| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
INSERM U461 (A.B., M.S., J.B., M.P., M.P.-W.) Faculté de
Pharmacie Paris-Sud 92296 Châtenay-Malabry, France
INSERM U459 (P.L.) Faculté de Médecine Henri
Warembourg 59045 Lille, France
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
and GRß. GR is
mainly located in the cytoplasm of unstimulated cells as part of a
large multiprotein inactive complex with heat shock proteins and
immunophilins. Hormone binding causes dissociation of this complex and
translocation of the receptor into the nucleus. Then, GR stimulates
hormone-dependent transcription through binding to a 15-bp
glucocorticoid responsive element (GRE) present in the regulatory
regions of responsive genes (1, 2). The function of GRß as a dominant
negative form of GR is still a matter of debate (3, 4). The mouse
mammary tumor virus (MMTV) promoter is one of the most studied
GC-responsive promoters. Its transcriptional activation requires
binding of GR to a cluster of four glucocorticoid response elements
(GREs) located within the U3 region of the long terminal repeat (LTR)
(5, 6). The promoter also contains binding sites for ubiquitous
transcription factors such as nuclear factor-1 (NF-1), octamer
transcription factors-1 and -2 (Oct-1 and Oct-2), and for other unknown
tissue-specific regulatory factors controlling the expression of MMTV
(7, 8, 9, 10). When stably transfected into mammalian cells and before
stimulation, the MMTV-LTR is reproducibly packaged into a phased array
of nucleosomes preventing NF-1 access to its binding site. Hormone
binding initiates chromatin remodeling, and the promoter becomes
accessible to NF-1. On transiently transfected MMTV-LTR plasmid DNA,
NF-1 binding occurs constitutively and is not affected by GR loading
(11). To mediate their effects, steroid hormones and cytokines activate various signaling pathways through intracellular or membrane receptors, respectively. In immune cells, cross-talks between these signaling pathways affect fundamental cellular processes such as proliferation, differentiation, or apoptosis. Indeed, GCs suppress interleukin-4 (IL-4)-induced proliferation of the murine CTLL-2 cytotoxic T cell line, without affecting the IL-2-driven growth of these cells (12). In mouse T helper (Th) cell lines, the synthetic GC dexamethasone (DEX) completely inhibits IL-2-induced cell proliferation, reduces IL-4-mediated cell growth, and has no effect on IL-9-response (13). Rat CD4+ T cells transiently exposed in vitro to DEX display an altered pattern of cytokine production and develop a Th2 response (14). IL-2 has a protective role against GC-induced cell death on T cell hybridomas (15) and T lymphocytes (16, 17, 18). Moreover, IL-4 protects Th2 cells from DEX-induced apoptosis and IL-2 rescues Th1 cells from the cytolytic effect of GCs, indicating that mature T cells can be saved by their own growth factor (19). IL-9 is a potent inhibitor of GC-induced apoptosis in thymic lymphoma cell lines (20). However, mechanisms underlying interactions between GCs and cytokines signal transduction pathways in immune cells remain poorly understood.
GR can establish protein-protein interactions, independently from
DNA-binding, with other transcription factors such as activated
protein-1 (AP-1), NF-
B, STAT-3 (signal transducer and activator of
transcription 3), and STAT-5 (21). These transcription factors are
activated by cytokines, resulting in positive or negative regulation of
GC-induced transcription. The outcome of cytokine stimulation on
GC-mediated transcription is dependent on the promoter context and on
the cell type. Indeed, in CEM, S49, and Jurkat lymphoid T cell lines,
phorbol myristate acetate (PMA), through induction of AP-1, enhances
DEX-induced transactivation of the MMTV-LTR, whereas it displays an
inhibitory effect in NIH 3T3 fibroblasts (22). In immune cells, GCs
inhibit NF-B activation induced by tumor necrosis factor-
(TNF), and GR was shown to physically interact with NF-B, thereby
preventing its binding to DNA (23, 24). Furthermore, GCs were shown to
induce the expression of the inhibitory protein IB, trapping
NF-B as an inactive complex in the cytoplasm (23, 24).
Overexpression of STAT-5 in PRL-activated COS-7 cells results in an
increased activity of the ß-casein gene promoter upon treatment with
GCs, whereas MMTV-LTR promoter activity was decreased under similar
conditions (25).
The aim of this work was to evaluate the effect of IL-2 on GR transcriptional activity in lymphoid cells. We show that IL-2 strongly inhibits GC-induced transcription from the MMTV promoter, whereas IL-2 alone weakly stimulates its activity. These effects were observed in cells stably or transiently transfected with MMTV-LTR-luciferase (MMTV-LTR-luc) constructs. We also demonstrate that IL-2 does not impede binding of NF-1 or Oct-1 to their specific DNA binding sites. A STAT5 binding site was identified in the 5'-region of the promoter, which proved to be responsible for the positive regulation of the promoter by IL-2. However, deletions and mutations within the MMTV promoter showed that neither the STAT5-responsive element, nor other uncharacterized DNA sequences, play a role in IL-2 inhibition of MMTV-LTR-luc transactivation. AP-1 is not involved in the inhibitory mechanism of IL-2 as shown by PMA treatment. In CTLL-2 cells overexpressing STAT5B, IL-2 inhibition of GC-induced transcription is enhanced when compared with normal cells, whereas overexpression of a dominant negative form of STAT5B (Y699F) abolishes the IL-2- inhibitory effect. Coimmunoprecipitation experiments indicated a physical association between GR and STAT5, occurring in CTLL-2 cells treated with DEX, or DEX and IL-2. Taken together, these results strongly suggest a role for IL-2-induced STAT5 in the inhibition of GR transcriptional activity in lymphocytes. The mechanism of inhibition does not rely on a competition for limiting amounts of CBP (CREB-binding protein) or SRC-1a (steroid receptor coactivator-1a), since overexpression of the coactivators increases the IL-2 inhibitory effect.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
). IL-2
alone caused a modest enhancement of the luciferase activity in the
absence of DEX. In the presence of DEX, a 12-fold induction of the
luciferase activity was observed. Incubation of the cells with DEX and
IL-2 resulted in more than 70% inhibition of the transcriptional
activity of the promoter when compared with DEX-only conditions.
|
).
These results suggest that the IL-2 inhibitory effect on GC-stimulated
MMTV promoter activity is probably promoter context dependent and is
not uniquely mediated through GREs.
We assessed the effect of IL-2 on GR DNA-binding activity by EMSA
(electrophoretic mobility shift assay) using in vitro
translated recombinant GC receptor (Fig. 1C). At least two major
complexes with different mobilities were detected, despite the unique
GRE sequence of the DNA probe. This could result from a partial
proteolysis of the GR in the rabbit reticulocyte lysate. These
complexes bound specifically to the GRE DNA-probe, since their
formation was competed by a 50-fold molar excess of unlabeled GRE DNA
probe, but not by the nonspecific random DNA fragment. Furthermore,
addition of the specific anti-GR antibody to the binding reaction did
not block complex formation, but generated an antibody-protein-DNA
ternary complex resulting in a further reduction of the mobility of all
the protein-DNA complexes.
Competition experiments were performed with increasing amounts (1, 2.5,
and 5 µg) of nuclear extracts of CTLL-2 cells deprived of IL-2 for
3 h and treated for 1 h with IL-2 (1 ng/ml), or left
untreated. Consistent with the results presented in Fig. 1B, IL-2
treatment did not impede the binding of GR to DNA (Fig. 1C) since GR
DNA-binding activity was similar in the presence of nuclear extracts
from nontreated or IL-2-treated cells. Addition of nuclear extracts did
not generate lower-mobility complexes, which could result from the
interaction of DNA-bound GR with other cellular proteins, induced or
not by IL-2. These results suggest that IL-2 stimulation of T cells
does not unmask activities able to perturb GR binding to DNA.
IL-2 Inhibits DEX-Induced MMTV-LTR Activity of Stably Transfected
Templates
Modifications of chromatin structure are known to affect
GC-induced transcriptional activity of the MMTV promoter. IL-2 could
interfere with chromatin remodeling by GR, thus affecting the
activation of transcription. We therefore compared stably integrated
templates and transiently transfected templates, which have been shown
to display, or not, a chromatin architecture, respectively, for the
effect of IL-2 on GC-induced transcriptional activity of MMTV-LTR. To
this end, we stably transfected the wild-type MMTV-LTR-luc plasmid in
CTLL-2 cells. Several clones were isolated and tested for their
responsiveness to GC treatment. All clones tested responded similarly
to IL-2 and GC stimulation, albeit with a different amplitude, which is
likely to be related to the number of integrated copies. The clone
showing the highest inducibility of GC-dependent transactivation
(CTLL-2 pLTR 2E) was selected for further experiments. Results
presented in Fig. 2 show that treatment
of these cells with DEX for 12 h results in a 20-fold increase in
luciferase activity when compared with the basal transcription level.
Simultaneous addition of IL-2 and DEX results in an IL-2 dose-dependent
inhibition of DEX induction of the MMTV promoter activity. At a
saturating dose of 1 ng/ml of IL-2, the promoter activity is almost
completely abolished (91% inhibition) and reaches basal level measured
in the absence of hormone stimulation. As previously observed with
transiently transfected templates, IL-2 alone positively regulated
MMTV-LTR activity. These results obtained with the chromatin templates
are thus comparable to the transiently transfected ones, suggesting
that the IL-2 effect is not affected by integration of the promoter
into chromatin.
|
) therefore
raised the hypothesis of the existence of one (or more) regulatory
element (s) necessary for the inhibitory effect of IL-2. Internal
deletions of the MMTV LTR (-967 to -220, and -860 to -220) allowed
us to localize a STAT5 consensus binding site
(TTCGGAGAA) (Fig. 3).
|
). However, after 1 h of IL-2
stimulation (1 ng/ml), a strong STAT5-DNA binding was observed
(Fig. 4A). Supershift experiments showed that STAT5A and STAT5B were
activated in CTLL-2 cells upon IL-2 treatment. We then evaluated by
EMSA whether IL-2-activated STAT5 proteins could bind to the
STAT-responsive element of the MMTV promoter (-923/-914) (Fig. 4B). A
DNA-bound complex was detected in IL-2-treated cells and could be
completely supershifted with the anti-STAT5 serum (Fig. 4B), suggesting that STAT5 was the only
component of the complex formed on the STAT/MMTV probe in CTLL-2 cells.
To assess the functional activity of this STAT5 element, we mutated
this sequence to disrupt STAT5-DNA binding on the MMTV-LTR distal
regions of the pLTR-luc and the p-1,180/-860-luc plasmids to generate
pLTRmut-luc and p-1180/-860 mut-luc plasmids. CTLL-2 cells were then
transiently transfected with wild-type or mutated constructs and
treated for 12 h with or without IL-2 (1 ng/ml). Results presented
in Fig. 4C show that mutation of the STAT5 site dramatically reduces
the basal activity of the promoter measured in untreated cells. In
addition, the activity of the mutated constructs was not significantly
affected by IL-2, strongly suggesting that this site plays a role in
the positive regulatory effect of the cytokine on the MMTV promoter
activity. We then investigated whether this STAT5-binding site could
play a role in the IL-2-inhibitory effect on DEX-induced MMTV-LTR
transactivation. To this end, CTLL-2 cells were transiently transfected
with the pLTR-luc and the pLTRmut-luc plasmids and treated for
12 h with IL-2 (1 ng/ml), with or without DEX (100
nM), or left untreated. Results show that the inducibility
of the mutated promoter is enhanced upon DEX treatment compared with
the wild-type promoter, due to the reduced level of the basal activity.
Nevertheless, the effect of IL-2 measured with pLTR-luc and pLTRmut-luc
(84% and 75%, respectively) is comparable with the two constructs
(Fig. 4D), suggesting that the STAT5-binding site does not contribute
to the IL-2-inhibitory effect.
|
)
to identify cis-acting sequences along the MMTV promoter
involved in the IL-2-mediated inhibition. CTLL-2 cells were transiently
transfected with these constructs and cultured for 12 h in the
presence or in the absence of 100 nM DEX and/or
IL-2 (1 ng/ml). Results show that the p104-luc plasmid, which lacks the
two distal GREs, does not retain any hormone inducibility. Sequences
located upstream of the -200 position can be deleted without impairing
the hormone response (Fig. 5B). The responsiveness of p325-luc,
p223-luc, and p200-luc plasmids to DEX stimulation is inhibited by
56%, 65%, and 62% upon IL-2 treatment (1 ng/ml), respectively. The
p223-luc and p200-luc deletion mutants show an increased GC-induced
transcriptional activity compared with the MMTV-LTR-luc plasmid,
suggesting that upstream negative regulatory sequence(s) were removed
(Fig. 5B).
|
) by
substituting two consensus synthetic GREs for the sequence spanning
from the 5' end to the -104 position within the MMTV-LTR promoter.
Results from Fig. 5C show that the basal luciferase activity of this
plasmid is very low and remains unaffected by IL-2. Very importantly,
addition of IL-2 reduced by 52% the DEX-induced transcriptional
activation of this promoter. These data demonstrate that the most
proximal region of the promoter, including the two proximal
hemi-palindromic GREs, the NF-1 and Oct-1 binding sites, as well as the
TATA box, are necessary and sufficient to convey the inhibitory effect
of IL-2.
IL-2 Does Not Impede NF-1 or Oct-1 Binding to Their Cognate DNA
Response Elements
Given that NF-1 binding to MMTV-LTR is an absolute prerequisite
for GC-induced MMTV transactivation, we evaluated whether IL-2 inhibits
NF-1 binding to its specific recognition site. We conducted EMSA using
a NF-1 probe whose sequence is similar to the NF-1 binding site present
in the MMTV promoter. Results from Fig. 6A show that IL-2 does not reduce NF-1
binding activity.
|
). To characterize the
nature of the Oct family member bound to the probe in CTLL-2 cells, we
performed supershift experiments with a specific anti-Oct-1 polyclonal
antibody directed against the carboxy terminus of the protein. The
anti-Oct-1 IgG, but not the control IgG, completely displaced the
specific complex for all tested treatments, indicating that Oct-1 is an
integral part of the complex formed on the Oct probe in CTLL-2
cells.
AP-1 Is Not Involved in the IL-2 Inhibition of GC-Induced
MMTV-LTR Transactivation
These results prompted us to determine which IL-2-dependent signal
transduction pathway was involved in the inhibition of GC-induced
transcription. The AP-1 transcription factor, which plays a crucial
role in cell cycle control and survival of lymphoid cells, has been
shown to physically interact with the GR, resulting in a mutual
transcriptional repression (29, 30, 31). IL-2 and PMA have been shown to
increase AP-1 DNA binding and transcriptional activities in CTLL-2
cells (Ref. 32 and Fig. 7A). Results
showed that PMA did not repress the GR transcriptional activity
assessed with the MMTV-LTR-luc plasmid (Fig. 7B). These results were
obtained in stably transfected CTLL-2 pLTR cells (Fig. 7B), as well as
in transiently transfected CTLL-2 cells (data not shown).
|
Overexpression of wild-type Myc-STAT5B protein was achieved after
48 h of culture in the absence of tetracycline. In the presence of
tetracycline (1 µg/ml), the expression of this protein was fully
repressed (Fig. 8A, right
insert). Cells were then transiently transfected with the
p2GRE104-luc construct and stimulated for 12 h in the presence or
in the absence of 100 nM DEX and/or IL-2 (1 ng/ml
and 10 ng/ml). In STAT5B overexpressing cells, IL-2 (1 ng/ml)
inhibition of GC-induced p2GRE104-luc transactivation was enhanced
by almost 20% (54% inhibition in overexpressing STAT5B cells compared
with 36% inhibition in wild-type cells). This difference was not
detectable when a 10-fold higher concentration of IL-2 (10 ng/ml) was
used, leading to a maximal inhibition of DEX-induced p2GRE104-luc
transactivation (Fig. 8A).
|
), since no clone displaying high regulatable
levels of Myc-STAT5B Y699F could be obtained. Analysis of STAT5
activation in these clones, by DNA affinity precipitation with a
5'-biotinylated IFN
-activated sequence (GAS) oligonucleotide,
revealed that a dominant negative effect of STAT5B Y699F was only
achieved with clone 5, which showed the highest level of expression of
this protein (Fig. 8C).
CTLL-2 Myc-STAT5B Y699F cells (clones 4, 5, and 9) were cultured
for 48 h in the absence of tetracycline. Cells were then
transiently transfected with the p2GRE104-luc construct and stimulated
for 12 h in the presence or in the absence of 100 nM
DEX and/or IL-2 (500 pg/ml). In clones 4 and 9, DEX-induced
p2GRE104-luc transactivation was still inhibited (-55% and -51%,
respectively) despite expression of the dominant negative form of
STAT5B (Fig. 8D). These results were correlated with STAT5 activation
levels (Fig. 8C). IL-2 inhibition of DEX-induced p2GRE104-luc
transactivation, however, was almost completely abolished (-6%) in
clone 5, which expresses the highest level of the Myc-STAT5B Y699F
protein (Fig. 8D). These results showed that activation of STAT5 is a
key event in the inhibition of GC-induced MMTV transactivation by
IL-2.
We postulated that STAT5 might physically interact with the GR,
explaining the transcriptional interference existing between these two
factors. Coimmunoprecipitation experiments were performed with an
anti-GR antibody using whole-cell extracts from CTLL-2 cells deprived
of IL-2 for 3 h and treated for 1 h with IL-2 (1 ng/ml), DEX
(100 nM), DEX and IL-2 (1 ng/ml), or left untreated.
Blotting of the GR immunocomplex with an anti-STAT5 antibody allowed us
to detect coprecipitated STAT5 in DEX- or DEX and IL-2-treated cells
(Fig. 8E). These results suggested that GR activation, but not STAT5
activation, is a prerequisite for GR-STAT5 complex formation. To
confirm this observation, we assessed whether STAT5B Y699F could
interact with the GR. For this purpose, CTLL-2 STAT5B Y699F cells
(clone 5) were cultured for 48 h in the absence of tetracycline,
deprived of IL-2 for 3 h, and then treated for 1 h with DEX
(100 nM) and IL-2 (1 ng/ml). Cell lysates were then
immunoprecipitated with an anti-GR antibody, and GR immunocomplexes
were separated by SDS-PAGE and probed with an anti-myc antibody.
Results presented in Fig. 8F show that STAT5B Y699F physically
interacts with GR, confirming that STAT5B tyrosine phosphorylation is
not necessary for GR-STAT5 association.
Role of Coactivators in the IL-2 Inhibition of GC-Induced
MMTV-LTR Transactivation
Since CBP was described to interact with both the GR and
STAT5, we tested the hypothesis of a competition between GR and STAT5
for limiting amounts of the coactivator. CTLL-2 cells were transiently
transfected with p2GRE104-luc and pCMV-2N3T-CBP or pCMV-2N3T plasmids,
cultured for 48 h in complete medium containing 1 ng/ml IL-2, and
then treated for 12 h with DEX (100 nM), IL-2 (1
ng/ml), DEX and IL-2, or left untreated (Fig. 9A). The transfection of increasing
quantities of CBP (1, 5, and 10 µg plasmid) resulted in an
enhancement of the IL-2 inhibitory effect on GC-induced p2GRE104-luc
transactivation (62%, 69%, and 72%, respectively, compared with 42%
with the control vector), ruling out the possibility of a squelching of
CBP by STAT5 as a mechanism of IL-2 inhibition (Fig. 9A). Moreover, we
could determine that transfection of 1 µg pCMV-2N3T-CBP leads to an
enhancement of DEX-induced p2GRE104-luc transactivation, showing that
CBP participates in GC-induced transactivation. However, transfection
of 10 µg of plasmid has a negative effect on the promoter activity.
Taken together, these results suggest that CBP is a modulator of MMTV
promoter activity. We then evaluated the role of another GR
coactivator, SRC-1a, in the IL-2 inhibitory effect (Fig. 9B).
Overexpression of SRC-1a causes a slight enhancement of the IL-2
inhibition of DEX-induced p2GRE104-luc transactivation (47%
vs. 35% for 5 µg of plasmid), indicating that the
mechanism of inhibition does not rely on a competition for limiting
amounts of SRC-1a (Fig. 9B). Again, transfection of high amounts of
SRC-1a leads to a down-regulation of the MMTV promoter activity.
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
DEX-induced MMTV-LTR transcriptional activity was significantly reduced upon IL-2 addition at saturating concentrations. However, GC-dependent transactivation measured with the GRE5-EBV-TATA-CAT plasmid (38) was not impaired by IL-2, suggesting that this effect depends on the promoter context, and that IL-2 did not prevent binding of DEX to the GR or translocation of the GR to the nucleus. Moreover, GR DNA binding activity was not affected in the presence of nuclear extracts of IL-2-treated CTLL-2 cells, showing that IL-2 did not induce the appearance of inhibitory factors.
In the MMTV-LTR promoter, binding of the activated GR to GREs initiates a remodeling of the chromatin that displaces nucleosome B and makes the promoter accessible to NF-1, a step necessary for GC-dependent transcription to occur. This mechanism is not observed on chromatin-free templates (39, 11). Binding of the octamer motifs in vivo was also observed to be strictly hormone dependent, and Oct/GR interactions result in a transcriptional cooperativity between these two factors (40). We observed that IL-2 inhibition was still effective when MMTV-LTR was stably integrated into chromatin, and that IL-2 did not alter NF-1 or Oct-1 binding capacity to their cognate DNA sequences, indicating that inhibition does not occur through modulation of either NF-1 or Oct-1 DNA binding activities.
Another mechanism that could account for IL-2 inhibition is the
activation of a trans-regulatory factor acting by binding to
the MMTV-LTR. Several sequences located within the U3 region of the
MMTV-LTR have been shown to regulate its activity (41). The MMTV
promoter contains four HREs (hormone responsive elements): two distal
palindromic sites (located between -184 and -114 positions) and two
proximal hemipalindromes (located between positions -98 and -78) (see
Fig. 3). Deletion of the two distal GREs/HREs is sufficient to abolish
GC-induced transactivation (see Fig. 5B) (42). A region around –1,090
to -900 within the MMTV-LTR has been delimited as an enhancer that
seems to be mainly involved in mammary specificity (41, 43). Deletion
of a regulatory element located between -294 to -200 termed AA
element has also been shown to decrease GC-induced transactivation
(42). Moreover, within this AA element a 20-bp region located between
-223 and -201 seems to play a regulatory role (44). At least three
negative regulatory elements have been described between -861 and
-364 (41). An AP-1 site has also been described between position -766
and -737 (45). We identified in this work a consensus STAT5-binding
site between positions -923 and -914 of the MMTV-LTR and showed the
involvement of this sequence in the IL-2 positive regulation of the
promoter activity, but not in the IL-2 inhibitory effect on DEX-induced
MMTV transactivation. This result rules out a role for a STAT5
DNA-binding element in the negative regulation of MMTV-LTR by IL-2.
Deletion of the entire 5'-end of the LTR up to the indicated position
allowed us to rule out the involvement of regions upstream of position
-325 (enhancer region, negative regulatory elements, AP-1 site),
between -325 and -223 and of the 20-bp sequence within the AA
element, respectively, in the IL-2-inhibitory effect. However, this
approach did not evaluate whether sequences located between GREs could
play a role. Indeed, a sequence located between the two distal GREs
(-163 to -147) and termed A element has been described to regulate
negatively MMTV-LTR activity upon binding of a
trans-negative modulator named C1 in 6.10.2 rat hepatoma
cells (46). However, data generated with the p2GRE-104 plasmid did not
confirm this hypothesis in CTLL-2 lymphocytes. Taken together, our
results obtained with these deletion mutants suggested that the
-104/+1 region is critical for the IL-2 negative response and that a
trans-acting IL-2-activated factor is not involved in the
regulation of MMTV-LTR activity by IL-2.
We then decided to identify which component of the IL-2 signal transduction pathway was involved in the IL-2-inhibitory effect. This approach could also clarify whether an IL-2-induced factor could interact with GR-transcriptional activity without binding on the DNA.
IL-2 leads to AP-1 activation in CTLL-2 cells (Ref. 32 and Fig. 7A).
AP-1 and GR have been shown to mutually interfere with their
transactivating functions (29, 30, 31). Elevated c-Jun or c-Fos levels
can inhibit GR-dependent transcription from the MMTV-LTR promoter or
from promoters carrying only GREs (29, 30, 31). However, when CTLL-2 cells
were treated with PMA and DEX, no inhibition of MMTV-LTR
transactivation was found, although AP-1 activity was clearly induced
as assayed by gene reporter assays. We note that the composition of the
AP-1 complex could differ between IL-2 or PMA treatment (32),
explaining this result. Indeed, cell-specific factors and AP-1
composition may affect the outcome of the effect of AP-1 on
GR-transcriptional activity (22, 47, 48).
The Jak-STAT pathway has been recently described to interact with
GC-dependent signaling. In COS-7 cells stimulated with PRL, STAT5
appears to synergize with GC on the ß-casein promoter (25, 49, 50)
but to antagonize GC-induced MMTV-LTR promoter activity (25, 51).
Moreover, in the rat hepatoma cell line H4IIE and in COS-7 cells
stimulated with IL-6, STAT3 was shown to synergize with GC on the rat
-fibrinogen promoter and on the MMTV-LTR promoter (52). Noticeably,
many of these studies have been performed with overexpressed factors
(STAT or GR) in cells that do not normally express either STAT or GR
proteins. Our experiments showed that, in CTLL-2 cells expressing
endogenous levels of STAT5 and GR, STAT5B is necessary for IL-2
inhibition of MMTV promoter activity. Moreover, tyrosine
phosphorylation of STAT5B seems to be an important step in this
mechanism. A physical association between GR and STAT5 was present only
if the GR was activated but independently of STAT5 activation.
Formation of complexes between STAT5 and GR has previously been
described in both COS-7 cells and in HC11 mammary cells, and this
association is dependent upon ligand-induced activation of STAT5 in COS
cells but not in HC11 cells (25, 53).
In light of these observations, STAT5 appears to play a major role in
the IL-2 inhibitory effect, without binding on a specific DNA sequence
located on the MMTV promoter. Our results with the deletion mutants of
the promoter argue that integrity of sequences from -104 to +1 is
required for IL-2 inhibition. However, the MMTV-LTR proximal promoter
does not contain STAT5-specific DNA-binding sequences, suggesting that
STAT5-mediated inhibition could occur through interference with
specific factors loading on the MMTV promoter rather than through
binding to a specific DNA sequence, probably by protein-protein
interaction. Indeed, we detected a physical association between GR and
STAT5 in CTLL-2 cells. GR/STAT5 complex formation could interfere with
contacts between the GR and coactivators of the basal transcription
machinery. It has been shown in COS cells that overexpression of
CBP/p300 did not alter STAT5 inhibition of MMTV-LTR activity, despite
interaction of p300 with STAT5 and p300-dependent enhancement of
MMTV-LTR activity (51). In CTLL-2 cells, we have also found that
overexpression of CBP or SRC-1a does not alter IL-2 inhibitory effect.
Although these results rule out a mechanism of squelching, where GR and
STAT5 could compete for a limiting amount of this coactivator, they do
not exclude the possibility that another mechanism involving CBP/p300
could take place. The tyrosine phosphorylation of STAT5 appears as an
essential step for IL-2 inhibition, either by altering the conformation
of the GR-STAT5 complex, or by allowing the recruitment of other
partners. GR and STAT5 could both interact with a cofactor like
CBP/p300, and/or SRC-1a, and form a complex devoid of transcriptional
activity, as recently hypothesized for the mutual antagonism between GR
and NF-B (54).
All together, these results indicate that in lymphocytes expressing endogenous levels of STAT5 and GR, STAT5 plays a critical role in IL-2 regulation of GC-dependent transactivation. This could be of importance in elucidating how cytokines modulate expression of GC-regulated genes in pathological situations, i.e. asthma or lymphoma, or in physiological situations, i.e. apoptosis (55).
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
Plasmids
pLTR-luc plasmid contains the entire MMTV-LTR
GC-responsive promoter from the C3H strain coupled to the
luciferase reporter gene (43). p325-luc, p223-luc, p200-luc, and
p104-luc plasmids display sequential 5'-end deletions up to positions
-325, -223, -200, and -104, respectively (Fig. 5A).
p-1,180/-860-luc plasmid displays an internal deletion between
positions -860 and -220 (43). p200-luc was obtained by deletion of
the HindIII/AflII fragment from the pLTR-luc
plasmid, filling with Klenow, and ligation. p2GRE-104-luc plasmid was
constructed by inserting two synthetic consensus GREs immediately
upstream of the -104 promoter. Briefly, the
HindIII/SacI fragment of the pLTR-luc plasmid was
ligated to the HindIII/SacI double-stranded
5'-phosphorylated oligonucleotide
(5'-AGCTTTGTACAGGATGTTCTAGATCTTGTACAGGATGTTCTGAGCT-3'). These
constructs were then verified by sequencing. The
GRE5-EBV-TATA-CAT plasmid was a kind gift of S.
Mader. The 5XTRE-tk-CAT contains five TRE sequences upstream from the
CAT (chloramphenicol acetyltransferase) gene and was a kind gift of B.
Binetruy.
The NdeI/HindIII fragment of STAT 5B was amplified by PCR by using RSV-STAT5B (a kind gift of F. Gouilleux) as template and oligonucleotides 5BSTART (5'-GGGAATTCCATATGGCTATGTGGATACAGGCTCAG-3') and 5BSTOP (5'-CCCAAGCTTGAATTCTCATGACTGTGCGTGAGGGAT-3') as primers. The NdeI/HindIII fragment of STAT5B was inserted at the NdeI/HindIII sites of pGEM-Myc. The EcoRI/EcoRI fragment of Myc-STAT5B was subsequently inserted into the PUHD 10–3 plasmid (33) at the EcoRI site in the sense orientation, generating the construct pTRE-Myc-STAT5B.
A point mutation in the STAT5B sequence from tyrosine to phenylalanine (Y699F) was introduced using the Quick-change site-directed mutagenesis kit (Stratagene, Amsterdam, The Netherlands). The pTRE-Myc-STAT5B construct was used as a template and the complementary oligonucleotides STAT5B Y699F sense (5'-TGACGGATTCGTGAAGCCACAGAT-3') and STAT5B Y699F antisense (5'-ATCTGTGGCTTCACGAATCCGTCA-3') as primers. The PCR was run under conditions recommended by the manufacturer for 12 cycles (30 sec at 95 C, 1 min at 55 C, and 13 min at 68 C). The mutation in the STAT/MMTV sequence (TTCGGAGAA 224 GGCGGAGAA) was introduced using the same kit, with the pLTR-luc and p-1,180/-860-luc plasmids as templates and complementary oligonucleotides STAT/MMTV sense (5'-ACAATCTAAACAAGGCGGAGAACTCGACCTTCCTCCTG-3') and STAT/MMTV antisense (5'-CAGGAGGAAGGTCGAGTTCTCCGCCTTGTTTAGATTGT-3') as primers. The PCR was run for 16 cycles (30 sec at 95 C, 1 min at 55 C, and 15 min at 68 C).
The pCMV-2N3T and pCMV-2N3T-CBP plasmids were a kind gift of A. Harel-Bellan, the pCR3.1 and pCR3.1-SRC-1a plasmids were a kind gift of S. Tsai.
Cell Culture, Transfections, and Tet-Off Gene Expression
System
The murine IL-2-dependent cytotoxic T cell line CTLL-2 was
cultured in complete medium: RPMI 1640 medium (Life Technologies, Inc.) containing 2 mM L-glutamine, 0.1
mg/ml streptomycin, 100 U/ml penicillin (Life Technologies, Inc.), 50 µM 2-mercaptoethanol
(Sigma), 1% sodium pyruvate (Life Technologies, Inc.), 10% FCS (Life Technologies, Inc.), and 1
ng/ml of human recombinant IL-2.
Transfections were performed using the electroporation method. Exponentially growing CTLL-2 cells (107) were washed in RPMI 1640 buffer and resuspended in 150 µl of RPMI 1640 containing 10 µg of plasmid. After 10 min incubation on ice, cells were electroporated using a Bio-Rad Laboratories, Inc. gene pulser (Ivry-sur-Seine, France) set at 250 V and 960 µF. Cells were then maintained on ice for 10 min and resuspended in complete medium.
For transient transfection assays with MMTV constructs, cells were cultured with 1 ng/ml IL-2 for 48 h before transfection. After electroporation, cells were stimulated with DEX (100 nM), IL-2 (1 ng/ml), IL-2 plus DEX, or left untreated. After 12 h of incubation, proteins were extracted and assayed for luciferase activity.
Selection of stably transfected cells was initiated 48 h after electroporation using 800 µg/ml G418 (Life Technologies, Inc.) for the cotransfections with the pMC1 plasmid (conferring resistance to neomycin) or 800 µg/ml hygromycin B (Life Technologies, Inc.) for the cotransfections with the pTK-hygromycin plasmid encoding an hygromycin resistance gene (CLONTECH Laboratories, Inc. Palo Alto, CA). Stably transfected CTLL-2 pLTR cells were selected by a 2-week treatment with G418 and cloned by limiting dilution. Clones were then screened for their GC-stimulated luciferase activity.
Development of a stable tTA (tetracycline transactivator) cell line was initiated by cotransfection of CTLL-2 cells with the plasmid pUHD 15–1 (33) and with pMC1, encoding a neomycin resistance gene. Stably transfected cells were selected in the presence of 800 µg/ml G418 for 2 weeks, after which CTLL-2 Tet-Off cells stably expressing the tTA were transfected with the plasmids pTRE-Myc-STAT5B or pTRE-Myc-STAT5B Y699F, and with pTK-hygromycin (CLONTECH Laboratories, Inc.). Cells were cultured in the presence of 1 µg/ml tetracycline and 800 µg/ml hygromycin. Expression of the fusion proteins Myc-STAT5B or Myc-STAT5B Y699F was turned on by removal of tetracycline for 48 h, and analyzed by Western blotting using the 9E10 anti-Myc antibody.
Reporter Gene Activity Assays
Luciferase Assay.
Luciferase levels were measured according to the manufacturer’s
protocol (Promega Corp.). Briefly, extracts were prepared
by three cycles of freezing and thawing of cells resuspended in a lysis
buffer containing 25 mM Tris-phosphate, pH 7.8, 2
mM CDTA, 2 mM dithiothreitol (DTT), and
10% glycerol. Protein extracts in equivalent protein concentration
samples were mixed with 100 µl of luciferase assay reagent
(Promega Corp.). Luciferase activity was determined at 25
C after 1 min with a luminometer (LKB Wallac, Inc. Turku, Finland). Results are expressed in relative
luciferase units (RLU) relative to the basal level to which the value 1
was arbitrarily affected. Fold induction was calculated as the ratio of
arbitrary luciferase units in cells treated with IL-2, DEX, or DEX and
IL-2 compared with untreated cells. Fold induction (DEX-treated
cells) = (DEX activity - basal activity)/basal activity.
Fold induction (DEX + IL-2-treated cells) = (DEX + IL-2 activity - IL-2 activity)/basal activity.
CAT Assay.
Extracts were prepared by three cycles of freezing and thawing of cells
resuspended in hypotonic buffer (0.25 M Tris HCl, pH 8).
Protein extracts (40 µg) were incubated with
(14C)-chloramphenicol (60 mCi/mmol,
Amersham Pharmacia Biotech, Orsay, France) in the presence
of 2 mM acetyl coenzyme A (Sigma) for 1 h
at 37 C. Acetylated chloramphenicol was extracted in ethyl acetate and
separated from unmodified chloramphenicol by TLC. Conversion of
chloramphenicol was quantified using a Storm 840 phosphorimager and
the Imagequant software (Molecular Dynamics, Inc.,
Sunnyvale, CA). Percentage of chloramphenicol conversion represents the
ratio between acetylated chloramphenicol and total chloramphenicol
(acetylated and nonacetylated).
Preparation of Nuclear Extracts
Nuclear extracts were prepared by a modification of the method
described by Dignam et al. (56). Cells were deprived of IL-2
for 3 h and stimulated at 37 C for 1 h with DEX, IL-2, or DEX
plus IL-2. Cells were then pelleted, washed with ice-cold 1x PBS, and
maintained for 10 min on ice in a hypotonic buffer containing 10
mM HEPES, pH 7.8, 15 mM
KCl, 2 mM MgCl2, 1
mM EDTA, 1 mM
phenylmethylsulfonyl fluoride, 1 µg/ml aprotinin, 1 µg/ml
leupeptin, 1 µg/ml pepstatin, 1 mM DTT.
Cytoplasmic membranes were lysed by 50 strokes using a Kontes all-glass
Dounce homogenizer (B type pestle). The lysate was centrifuged at
1,000 x g for 5 min at 4 C, and the nuclear pellet was
resuspended in a high-salt buffer (20 mM HEPES,
pH 7.8, 1.5 mM MgCl2, 0.2
mM EDTA, 25% (vol/vol) glycerol, 1
mM phenylmethylsulfonyl fluoride (PMSF), 1
µg/ml aprotinin, 1 µg/ml leupeptin, 1 µg/ml pepstatin, 1
mM DTT, 400 mM NaCl).
Nuclear extracts were centrifuged at 15,000 x g for 20
min at 4 C.
EMSA
Oligonucleotides were purchased from Oligo Express (Paris,
France). Complementary sequences were annealed at 80 C for 10 min and
65 C for 10 min and then were end-labeled using
(32P)-ATP with T4 polynucleotide kinase
(Life Technologies, Inc.) and used for EMSA after ethanol
precipitation.
The 5'-TCTTTTGGAATTTATCCAAATCTTAT-3' probe was used for
NF-1 binding. The
5'-ATCTTATGTAAATGCTTATGTAAACCAAGA- 3' probe was
used for Oct binding. These probes correspond, respectively, to the
NF-1 and Oct binding sites found in the MMTV promoter between positions
-80 and -55 for NF-1, -61 and -32 for Oct-1. The GAS probe from the
GAS site of the FcR promoter
(5'-GTATTTCCCAGAAAAGGAAC-3') was used for STAT5 binding,
and the STAT/MMTV probe
(5'-ATCTAAACAATTCGGAGAACTCGACCTTC-3') corresponds to the
STAT response element located between positions -923 and -914 of the
MMTV promoter. Specificity was determined by using a 25-fold molar
excess of cold probe or random probe
(5'-CCTCCATGACTCCAGAACTAACCTCCATGAC-3').
End-labeled oligonucleotides were incubated at 25 C for 30 min with 15 µg of nuclear proteins in the presence of 1 µg of sonicated salmon sperm DNA in 20 µl of binding buffer (12% glycerol, 12 mM HEPES, pH 7.8, 60 mM KCl, 1 mM EDTA, and 1 mM DTT). Protein-DNA complexes were separated from free probe on a 5% polyacrylamide gel in 0.5x TBE running buffer at 200 V. For supershift experiments, 3 µg of the control IgG or of the specific anti-Oct-1 IgG (C-21, Santa Cruz Biotechnology, Inc., Santa Cruz, CA), or 1.5 µl of the anti-STAT5A and anti-STAT5B sera were preincubated with nuclear extracts for 2 h at 4 C before addition of the probe.
Production and DNA Binding Activity of Recombinant GR
The recombinant rat GR was produced using the TNT Quick coupled
transcription/translation system as recommended by the manufacturer
(Promega Corp.). Briefly, 1 µg of pET30rGR (containing
the rat GR gene under the control of the T7 promoter) was added to an
aliquot of TNT Quick master mix and incubated in a 50 µl reaction
volume for 90 min at 30 C. The synthesized GR was then activated for
1 h at 4 C and 1 additional hour at 30 C in a buffer containing 10
mM HEPES, pH 7.4, 20 mM ß-mercaptoethanol,
5% glycerol, 50 mM NaCl, and 1 µM DEX. The
5'-ATCTCTGCAGAACAGGATGTTCTAGCTACTT-3' probe was
used for GR DNA binding. Specificity was determined by using a 50-fold
molar excess of cold probe or random probe. End-labeled GRE
oligonucleotides were incubated at 25 C for 1 h with 2 µl of
activated lysate in the presence of 1 µg of sonicated salmon sperm
DNA in 20 µl of binding buffer (12% glycerol, 12 mM
HEPES, pH 7.8, 60 mM KCl, 1 mM EDTA, and 1
mM DTT). For competition experiments, 1 µg, 2.5 µg, or
5 µg of nuclear extracts from CTLL-2 cells deprived of IL-2 for
3 h and treated or not with IL-2 for 1 h were incubated with
the activated lysate. For supershift experiments, 1 µg of the control
IgG or of the specific anti-GR (M-20, Santa Cruz Biotechnology, Inc., Santa Cruz, CA) were preincubated with GR for 30 min at 4
C before addition of the probe.
DNA Affinity Precipitation of STAT Proteins
Cells were deprived of IL-2 for 2 h and treated with IL-2
(500 pg/ml) for 12 h at 37 C. Cells were then collected by
centrifugation, washed in 1x PBS, and resuspended in NP40 buffer (50
mM Tris HCl, pH 8, 0.5% NP40, 150 mM NaCl, 0.1
mM EDTA, 10 mM NaF, 1 mM PMSF, 1
µg/ml aprotinin, 1 µg/ml leupeptin, 1 mM DTT). The
double-stranded 5'-biotinylated oligonucleotide GAS was coupled to
streptavidin-agarose beads (Sigma) for 1 h at 4 C.
Whole-cell extracts were then incubated with the precoated beads for
1 h at 4 C. The beads were then washed three times with the NP40
lysis buffer and boiled in reducing sample buffer to elute the
complexes. Bound proteins were then separated on 8% polyacrylamide gel
and electroblotted onto Amersham Pharmacia Biotech PVDF
(polyvinylidene difluoride) membranes. Western blot
analysis was performed with the specific anti-STAT5B antibody
(ref 06–969, Upstate Biotechnology, Inc. Lake Placid,
NY).
Coimmunoprecipitation Assays
CTLL-2 cells were deprived of IL-2 for 3 h and then
stimulated for 1 h with IL-2 (1 ng/ml) and/or DEX (100
nM), or left untreated. Cell lysates were first incubated
for 30 min at 4 C with protein A sepharose beads (Sigma)
and preimmune serum and then centrifuged at 4,000 rpm for 1 min. The
supernatant (precleared lysate) was incubated overnight at 4 C with the
anti-GR antibody (BuGR2, Affinity BioReagents, Inc., Golden, CO) precoupled to Protein A sepharose beads. Immune
complexes were washed three times with lysis buffer and analyzed by
SDS-PAGE using anti-STAT5 antiserum.
Western Blot
Cells were collected by centrifugation, washed in 1xPBS, and
resuspended in NP40 buffer (50 mM Tris HCl, pH 8, 0.5%
NP40, 150 mM NaCl, 0.1 mM EDTA, 10
mM NaF, 1 mM PMSF, 1 µg/ml aprotinin, 1
µg/ml leupeptin, 1 mM DTT). Cell lysates were resolved by
SDS-PAGE on 8% polyacrylamide gels and electroblotted onto
Amersham Pharmacia Biotech PVDF membranes. After
saturation of nonspecific binding sites with dry low-fat milk in
TBS-Tween 20 (0.2%) for 2 h, membranes were probed with the
monoclonal 9E10 anti-Myc monoclonal antibody and developed with ECL
(Amersham Pharmacia Biotech).
|
ACKNOWLEDGMENTS |
|---|
|
FOOTNOTES |
|---|
This research was supported by INSERM and by a fellowship from the Association pour la Recherche sur le Cancer to Armelle Biola.
1 Present address: INSERM, 101 rue de Tolbiac, 75654 Paris,
France. 
Received for publication June 9, 2000. Revision received March 7, 2001. Accepted for publication March 12, 2001.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
B
activity through induction of IB synthesis. Science 270:286–290B in mediation of
immunosuppression by glucocorticoids. Science 270:283–286
