REVIEW ARTICLE
Histones in functional diversification
Core histone variants
Rama-Haritha Pusarla and Purnima Bhargava
Centre for Cellular & Molecular Biology, Tarnaka, Hyderabad, India
Introduction
Eukaryotic cells package their DNA in the form of
chromatin to accommodate it in the small space provi-
ded by their nuclei [1]. In spite of the 10 000-fold com-
paction of DNA due to this packaging, minute details
of a local structure regulate the accessibility of any
small region. The folding of 147 bp of DNA over a
histone octamer (two molecules each of the four core
histones, H4, H3, H2A and H2B) surface gives a neat
organization of the DNA into a chromatin fibre of
10 nm diameter. The primary structure of 10 nm chro-
matin has a characteristic ‘beads on a string’ appear-
ance. This uniformity of the nucleosomal chain might
impose difficulties in region-specific, localized recogni-
tion and in uncoiling of the structure; both essential
for function. Thus, higher order folding of the chroma-
tin into a 30 nm fibre and larger domains could be an
attempt by the genome to demarcate itself into various
regions of activities.
Histones are abundant, basic, structural proteins
that bring in variety and novelty to the complicated
gene regulation mechanisms [1]. Apart from binding to
DNA and giving chromatin its strength, stability and
form, certain highly similar forms of histones, termed
‘histone variants’, have evolved to carry out many vital
functions. Though the focus on histone variants

nuclear processes and the structure of different variant nucleosome cores
shows that this may indeed be so. Histone variants may also be involved in
demarcating functional regions of the chromatin. We discuss in this review
why two of the four core histones show higher variation. A comparison of
the status of variants in yeast with those from higher eukaryotes suggests
that histone variants have evolved in synchrony with functional require-
ment of the cell.
Abbreviations
Cid, centromere identifier; DSB, double strand break; IRIF, irradiation induced foci; MSCI, meiotic sex chromosome inactivation;
NHEJ, nonhomologous end joining; RC, replication coupled; RI, replication independent.
FEBS Journal 272 (2005) 5149–5168 ª 2005 FEBS 5149
modifications or ATP-dependent chromatin remodel-
ling [8–10] are joined now by histone variants. This
review focusses mainly on new advances in chromatin-
related processes with reference to the ‘core histone
variants’ and their contribution to chromatin structure.
Other aspects, including the role of linker histone vari-
ants, can be found in other recent reviews [11–13].
Variation in high conservation – the
evolution of histone variants
Histones are among the most conserved proteins in
eukaryotes, and make the chromatin nonstatic and
parent nucleosomes regulatory. Folding of chromatin
domains is defined at a lower level by the compactness
of the basic units, guided and determined by the his-
tone–DNA as well as particle–particle interactions.
High conservation of core histone structure and their
contacts with each other and with DNA leaves little
scope for any heterogeneity. Therefore, apart from try-
ing to reshuffle or remove nucleosomes from the

cess throughout the cell cycle and quiescence (old age)
[21,22]. The variants have diverged from the normal
histones early in the course of evolution, acquiring
differential expression patterns. The structural hetero-
geneity conferred by the variants to chromatin can
potentially regulate various nuclear functions such as
transcription, gene silencing, chromosome segregation,
replication, repair and recombination. Such multiface-
ted regulatory activities of the nucleosomes through
variations in the subunits of the histone octamer would
not have been possible with a strict conservation of
histones at all the times and everywhere. Variants have
provided an added advantage.
Variants of H2A
Histones are proposed to have evolved from a com-
mon and simple ancestral archeal protein [23,24] and
followed three evolutionary histories. H2A and H2B
have diverged faster than H3 and H4. Different H2A
variants have arisen in two single events, while variants
of H3 have probably evolved through multiple inde-
pendent events [25]. They have evolved slowly in such
a way that they could not only fulfill the basic function
of DNA compaction and maintain the higher order
chromatin structure but also have gained functional
specialization due to the acquired changes [23,26]. Var-
iants of H2A show divergent functions in different
contexts (Table 1). H2A has the largest macro hetero-
geneous family of variants and all of them are found
to have a crucial role in gene expression and nuclear
dynamics [4]. Five human H2A genes encode proteins

histone region at the carboxyl end, which forms two
third of the protein’s molecular mass. The H2A region
of this variant is 50% identical to H2A.Z, both having
homology with the corresponding region of conven-
tional H2A. The nonhistone region, now termed as the
‘macrodomain’, contains a short, highly basic region
and a putative leucine zipper domain (Fig. 1; amino
acids 132–159 and 181–208, respectively, in rat liver
protein). Macrodomains may be associated with differ-
ent functions as they are found in diverse proteins such
as those containing poly(ADP-ribose) polymerase
activity and other single strand RNA viral proteins.
They show structural similarity to the DNA binding
domain of leucine aminopeptidases, suggesting that
DNA binding activity is associated with macrodomains
[30]. The exact functional status of the macrodomain
in mH2A is not known.
Variants of H3
Initial studies on histone H3 variants in mice have
helped to classify them according to their relationship
with DNA replication. The major, bulk histones are
deposited over newly synthesized DNA during replica-
tion in a replication-dependent chromatin assembly
pathway, whereas the replacement histone variants
undergo a replication-independent chromatin assembly
[31]. A replication coupled (RC) ⁄ dependent assembly
pathway involves a variety of components such as
CAF-1, RCAF (histone chaperones) and proliferating
cell nuclear antigen (PCNA), and deposits histones on
replicating DNA during the S-phase [32–34]. The repli-

gene expression. Of the three somatic H3 variants
known, H3.1 and H3.2 were classified as ‘strictly repli-
cation dependent’ and H3.3 as replication-independent
[1]. The RI variant accumulates as the tissue matures.
H3.1 and H3.2 are closely related, only differing in a
Cys-to-Ser substitution at amino acid 96, and belong
to the S-phase subtypes [35]. While only one type of
histone H3, similar to H3.3 is expressed [36] in yeast,
there are three variants of H3 in Drosophila; major
H3, H3.3 and centromeric centromere identifier (Cid).
H3.3 is almost identical to H3 and differs at only four
positions; one in the N-terminal tail (A31) and three in
the histone fold domain (S87, V89, M90) [37].
Centromere-specific H3 variants of all Drosophila
species are documented to show adaptive evolution
continuing for 25 million years [38]. Unlike H3.3, Cid
is characteristically a structural component of the
centromeres. It is very much diverged from H3, having
homologies only in histone fold domains although con-
served blocks are also seen in the N-terminal tail [38].
The evolutionary comparison of CenH3s from various
Drosophila species suggests a unique packaging func-
tion for the N-terminal tail at the cytological marker
of centromeres, the primary constriction [38]. In com-
parison, human centromeric H3-like protein, CENP-A,
shows 62% identity with H3 in its carboxy terminal
portion but there is no sequence similarity in the
N-terminus, which varies from 20 to 200 amino acids
in CENP-A as compared to 45 amino acids in the
N-terminus of H3 [39]. The histone fold domain of

of the repair machinery by the transcription complex
at the DNA damage site [40,41]. However, DNA may
be damaged under various conditions and cells have
several mechanisms for its repair [42]. Under nontran-
scribing conditions, recognition of DNA damage and
recruitment of the repair machinery may need other
signalling mechanisms [43,44]. For example, during
radiation-induced DNA damage or other events lead-
ing to double stranded breaks (DSBs) in DNA, a his-
tone variant present at the DNA damage point may
act as a marker for the quick recruitment of a repair
complex, thereby helping to maintain the eukaryotic
genome [45].
H2A.X is randomly incorporated into nucleosomes
and represents 10–15% of the total cellular H2A.
Phosphorylation of H2A.X is suggested to mark the
damaged DNA for recruitment of the repair machin-
ery, although it is not clear how the damage is indica-
ted in regions with bulk H2A. Nevertheless,
immunocytochemical analyses have shown that not
every contiguous H2A.X molecule is phosphorylated
[46]. The carboxy terminus of H2A.X differs from that
of bulk H2A in being longer and having a four amino
acid sequence element SQEL at the extreme end of the
protein (Fig. 1). Within this C-terminal motif, an aci-
dic residue follows the two relatively invariant amino
acids (SQ) while the last carboxy-terminal residue is
hydrophobic [27]. The SQE motif is part of the com-
mon consensus motif found in targets of the phospha-
tidylinositol kinases. Indeed, three members of the

phosphorylation of the serine residue in response to
DNA fragmentation facilitates NHEJ by decondensing
chromatin at the damaged DNA sites and making it
accessible to repair factors [47]. Deficiency of H2A.X
in mice leads to meiotic defects, such as retaining
unprocessed double stranded breaks after asynapsis
and increased predisposition to various tumours in the
absence of p53 [54]. Thus the rapid observed colocali-
zation of the p53 binding protein1 (53BP1) with
c-H2A.X foci after introduction of DNA double
strand breaks may have great clinical implications.
Phosphorylated H2A.X ensures an error-free process
by using the sister chromatid as a template in exclu-
ding the error-prone repair (single-strand annealing) at
chromosomal DSBs [55]. Furthermore, H2A.X phos-
phorylation by primary DNA damage checkpoint kin-
ases makes a large chromatin domain permissive for a
de novo recruitment of cohesins required for cohesion
of sister chromatids. Cohesins tether the broken DNA
ends, making them a preferred substrate for repair and
preventing the highly reactive DNA ends from aber-
rant translocations and large interstitial deletions [56].
Several examples from various species, including
Xenopus, Drosophila, mammals and S. cerevisiae, have
shown that ionizing radiations and other agents that
cause double-strand breaks result in rapid and massive
phosphorylation of the histone variant H2A.X. Effi-
cient, homologous recombinational repair of a chro-
mosomal DSB is evidently found to require Ser139 of
mammalian H2A.X. Recent studies with yeast have

way, which uses a general transcription factor [40,41].
Phosphorylation at the SQ motif of the variant may be
easier and more economical than developing a new
method of marking the damage site with the bulk
H2A.
ATP-dependent chromatin remodelling and covalent
histone modifications are two processes associated with
the regulation of gene expression from a chromatin
region. A close relationship between chromatin remod-
elling and DNA repair reported recently [61] is an
excellent example of the economy practiced by cells in
general. It suggests that chromatin remodelling may
not be a process related only to gene expression.
Rather, the same proteins may be active in other
DNA-related processes, coupling the two processes.
An HMG-like subunit, Nhp10, of the yeast chromatin
remodelling complex INO80, is shown to interact with
c-H2A.X at DSBs to recruit the INO80 complex. Gen-
etic evidence for the interaction of Nhp10 with mem-
bers of the RAD52-dependent repair pathway suggests
that INO80 may in turn recruit the repair machinery
at the damage site through Nhp10 [62]. In Drosophila,
the H2A variant H2Av, is a functional homologue of
both H2A.X as well as H2A.Z in mammals [63]. The
Drosophila Tip60 chromatin remodelling complex
acetylates nucleosomal phospho-H2Av. At the same
time, the ATPase activity of dTip60 exchanges the
R H. Pusarla and P. Bhargava Histone variants in various functions
FEBS Journal 272 (2005) 5149–5168 ª 2005 FEBS 5153
phospho-H2Av with the unmodified H2Av, presenting

known to associate with and promote the heterochro-
matin formation [66]. For example, Drosophila H2Av
is found to participate in heterochromatin formation
by marking the region for subsequent acetylation at
H4K12 and methylation at H3K9 with HP1 recruit-
ment [67]. It shows a nonuniform pattern of wide dis-
tribution in the genome and is present in thousands of
euchromatic bands as well as the heterochromatic
chromocentre of polytene chromosomes [28].
In mouse spermatocytes, c-H2A.X plays a crucial
role in sex chromosome condensation and transcrip-
tional inactivation under the process of meiotic sex
chromosome inactivation (MSCI). It regulates chroma-
tin remodelling and associated silencing of male sex
chromosomes by initiating heterochromatinization in
the sex body. Absence of H2A.X in mice results in
infertility in the male but not in the female, and several
sex body proteins such as XMR and macroH2A1 ⁄ 2
fail to localize to the sex chromosome [68]. The
absence of condensed sex body and the failure of
meiotic pairing by X and Y chromosomes in H2A.X
deficiency suggests that H2A.X is more important for
heterochromatinization in the male than the female.
Mammalian H2A.Z is also found to be essential for
establishing higher order chromatin structure at consti-
tutive heterochromatic domains, probably by control-
ling the localization of HP1a. It is localized along with
HP1a on chromosome arms but not on centromeric
regions [69]. Arrays of positioned nucleosomes con-
taining H2A.Z over the defined sequence 208–12 DNA

some, with methylated H3-K4 at a potential activation
boundary during metaphase [73], and with heterochro-
matin protein M31 during meiotic prophase [76], thus
suggesting that the association of macroH2A may not
be specific to the Barr body. It brings about X-chro-
mosome inactivation probably by stabilizing the bind-
ing of Xist to the X chromosome through its
nonhistone region [77].
Nucleosomes containing mH2A have altered struc-
ture owing to the high a-helical content in their C-ter-
minal nonhistone regions [78]. The unusual structure
of mH2A with a large C-terminal tail may give a
unique conformation to the nucleosome, as reflected
by their low sedimentation coefficient despite a 25%
increase in the mass. The core particles having mH2A
Histone variants in various functions R H. Pusarla and P. Bhargava
5154 FEBS Journal 272 (2005) 5149–5168 ª 2005 FEBS
show slower gel mobility but the same stability as that
of native nucleosomes, suggesting an asymmetric and
extended conformation. Presence of the nonhistone
region may be responsible for the observed DNaseI
hypersensitivity near the dyad axis and around
entry ⁄ exit sites of DNA in the nucleosome [78]. Macro-
H2A exerts its repressive action through control over
transcription and chromatin remodelling. The presence
of mH2A in a positioned nucleosome disrupts access
for NF-jB, as well as remodelling and mobilization of
variant nucleosomes by SW1 ⁄ SNF without affecting
either its binding or ATPase activity [79]. A macro-
H2A C-terminal region present near to a promoter

shows a constant turnover. The deposition of H3.3 is
directly linked to active transcription at the hsp70 gene
locus, as it stops replacing H3 after the induced gene is
switched off [81]. Constitutive synthesis replenishes
H3.3, which is shown to be short-lived compared to
bulk H3. The changing of one amino acid from his-
tone H3 to its H3.3 counterpart relieved the block to
RI assembly and further deposition of H3 outside S
phase [82]. Thus, while the N-terminal was required
for RC deposition, specific residues in the histone fold
could switch it to the RI deposition pathway, which
seems to be restricted to H3.3 deposition and targeted
to transcriptionally active chromatin.
In mice, the transcript levels of both H3.1 and H3.2
decrease as cell division slows down during differenti-
ation, whereas H3.3 continues to be synthesized and
maintained throughout differentiation. Similarly, Droso-
phila H3 is deposited only during S-phase, whereas
H3.3 is deposited both during and outside of S-phase,
suggesting that H3.3 might accumulate in nondividing
cells [2]. Excess accumulation of H3.3 in nerve cells
leads to further severity of Rett syndrome, a common
mental disorder directly related to the loss of MeCP2,
a methylated CpG binding protein. MeCP2 deficiency
leads to the loss of silencing mechanisms involving
H3K9 methylation and histone deacetylase activity.
Acetylation of H3K9 is associated with active chroma-
tin while H3K9 methylation marks inactive chromatin
regions. Thus, the unintended activation due to H3.3
accumulation (associated with transcribed regions) and

enzyme RNA polymerase. It was found in an in vitro
R H. Pusarla and P. Bhargava Histone variants in various functions
FEBS Journal 272 (2005) 5149–5168 ª 2005 FEBS 5155
study that RNA polymerase II (pol II) can transcribe
through a nucleosome without completely displacing
histones from it [87]. The protein complex facilitates
chromatin transcription (FACT) facilitates read-
through of the nucleosomal template by RNA polym-
erase II during transcription elongation [88]. Associ-
ated histone chaperone activity of FACT can help
remove as well as redeposit an H2A-H2B dimer during
the transcription [89]. Chromatin reassembly in yeast
becomes dependent on the Hir⁄ Hpc (human HIRA
homologue) pathway on the loss of yeast FACT activ-
ity [90], suggesting that both chaperones may be work-
ing on transcribed templates. Removal of H2A ⁄ H2B
by FACT may facilitate access of H3 for exchange
with H3.3 by HIRA in the next step. Nevertheless, a
recent study reports the exchange of H2A.Z with bulk
H2A on the c-myc gene during transcription [91].
These findings suggest that nucleosomes can indeed be
shuffled during read-through by RNA pol II in vivo
without displacing the histone octamer completely.
In the budding yeast S. cerevisiae, H2A.Z is found
to be important for both positive and negative gene
regulation [92–95]. Loss of Htz1 in yeast cells leads
to slow growth and formamide sensitivity at 28 °C
and lethality at 37 °C [96]. The PHO5 promoter is
found to be more open in the htz1 D⁄snf2D mutant
[95], suggesting this H2A variant in yeast acts with

not maintain the transcriptionally active state. In a
functional dynamic study, nucleosomes were found to
show two types of large motions in space; a stretch-
ing-compression along the dyad axis and the flipping,
bending sideways motions with respect to the dyad
axis, a result of the dynamism of the N-termini of H3
and the H2A.Z-H2B dimer. The nucleosomes with
variant histones show comparatively weaker correla-
tions between internal motions, resulting in the per-
turbation of interactions between the contact regions
of the variant histones with overlying DNA [19]. In
agreement with this, H2A.Z-H2B dimers in the vari-
ant nucleosomes dissociate with comparative ease,
correlating with the observation that chromatin
regions containing H2A.Z probably do not require
SW1 ⁄ SNF remodelling complexes [95]. However, in a
global analysis, a 13 protein complex, SWR-C, neces-
sary for promoting gene expression near silent hetero-
chromatic regions of yeast, is found to be required
for the recruitment of Htz1 to chromatin also [101].
Incorporation of Htz1 is facilitated by one of the
components of SWR-C, Swr1, an ATPase of Snf2
family, which acts as a histone exchanger and effi-
ciently replaces H2A with H2A.Z in nucleosome
arrays [94]. Genetic and biochemical approaches also
demonstrated the requirement of Swr1p for the depos-
ition of H2A.Z into euchromatic regions at several
sites [102]. Both groups identified a bromodomain
(which recognizes an acetyl group) containing protein
Bdf1 that also interacts with transcription factor IID

ants in generating a variety of chromatin structures.
Generation of the condensed chromatin domains
(Fig. 2H), starting from fully extended and relaxed
‘beads on a string’ (Fig. 2C), requires compaction of
the 10 nm fibre (Fig. 2B) followed by folding, conden-
sation and superfolding through the 30 nm stage to
higher order chromatin structure. The details of the
nucleosome structure in Fig. 2A depict the positions
where two of the core histones H3 and H2A can
acquire changes. H2A variants can lead to inactive or
condensed heterochromatin (Fig. 2D,E,G) as explained
above. However, they can also be found in active,
euchromatic regions as described in the following stud-
ies. Thus, H2A.Z is one of the variants that has been
found to induce both repressive and antisilencing
effects.
H2A.Z is essential for establishing the proper chro-
matin structure required for early development in
many organisms, including mice, Drosophila and Tetra-
hymena [105–107]. Absence of H2A.Z in mammals
leads to genome instability and defects in chromosome
segregation [69]. During embryonic differentiation sta-
ges, it is excluded from the nucleolus as well as the
inactive X chromosome and made its first appearance
A
B
C
D
E
F

stable folding and that the canonical H2A-H2B dimer
shows the most stable folding [110]. The 2.6 A
˚
resolu-
tion crystal structure of the variant nucleosome core
particle showed surprisingly small changes in the over-
all structure of H2A.Z [111]. However, distinct and
subtle destabilization of the interaction between the
H2A.Z-H2B dimer and the (H3-H4)
2
tetramer is seen.
The L1 loop domain of H2A (Fig. 2B), which ensures
incorporation of only one type of molecule, is altered
in H2A.Z. As a result, pairing of H2B with both
H2A.Z and H2A within the same nucleosome core
particle leads to steric imbalance that may favour
binding to another H2A.Z. A unique feature of the
acidic patch on the surface of normal H2A is extended
by replacement of Asn and Lys with Asp and Ser in
H2A.Z [111]. This enhanced charge patch at the C-ter-
minus is required for higher order chromatin forma-
tion and may offer a stronger docking domain for the
H4 tail of a neighbouring nucleosome [71], thereby
promoting interparticle folding in arrays (Fig. 2F).
Functional evidence of the implicit repressive role of
H2A.Z comes from a recent study demonstrating
replacement of the H2A.Z-H2B dimer by the
H2A-H2B dimer by transcribing RNA pol II [91].
While the acidic nature of the charged patch of
H2A is increased in H2A.Z, it is decreased in a newly

mobilize H2A.Bbd containing nucleosomes [114].
However, the lower stability of H2A.Bbd-containing
nucleosomes may facilitate the exchange of the
H2A.Bbd compared to H2A [115], probably promoting
transcription through nucleosomes during the elonga-
tion phase.
Similar to H3.3, the third H3 variant in Drosophila,
Cid, is deposited in an RI manner throughout the cell
cycle. An open chromatin configuration at both cen-
tromeres (due to the lack of H3K9 methylation in Cid)
as well as active chromatin is proposed to be the com-
mon basis of RI histone deposition at these sites [37].
Conserved blocks in the N-terminus and histone fold
of Cid may mediate essential protein–protein interac-
tions for recruitment of other centromeric proteins,
neutralize phosphates in linker DNA and further help
in higher order chromatin structure. Centromeric
nucleosomes of mice also are characterized by the pres-
ence of the centromeric H3 variant CENP-A [116]. It
is required for the recruitment of components essential
for kinetochore formation and chromosome segrega-
tion; disturbance in these important activities due to
targeted deletion of CENP-A in mice results in embryo-
nic death [117]. CENP-A competes with H3 for H4
during nucleosome formation and can be reconstituted
with DNA into nucleosomes with properties similar to
those of bulk nucleosomes [118]. CENP-A and H4
subnucleosome tetramers are more compact and con-
formationally rigid compared to normal tetramers
[119]. This tetrameric compaction in the nucleosomes

maintain inner centromere histones in a deacetylated
state and do not recruit CENP-A (Cnp1 ⁄ spCENP-A)
to centromeres. Similarly, human Mis16-like proteins,
RbAp46 and RbAp48, are also required for proper
CENP-A localization in human cells [127]. Further
studies will be necessary to reveal how Mis16 ⁄ Mis18
changes the chromatin environment at centromeres in
order to allow CENP-A loading.
In all of the above-mentioned nuclear processes
chromatin acquires a variety of configurations. For
inactivation of the X chromosome or heterochromati-
nization and the silencing of defined regions, the chro-
matin structure needs superfolding of the fibres,
extreme condensation through strong interfibre as well
as interparticle interactions. In contrast, for gene
expression from active regions as well as site-specific
DNA damage repair, it needs decondensation, expan-
sion and uncoiling of the regions by weakening of the
same interactions between its fibres or particles. Indi-
vidual nucleosomes contribute to these DNA–protein
and protein–protein interactions through the N-ter-
minal tail regions of their histones. Generating two
opposite end-results through the same set of inter-
actions can be made possible by regulating the para-
meters that define these interactions. It is conceivable
from the previous sections that the functional diversifi-
cation of chromatin is directly related to the structural
variety brought about by the variants. Thus the gen-
eration of functionally heterogeneous conditions may
become possible through variation in the histone pri-

are found in this region. While the random coil seg-
ments of N-terminal tails of both H3 and H2B pass
between gyres of the DNA superhelix, four amino
acids of the H2A N-terminal tail, close to the site of
H2B interaction, bind to the minor groove on the out-
side of the superhelix (Fig. 3A). Thus, N-terminal tails
are involved in deciding the DNA–histone interactions,
and to keep an intact nucleosome they need to be
spared from the changes that could destroy these inter-
actions. Changes in C-termini instead may give nucleo-
somes various properties without interfering with the
basic scheme of their structure.
Among the histone heterodimers of the core particle,
one of the partners is usually found to be more varied.
Varying only one of the partners at a time can give an
alteration in structure with the least perturbation, and
in the H2A-H2B dimer H2A could be the better
choice, due to the following reasons. Interaction of the
H3-H4 tetramer with the H2A-H2B dimer is esta-
blished through contacts made by H2B with H4 [128],
which is one of the important interactions in core par-
ticle assembly. Therefore, H2B may not be preferred
R H. Pusarla and P. Bhargava Histone variants in various functions
FEBS Journal 272 (2005) 5149–5168 ª 2005 FEBS 5159
for variation. Compared to other core histones, H2A
has a strategic placement in the nucleosome and con-
tains the largest consensus C-terminal tail (Fig. 2).
This tail protrudes on the outside of nucleosome near
the entry and exit sites of the DNA, and amino acids
105–117 link aN of the opposite H3 to the H3-H4 his-

appears that minor sequence variations in the C-ter-
minal proximal histone fold region of H3 that guide it
to actively transcribed chromatin regions can be toler-
ated easily. They do not disturb incorporation of H3
into the nucleosome, as shown by the similar overall
crystal structure of Xenopus and yeast nucleosomes
[131] with the latter having an H3 more akin to H3.3
of other eukaryotes. Small perturbations in H3 folding
due to the presence of a probe at its unique and cen-
trally placed cysteine (Cys96 or Cys110) in the a2of
the histone fold can generate different conformers of
the nucleosome; those with open conformations could
be better transcribed [132]. No specific changes in
structure are attributed [131] to the two different
amino acids at positions 89 and 90 of yeast H3, which
are found in the N-terminal halves of the a2 helices.
The location of these amino acids in the crystal struc-
ture of the yeast nucleosome core particle suggests that
they may influence the interaction of the H3-H4 dimer
with the H2A-H2B dimer, by altering its orientation in
AB
Fig. 3. Structural features of a nucleosome as revealed by the crystal structure analysis showing intranucleosomal interactions of histones.
(A) Half of a nucleosome (with one superhelical turn of 73 bp DNA) showing all domains of the four core histones and seven helical turns of
the DNA. The C-terminal tail of H2A with the maximum number of variations known is highligted. C and N indicate the C- and N-terminal
ends, respectively, of the individual histones. (B) Structure of the yeast nucleosome with both turns of the DNA, showing histones of the
lower half only partially. All four strands of DNA are shown in different shades for clarity; lighter shades are given to histones of the lower
half.
Histone variants in various functions R H. Pusarla and P. Bhargava
5160 FEBS Journal 272 (2005) 5149–5168 ª 2005 FEBS
space. Thus, it is probable that a compatible H3.3-H4

yeast. The yeast genome is reported to be largely active
with no pseudogenes or repetitive DNA, and it shows
structurally distinct promoter and nonpromoter
regions where promoters have a two- to threefold
lower nucleosome density covering them [133]. No lin-
ker histone H1 was found in yeast for a long time.
However, it is reported to have a higher order chro-
matin structure similar to that in higher eukaryotes [134].
Identification of a gene coding for a putative histone
H1 of yeast [135] suggests that this H1-like protein
may be involved in forming a higher order chromatin
similar to that in other metazoans. The differences in
the primary sequence of yeast histones from that of
higher eukaryotes may generate different particle–par-
ticle interactions. Thus, though the crystal structure of
yeast and Xenopus nucleosome core particles are sim-
ilar, sequence differences of individual histones may be
the cause of the observed crystal packing differences
and destabilization of the yeast core particle [131]. This
may also be the reason that yeast chromatin has a
similarly folded 30 nm fibre [134] but still an ‘open’
higher order chromatin structure.
Compared to mammals, fewer H2A variants in yeast
are known. Major H2A (90% of total H2A) itself
functions like H2A.X of mammals [65]. The amino
acid sequence of human H2A.X shows a C-terminal
region highly homologous to H2A species of S. cere-
visiae and Schizosaccharomyces pombe [136], suggesting
that yeast and human H2A may not have evolved
through the same pathway. Similarly, the presence of

repressively methylated at K9, and the methylation at
K4 is known to be associated with active chromatin.
The absence of the recently identified and universally
present H3K4Me-specific demethylase in S. cerevisiae
[140] may be related to the maintenance of this all-act-
ive state of the yeast genome, as demethylation of
H3K4Me may be counterproductive.
Yeast nucleosome assembly protein1 (Nap1; a his-
tone chaperone) was found to exchange the major
H2A-H2B dimer as well as variant dimers from nucle-
osomes [141]. On the other hand, two of the yeast cell
cycle-regulated histone gene repressors, Hir1p and
Hir2p, along with chromatin assembly proteins CAF1
R H. Pusarla and P. Bhargava Histone variants in various functions
FEBS Journal 272 (2005) 5149–5168 ª 2005 FEBS 5161
and Asf1, are involved in chromatin formation and
position-dependent gene silencing [142,143]. Hir pro-
teins are also reported to be required for kinetochore
function in both S. cerevisiae and Schizosaccharomyces
pombe [142,144]. It is not yet clear whether CAF-1 and
Hir proteins are the specific chaperones for Cse4 or
whether they also assemble other centromere-specific
proteins. The human homologue of Hir1p and Hir2p,
HIRA, is a substrate for cyclin-cdk2 and blocks the
S-phase [145], while the Xenopus homologue is an RI
pathway-specific histone chaperone [86]. Involvement
of the members of same protein family from different
sources in various activities suggests a simultaneous
evolution of functional diversification of histones as
well as their chaperones.

DNA is also by similar mechanisms. All of these rea-
sons together might have resulted in extreme conser-
vation of histones. However, a need for variations for
regulatory purposes would have also set in with evo-
lution. To form an octamer of the same organization,
conservation of histone fold regions needed for the
handshake contacts is essential. Their N-termini are
required for interaction with neighbouring DNA
while the C-termini provide docking domains for
internucleosomal interactions. Covalent modifications
of charged residues in the N-termini and a perturba-
tion of the C-terminus results in reduced interactions
of histones with DNA as well as interparticle inter-
actions. However, variations in primary sequence or
chain length give greater scope for changing the
target interactions in both directions. An additional
N-terminal sequence in CENP-A or the extra C-ter-
minal region in mH2A both result in inactive and
compact chromatin regions (Fig. 2G). In contrast,
H2A.Bbd with a shorter C-terminal tail is localized
to active chromatin regions.
A
B
C
D
Fig. 4. Histone variants may be involved in the demarcation of functional boundaries. (A) A typical chromosome showing its different
regions. (B) In yeast, H2A.Z prevents the spread of silent chromatin into the neighbouring regions. (C) Phosphorylation of Ser31 of mamma-
lian H3.3 surrounding the centromeric region. (D) Centromeric nucleosome having the centromeric H3 variant.
Histone variants in various functions R H. Pusarla and P. Bhargava
5162 FEBS Journal 272 (2005) 5149–5168 ª 2005 FEBS

stably modified structures will be questions for future
research.
Acknowledgements
We thank Durgadas Kasbekar for critical editing of
the manuscript.
References
1 van Holde KE (1988) Chromatin, 1st edn. Springer-
Verlag, New York.
2 Ahmad K & Henikoff S (2002) Epigenetic conse-
quences of nucleosome dynamics. Cell 111, 281–284.
3 Henikoff S, Furuyama T & Ahmad K (2004) Histone
variants, nucleosome assembly and epigenetic inheri-
tance. Trends Genet 20, 320–326.
4 Ausio J & Abbott DW (2002) The many tales of a tail:
carboxyl terminal tail heterogeneity specializes histone
H2A variants for defined chromatin function.
Biochemistry 41, 5945–5949.
5 Kamakaka RT & Biggins S (2005) Histone variants:
deviants? Genes Dev 19, 295–310.
6 Sarma K & Reinberg D (2005) Histone variants meet
their match. Nat Mol Cell Biol Rev 6, 139–149.
7 Smith MM (2002) Centromeres and variant histones:
what, where, when and why? Curr Opin Cell Biol 14,
279–285.
8 Cosgrove MS, Boeke JD & Wolberger C (2004) Regu-
lated nucleosome mobility and the histone code. Nat
Struct Mol Biol 11, 1037–1043.
9 Smith CL & Peterson CL (2004) ATP-dependent chro-
matin remodeling. Curr Top Dev Biol 65, 115–148.
10 Studisky VM, Walter W, Kireeva M, Kashlev M &

Proteins 58, 683–696.
20 Workman JL & Abmayr SM (2004) Histone H3 var-
iants and modifications on transcribed genes. Proc Natl
Acad Sci USA 101, 1429–1430.
21 Alvelo-Ceron D, Niu L & Collart DG (2000) Growth
regulation of human variant histone genes and acetyl-
ation of the encoded proteins. Mol Biol Report 27,
61–71.
22 Wolffe AP & Dimitrov S (1993) Histone–modulated
gene activity: developmental implications. Crit Rev
Eukaryot Gene Expr 3, 167–191.
23 Malik HS & Henikoff S (2003) Phylogenomics of the
nucleosome. Nat Struct Biol 10, 882–891.
R H. Pusarla and P. Bhargava Histone variants in various functions
FEBS Journal 272 (2005) 5149–5168 ª 2005 FEBS 5163
24 Sandman K, Krzycki JA, Dobrinski B, Lurz R &
Reeve JN (2000) HMf, a DNA binding protein isolated
from the hyperthermophilic archeon Methanothermus
fervidus, is most closely related to histones. Proc Natl
Acad Sci USA 7, 5788–5757.
25 Thatcher H & Gorovsky MA (1994) Phylogenetic ana-
lysis of the core histones H2A, H2B, H3 and H4.
Nucleic Acids Res 22, 174–179.
26 Wolffe AP & Pruss D (1996) Deviant nucleosomes:
The functional specialization of chromatin. Trends
Genet 12, 58–62.
27 Redon C, Pilch D, Rogakou EP, Sedelnikova O, New-
rock K & Bonner WM (2002) Histone H2A variants
H2AX and H2AZ. Curr Opin Genet Dev 12, 162–169.
28 Leach TJ, Mazzeo M, Chotkowski HL, Madigan JP,

38 Malik HS, Vermaak D & Henikoff S (2002) Recurrent
evolution of DNA binding motifs in the Drosophila
centromeric histone. Proc Natl Acad Sci USA 99,
449–1454.
39 Sullivan KF, Hechenberger M & Masri K (1994)
Human CENP-A contains a histone H3 related histone
fold domain that is required for targeting to the cen-
tromere. J Cell Biol 127, 581–592.
40 Bhargava P (1994) Repairing information lacunae:
coupling transcription with repair. Curr Sci 67, 10–13.
41 Leadon SA (2000) Transcription-coupled repair: a
multifunctional signaling pathway. Cold Spring Harb
Symp Quant Biol 65, 561–566.
42 Peterson CL & Cote J (2004) Cellular machineries for
chromosomal DNA repair. Genes Dev 18, 602–616.
43 Ehrenhorf-Murray AE (2004) Chromatin dynamics at
DNA replication, transcription and repair. Eur J
Biochem 271, 2335–2349.
44 Gontijo AM, Green CM & Almouzni G (2003) Repair-
ing DNA damage in chromatin. Biochimie 85, 1133–
1147.
45 Paull TT, Rogakou EP, Yamazaki V, Kirchgessner
CU, Gellert M & Bonner WM (2000) A critical role
for histone H2AX in recruitment of repair factors
to nuclear foci after DNA damage. Curr Biol 10,
886–895.
46 Rogakou EP, Pilch DR, Orr AH, Ivanova VS &
Bonner WM (1998) DNA double stranded breaks
induce H2AX phosphorylation on serine-139. J Biol
Chem 273, 5858–5868.

J, Eckersdorff M et al. (2003) Histone H2AX: a dosage
dependent suppressor of oncogenic translocations and
tumors. Cell 114, 359–370.
55 Xie A, Puget N, Shim I, Odate S, Jarzyna I, Bassing
CH, Alt FW & Scully R (2004) Control of sister
Histone variants in various functions R H. Pusarla and P. Bhargava
5164 FEBS Journal 272 (2005) 5149–5168 ª 2005 FEBS
chromatid recombination by histone H2AX. Mol Cell
16, 1017–1025.
56 Unal E, Arbel-Eden A, Sattler U, Shroff R, Lichten
M, Haber JE & Koshland D (2004) DNA damage
response pathway uses histone modification to assem-
ble a double-strand break-specific cohesion domain.
Mol Cell 16, 991–1002.
57 Celeste A, Fernandez-Capitello O, Kruhlak MJ, Pilch
DR, Staudt DW, Lee A, Bonner RF, Bonner WM &
Nussenzwig A (2003) Histone H2A.X phosphorylation
is dispensable for the initial recognition of DNA
breaks. Nat Cell Biol 5, 675–679.
58 Lisby M, Mortesen UH & Rothstein R (2003) Coloca-
lisation of multiple DNA double-strand breaks at a
single Rad52 repair centre. Nat Cell Biol 5, 572–577.
59 Aten JA, Stap J, Krawczyk PM, van Oven CH, Hoebe
RA, Essers J & Kanaar R (2004) Dynamics of DNA
double-strand breaks revealed by clustering of
damaged chromosome domains. Science 303, 92–95.
60 Bassing CH, Chua KF, Sekiguchi J, Suh H, Whitlow
SR, Fleming JC et al. (2002) Increased ionizing radi-
ation sensitivity and genomic instability in the absence
of histone H2A.X. Proc Natl Acad Sci USA 99, 8173–

68 Fernandez-Capetillo O, Mahadevaiah SK, Celeste A,
Romanienko PJ, Otero RDC, Bonner WM, Manova
K, Burgoyne P & Nussenzweig A (2003) H2AX is
required for chromatin remodeling and inactivation
of sex chromosomes in male mouse meiosis. Dev Cell
4, 497–508.
69 Rangasamy D, Greaves I & Tremethick DJ (2004)
RNA interference demonstrates a novel role for H2A.Z
in chromosome segregation. Nat Struct Mol Biol 11,
650–655.
70 Fan JY, Gordon F, Luger K, Hansen JC & Tre-
methick DJ (2002) The essential histone variant H2A.Z
regulates equilibrium between different chromatin
conformational states. Nat Struct Biol 9, 172–176.
71 Fan JY, Rangasamy D, Luger K & Tremethick DJ
(2004) H2A.Z alters the nucleosome surface to pro-
mote HP1a-mediated chromatin fibre folding. Mol Cell
16, 655–661.
72 Pehrson JR, Costanzi C & Dharia C (1997) Develop-
mental and tissue expression patterns of histone
macroH2A1 subtypes. J Cell Biochem 65, 107–113.
73 Chadwick BP & Willard HF (2002) Cell cycle-
dependent localization of macroH2A in chromatin of
the inactive X chromosome. J Cell Biol 157,
1113–1123.
74 Chadwick BP & Willard HF (2003) Chromatin of the
Barr body: histone and non-histone proteins associated
with or excluded from the inactive X chromosome.
Human Mol Genet 12, 2167–2178.
75 Perche P-Y, Vourch C, Konecny L, Souchier C,

H3.3 marks active chromatin by replication indepen-
dent nucleosome assembly. Mol Cell 9, 1191–1200.
83 Shahbazian MD, Young JI, Yuva-Paylor LA, Spencer
CM, Antalffy BA, Noebels JL, Armstrong DL, Paylor
R & Zoghbi HY (2002) Mice with truncated MeCP2
recapitulate many Rett Syndrome features and
display hyperacetylation of Histone H3. Neuron 35,
243–254.
84 Bulger M, Schubeler D, Bender MA, Hamilton J,
Farrell CM, Hardison RC & Groudine M (2003) A
complex chromatin landscape revealed by patterns of
nuclease sensitivity and histone modification within the
mouse b-globin locus. Mol Cell Biol 23, 5234–5244.
85 Tagami H, Ray-Gallet D, Almouzni G & Nakatani Y
(2004) Histone H3.1 and H3.3 complexes mediate
nucleosome assembly pathways dependent or indepen-
dent of DNA synthesis. Cell 116, 51–61.
86 Ray-Gallet DR, Quivy J-P, Scamps C, Martini EM-D,
Lipinski M & Almouzni G (2002) HIRA is critical for
a nucleosome assembly pathway independent of DNA
synthesis. Mol Cell 9, 1091–1100.
87 Bhargava P (1993) Dynamics of interaction of RNA
polymerase II with nucleosomes. II. During read-
through and elongation. Prot Sci 2, 2246–2258.
88 Orphanides G & Reinberg D (2000) RNA polymerase
II elongation though chromatin. Nature 407, 471–475.
89 Belotserkovskaya R, Oh S, Bondarenko VA, Orphan-
ides G, Studitsky VM & Reinberg D (2003) FACT
facilitates transcription-dependent nucleosome altera-
tion. Science 301, 1090–1093.

97 Meneghini MD, Wu M & Madhani HD (2003) Con-
served histone variant H2A.Z protects euchromatin
from the ectopic spread of silent heterochromatin. Cell
112, 725–736.
98 Larochelle M & Gaudreau L (2003) H2A.Z has a func-
tion reminiscent of an activator required for preferen-
tial binding to intergenic DNA. EMBO J 22, 4512–
4522.
99 Clarkson MJ, Wells JRE, Gibson F, Saint R & Tre-
methick DJ (1999) Regions of variant histone
His2AvD required for Drosophila development. Nature
399, 694–697.
100 Stargell LA, Bowen J, Dadd CA, Dedon PC, Davis M,
Cook RG, Allis CD & Gorovsky MA (1993) Temporal
and spatial association of histone H2A variant hv1
with transcriptionally competent chromatin during
nuclear development in Tetrahymena thermophila.
Genes Dev 7, 2641–2651.
101 Kobor MS, Venkatasubrahmanyam S, Meneghini MD,
Gin JW, Jennings JL, Link AJ, Madhani HD & Rine J
(2004) A protein complex containing the conserved
Swi2 ⁄ Snf2 related ATPase Swr1p deposits histone
variant H2A.Z into euchromatin. PLoS Biol 2,
587–598.
102 Krogan NJ, Keogh M-C, Datta N, Sawa C, Ryan
OW, Ding H et al. (2003) A Snf2 family ATPase com-
plex required for recruitment of the histone H2A vari-
ant Htz1. Mol Cell 12, 1565–1576.
103 Krogan NJ, Baetz K, Keogh M-C, Datta N, Sawa C,
Kwok TCY et al . (2003) Regulation of chromosome

14, 1–9.
111 Suto RK, Clarkson MJ, Tremethick DJ & Luger K
(2000) Crystal structure of a nucleosome core particle
containing the variant histone H2A.Z. Nat Struct Biol
7, 1121–1124.
112 Chadwick BP & Willard HF (2001) A novel chromatin
protein, distantly related to histone H2A, is largely
excluded from the inactive X chromosome. J Cell Biol
152, 375–382.
113 Bao Y, Konesky K, Park Y-J, Rosu S, Dyer PN,
Rangasamy D, Tremethick DJ, Laybourn PJ & Luger
K (2004) Nucleosomes containing the histone variant
H2A.Bbd organize only 118 base pairs of DNA.
EMBO J 23, 3314–3324.
114 Angelov D, Verdel A, An W, Bondarenko V, Hans F,
Doyen C-M, Studitsky VM, Hamiche A, Roeder RG,
Bouvet P & Dimitrov S (2004) SWI ⁄ SNF remodeling
and p300-dependent transcription of histone variant
H2A.Bbd nucleosomal arrays. EMBO J 23, 3815–3824.
115 Gautier T, Abbott DW, Molla A, Verdel A, Ausio J &
Dimitrov S (2004) Histone variant H2ABbd confers
lower stability to the nucleosome. EMBO Report 5,
715–720.
116 Palmer DK, O’Day K, Wener MH, Andrews BS &
Margolis RL (1987) A 17kDa centromere protein
(CENP-A) copurifies with nucleosome core particles
and with histones. J Cell Biol 104, 805–815.
117 Howman EV, Fowler KJ, Newsom AJ, Redward S,
MacDonald AC, Kaltisis P & Choo KHA (2000) Early
disruption of centromeric chromatin organization in

centromeres in budding yeast. Eukaryot Cell 3, 1533–
1543.
125 Chen Y, Baker RE, Keith KC, Harris K, Stoler S &
Fitgerald-Hayes M (2000) The N terminus of the cen-
tromere H3-like protein Cse4p performs an essential
function distinct from that of the histone fold domain.
Mol Cell Biol 18, 7037–7048.
126 Hsu JM, Hunag J, Meluh PB & Laurent BC (2003)
The yeast RSC chromatin-remodeling complex is
required for kinetochore function in chromosome
segregation. Mol Cell Biol 23, 3202–3215.
127 Hayashi T, Fujita Y, Iwasaki O, Adachi Y, Takahashi
K & Yanagida M (2004) Mis16 and Mis18 are required
for CENP-A loading and histone deacetylation at
centromeres. Cell 118, 715–729.
128 Luger K, Mader AW, Richmond RK, Sargent DF &
Richmond TJ (1997) Crystal structure of the nucleosome
core particle at 2.8 A
˚
resolution. Nature 389, 251–260.
129 Rhodes D (1997) The nucleosome core all wrapped up.
Nature 389, 231–233.
130 Baer BW & Rhodes D (1983) Eukaryotic RNA poly-
merase II binds to nucleosome cores from transcribed
genes. Nature 301, 482–448.
131 White CL, Suto RK & Luger K (2001) Structure of the
yeast nucleosome core particle reveals fundamental
changes in internucleosome interactions. EMBO J 20,
5207–5218.
132 Bhargava P (1993) Dynamics of interaction of RNA

LSD1. Cell 119, 941–953.
141 Park YJ, Chodaparmabil JV, Bao Y, McByrant SJ &
Luger K (2005) Nucleosome assembly protein 1
exchanges histone H2A-H2B dimers and assists nucleo-
some sliding. J Biol Chem 280, 1817–1825.
142 Sharp JA, Franco AA, Osley MA & Kaufman PD
(2002) Chromatin assembly factor 1 and Hir proteins
contribute to building functional kinetochores in S. cer-
evisiae. Genes Dev 16, 85–100.
143 Sharp JA, Fouts ET, Krawitz DC & Kaufman PD
(2001) Yeast histone deposition protein Asf1p requires
Hir proteins and PCNA for heterochromatic silencing.
Curr Biol 11, 463–473.
144 Blackwell C, Martin KA, Greenall A, Pidoux A, All-
shire RC & Whitehall SK (2004) The Schizosaccharo-
myces pombe HIRA-like protein Hip1 is required for
the periodic expression of histone genes and contri-
butes to the function of complex centromeres. Mol Cell
Biol 24, 4309–4320.
145 Hall C, Nelson DMYeX, Baker K, Decaprio JA, See-
holzer Z, Lipinski M & Adams PD (2001) HIRA, the
human homologue of yeast Hir1p and Hir2p, is a novel
cyclin-cdk2 substrate whose expression blocks S-phase
progression. Mol Cell Biol 21, 1854–1865.
146 Hake SB, Garcia BA, Kauer M, Baker SP, Shabano-
witz A, Hunt DF & Allis CD (2005) Serine 31 phos-
phorylation of histone variant H3.3 is specific to
regions bordering centromeres in metaphase chromo-
somes. Proc Natl Acad Sci USA 102, 6344–6349.
Histone variants in various functions R H. Pusarla and P. Bhargava