Centromeres put epigenetics in the
driver's seat
R. Kelly Dawe1,2
and Steven Henikoff3
1
Department of Plant Biology, University of Georgia, Athens, GA 30602, USA
2
Department of Genetics, University of Georgia, Athens, GA 30602, USA
3
Howard Hughes Medical Institute, Fred Hutchinson Cancer Research Center, 1100 Fairview Avenue N, Seattle, WA 98109-1024,
USA
A defining feature of chromosomes is the centromere,
the site for spindle attachment at mitosis and meiosis.
Intriguingly, centromeres of plants and animals are
maintained by both sequence-specific and sequenceindependent
(epigenetic) processes. Epigenetic inheritance
might enable kinetochores (the structures that
attach centromeres to spindles) to maintain an optimal
size. However, centromeres are susceptible to the evolution
of `selfish' DNA repeats that bind to kinetochore
proteins. We argue that such sequence-specific interactions
are evolutionarily unstable because they enable
repeat arrays to influence kinetochore size. Changes in
kinetochore size could affect the interaction of kinetochores
with the spindle and, in principle, skew Mendelian
segregation. We propose that key kinetochore
proteins have adapted to disrupt such sequence-specific
interactions and restore epigenetic inheritance.
The centromeric state
Every chromosome requires a kinetochore (see Glossary),
the proteinaceous structure that forms the interface
between centromeric DNA and the microtubules that pull
the chromosomes to the poles at mitosis. The kinetochore­
DNA interaction is mediated by a small group of proteins
that builds the kinetochore from the DNA up [1]. Recent
data show that both the DNA and proteins that comprise
plant and animal kinetochores are rapidly evolving, despite
the fact that the positions of centromeres are maintained for
millions of years. Moreover, human centromeres can form
spontaneously in chromosomal regions that lack centromere-specific
sequences, suggesting that centromeres are
maintained epigenetically. Here, we discuss how the evolutionary
forces that act on centromeres might have resulted
in their peculiar sequence organization.
The foundation of the kinetochore is the
centromere-specific nucleosome [2], which differs from
bulk nucleosomes by the presence of a histone H3 variant.
The first example of a centromere-specific H3 variant,
centromere protein (CENP)-A, was described as the smallest
of three centromere-specific proteins (CENP-A, -B and
-C) detected using anti-scleroderma antibodies [3,4].
CENP-A was later shown to be a core histone that replaces
ordinary histone H3 in centromeric nucleosomes [5,6]. In
yeast, all known components of the kinetochore are
dependent on the presence of a centromere-specific
nucleosome [1] and the available data from animals and
plants indicate that the same is true in complex eukaryotes.
Other conserved kinetochore proteins, CENP-C,
Mis12 and CENP-H, depend on the presence of centromeric
nucleosomes for localization [7­13]. These `foundation'
proteins [9] interact in an as yet unknown way to form a
higher-order chromatin complex that, in turn, recruits
dozens of transitory kinetochore proteins that are required
to ensure accurate chromosome segregation.
The high degree of divergence of CENP-A is surprising
in light of the absolutely conserved function of centromeric
nucleosomes and the near invariance of H3 in bulk chromatin.
CENP-A is significantly more closely related to H3
than to centromeric histones from distant clades, which
Review TRENDS in Biochemical Sciences Vol.31 No.12
Glossary
a satellite: The name of the 171-base-pair human centromeric repeat.
Centromeres: DNA sequences that interact with the kinetochore.
CEN chromatin: A term used here to describe the mixture of histone H3 and
CENH3 that underlies the kinetochore. The histone variant H3.3 might also be
present in CEN chromatin.
Cohesin: The protein complex that mediates sister-chromatid cohesion.
Epigenetics: Heritable changes in phenotype that are not caused by changes in
DNA sequence.
Gametogenesis: The formation of gametes, starting from meiosis and
proceeding through the differentiation of eggs and sperm.
Kinetochore: The proteinaceous structure that connects the centromere to the
spindle.
Kinetochore-foundation proteins: The small set of kinetochore proteins that
are intimately associated with centromeric DNA. Foundation proteins are
constitutive, that is, detectable at centromeres throughout the cell cycle. The
major (evolutionarily conserved) foundation proteins are CENH3, CENP-C,
Mis12 and CENP-H. Kinetochore-foundation proteins recruit outer kinetochore
proteins.
Meiotic drive: The preferential segregation of one chromosome over another.
The term is usually used to describe selfish DNA that accumulates by
circumventing Mendeľs rules.
Neocentromeres: Centromeres that arises spontaneously in a new position,
often in an area that does not contain repeat arrays.
Outer kinetochore proteins: A large collection of proteins that mediate,
facilitate or regulate the interaction of the kinetochore with the spindle. Outer
kinetochore proteins tend to be present only during cell division.
Pericentromeric heterochromatin: The deeply staining chromatin that flanks
CEN chromatin on either side. Unlike CEN chromatin, pericentromeric
heterochromatin does not have defined boundaries.
Repeat arrays: Long series of short DNA sequences arranged in tandem.
Centromeric repeat arrays are usually 150­180 base pairs and might extend
uninterrupted for hundreds of kilobases.
Selfish DNA: DNA that accumulates in the genome without affecting the
reproductive success of the organism.
Sister-chromatid cohesion: A term used to describe the tight association of
sister chromatids during mitosis and meiosis. At mitotic metaphase, sisterchromatid
cohesion is largely limited to pericentromeric regions.Corresponding author: Dawe, R.K. (kelly@plantbio.uga.edu).
Available online 30 October 2006.
www.sciencedirect.com 0968-0004/$ ­ see front matter ß 2006 Elsevier Ltd. All rights reserved. doi:10.1016/j.tibs.2006.10.004
reflects multiple evolutionary origins from a canonical
Histone 3. Evidently, these `CENH30
proteins in different
lineages have converged for centromeric function (Box 1).
CENH3s are homologous to histone H3 within the histone
core domain except for a small region that is required for
targeting CENH3 [14,15]. Another region of particular
interest is the N-terminal tail, which, at least on H3,
provides chromatin with a distinct identity that influences
gene expression. The N-terminal tail of CENH3 is strikingly
different from that of other H3s and displays almost
no homology across species. A tail is necessary for proper
centromere function [16] but, as yet, there is no evidence
that specific tail sequences are required (for instance, the
N-terminal tail of CENP-A can be replaced with unrelated
sequence [17]). Paradoxically, the CENH3 N-terminal tail
shows evidence of accelerated evolution, as if it were under
strong positive selection for an important function [18].
Similarly, CENP-C is evolving at remarkable rates and
shows evidence of adaptive evolution across all lineages
studied [19].
Why do eukaryotes tolerate (or favor) key kinetochore
proteins that change so rapidly? As we argue here, the
sequence variation in key kinetochore proteins is the outcome
of a complex interplay between histone deposition,
selfish DNA, and meiotic drive that enables the organism
to maintain Mendelian segregation of the chromosomal
DNA of the organism.
The unstable, transient nature of centromeric DNA
Simple repeats are the primary sequence of all
centromeres in complex eukaryotes. The a satellite (a
171-base-pair tandemly repeated sequence) in humans is
a well-studied example, but repeats of approximately the
same size can be found throughout the animal and plant
kingdoms. The basic structure of the repeat arrays is
similar among species but the base-pair level sequences
differ [20]. Thousands of simple repeats can evolve in
unison at rates that exceed those observed in non-coding
portions of genes [20,21]. This seems to be a dynamic
process because the lengths of the repeat arrays on a single
centromere can vary naturally by as much as tenfold
[22,23]. The most striking examples come from the rice
lineage, where entire centromere families have been
erased and replaced in unexpectedly short time frames
[24,25]. The primary mechanisms of change are presumed
to be unequal recombination and gene conversion, which
can cause the generation and spread of new repeats [26,27]
in addition to the loss of repeat arrays [28]. Most centromere
sequence data have been interpreted in the context of
unequal recombination [29­32], although direct measurements
of the rates and impact of recombination within
centromeres are as yet unavailable.
Although megabase-sized satellite sequence arrays are
characteristic features of higher eukaryotic centromeres,
they are not necessarily required for centromere function.
In human cells, new centromeres (`neocentromeres') can
appear sporadically at seemingly random sites that lack
satellite repeats [33], and, in other species, centromeres
move laterally to new sites that have little or no sequence
similarity [34,35]. Centromere plasticity is also revealed by
overexpressing human CENH3 (CENP-A), which leads to
expansion of centromeric chromatin [36]. Expansion can
occur over both higher-order a-satellite repeats and active
genes, indicating that there is no sequence or compositional
specificity requirements for CENH3-containing
(CEN) chromatin formation.
Spatial differentiation within CEN chromatin
The distinction between CEN chromatin that encompasses
the kinetochore-forming region and the surrounding pericentromeric
heterochromatin represents the first level of
chromatin differentiation at the centromere [37] (Figure 1).
At most centromeres, this chromatin distinction seems to
correspond to a differentiation between young and old
DNA sequences. The DNA sequences in CEN chromatin
are young, consisting of nearly identical copies of satellite
repeats that are constantly being homogenized by genetic
exchange [30,38]. CENH3 is usually found within the most
homogeneous satellite sequences [20], whereas the DNA in
pericentromeres is much less uniform and contains ancient
diverged satellite arrays that are peppered by transposons
Box 1. Evolutionary convergence of histones
Phylogeny is a highly successful predictor of function. For example,
kinesins comprise a family of microtubule-based motor proteins
that have diverged to acquire numerous roles in cellular processes;
by organizing kinesins according to subfamilies, the roles of
otherwise unknown kinesin family members have been successfully
predicted. When kinesins were renamed by the community to
enforce common usage, it made sense to classify them by the
subfamily, regardless of their various historical designations [81].
By contrast, the family of proteins related to histone H3 is unique in
that function is not predicted by phylogeny because multiple
lineages of histones have converged to the same function, rather
than diverging from a single ancestor [18]. This discrepancy
between phylogeny and function raises questions as to how these
histones should be classified and named.
In most eukaryotic lineages, there are three ancestrally related
forms of Histone 3 [82] with H3 being the usual designation for the
major form, H3.3 for the replacement form and CENP-A for the
centromeric form. Grouping all H3 histones together and all H3.3
histones together is convenient, but it violates phylogeny as there
seems to be at least four independent origins of the protein pair (in
animals and fungi, plants, ciliates and apicomplexans) [18]. Similarly,
there is evidence that centromeric forms have evolved multiple times
from an H3 or H3.3 ancestor [18]. For example, the original name for
the Arabidopsis centromeric histone was HTR12 (for histone three
number 12), assigned by the Plant Chromatin project, and HTR12 was
retained when its centromeric localization was demonstrated [83]. It
was clear, however, that a name that implies centromeric function is
preferred, and CENH3 was suggested as a suitable `umbrella' term
[83]. Unlike CENP-A, the CENH3 term does not imply a family
relationship. CENH3 retains the first syllable of CENP-A (for function
and localization) and unites it with H3 (for phylogeny), thus it follows
the precedence of using a prefix before the histone name to describe
unusual histones, as was done for the histone H2A variant,
macroH2A. CENH3 has since become the generally adopted term
for plants [84], ciliates [85] and apicomplexans [86]; however,
historical usage prevails for budding yeast, where the designation,
Cse4 is usually retained in the literature [1].
Therefore, we aim to promote the adoption of CENH3 in the field
while recognizing that, when searching the literature, researchers
should still be able to find articles using the alternative and
historical names for these proteins. To that end, the National Center
for Biotechnology Information (NCBI; http://www.ncbi.nlm.nih.gov)
has implemented an automatic link between the terms CENP-A and
CENH3 so that searching PubMed using either returns all references
for both.
Review TRENDS in Biochemical Sciences Vol.31 No.12 663
www.sciencedirect.com
[30,39]. For example, all human centromeres are
embedded within homogeneous 171-base-pair a-satellite
repeats and Arabidopsis centromeres are similarly
embedded within a different 178-base-pair satellite repetitive
array [21,40]. Pericentromeric regions tend to be
heterochromatic and enriched in cohesin, which holds
chromatids together until anaphase [41] (Figure 1). The
heterochromatic state might be required to recruit pericentromeric
cohesin in some species [41,42], but active
genes (presumably within euchromatin) can also exist
within centromeric and pericentromeric domains [43,44].
Recent work has highlighted the fact that H3-containing
nucleosomes are dispersed within the centromere core
domains [45], although the modifications on these specialized
histone H3s are not typical of heterochromatin
[37,46]. CENH3 and H3 seem to be dispersed in long
alternating arrays. A combination of fluorescent in situ
hybridization and antibody staining suggests that the sizes
of the individual CENH3- and H3-containing nucleosomal
arrays are 15­40 kb in Drosophila [45]. CENH3-containing
arrays are thought to face outwards towards microtubules,
whereas H3-containing arrays comprise the
internal regions where sister-chromatid cohesion
takes place [45,47] (Figure 1). Chromatin immunoprecipitation
analyses of two centromeres (the 10q25 human
neocentromere and rice centromere 8) have provided
important confirmation of the idea of alternating CENH3
and H3 arrays [44,48]. In the case of the human 10q25
neocentromere, the alternating arrays are 10­50 kilobases,
matching prior estimates from cytological analysis
[48].
Just how alternating histone patterns are maintained
in centromeres is unknown, although it is interesting to
consider the possibility that they are laid down during
nucleosome assembly. Like centromeres, chromosome
arms are differentiated by alternative H3 variants, which
are deposited by distinct nucleosomes-assembly pathways
[49,50]. Histone H3 is exclusively deposited at replication,
whereas H3.3 is deposited at transcriptionally active genes
throughout the cell cycle. H3 is frequently methylated at
Lys9 and Lys27, whereas H3.3 is usually acetylated at lys9
and methylated at Lys4 [51­54]. It is not known whether
the centromeric arrays of canonical nucleosomes contain
H3 or H3.3, although the fact that Schizosaccharomyces
pombe encodes only the H3.3 form [55] and their centromeres
show enrichment of dimethyl-Lys4 [56] indicates
that it could be the H3.3 replication-independent form
that separates CENH3 arrays. If so, then centromeric
chromatin would consist entirely of nucleosomes that
are deposited independently of replication (containing
CENH3 and H3.3), whereas the surrounding heterochromatin
consists of nucleosomes that are deposited during
replication (containing H3).
Dynamics of CEN chromatin
Biochemical purification of the soluble CENH3 nucleosome
pre-assembly complex from Drosophila has yielded a
simple trimolecular complex that consists of CID (the
Drosophila CENH3), histone H4 and RbAp48 (a histone
chaperone protein found also in the H3 and H3.3 assembly
complexes) [57]. The lack of any centromere-specific protein
in the soluble complex that deposits CENH3, except
for CENH3 itself, suggests either or both of two possible
explanations: (i) CENH3 could be deposited wherever
there is pre-existing CENH3 on chromatin, and/or (ii)
CENH3 could deposit promiscuously but can only gain
access for assembly onto DNA at centromeres. There is
evidence that is consistent with both of these possibilities.
On the one hand, biochemical fractionation of human
centromeric chromatin (as opposed to the soluble preassembly
complex) using either CENH3 (CENP-A) or
CENP-H reveals complex sets of centromere-specific proteins
[11,58], which could create a structure that would
enable recognition by the soluble CENP-A-containing
assembly complex. On the other hand, overexpression of
CENP-A or CID causes promiscuous assembly throughout
euchromatin [49,59,60], which might suggest a nonsequence-specific
gap-filling process. A promiscuous gapfilling
process could also explain the de novo formation of
artificial chromosomes [61] and neocentromeres [33­35]. It
seems likely that both mechanisms are used to assure the
faithful assembly of CENH3-containing nucleosomes [60].
It is now realized that chromatin is far more dynamic
than once thought. Heterochromatin-associated proteins
are in constant flux [62,63] and heterochromatin formation,
at least in S. pombe, requires cycles of transcription
Figure 1. Metaphase kinetochores in their chromosomal context. The larger
centromere­kinetochore domain includes pericentromeric heterochromatin and
cohesin. The CEN chromatin, which is poorly understood, is composed of a
mixture of histone H3 and CENH3 (and maybe histone H3.3). The relationship
between histone H3 and cohesion in plants and animals is unclear. There is at least
a spatial correlation between heterochromatin and cohesion, but it is not known
whether cohesin interacts with CEN chromatin.
664 Review TRENDS in Biochemical Sciences Vol.31 No.12
www.sciencedirect.com
followed by small interfering RNA (siRNA)-directed
silencing [64]. Similarly, maize centromeric repeats are
transcribed and the resulting centromeric transcripts are
bound to CENH3 [65]. When CENH3 is overexpressed,
centromeric chromatin can expand and invade surrounding
heterochromatin [36]; and, vice versa, when CENH3 is
underexpressed, heterochromatin invades centromeres
[45]. Expansion and contraction of epigenetic states might
also occur in normal centromeric chromatin because both
CENH3 and H3 can be detected on the same single-copy
DNA sequences in rice centromere 8 [66]. These data
indicate that the deposition machineries for CENH3-containing
and canonical nucleosomes compete with each
other during assembly. A built-in competition between
CENH3 and H3 (or H3.3) would cause regular jostling of
the two forms of nucleosomes during deposition that would
tend to break up large blocks of CENH3 (Figure 2). The
requirement for kinetochore attachment on one face of the
centromere and cohesion on the other [41,67­70] should
lead to a natural parity between CENH3- and H3-containing
arrays and strong selection against either assembly
process dominating.
Evolution of centromere repeats through epigenetic
selection
Given that tandem-repeat arrays are common at nearly all
centromeres, it seems likely that their size, sequence or
arrangement contributes in some manner to centromere
function. In human cells, arrays of a satellites are sufficient
to organize kinetochores within artificial chromosomes
[61,71]. The a satellite contains a key sequence
known as the CENP-B box, which binds in a sequencespecific
manner to the CENP-B protein and facilitates
kinetochore formation [72]. However, CENP-B (and its
DNA-binding motif) is absent at many human, plant and
numerous animal centromeres [73,74]. On CENP-B-free
human centromeres, and in species such as flies,
nematodes and flowering plants, the available data are
consistent with an epigenetic mode of centromere specification.
What drives the evolution of centromere-specific
repeats when epigenetics is the determining factor?
Most centromere repeats are approximately the
length of DNA that wraps around a single nucleosome
(150­180 base pairs), suggesting that satellites might
have evolved to facilitate regular nucleosome packaging.
Another possibility is that nucleosomes limit unequal
recombination to small regions of the packaged chromatin
[64,75] such that the repeat unit is approximately
equivalent to a recombination unit. However, although
such models help to explain the size of the repeats,
they do not offer an explanation for the uniformity of
the sequences, that is, the fact that the same repeats
are usually found among all centromeres within a species
(stochastic models such as those proposed by
Smith [26] only explain the spread of repeats within
centromeres).
The model we propose is that centromeric DNA behaves
selfishly and adapts to the local chromatin environment.
Species-specific differences in kinetochore-foundation proteins
would be expected to present unique DNA-binding
faces and the opportunity for evolving to better fit those
faces. Furthermore, there should be strong competition
among repeats (or arrays of repeats) to adapt more efficiently
because the repeats that bind tightest are most
likely to be connected to the spindle apparatus. Because all
kinetochores within a species will have the same protein
features, such a competition mechanism would be expected
to drive centromeric repeat arrays to uniformity. In this
model, the CENH3-centered epigenetic state is the first
stage in centromere evolution and the formation and
spread of centromere-repeat arrays is a consequence
[25,66].
Figure 2. An epigenetic-competition model for centromere organization. (a) (i) We envision a situation whereby CENH3- and H3-containing nucleosomes compete or `jostle'
for placement over large expanses of similar centromere repeats. Arrowheads indicate that the two forms of chromatin often invade each other's territory. (ii) During
metaphase, the CENH3-containing portion interacts with kinetochore-foundation proteins (right). (b) Three arrangements of chromatin over the same DNA sequence (i),
together with representations of how the chromatin might fold into metaphase kinetochores in each case (ii). Dots represent satellite repeat arrays and solid lines represent
the older, more heterogeneous sequences typical of pericentromeric regions.
Review TRENDS in Biochemical Sciences Vol.31 No.12 665
www.sciencedirect.com
Meiotic drive promotes the epigenetic mode of
centromere inheritance
Why did this odd epigenetic mechanism evolve in
complex eukaryotes but not in small genome species
such as S. cerevisiae and S. pombe? We suspect that
the answer lies in the mechanics of gametogenesis. In
fungi, all four products of meiosis are functional,
whereas in animals and plants only one product of
female meiosis survives to make an egg. Meiotic inequality
presents the opportunity for the centromeres to be
`abuseď by meiotic drive [20,76,77]. In principle,
increases in the size of centromere-repeat arrays could
lead to increases in kinetochore size, increased capacity
of the kinetochores to interact with microtubules and an
increased likelihood that a chromosome is segregated to
the egg (Figure 3a). Such deviation from Mendelian
inheritance limits the capacity of a species to adapt to
different environments.
Under conditions that enable centromere-mediated
meiotic drive (Figure 3), the organism might find itself
with a bad (unfit) chromosome that it cannot get rid of
(segregate away). Mutations in key foundation proteins
such as CENH3 or CENP-C could relieve the imbalance
between strong and weak centromeres. The example illustrated
in Figure 3 shows a mutated CENP-C that binds to
existing centromere sequence in addition to a new (neocentromeric)
sequence near the normal centromere
(Figure 3b). The neocentromere sequence might be a
slightly different variant of the native centromeric array,
a dissimilar array that has transposed from another chromosome
[78], or even a completely novel satellite repeat,
such as is found at Drosophila centromeres [79]. The
combined pull of the normal centromere and the nearby
neocentromere would restore Mendelian segregation and
provide a fitness advantage for both the neocentromerecontaining
chromosome and the mutant foundation protein.
Both the new centromere and mutant kinetochore
protein would then rapidly drive to fixation.
The mutation that best fits this scenario is one that
reduces sequence specificity of a kinetochore-foundation
protein, so that the protein then binds to both the normal
centromere and the neocentromere. Any reduction in
sequence specificity will effectively loosen the influence
of DNA on kinetochore size and favor the epigenetic mode
that promotes uniform kinetochores (Figure 2). CENP-C is
a prime candidate for mediating this process because it is a
DNA-binding protein [80] that shows evidence of recurrent
positive selection over most of its length in both plants and
animals [19]. In organisms in which all four products of
meiosis are functional, genetically specified centromeres
have evolved and the associated foundation proteins show
strong purifying selection. However, where centromere
drive poses the threat of extinction, any tendency towards
genetic specification would be counteracted by the need to
Figure 3. Meiotic drive favors epigenetic centromeres. A centromere that expands is proposed to gain an advantage at female meiosis by achieving an orientation towards
the egg pole [20]. (a) The expanded centromere increases its representation in the population by accumulating more foundation proteins (represented by CENP-C; green)
and making more microtubule connections, causing it to orientate towards the egg pole (curved arrow) even when it starts out in an unfavorable position. However, the
resulting departure from Mendelian segregation would be expected to limit the capacity of the species to adapt to new environments. (b) Normal segregation can be
restored if the driving centromere encounters a mutant CENP-C (blue) with reduced DNA-binding specificity that binds to previously non-centromeric sequence. This results
in meiotic parity, providing an advantage for the gene encoding the mutant foundation protein.
666 Review TRENDS in Biochemical Sciences Vol.31 No.12
www.sciencedirect.com
relieve drive, thus putting epigenetic centromeres back in
the driver's seat.
The process should be cyclical (Figure 4). The most
stable state is presumably epigenetic, but the opportunity
always exists for selfish repeats to evolve within the
centromeric chromatin environment. Competition among
selfish DNAs will push the system towards genetic specification
and again make centromeres susceptible to meiotic
drive. Recurrent cycles of drive and suppression should
cause the rapid divergence of both centromeric sequence
arrays and foundation proteins. Such an evolutionary
interplay between epigenetic and genetic states can
explain most of the available data on centromeres. However,
many of the underlying assumptions have yet to be
rigorously tested. Future work will be focused on showing
that chromosome transmission can be influenced by
inequities in kinetochore size, that gene conversion and/
or unequal recombination occur at high frequencies within
centromeric repeat arrays, and that some kinetochorefoundation
proteins (in addition to CENP-B) have
sequence-specific binding affinities for centromeric
repeats.
Acknowledgements
We thank David Landsman and Kathi Canese at the NCBI for their
assistance in getting modifications described in Box 1 made to the
PubMed query engine.
References
1 Collins, K.A. et al. (2005) De novo kinetochore assembly requires the
centromeric histone H3 variant. Mol. Biol. Cell 16, 5649­5660
2 Palmer, D.K. and Margolis, R.L. (1985) Kinetochore components
recognized by human autoantibodies are present on
mononucleosomes. Mol. Cell. Biol. 5, 173­186
3 Guldner, H.H. et al. (1984) Human anti-centromere sera recognise a
19.5 kD non-histone chromosomal protein from HeLa cells. Clin. Exp.
Immunol. 58, 13­20
4 Earnshaw, W.C. and Rothfield, N. (1985) Identification of a family of
human centromere proteins using autoimmune sera from patients with
schleroderma. Chromosoma 91, 313­321
5 Palmer, D.K. et al. (1987) A 17-kD centromere protein (CENP-A)
copurifies with nucleosome core particles and with histones. J. Cell
Biol. 104, 805­815
6 Palmer, D.K. et al. (1991) Purification of the centromere-specific
protein CENP-A and demonstration that it is a distinctive histone.
Proc. Natl. Acad. Sci. U. S. A. 88, 3734­3738
7 De Wulf, P. et al. (2003) Hierarchical assembly of the budding
yeast kinetochore from multiple subcomplexes. Genes Dev. 17,
2902­2921
8 Goshima, G. et al. (2003) Human centromere chromatin protein
hMis12, essential for equal segregation, is independent of CENP-A
loading pathway. J. Cell Biol. 160, 25­39
9 Amor, D.J. et al. (2004) Building the centromere: from foundation
proteins to 3D organization. Trends Cell Biol. 14, 359­368
10 Kline, S.L. et al. (2006) The human Mis12 complex is required for
kinetochore assembly and proper chromosome segregation. J. Cell Biol.
173, 9­17
11 Okada, M. et al. (2006) The CENP-H-I complex is required for the
efficient incorporation of newly synthesized CENP-A into centromeres.
Nat. Cell Biol. 8, 446­457
12 Emanuele, M.J. et al. (2005) Measuring the stoichiometry and physical
interactions between components elucidates the architecture of the
vertebrate kinetochore. Mol. Biol. Cell 16, 4882­4892
13 Meraldi, P. et al. (2006) Phylogenetic and structural analysis
of centromeric DNA and kinetochore proteins. Genome Biol. 7,
R23
14 Vermaak, D. et al. (2003) Maintenance of chromatin states: an
open-and-shut case. Curr. Opin. Cell Biol. 15, 266­274
15 Black, B.E. et al. (2004) Structural determinants for generating
centromeric chromatin. Nature 430, 578­582
16 Van Hooser, A.A. et al. (2001) Specification of kinetochore-forming
chromatin by the histone H3 variant CENP-A. J. Cell Sci. 114,
3529­3542
Figure 4. Epigenetics is in the driver's seat of a cyclical process. The fact that centromeres are directly attached to spindles promotes the evolution of selfish repeats that
bind tightly to foundation proteins. The bent, tightly bound region of DNA is shown interacting with CENP-C as an example (green). In principle, any of the foundation
proteins could be involved in this process. Such sequence-dependent (genetic) interactions between centromeric DNA and the kinetochore are susceptible to meiotic drive.
A mutation that changes the structure of CENP-C (blue) would disrupt meiotic drive and return the centromeres to a more stable epigenetic mode of centromere
specification. Adaptive mutations are presumed to change the shape of foundation proteins such that the selfish DNA no longer binds with high affinity. The process should
be cyclical, driving the rapid evolution of both centromeric repeats and kinetochore-foundation proteins.
Review TRENDS in Biochemical Sciences Vol.31 No.12 667
www.sciencedirect.com
17 Wieland, G. et al. (2004) Functional complementation of human
centromere protein A (CENP-A) by Cse4p from Saccharomyces
cerevisiae. Mol. Cell. Biol. 24, 6620­6630
18 Malik, H.S. and Henikoff, S. (2003) Phylogenomics of the nucleosome.
Nat. Struct. Biol. 10, 882­891
19 Talbert, P.B. et al. (2004) Adaptive evolution of centromere proteins in
plants and animals. J. Biol. 3, 18
20 Henikoff, S. et al. (2001) The centromere paradox: stable inheritance
with rapidly evolving DNA. Science 293, 1098­1102
21 Hall, S.E. et al. (2003) Centromere satellites from Arabidopsis
populations: maintenance of conserved and variable domains.
Genome Res. 13, 195­205
22 Mahtani, M.M. and Willard, H.F. (1990) Pulsed-field gel analysis
of alpha-satellite DNA at the human X chromosome centromere:
high-frequency polymorphisms and array size estimate. Genomics 7,
607­613
23 Kato, A. et al. (2004) Chromosome painting using repetitive DNA
sequences as probes for somatic chromosome identification in maize.
Proc. Natl. Acad. Sci. U. S. A. 101, 13554­13559
24 Lee, H.R. et al. (2005) Chromatin immunoprecipitation cloning reveals
rapid evolutionary patterns of centromeric DNA in Oryza species. Proc.
Natl. Acad. Sci. U. S. A. 102, 11793­11798
25 Dawe, R.K. (2005) Centromere renewal and replacement in the plant
kingdom. Proc. Natl. Acad. Sci. U. S. A. 102, 11573­11574
26 Smith, G.P. (1976) Evolution of repeated DNA sequences by unequal
crossover. Science 191, 528­535
27 Schindelhauer, D. and Schwarz, T. (2002) Evidence for a fast,
intrachromosomal conversion mechanism from mapping of
nucleotide variants within a homogeneous a-satellite DNA array.
Genome Res. 12, 1815­1826
28 Walsh, J. (1987) Peristence of tandem arrays: implications for satellite
and simple-sequence DNAs. Genetics 115, 553­567
29 Warburton, P.E. and Willard, H.F. (1992) PCR amplification of
tandemly repeated DNA: analysis of intra- and interchromosomal
sequence variation and homologous unequal crossing-over in human
a satellite DNA. Nucleic Acids Res. 20, 6033­6042
30 Schueler, M.G. et al. (2001) Genomic and genetic definition of a
functional human centromere. Science 294, 109­115
31 Hall, S.E. et al. (2005) Differential rates of local and global
homogenization in centromere satellites from Arabidopsis relatives.
Genetics 170, 1913­1927
32 Ma, J. and Bennetzen, J.L. (2006) Recombination, rearrangement,
reshuffling, and divergence in a centromeric region of rice. Proc.
Natl. Acad. Sci. U. S. A. 103, 383­388
33 Warburton, P.E. (2004) Chromosomal dynamics of human
neocentromere formation. Chromosome Res. 12, 617­626
34 Maggert, K.A. and Karpen, G.H. (2001) The activation of a
neocentromere in Drosophila requires proximity to an endogenous
centromere. Genetics 158, 1615­1628
35 Nasuda, S. et al. (2005) Stable barley chromosomes without
centromeric repeats. Proc. Natl. Acad. Sci. U. S. A. 102, 9842­9847
36 Lam, A.L. et al. (2006) Human centromeric chromatin is a dynamic
chromosomal domain that can spread over noncentromeric DNA. Proc.
Natl. Acad. Sci. U. S. A. 103, 4186­4191
37 Sullivan, B.A. and Karpen, G.H. (2004) Centromeric chromatin
exhibits a histone modification pattern that is distinct from
both euchromatin and heterochromatin. Nat. Struct. Mol. Biol. 11,
1076­1083
38 Roizes, G. (2006) Human centromeric alphoid domains are periodically
homogenized so that they vary substantially between homologues.
Mechanism and implications for centromere functioning. Nucleic
Acids Res. 34, 1912­1924
39 Copenhaver, G.P. et al. (1999) Genetic definition and sequence analysis
of Arabidopsis centromeres. Science 286, 2468­2474
40 Hosouchi, T. et al. (2002) Physical map-based sizes of the centromeric
regions of Arabidopsis thaliana chromosomes 1, 2, and 3. DNA Res. 9,
117­121
41 Bernard, P. et al. (2001) Requirement of heterochromatin for cohesion
at centromeres. Science 294, 2539­2542
42 Nakashima, H. et al. (2005) Assembly of additional heterochromatin
distinct from centromere-kinetochore chromatin is required for de
novo formation of human artificial chromosome. J. Cell Sci. 118,
5885­5898
43 Saffery, R. et al. (2003) Transcription within a functional human
centromere. Mol. Cell 12, 509­516
44 Yan, H. et al. (2005) Transcription and histone modifications in the
recombination-free region spanning a rice centromere. Plant Cell 17,
3227­3238
45 Blower, M. et al. (2002) Conserved organization of centromeric
chromatin in flies and humans. Dev. Cell 2, 319­330
46 Shi, J. and Dawe, R.K. (2006) Partitioning of the maize epigenome by
the number of methyl groups on histone H3 lysines 9 and 27. Genetics
173, 1571­1583
47 Zinkowski, R. et al. (1991) The centromere-kinetochore complex: a
repeat subunit model. J. Cell Biol. 113, 1091­1110
48 Chueh, A.C. et al. (2005) Variable and hierarchical size distribution of
L1-retroelement-enriched CENP-A clusters within a functional human
neocentromere. Hum. Mol. Genet. 14, 85­93
49 Ahmad, K. and Henikoff, S. (2002) Histone H3 variants specify
modes of chromatin assembly. Proc. Natl. Acad. Sci. U. S. A. 99,
6477­6484
50 Tagami, H. et al. (2004) Histone H3. 1 and H3. 3 complexes mediate
nucleosome assembly pathways dependent or independent of DNA
synthesis. Cell 116, 51­61
51 Waterborg, J.H. (1990) Sequence analysis of acetylation and
methylation in two histone H3 variants of alfalfa. J. Biol. Chem.
265, 17157­17161
52 Johnson, L. et al. (2004) Mass spectrometry analysis of Arabidopsis
histone H3 reveals distinct combinations of post-translational
modifications. Nucleic Acids Res. 32, 6511­6518
53 McKittrick, E. et al. (2004) Histone H3.3 is enriched in covalent
modifications associated with active chromatin. Proc. Natl. Acad.
Sci. U. S. A. 101, 1525­1530
54 Hake, S.B. et al. (2006) Expression patterns and post-translational
modifications associated with mammalian histone H3 variants. J. Biol.
Chem. 281, 559­568
55 Choi, E.S. et al. (2005) Dynamic regulation of replication independent
deposition of histone H3 in fission yeast. Nucleic Acids Res. 33,
7102­7110
56 Cam, H.P. et al. (2005) Comprehensive analysis of heterochromatinand
RNAi-mediated epigenetic control of the fission yeast genome. Nat.
Genet. 37, 809­819
57 Furuyama, T. et al. (2006) Chaperone-mediated assembly of
centromeric chromatin in vitro. Proc. Natl. Acad. Sci. U. S. A. 103,
6172­6177
58 Foltz, D.R. et al. (2006) The human CENP-A centromeric
nucleosome-associated complex. Nat. Cell Biol. 8, 458­469
59 Shelby, R.D. et al. (1997) Assembly of CENP-A into centromeric
chromatin requires a cooperative array of nucleosomal DNA contact
sites. J. Cell Biol. 136, 501­513
60 Heun, P. et al. (2006) Mislocalization of the Drosophila centromerespecific
histone CID promotes formation of functional ectopic
kinetochores. Dev. Cell 10, 303­315
61 Harrington, J.J. et al. (1997) Formation of de novo centromeres and
construction of first-generation human artificial microchromosomes.
Nat. Genet. 15, 345­355
62 Cheutin, T. et al. (2003) Maintenance of stable heterochromatin
domains by dynamic HP1 binding. Science 299, 721­725
63 Festenstein, R. et al. (2003) Modulation of heterochromatin protein 1
dynamics in primary mammalian cells. Science 299, 719­721
64 Verdel, A. et al. (2004) RNAi-mediated targeting of heterochromatin by
the RITS complex. Science 303, 672­676
65 Topp, C.N. et al. (2004) Centromere-encoded RNAs are integral
components of the maize kinetochore. Proc. Natl. Acad. Sci. U. S. A.
101, 15986­15991
66 Nagaki, K. et al. (2004) Sequencing of a rice centromere reveals active
genes. Nat. Genet. 36, 138­145
67 Wines, D.R. and Henikoff, S. (1992) Somatic instability of a Drosophila
chromosome. Genetics 131, 683­691
68 Hoque, M.T. and Ishikawa, F. (2002) Cohesin defects lead to
premature sister chromatid separation, kinetochore dysfunction, and
spindle-assembly checkpoint activation. J. Biol. Chem. 277,
42306­42314
69 Shibata, F. and Murata, M. (2004) Differential localization of the
centromere-specific proteins in the major centromeric satellite of
Arabidopsis thaliana. J. Cell Sci. 117, 2963­2970
668 Review TRENDS in Biochemical Sciences Vol.31 No.12
www.sciencedirect.com
70 Zhang, X. et al. (2005) Phosphoserines on maize centromeric histone H3
and histone H3 demarcate the centromere and pericentromere during
chromosome segregation. Plant Cell 17, 572­583
71 Larin, Z. and Mejia, J.E. (2002) Advances in human artificial
chromosome technology. Trends Genet. 18, 313­319
72 Masumoto, H. et al. (2004) The role of CENP-B and a-satellite DNA: de
novo assembly and epigenetic maintenance of human centromeres.
Chromosome Res. 12, 543­556
73 Earnshaw, W.C. et al. (1989) Visualization of centromere proteins
CENP-B and CENP-C on a stable dicentric chromosome in
cytological spreads. Chromosoma 98, 1­12
74 Saffery, R. et al. (2000) Human centromeres and neocentromeres show
identical distribution patterns of >20 functionally important
kinetochore-associated proteins. Hum. Mol. Genet. 9, 175­185
75 Baumann, M. et al. (2003) Regulation of V(D)J recombination by
nucleosome positioning at recombination signal sequences. EMBO J.
22, 5197­5207
76 Dawe, R.K. and Hiatt, E.N. (2004) Plant neocentromeres: fast, focused,
and driven. Chromosome Res. 12, 655­669
77 Burt, A. and Strivers, R. (2006) Female drive. In Genes in Conflict:
The Biology of Selfish Genetic Elements, Harvard University Press, pp.
301­321
78 Puechberty, J. et al. (1999) Genetic and physical analyses of the
centromeric and pericentromeric regions of human chromosome 5:
recombination across 5cen. Genomics 56, 274­287
79 Lohe, A.R. et al. (1993) Mapping simple repeated DNA sequences in
heterochromatin of Drosophila melanogaster. Genetics 134, 1149­1174
80 Sugimoto, K. et al. (1994) Human centromere protein C (CENP-C) is a
DNA-binding protein which possesses a novel DNA-binding motif. J.
Biochem. (Tokyo) 116, 877­881
81 Lawrence, C.J. et al. (2004) A standardized kinesin nomenclature. J.
Cell Biol. 167, 19­22
82 Luger, K. et al. (1997) Crystal structure of the nucleosome core particle
at 2.8 A° resolution. Nature 389, 251­260
83 Talbert, P.B. et al. (2002) Centromeric localization and adaptive
evolution of an Arabidopsis histone H3 variant. Plant Cell 14,
1053­1066
84 Houben, A. and Schubert, I. (2003) DNA and proteins of plant
centromeres. Curr. Opin. Plant Biol. 6, 554­560
85 Cervantes, M.D. et al. (2006) The CNA1 histone of the ciliate
Tetrahymena thermophila is essential for chromosome segregation
in the germline micronucleus. Mol. Biol. Cell 17, 485­497
86 Miao, J. et al. (2006) The malaria parasite Plasmodium falciparum
histones: organization, expression, and acetylation. Gene 369, 53­65
Five things you might not know about Elsevier
1.
Elsevier is a founder member of the WHO's HINARI and AGORA initiatives, which enable the
worlďs poorest countries to gain free access to scientific literature. More than 1000 journals,
including the Trends and Current Opinion collections and Drug Discovery Today, are now available
free of charge or at significantly reduced prices.
2.
The online archive of Elsevier's premier Cell Press journal collection became freely available in
January 2005. Free access to the recent archive, including Cell, Neuron, Immunity and Current
Biology, is available on ScienceDirect and the Cell Press journal sites 12 months after articles are
first published.
3.
Have you contributed to an Elsevier journal, book or series? Did you know that all our authors are entitled to a 30%
discount on books and stand-alone CDs when ordered directly from us? For more information, call our sales offices:
+1 800 782 4927 (USA) or +1 800 460 3110 (Canada, South and Central America)
or +44 (0)1865 474 010 (all other countries)
4.
Elsevier has a long tradition of liberal copyright policies and for many years has permitted both the
posting of preprints on public servers and the posting of final articles on internal servers. Now,
Elsevier has extended its author posting policy to allow authors to post the final text version of
their articles free of charge on their personal websites and institutional repositories or websites.
5.
The Elsevier Foundation is a knowledge-centered foundation that makes grants and contributions
throughout the world. A reflection of our culturally rich global organization, the Foundation has,
for example, funded the setting up of a video library to educate for children in Philadelphia,
provided storybooks to children in Cape Town, sponsored the creation of the Stanley L. Robbins
Visiting Professorship at Brigham and Women's Hospital, and given funding to the 3rd
International Conference on Children's Health and the Environment.
Review TRENDS in Biochemical Sciences Vol.31 No.12 669
www.sciencedirect.com