- Email: [email protected]
The mammalian centromere: Its molecular architecture
Fundamental Mechanisms
ELSEVIER
Mutation
Research
372 (1996)
and Molecular of Mutagenesis
153-167
The mammalian centromere: Its molecular architecture Arthur R. Mitchell
*
Abstract The DNA and protein composition of the centromeric domains in mammalian chromosomes is now relatively well characterised. The major families of repeated DNAs, i.e., the simple-sequence and alphoids in man and the satellite sequences (both minor and major) in the mouse have been sequenced and long-range maps using pulse-field gels of some centromeres have been carried out. Autoimmune antibodies have provided an insight into some of the proteins which interact with these DNA sequences. Although the individual components of the mammalian centromere may have been identified, how they interact with each other to give the functional structure visualised by electron microscopy is yet to be determined. This review examines our understanding of these separate components. Keyords;
Mammalian:
Centromere:
Centromere
architecture
1. Introduction In mammalian chromosomesthe position of the centromere has been recognised for many years by cytologists as a physical constriction or narrowing of the chromatin at a specific place along the chromosome. Because of this narrowing feature the centromere is referred to as the primary constriction of the chromosome.This position is of particular importance since it is thought to be the site of attachment of the tubulin microtubules formed during cell division. This leads to the correct alignment of the chromosomeson the metaphaseplate at cell division which in turn gives rise to the correct segregationof chromosomesto each daughter cell at anaphase.Loss of centromere function is normally associatedwith chromosomeinstability leading to chromosomeloss
Corresponding author. Tel.: 2620: E-mail: [email protected] 0027-S 107/96/$lS.O0 PII
SOO27-5107~96~00135-2
Copyright
0131 467 8319;
6 1996 Elsevier
Fax: 0131 343
in subsequentcell divisions. This article will examine our understandingof the molecular nature of the mammaliancentromere at both the protein and DNA level. Most structural evidence in mammalshasbeen derived from studies on the chromosomesof man and mouse. 1 will also attempt to see if our understanding of the structural components of a mammalian centromere has led to an insight into a functional role for one or more of the molecular components. Before beginning it would be wise to define the terminology of the centromere. The terms ‘ kinetochore’ and ‘centromere’ were originally coined to describe the spindle attachment region on chromosomes(Sharp. 1934: Darlington, 1937). Later by using the electron microscopeit was shown that a trilaminar plate or disc was present on somechromosomes at this site (Jokelainen. 1967; Comings and Akada, 197I) and the term ‘kinetochore’ was used by some to describe this visible structure. In this
Science B.V. All rights rewrvrd
present article I will use ‘centromere’ as the attachment site. The term ‘centromeric domain’ will be used to define the chromatin immediately surrounding and underlying the centromere. This centromeric chromatin can be recognised at both the cytological and molecular level by being composed of specific well-characterised DNA sequences. ‘Kinetochore domain’ (Eamshaw, 1991) will be used solely to describe those DNA sequences that have been shown to lie directly beneath a visible trilaminar structure. The kinetochore domain will therefore be a specific region within the centromeric domain of the chromosome.
bask repr
unii
Fig. I. A model showing the amplification of the basic 170 bp repeat unit of alphoid DNA sequences leading to the formation of large chromosome specific arrays at the centromeres of human chromosomes.
2. The DNA content Although it was known from kinetic studies that the genome of eukaryotes contained DNA sequences of different complexity and copy number, it was not until Pardue and Gall (1969) and Jones (1970) localised mouse major satellite DNA to the centromeric domain of mouse chromosomes that it became possible to correlate cytological landmarks on the chromosome with specific repetitive DNA families. Rapid developments using the technique of in situ hybridisation followed on from these initial studies. Thus the simple-sequence satellites from man (i.e., those DNAs isolated from the bulk of genomic DNA using density gradient centrifugation) were mapped to the centromeric domains of certain human chromosomes (Gosden et al.. 1975). The use of restriction endonucleases led to the identification of other repetitive DNAs such as the SAU3A (Agresti et al., 1987) and HAE3 (Agresti et al., 1989) families. One such family, the alphoid DNA family originally described by Maio et al. (1981). was particularly interesting since in man chromosome specific subsets were isolated using different restriction enzymes. Manuelidis (1978) using EcoRI identified a member of this family which hybridised predominantly to chromosomes I, 3, 7. IO and 19. Willard et al. (1983) and Yang et al. (1982) found an X specific member of the alphoid DNA family. Jabs et al. (1984) isolated a subset using XbaI which hybridised to the centromeric domains of human chromosomes 2. 4, 14, 15. 18, 20 and 22. The alphoid
repeated DNA family is comprised of a short monomeric unit of - 170 bp. This has undergone multiple rounds of amplification so today it represents somewhere around 5% of the human genome, i.e., 3 X lo* bp. During this process mutations are incorporated giving rise to different but related subsets of sequences (Fig. I). It is this process which gives rise to the subsets of alphoid arrays which are chromosome specific. Mitchell et al. (1985) first showed that alphoid DNAs were present at the centromeric domains of all chromosomes in man. The p82H sequence (Mitchell et al., 1985) was heterogeneous in sequence with adjacent monomeric mismatches varying from 8-26s. In all probability it was this high degree of sequence mismatch which allowed the p82H clone to hybridise to the alphoid sequences on all human chromosomes. Thus for the first time a family of repeated DNAs were associated with the centromeric domains of all chromosomes in the human genome. In comparison, mouse major satellite was not detectable on the Y chromosome (Jones, 1970). The position of the simple-sequence satellites of man. as determined by in situ studies, positioned them further from the primary constriction than alphoid sequences. Sequences homologous to the human alphoid DNAs were also detected in the higher primates (Miller et al.. 1988; Baldini et al., 1991) suggestive of some degree of evolutionary conservation. Pietras et al. (1983) had previously cloned, using DNA reassociation kinetics as the tool, a second
repeated DNA sequence from the mouse genome. In comparison to the major satellite in the mouse (which constitutes around 10% of the genome) this satellite represented less than 1% of the genome with a consequence that it was known as the mouse minor satellite. The monomeric repeat of this satellite was 120 bp. Initially in situ hybridisation suggested that it was present on only some of the chromosomes in the mouse but that it was centromerically located (Pietras et al.. 19831. Later Wong and Rattner (19881 and Joseph et al. (19891 showed that this DNA family was present on all chromosomes except the Y chromosome in mouse. Both the human alphoid and mouse minor satellite DNAs shared a common feature which was that they were situated closer to the primary constriction of chromosomes than either the human simple sequence DNAs or mouse major satellite DNA (Joseph et al., 1989; Mitchell et al., 19921. The centromeric domain of chromosomes in man and mouse appeared therefore to be comprised of repetitive DNAs. At the structural level as shown by in situ hybridisation a similar type of organisation was present within the centromeric domains of both species. However, at the DNA sequence level little homology between these two DNA t’dmilies could be recognised. On the basis of this it was difficult to establish a functional relationship between these two repeated DNAs. It was known that somatic cell hybrids between mouse and human cells could be successfully maintained implying a common feature within the centromeres of the chromosomes in both species for spindle attachment at cell division.
3. The protein content The serum from some patients suffering from the autoimmune disease CREST contains autoimmune antibodies which react specifically with centromeric proteins (Moroi et al.. 19801. CREST ACA (anticentromere autoantibody) staining of chromosomes produces a double dot signal at the primary constriction in all mammalian species tested. Electron microscopy has mapped the CREST antigens close to the centromere (Brenner et al., 198 1). A combination of CREST ACA staining with PRINS in situ hybridisation (Mitchell et al., 1992) confirmed the earlier observations that the alphoid repeats were physically
Fig. 2. PRINS in situ hybridisation using (a) a human alphoid wquence in combination with CREST staining (b). Note that the in situ signal of the alphoid sequence and the CREST signal co-localise to the same position in human centromeres. The arrow points to a chromosome 9. In (c) the chromosomes are stained with DAPl.
Fig. 3. This shows PRINS in situ hybridisation using a human simple satellite DNA sequence (a). The CREST signal is again shown in (b) and, the DAPI stain is shown in (cl. kote that the in Gtu signal for the simple-cequence DNA and CREST do not co-localise to the same position within the centromeric domain on human chromosome 9.
close to the centromere (Fig. 2). Simple-sequence DNAs by comparison were mapped some distance from the centromere (as defined by CREST ACA indirect immunofluorescence) (Fig. 3). The major protein antigens recognised by patients with CREST anticentromere antibodies (ACA) have been identified. Three proteins termed CENP-A, CENP-B and CENP-C (Earnshaw and Rothfield, 1985) with molecular weights of 17, 80 and 140 kDa have been studied in detail. More recently two other centromere proteins CENP-E (312 kDa); Yen et al., 1991, 1992) and (CENP-F; (400 kDa); Rattner et al., 1993) have also been identified (CENP stands for centromere protein). A related group of proteins have been termed the itmer ccntromere proteins (INCENPS) (Mackay andzarnshaw, 1993). CENP-E and CENP-F are members of this group. They are sometimes called passenger proteins since they associate with chromosomes at specific stages during cell division. During interphase they may be associated with the nuclear cytoskeleton (Mackay and Earnshaw, 1993). Some like CENP-E are motor proteins related to the kinesins whose role is to promote microtubule and thus chromosome movement during mitosis (Thrower et al., 1995). CENP-A is a histone like molecule (Palmer and Margolis. 1985; Palmer et al., 1987) related most closely to histone H3 (Sullivan et al., 1994). Full length cDNA of CENP-A shows the sequence of this protein to be 62% identical to histone H3. CENP-A is found in nucleosomes and one possible role may be to replace the H3 molecule specifically in those nucleosomes which are directly associated with the centromere itself. Why the centromere should require nucleosomes of different histone octamer content from the rest of the DNA is not yet clear. One hypothesis is that CENP-A may provide a structure which can withstand the stronger stress forces present during the process of cell division. CENP-C was found to be a highly basic protein of 943 amino acids (Saitoh et al., 1992). Some conflicting data on its properties exist. Saitoh et al. (1992) found no evidence for any DNA binding motifs whereas Sugimoto et al. (1994) reported that a specific DNA binding domain of 101 amino acids is present by expressing truncated *in frame deletions’ of the protein in E. co/i. Although not resolved, there is little doubt that this protein is part of the kinetochore structure. Saitoh et al. (1992) using im-
157
munoelectron microscopy localised the protein to the inner plate of the kinetochore. In terms of understanding the interactions of the CENPs with centromeric DNAs then CENP-B has provided the most insight. Masumoto et al. (1989) demonstrated that human alphoid DNA could be immunoprecipitated using CREST ACA serum. DNasel footprinting showed that a 17 bp motif (CTTCGTTGGAAACGGGA) within some alphoid monomers directly bound the CENP-B protein. This 17 bp motif was called the CENP-B box. Database searches showed a similar CENP-B box (ATTCGTTGGAAACGGGA) to be present in some of the mouse minor satellite sequences. Thus a direct link was established between centromerically located DNAs and the CENP-B protein. This finding also implied that both DNA sequences may carry out similar functions at the centromere of chromosomes, a conclusion which could not be reached from DNA sequence analysis alone nor from the implied functional role indirectly suggested by the in situ results which were described earlier. Further analysis of the interactions between the alphoid DNA and the CENP-B protein suggested that two alphoid monomers containing the CENP-B motif interact with the protein giving rise to a stable dimer structure maintained by protein-protein hydrophobic bonds (Yoda et al.. 1992; Kitagawa et al., 1995). The DNA binding domain was localised to the amino end whereas the protein:protein binding domain resided at the carboxyl end of the CENP-B molecule. A function for the CENP-B protein was predicted from this result which showed that one possible role for this protein within the chromatin containing the alphoid DNAs was to cause condensation of the long arrays (see Fig. 4). Because of the repetitive nature of the alphoid DNA and the fact that many CENP-B binding sites would be present, then this could result in the formation of higher ordered structures of this chromatin. Some support for this model came from the finding that the distribution of CENP-B binding sites was not uniform throughout the alphoid arrays on human chromosomes. One example is human chromosome 21 where two separate blocks of alphoid arrays are found. One array (o 21- 1) comprising around 1.3 Mb of DNA had a CENP-B binding site every alternate monomer. The second array (o 2 1- 11I was composed of more divergent alphoid monomers
Fig. 4. Possible role for the CENP-B protein in chromatin condensation. Alphold monomers with CENP-B motifs are \hown in black. The CENP-B protein (open circles) causes the long alphoid arrays to package into tight structures.
with less frequent binding sites (1 site per 100 alphoid monomers> for the protein (Ikeno et al., 1994). Furthermore in situ hybridisation in combination with indirect immunofluorescence with CREST ACA serum suggested that the cx21- 1 chromatin appeared more compact than the (Y21- 11 chromatin in support of the proposed model. Binding sites for the CENP-B protein are also present within the monomers of mouse minor satellite DNA and the assumption is that the CENP-B protein complexes to minor satellite monomers in a manner similar to that found for alphoid DNAs. Interestingly, Kipling et al. (1995) showed that in the mouse species MU ca~li, which lacks DNA sequences homologous to the Mus mz4.sc~ulus minor satellite, functional CENP-B binding sites are present within the centromerically located repeated DNAs in this animal. The CENP-B binding site in the Mus caroli satellite is diverged from the sequence found in minor satellite DNA although it binds human CENP-B protein. Nine bases in the CENP-B binding site have been conserved and appear absolute requirements for binding to occur. Thus the diverged sequence CENP-B box has a xTTCGxxxxAxxCGGGx. It appears that any base can be substituted for x. Nuclear proteins from Drosophila form complexes in gel retardation assays
with alphoid monomers containing CENP-B motifs. When challenged with CENP-B oligos, the complex is dissociated suggesting that the Drosophila proteins are homologs of the human CENP-B protein (Avides and Sunkel, 1994). Evolutionary conservation of both binding site and the possibility of homologous proteins in other organisms strongly argues in favour of the CENP-B protein having some common functional role in centromeric chromatin structure. Although the above evidence favours a role for the CENP-B protein in centromeric chromatin structure, it is far from clear if this protein is required for centromere function. At the EM level this protein underlies the kinetochore in the kinetochore domain chromatin (Cooke et al., 1990; Pluta et al.. 1990) and does not appear to be part of the proteinaceous trilaminar structure of the kinetochore plate itself. implying more a structural or supportive role for these DNA sequences in kinetochore domain chromatin. Studies on dicentric chromosomes have shown quite clearly that CENP-C is always associated with the active centromere and never with the inactive centromere, whereas CENP-B is found at both the active and inactive centromere in the dicentric chromosomes (Eamshaw et al., 1989; Page et al., 1995; Sullivan and Schwartz, 1995). Similarly, in the mouse examples have been reported of stable chromosomes without any detectable minor satellite DNA sequences and hence no CENP-B binding sites (Vig et al., 1994: Vig and Richards, 1992) and. at the molecular level, the alphoid sequences on the human Y chromosome have been extensively sequenced with no evidence that CENP-B binding motifs are present. Whether all available CENP-B binding sites are in fact used for binding the protein seems in doubt from studies on the DBA/2 mouse. In this animal chromosome I is polymorphic. containing two separate blocks of minor satellite DNA sequences (Mitchell et al., 1993). Although each block of minor satellite arrays contain CENP-B binding sites, the centromere is always associated with the more terminal block never with the internal array of minor satellite sequences (Fig. 5a). Thus even where CENP-B binding
sites are available it would appear that not all are utilised to bind the CENP-B protein. Although this figure shows the position of the centromere using CREST ACA serum. antibodies raised against the human CENP-B protein give a similar result. Because the CENP-B motif contains CpG dinucleotides we proposed (Mitchell et al., 1993) that methylation of cytosines within the arrays of minor satellite on chromosome 1 could have a role in determining the position of the centromere. When the minor satellite sequences on this chromosome are demethylated a redistribution of the CENP-B protein is seen (Mitchell et al.. in press) resulting in a labelling of both blocks of minor satellite sequences (Fig. 5b). Our interpretation of this result is that only some CENP-B motifs remain unmethylated in chromatin and hence capable of binding the protein. The majority of sites are methylated at the CpG dinucleotide and are incapable of binding the protein. In vitro methylation experiments (Masumoto, pers commun.) support our conclusions. Interestingly, making more sites within chromatin accessible to the CENP-B protein does not affect the position of the centromere nor does it apparently interfere with chromosome segregation. It has to be borne in mind that these experiments were carried out using short-term cultures and it remains to be seen whether demethylation of CENP-B binding sites has longer term effects on the stability of chromosomes.
4. The size of the mammalian
centromere
Much of the discussion above has centred on the role of the human alphoid and mouse minor satellite repeats in the formation of an active centromere. Both DNAs occur as tandem arrays which would tend to suggest that if they are indeed involved in centromere function then their interaction with centromeric proteins, whether the CENPs or other proteins which are yet to be identified, might predictably lead to a structure which is repetitive in
Fig. 5. (a) DBA/Z chromosome I (at-rowed) contains two blocks of mmor satellite arrays (red). The CREST signal (green) is always associated wtith the terminal block. (b) Redistribution of CENP-B signal on DBA/2 chromosome I when minor satellite DNA sequences are demethylated. (A) combined image of minor satellite in situ hybridisation (red) and CENP-B localisation (green). Blue is the DAPI stained chromosomes. The split images are shown in: (B) CENP-B signal (C) in situ minor satellite signal and (D) DAPI stained chromosomes.
160
AR. Mitchell/Mutation
nature. Zinkowski et al. (199 1) did indeed detect such a repeat-like structure at the centromeres in Indian muntjac chromosomes. Using physical methods such as a long hypotonic treatment followed by centrifugation onto glass slides centromeres on occasions were physically stretched. CREST ACA staining in these situations gave rise to beaded structures interpreted to be from a centromere whose subunits had been separated during this procedure. In situ hybridisation of human alphoid DNAs to extended chromatin fibres also suggest that the CREST ACA bind bead-like to some but not all of the chromatin containing the alphoid DNA (Haaf and Ward, 1994). Although CENP-C antibody normally gives a double dot signal on human centromeres, when it is applied to centromeres from cells with extended chromatin fibres 6-8 smaller signals are seen (Mitchell unpublished). This result would also suggest that the centromere consists of a structure of a repetitive nature. The mammalian kinetochore as seen under the EM (Ris and Witt, 1981) could be interpreted as a proteinaceous structure (i.e., the inner and outer plates) into which DNA sequences (alphoids? or mouse minor satellite DNA?) intercalate, binding some of the centromere proteins to give a stable structure capable of binding other proteins through protein:protein interactions similar to those seen between CENP-B molecules. The outer proteins in this complex are responsible for capturing microtubules at cell division. The number of microtubules required for this process is somewhat of a guessing game with figures of 20-25 separate tubules being required for attachment to any one kinetochore in human chromosomes. The most compelling evidence in favour of the alphoid sequences comprising the DNA content of the human centromere comes from work on the human Y chromosome (Tyler-Smith et al., 1993; Brown et al., 1994). Tyler-Smith et al. (1993) looked at a series of Y chromosomes which had undergone rearrangements and characterised both the stability of the rearranged chromosomes and analysed their DNA content by pulse-field gel in order to pinpoint which sequence was responsible for chromosome stability and hence centromere activity. One such chromosome was found to have a deletion within an alphoid array and the resulting phenotype was chromosome instability. Their data ruled out other sequences within this array which strongly pointed to
RrJmrch
372 (19%)
153-162
this alphoid array having a functional role in centromere activity. Size estimates for the centromere on the Y chromosome came up with a figure 500 kb. Of this they estimate that 300 kb is composed of alphoid sequences. Brown et al. (1994) used the technique of ’ telomere associated chromosome fragmentation’ (Bamett et al., 1993) to develop deletion of Y chromosomes. One, bYq74, appeared mitotitally stable with 140 kb of alphoid array within its centromere. A second construct, AYp 134, although having 540 kb of alphoid DNA was not stable. If only alphoid DNA is required for centromere activity then the 140 kb of alphoid sequence in AYq74 is sufficient. However, it cannot be ruled out that some other DNA sequence essential for centromere activity remains embedded in this array. This sequence if present would have to be nonrepetitive in nature. In Drosophila a minimum centromeric core size appears to be around 240 kb of DNA. This essential core element termed ‘ bora bora’ is a complex consisting not only of repetitive DNA elements but also middle repetitive and single-copy DNA sequences quite unlike the model envisaged for the human Y centromere described above (Murphy and Karpen, 1995). In contrast to both of the above, Garagna et al. ( 1995) found stable Robertsonian chromosomes in the mouse where only SO kb of minor satellite at the centromere remained after fusion. It is clear that we have some way to go before the definitive DNA sequence which is indispensible for centromere function is found. I leave you with the thought that there may not necessarily be one DNA sequence which confers this property. If a centromere is formed by the deposition of specific proteins then the important factor may simply be that of the DNA molecule adopting the correct 3-dimensional shape for a protein to recognise. Once this initial complex is formed then the rest of the reactions, like many cascade mechanisms in enzyme kinetics, is determined more by the energetics of the process than by sequence composition alone. Thus each chromosome may have developed its own solution to its problem of survival in the cell. Acknowledgements I would like to thank my colleagues Drs Peter Jeppesen and David Kipling for their helpful advice
A.R.
Mitchrll/Mutution
in preparing this paper. The photographic department of the Unit is also thanked for professional help.
References Agresti, A., Meneveri. R., Siccardi. A.G.. Marozzi, A., Comeo, G.. Guadi, S. and Ginelli. E. (1989) Linkage in human heterochromatin between highly divergent Sau 3A repeats and a new family of repeated DNA sequences (HAE 3 family). J. Mol. Biol., 205, 625-63 I. Agresti. A., Rainabli, G., Lobbiani, A., Magnani, I., Dilernia, R.. Meneveri, R.. Siccardi. A.G. and Ginelli, E. (1987) Chromosomal location by in situ hybridisation of the human sau3A family of DNA repeats. Hum. Genet., 75, 326-332. Avides. Maria C. and Sunkel, Claudia E. (1994) Isolation of chromosome-associated proteins from Drosophila mekdno~asrrr that bind a human centromere DNA sequence. .I. Cell Biol.. 127, 1159-i 171. Baldini, A., Miller, Dorothy A.. Miller. O.J., Ryder, O.A. and Mitchell A.R. (I 99 1J A chimpanzee-derived chromosomespecific alpha satellite DNA sequnce conserved between chimpanzee and human. Chromosoma, 100, 156-162. Barnett, M.A., Buckle, V.J.. Evans, E.P.. Porter. A.C.G., Rout, D., Smith. A.G. and Brown. W.R.A. (1993) Telomere directed fragmentation of mammalian chromosomes. Nucleic Acids Res.. 2 I( 27-36. Brenner, S., Pepper, D.. Berns, M.W., Tan, E. and Brinkley, B.R. (198 1I Kinetochore structure. duplication, and distribution in mammalian calls: analysis by human autoantibodies from scleroderma patients. J. Cell Biol.. 91. 95- 102. Brown. K.E.. Barnett. M.A.. Burgtorf, C., Shaw, P., Buckle, V.J. and Brown, W.R.A. (1994) Dissecting the centromere of the human Y chromosome with cloned telomeric DNA. Hum. Mol. Cen.. 3. 1227-1237. Comings, D.E. and Akada. T.A. (1971) Fine structure of kinetochore in Indian Muntjac. Exp. Cell Res.. 67, 97-110. Cooke, C.A.. Bernat, R.L. and Earnshaw, W.C. (I 990) CENP-B: a major human centromere protein located beneath the kinetochore. J. Cell Biol.. 110. 1475-1488. Darlington, CD. (1937) Recent advances in cytology. Philadelphia: Blakiston. Eamshaw, W.C. (1991) When is a centromere not a kinetochore? J. Cell Sci., 99, 1-4. Earnshaw. W.C. and Rothfield, N. (1985) Identification of a family of human centromere proteins using autoimmune sera from patients with scleroderma. Chromosoma. 91, 313-321. Earnshaw, W.C., Ratrie. H. and Stetten, G. (1989) Visualisation of centromere proteins CENP-B and CENP-C on a stable dicentric chromosome in cytological spreads. Chromosoma, 98, I-12. Garagna, S.. Broccoli. D.. Redi, C.A., Searle. J.B., Cooke. H.J. and Capanna, E. (1995) Robertsonian metacentrics of the house mouse lose telomeric sequences but retain some minor satellite DNA in the pericentromeric region. Chromosoma, 103, 685-692.
Reseurch
372
~IYYfil
153-162
161
Gosden, J.R., Mitchell, A.R., Buckland, R.A., Clayton, R.P. and Evans H.J. (1975) The location of four human satellite DNAs on human chromosomes. Exp. Cell Res.. 92. 148-152. Haaf. T. and Ward. D.C. (1994) Structural analysis of a-satellite DNA and centromere proteins using extended chromatin and chromosomes. Hum. Mol. Gen., 3, 697-709. Ikeno. M.. Masumoto. H. and Okazaki, T. (1994) Distribution of CENP-B boxes reflected in CREST centromere antigenic sites on long-range a-satellite DNA arrays of human chromosome 21. Hum. Mol. Genet., 3, 1245-1257. Jabs, E.W., Wolf, S.F. and Midgeon, B.R., (1984) Characterisation of a cloned DNA sequence that is present at the centromeres of all human autosomes and the X chromosome shows polymorphic variation. Proc. Nat]. Acad. Sci. USA, 81, 4884-4888. Jokelainen. P.T. (1967) The ultra structure and spatial organization of the metaphase kinetochore in mitotic rat cells. J. Ultrastruct. Res., 19, 19-44. Jones, K. (1970) Chromosomal and nuclear location of mouse satellite DNA in individual cells. Nature, 225. 912-915. Joseph, A., Mitchell, A.R. and Miller, O.J. (1989) The organisation of the mouse satellite DNA at centromeres. Exp. Cell Res.. 183, 494-500. Kipling, D., Mitchell, A.R., Masumoto, H., Wilson, E., Nicol, L. and Cooke, H.J. (1995) CENP-B binds a novel centromeric sequence in the Asian mouse Mus carob. Mol. Cell Biol., 15, 4009-4020. Kitagawa. K., Masumoto, H., Ikeda, M. and Okazaki, T. (1995) Analysis of protein-DNA and protein-protein interactions of centromere protein B (CENP-B) and properties of the DNACENP-B complex in the cell cycle. Mol. Cell. Biol., 15. 1602-1612. Mackay, A.M. and Earnshaw, W.C. (1993) The INCENPs: Structural and functional analysis of a family of chromosome passenger proteins. C.S.H. Sym. LVl 11, 697-706. Maio, J.J., Brown, F.I. and Musich, P.R. (1981) Towards a molecular paleontology of primate genomes. Chromosoma. 83, 103-125. Manuelidis, L. (1978) Chromosomal localisation of complex and simple repeated human DNAs. Chromosoma, 66, 23-32. Masumoto, H., Masukata, H.. Muro, Y., Noaki. N. and Okasaki, T. (1989) A human centromere antigen (CENP-B) interacts with a short sequence in alphoid DNA, a human centromeric satellite. J. Cell Biol., 109. 1963-1973. Miller, Dorothy A., Sharma, Vasundhara and Mitchell, A.R. (1988) A human-derived probe, p82H, hybridizes to the centromeres of gorilla, chimpanzee and orangutan. Chromosoma, 96, 270274. Mitchell, A.R., Gosden. J.R. and Miller. D.A. (1985) A cloned sequence, p82H, of the alphoid repeated DNA family found at the centromeres of all human chromosomes. Chromosoma, 92, 369-377. Mitchell, A.. Jeppesen. P., Hanrdtty, D. and Gosden, J. (1992) The organisation of repetitive DNA sequences on human chromosomes with respect to the kinetochore analysed using a combination of oligonucleotide primers and CREST anticentromere serum. Chromosoma, 101. 333-341.
162
A. R. Mit~hrll/M~~tntion
Mitchell. A.R., Nicol, L., Malloy, P. and Kipling, D. (1993) Novel structural organisation of a Mua musculus DBA/Z chromosome shows a fixed position for the centromere. J. Cell Sci.. 106, 79-85. Moroi, Y., Peebles, C., Fritzler, M.J., Steigerwald, J. and Tan, E.M. (1980) Autoantibody to centromere (kinetochore) in scleroderma sera. Proc. Natl. Acad. Sci., USA 77. 1627- 163 I. Murphy, Terence D. and Karpen. Gary H. (1995) Localisation of centromere function in a Drosophila minichromosome. Cell, 82, 599-609. Page, S.C., Eamshaw. W.C.. Choo. K.H. and Shaffer, Lisa G. (1995) Further evidence that CENP-C is a necessary component of active centromeres: studies of a dic(X;lS) with simultaneous immunofluorescence and FISH. Hum. Mol. Gen., 4. 289-294. Palmer, D.K. and Margolis. R.L. (1985) Kinetochore components recognised by human autoantibodies are present on mononucleosomes. Mol. Cell Biol., 5, 173- 186. Palmer. D.K., O’Day, K., Wener. M.H. andrews, B.S. and Margolis. R.L. (1987) A 17-kD centromeric protein (CENP-A) copurifies with nucleosome core particles and with histones. J. Cell Biol.. 104, 805-815. Pardue, M.L. and Gall, J.G. (1969) Chromosomal localisation of mouse satellite DNA. Science, 168, 1356- 1358. Pietras. D.F., Bennet, K.L., Siracusa, L.D., Woodworth-Gutai. M.. Chapman V.M., Gross. K.W., Kane-Hass. C. and Hastie, N.D. (1983) Construction of a small Mus musculus repetitive DNA library: identification of a new satellite sequence in Mus musculus. Nucleic Acids Res.. I I, 6965-6983. Pluta, A.F., Cooke, C.A. and Earnshaw, W.C. (1990) Structure of the human centromere at metaphase. Trends Biochem. Sci.. 15, 181-185. Rattner, J.B., Rao, A., Fritzler, M.J., Valencia, D.W. and Yen. T.J. (1993) CENP-F ia a ca. 400kDa kinetochore protein that exhibits a cell-cycle dependent localization. Cell Motil. Cytoskel., 26, 214-226. Ris, H. and Witt. P.L. (1981) Structure of the mammalian kinetochore. Chromosoma, 82, 153-167. Saitoh, H., Tomkiel. J., Cooke, C.A., Ratrie, H.. III, Maurer, M.. Rothfield, N.F. and Eamshaw. W.C. (1992) CENP-C, an autoantigen in scleroderma is a component of human inner kinetochore plate. Ceil, 70, 115-125. Sharp, L.W. (1934) Introduction to cytology. New York: McGraw-Hill 1934. Sugimoto. K.. Yata, H., Muro, Y. and Himeno. M. (1994) Human centromere protein C (CENP-C) is a DNA-binding protein which possesses a novel DNA-binding motif. J. Biochem.. 116. 877-881.
Resecmh
372 1 I9961 153-162
Sullivan, B.A. and Schwartz. S. (1995) Identification of centromeric antigens in dicentric Robertsonian translocatons: CENP-C and CENP-E are necessary components of functional centromeres. Hum. Mol. Gen.. 4, 2 189-2 197. Sullivan, K.F., Hechenberger. M. and Masri. K. (1994) Human CENP-A contains a histone H3 related histone fold domain that is required for targeting to the centromere. J. Cell Biol., 127. 581-592. Thrower. D.A., Jordan, Mary A.. Schaar. B.T., Yen, T.J. and Wilson, L. (1995) EMBO J., 14. 918-926. Tyler-Smith. C., Oakley. Rebecca J.. Larin. Zoia., Fisher, R.B., Cracker. M.. Affara. Nabeel A.. Ferguson-Smith. M.A., Muenke, M., Zuffardi. Orsetta and Jobling, M.A. (1993) Localisation of DNA sequnces required for human centromere function through an analysis of rearranged Y chromosomes. Nature Genet., 5, 368-375. Vig, B.K. and Richards, B.T. (1992) Formation of primary constriction and heterochromatin in mouse does not require minor satellite DNA. J. Exp. Res.. 201, 292-298. Vig, B.K., Latour, D. and Frankovich, Jennifer (1994) Dissociation of minor satellite from the centromere in mouse. J. Cell Sci., 107, 3091-3095. Willard, H.F.. Smith, K.D. and Sutherland J. (I 983) Isolation and characterisation of a major tandem repeat family from the human X chromosome. Nucleic Acids Res.. I1 2017-2033. Wong. A.K.C. and Rattner. J.B. (I 988) Sequence organisation and cytological organisation of the minor satellite of mouse. Nucleic Acids Res.. 16. 11645-l 1661. Yang, T.P.. Hansen S.K.. Oishi K.K., Ryder, O.R. and Hamkalo, B.A. (1982) Characterisation of a cloned repetitive DNA sequnce concentrated on the human X chromosome. Proc. Natl. Acad. Sci. USA. 79. 6593-6597. Yen, T.J., Compton. D.A.. Wise, D., Zinkowski. R.P., Brinkley. B.R., Earnshaw. W.C. and Cleveland D.W. (1991) CENP-E a novel human centromere-associated protein required for progression from metaphase to anaphase. EMBO J., 10, 12451254. Yen, T.J., Li, G., Schaar. B.T., Szilak, I. and Cleveland, D.W. (1992) CENP-E is a putative kinetochore motor that accumulates just before mitosis. Nature. 359, 536-539. Yoda, K., Kitagawa. K.. Masumoto, H.. Muro. Y. and Okazaki, T. (1992) A human centromere protein, CENP-B, has a DNA binding domain containing four potential a-helices at the NH2 terminus which is separable from dimerising activity. J. Cell Biol., 119, 1413-1327. Zinkowskl, R.P., Meyne, J. and Brinkley, B.R. (1991) The centromere-kinetochore complex: a repeat cubunit model’? J. Cell Biol., 113, 1091-I 1 IO.
ELSEVIER
Mutation
Research
372 (1996)
and Molecular of Mutagenesis
153-167
The mammalian centromere: Its molecular architecture Arthur R. Mitchell
*
Abstract The DNA and protein composition of the centromeric domains in mammalian chromosomes is now relatively well characterised. The major families of repeated DNAs, i.e., the simple-sequence and alphoids in man and the satellite sequences (both minor and major) in the mouse have been sequenced and long-range maps using pulse-field gels of some centromeres have been carried out. Autoimmune antibodies have provided an insight into some of the proteins which interact with these DNA sequences. Although the individual components of the mammalian centromere may have been identified, how they interact with each other to give the functional structure visualised by electron microscopy is yet to be determined. This review examines our understanding of these separate components. Keyords;
Mammalian:
Centromere:
Centromere
architecture
1. Introduction In mammalian chromosomesthe position of the centromere has been recognised for many years by cytologists as a physical constriction or narrowing of the chromatin at a specific place along the chromosome. Because of this narrowing feature the centromere is referred to as the primary constriction of the chromosome.This position is of particular importance since it is thought to be the site of attachment of the tubulin microtubules formed during cell division. This leads to the correct alignment of the chromosomeson the metaphaseplate at cell division which in turn gives rise to the correct segregationof chromosomesto each daughter cell at anaphase.Loss of centromere function is normally associatedwith chromosomeinstability leading to chromosomeloss
Corresponding author. Tel.: 2620: E-mail: [email protected] 0027-S 107/96/$lS.O0 PII
SOO27-5107~96~00135-2
Copyright
0131 467 8319;
6 1996 Elsevier
Fax: 0131 343
in subsequentcell divisions. This article will examine our understandingof the molecular nature of the mammaliancentromere at both the protein and DNA level. Most structural evidence in mammalshasbeen derived from studies on the chromosomesof man and mouse. 1 will also attempt to see if our understanding of the structural components of a mammalian centromere has led to an insight into a functional role for one or more of the molecular components. Before beginning it would be wise to define the terminology of the centromere. The terms ‘ kinetochore’ and ‘centromere’ were originally coined to describe the spindle attachment region on chromosomes(Sharp. 1934: Darlington, 1937). Later by using the electron microscopeit was shown that a trilaminar plate or disc was present on somechromosomes at this site (Jokelainen. 1967; Comings and Akada, 197I) and the term ‘kinetochore’ was used by some to describe this visible structure. In this
Science B.V. All rights rewrvrd
present article I will use ‘centromere’ as the attachment site. The term ‘centromeric domain’ will be used to define the chromatin immediately surrounding and underlying the centromere. This centromeric chromatin can be recognised at both the cytological and molecular level by being composed of specific well-characterised DNA sequences. ‘Kinetochore domain’ (Eamshaw, 1991) will be used solely to describe those DNA sequences that have been shown to lie directly beneath a visible trilaminar structure. The kinetochore domain will therefore be a specific region within the centromeric domain of the chromosome.
bask repr
unii
Fig. I. A model showing the amplification of the basic 170 bp repeat unit of alphoid DNA sequences leading to the formation of large chromosome specific arrays at the centromeres of human chromosomes.
2. The DNA content Although it was known from kinetic studies that the genome of eukaryotes contained DNA sequences of different complexity and copy number, it was not until Pardue and Gall (1969) and Jones (1970) localised mouse major satellite DNA to the centromeric domain of mouse chromosomes that it became possible to correlate cytological landmarks on the chromosome with specific repetitive DNA families. Rapid developments using the technique of in situ hybridisation followed on from these initial studies. Thus the simple-sequence satellites from man (i.e., those DNAs isolated from the bulk of genomic DNA using density gradient centrifugation) were mapped to the centromeric domains of certain human chromosomes (Gosden et al.. 1975). The use of restriction endonucleases led to the identification of other repetitive DNAs such as the SAU3A (Agresti et al., 1987) and HAE3 (Agresti et al., 1989) families. One such family, the alphoid DNA family originally described by Maio et al. (1981). was particularly interesting since in man chromosome specific subsets were isolated using different restriction enzymes. Manuelidis (1978) using EcoRI identified a member of this family which hybridised predominantly to chromosomes I, 3, 7. IO and 19. Willard et al. (1983) and Yang et al. (1982) found an X specific member of the alphoid DNA family. Jabs et al. (1984) isolated a subset using XbaI which hybridised to the centromeric domains of human chromosomes 2. 4, 14, 15. 18, 20 and 22. The alphoid
repeated DNA family is comprised of a short monomeric unit of - 170 bp. This has undergone multiple rounds of amplification so today it represents somewhere around 5% of the human genome, i.e., 3 X lo* bp. During this process mutations are incorporated giving rise to different but related subsets of sequences (Fig. I). It is this process which gives rise to the subsets of alphoid arrays which are chromosome specific. Mitchell et al. (1985) first showed that alphoid DNAs were present at the centromeric domains of all chromosomes in man. The p82H sequence (Mitchell et al., 1985) was heterogeneous in sequence with adjacent monomeric mismatches varying from 8-26s. In all probability it was this high degree of sequence mismatch which allowed the p82H clone to hybridise to the alphoid sequences on all human chromosomes. Thus for the first time a family of repeated DNAs were associated with the centromeric domains of all chromosomes in the human genome. In comparison, mouse major satellite was not detectable on the Y chromosome (Jones, 1970). The position of the simple-sequence satellites of man. as determined by in situ studies, positioned them further from the primary constriction than alphoid sequences. Sequences homologous to the human alphoid DNAs were also detected in the higher primates (Miller et al.. 1988; Baldini et al., 1991) suggestive of some degree of evolutionary conservation. Pietras et al. (1983) had previously cloned, using DNA reassociation kinetics as the tool, a second
repeated DNA sequence from the mouse genome. In comparison to the major satellite in the mouse (which constitutes around 10% of the genome) this satellite represented less than 1% of the genome with a consequence that it was known as the mouse minor satellite. The monomeric repeat of this satellite was 120 bp. Initially in situ hybridisation suggested that it was present on only some of the chromosomes in the mouse but that it was centromerically located (Pietras et al.. 19831. Later Wong and Rattner (19881 and Joseph et al. (19891 showed that this DNA family was present on all chromosomes except the Y chromosome in mouse. Both the human alphoid and mouse minor satellite DNAs shared a common feature which was that they were situated closer to the primary constriction of chromosomes than either the human simple sequence DNAs or mouse major satellite DNA (Joseph et al., 1989; Mitchell et al., 19921. The centromeric domain of chromosomes in man and mouse appeared therefore to be comprised of repetitive DNAs. At the structural level as shown by in situ hybridisation a similar type of organisation was present within the centromeric domains of both species. However, at the DNA sequence level little homology between these two DNA t’dmilies could be recognised. On the basis of this it was difficult to establish a functional relationship between these two repeated DNAs. It was known that somatic cell hybrids between mouse and human cells could be successfully maintained implying a common feature within the centromeres of the chromosomes in both species for spindle attachment at cell division.
3. The protein content The serum from some patients suffering from the autoimmune disease CREST contains autoimmune antibodies which react specifically with centromeric proteins (Moroi et al.. 19801. CREST ACA (anticentromere autoantibody) staining of chromosomes produces a double dot signal at the primary constriction in all mammalian species tested. Electron microscopy has mapped the CREST antigens close to the centromere (Brenner et al., 198 1). A combination of CREST ACA staining with PRINS in situ hybridisation (Mitchell et al., 1992) confirmed the earlier observations that the alphoid repeats were physically
Fig. 2. PRINS in situ hybridisation using (a) a human alphoid wquence in combination with CREST staining (b). Note that the in situ signal of the alphoid sequence and the CREST signal co-localise to the same position in human centromeres. The arrow points to a chromosome 9. In (c) the chromosomes are stained with DAPl.
Fig. 3. This shows PRINS in situ hybridisation using a human simple satellite DNA sequence (a). The CREST signal is again shown in (b) and, the DAPI stain is shown in (cl. kote that the in Gtu signal for the simple-cequence DNA and CREST do not co-localise to the same position within the centromeric domain on human chromosome 9.
close to the centromere (Fig. 2). Simple-sequence DNAs by comparison were mapped some distance from the centromere (as defined by CREST ACA indirect immunofluorescence) (Fig. 3). The major protein antigens recognised by patients with CREST anticentromere antibodies (ACA) have been identified. Three proteins termed CENP-A, CENP-B and CENP-C (Earnshaw and Rothfield, 1985) with molecular weights of 17, 80 and 140 kDa have been studied in detail. More recently two other centromere proteins CENP-E (312 kDa); Yen et al., 1991, 1992) and (CENP-F; (400 kDa); Rattner et al., 1993) have also been identified (CENP stands for centromere protein). A related group of proteins have been termed the itmer ccntromere proteins (INCENPS) (Mackay andzarnshaw, 1993). CENP-E and CENP-F are members of this group. They are sometimes called passenger proteins since they associate with chromosomes at specific stages during cell division. During interphase they may be associated with the nuclear cytoskeleton (Mackay and Earnshaw, 1993). Some like CENP-E are motor proteins related to the kinesins whose role is to promote microtubule and thus chromosome movement during mitosis (Thrower et al., 1995). CENP-A is a histone like molecule (Palmer and Margolis. 1985; Palmer et al., 1987) related most closely to histone H3 (Sullivan et al., 1994). Full length cDNA of CENP-A shows the sequence of this protein to be 62% identical to histone H3. CENP-A is found in nucleosomes and one possible role may be to replace the H3 molecule specifically in those nucleosomes which are directly associated with the centromere itself. Why the centromere should require nucleosomes of different histone octamer content from the rest of the DNA is not yet clear. One hypothesis is that CENP-A may provide a structure which can withstand the stronger stress forces present during the process of cell division. CENP-C was found to be a highly basic protein of 943 amino acids (Saitoh et al., 1992). Some conflicting data on its properties exist. Saitoh et al. (1992) found no evidence for any DNA binding motifs whereas Sugimoto et al. (1994) reported that a specific DNA binding domain of 101 amino acids is present by expressing truncated *in frame deletions’ of the protein in E. co/i. Although not resolved, there is little doubt that this protein is part of the kinetochore structure. Saitoh et al. (1992) using im-
157
munoelectron microscopy localised the protein to the inner plate of the kinetochore. In terms of understanding the interactions of the CENPs with centromeric DNAs then CENP-B has provided the most insight. Masumoto et al. (1989) demonstrated that human alphoid DNA could be immunoprecipitated using CREST ACA serum. DNasel footprinting showed that a 17 bp motif (CTTCGTTGGAAACGGGA) within some alphoid monomers directly bound the CENP-B protein. This 17 bp motif was called the CENP-B box. Database searches showed a similar CENP-B box (ATTCGTTGGAAACGGGA) to be present in some of the mouse minor satellite sequences. Thus a direct link was established between centromerically located DNAs and the CENP-B protein. This finding also implied that both DNA sequences may carry out similar functions at the centromere of chromosomes, a conclusion which could not be reached from DNA sequence analysis alone nor from the implied functional role indirectly suggested by the in situ results which were described earlier. Further analysis of the interactions between the alphoid DNA and the CENP-B protein suggested that two alphoid monomers containing the CENP-B motif interact with the protein giving rise to a stable dimer structure maintained by protein-protein hydrophobic bonds (Yoda et al.. 1992; Kitagawa et al., 1995). The DNA binding domain was localised to the amino end whereas the protein:protein binding domain resided at the carboxyl end of the CENP-B molecule. A function for the CENP-B protein was predicted from this result which showed that one possible role for this protein within the chromatin containing the alphoid DNAs was to cause condensation of the long arrays (see Fig. 4). Because of the repetitive nature of the alphoid DNA and the fact that many CENP-B binding sites would be present, then this could result in the formation of higher ordered structures of this chromatin. Some support for this model came from the finding that the distribution of CENP-B binding sites was not uniform throughout the alphoid arrays on human chromosomes. One example is human chromosome 21 where two separate blocks of alphoid arrays are found. One array (o 21- 1) comprising around 1.3 Mb of DNA had a CENP-B binding site every alternate monomer. The second array (o 2 1- 11I was composed of more divergent alphoid monomers
Fig. 4. Possible role for the CENP-B protein in chromatin condensation. Alphold monomers with CENP-B motifs are \hown in black. The CENP-B protein (open circles) causes the long alphoid arrays to package into tight structures.
with less frequent binding sites (1 site per 100 alphoid monomers> for the protein (Ikeno et al., 1994). Furthermore in situ hybridisation in combination with indirect immunofluorescence with CREST ACA serum suggested that the cx21- 1 chromatin appeared more compact than the (Y21- 11 chromatin in support of the proposed model. Binding sites for the CENP-B protein are also present within the monomers of mouse minor satellite DNA and the assumption is that the CENP-B protein complexes to minor satellite monomers in a manner similar to that found for alphoid DNAs. Interestingly, Kipling et al. (1995) showed that in the mouse species MU ca~li, which lacks DNA sequences homologous to the Mus mz4.sc~ulus minor satellite, functional CENP-B binding sites are present within the centromerically located repeated DNAs in this animal. The CENP-B binding site in the Mus caroli satellite is diverged from the sequence found in minor satellite DNA although it binds human CENP-B protein. Nine bases in the CENP-B binding site have been conserved and appear absolute requirements for binding to occur. Thus the diverged sequence CENP-B box has a xTTCGxxxxAxxCGGGx. It appears that any base can be substituted for x. Nuclear proteins from Drosophila form complexes in gel retardation assays
with alphoid monomers containing CENP-B motifs. When challenged with CENP-B oligos, the complex is dissociated suggesting that the Drosophila proteins are homologs of the human CENP-B protein (Avides and Sunkel, 1994). Evolutionary conservation of both binding site and the possibility of homologous proteins in other organisms strongly argues in favour of the CENP-B protein having some common functional role in centromeric chromatin structure. Although the above evidence favours a role for the CENP-B protein in centromeric chromatin structure, it is far from clear if this protein is required for centromere function. At the EM level this protein underlies the kinetochore in the kinetochore domain chromatin (Cooke et al., 1990; Pluta et al.. 1990) and does not appear to be part of the proteinaceous trilaminar structure of the kinetochore plate itself. implying more a structural or supportive role for these DNA sequences in kinetochore domain chromatin. Studies on dicentric chromosomes have shown quite clearly that CENP-C is always associated with the active centromere and never with the inactive centromere, whereas CENP-B is found at both the active and inactive centromere in the dicentric chromosomes (Eamshaw et al., 1989; Page et al., 1995; Sullivan and Schwartz, 1995). Similarly, in the mouse examples have been reported of stable chromosomes without any detectable minor satellite DNA sequences and hence no CENP-B binding sites (Vig et al., 1994: Vig and Richards, 1992) and. at the molecular level, the alphoid sequences on the human Y chromosome have been extensively sequenced with no evidence that CENP-B binding motifs are present. Whether all available CENP-B binding sites are in fact used for binding the protein seems in doubt from studies on the DBA/2 mouse. In this animal chromosome I is polymorphic. containing two separate blocks of minor satellite DNA sequences (Mitchell et al., 1993). Although each block of minor satellite arrays contain CENP-B binding sites, the centromere is always associated with the more terminal block never with the internal array of minor satellite sequences (Fig. 5a). Thus even where CENP-B binding
sites are available it would appear that not all are utilised to bind the CENP-B protein. Although this figure shows the position of the centromere using CREST ACA serum. antibodies raised against the human CENP-B protein give a similar result. Because the CENP-B motif contains CpG dinucleotides we proposed (Mitchell et al., 1993) that methylation of cytosines within the arrays of minor satellite on chromosome 1 could have a role in determining the position of the centromere. When the minor satellite sequences on this chromosome are demethylated a redistribution of the CENP-B protein is seen (Mitchell et al.. in press) resulting in a labelling of both blocks of minor satellite sequences (Fig. 5b). Our interpretation of this result is that only some CENP-B motifs remain unmethylated in chromatin and hence capable of binding the protein. The majority of sites are methylated at the CpG dinucleotide and are incapable of binding the protein. In vitro methylation experiments (Masumoto, pers commun.) support our conclusions. Interestingly, making more sites within chromatin accessible to the CENP-B protein does not affect the position of the centromere nor does it apparently interfere with chromosome segregation. It has to be borne in mind that these experiments were carried out using short-term cultures and it remains to be seen whether demethylation of CENP-B binding sites has longer term effects on the stability of chromosomes.
4. The size of the mammalian
centromere
Much of the discussion above has centred on the role of the human alphoid and mouse minor satellite repeats in the formation of an active centromere. Both DNAs occur as tandem arrays which would tend to suggest that if they are indeed involved in centromere function then their interaction with centromeric proteins, whether the CENPs or other proteins which are yet to be identified, might predictably lead to a structure which is repetitive in
Fig. 5. (a) DBA/Z chromosome I (at-rowed) contains two blocks of mmor satellite arrays (red). The CREST signal (green) is always associated wtith the terminal block. (b) Redistribution of CENP-B signal on DBA/2 chromosome I when minor satellite DNA sequences are demethylated. (A) combined image of minor satellite in situ hybridisation (red) and CENP-B localisation (green). Blue is the DAPI stained chromosomes. The split images are shown in: (B) CENP-B signal (C) in situ minor satellite signal and (D) DAPI stained chromosomes.
160
AR. Mitchell/Mutation
nature. Zinkowski et al. (199 1) did indeed detect such a repeat-like structure at the centromeres in Indian muntjac chromosomes. Using physical methods such as a long hypotonic treatment followed by centrifugation onto glass slides centromeres on occasions were physically stretched. CREST ACA staining in these situations gave rise to beaded structures interpreted to be from a centromere whose subunits had been separated during this procedure. In situ hybridisation of human alphoid DNAs to extended chromatin fibres also suggest that the CREST ACA bind bead-like to some but not all of the chromatin containing the alphoid DNA (Haaf and Ward, 1994). Although CENP-C antibody normally gives a double dot signal on human centromeres, when it is applied to centromeres from cells with extended chromatin fibres 6-8 smaller signals are seen (Mitchell unpublished). This result would also suggest that the centromere consists of a structure of a repetitive nature. The mammalian kinetochore as seen under the EM (Ris and Witt, 1981) could be interpreted as a proteinaceous structure (i.e., the inner and outer plates) into which DNA sequences (alphoids? or mouse minor satellite DNA?) intercalate, binding some of the centromere proteins to give a stable structure capable of binding other proteins through protein:protein interactions similar to those seen between CENP-B molecules. The outer proteins in this complex are responsible for capturing microtubules at cell division. The number of microtubules required for this process is somewhat of a guessing game with figures of 20-25 separate tubules being required for attachment to any one kinetochore in human chromosomes. The most compelling evidence in favour of the alphoid sequences comprising the DNA content of the human centromere comes from work on the human Y chromosome (Tyler-Smith et al., 1993; Brown et al., 1994). Tyler-Smith et al. (1993) looked at a series of Y chromosomes which had undergone rearrangements and characterised both the stability of the rearranged chromosomes and analysed their DNA content by pulse-field gel in order to pinpoint which sequence was responsible for chromosome stability and hence centromere activity. One such chromosome was found to have a deletion within an alphoid array and the resulting phenotype was chromosome instability. Their data ruled out other sequences within this array which strongly pointed to
RrJmrch
372 (19%)
153-162
this alphoid array having a functional role in centromere activity. Size estimates for the centromere on the Y chromosome came up with a figure 500 kb. Of this they estimate that 300 kb is composed of alphoid sequences. Brown et al. (1994) used the technique of ’ telomere associated chromosome fragmentation’ (Bamett et al., 1993) to develop deletion of Y chromosomes. One, bYq74, appeared mitotitally stable with 140 kb of alphoid array within its centromere. A second construct, AYp 134, although having 540 kb of alphoid DNA was not stable. If only alphoid DNA is required for centromere activity then the 140 kb of alphoid sequence in AYq74 is sufficient. However, it cannot be ruled out that some other DNA sequence essential for centromere activity remains embedded in this array. This sequence if present would have to be nonrepetitive in nature. In Drosophila a minimum centromeric core size appears to be around 240 kb of DNA. This essential core element termed ‘ bora bora’ is a complex consisting not only of repetitive DNA elements but also middle repetitive and single-copy DNA sequences quite unlike the model envisaged for the human Y centromere described above (Murphy and Karpen, 1995). In contrast to both of the above, Garagna et al. ( 1995) found stable Robertsonian chromosomes in the mouse where only SO kb of minor satellite at the centromere remained after fusion. It is clear that we have some way to go before the definitive DNA sequence which is indispensible for centromere function is found. I leave you with the thought that there may not necessarily be one DNA sequence which confers this property. If a centromere is formed by the deposition of specific proteins then the important factor may simply be that of the DNA molecule adopting the correct 3-dimensional shape for a protein to recognise. Once this initial complex is formed then the rest of the reactions, like many cascade mechanisms in enzyme kinetics, is determined more by the energetics of the process than by sequence composition alone. Thus each chromosome may have developed its own solution to its problem of survival in the cell. Acknowledgements I would like to thank my colleagues Drs Peter Jeppesen and David Kipling for their helpful advice
A.R.
Mitchrll/Mutution
in preparing this paper. The photographic department of the Unit is also thanked for professional help.
References Agresti, A., Meneveri. R., Siccardi. A.G.. Marozzi, A., Comeo, G.. Guadi, S. and Ginelli. E. (1989) Linkage in human heterochromatin between highly divergent Sau 3A repeats and a new family of repeated DNA sequences (HAE 3 family). J. Mol. Biol., 205, 625-63 I. Agresti. A., Rainabli, G., Lobbiani, A., Magnani, I., Dilernia, R.. Meneveri, R.. Siccardi. A.G. and Ginelli, E. (1987) Chromosomal location by in situ hybridisation of the human sau3A family of DNA repeats. Hum. Genet., 75, 326-332. Avides. Maria C. and Sunkel, Claudia E. (1994) Isolation of chromosome-associated proteins from Drosophila mekdno~asrrr that bind a human centromere DNA sequence. .I. Cell Biol.. 127, 1159-i 171. Baldini, A., Miller, Dorothy A.. Miller. O.J., Ryder, O.A. and Mitchell A.R. (I 99 1J A chimpanzee-derived chromosomespecific alpha satellite DNA sequnce conserved between chimpanzee and human. Chromosoma, 100, 156-162. Barnett, M.A., Buckle, V.J.. Evans, E.P.. Porter. A.C.G., Rout, D., Smith. A.G. and Brown. W.R.A. (1993) Telomere directed fragmentation of mammalian chromosomes. Nucleic Acids Res.. 2 I( 27-36. Brenner, S., Pepper, D.. Berns, M.W., Tan, E. and Brinkley, B.R. (198 1I Kinetochore structure. duplication, and distribution in mammalian calls: analysis by human autoantibodies from scleroderma patients. J. Cell Biol.. 91. 95- 102. Brown. K.E.. Barnett. M.A.. Burgtorf, C., Shaw, P., Buckle, V.J. and Brown, W.R.A. (1994) Dissecting the centromere of the human Y chromosome with cloned telomeric DNA. Hum. Mol. Cen.. 3. 1227-1237. Comings, D.E. and Akada. T.A. (1971) Fine structure of kinetochore in Indian Muntjac. Exp. Cell Res.. 67, 97-110. Cooke, C.A.. Bernat, R.L. and Earnshaw, W.C. (I 990) CENP-B: a major human centromere protein located beneath the kinetochore. J. Cell Biol.. 110. 1475-1488. Darlington, CD. (1937) Recent advances in cytology. Philadelphia: Blakiston. Eamshaw, W.C. (1991) When is a centromere not a kinetochore? J. Cell Sci., 99, 1-4. Earnshaw. W.C. and Rothfield, N. (1985) Identification of a family of human centromere proteins using autoimmune sera from patients with scleroderma. Chromosoma. 91, 313-321. Earnshaw, W.C., Ratrie. H. and Stetten, G. (1989) Visualisation of centromere proteins CENP-B and CENP-C on a stable dicentric chromosome in cytological spreads. Chromosoma, 98, I-12. Garagna, S.. Broccoli. D.. Redi, C.A., Searle. J.B., Cooke. H.J. and Capanna, E. (1995) Robertsonian metacentrics of the house mouse lose telomeric sequences but retain some minor satellite DNA in the pericentromeric region. Chromosoma, 103, 685-692.
Reseurch
372
~IYYfil
153-162
161
Gosden, J.R., Mitchell, A.R., Buckland, R.A., Clayton, R.P. and Evans H.J. (1975) The location of four human satellite DNAs on human chromosomes. Exp. Cell Res.. 92. 148-152. Haaf. T. and Ward. D.C. (1994) Structural analysis of a-satellite DNA and centromere proteins using extended chromatin and chromosomes. Hum. Mol. Gen., 3, 697-709. Ikeno. M.. Masumoto. H. and Okazaki, T. (1994) Distribution of CENP-B boxes reflected in CREST centromere antigenic sites on long-range a-satellite DNA arrays of human chromosome 21. Hum. Mol. Genet., 3, 1245-1257. Jabs, E.W., Wolf, S.F. and Midgeon, B.R., (1984) Characterisation of a cloned DNA sequence that is present at the centromeres of all human autosomes and the X chromosome shows polymorphic variation. Proc. Nat]. Acad. Sci. USA, 81, 4884-4888. Jokelainen. P.T. (1967) The ultra structure and spatial organization of the metaphase kinetochore in mitotic rat cells. J. Ultrastruct. Res., 19, 19-44. Jones, K. (1970) Chromosomal and nuclear location of mouse satellite DNA in individual cells. Nature, 225. 912-915. Joseph, A., Mitchell, A.R. and Miller, O.J. (1989) The organisation of the mouse satellite DNA at centromeres. Exp. Cell Res.. 183, 494-500. Kipling, D., Mitchell, A.R., Masumoto, H., Wilson, E., Nicol, L. and Cooke, H.J. (1995) CENP-B binds a novel centromeric sequence in the Asian mouse Mus carob. Mol. Cell Biol., 15, 4009-4020. Kitagawa. K., Masumoto, H., Ikeda, M. and Okazaki, T. (1995) Analysis of protein-DNA and protein-protein interactions of centromere protein B (CENP-B) and properties of the DNACENP-B complex in the cell cycle. Mol. Cell. Biol., 15. 1602-1612. Mackay, A.M. and Earnshaw, W.C. (1993) The INCENPs: Structural and functional analysis of a family of chromosome passenger proteins. C.S.H. Sym. LVl 11, 697-706. Maio, J.J., Brown, F.I. and Musich, P.R. (1981) Towards a molecular paleontology of primate genomes. Chromosoma. 83, 103-125. Manuelidis, L. (1978) Chromosomal localisation of complex and simple repeated human DNAs. Chromosoma, 66, 23-32. Masumoto, H., Masukata, H.. Muro, Y., Noaki. N. and Okasaki, T. (1989) A human centromere antigen (CENP-B) interacts with a short sequence in alphoid DNA, a human centromeric satellite. J. Cell Biol., 109. 1963-1973. Miller, Dorothy A., Sharma, Vasundhara and Mitchell, A.R. (1988) A human-derived probe, p82H, hybridizes to the centromeres of gorilla, chimpanzee and orangutan. Chromosoma, 96, 270274. Mitchell, A.R., Gosden. J.R. and Miller. D.A. (1985) A cloned sequence, p82H, of the alphoid repeated DNA family found at the centromeres of all human chromosomes. Chromosoma, 92, 369-377. Mitchell, A.. Jeppesen. P., Hanrdtty, D. and Gosden, J. (1992) The organisation of repetitive DNA sequences on human chromosomes with respect to the kinetochore analysed using a combination of oligonucleotide primers and CREST anticentromere serum. Chromosoma, 101. 333-341.
162
A. R. Mit~hrll/M~~tntion
Mitchell. A.R., Nicol, L., Malloy, P. and Kipling, D. (1993) Novel structural organisation of a Mua musculus DBA/Z chromosome shows a fixed position for the centromere. J. Cell Sci.. 106, 79-85. Moroi, Y., Peebles, C., Fritzler, M.J., Steigerwald, J. and Tan, E.M. (1980) Autoantibody to centromere (kinetochore) in scleroderma sera. Proc. Natl. Acad. Sci., USA 77. 1627- 163 I. Murphy, Terence D. and Karpen. Gary H. (1995) Localisation of centromere function in a Drosophila minichromosome. Cell, 82, 599-609. Page, S.C., Eamshaw. W.C.. Choo. K.H. and Shaffer, Lisa G. (1995) Further evidence that CENP-C is a necessary component of active centromeres: studies of a dic(X;lS) with simultaneous immunofluorescence and FISH. Hum. Mol. Gen., 4. 289-294. Palmer, D.K. and Margolis. R.L. (1985) Kinetochore components recognised by human autoantibodies are present on mononucleosomes. Mol. Cell Biol., 5, 173- 186. Palmer. D.K., O’Day, K., Wener. M.H. andrews, B.S. and Margolis. R.L. (1987) A 17-kD centromeric protein (CENP-A) copurifies with nucleosome core particles and with histones. J. Cell Biol.. 104, 805-815. Pardue, M.L. and Gall, J.G. (1969) Chromosomal localisation of mouse satellite DNA. Science, 168, 1356- 1358. Pietras. D.F., Bennet, K.L., Siracusa, L.D., Woodworth-Gutai. M.. Chapman V.M., Gross. K.W., Kane-Hass. C. and Hastie, N.D. (1983) Construction of a small Mus musculus repetitive DNA library: identification of a new satellite sequence in Mus musculus. Nucleic Acids Res.. I I, 6965-6983. Pluta, A.F., Cooke, C.A. and Earnshaw, W.C. (1990) Structure of the human centromere at metaphase. Trends Biochem. Sci.. 15, 181-185. Rattner, J.B., Rao, A., Fritzler, M.J., Valencia, D.W. and Yen. T.J. (1993) CENP-F ia a ca. 400kDa kinetochore protein that exhibits a cell-cycle dependent localization. Cell Motil. Cytoskel., 26, 214-226. Ris, H. and Witt. P.L. (1981) Structure of the mammalian kinetochore. Chromosoma, 82, 153-167. Saitoh, H., Tomkiel. J., Cooke, C.A., Ratrie, H.. III, Maurer, M.. Rothfield, N.F. and Eamshaw. W.C. (1992) CENP-C, an autoantigen in scleroderma is a component of human inner kinetochore plate. Ceil, 70, 115-125. Sharp, L.W. (1934) Introduction to cytology. New York: McGraw-Hill 1934. Sugimoto. K.. Yata, H., Muro, Y. and Himeno. M. (1994) Human centromere protein C (CENP-C) is a DNA-binding protein which possesses a novel DNA-binding motif. J. Biochem.. 116. 877-881.
Resecmh
372 1 I9961 153-162
Sullivan, B.A. and Schwartz. S. (1995) Identification of centromeric antigens in dicentric Robertsonian translocatons: CENP-C and CENP-E are necessary components of functional centromeres. Hum. Mol. Gen.. 4, 2 189-2 197. Sullivan, K.F., Hechenberger. M. and Masri. K. (1994) Human CENP-A contains a histone H3 related histone fold domain that is required for targeting to the centromere. J. Cell Biol., 127. 581-592. Thrower. D.A., Jordan, Mary A.. Schaar. B.T., Yen, T.J. and Wilson, L. (1995) EMBO J., 14. 918-926. Tyler-Smith. C., Oakley. Rebecca J.. Larin. Zoia., Fisher, R.B., Cracker. M.. Affara. Nabeel A.. Ferguson-Smith. M.A., Muenke, M., Zuffardi. Orsetta and Jobling, M.A. (1993) Localisation of DNA sequnces required for human centromere function through an analysis of rearranged Y chromosomes. Nature Genet., 5, 368-375. Vig, B.K. and Richards, B.T. (1992) Formation of primary constriction and heterochromatin in mouse does not require minor satellite DNA. J. Exp. Res.. 201, 292-298. Vig, B.K., Latour, D. and Frankovich, Jennifer (1994) Dissociation of minor satellite from the centromere in mouse. J. Cell Sci., 107, 3091-3095. Willard, H.F.. Smith, K.D. and Sutherland J. (I 983) Isolation and characterisation of a major tandem repeat family from the human X chromosome. Nucleic Acids Res.. I1 2017-2033. Wong. A.K.C. and Rattner. J.B. (I 988) Sequence organisation and cytological organisation of the minor satellite of mouse. Nucleic Acids Res.. 16. 11645-l 1661. Yang, T.P.. Hansen S.K.. Oishi K.K., Ryder, O.R. and Hamkalo, B.A. (1982) Characterisation of a cloned repetitive DNA sequnce concentrated on the human X chromosome. Proc. Natl. Acad. Sci. USA. 79. 6593-6597. Yen, T.J., Compton. D.A.. Wise, D., Zinkowski. R.P., Brinkley. B.R., Earnshaw. W.C. and Cleveland D.W. (1991) CENP-E a novel human centromere-associated protein required for progression from metaphase to anaphase. EMBO J., 10, 12451254. Yen, T.J., Li, G., Schaar. B.T., Szilak, I. and Cleveland, D.W. (1992) CENP-E is a putative kinetochore motor that accumulates just before mitosis. Nature. 359, 536-539. Yoda, K., Kitagawa. K.. Masumoto, H.. Muro. Y. and Okazaki, T. (1992) A human centromere protein, CENP-B, has a DNA binding domain containing four potential a-helices at the NH2 terminus which is separable from dimerising activity. J. Cell Biol., 119, 1413-1327. Zinkowskl, R.P., Meyne, J. and Brinkley, B.R. (1991) The centromere-kinetochore complex: a repeat cubunit model’? J. Cell Biol., 113, 1091-I 1 IO.