| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Genetic Institute, Justus Liebig University, Heinrich Buff Ring 58–62, D-35392 Giessen, Germany
1Correspondence: Genetic Institute, Justus Liebig University, Heinrich Buff Ring 58–62, D-35392 Giessen, Germany. E-mail: Aria.Baniahmad{at}gen.bio.uni-giessen.de
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONBIOLOGICAL ROLE OF CO-REPRESSORSCO-REPRESSORS AND DISEASESCO-REPRESSOR ACTIONMECHANISMS OF SILENCER-CO...OUTLOOKREFERENCES |
|---|
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONBIOLOGICAL ROLE OF CO-REPRESSORSCO-REPRESSORS AND DISEASESCO-REPRESSOR ACTIONMECHANISMS OF SILENCER-CO...OUTLOOKREFERENCES |
|---|
A large number of co-repressors have been identified (Table 1
). Co-repressors are not just recruited by a small selection of
transcriptional silencers. Rather, they bind to and mediate the
function of a wide range of transcriptional regulators. Specific
co-repressors are used for a particular transcriptional silencer.
Furthermore, specific interactions of particular co-repressors with
silencers can lead to additional recruitment of protein complexes that
also contain intrinsic repression function, thus increasing the
complexity of transcriptional silencing.
|
This is perhaps best illustrated by the various mechanisms used to
recruit histone deacetylase (HDAC) enzymes, which are often associated
with co-repressor function. Indirect recruitment of HDAC activity
involves different interaction partners depending on the
transcriptional regulator. One such targeting complex is the
SIN3-protein complex, which includes HDAC1 and 2. The SIN3 proteins
SIN3A and SIN3B bind directly to various silencers and target the HDAC
activity to these repressors (1
2
3
4
5
6
7
8
9
10
11
12
13
14)
. The SIN3 complex and
the associated HDAC activity can also be recruited by the nuclear
receptor co-repressor (NCoR)/silencing mediator for retinoic acid and
thyroid hormone receptor (SMRT) and Alien classes of co-repressors to
different silencers (5
, 6
, 15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30)
.
On the other hand, some transcriptional silencers are able to bind
directly to HDACs or use both direct and indirect recruitment of HDAC
to silence transcription. The retinoblastoma protein (RB) can directly
bind to HDACs independently of SIN3 proteins (31
32
33)
. In
addition, RB binds directly to Ski that associates with the SIN3
complex, thus indirectly recruiting HDAC activity (50)
.
Similarly, the NCoR and SMRT co-repressors bind directly to SIN3 and to
HDAC4, 5, and 7, whereas HDAC7 can then recruit the SIN3 proteins
(93
94
95)
. Furthermore, complexes containing HDAC activity
are not restricted to SIN3-containing complexes. Another co-repressor
complex, the Mi-2/NuRD (nucleosome remodeling and deacetylase) complex,
also contains HDAC activity, which is mediated by HDAC1 and 2 but is
not associated with the SIN3 proteins (reviewed in refs 96
, 97
). Thus, as seen in Table 1
, the same co-repressor can
associate specifically with different transcription factors and
transcriptional silencers can use multiple co-repressors to exert their
effect.
|
BIOLOGICAL ROLE OF CO-REPRESSORS |
|---|
|
TOPABSTRACTINTRODUCTIONBIOLOGICAL ROLE OF CO-REPRESSORSCO-REPRESSORS AND DISEASESCO-REPRESSOR ACTIONMECHANISMS OF SILENCER-CO...OUTLOOKREFERENCES |
|---|
Essential functions of co-repressors in development have been revealed
through knockout or mutant alleles in Drosophila and
vertebrates (98)
. Mutants of groucho, a
co-repressor for many of the repressor proteins in the
Drosophila embryo, including hairy and
dorsal, exhibit severe defects in neurogenesis,
segmentation, and initiation of sex determination (63
, 68)
. Reduction of the levels of Drosophila CtBP
(carboxy-terminal binding protein) co-repressor results in severe
segmentation and dorsoventral patterning defects, which can be
attributed to loss of repression activity of Krüppel, Knirps, and
Snail (77)
. The ecdysone receptor/ultraspiracle
heterodimer, a critical regulator of molting and metamorphic processes,
silences transcription by recruitment of the co-repressors Alien and
SMRTER (SMRT-related ecdysone receptor interacting protein) (30
, 57)
. A mutation in the ecdysone receptor that results in loss of
co-repressor interaction and silencing is characterized by
developmental abnormalities and lethality (57)
. In
Xenopus the homologue of dCtBP, XCtBP, is a co-repressor for
XTcf-3 that mediates dorsal ventral axis formation. Expression of a
mutant XCtBP that has activation rather than silencing function
indicates a critical role for XCtBP in the regulation of head and
notochord development (80)
.
In vertebrates, the SIN3 proteins and the associated HDAC activity have
been shown to associate directly with a number of transcriptional
repressors involved in essential cell functions. The Myc/Mad family of
basic helix-loop-helix (bHLH) proteins are transcription factors that
mediate diverse cellular functions including proliferation,
differentiation and apoptosis (reviewed in ref 99
).
Mad-Max heterodimers silence genes that are activated by Myc-Max and
thus antagonize Myc action. The Mad/Max-silencer binds to the SIN3
co-repressor complex, which also includes the co-repressors NCoR, Ski,
and HDACs (1
, 2
, 100
101
102)
. The essential role of the two
SIN3 proteins, SIN3A and SIN3B, in cellular transformation by the Mad
transcription factor is underlined by the observation that deletion of
the carboxyl terminus of SIN3B results in loss of transforming
potential of the Mad family (100)
.
SIN3A itself and the SIN3 co-repressor complex are also associated with
the Ikaros family of proteins, which regulate B and T cell
differentiation and proliferation of differentiated thymocytes and
mature T cells (reviewed in ref 103
). The Ikaros family of
proteins have been shown to be associated with two types of HDAC
complexes, the SIN3 co-repressor complex and the Mi-2/NuRD complex;
Ikaros can recruit the Mi-2/NuRD complex to the heterochromatic regions
(11
, 52)
. The role of these complexes in the functions of
the Ikaros family of proteins, however, is at present unclear.
The two co-repressor complexes SIN3 and Mi-2/NuRD are also involved in
transcriptional silencing mediated by methylated DNA. The presence of
methyl-cytosines in CpG groups has long been associated with tissue and
developmental specific gene regulation. Recently, factors containing a
methyl binding domain (MBD) have been shown to recruit HDAC activity
and chromatin remodeling activities to methyl-CpGs (reviewed in ref
104
). MeCP2, which specifically binds methyl-CpGs,
mediates transcriptional repression by a carboxyl-terminal repression
domain through interaction with the co-repressor SIN3A and recruits
HDACs (7
, 8)
. This provided the first functional link
between the presence of methyl-CpGs and transcriptional repression.
Since then, other repressing complexes involving MBD-containing
proteins have been identified. The MeCP1 complex that binds to
methylated DNA contains the methyl binding protein MBD2 along with HDAC
activity (105)
, whereas MBD3 is part of the Mi-2/NuRD
complex (54
, 106)
. The de novo DNA
methyltransferase, Dnmt1, has also been shown to contain a repression
domain that associates with HDACs (42)
.
The two tumor suppressor proteins, p53 and RB, which are involved in
cell cycle control and apoptosis, also rely on the SIN3 co-repressor
complex for their action. Both tumor suppressor proteins are targets of
viral oncoproteins. RB is a major regulator of the cell cycle at the
G1/S phase boundary. RB forms a complex with c-Ski, Sin3A, and HDACs
via direct binding to c-Ski and HDACs (31
32
33
, 50)
.
Furthermore, RB and the related protein, p130, also recruit the
CtBP/CtIP complex (82)
, indicating that there are multiple
mechanisms operating at the level of RB-mediated transcriptional
repression (Table 1)
. The p53 protein is often mutated in tumors and
has the dual function to activate and repress genes. The activity of
p53 is stimulated by a variety of agents to induce apoptosis or
programmed cell death (reviewed in ref 107
). This
p53-dependent apoptosis is specifically inhibited by the HDAC inhibitor
trichostatin A (TSA) (10)
, indicating a critical role of
HDACs in the apoptotic process. During apoptosis, p53 silences the
expression of various genes, including the microtubule-associated
protein Map4 (108)
. This silencing of Map4 expression
involves the recruitment of SIN3 proteins and the associated HDAC
activity that leads to a p53-dependent deacetylation of nucleosomes of
the Map4 promoter (10)
.
The SIN3 co-repressor complex is also associated with gene silencing by
nuclear hormone receptors (NHRs). NHRs are ligand-regulated
transcription factors that control key processes in development,
differentiation, and homeostasis. Some members of this superfamily have
the dual role to silence gene expression in the absence of ligand and
to activate gene transcription in the presence of their cognate hormone
(109)
. Silencing is mediated, at least in part, by the
hormone-sensitive interaction with the co-repressor proteins NCoR
(16)
, SMRT (15
, 110
, 111)
, and with Alien
(30)
, which bind to the SIN3 proteins and recruit HDAC
activity.
The SMRT/NCoR co-repressor class is not specific for the NHR
superfamily (Table 1)
. SMRT is also a co-repressor for POZ
domain-containing transcriptional silencers such as PLZF, which is
involved in leukemia-associated translocations, and BCL6, which is
involved in germinal center formation (5
, 19
20
21
22
23
24)
. The
homeo domain-containing protein Rpx, which plays a role in pituitary
determination, requires NCoR, the SIN3 proteins, and HDAC2 for
repression activity (27)
. NCoR is also necessary for the
activity of the Mad protein/SIN3 complex (100
, 101)
. In
addition, NCoR interacts with the myogenic factor MyoD and represses
MyoD-mediated trans-activation by an HDAC-dependent
mechanism (25)
. Moreover, specific functions of MyoD,
including differentiation of C2C12 cells and MyoD-mediated myogenic
conversion of pluripotent precursor cells, can be blocked by NCoR.
Co-repressors are also a major determinant for cell-specific gene
expression. Changes in the relative abundance of co-repressors or the
cell-specific expression of co-repressors may determine the relative
contribution of co-repressors to transcriptional repression mediated by
specific receptors. It has been demonstrated that limiting amounts of
the SMRT co-repressor compromises RAR-mediated but not TR-mediated
silencing (112)
. Thus, the level of certain cellular
co-repressors can control the hormonal response, and therefore
modulating target gene expression. Furthermore, cell-specific
expression of mSiah2 that promotes NCoR degradation also determines the
ability of NHRs to function as repressors in specific cell types
(113)
.
Thus, co-repressors are essential factors in gene regulation at all levels of biological pathways.
|
CO-REPRESSORS AND DISEASES |
|---|
|
TOPABSTRACTINTRODUCTIONBIOLOGICAL ROLE OF CO-REPRESSORSCO-REPRESSORS AND DISEASESCO-REPRESSOR ACTIONMECHANISMS OF SILENCER-CO...OUTLOOKREFERENCES |
|---|
An enhanced co-repressor–transcription factor interaction has been
associated with many forms of leukemia. Several types of leukemia
result from translocations involving fusions to transcriptional
silencers (for review, see ref 114
). Increased
co-repressor binding to these fusion proteins leads to inappropriate
repression of target genes important for normal cellular
differentiation.
The AML-1 gene is the target of chromosomal translocations that have
been associated with leukemia. Acute myelogenic leukemia (AML) has been
shown to involve the chromosomal translocation t(8;21), which fuses the
DNA binding domain of the AML protein to the transcriptional repressor
ETO/MTG8. Similarly, B cell acute lymphoblastic leukemia (ALL) involves
the chromosomal translocation t(12;21), which results in the fusion of
the repression domain of the ets-leukemia protein TEL to the large form
of AML, AML-1B. AML-1 normally activates and represses target gene
expression when bound to DNA. However, fusion proteins of AML-1 display
abnormal transcriptional properties and act mainly as repressors on
promoters. This function arises from the recruitment of NCoR and the
SIN3–HDAC complex by ETO (6
, 17
, 18)
and TEL (9
, 115)
. An inhibitor of HDAC1, phenylbutyrate, partially reverses
this ETO-mediated transcriptional repression and is also able to induce
partial differentiation of an AML1-ETO cell line (116)
.
This highlights the importance of the transcriptional repression by
HDACs in the development of leukemia.
The three chromosomal translocations t(11;17), t(15;17), and
t(5;17) involved in acute promyelocytic leukemia (APL) result in fusion
of the promyelocytic leukemia zinc finger protein PLZF, the
promyelocytic leukemia protein PML, or nucleophosmin, respectively
(117)
, to RAR (reviewed in ref 118
). All of
the resulting fusion proteins have been shown to play important roles
in leukemogenesis. These fusion proteins are able to bind to RAR target
sequences and repress transcription. The wild-type, unfused RAR
represses genes in the absence of its cognate ligand due to recruitment
of co-repressors, whereas in the presence of its ligand,
all-trans retinoic acid (t-RA), RAR activates target genes
by recruitment of coactivators. As expected, both the RAR-PML and
RAR-PLZF fusion proteins recruit the NCoR/SMRT co-repressors and the
associated HDACs via the SIN3 co-repressor in the absence of ligand
(24
, 119
, 120)
. However, the fusion receptors exhibit a
reduced responsiveness to retinoic acid and thus target genes remain
repressed (Fig. 1
) even in the presence of endogenous blood levels of t-RA. After
treatments with pharmacological levels of t-RA, patients with the
PML-RAR
fusion protein have complete, if somewhat transient,
remission of leukemia. This correlates with co-repressor dissociation
from RAR at higher concentrations of t-RA. However, patients with
the PLZF-RAR translocation are unresponsive to higher doses of t-RA
(24
, 120
, 121)
. This is related to the ability of the PLZF
moiety to retain co-repressors. Upon treatment with t-RA, co-repressors
dissociate from the RAR part of the fusion protein, whereas the PLZF
part remains attached to the co-repressor SMRT (23)
.
|
Furthermore, HDAC inhibitors, in combination with t-RA, can overcome
the transcriptional repressor activity as well as the unresponsiveness
of PLZF-RAR-expressing leukemic cells to t-RA (24
, 119
, 120)
. After a period of time, patients with PML-RAR fusions
become refractive to t-RA treatment. This refractiveness can be
overcome with sodium phenylbutyrate, an inhibitor of HDACs, which
results in clinical and cytogenetic remission shortly after treatment
(122)
. This points to the potential of a combined
treatment of APL patients with t-RA and HDAC inhibitors for more
successful therapy.
Increased stability of a co-repressor–silencer complex is also
involved in the thyroid hormone resistance syndrome (THR). This
syndrome is inherited in an autosomal dominant manner and is due to
mutations of the thyroid hormone receptor (TR) ß, a member of the NHR
superfamily (118
, 123
124
125)
. Patients with this syndrome
have a short physical stature due to a delay in bone growth and exhibit
hearing defects, learning disabilities, and mental retardation.
Furthermore, patients have higher levels of thyroid hormone, although
symptoms of hypothyroidism are manifested. Point mutations in TRß
reduce or completely abolish thyroid hormone binding capability,
whereby the receptor remains in the repressive state (126)
and remains attached to co-repressors (127
, 128)
despite
higher levels of blood thyroid hormone (Fig. 1)
. It is thought that
important genes required for normal development and cellular
differentiation remain abnormally silenced. Similarly, the oncogene
v-ErbA, which is a mutated form of TR, induces erythroblastosis and
sarcomas through repression of genes that are essential for normal
differentiation. The oncogene exhibits constitutive and dysregulated
binding to co-repressors, which leads to constitutive silencing
function (reviewed in ref 129
).
Examples of human diseases based on loss of co-repressor interaction
are represented by mutations of MeCP2, RB, and DAX-1. The latter is an
unusual member of the NHR superfamily in that it has a conserved
carboxyl-terminal domain, whereas the DNA binding domain differs from
other members. Mutations of the DAX-1 gene are associated with adrenal
hypoplasia and hypogonadism (130
131
132
133)
. DAX-1 is a potent
transcriptional silencer of the orphan nuclear receptor steroidogenic
factor 1, SF-1, and recruits the nuclear receptor co-repressors NCoR
and Alien, but not SMRT. Naturally occurring DAX-1 mutants have lost
silencing function, which correlates with loss of interaction with the
co-repressors NCoR and Alien (134
, 135)
(Fig. 1)
.
The Rett syndrome is an X-linked progressive neurodevelopmental
disorder. The genetic basis of this syndrome are mutations in the
X-linked methyl-CpG binding protein 2, MeCP2, that affect either the
methyl binding domain or the trans-repression domain
(136)
. These mutations result in loss of silencing
function and loss of recruitment of the SIN3 co-repressor to target
promoters. In these cases, transcriptional silencing of important genes
is lost (Fig. 1)
.
Naturally occurring mutants of the tumor suppressor protein RB from
patients who have developed cancer show that the ability of these
mutant RB proteins to repress E2F trans-activation is lost.
This results in loss of cell cycle regulation. Accordingly, mutant RBs
have been shown to lack interaction with HDACs and have lost silencing
function (reviewed in ref 137
). Thus, one
mechanism of generating cancer by mutation of the RB gene is the loss
of co-repressor interaction and loss of gene silencing.
Other viral oncoproteins also recruit co-repressors and this
association is important for disease formation. The oncoprotein E7 is
the main transforming protein of human papillomavirus 16, which plays a
major role in cervical cancer. E7 has been shown to bind to Mi-2 and
HDACs, and this HDAC association is necessary for the growth-promoting
effects mediated by the E7 protein (53)
.
Thus, co-repressors play an important role in human diseases and the co-repressor–silencer complex formation must be controlled in a stringent fashion.
|
CO-REPRESSOR ACTION |
|---|
|
TOPABSTRACTINTRODUCTIONBIOLOGICAL ROLE OF CO-REPRESSORSCO-REPRESSORS AND DISEASESCO-REPRESSOR ACTIONMECHANISMS OF SILENCER-CO...OUTLOOKREFERENCES |
|---|
).
It is generally accepted that the acetylation of the lysine residues of
the core histone tails is associated with transcriptionally active DNA
whereas the deacetylation of these lysine residues is associated with
transcriptional repression. The most widely accepted proposal
surrounding this observation is that the acetylation of the lysine
residues in the core histone tails results in a more open structure of
the chromatin. Thereby, chromatin is more readily accessible to
transcription factors. Deacetylation results in a more compact
chromatin structure and thus decreases accessibility of the chromatin
for transcription factors (for review, see ref 139
)
(Fig. 2
).
|
Many enzymes that catalyze either histone acetylation (HATs) or histone
deacetylation (HDACs) have been identified. The importance of HDAC
activity in mediating gene silencing is shown in cases of HDAC1 and
HDAC2/mRPD3 for which the HDAC activity is essential for the full
functional repression (140
, 141)
. In higher eukaryotes,
seven proteins have been identified so far that possess HDAC activity
(37
, 38
, 94
, 142
, 143)
. These can be divided into two
classes based on their sequence homology to yeast proteins and the
observation that the homologous proteins in yeast form distinct
complexes (144)
. Class I HDACs, consisting of HDAC 1–3,
share homologies to the yeast RPD3 whereas class II HDACs, consisting
of HDAC 4 to 7, share homologies to yeast Hda1 (94)
.
HDAC1 and 2 exist as core components of the two deacetylase
complexes characterized in eukaryotes: the SIN3 complex and the
Mi-2/NuRD complex (reviewed in refs 96
, 97
). These
complexes both contain the four core proteins HDAC1, HDAC2, which
possess deacetylase activity (37
, 142)
, and the
retinoblastoma-associated proteins RbAp46 and RbAp48, which appear to
target the complex by binding to the histones (145)
. The
SIN3 complex is composed of the four core proteins and a number of
other factors, including the SIN-associated proteins SAP18 and SAP30.
The mechanism of action and the biological role of these associated
factors in the context of the complex are largely unknown.
The Mi-2/NuRD complex, apart from the core proteins, contains MTA2, a
protein that is similar to the metastasis-associated protein MTA-1, the
methyl binding domain proteins MBD3a/b, and a protein with chromatin
remodeling activities, Mi-2ß/CHD4 (106
, 146
147
148
149)
. The
Mi-2ß protein is a member of the SWI/SNF family of proteins, which
are known to change DNA-histone interactions in an ATP-dependent
manner. Traditionally, these ATP-dependent chromatin remodeling
activities have been associated with transcriptional activation, but
have recently been shown to facilitate transcriptional repression by a
yet unknown mechanism. The ability of the NuRD complex to remodel
nucleosomes is necessary for deacetylation of oligonucleosomal histone
in vitro (149)
, suggesting that the remodeling
activity in some way influences the access of the deacetylating
components of the Mi-2/NuRD complex. Therefore, it has been postulated
that the remodeling activities of the Mi-2/NuRD components make the DNA
more accessible to the deacetylase and thus promote deacetylation.
Thus, co-repressors that recruit these HDACs mediate gene silencing by
targeting and modifying chromatin (for review, see ref
(150)
.
Mechanisms of co-repressor action on the basal transcription
machinery
The use of the HDAC inhibitors sodium butyrate and (more recently)
the specific inhibitor TSA have been instrumental in the identification
of proteins that have an associated HDAC activity. However, it has also
been revealed that in some cases not all of the silencing ability of
co-repressors could be relieved by these inhibitors (30
, 93
, 101
, 140
, 151)
. Accordingly, some co-repressors have been shown to
use multiple mechanisms to repress gene expression. This includes
interaction of co-repressors with factors of the basal transcriptional
machinery (Fig. 2)
.
RevErbß, a member of the NHR superfamily, can bind a complex
containing the co-repressor NcoR and the basal transcription factors
TAFII32, TAFII70, TFIIB, SIN3, and HDAC1. It has been shown that
multiple and distinct domains of NCoR are required for the interaction
with different members of the basal transcriptional apparatus
(152)
. Furthermore, the amino-terminal repression domain
of NCoR, which can bind to both TFIIB and TAFII32, can disrupt the
functional interaction between TAFII32 and TFIIB that is required for
transcriptional activation. Similarly, SMRT also harbors multiple
repression domains and can interact with both the SIN3 co-repressor and
the basal factor TFIIB. This TFIIB interaction correlates with the
repression ability of SMRT (153)
.
Members of the basal transcriptional machinery, including TFIIE and
TFIIF, can also be acetylated by histone acetyltransferases
(154)
. The action of HDACs on the acetylation status of
these proteins during transcriptional repression is unknown, although
it is conceivable that these proteins may also be targets of HDACs.
Thus, co-repressors act via multiple mechanisms with multiple
interaction partners, including the transcriptional apparatus. This
implies that a large complex that involves both histone deacetylation
and repression of the basal transcriptional machinery (Fig. 2)
is very
potent for full transcriptional repression.
Repression of transcriptional activators
Not only can co-repressors mediate the silencing effects of
DNA-bound silencers, but more and more examples are being presented
where co-repressors also associate with previously described
transcriptional activators. This association may result in the masking
of the activation domain and the displacement of coactivators. In these
cases, not only is the activation of the transcription factor reduced,
but active gene silencing also results. Thus, under certain conditions
a transcription factor can act as a repressor or an activator.
One such example is the modulation of E2F activity by the RB
tumor suppressor. In the absence of RB binding, E2F acts as a
transcriptional activator. Binding of RB to E2F renders E2F response
elements from an activating to a repressive sequence. RB has a dual
role in this regulation in that the activation domain of E2F is blocked
by RB binding and RB also actively represses transcription (155
, 156)
by direct recruitment of HDAC1 (31
32
33)
, c-Ski
(50)
and CtBP (82)
.
|
MECHANISMS OF SILENCER-CO-REPRESSOR FORMATION |
|---|
|
TOPABSTRACTINTRODUCTIONBIOLOGICAL ROLE OF CO-REPRESSORSCO-REPRESSORS AND DISEASESCO-REPRESSOR ACTIONMECHANISMS OF SILENCER-CO...OUTLOOKREFERENCES |
|---|
).
|
NHRs that bind ligands recruit co-repressors in a ligand-sensitive
manner (reviewed in ref 157
). In the absence of ligand,
these NHRs are bound to response elements in the regulatory regions of
target genes in association with co-repressors (NCoR, SMRT, and Alien)
and actively repress transcription. After ligand binding, the receptors
undergo conformational changes that lead to the dissociation of these
co-repressors (15
, 16
, 158)
.
Co-repressor–silencer complex formation can also depend on the
architecture of the specific DNA sequence, which is exemplified in case
of NHRs (39
, 159
, 160)
. When RAR is bound to a DR5
response element, a direct repeat with five base pair spacing,
co-repressors dissociate on ligand binding. On this element RAR is a
silencer and a hormone-dependent transcriptional activator. However, on
a DR1 element, a direct repeat with one base pair spacing,
co-repressors remain bound to the receptor even after binding to
ligand. Since coactivators are recruited and co-repressors are not
dissociated upon hormone binding, this leads to a transcriptionally
inert receptor (159)
. Also, on negative TREs, which are
named for their ability to repress genes in the presence of thyroid
hormone, liganded TR recruits HDAC2 in the presence of ligand
(39)
. Opposite to that, on classical elements ligand
binding to TR leads to dissociation of the co-repressor complex
(15
, 16
, 158)
.
Phosphorylation of the silencer proteins can lead to changes in the
recruitment of the co-repressors. In early G1
phase, hypophosphorylated RB represses E2F gene activation by
recruitment of HDAC activity (31
32
33
; reviewed in ref
137
). Activation of cdk4/6 cyclin-dependent kinases in
late G1 phase results in hyperphosphorylation of
RB and dissociation of the HDAC activity resulting in inactivation of
transcriptional repression mediated by RB bound to E2F
(161)
. NHR co-repressor interactions are also influenced
by kinases. Activation of the epidermal growth factor receptor signal
transduction pathway inhibits the ability of TR and v-ErbA to repress
transcription and to interact with SMRT (162)
.
Furthermore, this tyrosine kinase activation also inhibits the ability
of PLZF to repress transcription and to interact with SMRT, implying
that the involvement of tyrosine phosphorylation in regulation of SMRT
recruitment is more widespread than just by NHRs.
Proteolysis of co-repressors is also a mechanism that regulates the
duration of the co-repressor–silencer complex at the promoter region
of the targeted gene. The NH2 terminus of NCoR
can be targeted by mSiah2 for degradation (113)
. This
mechanism of co-repressor removal, therefore, operates like an
antirepressor mechanism. Since the expression of mSiah2 is cell
specific, this regulates the ability of NCoR specific receptors such as
RevErb to function as repressors in specific cell types. Similarly, the
Ski family of proteins are regulated by proteolysis. This family are
co-repressors for the SMAD proteins that are involved in tumor growth
factor ß (TGFß) -mediated cell signaling (reviewed in ref
163
). TGFß can induce the rapid degradation of the two
Ski family proteins c-Ski and SnoN probably by the proteosomal pathway,
thus allowing activation of SMAD responsive target genes
(49)
.
Competition for cellular components also regulates the ability of
silencers to associate with co-repressors. The proteins encoded by the
Ski proto-oncogene family have been shown to be part of a histone
deacetylase complex and bind to mSin3A and NCoR/SMRT
(102)
. Transcriptional silencing by RB and the Mad family
requires a complex containing c-Ski, which binds directly to RB and the
Mad proteins (50
, 102)
. The viral counterpart of c-Ski,
v-Ski, lacks the SIN3 binding region but can still bind to RB and Mad.
After v-Ski expression, RB and Mad proteins will bind v-Ski but fail to
recruit the SIN3 factor, leading to abrogation of the silencing
mediated by RB or Mad.
Cellular localization of the repressor also regulates the binding of
co-repressors. TGFß receptor activation induces the phosphorylation
of SMAD2. This results in the heterodimerization with SMAD4 and the
nuclear localization of the SMAD2/4 complex. Once in the nucleus, the
SMAD2/4 complex forms a complex with FAST2 on DNA and recruits the
homeodomain protein TGIF and histone deacetylase activity to repress
SMAD2/4-responsive genes (92)
.
Thus, nature has generated multiple mechanisms to regulate the protein complex formation of transcriptional silencers with co-repressors on DNA response elements.
|
OUTLOOK |
|---|
|
TOPABSTRACTINTRODUCTIONBIOLOGICAL ROLE OF CO-REPRESSORSCO-REPRESSORS AND DISEASESCO-REPRESSOR ACTIONMECHANISMS OF SILENCER-CO...OUTLOOKREFERENCES |
|---|
, 165)
. In general, co-repressors play
a major part in gene repression mechanisms. An important insight was
gained by revealing that co-repressors harbor or recruit HDAC activity
as one major player of gene repression. At this stage, the contribution
of either the HDAC activity or the interaction with the basal
transcription machinery is still unclear. Using the in vitro
transcription system with naked DNA or chromatin templates, the
detailed influence of each HDAC activity or the inhibition of basal
transcription factors on gene silencing will be able to be determined.
Also, silencing mediated by heterochromatic areas, telomeric regions,
and other aspects of higher order chromosome structure are presumably
based on the involvement of co-repressors. The fact that HDACs are
expressed in a tissue and stage-specific manner raises questions of the
specificity of HDACs in gene control and the role of each HDAC in
development. Future experiments including mice model systems will
reveal the biological role of each co-repressor. These findings are
also important for cancer therapy by using inhibitors of HDACs. It will
be important to develop drugs specific for a particular HDAC and/or
peptides to effectively inhibit a specific co-repressor–silencer
interaction.
|
ACKNOWLEDGMENTS |
|---|
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONBIOLOGICAL ROLE OF CO-REPRESSORSCO-REPRESSORS AND DISEASESCO-REPRESSOR ACTIONMECHANISMS OF SILENCER-CO...OUTLOOKREFERENCES |
|---|
