Centromere

Prévia do material em texto

The Centromere: Hub of Chromosomal Activities
A. F. Pluta, A. M. Mackay, A. M. Ainsztein, 1. G. Goldberg,
W. C. Earnshaw*
Centromeres are the structures that direct eukaryotic chromosome segregation in mitosis and is the binding site for an essential pro-
and meiosis. There are two major classes of centromeres. Point centromeres, found in tein complex, termed CBF3 (10). Muta-
the budding yeasts, are compact loci whose constituent proteins are now beginning to tions in CDEIII or in genes encoding the
yield to biochemical analysis. Regional centromeres, best described in the fission yeast CBF3 polypeptides disrupt chromosome
$chizosaccharomycespombe, encompass many kilobases of DNA and are packaged into segregation (3-5). The 8-bp CDEI sequence
heterochromatin. Their associated proteins are as yet poorly understood. In addition to also appears in the promoters of a number of
providing the site for microtubule attachment, centromeres also have an important role genes. Its binding factor, Cpflp (I1), is a
in checkpoint regulation during mitosis. transcriptional activator when bound to
these noncentromeric sequences. This is
paradoxical because transcription directed
toward the core CEN sequences abolishes
Centromeres have multiple roles during entire length of the chromatid, to chromo- centromere function in vivo (12). CDEI
mitosis and may also have essential func- somes with localized centromeres, where and CDEIII are separated by CDEII, an
tions in interphase. (i) The centromere is microtubules attach to a single region that A:T-rich region whose primary sequence is
the site of formation of the kinetochore, the often appears constricted relative to the not well conserved but whose length is
buttonlike structure at the chromosomal chromosome arms. Although holocentric important for centromere function.
surface that binds spindle microtubules and chromosomes occur in many arthropods The point centromere is packaged into a
regulates chromosome movements in mito- and plants as well as in the well-studied compact nuclease-resistant chromatin struc-
sis. The DNA sequence that specifies cen- genetic model organism Caenorhabditis el- ture of -250 bp (13) that binds a single
tromere location on the chromosome is re- egans (5), their structure remains poorly microtubule (14). Several findings suggest a
ferred to as the CEN locus. (ii) The cen- understood. This review will therefore focus model in which this structure may be assem-
tromere is the final locus of sister chromatid on localized centromeres. bled around a modified nucleosome. A reduc-
pairing in mitosis and, therefore, must re- There are two classes of localized centro- tion in amounts of core histones H2B or H4
ceive the signal that triggers the release of meres, point and regional centromeres. alters centromeric chromatin (15), and CSE4,
sister chromatids at the metaphase-an- Point centromeres have been identified in a gene required for chromosome segregation,
aphase transition. Centromeres presumably the budding yeasts Saccharomyces cerevisiae encodes a highly divergent histone H3-related
contain the machinery necessary to effect (3-5), Schizosaccharomyces uvarum (6), and molecule (16). The human centromeric
this separation. (iii) The centromere is in- Kluyveromyces lactii (7). These compact loci polypeptide CENP-A is also a very divergent
volved in cell cycle checkpoint control. In encompass three conserved DNA elements histone H3 (17). Because the preferred length
most cells, centromeres of mitotic chromo- (CDEs): CDEI, CDEII, and CDEIII (8). for CDEII, 78 to 86 bp, corresponds to roughly
somes that have not yet achieved a stable [The point centromeres of Yarrowia lipoly- one turn of the DNA around the nucleosomal
bipolar orientation on the spindle send a tica have a different sequence organization core, this model predicts that CDEI and
signal that delays the onset of the met- (9).] The 25-base pair (bp) CDEIII is ab- CDEIII might be juxtaposed in space (Fig. 2).
aphase-anaphase transition. (iv) The cen- solutely required for centromere function Regional centromeres, the other class of
tromere acts as a marshaling area for the
chromosomal passenger proteins, proteins Fig. 1. Different behavior
that transfer from the chromosomes to the of intrinsic proteins of the
mitotic spindle during metaphase or an- centromere and chromo-
aphase (1). These are thought to be mitosis- somal passenger pro-
specific cytoskeletal proteins (Fig. 1). teins during mitosis.
In this review, we discuss several aspects Both the CENP antigens
of centromere structure and function that [(A), labeled red with au-
have received particular attention in recent toimmune serum] and
years. For other reviews we refer the reader passenger proteins, the
to references (2-4). INCENPs 1(B), labeled redwith rabbit serum], con- ,
centrate at centromeres 0 i
The Two Classes of Centromeres: during metaphase. The
Point and Regional CENP antigens remain at
centromeres during an-
Although centromere functions are con- aphase (C), whereas the
served in all cells, centromeres exhibit be- INCENPs transfer to the
wildering structural variability between spe- overlapping microtubules
cies. This ranges from holocentric chromo- of the central spindle (D).
somes, where microtubules attach along the The position of the CENPand INCENP antigens In "
____________________________________ mitosis is indicated by yel-
The authors are in the Department of Cell Biology and low arrows. Microtubules J 1lif : if l:d.
Anatomy, Johns Hopkins School of Medicine, Baltimore, are labeled green, and the
MD 21205, USA. blue arrows indicate the
*To whom correspondence should be addressed. position of the chromosomes during anaphase and telophase.
SCIENCE * VOL. 270 * 8 DECEMBER 1995 1591
 
o
n
 M
ay
 2
0,
 2
01
0 
w
w
w
.s
ci
en
ce
m
ag
.o
rg
D
ow
nl
oa
de
d 
fro
m
 
localized centromeres, encompass kilobases gional centromeres contain point CEN loci onies (30). Thus, centromere function may
of DNA and include both unique and re- that have yet to be identified. require not only a minimal set of sequences,
peated DNA elements. The organization of In most organisms, centromeres are com- but the acquisition of specific epigenetic
the repeated DNA component can exhibit posed of heterochromatin, a specialized modifications by those sequences.
considerable polymorphisms between form of chromatin that remains condensed A recent study suggests a fundamental
strains, as it does in S. pombe (18), or throughout the cell cycle, replicates late in similarity of organization between the S.
between individuals, as it does in humans the S phase, and suppresses the expression pombe and Drosophila melanogaster centro-(19). Whereas a single CEN locus accom- of most genes. Like classic heterochroma- meres. Drosophila centromere function re-
plishes all aspects of centromere function in tin, S. pombe centromeres are not tran- quires Bora Bora, a 220-kb "island" contain-
the point centromere, in regional centro- scribed and can exert a position effect on ing complex sequence DNA (31). For full
meres different sequences may have become reporter genes that are inserted into the centromere function, Bora Bora must be
specialized to regulate kinetochore assembly centromeric array (24). Certain mutants flanked on either the 5' or 3' side by an-
and sister chromatid pairing. that affect silencing of the mating type loci other -200 kb of simple sequence satellite
Centromeres of S. pombe are the para- (clr4, riki, and swi6) also affect centromeric DNA. It has been suggested that Bora Bora
digm for the regional centromere. These loci silencing (25). These mutations also com- nucleates kinetochore formation, whereas
span 40 to 100 kb (20), with the smallest promise the fidelity of chromosome segre- the simple-sequence DNA regulates sister
functional centromeres containing -19 kb gation,suggesting that transcriptional sup- chromatid pairing.(21). The chromatin structure and function- pression and centromere function are Several groups are attempting to define
al organization of these centromeres are mechanistically linked. Recent work on the functional CEN sequences in mammals,
complex. The centromeres of S. pombe all swi6+ gene reveals that Swi6p is enriched with the ultimate goal of constructing sta-
contain a 4- to 7-kb central core sequence in centromeric and telomeric heterochro- ble artificial chromosomes. This approach is
embedded within a domain containing both matin (26). Swi6p has a sequence motif, the hindered by the extraordinary variability of
inverted and direct repeats. This central core chromo domain, that has been found in centromeric DNA sequences between spe-
contains several functionally redundant do- several proteins that are involved in the cies. Only one centromeric sequence ele-
mains (21). A 2.1-kb region of the K/dg-type stable repression of gene expression (27). ment is known to be conserved between
flanking repeat is also required for centro- One of these proteins, HP1, is concentrated primates and rodents. This is the 17-bp
mere function (20). This region confers a in heterochromatin (28), particularly at binding site for CENP-B (32, 33). A sum-
special chromatin structure on the central centromeres in mammals (29). The other mary of the centromere components of
core sequences (22) and has a number of protein, polycomb, is thought to stabilize mammals is presented in Fig. 3. The
binding sites for proteins that have yet to be chromatin domains in a transcriptionally "CENP-B box" is found in human ox-satel-
identified. It is not known if the bundles of inactive state. lite DNA, Mus musculus minor satellite
two to four microtubules bound by S. pombe In a further parallel with heterochroma- DNA, and in a minimal functional form (9
kinetochores (23) are associated with specif- tin, the establishment of centromeric activ- of 17 bp conserved) in the 79-bp centro-
ic DNA sequences, as is the case in the point ity in S. pombe is subject to functional meric satellite DNA of Mus caroli (34). In
centromere, or if the regional centromere variegation. When certain artificial chro- African green monkey ox-satellite DNA the
adopts a three-dimensional configuration mosomes are introduced into S. pombe cul- CENP-B box occurs at a frequency .0.1%
that nucleates assembly of a proteinaceous tures, a substantial difference in chromo- of that found in human cells (35), raising
kinetochore. It cannot be excluded that re- some stability is seen between different col- doubts as to whether the binding of
CENP-B to this sequence can be a major
determinant of centromere structure andFig. 2. Hypothetical model
_ 1
forthe budding yeast kineto- crotuul e function. Such concerns were previouslyforthebudin yest ineo-
sequence raised by the demonstration that CENP-Bchore. The 1 25-bp CEN similarity to CFcopntsrtein and CENP-B box DNA are bothDNA is shown wrapped human CBF3 components prot
around a nucleosomal core, CENP-C _ and genes undetectable on the human Y chromosome
which may contain the his- - p110 NDC10/CBF2/CTF14 (36).
tone H3 variant Cse4p (16). Mif2p p64 CBF3b CEP3 Current evidence indicates that oa-satel-
Whilst this model is consist- interactions p58 SKPC1 lite DNA does have an important role in
ent with the results of muta- Cpf1p centromere function. When incorporatedtional analysis of the his- into ectopic locations on chromosome arms,
tones and length constraints DEI CDEIII transfected ox-satellite DNA can perturb the
on the CDEII sequence, di- segregation of sister chromatids (37).rect evidence for such a W_ o bconfiguration has yet to be epWhether this occurs because of the assembly
obtained. Conserved DNA binding? of extra kinetochores or because of effects on
element (CDEI, 8 bp) is < histoneH3 sister chromatid separation remains to be
complexed with the 39-kD CDEII variant determined. One approach to identifying hu-Cpf1p [also known as CP1 man centromeric DNA involves the use of
and CBF1 (11)]. The four-subunit CBF3 complex binds the 25-bp element CDEIII. The CBF3 subunits telomere-mediated fragmentation of chro-include p1 0 [encoded by the NDC10/CBF2/CTF14 gene (68, 69)], p64 [CBF3b/CEP3 (70)], p58 [CTF13 mosomes within the centromere. This ap-(69)], and p23 [SKP1 (71)]. The detailed functions of these components are unknown, although p64 proach has enabled the localization of the Y
contains zinc finger motifs and has therefore been suggested to be the DNA binding subunit. p23 is chromosome centromere to a region of
-70
conserved in species from Arabadopsis to mammals (71). The hypothetical linker molecule iS inserted as
a result of experiments demonstrating that CBF3 alone is not sufficient to cause CEN DNA to associate kb within the ax-satellite array (38). The
with microtubules in vitro (72). The kinesin-related protein Kar3p appears to be responsible for centro- most stable of these fragmented chromo-
mere movement on microtubules under in vitro conditions (73). Mif2p, which is required for chromosome somes also retain unique sequences from the
segregation and spindle assembly (74, 75), interacts genetically with p39, p64, and p110. Cbf5p, which p arm of the Y chromosome. The parallel to
can bind to microtubules, also interacts genetically with p1 10 (56). Mif2p shares several short regions of S. pombe and Drosophila regional centro-
amino acid similarity with the human kinetochore protein CENP-D (75, 76). meres is noteworthy.
1592 SCIENCE * VOL. 270 * 8 DECEMBER 1995
 
o
n
 M
ay
 2
0,
 2
01
0 
w
w
w
.s
ci
en
ce
m
ag
.o
rg
D
ow
nl
oa
de
d 
fro
m
 
Centromeres May Function changes during cell cycle progression (46) Centromeres Signal the "Wait
During Interphase and in response to the functional (47) and Anaphase" Checkpoint
transcriptional states of the cell (48, 49).
Whereas the functions of mitotic centromeres This has led to the thinking that highly When vertebrate cells enter mitosis, the nu-
are well documented, the role of centromeres repeated sequences such as those at centro- clear envelope breaks down and chromosomes
during interphase remains enigmatic. The meres may organize the interphase nucleus randomly encounter microtubules that ema-
persistence of condensed centromeres during in ways that influence gene expression (50). nate from opposing centrosomes. The ensuing
interphase was first revealed by the staining of The association of interphase centromeres processes of microtubule capture by kineto-
discrete foci in interphase nuclei with serum with nucleoli in human cells is true not only chores and the gradual establishment of a
from autoimmune patients (39). These foci for the acrocentric chromosomes in which balance of forces between opposing kineto-
correspond to specialized heterochromatin centromeres and nucleolus organizer regions chores and centrosomes ultimately result in
domains (40). The demonstration that these (NOR) are in relatively close proximity, but the establishment of the metaphase configu-
foci of antibody staining colocalize with a-sat- also for centromeres of non-NOR-bearing ration with all chromosomes lined up at the
ellite DNA suggested a tight association of at chromosomes (49, 50). Centromere proteins midzone of a bipolar spindle and awaiting the
least some "intrinsic" proteins with the cen- are detectable both ultrastructurally and bio- onset of anaphase. If aneuploidy is to be
tromere throughout the cell cycle (41). Asso- chemically in isolated human nucleoli (51). avoided, it is crucial that the cell delay an-
ciation of these proteins with centromeres Furthermore, preliminary evidence has been aphase onset until all chromosomes have
during interphase is apparently required for obtained for a direct biochemical interac- achieved this balanced alignment. In recent
assembly offunctional kinetochores. Microin- tion between human kinetochore protein years, significant advances have been made injection of centromere antibodies during inter- CENP-C (52) and the nucleolar transcription the characterization of this checkpoint.
phase disrupts kinetochore assembly and factor UBF (also known as NOR90) (53). In 1970, Zirkle proposed that a signal em-
blocks progression through mitosis (42, 43). Whereas the functional significance of these anating from the chromosomes or spindle
Thus, centromeres are not quiescent during associations remains unclear, studies of hu- could prolong metaphase until both ki-
interphase but require specific interactions man autoantibodies suggest that centromere netochores of every chromosome had been
with other proteins to prepare for their subse- associations with nucleoli may have impor- attached to microtubules and brought to the
quent roles in chromosome segregation during tant consequences. In particular, it has been metaphase plate (57). Dietz (58) had original-
mitosis. postulated that a complex between nucleoli ly proposed and Nicklas (59) demonstrated
Mammalian centromeres occupy dis- and centromeres is the dominant autoantigen that meiotic kinetochores could somehow
tinct, nonrandom positions in the inter- in scleroderma spectrum disease (54). sense tension. This led to the proposal that in
phase nucleus. However, these centromeres A further link between centromeres and the absence of tension produced by a stable
are not entirely clustered or polarized as has nucleoli involves Nap57p, a nucleolar pro- bipolar attachment, kinetochores transmit an
been described in fission yeast (44) and tein from rat liver, which is thought to inhibitory signal that delays the metaphase-
Drosophila (45). Instead, mammalian cen- function as a chaperone for nucleolar pro- anaphase transition (60). Results consistent
tromeres tend to congregate near the nucle- teins (55). Nap57p closely resembles Cbf5p, with this view have been obtained in S. cer-
ar periphery and around nucleoli, although a S. cerevisiae protein of unknown function evisiae, where cells with mutant centromeres
they can also be found throughout the nu- (55, 56) that interacts genetically with exhibit a mitotic delay (61), and in humans,
clear volume. These interphase centromere CBF3 (Fig. 2) and is also localized to the where microinjection of antibodies to centro-
positions are not static, nor does their ar- nucleolus (56). The detection of biochem- meres also causes a mitotic delay (42).
rangement appear to conform to one pat- ical associations between centromeres and A recent study provides direct evidence
tern for all cell types. Rather, within a cell nucleolar proteins suggests that centro- for the involvement of kinetochore tension
type, centromere distribution with regard to meres may have as yet undiscovered func- in releasing cells from the "wait anaphase"
these two nuclear landmarks reproducibly tions during interphase. checkpoint. In mantid spermatocytes, the
presence of a single mono-oriented X chro-
mosome results in a permanent block to
Fig. 3. Components of the Centromere anaphase entry, leading to eventual cell
mammalian centromere. The K death. However, if such a chromosome is
condensed centromeric het- CENP-A Kinetochore placed under gentle tension by pulling it
erochromatin, which may CLation? away from the pole with a microneedle, the
function as a structural foun- Function? MitubuIecells overcome their metaphase block anddation for the kinetochore, is proceed through anaphase (62). Presumably
rich in (-satellite DNA and its Centromeric Fibrous corona the external application of tension switchesbinding protein CENP-B (33, heterochromatin c_toplasm CENP-E o the "at aphase" siglering
40) but also contains the ki- cytoplasmiNdynein off theh wait anaphase" signal-generating
nesin-related MCAK (77) and CENP-B _ X microtubule binding mechanism at the kinetochore.
the INCENP chromosomal MCAKENP-B anaphase motors In cultured vertebrate cells, this "wait
passengers (78). At the sur- INCENPs lat Outer plate anaphase" signal can be transmitted by un-face of the heterochromatin chromatid pairing Interzop C attached kinetochores. In the rat kangaroo
is the inner kinetochore plate. stFucuraantipoensNACENP-F cell line Ptkl, attachment of the last free
This structure contains both kinetochore sizelFtension receptors? tubule binding kinetochore to the spindle precedes an-DNA (79) and CENP-C (52), Ideinaio cell cyclesignaling? aphase onset by about 20 min, and the
with the latter being essentialprsneooeormemn-ritd
for kinetochore assembly (43). The outer kinetochore plate, which contains the kinesin-related CENP-E (80) choooePlcs heahs-n
and CENP-F (also termed mitosin) (81) is involved in microtubule binding. The interzone between the two crmsmsbok heahs-n
plates contains the 3F3/2 antigens (65) and may be involved in tension sensing and cell cycle signaling. The aphase transition (63). However, if the un-
fibrous corona contains microtubule motor proteins, both of the kinesin [CENP-E (80)] and cytoplasmic attached kinetochore of such a mono-ori-
dynein (82) families. The detailed localizations of CENP-A [a divergent histone H3 that resembles yeast ented chromosome is destroyed with a laser,
Cse4p but is not functionally interchangeable with it (16, 17)] and CENP-D [which appears to correspond to the cell then enters anaphase -20 mmn later
the RCC1 protein, a regulator of nuclear transport (83)] are not known. (63). This was interpreted to show that
SCIENCE * VOL. 270 * 8 DECEMBER 1995 1593
 
o
n
 M
ay
 2
0,
 2
01
0 
w
w
w
.s
ci
en
ce
m
ag
.o
rg
D
ow
nl
oa
de
d 
fro
m
 
destruction of the unattached kinetochore 3. L. Clarke, Trends Genet. 6, 150 (1990); 1. Schulman 43. J. E. Tomkiel, C. A. Cooke, H. Saitoh, R. L. Bernat,
abolished transmission of an inhibitory and K. S. Bloom, Annu. Rev. Cell BioL 7, 311 (1991); W. C. Earnshaw, ibid. 125, 531 (1994).
it ~~~~~~~~~~~~~~~J.H. Hegemann and U. N. Fleig, BioEssays 15, 451 44. H. Funabiki, I. Hagan, S. Uzawa, M. Yanagida, ibid.
wait anaphase" signal. (1993). 121, 961 (1993).
A fortuitous observation provided the 4. C. Tyler-Smith and H. F. Willard, Curr. Op/n. Genet. 45. V. E. Foe and B. M. Alberts, ibid. 100, 1623 (1985).
first evidence for a link between the attach- Dev. 3, 390 (1993); W. C. Earnshaw and J. E. Tom- 46. M. Ferguson and D. C. Ward, Chromosoma (Bert.)kiel, Curr. Op/n. Cell B/of. 4, 86 (1992); G. J. Gorbsky, 101, 557 (1992).
ment status of a kinetochore and its bio- Trends Cell B/of. 5,143 (1995). 47. P. C. Park and U. De Boni, Exp. Cell Res. 203, 222
chemical makeup. In addition to staining 5. D. G. Albertson and J. N. Thomson, Chromosoma (1992).
other cellular structures, the phosphoepitope- (Berl.) 86, 409 (1982). 48. J. Janevski, P. C. Park, U. De Boni, ib/d. 217, 227
spcfcmonoclonal antibody 3F3/2 preferen- 6. J. A. Huberman, R. D. Pridmore, 0. Jaiger, B. Zon- (1995).specific
~~~~~~~~~~~~~neveld,P. Philippsen, ibid. 94, 162 (1986). 49. 1. Leger, M. Guillaud, B. Krief, G. Brugal, Cytometry
tially stains kinetochores that are not under 7. J. J. Heus, B. J. M. Zonneveld, H. Y. Steensma, J. A. 16, 313 (1994).
tension (64). Injection of this antibody into Van den Berg, Curr. Genet. 18, 517 (1990). 50. L. Manuelidis and J. Borden, Chromosoma (Berl.)
cells delays the metaphase-anaphase transi- 8. M. Fitzgerald-Hayes, L. Clarke, J. Carbon, Cell 29, 96, 397 (1988).
without affeting chromoome move- ~ 235 (1982). 51. R. L. Ochs and R. I. Press, Exp. Cell Res. 200, 339tion wtotafcigcrmsm moe 9.P. Fournier et at., Proc. Nati. Acad. Sci. U.S.A. 90, (1992).
ments (65). The kinase-phosphatase balance 4912 (1993). 52. H. Saitoh et at., Cell 70, 115 (1992).
responsible for expression of this phospho- 10. J. Lechner and J. Carbon, Cell 64, 717 (1991). 53 A. F. Pluta and W. C. Earnshaw, in preparation; E.K.11. R. J. Brain and R. 0. Kornberg, Mot. Cell. B/of. 7,403 L.Ca,HImiJ..Hme,EM.TnJ.xp
epitoe mayeither be indicative of, or actu- (1987); R. E. M. Baker, M. Fitzgerald-Hayes, T. C. L.Cn,HlmiJ..Hme,EM.T,J.xpepitope may ~~~~~~~~~~~~~~~~~~~~~~~Med.174,1239 (1991).
ally contribute to, the transmission of the O'Brien, J. B/of. Chem. 264, 10843 (1989); M. Cai' 54. E. M. Tan, E. K. L. Chan, K. F. Sullivan, R. L. Rubin,
"twait anaphase" signal (64, 65). and R. W. Davis, Cell 61, 437 (1990); J. Mellor et at., Clin. Immunof. Immunopathof. 47,121 (1988).
Recenly, dirct eperienta lin was EMBO J. 8, 4017 (1990). 5U.TMeranG.Bol,JCe io12 50Recently adiret expermental ink was 12. A. Hill and K. Bloom, Mol. Cell. B/of. 7, 2397 (1987) 55 U194.T.MiradGBob,J.ClB/f12,50
shown between the expression of the 3F3/2 13. K. S. Bloom and J. Carbon, Cell 29, 305 (1982). (1994).Jin,KMdleo,H-.Yn CFuqt
eioeand kinetochore tension in grasshop- 14. J. B. Peterson and H. Ris, J. Cell Sc/. 22, 219 (1976). 56C.Jangr.Midbon,Ho.-Jl.Bo13 Yoon4C193.Fuqt,Jpepitopermtcts hnbiaet eed- 15. M. J. Saunders, E. Yeh, M. Grunstein, K. Bloom, Ca .Zrbon,MJ. Cell. B/of. 13, 4884(1993).pacer spromatocyes Weahaen bpivalentsheire de- Mol. Cell. B/of. 10, 5721 (1990). 57. R. ED ietz, JCelloom(B/of.)47,2359(1970).tached romth metaphse spidle, thir ki- 16. 5. Stoler, K. C. Keith, K. E. Curnick, M. Fitzgerald- 8 .Dez hoooa(et)9 5 15)
netochores acquired strong 3F3/2 reactivity. If Hayes, Genes Dev. 9, 573 (1995). 59. R. B. Nicklas and C. A. Koch, J. Cell BioL 43, 40
these bivalents were then directed to reattach 17. K. F. Sullivan, M. Hechenberger, K. Masri, J. Cell (1969).B/of. 127, 581 (1994). 60. J. R. McIntosh, Cold Spring Harbor Symp. Quant.
to a single spindle pole (no kinetochore ten- 18. N. Steiner, K. M. Hahnenberger, L. Clarke, Mol. Cell. B/of. 56, 613 (1991).
sion), 3F3/2 reactivity remained high. How- B/of. 13, 4578 (1993). 61. F. Spencer and P. Hieter, Proc. Natl. Acad. Sci.
ever, if such mono-oriented bivalents were 19. R. Wevrick and H. F. Willard, Proc. Natl. Acad. Sci. U.S.A. 89, 8908 (1992).U.S.A. 86, 9394 (1989). 62. X. Li and R. B. Nicklas, Nature 373, 630 (1995).
then stretched with a microneedle 50 as to 20. B. Fishel, H. Amstitz, M. Baum, J. Carbon, L. Clarke, 63. C. L. Rieder, R. W. Cole, A. Khodjakov, G. Sluder, J.
bring one kinetochore under tension, the Mof. Cell. Biol. 8, 754 (1988); Y. Chikashige et at., Cell B/of. 130, 941 (1995).
amount of 3F3/2 epitope at the tense kineto- Cell 57, 739 (1989). 64. 0. J. Gorbsky and W. A. Ricketts, ibid. 122, 131121. M.- Baum, V. K. Ngan, L. Clarke, Mof. B/of. Cell5, 747 (1993).
chore underwent a marked decrease (66). (1994). 65. M. S. Campbell and G. J. Gorbsky, ibid. 129, 1195
Whereas recent studies have made 22. J. G. Marschall and L. Clarke, J. Cell B/of. 128, 445 (1995).
progress toward proving that kinetochores (1995). 66. R. B. Nicklas, S. C. Wards, G. J. Gorbsky, ibid. 130,
cnsend signals and that spindle tension 23. R. Ding, K. L. McDonald, J. R. McIntosh, ibid. 120, 929 (1995).can
~~~~~~~~~~~~~~~~141 (1993). 67. G. Sluder, F. J. Miller, E. A. Tomson, D. E. Wolf, ibid.
status affects kinetochore biochemistry, fun- 24. R. C. Allshire, J.-P. Javerzat, N. J. Redhead, G. 126,189 (1994).
damental questions remain. On the one Cranston, Cell 76, 157 (1994). 68. P.-Y. Goh and J. V. Kilmartin, ibid. 121, 503 (1993);
hn,studies with 3F3/2 and micromanipu- 25. R. C. Al/shire, E. R. Nimmo, K. Ekwall, J.-P. Javerzat, W. Jiang, J. Lechner, J. Carbon, ibid., p. 513.hand, ~~~~~~~~~~~~~~~G.Cranston, Genes Dev. 9, 218 (1995). 69. K. F. Doheny et at., Cell 73, 761 (1993).lation point to tension as a key factor in the 26. K. Ekwall et at., Science 269, 1429 (1995). 70. J. Lechner, EMBO J. 13, 5203 (1994); A. V. Strunni-
"wait anaphase" signal. On the other hand, 27. R. Paro and D. Hogness, Proc. Natl. Acad. Sci. kv .Knsuy .Ksln,J elBo.18 4
the laser ablation studies suggest that kinet- U.S.A, .88, mdt263 n1339(1991)4.oenz .Osemn,0 (1995).
ochore attachment rather than tension is 28. Flck H. Schmisadt, CGen 143,13 (1994.Cl.Bo6, 71. C. Connelly and P. Hieter, unpublished data.
imotn.When the unattached kineto- T3C.Jme6 ndS2C .19gn8Mt6Cl. /f.6 72. P. K. Sorger, F. F. Sever/n, A. A. Hyman, J. Cell B/of.important. 3862~~~~~~~~~29 (1986)ad . Jepsn hoooeRs ,25 127, 995 (1994).chore of a mono-oriented chromosome is 2 19.4L.NioanP.Jpen,CrmseRe.224 73. K. Middleton and J. Carbon, Proc. Natl. Acad. Sci.
abae,this does not put the attached kinet- 30
.SenradL Cak,Cl 9 6(1994). U.S.A. 91, 7212 (1994).
ochore under tension, yet it does tum off the 31. T. 0. Murphy and G. H. Karpen, ibid. 82, 599 (1995). 123,.T 387wn(193.GotcLH.arwl,JCelBo
ttwait anaphase" signal (63). These incon- 32. W. C. Earnshaw and N. Rothfield, Chromosoma 751P2B3Mlu and0993. Ksln,Mf /f el69
s i s t e n c i e s m a y b e d u e t o i n t r i n s i c differences). 5 P B Me uh a(B ert.)hla91 ,M o31 3o(19 8 5 ).79sistncie mabe ue t inrinsc difernces 33.H. Masumoto, H. Masukata, Y. Muro, N. Nozaki, T. (1995).
between meiosis and mitosis (67), or more Okazaki, J. Cell B/of. 109, 1963 (1989). 76. M. T. Brown, Gene 160,111 (1995).
subtle factors may be involved. 34. 0. Kipling et af., Mot. Cell. B/of. 15, 4009 (1995). 77. L. Wordeman and T. J. Mitch/son, J. Cell Biof. 128,
Clearly, no single structure or mechanism 35. 1. G. Goldberg, H. Sawhney, A. F. Pluta, P. W. War- 95 (1995).burton, W. C. Earnshaw, in preparation. 78. W. C. Earnshaw and C. A. Cooke, J. Cell Sci. 98,443
can yet be used to define all centromeres. This 36. W. C. Earnshaw, R. L. Bernat, C. A. Cooke, N. F. (1991).
variability reflects not only the complexity of Rothfield, Cold Spring Harbor Symp. Quant. B/of. 56, 79. C. A. Cooke, 0. P. Bazett-Jones, W. C. Earnshaw, J.
the functions attributed to the centromere, 675 (1991); H. Masumoto etal., in Chromosome Seg- B. Rattner, J. Cell B/of. 120, 1083 (1993).
but aso,udoubtdly, he reulatoy ciruits regation and Anneuploidy, B. K. Vig, Ed. (Springer- 80. C. A. Cooke, T. Yen, W. C. Earnshaw, unpublished
circuitsVerlag, Berlin, 1993), vol. H72, p. 31.ob e v t ns
involved in cell cycle checkpoints that are 3.7. T. Haaf, P. E. Warburton, H. F. Willard, Cell 70, 1 81. J. B. Rattner, A. Rao, M. J. Fritzler, D. W. Valencia, T.
influenced by the centromere. Continued (1992); Z. Larin, M. D. Fricker, C. Tyler-Smith, Hum. J. Yen, Cell Motif. Cytoskeleton 26, 214 (1993); C. A.
progessin tudingboth point and regional 38Mol. Genet. 3, 689 (1994). Casiano, G. Landberg, R. L. Ochs, E. M. Tan, J. Cell
.
38.K. E. Brown et at., ibid., p. 1227. Sc/. 106, 1045 (1993); X. Zhu et at., Mof. Cell. B/of.
 
o
n
 M
ay
 2
0,
 2
01
0 
w
w
w
.s
ci
en
ce
m
ag
.o
rg
D
ow
nl
oa
de
d 
fro
m