Oncogene (2001) 20, 3047 ± 3054
ã 2001 Nature Publishing Group All rights reserved 0950 ± 9232/01 $15.00
www.nature.com/onc
Nuclear receptors coordinate the activities of chromatin remodeling
complexes and coactivators to facilitate initiation of transcription
F Je€rey Dilworth1 and Pierre Chambon*,1
1
Institut de GeÂneÂtique et de Biologie MoleÂculaire et Cellulaire (IGBMC), CNRS/INSERM/ULP/ColleÁge de France, BP163, 67404
Illkirch Cedex, CU de Strasbourg, France
Recent advances in the ®eld of in vitro chromatin
assembly have led to in vitro transcription systems which
reproduce in the test tube, in vivo characteristics of
ligand-dependent transcriptional activation by nuclear
receptors. Dissection of these systems has begun to
provide us with information concerning the underlying
molecular mechanisms. Through recruitment of coactivator proteins, nuclear receptors act ®rst to remodel
chromatin within the promoter region and then to recruit
the transcriptional machinery to the promoter region in
order to initiate transcription. Here we present a possible
sequential mechanism for ligand-dependent transcriptional activation by nuclear receptors and discuss the in
vitro and in vivo data that support this model. Oncogene
(2001) 20, 3047 ± 3054.
Keywords: nuclear receptors; histone acetyltransferases;
ATP-dependent chromatin remodeling; coactivators; in
vitro transcription; RAR/RXR heterodimers
Introduction
Lipophilic ligands, such as steroid and non-steroid
hormones, play a critical role in the development and
homeostasis of virtually every tissue of vertebrates
through their regulatory e€ects on cell di€erentiation,
proliferation, apoptosis, and metabolism. It has long
been established that these ligands exert their e€ects by
regulating the expression of speci®c subsets of genes
within target tissues (Thompson et al., 1966). Modi®cation of chromatin structure was suggested to be
involved in this regulation of gene expression, as
treatment of insect larvae with the steroid hormone
ecdysone resulted in the pung of speci®c loci within
polytene chromosomes (Clever, 1965), but the mechanism by which this signal was transduced from hormone
to chromatin was unknown. The isolation of nuclear
receptors (NRs) demonstrating high anity binding for
speci®c steroids gave some insight into the mechanism
by which the cellular response to these hormone
signaling pathways could be mediated. Characterization of promoter regions within hormone-responsive
genes furthered our understanding when it was shown
*Correspondence: P Chambon
that they possess hormone response elements (REs)
which exhibiting a high anity binding for cognate
NRs.
Cloning of several members of the NR superfamily
allowed the classi®cation of these proteins into three
subtypes: type I (i.e., ER, GR, PR) ± which bind as
homodimers to REs composed of inverted repeats; type
II (i.e., RAR, RXR, TR, VDR) ± which bind as
heterodimers to REs composed of direct repeats; and
type III (i.e., SF1) ± which bind as monomers to
extended REs. Comparison of the primary structure
between these proteins revealed that all three NR types
share a highly conserved domain structure (reviewed in
Chambon, 1996; Evans, 1988; Green and Chambon,
1988). Functional analysis of these domains demonstrated that in addition to the ligand binding (LBD)
and DNA binding (DBD) domains, NRs possess two
autonomous transcriptional activation (AF-1 and AF2) functions (Chambon, 1996; Evans, 1988; Giguere,
1990; Green and Chambon, 1988; Gronemeyer, 1991;
Tassetet al., 1990). A transcription repression domain
was later identi®ed (Horlein et al., 1995). Threedimensional structural analysis of the LBD of several
NRs suggested how these regulatory proteins could
exhibit both transcriptional activator and repressor
activity, as it was shown that ligand binding caused a
transconformational change in the protein to induce
the AF-2 function (Bourguet et al., 1995; Chambon,
1996; Egea et al., 2000; Moras and Gronemeyer, 1998;
Shiau et al., 1998; Wurtz et al., 1996).
Experiments in transfected cells suggested how the
activation functions exert their e€ect on transcription
when competition between multiple NRs for a limiting
transcription factor indicated the existence of intermediary factors acting as coactivators (Meyer et al.,
1989; Tasset et al., 1990). It was assumed that there was
a similar need for intermediary factors to mediate
transcriptional repression. The use of in vivo (e.g., Le
Douarin et al., 1995) and in vitro (e.g., Cavailles et al.,
1994) techniques as tools to identify NR coactivators
(or corepressors) resulted in the isolation of a large
number of proteins interacting with NRs in the presence
(or absence) of their cognate ligand. Several of these
proteins have been con®rmed as either NR coactivators,
including the p160 (SRC-1, TIF2/GRIP1/SRC-2, pCIP/
AIB1/ACTR/RAC3/SRC-3) family of proteins, p300/
CBP, GCN5/PCAF, SWI/SNF and SMCC/TRAP/
Nuclear receptor-mediated transactivation
FJ Dilworth and P Chambon
3048
DRIP/ARC/PC2/CRSP (hereafter called SMCC); or
NR corepressors, including NCoR and SMRT (for
reviews see Glass and Rosenfeld, 2000; McKenna et al.,
1999).
Functional analysis of known nuclear receptor
coactivators has suggested two roles for these proteins
in mediating ligand-dependent transcriptional activation by NRs:
1. To modify chromatin structure, through the histone
acetyltransferase activity of p160 proteins (Chen et
al., 1997; Spencer et al., 1997), GCN5/PCAF (Yang
et al., 1996), and p300/CBP (Bannister and Kouzarides, 1996; Ogryzko et al., 1996) and/or the ATPdependent chromatin remodeling activity of the
SWI/SNF complexes (Hirschhorn et al., 1992; Kwon
et al., 1994), in order to generate a transcriptionally
permissive state within the promoter region.
2. To bridge the transcriptional machinery to NRs,
through the coactivator protein:protein interaction
domains of either p300/CBP (Nakajima et al., 1997)
or SMCC (Chiba et al., 2000., Malik and Roeder,
2000), in order to recruit the Pol II holoenzyme and
its associated TFII factors to promoter region.
As for transcriptional corepressors, NCoR- and
SMRT-containing complexes, are suggested to antagonize the chromatin modifying activities of coactivators
to generate a transcriptionally repressed state within
the promoter region (for review see Hu and Lazar,
2000).
While these `in cell' studies have provided us with
signi®cant insight into the mechanism by which
hormones regulate gene expression at speci®c promoters, the complexity of cellular systems makes it
dicult to dissect this process at the molecular level.
Such an understanding requires examination of this
process in a highly de®ned system, something which is
most easily achieved in the test tube. Crude in vitro
transcription systems have been established to study
ligand-dependent transcriptional activation by several
NRs (Dilworth et al., 1999, 2000; Kraus and
Kadonaga, 1998; Kraus et al., 1999; Liu et al., 1999;
Rachez et al., 1999). Dissection of these systems have
begun providing information with respect to the
molecular mechanism by which NRs mediate initiation
of transcription (Dilworth et al., 2000 and references
therein). Based on these in vitro transcription systems,
and other in vivo and in vitro results, we propose a
model (Figure 1) where NR-mediated transactivation
occurs through seven temporally-ordered steps at the
promoter of hormone responsive genes.
Step 1
The initial step in transactivation is the binding of
receptor dimers to their RE within the regulatory
region of hormone responsive genes. DNase I footprinting data has demonstrated that RAR/RXR heterodimers are only weakly associate with their RE when
the promoter is incorporated into a nucleosomal
template (Dilworth et al., 2000). Tight binding of both
RAR/RXR heterodimers (Dilworth et al., 2000) and
Oncogene
PR homodimers (Di Croce et al., 1999) to their
responsive elements in a chromatin context requires
the activity of ISWI containing ATP-dependent
chromatin remodeling complexes. In the case of the
PR, it has been shown that the presence of the receptor
homodimers can recruit ISWI to the chromatin
template (Di Croce et al., 1999). While the need for
ligand to mediate this recruitment was not examined,
several in vitro studies have demonstrated that tight
binding of both type I and II NRs to their cognate REs
occurs in a ligand-independent manner (Dilworth et al.,
2000; Kraus and Kadonaga, 1998; Minucci et al., 1998).
Under certain cellular conditions NR occupancy of
response elements has been shown to be liganddependent (Bhattacharyya et al., 1997; Dey et al.,
1994; Savoldi et al., 1997). However, quantitative
studies have demonstrated that, in the case of the
ER, ligand binding does not change the anity of the
receptor homodimer for its speci®c responsive element
(Driscoll et al., 1996), thus indicating that binding of
the receptor to the template may be in¯uenced by other
factors within the cell. In the case of type I NRs, the
cytosolic localization of the activators is maintained by
associated heat shock proteins in the absence of ligand,
thus ensuring a ligand-inducible binding of these
proteins to their responsive element (reviewed in Pratt
and Toft, 1997). Interestingly, binding of antagonistic
ligands to either ER or PR permit the binding of the
homodimers to their respective response elements in
vivo without inducing an active AF-2, further demonstrating that this activation function does not play a
role in receptor binding to the template (Lavinsky et
al., 1998; Smith et al., 1997). For type II NRs, it is
unclear what may be limiting the receptors from
interaction with their RE under certain cellular
conditions. As ligand-dependent binding of NRs to
their RE is observed in a cell- and promoter-speci®c
context, we speculate that heterodimer formation or
DNA binding may be regulated through other cellular
events or signaling pathways. Indeed, phosphorylation
of NRs has been shown to regulate dimerization (Bhat
et al., 1994; Chen et al., 1999a; Delmotte et al., 1999),
and other post-translational modi®cations have been
shown to a€ect DNA binding of transcriptional
activators (Gu and Roeder, 1997).
Interestingly, the need for an ATP-dependent
chromatin remodeling activity to mediate tight binding
of an activator to its RE is not universal, as ISWI
containing complexes were determined to be dispensable for transactivation by the Drosophila transcriptional activator Zeste (Kal et al., 2000). Not only might
the need for an ATP-dependent chromatin remodeling
activity to mediate tight binding of the proteins to their
response elements vary from activator to activator, but
it may also vary from promoter to promoter. It should
be noted that the receptor binding studies have been
performed on promoters in a transcriptionally repressed state. It remains to be seen whether there is a
need for an ATP-dependent chromatin remodeling
activity to mediate tight binding of NRs to a promoter
which is already in a `poised' transcriptionally
Nuclear receptor-mediated transactivation
FJ Dilworth and P Chambon
3049
Figure 1 Sequential model for retinoic acid-induced initiation of transcription mediated by RAR/RXR heterodimers. Step 1a:
ligand-independent binding of RAR/RXR heterodimer to its response element; Step 1b: localized chromatin remodeling mediated
by ISWI-containing complexes to permit tight binding of the heterodimer to its response element; Step 2: RA-dependent recruitment
of coactivators possessing histone acetyltransferase activity (p160 proteins and p300/CBP) to the promoter region via the AF-2
function of RAR; Step 3: acetylation of histones within the promoter region by p300/CBP and p160 proteins; Step 4: displacement
of p160 proteins from AF-2 of RAR to permit recruitment of SMCC to the same activation domain; Step 5: recruitment of the Pol
II holoenzyme to the promoter region by SMCC complex; Step 6: recruitment of SWI/SNF complex to the promoter region; Step 7:
recruitment of the basal transcription factors, SWI/SNF-mediated chromatin remodeling to permit their tight binding to the
promoter, and initiation of transcription
permissive state. Thus, the need for ATP-dependent
chromatin remodeling activities to mediate tight
binding of NRs to their response element as a general
rule for transcriptional regulation by these proteins will
have to be examined.
Step 2
The second step in NR-mediated transcriptional
initiation appears to be recruitment of coactivators
from the p160 and p300/CBP protein families to the
regulatory region. Indeed, the p160 proteins TIF2 and
SRC-1 have been demonstrated to enhance transactivation by RAR/RXR (Dilworth et al., 2000) and PR (Liu
et al., 1999), respectively, while p300 has been shown
to enhance ligand-dependent transcriptional activation
by both RAR/RXR (Dilworth et al., 1999, 2000) and
ER (Kraus and Kadonaga, 1998; Kraus et al., 1999).
Direct interactions between the p160 proteins and the
AF-2 function of NRs have been clearly demonstrated
Oncogene
Nuclear receptor-mediated transactivation
FJ Dilworth and P Chambon
3050
(Anzick et al., 1997; Chen et al., 1997; OnÄate et al.,
1995; Voegel et al., 1996) and would account for their
recruitment to the regulatory region. Once bound to
the NR, p160 proteins could then aid in the
recruitment of p300/CBP. Several studies have suggested that p300/CBP interacts directly with NRs
(Chakravarti et al., 1996; Kamei et al., 1996), though
the NR interacting domain of p300 is not necessary for
mediating maximal ligand-dependent transactivation
by ER homodimers in vitro (Kraus et al., 1999) and
TR/RXR heterodimers in vivo (Li et al., 2000). In
contrast, deletion of the p160 interacting domain of
p300 signi®cantly impaired the ability of p300 to
enhance ligand-dependent transactivation by the same
two NRs (Kraus et al., 1999; Li et al., 2000).
Furthermore, CBP was recruited to an antagonist
bound ER in the presence, but not in the absence, of a
genetically engineered p160 protein in which the NR
boxes were replaced by CoRNR boxes (Shang et al.,
2000). This suggests that p300 is recruited to hormone
REs through its interaction with p160 proteins, though
the direct interaction between the coactivators and the
NR may stabilize the tertiary complex. Once tethered
to the DNA via the NR, p300/CBP have been
suggested to increase the number of productive
preinitiation complexes present within the promoter
(Kraus and Kadonaga, 1998; Yie et al., 1999). The
mechanism by which p300/CBP achieves this increase
is not clear. Both p300 and CBP possess potent
acetyltransferase activities capable of covalently modifying histones, transcriptional activators, and coactivators (for review see Sterner and Berger, 2000).
Furthermore, p300/CBP can interact with Pol II
through the intermediate of RNA helicase A (Nakajima et al., 1997), and may recruit the Pol II holoenzyme
complex to the promoter region. If indeed, p300/CBP
carries out both functions, order of addition experiments have clearly demonstrated that acetylation of
histones by p300 precedes recruitment of basal
transcriptional machinery to the promoter (Dilworth
et al., 2000).
Step 3
Thus, in the third step, tethering of p160 and p300/
CBP to the promoter region leads to an increased level
of histone acetylation within the promoter region.
Consistent with this, the acetyltransferase activity of
p300/CBP has been shown to be essential to mediate
transactivation by both ER (Kraus et al., 1999) and
TR/RXR (Li et al., 2000) from repressive chromatin.
While other proteins may be acetylated during the
transcription initiation process (Imhof et al., 1997),
chromatin immunoprecipitation (ChIP) analysis of a
retinoic acid-responsive promoter in vitro clearly
demonstrated that p300 was recruited to the promoter
region by liganded RAR/RXR heterodimers to mediate
acetylation of histones within the template (Dilworth et
al., 2000). Ligand-dependent acetylation of histones
within the RA-responsive template could be further
increased by addition of the p160 protein TIF2, though
Oncogene
addition of TIF2 on its own did not result in signi®cant
acetylation of the template (Dilworth and Chambon,
unpublished results). These results are consistent with
the fact that lipophilic hormones induce hyperacetylation of histones at hormone responsive promoters in
cultured cells (Chen et al., 1999b; Shang et al., 2000)
and Xenopus embryos (Sachs and Shi, 2000). Interestingly, in the case of the cathepsin D gene, ER
recruitment of p300, but not GCN5/PCAF, coincided
with maximal levels of histone acetylation within the
promoter region (Shang et al., 2000).
While less evidence exists for other events occurring
during the third step, cell transfection studies have
suggested that the arginine methyltransferase CARM1
can synergize with p160 proteins and p300 to enhance
ligand-dependent transcriptional activation by NRs,
possibly through its ability to methylate histone H3
(Chen et al., 2000). A second study demonstrated that
p300 can also interact with the histone H2A/H2B
chaperone protein NAP1 (Shikama et al., 2000).
Recruitment of NAP1 to the nucleosomal template
by p300 may facilitate transfer of H2A/H2B dimers
from the histone octamer to the chaperone protein (Ito
et al., 2000; Shikama et al., 2000), through a
mechanism similar to that proposed for the transcription elongation factor FACT (Orphanides et al., 1999).
As acetylation of H3/H4 tetramers has been shown to
permit DNA to adopt a relaxed conformation
(Morales and Richard-Foy, 2000), it is interesting to
speculate that NAP1 may act to generate a transient
H3/H4 tetramer within the proximal promoter region
to facilitate access of basal transcription factors to the
DNA. However, this is only speculation as no direct
evidence exists for histone octamers disruption within
the promoter region of actively transcribing genes. As
all of the preceding events occur in the absence of the
basal transcriptional machinery, the `raison d'eÃtre' of
the ®rst three steps appear to be to prepare the
chromatin template for the recruitment of basal
transcription factors.
Step 4
The events following covalent modi®cation of histones
can be regarded as the recruitment step of transcriptional activation. While the order of events occurring
during this process and the mechanism by which it
occurs are less well de®ned, several studies have
provided information which allow us to speculate as
to what may be occurring. Indeed, roles for the SWI/
SNF (Dilworth et al., 2000) and SMCC (Fondell et al.,
1996; Rachez et al., 1998, 1999) complexes have been
shown during this stage of the initiation process in
vitro. Based on these studies, we suggest that the fourth
step could be an exchange of coactivators at the AF-2
surface of the NR-LBD. In vitro experiments have
demonstrated that both p160 proteins and TRAP220
compete for this protein interacting surface (Rachez et
al., 2000; Treuter et al., 1999). However, the mechanism underlying this coactivator exchange is still
unclear. One possibility is that p300 could acetylate
Nuclear receptor-mediated transactivation
FJ Dilworth and P Chambon
p160 proteins at lysine residues adjacent to their NR
interacting domains (Chen et al., 1999b). This acetylation of the p160 protein may destabilize the p160:NR
interaction, and thus could permit its displacement by
TRAP220 to allow binding of the SMCC complex to
the NR. The physiological relevance of this acetylation
has yet to be con®rmed.
Interestingly, ChIP analysis performed in cultured
cells have suggested that a single ER homodimer can
interact with TRAP220 and p160 proteins simultaneously (Shang et al., 2000). As ER homodimers
possess two AF-2 interaction surfaces, it is not clear
whether both coactivators are interacting with the same
receptor monomer. In vitro interaction studies performed using PPAR/RXR heterodimers have suggested
that in the presence of DNA and an agonistic ligand
for both receptors, the TRAP220 protein selectively
interacts with the PPAR moiety of the heterodimer
while p160 proteins interact with the RXR moiety
(Yang et al., 2000). However, activation of transcription by type II NR heterodimers is possible in the
presence of a single liganded partner (Dilworth et al.,
1999; Rachez et al., 1999). Thus, one could hypothesize
that in the presence of a single ligand, a coactivator
exchange must take place at the AF-2 function of type
II NRs, slowing the process of preinitiation complex
assembly. In contrast, the presence of an agonistic
ligand for the second moiety of the heterodimer could
allow simultaneous recruitment of p300/CBP, p160
protein and SMCC complexes to the promoter,
permitting a much faster assembly of the preinitiation
complex. Such a possibility could explain the synergistic activation of transcription observed in the presence
of ligands for both moieties of a heterodimer (Dilworth
et al., 1999 and references therein). Further experiments are required to validate the proposed coactivator
exchange and/or co-recruitment models.
Step 5
In the ®fth step, the DNA tethered SMCC complex
probably acts to recruit the Pol II holoenzyme to the
proximal promoter. An association between the SMCC
complex and Pol II has been demonstrated in vitro
(Chiba et al., 2000; Malik and Roeder, 2000). This
recruitment could be facilitated through a second
interaction between p300/CBP and the Pol II holoenzyme (Nakajima et al., 1997). Though SMCC and CBP
have been shown to be associated with the Pol II
holoenzyme in nuclear extracts, ChIPs analysis have
demonstrated di€erent kinetics for association of these
proteins with the promoter region (Agalioti et al., 2000;
Shang et al., 2000). Thus, we suggest that the NRs do
not recruit a preformed coactivator/Pol II complex but
rather the promoter tethered coactivators act to recruit
the Pol II holoenzyme to the DNA template.
Step 6
The sixth step in the process of NR-mediated initiation
of transcription is likely to be the recruitment of SWI/
SNF to the proximal promoter. In vitro order of
addition experiments demonstrate that maximal levels
of transactivation mediated by RAR/RXR can occur
when SWI/SNF is added to the reactions at the same
time as the basal transcription factors (Dilworth et al.,
2000). This is consistent with ChIP experiments
performed in vivo which suggest that SWI/SNF arrives
at IFNb promoter at the same time as TFIID (Agalioti
et al., 2000). Recruitment of SWI/SNF to the promoter
region is likely to be facilitated by NRs as transfected
cell systems have shown that the chromatin remodeling
complex interacts with GR in a ligand-dependent
manner (Fryer and Archer, 1998). Such an interaction
could be mediated through either the SNF2 (a or b) or
BAF250 subunits of SWI/SNF which have been shown
to interact in a ligand-stimulated manner with the ER
LBD (Ichinose et al., 1997) or with full-length GR (Nie
et al., 2000), respectively. In addition, the SNF2/
BAF250-containing subtype of SWI/SNF complexes
also contains the subunits BAF155 and BAF170 which
have been shown to interact in vitro with the zinc ®nger
DBD of the transcriptional regulators EKLF and
GATA1 (Kadam et al., 2000). Consistent with the fact
that NRs may interact with SWI/SNF through a
similar mechanism, experiments in transfected cells
have shown that the DBD of the GR was required to
mediate the interaction between the NR and the ATPdependent chromatin remodeling complex (Muchardt
and Yaniv, 1993). Alternatively, SWI/SNF could be
recruited to the promoter region, through an interaction between the bromodomain of the SNF2 subunit
and nucleosomal histones previously acetylated by
p300/CBP (Agalioti et al., 2000). Consistent with this
possibility, prior acetylation of nucleosomal histones
by p300 has been shown to facilitate the activity of
SWI/SNF on RAR/RXR mediated transactivation
(Dilworth et al., 2000). Thus, several distinct interactions between NRs, SWI/SNF and acetylated histones
within the DNA template are likely to occur to ensure
ecient recruitment of SWI/SNF to hormone responsive promoter.
3051
Step 7
Once stably associated with the promoter, SWI/SNF
may mediate the ®nal event in the transcriptional
initiation process. In this step, SWI/SNF probably acts
to remodel the proximal promoter region in order to
permit binding of the basal transcriptional machinery
to the DNA template. In vitro studies have demonstrated that TBP has a relatively weak anity for the
TATA box when the promoter element is incorporated
into a nucleosomal template (Imbalzano et al., 1994).
However, the association between TBP and the TATA
box can be facilitated by the enzymatic activity of
SWI/SNF (Imbalzano et al., 1994). Consistent with
this, ChIP analysis at the IFNb promoter has
demonstrated that recruitment of TFIID and SWI/
SNF to the promoter coincides with a remodeling of
nucleosomes within this region of the template and
initiation of transcription (Agalioti et al., 2000). Thus,
Oncogene
Nuclear receptor-mediated transactivation
FJ Dilworth and P Chambon
3052
using its ATP-dependent chromatin remodeling activity
to displace impeding nucleosomes within the proximal
promoter region, SWI/SNF likely stimulates tight
binding of the basal transcriptional machinery to the
DNA template. Once tightly associated with the
proximal promoter, the basal transcriptional machinery
can then initiate transcription at the hormone
responsive gene.
Conclusions/perspectives
Our current model (Figure 1) represents a possible
basic mechanism for ligand-dependent transcriptional
activation by NRs. This model is highly consistent with
that recently proposed for initiation of transcription
from the IFNb promoter by the enhancesome (Agalioti
et al., 2000) which suggests that it may correspond to a
general mechanism for activator-mediated transcriptional activation from repressive chromatin. Furthermore, the similarity of results obtained between
experiments performed on natural (Agalioti et al.,
2000) and synthetic (Dilworth et al., 2000) promoters
validates, at least to some extent, the use of `arti®cial'
DNA templates to study the basic mechanisms of
transcriptional activation.
While the sequence of events for the chromatin
modifying steps during ligand-dependent transcriptional activation have been clearly established (Dilworth et al., 2000), future studies are needed to con®rm
hypotheses put forth as to the order of events
occurring during the recruitment steps of NR-mediated
transactivation. In addition, current studies have not
delineated roles for several coactivators implicated in
NR signaling, including the GCN5/PCAF containing
complex TFTC (Brand et al., 1999; Ogryzko et al.,
1998) and the TAFs associated with TBP (Albright and
Tjian, 2000). Further characterization of the established in vitro transcription systems will be required to
elucidate the role of these proteins in NR-mediated
transactivation.
Even though the current in vitro transcription
systems are proving useful for elucidation of the basic
mechanism by which NRs mediate ligand-dependent
activation of transcription, the mechanism by which
NRs act in vivo to regulate di€erent cellular processes
obviously must be more complex. It is not conceivable
that the mechanism by which NRs mediate upregulation of a gene that is already transcribing would be the
same as that in which NRs mediate initiation of
transcription from a repressed promoter. Furthermore,
transcriptional response to hormonal stimuli are often
temporally distinct for di€erent loci in the same cell,
while the complement of genes activated by a single
hormonal stimuli may vary from one cell type to
another. Indeed, the key to NR-mediated regulation of
cellular processes is likely to lie in the complexity of the
promoters of the various hormonal responsive genes.
The precise positioning of hormonal REs with respect
to those of other transcriptional regulators, and
nucleosomal architecture, most likely permits a speci®c
combinatorial integration of several signaling pathways
to control the expression of genes in the cell.
Furthermore, the need for multiple NR coactivators
(and corepressors) will probably be shown to re¯ect
cell-speci®c promoter arrangements depending on the
actual cellular complement of transcriptional regulators. Thus, future research aimed at elucidating the
molecular mechanism(s) by which NRs regulate
developmental processes must focalize on understanding how transcription is regulated in the context of
complex natural promoters.
Abbreviations
AF-1: ligand-independent activation function; AF-2: ligand-dependent activation function; ChIP: chromatin
immunoprecipitation; DBD: DNA binding domain; ER:
estrogen receptor; GR: glucocorticoid receptor; LBD:
ligand binding domain; NR: nuclear receptor; PolII:
RNA polymerase II; PPAR: peroxisome proliferatoractivator receptor; PR: progesterone receptor; RA: retinoic
acid; RAR: retinoic acid receptor; RE: response element;
RXR: retinoid X receptor; SF1: steroidogenic factor 1; TR:
thyroid hormone receptor; VDR: vitamin D receptor.
Acknowledgments
We thank Marjorie Brand, Catherine Fromental-Ramain,
Laurent Gelman and Frank Vreede for useful discussions
and for critically reading the manuscript. Work performed
at the IGBMC was supported by grants from the Centre
National de la Recherche Scienti®que, the Institut National
de la Sante et de la Recherche MeÂdicale, the ColleÁge de
France, the Centre Hospitalier Universitaire ReÂgional, the
Association pour la Recherche sur la Cancer, the Fondation pour la Recherche Me dicale and Bristol-Myers
Squibb.
References
Agalioti T, Lomvardas S, Parehk B, Yie J, Maniatis T and
Thanos D. (2000). Cell, 103, 667 ± 678.
Albright SR and Tjian R. (2000). Gene, 242, 1 ± 13.
Anzick SL, Kononen J, Walker RL, Azorsa DO, Tanner
MM, Guan X-Y, Sauter G, Kallioniemi O-P, Trent JM
and Meltzer PS. (1997). Science, 277, 965 ± 968.
Bannister AJ and Kouzarides T. (1996). Nature, 384, 641 ±
643.
Oncogene
Bhat MK, Ashizawa K and Cheng SY. (1994). Proc. Natl.
Acad. Sci. USA, 91, 7927 ± 7931.
Bhattacharyya N, Dey A, Minucci S, Zimmer A, John S,
Hager G and Ozato K. (1997). Mol. Cell. Biol, 17, 6481 ±
6490.
Bourguet W, Ru€ M, Chambon P, Gronemeyer H and
Moras D. (1995). Nature, 375, 377 ± 382.
Nuclear receptor-mediated transactivation
FJ Dilworth and P Chambon
Brand M, Yamamoto K, Staub A and Tora L. (1999). J. Biol.
Chem., 274, 18285 ± 18289.
Cavailles V, Dauvois S, Danielian PS and Parker MG.
(1994). Proc. Natl. Acad. Sci. USA, 91, 10009 ± 10013.
Chakravarti., LaMorte VJ, Nelson MC, Nakajima T,
Schulman IG, Juguilon H, Montminy M and Evans RM.
(1996). Nature, 383, 99 ± 103.
Chambon P. (1996). FASEB J, 10, 940 ± 954.
Chen D, Huang SM and Stallcup MR. (2000). J. Biol.
Chem., 275, 40810 ± 40816.
Chen D, Pace PE, Coombes RC and Ali S.(1999a). Mol. Cell.
Biol., 19, 1002 ± 1015.
Chen H, Lin RJ, Schiltz RL, Chakravarti D, Nash A, Nagy
L, Privalsky ML, Nakatani Y and Evans RM. (1997). Cell,
90, 569 ± 580.
Chen H, Lin RJ, Xie W, Wilpitz D and Evans RM. (1999b).
Cell, 98, 675 ± 686.
Chiba N, Suldan Z, Freedman LP and Parvin JD. (2000). J.
Biol. Chem., 275, 10719 ± 10722.
Clever U. (1965). Chromosoma, 17, 309 ± 322.
Delmotte MH, Tahayato A, Formstecher and Lefebvre P.
(1999). J. Biol. Chem., 274, 38225 ± 38231.
Dey A, Minucci S and Ozato K. (1994). Mol. Cell. Biol., 14,
8191 ± 8201.
Di Croce L, Koop R, Venditti P, Westphal HM, Nightingale
KP, Corona DF, Becker PB and Beato M. (1999). Mol.
Cell, 4, 45 ± 54.
Dilworth FJ, Fromental-Ramain C, Remboutsika E, Benecke A and Chambon P. (1999). Proc. Natl. Acad. Sci.
USA, 96, 1995 ± 2000.
Dilworth FJ, Ramain-Fromental C, Yamamoto K and
Chambon P (2000). Mol. Cell, 6, 1049 ± 1058.
Driscoll MD, Klinge CM, Hilf R and Bambara RA. (1996).
J. Steroid Biochem. Mol. Biol., 58, 45 ± 61.
Egea PF, Mitschler A, Rochel N, Ru€ M, Chambon P and
Moras D. (2000). EMBO J., 19, 2592 ± 2601.
Evans RM. (1988). Science, 240, 889 ± 895.
Fondell JD, Ge H and Roeder RG. (1996). Proc. Natl. Acad.
Sci. USA, 93, 8329 ± 8333.
Fryer CJ and Archer TK. (1998). Nature, 393, 88 ± 91.
Giguere V. (1990). Genet. Eng., 12, 183 ± 200.
Glass CK and Rosenfeld MG. (2000). Genes. Dev., 14, 121 ±
141.
Green S and Chambon P (1988). Trends Genet., 4, 309 ± 314.
Gronemeyer H. (1991). Annu. Rev. Genet., 25, 89 ± 123.
Gu W and Roeder RG. (1997). Cell, 90, 595 ± 606.
Hirschhorn JN, Brown SA, Clark CD and Winston F.
(1992). Genes Dev., 6, 2288 ± 2298.
Horlein AJ, Naar AM, Heinzel T, Torchia J, Gloss B,
Kurokawa R, Ryan A, Kamei Y, Soderstrom M, Glass
CK and Rosenfeld MG. (1995). Nature, 377, 397 ± 404.
Hu I and Lazar MA. (2000). Trends Endocrinol. Metab,, 11,
6 ± 10.
Ichinose H, Garnier J-M, Chambon P and Losson R. (1997).
Gene, 188, 95 ± 100.
Imbalzano AN, Kwon H, Green MR and Kingston RE.
(1994). Nature, 370, 481 ± 485.
Imhof A, Yang X-J, Ogryzko VV, Nakatani Y, Wol€e AP
and Ge H. (1997). Curr. Biol., 7, 689 ± 692.
Ito T, Ikehara T, Nakagawa T, Kraus WL and Muramatsu
M. (2000). Genes Dev., 14, 1899 ± 1907.
Kadam S, McAlpine GS, Phelan ML, Kingston RE, Jones
KA and Emerson BM. (2000). Genes Dev., 14, 2441 ± 2451.
Kal AJ, Mahmoudi T, Zak NB and Verrijzer CP. (2000).
Genes Dev., 14, 1058 ± 1071.
Kamei Y, Xu L, Heinzel T, Torchia J, Kurokawa R, Gloss B,
Lin S-C, Heyman RA, Rose DW, Glass CK and Rosenfeld
MG. (1996). Cell, 85, 403 ± 414.
Kraus WL and Kadonaga JT. (1998). Genes Dev., 12, 331 ±
342.
Kraus WL, Manning ET and Kadonaga JT. (1999). Mol.
Cell. Biol., 19, 8123 ± 8135.
Kwon H, Imbalzano AN, Khavari PA, Kingston RE and
Green MR. (1994). Nature, 370, 477 ± 481.
Lavinsky RM, Jepsen K, Heinzel T, Torchia ., Mullen TM,
Schi€ R, Del-Rio AL, Ricote M, Ngo S, Gemsch J,
Hilsenbeck SG, Osborne CK, Glass CK, Rosenfeld MG
and Rose DW. (1998). Proc. Natl. Acad. Sci. USA, 95,
2920 ± 2925.
Le Douarin B, Pierrat B, vom Baur E, Chambon P and
Losson R. (1995). Nucleic Acid Res., 23, 876 ± 878.
Li J, O'Malley BW and Wong J. (2000). Mol. Cell. Biol., 20,
2031 ± 2042.
Liu Z, Wong J, Tsai SY, Tsai MJ and O'Malley BW. (1999).
Proc. Natl. Acad. Sci. USA, 96, 9485 ± 9490.
Malik S and Roeder RG. (2000). Trends Biochem. Sci., 25,
277 ± 283.
McKenna NJ, Xu J, Nawaz Z, Tsai SY, Tsai MJ and
O'Malley BW. (1999). J. Steroid Biochem. Mol. Biol., 69,
3 ± 12.
Meyer ME, Gronemeyer H, Turcotte B, Bocquel MT, Tasset
D and Chambon P. (1989). Cell, 57, 433 ± 442.
Minucci S, Wong, Blanco JC, Shi YB, Wol€e AP and Ozato
K. (1998). Molec. Endocrin., 12, 315 ± 324.
Morales V and Richard-Foy H. (2000). Mol. Cell. Biol., 20,
7230 ± 7237.
Moras D and Gronemeyer H. (1998). Curr. Opin. Cell. Biol.,
10, 384 ± 391.
Muchardt C and Yaniv M. (1993). EMBO J., 12, 4279 ± 4290.
Nakajima T, Uchida C, Anderson SF, Lee C-G, Hurwitz J,
Parvin JD and Montminy M. (1997). Cell, 90, 1107 ± 1112.
Nie Z, Xue Y, Yang D, Zhou S, Deroo BJ, Archer TK and
Wang W. (2000). Mol. Cell. Biol., 20, 8879 ± 8888.
Ogryzko VV, Schiltz RL, Russanova V, Howard BH and
Nakatani Y. (1996). Cell, 87, 953 ± 959.
Ogryzko VV, KotaniT, Zhang X, Schlitz RL, Howard T,
Yang XJ, Howard BH, Qin and Nakatani Y. (1998). Cell,
94, 35 ± 44.
OnÄate SA, Tsai SY, Tsai MJ and O'Malley BW. (1995).
Science, 270, 1354 ± 1357.
Orphanides G, Wu WH, Lane WS, Hampsey M and
Reinberg D. (1999). Nature, 400, 284 ± 288.
Pratt WB and Toft DO. (1997). Endoc. Rev., 18, 306 ± 360.
Rachez C, Gamble M, Chang CP, Atkins GB, Lazar MA and
Freedman LP. (2000). Mol. Cell. Biol., 20, 2718 ± 2726.
Rachez C, Lemon BD, Suldan Z, Bromleigh V, Gamble M,
Naar AM, Erdjument-Bromage H, Tempst P and Freedman LP. (1999). Nature, 398, 824 ± 828.
Rachez C, Suldan Z, Ward J, Chang C-PB, Burakov D,
Erdjument-Bromage H, Tempst P and Freedman LP.
(1998). Genes Dev., 12, 1787 ± 1800.
Sachs LM and Shi YB. (2000). Proc. Natl. Acad. Sci. USA.,
97, 13138 ± 13143.
Savoldi G, Fenaroli A, Ferrari F, Rigaud G, Albertini A and
Di Lorenzo D. (1997). DNA Cell. Biol., 16, 1467 ± 1476.
Shang Y, Hu X, DiRenzo J, Lazar MA and Brown M.
(2000). Cell, 103, 843 ± 852.
Shiau AK, Barstad D, Loria PM, Cheng L, Kushner PJ,
Agard DA and Greene GL. (1998). Cell, 95, 927 ± 937.
3053
Oncogene
Nuclear receptor-mediated transactivation
FJ Dilworth and P Chambon
3054
Oncogene
Shikama N, Chan HM, Krstic-Demonacos M, Smith L, Lee
CW, Cairns W and La Thangue B. (2000). Mol. Cell. Biol.,
20, 8933 ± 8943.
Smith CL, Nawaz Z and O'Malley BW. (1997). Molec.
Endoc., 11, 657 ± 666.
Spencer TE, Jenster G, Burcin MM, Allis CD, Zhou J,
Mizzen CA, McKenna NJ, Onate SA, Tsai SY, Tsai M-J
and O'Malley BW. (1997). Nature, 389, 191 ± 198.
Sterner DE and Berger SL. (2000). Microbiol. Mol. Biol.
Rev., 64, 435 ± 459.
Tasset D, Tora L, Fromental C, Scheer E and Chambon P.
(1990). Cell, 62, 1177 ± 1187.
Thompson EB, Tomkins GM and Curran JF. (1966). Proc.
Natl. Acad. Sci. USA, 56, 296 ± 303.
Treuter E, Johansson L, Thomsen JS, Warnmark A, Leers J,
Pelto-Huikko M, Sjoberg M, Wright AP, Spyrou G and
Gustafsson JA. (1999). J. Biol. Chem., 274, 6667 ± 6677.
Voegel JJ, Heine MJS, Zechel C, Chambon P and
Gronemeyer H. (1996). EMBO J., 15, 3667 ± 3675.
Wurtz J-M, Bourguet W, Renaud J-P, Vivat V, Chambon P,
Moras D and Gronemeyer H. (1996). Nature Struct. Biol.,
3, 87 ± 94.
Yang W, Rachez C and Freedman LP. (2000). Mol. Cell.
Biol., 20, 8008 ± 8017.
Yang, X-J, Ogryzko VV, Nishikawa J-I, Howard BH and
Nakatani Y. (1996). Nature, 382, 319 ± 324.
Yie J, Senger K and Thanos D. (1999). Proc. Natl. Acad. Sci.
USA, 96, 13108 ± 13113.