterochromatininteractions. Nat. Cell Biol., 3, 730–739.

13. Brown, K.E., Guest, S.S., Smale, S.T., Hahm, K., Merkenschlager, M.and Fisher, A.G. (1997) Association of transcriptionally silent genes withIkaros complexes at centromeric heterochromatin. Cell, 91, 845–854.

14. Francastel, C., Magis, W. and Groudine, M. (2001) Nuclear relocation of atransactivator subunit precedes target gene activation. Proc. Natl Acad. Sci.USA, 98, 12120–12125.

15. Nielsen, S.J., Schneider, R., Bauer, U.M., Bannister, A.J., Morrison, A.,O’Carroll, D., Firestein, R., Cleary, M., Jenuwein, T., Herrera, R.E. et al.(2001) Rb targets histone H3 methylation and HP1 to promoters. Nature,412, 561–565.

16. Henikoff, S., Ahmad, K., Platero, J.S. and van Steensel, B. (2000)Heterochromatic deposition of centromeric histone H3-like proteins.Proc. Natl Acad. Sci. USA, 97, 716–721.

17. Felsenfeld, G. and Groudine, M. (2003) Controlling the double helix.Nature, 421, 448–453.

18. Narlikar, G.J., Fan, H.Y. and Kingston, R.E. (2002) Cooperation betweencomplexes that regulate chromatin structure and transcription. Cell,108, 475–487.

19. Meehan, R.R. (2003) DNA methylation in animal development. Semin.Cell. Dev. Biol., 14, 53–65.

20. Turner, J. and Crossley, M. (2001) The CtBP family: enigmatic andenzymatic transcriptional co-repressors. Bioessays, 23, 683–690.

21. Dellaire, G., Farrall, R. and Bickmore, W.A. (2003) The Nuclear ProteinDatabase (NPD): sub-nuclear localisation and functional annotation of thenuclear proteome. Nucl. Acids Res., 31, 328–330.

22. Wade, P.A. (2001) Methyl CpG-binding proteins and transcriptionalrepression. Bioessays, 23, 1131–1137.

23. Keller, C., Ladenburger, E.M., Kremer, M. and Knippers, R. (2002) Theorigin recognition complex marks a replication origin in the human TOP1gene promoter. J. Biol. Chem., 277, 31430–31440.

24. Hsu, T.C. and Arrighi, F.E. (1971) Distribution of constitutive hetero-chromatin in mamallian chromosomes. Chromosoma, 34, 243–253.

25. Choo, K.H.A (1997) The Centromere. Oxford University Press, Oxford.26. Page, S.L., Earnshaw, W.C., Choo, K.H.A. and Shaffer, L.G. (1995) Further

evidence that CENP-C is a necessary component of active centromeres:studies of a dic(X; 15) with simultaneous immunofluorescence and FISH.Hum. Mol. Genet., 4, 289–294.

27. du Sart, D., Cancilla, M.R., Earle, E., Mao, J.I., Saffery, R., Tainton, K.M.,Kalitsis, P., Martyn, J., Barry, A.E. and Choo, K.H.A. (1997) Afunctional neo-centromere formed through activation of a latent humancentromere and consisting of non-alpha-satellite DNA. Nat. Genet.,16, 144–153.


Dow


ic.oup.com/hm


28. Craig, J.M., Wong, L.H., Lo, A.W., Earle, E. and Choo, K.H.A. (2003)Centromeric chromatin pliability and memory at a human neocentromere.EMBO J., 22, 2495–2504.

29. Saffery, R., Sumer, H., Hassan, S., Wong, L.H., Craig, J.M., Todokoro, K.,Anderson, M., Stafford, A.J. and Choo, K.H.A. (2003) Transcription withina functional human centromere. Mol. Cell, 12, 509–516.

30. Warburton, P.E., Cooke, C.A., Bourassa, S., Vafa, O., Sullivan, B.A.,Stetten, G., Gimelli, G., Warburton, D., Tyler-Smith, C., Sullivan, K.F. et al.(1997) Immunolocalization of CENP-A suggests a distinct nucleosomestructure at the inner kinetochore plate of active centromeres. Curr. Biol.,7, 901–904.

31. Jablonski, S.A., Chan, G.K., Cooke, C.A., Earnshaw, W.C. and Yen, T.J.(1998) The hBUB1 and hBUBR1 kinases sequentially assemble ontokinetochores during prophase with hBUBR1 concentrating at thekinetochore plates in mitosis. Chromosoma, 107, 386–396.

32. Nielsen, A.L., Ortiz, J.A., You, J., Oulad-Abdelghani, M., Khechumian, R.,Gansmuller, A., Chambon, P. and Losson, R. (1999) Interaction withmembers of the heterochromatin protein 1 (HP1) family andhistone deacetylation are differentially involved in transcriptionalsilencing by members of the TIF1 family. EMBO J., 18, 6385–6395.

33. Ng, H.H., Jeppesen, P. and Bird, A. (2000) Active repression ofmethylated genes by the chromosomal protein MBD1. Mol. Cell. Biol.,20, 1394–1406.

34. McDowell, T.L., Gibbons, R.J., Sutherland, H., O’Rourke, D.M.,Bickmore, W.A., Pombo, A., Turley, H., Gatter, K., Picketts, D.J.,Buckle, V.J. et al. (1999) Localization of a putative transcriptionalregulator (ATRX) at pericentromeric heterochromatin and the short armsof acrocentric chromosomes. Proc. Natl Acad. Sci. USA, 96,13983–13988.

35. Aagaard, L., Laible, G., Selenko, P., Schmid, M., Dorn, R., Schotta, G.,Kuhfittig, S., Wolf, A., Lebersorger, A., Singh, P.B. et al. (1999)Functional mammalian homologues of the DrosophilaPEV-modifier Su(var)3-9 encode centromere-associated proteinswhich complex with the heterochromatin component M31. EMBO J.,18, 1923–1938.

36. Lewis, J.D., Meehan, R.R., Henzel, W.J., Maurer-Fogy, I., Jeppesen, P.,Klein, F. and Bird, A. (1992) Purification, sequence, and cellularlocalization of a novel chromosomal protein that binds to methylated DNA.Cell, 69, 905–914.

37. Rountree, M.R., Bachman, K.E. and Baylin, S.B. (2000) DNMT1 bindsHDAC2 and a new co-repressor, DMAP1, to form a complex at replicationfoci. Nat. Genet., 25, 269–277.

38. Bachman, K.E., Rountree, M.R. and Baylin, S.B. (2001) Dnmt3a andDnmt3b are transcriptional repressors that exhibit unique localizationproperties to heterochromatin. J. Biol. Chem., 276, 32282–32287.

39. Taplick, J., Kurtev, V., Kroboth, K., Posch, M., Lechner, T. and Seiser, C.(2001) Homo-oligomerisation and nuclear localisation of mouse histonedeacetylase 1. J. Mol. Biol., 308, 27–38.

40. Hendzel, M.J., Kruhlak, M.J. and Bazett-Jones, D.P. (1998) Organization ofhighly acetylated chromatin around sites of heterogeneous nuclear RNAaccumulation. Mol. Biol. Cell., 9, 2491–2507.

41. Yamagoe, S., Kanno, T., Kanno, Y., Sasaki, S., Siegel, R.M., Lenardo, M.J.,Humphrey, G., Wang, Y., Nakatani, Y., Howard, B.H. et al. (2003)Interaction of histone acetylases and deacetylases in vivo. Mol. Cell. Biol.,23, 1025–1033.

42. Hansen, R.S., Wijmenga, C., Luo, P., Stanek, A.M., Canfield, T.K.,Weemaes, C.M. and Gartler, S.M. (1999) The DNMT3B DNAmethyltransferase gene is mutated in the ICF immunodeficiencysyndrome. Proc. Natl Acad. Sci. USA, 96, 14412–14417.

43. Robertson, K.D. and Wolffe, A.P. (2000) DNA methylation in health anddisease. Nat. Rev. Genet., 1, 11–19.

44. Amir, R.E., Van den Veyver, I.B., Wan, M., Tran, C.Q., Francke, U. andZoghbi, H.Y. (1999) Rett syndrome is caused by mutations in X-linkedMECP2, encoding methyl-CpG-binding protein 2. Nat. Genet., 23, 185–188.

45. Netzer, C., Rieger, L., Brero, A., Zhang, C.D., Hinzke, M., Kohlhase, J. andBohlander, S.K. (2001) SALL1, the gene mutated in Townes-Brockssyndrome, encodes a transcriptional repressor which interacts withTRF1/PIN2 and localizes to pericentromeric heterochromatin. Hum. Mol.Genet., 10, 3017–3024.

46. Xue, Y., Canman, J.C., Lee, C.S., Nie, Z., Yang, D., Moreno, G.T.,Young, M.K., Salmon, E.D. and Wang, W. (2000) The human SWI/SNF-Bchromatin-remodeling complex is related to yeast rsc and localizes atkinetochores of mitotic chromosomes. Proc. Natl Acad. Sci. USA,97, 13015–13020.

47. Cowell, I.G., Aucott, R., Mahadevaiah, S.K., Burgoyne, P.S., Huskisson, N.,Bongiorni, S., Prantera, G., Fanti, L., Pimpinelli, S., Wu, R. et al. (2002)Heterochromatin, HP1 and methylation at lysine 9 of histone H3 in animals.Chromosoma, 111, 22–36.

48. Rangasamy, D., Berven, L., Ridgway, P. and Tremethick, D.J. (2003)Pericentric heterochromatin becomes enriched with H2A.Z during earlymammalian development. EMBO J., 22, 1599–1607.

49. Saffery, R., Earle, E., Irvine, D.V., Kalitsis, P. and Choo, K.H.A. (1999)Conservation of centromere protein in vertebrates. Chromosome Res.,7, 261–265.

50. Lo, A.W., Magliano, D.J., Sibson, M.C., Kalitsis, P., Craig, J.M. andChoo, K.H.A. (2001) A novel chromatin immunoprecipitation and array(CIA) analysis identifies a 460-kb CENP-A-binding neocentromere DNA.Genome Res., 11, 448–457.

51. Jeppesen, P., Mitchell, A., Turner, B. and Perry, P. (1992) Antibodies todefined histone epitopes reveal variations in chromatin conformationand underacetylation of centric heterochromatin in human metaphasechromosomes. Chromosoma, 101, 322–332.

52. Sullivan, B.A. and Warburton, P.E. (1999) Studying progression ofvertebrate chromosomes through mitosis by immunofluorescence andFISH. In Bickmore, W.A. (ed.), Chromosome Structural Analysis: aPractical Approach. Oxford University Press, Oxford.

53. Silverstein, R.A., Richardson, W., Levin, H., Allshire, R. and Ekwall, K.(2003) A new role for the transcriptional corepressor SIN3; regulation ofcentromeres. Curr. Biol., 13, 68–72.

54. Minc, E., Allory, Y., Worman, H.J., Courvalin, J.C. and Buendia, B. (1999)Localization and phosphorylation of HP1 proteins during the cell cycle inmammalian cells. Chromosoma, 108, 220–234.


Dow


ic.oup.com/hm


ddg330.pdf - Oxford Academic

Documents