ã
Oncogene (2001) 20, 3094 ± 3099
2001 Nature Publishing Group All rights reserved 0950 ± 9232/01 $15.00
www.nature.com/onc
Chromatin alteration, transcription and replication: What's the opening line
to the story?
Michelle Craig Barton*,1 and Alison J Crowe2
1
Department of Biochemistry and Molecular Biology, University of Texas, M.D. Anderson Cancer Center, 1515 Holcombe Blvd.,
Houston, Texas, TX 77030, USA; 2Department of Zoology, University of Washington, Box 351800, Seattle, Washington, WA
98195-1800, USA
Polymerase accessibility to chromatin is a limiting step
in both RNA and DNA synthesis. Unwinding DNA and
nucleosomes during polymerase complex binding and
processing likely requires priming by chromatin restructuring. The initiating step in these processes remains an
area of speculation. This review focuses on the physical
handling of chromatin during transcription and replication, the fate of nucleosomes assembled on DNA during
unwinding and processing the chromatin substrate, and
how these alterations in chromatin structure may a€ect
gene expression. Transcription or replication may alter
chromatin structure during synthesis, enabling regulatory
factor binding and, potentially, future rounds of
transcription. As chromatin remodeling and transcription
factor binding augment transcription and replication, and
are themselves increased by these processes, a temporal
model of structural alterations and gene activation is
built that may be more circular than linear. Oncogene
(2001) 20, 3094 ± 3099.
Keywords: chromatin; replication; transcription; polymerase; remodeling
Introduction
A ¯urry of scienti®c attention has focused on
chromatin remodeling as a requisite step in gene
activation. Interactions between previously de®ned
chromatin remodeling complexes and nucleosomes/
histones are linked to enhanced gene transcription
(recently reviewed in Kingston and Narlikar, 1999;
Peterson and Workman, 2000; Sudarsanam and
Winston, 2000; Wol€e and Guschin, 2000), as well as
increased DNA replication (Flanagan and Peterson,
1999; Hu et al., 1999; Li, 1999; Li et al., 1998). Both of
these operations rely on large, megadalton-size complexes of proteins, physically processing a chromatin
substrate by lumbering along a nucleosome-assembled
DNA template or serving as nuclear matrix-associated
factory sites of synthesis (reviewed in Cook, 1999).
Could physical machination of chromatin in this way,
*Correspondence: MC Barton, Department of Biochemistry and
Molecular Biology, Box 117, University of Texas, M.D. Anderson
Cancer Center, 1515 Holcombe Blvd., Houston, TX 77030, USA
alter its structure and a€ect replication and transcription eciency? We will consider the fate of nucleosomes and chromatin structure during transcription
and replication, and review the consequences for gene
expression.
The substrate is chromatin
Chromatin as a hindrance to transcription activation
can inhibit binding of upstream (or downstream)
regulatory proteins as well as assembly of a preinitiation complex. Once the polymerase complex is
engaged, and an open complex formed, ecient
elongation through the gene body requires unwrapping
and decoding a nucleosome-assembled substrate. A
polymerase complex, whether RNA or DNA synthesizing, must somehow gain access to the correct strand of
DNA, reliably copy and transduce the information and
continue along the template (reviewed in Kodadek,
1998). During elongation, the encroaching polymerase
complex could either (a) force complete release of
nucleosomes, with re-assembly occurring de novo in the
wake of the processive complex, or (b) maintain
interaction between at least one strand of DNA and
nucleosomes or component histones in a type of
ternary complex. This interaction could facilitate a
continuum of regulation where a nucleosomal binding
pattern is maintained or re-established quickly. However, the time and modi®cation processes required to
reset chromatin to its ground state could provide a
window of structural plasticity during DNA and/or
RNA polymerase operations.
The transcription elongation process is one of starts,
stops and stutters; the consequences of which may be
modulated by trans-acting elongation factors such as
TFIIS, Elongin, ELL factors and others (reviewed in
Conaway et al., 2000; Orphanides and Reinberg, 2000).
Over a number of years, elegant studies by the Felsenfeld
laboratory and their collaborators support a model
where contact between histones and a DNA template is
maintained throughout elongation (Bednar et al., 1999;
Clark and Felsenfeld, 1992; Studitsky et al., 1994, 1995,
1997; Williamson and Felsenfeld, 1978). In vitro
investigations with E. coli RNA polymerase, SP6
bacteriophage RNA polymerase, S. cerevisiae RNA
polymerase III and histone/mononucleosome-associated
Gene activation by replication and transcription
MC Barton and AJ Crowe
DNA templates of varying lengths and sequence
composition, reveal that a nucleosome virtually steps
around the processive polymerase. A nucleosome from
the leading edge of the polymerase machinery is
internally transferred via a DNA loop to promoterproximal DNA. Arrested, short-lived RNA polymerase/
DNA intermediates have been visualized by electron
cryomicroscopy at the predicted positions along the
DNA template. Traversing the intranucleosomal DNA
loop substrate, RNA polymerases pause with a periodic
10-11 base pair frequency (Bednar et al., 1999).
The elongating RNA polymerase II complex:
Nimble dancer or chromatin snowplow?
The collective e€ects of nucleosome-mediated pausing
on a higher-order chromatin substrate are unknown.
Multiple lines of evidence reveal that chromatin
remodeling and/or histone modi®cation expedites the
rather inecient transcription elongation on a multinucleosome DNA template. ATP-dependent chromatin
remodeling complexes and histone modi®ers may
render transcription more ecient by modifying or
remodeling chromatin along the gene body or stabilizing nucleosome/loop formation (Brown and Kingston,
1997; Walia et al., 1998; reviewed in Workman and
Kingston, 1998). Biochemical studies revealed the
importance of the protein FACT (facilitates chromatin
transcription), which was puri®ed by its ability to
augment elongation processing of chromatin-assembled
templates (Orphanides et al., 1998; reviewed in
Orphanides and Reinberg, 2000). FACT works in a
two-step, ordered combination with chromatin-remodeling factor RSF, which promotes transcription
initiation. RSF-mediated nucleosome binding disruption, is localized to sites near the promoter while
FACT acts more downstream, binding and modifying
histones H2A and H2B interactions (LeRoy et al.,
1998; Orphanides et al., 1999).
Several proteins categorized originally by their roles in
transcription elongation are also modi®ers of chromatin
structure or associated with chromatin regulation. These
include the elongation factor DSIF, a human homologue
of yeast Spt4 and Spt5 proteins (Wada et al., 1998), and
elongation factor TFIIS, which was associated with the
SWI/SNF chromatin remodeling complex by a synthetic
lethal screen in S. cerevisiae (Davie and Kane, 2000). In
combination with disabled TFIIS activity (elongation
promotion), chromatin remodeling and/or histone modi®cation is essential for transcription of a greater number
of genes than seen with SWI/SNF disruption alone
(Davie and Kane, 2000). Additionally, the SWI/SNF
complex has a continued requirement in downstream
events following transcription initiation, implying a
potential role for chromatin remodeling during elongation (Biggar and Crabtree, 1999).
Perhaps the most intriguing link between chromatin
structure alteration and an elongating RNA polymerase complex is that remodeling and modi®er proteins
have been identi®ed as members of the polymerase
holoenzyme complex. Various laboratories have puri®ed RNA polymerase II holoenzyme preparations
with SWI/SNF complex members and histone acetyltransferases CBP and P/CAF among their components
(Cho et al., 1998; Neish et al., 1998; Wilson et al.,
1996). This puri®ed holoenzyme must necessarily be
disengaged from its substrate, and not every biochemical preparation of the holoenzyme RNA polymerase II
complex contains these remodeling factors (Cairns et
al., 1996; Kershnar et al., 1998; Neish et al., 1998).
However, additional studies have focused on isolation
of an engaged, elongating form of an RNA polymerase
II complex (Otero et al., 1999; Wittschieben et al.,
1999). Associated with the hyperphosphorylated RNA
polymerase II C-terminal domain (CTD) in a complex
with yeast chromatin and RNA, is a transcription
elongation factor called Elongator. One of the
Elongator subunits is a novel and highly conserved
histone acetyltransferase that is shared with the SAGA
complex, and can acetylate all four histone proteins
(Wittschieben et al., 1999, 2000).
Chromatin remodeling proteins and histone modi®ers in alliance with RNA polymerase II evoke a
previously used metaphor of an elongating RNA
polymerase as a chromatin snowplow. Does chromatin
structure remain altered in a measurable way in the
wake of the snowplow? Are there opposing forces in
equilibrium with remodeling factors and histone
acetyltransferases that re-establish a resting state to
transcribed chromatin after passage of a polymerase
complex?
3095
The trail of transcription
The relative depletion of histones H2A and H2B in
transcribed chromatin has been invoked as intriguing
evidence for nucleosome alteration post-transcription
(Baer and Rhodes, 1983). Sogo and colleagues, using
accessibility to psoralen crosslinking and 2D-gel
eletrophoresis analysis, show that chromatin structure
`opening' is initiated by RNA polymerase I transcribing ribosomal RNA gene repeats. The altered, active
chromatin is disrupted locally, and occupies relatively
short stretches of DNA along the path of the
processing polymerase complex (Lucchini and Sogo,
1995). These results imply that active chromatin
structure, which contains disrupted nucleosomes and
modi®ed histones, is a short-lived intermediate that is
not transmitted or copied during replication. Opposing
forces stripping histones of their modi®cation and
returning nucleosomes to full octamer composition
must therefore be targeted to recently transcribed
chromatin. However, timing may be very important
in interactions between regulatory factors and modi®ed
chromatin; and, di€erent measurements of chromatin
alteration induced by RNA polymerase II have
revealed further information as well. More recent
functional studies of RNA polymerase II action reveal
that ongoing transcription may potentiate further
chromatin activation (Gribnau et al., 2000).
Oncogene
Gene activation by replication and transcription
MC Barton and AJ Crowe
3096
Convincing evidence from Fraser and colleagues
supports a model where transcription begat transcription, possibly by physically perturbing chromatin as a
template (Ashe et al., 1997; Gribnau et al., 2000).
Building on observations made in the 1970's that large,
likely noncoding, transcripts exist across the globin
genes of chick and murine loci (Bastos and Aviv, 1977;
Imaizumi et al., 1973), these authors ®nd that
intergenic transcription is unidirectional, regulated
and initiates just prior to developmental activation of
stage-speci®c globin genes. Mapping nascent RNA
transcripts using in situ hybridization reveals three
subdomains of intergenic transcription within the
human b-globin locus (Gribnau et al., 2000). Termed
domain transcripts, these RNAs initiate at speci®c,
intergenic start sites, terminate within sharp boundaries
of closed chromatin structure and are not dependent
on concomitant globin gene-speci®c transcription.
Further, deletion of a domain transcript initiation
region results in considerable loss of downstream
globin gene expression. Taken together, a picture is
built of RNA polymerase II holoenzyme travel and
transcription across a gene locus that opens chromatin
structure for further alteration and speci®c, regulated
gene expression (reviewed in Travers, 1999).
Chromatin with potential
How could `priming' transcription occur in a repressed
chromatin domain? In a ®eld concerned with initiating,
regulatory events in chromatin structural alteration
and gene expression, transcription activating transcription poses a prototypical chicken-and-egg problem.
The concept of `transcriptional competence' across an
entire gene locus, distinct from transcriptional activity,
is supported by structural studies of the chick b-globin
locus. Locus-wide, tissue-speci®c sensitivity to DNaseI
digestion, 5 ± 10-fold greater than ¯anking genomic
DNA and globin loci in non-expressing tissues, comaps with broad, locus-wide histone hyperacetylation.
These tissue-speci®c alterations in globin locus chromatin structure occur prior to more localized, developmentally regulated DNaseI hypersensitive sites
where regulatory factors interact with sequence-speci®c
DNA elements (Hebbes et al., 1988, 1994).
Utilizing chromatin immunopreciptiation analysis of
the S. cerevisiae genome, the Grunstein laboratory
®nds that widespread, global histone modi®cation
exists across gene loci (Vogelauer et al., 2000). Unique
to these analyses is the level of quantitation applied
over a relatively broad region of the genome, and the
manipulation of histone acetyltransferase and deacetylase expression levels by gene deletion. These authors
®nd that multiple chromosomal domains exhibit an
equilibrium between acetylation and deacetylation, one
that can be shifted rapidly but favors a general
underacetylated condition. In a state dictated by a
balance of enzymatic actions, chromatin structure may
be more dynamic than previously considered. Targeting
by a minimal number of DNA-bound regulatory
Oncogene
factors could shift the equilibrium toward chromatin
activation. Further analyses of yeast and mammalian
genomes in this way should reveal whether all
euchromatin can be characterized by histone modi®cation at a quanti®able level, in¯uenced by tissue-,
developmental-, and ligand/signal-speci®c factors that
tip the balance toward activation or repression of
chromatin structure.
The case for DNA polymerase: a matter of timing
A recent review by Orphanides and Reinberg (Orphanides and Reinberg, 2000) discusses the concept of a
`pioneer [RNA] polymerase'. This polymerase would be
specially equipped to pry open unmodi®ed, fully
repressed chromatin, between borders of transcription
silencing, for future rounds of transcription. By
modifying histones and remodeling chromatin structure
with its appropriate protein subunits, the pioneer
polymerase complex could initiate chromatin structure
alteration to a level of readiness amenable to
regulatory factor interactions with DNA. As noted
by these same authors, there is evidence for primed
chromatin structure (discussed above), but little
experimental support for a special-forces RNA polymerase. Another candidate to consider for the title of
pioneer polymerase, ful®lling the quali®cations of
multi-subunit protein complex that traverses gene loci
and manipulates chromatin, is DNA polymerase.
The correlative links between DNA polymerase
action and gene expression are numerous and attractive
at several levels. Surveys of housekeeping genes, tissuespeci®c and developmentally expressed genes established the dogma that early replication (in S phase of
the cell cycle) means active gene expression (Goldman
et al., 1984). Comparison of a single gene in highly
expressing cells versus non-expressing cells, such as bglobin in red blood cells versus HeLa cells, reveals that
b-globin gene replication timing di€ers markedly
between the two cell types. Further, during red cell
development, the timing of b-globin replication
switches concomitantly with developmental expression
of the gene(s) (reviewed in Groudine et al., 1989;
Groudine and Weintraub, 1981; Kitsberg et al., 1993).
Replication late in the cell cycle correlates with silenced
gene expression, such as the mammalian inactive X
chromosome and yeast silenced mating type genes
compared to their actively expressed counterparts,
which are replicated early (reviewed in Fangman and
Brewer, 1992; Raghuraman et al., 1997).
Direct cause-and-e€ect between DNA replication
and gene transcription activation is more dicult to
establish. Recent analyses of inactive X chromosomes
isolated from patients with ICF (Immunode®ciency,
centromeric region instability, and facial anomalies)
syndrome indicate that late replication is sucient to
confer a silenced state even in the presence of
transactivator binding (Hansen et al., 2000). ICF is
caused by a point mutation in the methyl transferase
gene DNMT3B. The resulting hypomethylation of
Gene activation by replication and transcription
MC Barton and AJ Crowe
inactive X chromosomes and escape from silencing of
X-inactivated genes occurs concomitantly with a shift
in replication timing, linking the processes of replication and transcription. The switch from silenced to
activated requires multiple steps, including advanced
replication timing, hypomethylation, chromatin remodeling and transcription factor binding (Litt et al.,
1997; Sasaki et al., 1992). Again, it is not clear in what
order these steps must proceed but, with each
progressive alteration, there is an increased probability
that a gene will escape silencing.
The consequences of replication
DNA replication results in the transient disruption of
1 ± 2 nucleosomes at the replication fork; additionally,
nucleosomes assembled de novo after replication fork
passage may be transiently hyperacetylated and
depleted in histones H2A and H2B (reviewed in
Verreault, 2000; Wol€e, 1991). As a result, there may
be a brief opportunity for activators and/or repressors
to bind DNA regulatory sites normally occluded by
nucleosomes, and establish chromatin structure that
di€ers from the original replication substrate. Chromatin replication in synthetic nuclei directly links
alterations in chromatin structure mediated by passage
of a replication fork and activation of gene expression
(Barton and Emerson, 1994; Crowe et al., 2000).
Cytoplasmic Xenopus laevis egg extracts assemble
chromatin and an encapsulating, nuclear membrane
on DNA templates in vitro (Newport, 1987). Coupled
transcription analyses of chick b-globin and mouse
alpha-fetoprotein genes replicated in synthetic nuclei,
demonstrates replication-dependent transcription activation in the presence of tissue- and developmentalspeci®c trans-activators. Chromatin assembled in the
presence of trans-activators during primed, singlestranded DNA replication is activated for expression
and repressed when uncoupled from replication
(Almouzni et al., 1990; Kamakaka et al., 1993).
Returning to the previous study of intergenic domain
transcription across de®ned regions of the b-globin
locus (Gribnau et al., 2000), a temporal order of
transcription was established. In situ hybridization and
quanti®ed measurements of nascent transcripts at
speci®c times during the cell cycle reveals periodic
domain transcription at the time of globin replication
early in S phase, but also in late G1. The long-term
e€ects on chromatin structure and globin gene-speci®c
expression follow the early replication and transcription of locus subdomains, perhaps working in concert.
Replication-independent gene activation
Studies by Sogo and colleagues, using techniques of
psoraren accessibility, crosslinking, 2-dimensional gel
electrophoresis and electron microscopy, focus on the
relationship between DNA replication and transcription of ribosomal RNA genes (Lucchini and Sogo,
1995; Muller et al., 2000). Several ®ndings are of
interest: (1) there is no measurable di€erence between
the lengths of the short nucleosome-free regions at
replication forks within transcriptionally active and
transcriptionally inactive chromatin; (2) the amount of
time it takes to re-assemble nucleosomes in the wake
of DNA synthesis is also equal between the two types
of chromatin; and (3) active origins of replication
occur primarily downstream of an actively transcribing rRNA gene and not upstream. These authors
further suggest that it is transcription factor binding
at a rRNA enhancer, maintained in an open
nucleosome-free structure, that activates a downstream
replication origin and blocks replication fork movement upstream from the origin (Muller et al., 2000).
The open chromatin structure of the rRNA enhancer
does not appear to be dependent on ongoing
transcription or replication, but may be highly
sequence-directed. How these activities compare to
the majority of RNA polymerase II-transcribed genes
with ill-de®ned, broad regions of replication origin
needs to be determined.
A recent study of the yeast heat shock factor (HSF)
transcriptional activator found that HSF binding to its
nucleosomal site in vivo requires an S phase-speci®c
event that is replication-independent (Venturi et al.,
2000). These results indicate that, in addition to DNA
replication, there are other cell cycle-linked cellular
processes that facilitate factor binding in early S phase.
There is precedent for cell cycle regulation of
chromatin remodeling complexes: Human SWI/SNF
complexes are inactivated during mitosis through
phosphorylation, potentially to allow formation of a
transcriptionally repressed chromatin structure (Sif et
al., 1998). Inactivation of SWI/SNF during mitosis
may be necessary to allow chromosome condensation
to proceed in the absence of active remodeling. It is
possible, therefore, that S phase-speci®c induction of
SWI/SNF or related remodeling activities may enhance
transcription factor binding during this critical stage of
the cell cycle.
3097
The question remains
Though DNA replication a€ords an opportunity for
initiating changes in chromatin structure and gene
expression as replication forks move through the
genome, there is no unifying picture of DNA
replication as the global, primary modi®er of chromatin structure. Like transcription activation, DNA
replication can be ampli®ed through binding or
tethering a diverse array of transcription factors at
certain ARS (autonomously replicating sequence)
elements in S. cerevisiae (Hu et al., 1999; Li, 1999; Li
et al., 1998). Chromatin remodeling occurs at transcription factor-bound ARS elements, presumably
through targeting of ATP-dependent chromatin remodeling complexes such as SWI/SNF (Flanagan and
Peterson, 1999). We return again to the question of
how these transcription factors gain accessibility to
Oncogene
Gene activation by replication and transcription
MC Barton and AJ Crowe
3098
repressed chromatin to begin the cycle of chromatin
remodeling, replication and transcription.
The ability of transcription factors to bind their sites
when nucleosome-assembled is limited but highly
variable (Adams and Workman, 1995). Proteins such
as Swi5p may readily interact with chromatin and bind,
targeting the SWI/SNF chromatin remodeling complex, which then interacts with the SAGA complex,
permitting further interaction with transcription factors
at the HO gene locus in yeast (Cosma et al., 1999). As
other gene regulatory models are assessed for the
temporal process of transcription activation (Dilworth
et al., 2000; Gribnau et al., 2000), there will likely
emerge gene- or gene locus-speci®c initiators of a
sequential order of stages in gene expression. It is likely
that individual genes depend on highly regulated ways
of accessing chromatin structure, the combination of
which may be unique. An imbalance in any one of the
means of altering chromatin structure, e.g. increased
DNA replication, unregulated transcription or disrupted chromatin modi®er interactions, could feed into
a cascade of gene regulatory dysfunction through
escalating changes in chromatin structure.
References
Adams CC and Workman JL. (1995). Mol. Cell. Biol., 15,
1405 ± 1421.
Almouzni G, Mechali M and Wol€e AP. (1990). EMBO J., 9,
573 ± 582.
Ashe JL, Monks J, Wijgerde M, Fraser P and Proudfoot NJ.
(1997). Genes Dev., 11, 2494 ± 2509.
Baer BW and Rhodes D. (1983). Nature, 301, 482 ± 488.
Barton MC and Emerson BM. (1994). Genes Dev., 8, 2453 ±
2465.
Bastos RM and Aviv H. (1977). Cell, 11, 641 ± 650.
Bednar J, Studitsky VM, Grigoryev SA, Felsenfeld G and
Woodcock CL. (1999). Mol. Cell, 4, 377 ± 386.
Biggar SR and Crabtree GR. (1999). EMBO J., 18, 2254 ±
2264.
Brown SA and Kingston RE. (1997). Genes Dev., 11, 3116 ±
3121.
Cairns BR, Lorch Y, Li Y, Zhang M, Lacomis L, ErdjumentBromage H, Tempst P, Du J, Laurent B and Kornberg
RD. (1996). Cell, 87, 1249 ± 1260.
Cho H, Orphanides G, Sun X, Yang XJ, Ogryzko V, Lees E,
Nakatani Y and Reinberg D. (1998). Mol. Cell. Biol., 18,
5355 ± 5363.
Clark DJ and Felsenfeld G. (1992). Cell, 71, 11 ± 22.
Conaway JW, Shilatifard A, Dvir A and Conaway RC.
(2000). Trends Biochem. Sci., 25, 375 ± 380.
Cook PR. (1999). Science, 284, 1790 ± 1795.
Cosma MP, Tanaka T and Nasmyth K. (1999). Cell, 97,
299 ± 311.
Crowe AJ, Piechan JL, Sang L and Barton MC. (2000). Mol.
Cell. Biol., 20, 4169 ± 4180.
Davie JK and Kane CM. (2000). Mol. Cell. Biol., 20, 5960 ±
5973.
Dilworth FJ, Fromental-Ramain C, Yamamoto K and
Chambon P. (2000). Mol. Cell, 6, 1049 ± 1058.
Fangman WL and Brewer BJ. (1992). Cell, 71, 363 ± 366.
Flanagan JF and Peterson CL. (1999). Nucleic Acids Res., 27,
2022 ± 2028.
Goldman MA, Holmquist GP, Gray MC, Caston LA and
Nag A. (1984). Science, 224, 686 ± 692.
Gribnau J, Diderich K, Pruzina S, Calzolari R and Fraser P.
(2000). Mol. Cell, 5, 377 ± 386.
Groudine M, Forrester WC, Novak U and Epner E. (1989).
Prog. Clin. Biol. Res., 316A, 15 ± 35.
Groudine M and Weintraub H. (1981). Cell, 24, 393 ± 401.
Hansen RS, Stoger R, Wijmenga C, Stanek AM, Can®eld
TK, Luo P, Matarazzo MR, D'Espositio M, Feil R,
Gimelli G, Weemaes CM, Laird CD and Gartler SM.
(2000). Hum. Mol. Genet., 9, 2575 ± 2587.
Hebbes TR, Clayton AL, Thorne AW and Crane-Robinson
C. (1994). EMBO J., 13, 1823 ± 1830.
Oncogene
Hebbes TR, Thome AW and Crane-Robinson C. (1988).
EMBO J., 7, 1395 ± 1402.
Hu YF, Hao ZL and Li R. (1999). Genes Dev., 13, 637 ± 642.
Imaizumi T, Diggelmann H and Scherrer K. (1973). Proc.
Natl. Acad. Sci. USA, 70, 1122 ± 1126.
Kamakaka RT, Bulger M and Kadonaga JT. (1993). Genes
Dev., 7, 1779 ± 1795.
Kershnar E, Wu SY and Chiang CM. (1998). J. Biol. Chem.,
273, 34444 ± 34453.
Kingston RE and Narlikar GJ. (1999). Genes Dev., 13,
2339 ± 2352.
Kitsberg D, Selig S, Keshet I and Cedar H. (1993). Nature,
366, 588 ± 590.
Kodadek T. (1998). Trends Biochem. Sci., 23, 79 ± 83.
LeRoy G, Orphanides G, Lane WS and Reinberg D. (1998).
Science, 282, 1900 ± 1904.
Li R. (1999). J. Biol. Chem., 274, 30310 ± 30314.
Li R, Yu DS, Tanaka M, Zheng L, Berger SL and Stillman B.
(1998). Mol. Cell. Biol., 18, 1296 ± 1302.
Litt MD, Hansen RS, Hornstra IK, Gartler SM and Yang
TP. (1997). J. Biol. Chem., 272, 14921 ± 14926.
Lucchini R and Sogo JM. (1995). Nature, 374, 276 ± 280.
Muller M, Lucchini R and Sogo JM. (2000). Mol. Cell., 5,
767 ± 777.
Neish AS, Anderson SF, Schlegel BP, Wei W and Parvin JD.
(1998). Nucleic Acids Res., 26, 847 ± 853.
Newport J. (1987). Cell, 48, 205 ± 217.
Orphanides G, LeRoy G, Chang CH, Luse DS and Reinberg
D. (1998). Cell, 92, 105 ± 116.
Orphanides G and Reinberg D. (2000). Nature, 407, 471 ±
475.
Orphanides G, Wu WH, Lane WS, Hampsey M and
Reinberg D. (1999). Nature, 400, 284 ± 288.
Otero G, Fellows J, Li Y, de Bizemont T, Dirac AM,
Gustafsson CM, Erdjument-Bromage H, Tempst P and
Svejstrup JQ. (1999). Mol. Cell, 3, 109 ± 118.
Peterson CL and Workman JL. (2000). Curr. Opin. Genet.
Dev., 10, 187 ± 192.
Raghuraman MK, Brewer BJ and Fangman WL. (1997).
Science, 276, 806 ± 809.
Sasaki T, Hansen RS and Gartler SM. (1992). Mol. Cell.
Biol., 12, 3819 ± 3826.
Sif S, Stukenberg PT, Kirschner MW and Kingston RE.
(1998). Genes Dev., 12, 2842 ± 2851.
Studitsky VM, Clark DJ and Felsenfeld G. (1994). Cell, 76,
371 ± 382.
Studitsky VM, Clark DJ and Felsenfeld G. (1995). Cell, 83,
19 ± 27.
Studitsky VM, Kassavetis GA, Geiduschek EP and Felsenfeld G. (1997). Science, 278, 1960 ± 1963.
Gene activation by replication and transcription
MC Barton and AJ Crowe
Sudarsanam P and Winston F. (2000). Trends Genet., 16,
345 ± 351.
Travers A. (1999). Proc. Natl. Acad. Sci. USA, 96, 13634 ±
13637.
Venturi CB, Erkine AM and Gross DS. (2000). Mol. Cell.
Biol., 20, 6435 ± 6448.
Verreault A. (2000). Genes Dev., 14, 1430 ± 1438.
Vogelauer M, Wu J, Suka N and Grunstein M. (2000).
Nature, 408, 495 ± 498.
Wada T, Takagi T, Yamaguchi Y, Ferdous A, Imai T, Hirose
S, Sugimoto S, Yano K, Hartzog GA, Winston F,
Buratowski S and Handa H. (1998). Genes Dev., 12,
343 ± 356.
Walia H, Chen HY, Sun JM, Holth LT and Davie JR. (1998).
J. Biol. Chem., 273, 14516 ± 14522.
Williamson P and Felsenfeld G. (1978). Biochemistry, 17,
5695 ± 5705.
Wilson CJ, Chao DM, Imbalzano AN, Schnitzler GR,
Kingston RE and Young RA. (1996). Cell, 84, 235 ± 244.
Wittschieben BO, Fellows J, Du W, Stillman DJ and
Svejstrup JQ. (2000). EMBO. J., 19, 3060 ± 3068.
Wittschieben BO, Otero G, de Bizemont T, Fellows J,
Erdjument-Bromage H, Ohba R, Li Y, Allis CD, Tempst P
and Svejstrup JQ. (1999). Mol. Cell, 4, 123 ± 128.
Wol€e AP. (1991). J. Cell. Sci., 99, 201 ± 206.
Wol€e AP and Guschin D. (2000). J. Struct. Biol., 129, 102 ±
122.
Workman JL and Kingston RE. (1998). Annu. Rev.
Biochem., 67, 545 ± 579.
3099
Oncogene