Multicentric chromosomes are often found in tumor cells and certain cell lines. How they are generated is not fully understood, though their stability suggeststhat they are non-functional during chromosome segregation. Growing evidence has implicated microtubule motor proteins in attachment of chromosomes to the mitotic spindle and in chromosome movement. To better understand the molecular basis for the inactivity of centromeres associated with secondary constrictions, we have tested these structures by immunofluorescence microscopy for the presence of motor complexes and associated proteins. We find strong immunoreactivity at the active, but not inactive, centromeres of prometaphase multicentric chromosomes using antibodies to the cytoplasmic dynein intermediate chains, three components of the dynactin complex (dynamitin, Arp1 and p150Glued), the kinesin-related proteins CENP-E and MCAK and the proposed structural and checkpoint proteins HZW10, CENP-F and Mad2p. These results offer new insight into the assembly and composition of both primary and secondary constrictions and provide a molecular basis for the apparent inactivity of the latter during chromosome segregation.
The origin and function of inactive, or accessory, centromeres in eukaryotic chromosomes is uncertain. These structures are observed most often in neoplastic cells and may originate from duplication events involving active centromere sequences (1). Consistent with this possibility, inactive centromeres are heterochromatic and contain centromere-specific DNA sequence motifs. They react strongly with CREST autoantisera (2,3) as well as specific antibodies raised against CENP-B, a DNA binding centromere protein (4,5).However, most inactive centromeres do not contain CENP-C (6) and do not show the trilaminar ultrastructure characteristic of functional kinetochores (7,8). Accessory centromeres do not appear to attach to microtubule ends and, like the surrounding chromatid arms, can be observed to separate prior to anaphase onset (9). These findings suggest that inactive centromeres do not participate in the process of chromosome segregation, a conclusion which is also supported by the long-term stability of these structures in cultured cell lines (10).
Work during the past decade has led to identification of a variety of microtubule motor proteins, several of which have been implicated in kinetochore function (reviewed in 11,12). The specific role of these motor proteins in kinetochore function continues to be an area of intense investigation which, to date, has focused primarily on cytoplasmic dynein and two kinesin-related proteins, centromere protein E (CENP-E) and mitotic centromere-associated kinesin (MCAK).
Cytoplasmic dynein is a large (>1 MDa) multi-subunit complex responsible for force production towards the minus ends of microtubules (13,14). It binds to another large multi-subunit complex, dynactin (15,16), which has been proposed to mediate the interaction of cytoplasmic dynein with subcellular structures (17-20). Cytoplasmic dynein (21,22) and dynactin (19) have both been found to accumulate at kinetochores, reaching a maximum level during prometaphase, a period during which rapid, poleward (minus end-directed) microtubule-based movements by newly captured chromosomes have been observed in favorable cell types (23). Dynein staining has been localized within the prometaphase kinetochore to the outermost region, the fibrous corona (24). Overexpression of the 50 kDa dynamitin subunit of dynactin has been found to dissociate the dynactin complex (19) and block mitosis in a prometaphase-like state. The arrested cells exhibit a decreased level of both dynactin and cytoplasmic dynein at kinetochores (19). Together these data support a role for cytoplasmic dynein and dynactin in kinetochore function during prometaphase.
CENP-E is a 312 kDa polypeptide which contains a kinesin-like motor domain located near the N-terminus (25,26). CENP-E has been observed to assemble onto kinetochores following nuclear envelope breakdown, reaching a maximum level during prometaphase, declining after metaphase and redistributing to the spindle midzone during late anaphase (reviewed in 11). It too, like cytoplasmic dynein, has been localized to the fibrous corona region of the kinetochore (27). The function of CENP-E is uncertain. Microinjection of anti-CENP-E antibody into dividing cells interfered with proper congression to the metaphase plate and delayed anaphase onset (25). Anti-CENP-E antibody also blocked ATP-independent chromosome movement driven by microtubule depolymerization, suggesting a role in maintaining microtubule attachment (28). Although CENP-E was first deduced to be a minus end motor based on immunodepletion experiments (29), the purified protein has recently been shown to exhibit plus end motor activity(30). In an earlier study CENP-E was reported to be restricted to the active centromere in a majority of Robertsonian dicentric translocations examined and in some cases at both centromeres (31). Participation of the inactive centromeres in chromosome segregation was shown to correlate with CENP-E immunoreactivity.
MCAK is a 90 kDa member of the KIF2 subfamily of kinesins, which contains a centrally located motor domain (32). It has been localized to the outer face of the kinetochore, as well as to the central region of the centromere between sister kinetochores(32). Although sequence analysis reveals homology to a subfamily of plus end kinesin-related proteins, in vitro microtubule-based motility has not been demonstrated for the purified polypeptide. The Xenopus homolog of MCAK, XKCM1, has been shown to regulate the frequency of microtubule catastrophe events in an in vitro spindle assembly assay (33). These data suggest that MCAK may coordinate microtubule dynamics with chromosome movement during mitosis in vivo.
Knowledge of the motor protein content of inactive centromeres should provide further insight into their functional relationship to active centromeres. We have investigated this question by immunofluorescence microscopy using antibodies to cytoplasmic dynein, dynactin, CENP-E and MCAK, as well as to HZW10 and CENP-F, recently identified binding partners of dynactin and CENP-E respectively (34,35), and Mad2p, a cell cycle checkpoint protein (36). We find that all motor-related proteins, with the possible exception of MCAK, are restricted to active centromeres and deficient in inactive centromeres, providing a mechanistic basis for understanding the loss of function of these stable accessory centromeres.
To compare the antigenic properties of active and inactive centromeres cell lines which contain stable multicentric chromosomes were first treated with nocodazole to arrest them in prometaphase. Under these conditions fully developed kinetochores are observed (8). For clear identification of multicentrics, chromosome spreads were examined. In each case we identified the active centromere on the basis of two criteria. First, it was the sole site of attachment between sister chromatids, a property attributed to the active centromere in fixed cell studies (37; see also for example Fig. 5e-h and i-l). Second, it was also generally seen to correspond to the more pronounced constriction (see for example Fig. 3).
Figure 1 shows triple labeling of a dicentric chromosome from MDA-435 cells using anti-cytoplasmic dynein intermediate chain antibody, CREST autoimmune serum and a DNA-specific stain (Oligreen). CREST staining was observed throughout both centromeres (Fig. 1b). In contrast, anti-cytoplasmic dynein staining was seen only at the primary constriction (Fig. 1a). The two cytoplasmic dynein clusters lie on either side of the CREST-positive region, consistent with the previously reported localization to the outermost layer of prometaphase kinetochores (24).
We have found that antibodies directed against several different motor-related antigens, cytoplasmic dynein intermediate chain, the p150Glued, Arp1, and dynamitin (p50) components of the dynactin complex and CENP-E, as well as HZW10, CENP-F and Mad2p, localize to a single centromere in all MDA-435 dicentric or multicentric chromosomes (summarized in Table 1). These data add strong support to the specificity of kinetochore staining reported for antibodies directed against these polypeptides (19,21,22,24,31,32,38-42,44,45) and further distinguish active from inactive centromeres at the molecular level.
Table 1.
The one exception to the selective localization pattern was that observed using anti-MCAK. In chromosome spreads prepared by conventional means staining at neither primary nor secondary constriction was observed, suggesting dissociation of MCAK from centromeres during sample preparation. Fixation did, indeed, preserve primary constriction staining (Fig. 5). Intriguingly, in some of the few cases of free chromosomes that could be found in preparations fixed at high formaldehyde concentrations, secondary constriction staining was also observed. It is uncertain whether this aspect of the staining pattern is specific. Because secondary constriction staining is observed under harder fixation conditions, it may be argued that it more closely reflects the physiological state. On the other hand, the diffuse nature of the secondary constriction staining may be an indication of a non-specific interaction produced as a result of our experimental manipulations.
The contributions of the motor-related or checkpoint proteins to the morphological features of the kinetochore remain uncertain, but it seems reasonable to expect that the laminar elements of the kinetochore represent distinct proteinaceous domains. CENP-B has been localized to the centromeric chromatin beneath the kinetochore (4), consistent with a direct interaction with centromeric DNA (5). Cytoplasmic dynein, CENP-E and CENP-F have been localized to the outer kinetochore plate and the fibrous corona (27,32,41), a structure observed during prometaphase. Dynactin, which we expect to co-localize with cytoplasmic dynein based on the biochemical interaction of the two complexes (17,18), HZW10 and Mad2p all also appear from our light microscopic analysis to be associated with the outer portion of the kinetochore.What fraction of the mass of the kinetochore these proteins represent is unknown, but their absence from inactive centromeres could largely account for the limited kinetochore structure observed at these sites(7,46,47).
The mechanism by which these proteins are lost during the evolution of inactive centromeres from active centromeres remains an intriguing issue. It seems unlikely that theseveral proteins and protein complexes evaluated in this study are lost independently. Rather, it seems reasonable to assume the existence of an organizing factor, the loss of which leads to the large scale eliminationof the outer kinetochore structure. One candidate for this role is CENP-C, a centromere protein needed for maintenance of kinetochore size and stability (48) and also reported to interact directly with human alphoid DNA (49). In the light of evidence that HZW10 is involved in the association of dynactin with the kinetochore (34), which in turn mediates cytoplasmic dynein binding, the absence of HZW10 from inactive centromeres must be considered significant. Future work will be directed at defining the stages in the evolution of inactive centromeres and, a related issue, the organizational hierarchy of kinetochore proteins.
The cells used in this study were a human breast ductal carcinoma line which contains a stable dicentric chromosome (MDA435, ATCC no. HTB129) and a distinct MDA435-derived line which contains a stable pentacentric chromosome (10). Both lines were maintained at subconfluence in MEM (Gibco/BRL Life Technologies, Gaithersburg, MD), 12% fetal bovine serum (Gibco/BRL), 100 U/ml penicillin, 100 mg/ml streptomycin (Sigma Chemical Co., St Louis, MO).
Primary antibodies used were CREST-SH human autoimmune serum (50; provided by Dr B.R.Brinkley, Baylor College of Medicine), pAb-1.6 rabbit polyclonal anti-CENP-E antiserum (28; provided Dr T.J.Yen, Fox Chase Cancer Center), polyclonal anti-CENP-F antiserum (51; provided Dr T.J.Yen), 50-1 monoclonal anti-dynamitin antibody (16), polyclonal anti-p150Glued serum (17; provided by Dr K.T.Vaughan, University of Massachusetts Medical Center), 74.1 monoclonal anti-cytoplasmic dynein intermediate chain antibody (52; provided by Dr K.Pfister, University of Virginia), MCAK polyclonal anti-mitotic centromere-associated kinesin antibody (32), A27 polyclonal anti-Arp1 antibody (provided by Drs S.Clark and D.Meyer, University of California, Los Angeles), monoclonal Mad2p antibody (provided by J.Waters, University of North Carolina, Chapel Hill, and Drs R.Chen and A.W.Murray, University of California, San Francisco), human JH autoimmune serum against TD60 (43; provided by Drs S.Wheatley and Y.-L.Wang, University of Massachusetts Medical Center) and affinity-purified anti-HZW10 antibody (39; provided by D.A.Starr and Dr M.L.Goldberg, Cornell University).
Chromosome spreads were prepared as follows. Cells were exposed to 10 mM nocodazole for 6-8 h or to 0.01 mg/ml colcemid for 1 h for the pentacentric-containing subline, rinsed in phosphate-buffered saline (PBS), incubated in 0.075 M KCl for 10 min at room temperature and centrifuged in a swinging bucket rotor at 3000 r.p.m. for 10 min at 4°C. For C banding the chromosome spreads were fixed with 3:1 methanol:acetic acid and treated with 0.2 N HCl for 50 min, saturated Ba(OH)2 for 6-10 min and 2* SSC at 65°C for 3 h. For most of the experiments involving immunofluorescence microscopy the spreads were fixed in methanol at -20°C for 10 min. For MCAK-labeled samples cells were centrifuged as above but in the presence of 0.1-4% formaldehyde in PBS for 15 min at 4°C and then fixed in methanol. The spreads were rinsed in PBS, incubated in primary antibody for 30-45 min, followed by secondary antibody for 30-45 min, exposed to Oligreen (Molecular Probes Inc., Eugene, OR) and mounted in 0.1% p-phenylene diamine in PBS, 50% glycerol. Cy3- and Cy5-conjugated secondary antibodies were obtained from Jackson Immunoresearch (Westgrove, PA). For sequential labeling (Fig. 2) chromosome spreads were incubated in the first set of primary and secondary antibodies (monoclonal anti-dynamitin followed by FITC-conjugated goat anti-mouse IgG), exposed to ethidium bromide, mounted in Dabco and photographed. The chromosomes were then incubated in a second set of antibodies (CREST-SH followed by FITC-conjugated goat anti-human IgG), exposed again to ethidium bromide and relocated for photography.
For most experiments laser scanning confocal immunofluorescence microscopy was carried out using a MRC1024 system (BioRad, Hercules, CA) equipped with a Kr/Ar laser mounted on a Nikon Diaphot 200 photomicroscope. For Figure 2 photographs were taken with black and white film (Kodachrome 400 ASA, 12 s exposure) using a Zeiss photomicroscope and the negatives were scanned using a Nikon Coolscan Scanner (Nikon Inc., Melville, NY). Cropping and color overlays of digitized images were carried out using Adobe Photoshop v.4.0 (Adobe Systems Inc., Mountain View, CA) and figures were assembled in CorelDraw v.7.0 (Corel Corp., Ottawa, Canada). All figures were printed on a Kodak ColorEase PS color printer (Eastman Kodak Co., Rochester, NY). Unless noted in the figure legends, green was assigned to the motor components, red to CREST and blue to the DNA.
We thank Dr B.R.Brinkley, Dr R.Chen, Dr S.Clark, Dr M.L.Goldberg, Dr D.Meyer, Dr A.W.Murray, Dr K.Pfister, D.A.Starr, Dr K.T.Vaughan, Dr Yu-Li Wang, J.Waters, Dr S.Wheatley and Dr T.J.Yen for their generous gift of antibodies. This work was supported by NIH grant GM478434 to RBV.
Human Molecular Genetics
Pages
Introduction
Results
Discussion
Materials And Methods
Cells
Immunological reagents
Cytochemical methods
Microscopy and photography
Acknowledgements
References
Antigen
Active
centromereInactive
centromere
CREST (SH)a
+
+
CENP-Aa
+
-
CENP-Ba
+
+
CENP-Ca
+
-/+b
Cytoplasmic dynein intermediate chainc
+
-
Dynactin subunitsc
p150Glued
+
-
Arp1
+
-
Dynamitin (p50)
+
-
CENP-Ea,c
+
-/+b
MCAKc
+
-/+d
CENP-Fc
+
-
HZW10a,c
+
-
Mad2pc
+
-
TD60c
+
-
REFERENCES
This page is run by Oxford University Press, Great Clarendon Street, Oxford OX2 6DP, as part of the OUP Journals
Comments and feedback: www-admin{at}oup.co.uk
Last modification: 14 Mar 1998
Copyright© Oxford University Press, 1998.
![]()
CiteULike
Connotea
Del.icio.us What's this?
This article has been cited by other articles:
![]() |
Y. Mimori-Kiyosue and S. Tsukita "Search-and-Capture" of Microtubules through Plus-End-Binding Proteins (+TIPs) J. Biochem., September 1, 2003; 134(3): 321 - 326. [Abstract] [Full Text] [PDF] |
||||
![]() |
K. A. Maggert and G. H. Karpen The Activation of a Neocentromere in Drosophila Requires Proximity to an Endogenous Centromere Genetics, August 1, 2001; 158(4): 1615 - 1628. [Abstract] [Full Text] [PDF] |
||||
![]() |
A. A. Van Hooser, I. I. Ouspenski, H. C. Gregson, D. A. Starr, T. J. Yen, M. L. Goldberg, K. Yokomori, W. C. Earnshaw, K. F. Sullivan, and B. R. Brinkley Specification of kinetochore-forming chromatin by the histone H3 variant CENP-A J. Cell Sci., January 10, 2001; 114(19): 3529 - 3542. [Abstract] [Full Text] [PDF] |
||||
![]() |
N. Sugata, S. Li, W. C. Earnshaw, T. J. Yen, K. Yoda, H. Masumoto, E. Munekata, P. E. Warburton, and K. Todokoro Human CENP-H multimers colocalize with CENP-A and CENP-C at active centromere-kinetochore complexes Hum. Mol. Genet., November 1, 2000; 9(19): 2919 - 2926. [Abstract] [Full Text] [PDF] |
||||
![]() |
R. Saffery, D. V. Irvine, B. Griffiths, P. Kalitsis, L. Wordeman, and K.H. A. Choo Human centromeres and neocentromeres show identical distribution patterns of >20 functionally important kinetochore-associated proteins. Hum. Mol. Genet., January 22, 2000; 9(2): 175 - 185. [Abstract] [Full Text] [PDF] |
||||
![]() |
D. Starr, R Saffery, Z Li, A. Simpson, K. Choo, T. Yen, and M. Goldberg HZwint-1, a novel human kinetochore component that interacts with HZW10 J. Cell Sci., January 6, 2000; 113(11): 1939 - 1950. [Abstract] [PDF] |
||||
![]() |
A. A. VAN HOOSER, M. A. MANCINI, C. D. ALLIS, K. F. SULLIVAN, and B. R. BRINKLEY The mammalian centromere: structural domains and the attenuation of chromatin modeling FASEB J, December 1, 1999; 13(9002): 216S - 220S. [Abstract] [Full Text] [PDF] |
||||
![]() |
H. R Slater, S. Nouri, E. Earle, A. W I Lo, L. G Hale, and K H A. Choo Neocentromere formation in a stable ring 1p32-p36.1 chromosome J. Med. Genet., December 1, 1999; 36(12): 914 - 918. [Abstract] [Full Text] [PDF] |
||||
![]() |
A. W.I. Lo, G. C.-C. Liao, M. Rocchi, and K.H. A. Choo Extreme Reduction of Chromosome-Specific alpha -Satellite Array Is Unusually Common in Human Chromosome 21 Genome Res., October 1, 1999; 9(10): 895 - 908. [Abstract] [Full Text] |
||||
![]() |
J. T. Robinson, E. J. Wojcik, M. A. Sanders, M. McGrail, and T. S. Hays Cytoplasmic Dynein Is Required for the Nuclear Attachment and Migration of Centrosomes during Mitosis in Drosophila J. Cell Biol., August 9, 1999; 146(3): 597 - 608. [Abstract] [Full Text] [PDF] |
||||
![]() |
C.-Y. F. Huang, C.-P. B. Chang, C.-L. Huang, and J. E. Ferrell Jr. M Phase Phosphorylation of Cytoplasmic Dynein Intermediate Chain and p150Glued J. Biol. Chem., May 14, 1999; 274(20): 14262 - 14269. [Abstract] [Full Text] [PDF] |
||||
![]() |
T. Maney, A. W. Hunter, M. Wagenbach, and L. Wordeman Mitotic Centromere-associated Kinesin Is Important for Anaphase Chromosome Segregation J. Cell Biol., August 10, 1998; 142(3): 787 - 801. [Abstract] [Full Text] [PDF] |
||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||









