Content uploaded by Danny Reinberg
Author content
All content in this area was uploaded by Danny Reinberg on Dec 29, 2013
Content may be subject to copyright.
review article
NATURE
|
VOL 407
|
28 SEPTEMBER 2000
|
www.nature.com 471
T
he need to compact 2 m of DNA into the nucleus of a
eukaryotic cell has severe consequences for processes
that require access to DNA. This compaction is achieved
by the binding of proteins that mediate successive orders
of DNA folding: two copies each of histones H2A, H2B,
H3 and H4 form a protein core that wraps 146 base pairs of DNA
tightly on its surface to form a nucleosome
1
. With the aid of
additional proteins, nucleosomes are compacted to form
chromatin—a complex, highly ordered nucleoprotein assembly
that is inaccessible to DNA-binding proteins
2
. Decompaction of
chromatin to facilitate access to DNA has been most widely studied
for RNA polymerase II (RNAP II)-mediated transcription of
protein-coding genes, a process that requires rapid access to genes
for the response to environmental signals and programmed cellular
events, but the underlying principles are equally applicable to any
process requiring interaction with DNA.
Transcription can be divided into two basic stages: initiation,
which involves the binding of transcription factors and RNAP II to
specific sites in DNA (promoter regions) and the onset of RNA syn-
thesis; and elongation, during which the polymerase tracks along the
DNA and makes an RNA copy
3
. Until recently, interest in chromatin
accessibility has focused on transcription initiation; however, initia-
tion is thought only to result in the decompaction of chromatin sur-
rounding the promoter, leaving transcribed regions in a compacted
and repressive conformation. Recent reports suggest that eukaryotic
cells are equipped with specialized proteins that help RNAP II pass
through chromatin during transcription elongation. Here we discuss
how these new findings implicate the elongation process in propagat-
ing the disruption of chromatin structure from the promoter region
of a gene to the entire transcribed domain.
Proteins that decompact chromatin structure
Chromatin decompaction has been the subject of three decades of
intense study. The lack of knowledge regarding the complex architec-
ture of chromatin has hampered research in this area; however, it is
clear that regions of the genome that are actively transcribed have a
more open and accessible chromatin structure than non-transcribed
regions
4
. Transcriptionally active, accessible regions are associated
with a loss of proteins involved in the maintenance of higher-order
chromatin structure. For example, histone H1, which binds to nucle-
osomes and promotes chromatin folding, is depleted in transcribed
chromatin
2
. Closer examination of the chromatin of transcribed
regions reveals alterations in histone proteins themselves. Acetyla-
tion of lysine residues in the amino-terminal ‘tails’ of histones has
long been correlated with transcriptional competence
5
. A good
example of this phenomenon is found at the transcriptionally active
b-globin locus of chicken erythrocytes, which contains 33 kilobases
(kb) of accessible chromatin and is enriched in acetylated histones
6
.
Structurally, histone tail acetylation disrupts histone—DNA and
inter-nucleosomal interactions by neutralizing positively charged
lysine residues. The enzymes that catalyse these covalent modifica-
tions are the histone acetyltransferases (HATs) and histone deacety-
lases (HDACs), which add or remove an acetyl group, respectively
7
.
The recent finding that many proteins that regulate transcription are,
or can recruit, HATs and HDACs has reinforced the link between his-
tone acetylation/deacetylation and gene activity.
Another class of factors that manipulate chromatin structure and
have roles in transcriptional regulation are the chromatin-remodel-
ling complexes, which use energy from ATP hydrolysis to disrupt
chromatin and make DNA more accessible to DNA-binding
proteins
8
. Chromatin remodelling enzymes have been purified from
a variety of organisms, and most cells contain more than one type of
complex
1,8
. These complexes contain structurally related catalytic
subunits, but differ in the way in which they manipulate chromatin.
Most of these enzymes can alter the conformation of a nucleosome to
increase access to DNA, and one complex, RSC
9
, can transfer an
entire octamer of histone proteins from one region of DNA to anoth-
er. Chromatin remodelling and HAT activities probably cooperate to
overcome the transcriptional repression imposed by chromatin pack-
aging.
RNA polymerase meets the nucleosome
RNAP II transcription is a complex process that is subject to chro-
matin repression at several stages (see Box 1). DNA binding by acti-
vator proteins is prevented by chromatin packaging
10
, but disruption
of histone—DNA contacts by acetylation of histone tails can over-
come this repression
8
. Similarly, the transient disruption of his-
tone—DNA contacts by ATP-dependent chromatin-remodelling
enzymes also facilitates DNA binding. Therefore, disruption of his-
tone—DNA contacts by histone acetylation or chromatin remodel-
ling allows DNA-binding proteins to compete with histones for
DNA. Does this simple paradigm also apply to transcription elonga-
tion? Will weakening of histone—DNA contacts in a nucleosome
allow its passage by an RNA polymerase? Initial experiments examin-
ing how an RNA polymerase copes with a nucleosome used small,
single-subunit prokaryotic RNA polymerases from bacteriophages
T7 and SP6 (refs 11, 12). Surprisingly, these small polymerases could
transcribe completely through a nucleosome, albeit at a reduced rate
compared with free DNA. Using a defined system containing a single
nucleosome on a short linear DNA fragment, Felsenfeld and co-
workers
13,14
found that during elongation the entire octamer of his-
tones is transferred backwards on the DNA fragment through a tran-
siently formed DNA loop (Fig. 1). The observation that the histone
octamer does not leave the DNA during the elongation process is
consistent with the finding that histones remain associated with the
DNA of genes being transcribed in the cell
15
. Although these studies
have provided useful mechanistic insights, the templates used con-
tained only a single nucleosome that is not subject to repressive inter-
RNA polymerase II elongation through chromatin
George Orphanides* & Danny Reinberg†
* Zeneca Central Toxicology Laboratory, Alderley Park, Cheshire, SK10 4TJ, UK
† Howard Hughes Medical Institute, Division of Nucleic Acids Enzymology, Department of Biochemistry, University of Medicine and Dentistry of New Jersey,
Robert Wood Johnson Medical School, Piscataway, New Jersey 08854, USA
The machinery that transcribes protein-coding genes in eukaryotic cells must contend with repressive chromatin structures
in order to find its target DNA sequences. Diverse arrays of proteins modify the structure of chromatin at gene promoters to
help transcriptional regulatory proteins access their DNA recognition sites. The way in which disruption of chromatin
structure at a promoter is transmitted through a whole gene has not been defined. Recent breakthroughs suggest that the
passage of an RNA polymerase through a gene is coupled to mechanisms that propagate the breakdown of chromatin.
© 2000 Macmillan Magazines Ltd
review article
472 NATURE
|
VOL 407
|
28 SEPTEMBER 2000
|
www.nature.com
nucleosome interactions that occur in natural chromatin. More
physiological templates, containing long arrays of 15 or more nucleo-
somes, slow the progress of prokaryotic T7 RNA polymerase sub-
stantially, suggesting that inter-nucleosome contacts significantly
repress elongation (for example, see ref. 16).
These experiments demonstrate that a single-subunit prokaryotic
RNA polymerase can bypass a nucleosomal roadblock under certain
conditions. Is this likely to be true for the large and cumbersome
eukaryotic RNA polymerases with 12 or more protein subunits? The
eukaryotic RNA polymerase III passes through a single nucleosome
with pronounced pausing, suggesting that it is hindered by its large
bulk
17
. Moreover, RNAP II is brought to an almost complete halt by
arrays of nucleosomes, with strong polymerase-pausing sites in the
nucleosome reflecting sites that induce natural pausing on free
DNA
18,19
. Furthermore, elongation factors that accelerate elongation
on free DNA cannot overcome this chromatin block
18,19
. The inability
of eukaryotic RNA polymerases to elongate through nucleosomes in
vitro may seem surprising, as they have evolved with their DNA tem-
plate packaged into chromatin and, therefore, would be expected to
have developed mechanisms to deal with nucleosomes. However,
recent evidence indicates that in a cell these RNA polymerases recruit
the assistance of cellular factors that can disrupt chromatin structure.
Helping RNAP II to elongate through chromatin
RNAP II in a cell travels at a rate of 25 nucleotides per second, an elon-
gation rate that can only be achieved in vitro on free DNA templates
(for example, see ref. 20). How does the polymerase achieve these
rates in the repressive context of chromatin? This question has
prompted researchers to seek conditions that will facilitate transcrip-
tion elongation through chromatin in vitro. Chromatin remodelling
by the SWI/SNF complex can promote RNAP II elongation through a
nucleosome
21
, presumably by disrupting histone—DNA interac-
tions. Similarly, the chromatin-associated HMG14 protein, which is
found in the chromatin of active genes in vivo
22
, can modestly
enhance RNAP II elongation through chromatin in vitro
23
.
Another complex that can facilitate RNAP II elongation through
nucleosomes is FACT. When added to a completely defined tran-
scription system using chromatin as the template, the heterodimeric
FACT complex enables RNAP II to elongate through nucleosomes
19
.
One of the subunits of FACT is a human homologue of the Saccha-
romyces cerevisiae Spt16 protein, which has been implicated in
modulating chromatin structure for transcription elongation by
genetic experiments
24-26
. The smaller subunit of FACT is the
HMG-1-like protein SSRP1. Biochemical analysis suggests that FACT
stably and specifically interacts with histones H2A and H2B, and
covalent crosslinking of histones in a nucleosome, to prevent removal
of these histones, abrogates FACT activity
26
. These properties
suggest that FACT disrupts nucleosomes during RNAP II elongation
by transiently binding and removing histones H2A and H2B. Con-
sistent with this model, yeast strains with mutations in histone H4,
which alter the interaction of histones H2A and H2B with other
histones in the nucleosome, exhibit identical phenotypes to
strains carrying mutations in the Spt16 subunit of FACT
27
. Further-
more, chromatin that contains transcribed sequences is deficient
in histones H2A and H2B and is preferentially bound by RNAP II
(ref. 28).
Another group of proteins implicated in relieving the chromatin
block to transcription are the Spt4, Spt5 and Spt6 proteins. Yeast
strains with mutations in the genes encoding these proteins exhibit
phenotypes consistent with defects in transcription elongation
29
and
share many phenotypes with strains containing mutations in the
Spt16 subunit of FACT and in histone proteins
30
. A human complex
of Spt4 and Spt5 proteins binds to RNAP II and modulates its elon-
gation activity on naked DNA templates in vitro
31
. In addition, this
complex, called DSIF, can promote RNAP II elongation through
chromatin templates assembled in vitro (T. Wada, G.O., H. Handa
and D.R., unpublished data). The third member of this group of pro-
teins, Spt6, can bind to histones and alters chromatin structure in
vitro
32
. The roles played by FACT, the SWI/SNF complex, HMG14
and Spt4, Spt5 and Spt6 in facilitating elongation through chromatin
1
2
3
4
5
pol
pol
pol
pol
pol
Figure 1 Model for transcription through a nucleosome by a single-subunit
prokaryotic RNA polymerase
13,14
. The RNA polymerase approaches the nucleosome
and begins to synthesize a transcript (1; the DNA present in the nucleosome is
shaded black). Transcription of the first 25 bp of DNA in the nucleosome is rapid and
results in the displacement of DNA from the central histone octamer (2). The DNA
segment behind the polymerase binds to the freshly exposed surface of the histone
octamer, forming a DNA loop (3). Further polymerase elongation is hindered by this
tight loop, which prevents the polymerase from rotating around the DNA as it reads
the nucleotide sequence. The DNA behind the polymerase then transiently
dissociates from the histone octamer, breaking the loop and allowing further
elongation. This results in progression of the loop further into the nucleosome (4).
This cycle of loop formation and breaking is repeated until the polymerase passes
completely through the nucleosome (5; for details, see ref. 14). This process results
in the transfer of the entire octamer of histone proteins backward on the same DNA
segment. Figure adapted with permission from ref. 48.
In its most general form, transcription activation by RNA polymerase
II (RNAP II) begins with the binding of activator proteins to DNA sites
adjacent to the start site of transcription (the promoter region). Once
bound, the activators can recruit RNAP II in a step involving a set of
accessory proteins---the general transcription factors (GTFs)---that
accurately position the polymerase over the transcription start
site
3,49
. RNAP II then begins transcription, clears the promoter and
enters the elongation phase. Finally, directed by DNA sequences at
the end of a gene, RNAP II terminates transcription and releases
newly synthesized RNA. Transcription is regulated primarily at
initiation, but a set of factors can stimulate the elongation process by
suppressing transient polymerase pausing or by increasing the rate
of RNA polymerization
50
.
Soon after initiating transcription, RNAP II is hyperphosphorylated on
its C-terminal domain (CTD), an unusual structure composed of
several repeats of a heptapeptide motif. Mounting evidence
suggests that CTD phosphorylation regulates the interaction of
RNAP II with proteins involved in transcriptional regulation and
mRNA maturation
51
. RNAP II containing an unphosphorylated CTD
is bound by the mediator complex, a multi-protein assembly
involved in the recruitment of RNAP II to active promoters
52
. Once
elongation is underway, the polymerase breaks its interactions with
transcription initiation factors and its CTD becomes a docking site
for enzymes that catalyse RNA polyadenylation, capping and
splicing.
Box 1
Transcription of protein-coding genes is a complex
process
© 2000 Macmillan Magazines Ltd
review article
NATURE
|
VOL 407
|
28 SEPTEMBER 2000
|
www.nature.com 473
in a cell are unknown, but are likely to be determined through a com-
bination of genetic and biochemical approaches.
Hitching a ride on RNAP II
If the factors described in the previous section facilitate elongation
through chromatin in a cell, how are they likely to be targeted to tran-
scribed regions of the genome? It is established that activities that
modify chromatin structure can be recruited to promoter regions to
facilitate transcription initiation through direct interactions with
DNA-bound activator proteins
8
, but how are they directed to down-
stream regions to facilitate elongation? One way in which these activ-
ities could gain access to the entire transcribed region is by hitching a
ride on the polymerase as it travels on its journey through a gene
33
. In
this way, the chromatin modifier would gain access to, and could pro-
mote the disruption of, the whole transcribed region (Fig. 2).
For this targeting mechanism to be effective, the modifying activi-
ties must be able to recognize and bind to polymerases that are elon-
gating, and not to polymerases that are free in the nucleus or are at
gene promoters. It is likely that the ‘tag’ that distinguishes an elongat-
ing polymerase is phosphorylation of its carboxy-terminal domain
(CTD) tail (see Box 1). The first evidence that a chromatin-remodel-
ling activity can recognize and bind to an elongating polymerase was
the finding that the PCAF HAT, which acetylates histones H3 and H4
in a nucleosome
34
, binds specifically to the phosphorylated, elongat-
ing form of RNAP II (ref. 35). More recently, the Svejstrup
laboratory
36,37
has characterized the composition of the elongating S.
cerevisiae RNAP II complex. This resulted in the isolation of a het-
erotrimeric complex, called ‘elongator’, that associates only with the
phosphorylated, elongating form of RNAP II and can also be found
in a free form
36,37
. The 60-kDa subunit of the elongator is the Elp3
protein, which contains HAT activity towards all four histones. These
characteristics of elongator suggest that it is recruited to elongating
RNAP II to acetylate the histones of transcribed chromatin. Genetic
experiments reveal that the elongator complex is dispensable for yeast
survival, but has an important function in the induction of certain
genes.
Transcription elongation and histone acetylation
Although the neutralization of negatively charged histone tails by
acetylation of lysines can disrupt chromatin structure, there is evi-
dence that this is not how the transcriptional effects of histone acety-
lation are propagated in a cell. For example, the observation that
acetylation of histone H3 is associated with both transcriptional acti-
vation and the generation of repressed chromatin structures suggests
that charge neutralization by acetylation is not the direct cause of
transcriptional activation
38
. A more appealing model for the func-
tion of histone acetylation proposes that the pattern of histone tail
modifications serves as a recognition code for the binding of proteins
that modulate chromatin structure (the ‘histone code’ hypothesis
38
).
Therefore, acetylation of different histone tail lysines by one or more
HATs travelling with elongating RNAP II may provide a combination
of recognition sites for chromatin-modifying complexes that assist
the elongation process, such as FACT, HMG14 and additional HAT
complexes.
The overall picture that emerges is one in which the modification
of chromatin by HATs that track with RNAP II leads to derepression
of a whole transcription unit. This model also implies a mechanism
for the way in which active transcription can be turned off. Competi-
tion with HDACs that are free in the nucleus means that maintaining
histones in an acetylated state requires constant transcription, a pro-
posal supported by experiments showing that the establishment of an
unfolded chromatin domain in vivo requires transcription elongation
and histone acetylation
39
. The state of histone tail acetylation is a
dynamic equilibrium determined by the activities of HATs bound to
elongating RNAP II and HDACs. Once the RNAP II traffic along a
gene is decreased, which is governed by signals at the promoter, the
equilibrium shifts in favour of the HDACs. Loss of histone tail acety-
lation may then result in the rapid conversion of chromatin structure
to a repressed conformation.
Is a ‘pioneer’ polymerase required?
Activities such as chromatin-remodelling complexes, HATs and
FACT can disrupt chromatin structure at the level of the nucleosome
to facilitate transcription elongation in vitro. But the nucleosome is
only the first level in chromatin compaction. The precise degree of
DNA compaction faced by elongating RNAP II in a cell is not known
and, therefore, it is unclear whether these activities are sufficient for
elongation through chromatin in vivo. Two models, which differ in
the extent of chromatin decompaction that occurs following the
binding of activators to promoters, are possible (Fig. 3). In the first
model, recruitment of chromatin-modifying activities by activators
results in complete decompaction of chromatin surrounding the
activator-binding sites and only partial decompaction elsewhere in
the gene. In this case, elongating RNAP II faces a compacted chro-
matin template. In the second model, activators promote decom-
1
3
5
pol II
P
P
P
P
P
P
P
P
P
P
P
P
P
P
P
P
P
P
P
P
pol II
P
P
P
P
P
P
P
P
P
P
4
2
pol II
= Nucleosome
Compacted
chromatin
Activator binding
and local chromatin
disruption
pol II
P
P
P
P
P
P
P
P
P
P
Transcription initiation and
phosphorylation of the
CTD of RNAP II
Transcription elongation and
propagation of chromatin disruption
RNAP II binding to
promoter region
Binding of chromatin
modifiers to RNAP II
containing a phosphorylated CTD
= Remodelled nucleosome
= Activator protein
= Chromatin modifiers
= RNA polymerase II
= RNA transcript
= CTD tail of
polymerase
= Phosphorylated CTD
'tail' of polymerase
Figure 2 Propagation of chromatin disruption by chromatin-modifying activities that
track with elongating RNAP II. One way in which chromatin-modifying activities can
gain access to the whole transcribed region is by binding to RNAP II and travelling
with it during elongation. Binding of activator proteins to the promoter results in
recruitment of chromatin-modifying activities and local chromatin disruption (1). With
the assistance of the general transcription factors, RNAP II then binds to the promoter
region (2). Shortly after the initiation of transcription, the C-terminal domain (CTD) of
RNAP II is hyperphosphorylated (3). RNAP II with a hyperphosphorylated CTD is
recognized and bound by chromatin-modifying factors such as PCAF and/or
elongator (4). The chromatin-modifying activities then travel with elongating RNAP II,
leading to the propagation of chromatin disruption (5). This mechanism would enable
the chromatin-modifying activities to gain access to the entire transcribed region and
would result in the propagation of chromatin disruption through the entire length of
the gene. For illustrative purposes only, the template ahead of the polymerase is
shown here as a compacted chromatin fibre.
© 2000 Macmillan Magazines Ltd
review article
474 NATURE
|
VOL 407
|
28 SEPTEMBER 2000
|
www.nature.com
paction of chromatin over the whole gene. In this case, the elongating
RNAP II would find partially decompacted nucleosomes in its path
(Fig. 3).
Very little experimental evidence exists to support either of these
models. If the first one is correct, RNAP II must penetrate and
unpackage a repressive chromatin fibre. To facilitate this, it is possible
that the first polymerase to transcribe a gene is a specialized ‘pioneer’
polymerase equipped with additional tools that break down higher-
order chromatin structures. Histone acetyltransferases that travel
with subsequent elongating polymerases would then function to
maintain transcribed chromatin in an accessible conformation. If a
pioneer RNAP II complex exists, it is likely to be in low abundance
and, therefore, may be difficult to purify by conventional means. Nev-
ertheless, experiments in S. cerevisiae suggest that the first polymerase
to transcribe a given gene is different from subsequent polymerases.
Upon transcription initiation, the mediator complex of S. cerevisiae
RNAP II, which is associated with a subset of RNAP II complexes in
the cell, is left at the promoter, suggesting that subsequent polymeras-
es loaded onto the gene at the promoter do not require associated
mediator complexes (S. Hahn, personal communication). A predic-
tion of this model is that the pioneer polymerase, faced with the
repressive, compacted chromatin fibre, will elongate more slowly than
will the polymerases that follow. This does not seem to be the case: on
the Drosophila HSP70 gene, the first polymerase synthesizes a tran-
script at a rate equal to subsequent polymerases
40
. Unlike most pro-
moters, however, the unactivated form of the HSP70 promoter
contains a transcriptionally engaged, paused RNAP II, and one can-
not rule out that a pioneer polymerase has already transcribed the
gene to establish an accessible chromatin conformation. Moreover,
the transcription complex formed at heat shock promoters, such as
the HSP70 gene promoter, appears to be different from that formed
on most promoters
41,42
. Thus, further definition of the conformation
of the chromatin template faced by elongating RNAP II requires
direct examination of the structure of transcribed chromatin subse-
quent to transcription initiation, both before and after elongation.
Does intergenic transcription disrupt chromatin
In some cases, chromatin modification and disruption are not
restricted to promoters and genes, but are spread over complete gene
loci. For example, the entire 33 kb of the chicken b-globin locus,
including intergenic regions, contains disrupted chromatin with
acetylated histones
6
. The human b-globin locus consists of five devel-
opmentally regulated genes whose expression is controlled by a
stretch of DNA known as the locus control region (LCR). The LCR
contains clusters of transcription-factor-binding sites and can pro-
mote transcription of genes in an adjacent chromatin region when
placed at any chromosomal location
43
. Two hypotheses have been put
forward to explain the way in which an open chromatin conforma-
tion is established at the b-globin locus to facilitate transcription. The
first proposes that proteins bound to the LCR recruit factors that dis-
rupt chromatin over the entire b-globin locus; the second hypothesis
directly implicates transcription elongation in generating an open
chromatin conformation
33,44
.
If transcription elongation is directly involved in disrupting chro-
matin at the b-globin locus, one would predict the existence of large
intergenic transcripts in cells containing active b-globin loci. Striking-
ly, large and rare intergenic transcripts derived from the human b-
globin locus have been detected
44
. Recently, Frazer and co-workers
45
examined the link between intergenic transcription and chromatin
disruption and found the locus to be divided into three distinct chro-
matin domains, each of which corresponds to a specific intergenic
transcript. A tight temporal correlation exists between the synthesis of
these transcripts and chromatin structural changes associated with
activation of specific genes in the locus. Moreover, the intergenic tran-
scripts appear to originate from precise regions, at least one of which
has been implicated in chromatin disruption. These observations
have challenged the long-standing hypothesis that the LCR of the b-
globin locus is necessary and sufficient for chromatin disruption.
Indeed, recent studies have demonstrated that the LCR of the mouse
b-globin locus is not required for chromatin disruption, but is
required for high-level transcription of the b-globin genes
46
. It will
be interesting to determine whether intergenic transcription is
involved in chromatin disruption at other gene loci and whether
a distinct form of the polymerase, or a pioneer polymerase, is
responsible.
Future directions
Once thought to be an inert structural feature, chromatin is now
known to be an integral part of the machinery that controls gene
expression. The progress described here sets the stage for the next
decade of experiments aimed at understanding how chromatin is
disrupted and re-structured during transcription. Future work will
reveal the molecular details of these chromatin changes, including
the pattern of histone acetylation, changes in nucleosome confor-
mation and the binding of accessory proteins. Attempts must be
made to determine the role of histone tail acetylation by HAT activi-
ties that track with RNAP II. This includes the mapping of histone
tail lysine residues acetylated during elongation, and the identifica-
tion and characterization of proteins that recognize these modifica-
tions. Engineering strains of the yeast S. cerevisiae with mutations
that inactivate these HAT activities, or replace their target lysine
Activator binding and chromatin
disruption over the entire gene
Activator binding and local
chromatin disruption only
Propagation of chromatin disruption
by pioneer polymerase
Transcription elongation
and maintainance of disrupted
chromatin conformation
= Nucleosome
= Activator
protein
= Elongating
polymerase
= Pioneer
polymerase
= Remodelled
nucleosome
1
*
2
Figure 3 Models for the order of events in the decompaction of transcribed
chromatin. In model 1, activator proteins bound to promoter regions promote only a
local decompaction of chromatin surrounding their DNA-binding sites. This leaves the
remainder of the gene in a compacted or partially decompacted conformation. In this
case, the elongating RNAP II will face a repressive chromatin structure. If this model
is correct, disruption of transcribed chromatin may be accomplished by a dedicated
‘pioneer' polymerase equipped with tools to penetrate and decompact higher-order
chromatin structure (asterisk). Transcribed chromatin may then be maintained in an
accessible conformation by HAT complexes associated with subsequent
polymerases. In an alternative model (2), activators promote chromatin decompaction
over the entire gene. In this case, RNAP II will find only nucleosomes in its path.
© 2000 Macmillan Magazines Ltd
review article
NATURE
|
VOL 407
|
28 SEPTEMBER 2000
|
www.nature.com 475
21.Brown, S. A., Imbalzano, A. N. & Kingston, R. E. Activator-dependent regulation of transcriptional
pausing on nucleosomal templates. Genes Dev. 10, 1479–1490 (1996).
22.Bustin, M. & Reeves, R. High-mobility-group chromosomal proteins: architectural components that
facilitate chromatin function. Prog. Nucleic Acid Res. Mol. Biol. 54, 35–100 (1996).
23.Ding, H. F., Rimsky, S., Batson, S. C., Bustin, M. & Hansen, U. Stimulation of RNA polymerase II
elongation by chromosomal protein HMG-14. Science 265, 796–799 (1994).
24.Rowley, A., Singer, R. A. & Johnston, G. C. CDC68, a yeast gene that affects regulation of cell
proliferation and transcription, encodes a protein with a highly acidic carboxyl terminus. Mol. Cell.
Biol. 11, 5718–5726 (1991).
25.Malone, E. A., Clarke, C. D., Chiang, A. & Winston, F. Mutations in SPT16/CDC68 suppress cis- and
trans-acting mutations that affect promoter function in Saccharomyces cerevisiae. Mol. Cell. Biol. 11,
5710–5717 (1991).
26.Orphanides, G., Wu, W. H., Lane, W. S., Hampsey, M. & Reinberg, D. The chromatin-specific
transcription elongation factor FACT comprises human SPT16 and SSRP1 proteins. Nature 400,
284–288 (1999).
27.Santisteban, M. S., Arents, G., Moudrianakis, E. N. & Smith, M. M. Histone octamer function in vivo:
mutations in the dimer-tetramer interfaces disrupt both gene activation and repression. EMBO J. 16,
2493–2506 (1997).
28.Baer, B. W. & Rhodes, D. Eukaryotic RNA polymerase II binds to nucleosome cores from transcribed
genes. Nature 301, 482–488 (1983).
29.Hartzog, G. A., Wada, T., Handa, H. & Winston, F. Evidence that Spt4, Spt5, and Spt6 control
transcription elongation by RNA polymerase II in Saccharomyces cerevisiae. Genes Dev. 12, 357–369
(1998).
30.Winston, F. & Carlson, M. Yeast SNF/SWI transcriptional activators and the SPT/SIN chromatin
connection. Trends Genet. 8, 387–391 (1992).
31.Wada, T. et al. DSIF, a novel transcription elongation factor that regulates RNA polymerase II
processivity, is composed of human Spt4 and Spt5 homologs. Genes Dev. 12, 343–356 (1998).
32.Bortvin, A. & Winston, F. Evidence that Spt6p controls chromatin structure by a direct interaction
with histones. Science 272, 1473–1476 (1996).
33.Travers, A. Chromatin modification by DNA tracking. Proc. Natl Acad. Sci. USA 96, 13634–13637
(1999).
34.Yang, X. J., Ogryzko, V. V., Nishikawa, J., Howard, B. H. & Nakatani, Y. A p300/CBP-associated factor
that competes with the adenoviral oncoprotein E1A. Nature 382, 319–324 (1996).
35.Cho, H. et al. A human RNA polymerase II complex containing factors that modify chromatin
structure. Mol. Cell. Biol. 18, 5355–5363 (1998).
36.Otero, G. et al. Elongator, a multisubunit component of a novel RNA polymerase II holoenzyme for
transcriptional elongation. Mol. Cell 3, 109–118 (1999).
37.Wittschieben, B. O. et al. A novel histone acetyltransferase is an integral subunit of elongating RNA
polymerase II holoenzyme. Mol. Cell 4, 123–128 (1999).
38.Strahl, B. D. & Allis, C. D. The language of covalent histone modifications. Nature 403, 41–45 (2000).
39.Walia, H., Chen, H. Y., Sun, J. M., Holth, L. T. & Davie, J. R. Histone acetylation is required to
maintain the unfolded nucleosome structure associated with transcribing DNA. J. Biol. Chem. 273,
14516–14522 (1998).
40.O’Brien, T. & Lis, J. T. Rapid changes in Drosophila transcription after an instantaneous heat shock.
Mol. Cell. Biol. 13, 3456–3463 (1993).
41.Lee, D. & Lis, J. T. Transcriptional activation independent of TFIIH kinase and the RNA polymerase II
mediator in vivo. Nature 393, 389–392 (1998).
42.McNeil, J. B., Agah, H. & Bentley, D. Activated transcription independent of the RNA polymerase II
holoenzyme in budding yeast. Genes Dev. 12, 2510–2521 (1998).
43.Fraser, P. & Grosveld, F. Locus control regions, chromatin activation and transcription. Curr. Opin.
Cell Biol. 10, 361–365 (1998).
44.Ashe, H. L., Monks, J., Wijgerde, M., Fraser, P. & Proudfoot, N. J. Intergenic transcription and
transinduction of the human b-globin locus. Genes Dev. 11, 2494–2509 (1997).
45.Gribnau, J., Diderich, K., Pruzina, S., Calzolari, R. & Fraser, P. Intergenic transcription and
developmental remodelling of chromatin subdomains in the human
b
-globin locus. Mol. Cell 5,
377–386 (2000).
46.Bender, M. A., Bulger, M., Close, J. & Groudine, M. b-globin switching and DNase I sensitivity of the
endogenous b-globin locus in mice do not require the locus control region. Mol. Cell 5, 387–393
(2000).
47.Kuo, M. -H., Zhou, J., Jambeck, P., Churchill, M. E. A. & Allis, C. D. Histone acetyltransferase activity
of yeast Gcn5p is required for activation of target genes in vivo. Genes Dev. 12, 627–639 (1998).
48.Felsenfeld, G., Boyes, J., Chung, J., Clark, D. & Studitsky, V. Chromatin structure and gene expression.
Proc. Natl Acad. Sci. USA 93, 9384–9388 (1996).
49.Roeder, R. G. The role of the general initiation factors in transcription by RNA polymerase II. Trends
Biochem. Sci. 21, 327–335 (1996).
50.Uptain, S. M., Kane, C. M. & Chamberlin, M. J. Basic mechanisms of transcript elongation and its
regulation. Annu. Rev. Biochem. 66, 117–172 (1997).
51.Neugebauer, K. M. & Roth, M. B. Transcription units as RNA processing units. Genes Dev. 11,
3279–3285 (1997).
52.Koleske, A. J. & Young, R. A. The RNA polymerase II holoenzyme and its implications for gene
regulation. Trends Biochem. Sci. 20, 113–116 (1995).
Acknowledgements
We thank M. Hampsey, D. Luse and S. Hahn for helpful comments, and R. Sternglanz, J. Workman and S.
Hahn for communication of unpublished results. Work in D.R.’s laboratory is supported by grants from
the NIH and the Howard Hughes Medical Institute.
Correspondence should be addressed to D.R. (e-mail: reinbedf@umdnj.edu).
residues in histones with inert amino acids, will be crucial in unravel-
ing these problems. The histone proteins of chromatin are subject to a
plethora of covalent modifications, including phosphorylation and
methylation, and it remains to be seen whether other histone modifi-
cations are involved in transcription elongation. It is unlikely that a
general mechanism exists for the derepression of chromatin for tran-
scription: for some genes (for example, those regulated by the GCN5
HAT
47
), histone acetylation appears to be restricted to the promoter
region. Deciphering the underlying reasons for this diversity will be a
major task. Biochemical studies have identified factors (for example,
FACT, HMG14 and remodelling complexes) that help RNAP II elon-
gate through chromatin. The observation that, even in the presence
of these factors, elongation by RNAP II proceeds much more slowly
in vitro than it does in vivo, suggests that other critical factors remain
to be discovered.
Our progress in understanding mechanisms that regulate
chromatin dynamics is limited by the lack of information regarding
the overall organization of compacted chromatin; we cannot fully
comprehend how chromatin is taken apart unless we first know how
it is put together. Currently, our knowledge of chromatin decom-
paction extends only to the disruption of nucleosomes—the first
level of chromatin compaction. The DNA template goes through
many more degrees of folding before a mature chromatin fibre is
formed, and we have yet to discover the machinery that mediates this
folding. Transcription is proving to be an ideal model system for
studying transitions in chromatin structure. It is likely that the
lessons learned from transcription will be relevant to other processes
that use DNA as a template and will have implications for diverse
biological processes. ■■
1. Kornberg, R. D. & Lorch, Y. Twenty-five years of the nucleosome, fundamental particle of the
eukaryotic chromosome. Cell 98, 285–294 (1999).
2. van Holde, K. E. Chromatin (Springer, New York, 1989).
3. Orphanides, G., Lagrange, T. & Reinberg, D. The general transcription factors of RNA polymerase II.
Genes Dev. 10, 2657–2683 (1996).
4. Gross, D. S. & Garrard, W. T. Nuclease hypersensitive sites in chromatin. Annu. Rev. Biochem. 57,
159–157 (1988).
5. Struhl, K. Histone acetylation and transcriptional regulatory mechanisms. Genes Dev. 12, 599–606
(1998).
6. Hebbes, T. R., Clayton, A. L., Thorne, A. W. & Crane-Robinson, C. Core histone hyperacetylation co-
maps with generalized DNase I sensitivity in the chicken beta-globin chromosomal domain. EMBO J.
13, 1823–1830 (1994).
7. Kuo, M.-H. & Allis, C. Roles of histone acetyltransferases and deacetylases in gene regulation.
BioEssays 20, 615–626 (1998).
8. Kingston, R. E. & Narlikar, G. J. ATP-dependent remodelling and acetylation as regulators of
chromatin fluidity. Genes Dev. 13, 2339–2352 (1999).
9. Lorch, Y., Zhang, M. & Kornberg, R. D. Histone octamer transfer by a chromatin-remodelling
complex. Cell 96, 389–392 (1999).
10.Owen-Hughes, T. & Workman, J. L. Experimental analysis of chromatin function in transcription
control. Crit. Rev. Euk. Gene Exp. 4, 403–441 (1994).
11.Lorch, Y., LaPointe, J. W. & Kornberg, R. D. Nucleosomes inhibit the initiation of transcription but
allow chain elongation with the displacement of histones. Cell 49, 203–210 (1987).
12.Clark, D. J. & Felsenfeld, G. A nucleosome core is transferred out of the path of a transcribing
polymerase. Cell 71, 11–22 (1992).
13.Studitsky, V. M., Clark, D. J. & Felsenfeld, G. A histone octamer can step around a transcribing
polymerase without leaving the template. Cell 76, 371–382 (1994).
14.Studitsky, V. M., Clark, D. J. & Felsenfeld, G. Overcoming a nucleosomal barrier to transcription. Cell
83, 19–27 (1995).
15.Nacheva, G. A. et al. Change in the pattern of histone binding to DNA upon transcriptional
activation. Cell 58, 27–36 (1989).
16.O’Neill, T. E., Roberge, M. & Bradbury, E. M. Nucleosome arrays inhibit both initiation and
elongation of transcripts by bacteriophage T7 RNA polymerase. J. Mol. Biol. 223, 67–78 (1992).
17.Studitsky, V. M., Kassavetis, G. A., Geiduschek, E. P. & Felsenfeld, G. Mechanism of transcription
through the nucleosome by eukaryotic RNA polymerase. Science 278, 1960–1963 (1997).
18.Izban, M. G. & Luse, D. S. Transcription on nucleosomal templates by RNA polymerase II in vitro:
inhibition of elongation with enhancement of sequence-specific pausing. Genes Dev. 5, 683–696
(1991).
19.Orphanides, G., LeRoy, G., Chang, C. -H., Luse, D. S. & Reinberg, D. FACT, a factor that facilitates
transcript elongation through nucleosomes. Cell 92, 105–116 (1998).
20.Izban, M. J. & Luse, D. S. Factor-stimulated RNA polymerase II transcribes at physiological
elongation rates on naked DNA but very poorly on chromatin templates. J. Biol. Chem. 267,
13647–13655 (1992).
© 2000 Macmillan Magazines Ltd