| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
*
Program in Immunology, Department of Pathology, Tufts University School of Medicine, Boston, MA 02111;
Howard Hughes Medical Institute and Department of Genetics, University of Pennsylvania School of Medicine, Philadelphia, PA 19704;
Department of Pathology, School of Medicine, University of New Mexico, Albuquerque, NM 87131; and
§
Trudeau Institute, Saranac Lake, NY 12983
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
, IL-2, and IL-4 upon activation. Repeated
stimulation of Th precursor cells in the presence of IL-4 leads to
terminally differentiated Th2 cells that have lost the ability to
transcribe the IL-2 gene. We provide evidence that repression of IL-2
gene expression in Th2 cells and partial repression in Th1 cells are
mediated by ZEB, a zinc finger, E box-binding transcription factor.
This factor binds to a negative regulatory element, NRE-A, in the IL-2
promoter, thereby acting as a potent repressor of IL-2 transcription. |
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
, IL-12, and other cytokines when stimulated; in contrast,
the cytokines produced by Th2 cells include IL-4, IL-5, and IL-10
(2, 3, 4). The induction of cytokine-mediated Th1 immune responses,
characterized by macrophage and NK cell activation and IgG2 production,
or Th2 immune responses, characterized by eosinophil activation and IgE
production, can determine disease outcomes in humans (5) and mice (3, 4, 6). Thus, it is important to understand the mechanisms involved in
the induction and maintenance of cytokine gene expression in Th cell
subsets. Although the mechanisms that silence IL-2 production in Th2 cells during the differentiation process are unknown, it is clear that the primary control of IL-2 expression is at the transcriptional level (7). In Th1 cells, IL-2 transcription is induced as a result of the cognate interaction of the TCR with antigenic peptides presented by MHC molecules on APC (1), combined with a second signal, termed costimulation, induced by ligation of Th cell surface CD28 molecules by APC B7 proteins (8, 9, 10). In the absence of costimulation, a state of unresponsiveness or clonal anergy is induced in CD4+ Th cells by TCR ligation (11).
Full activation of Th1 cells induces the binding of transcriptional
activators to the IL-2 promoter region (7, 12). In particular, the
binding of NF-AT3 proteins is
thought to be essential, but not sufficient, for IL-2 transcription
(7, 13). NF-AT binding sites are present in the promoter region of
many other cytokine genes, including IL-4 (14). In fact, the
transcription factors thought to be necessary for IL-2 expression,
which include NF-AT, AP-1, and NF-
B, are present in the nucleus of
activated Th1 and Th2 cells (15). While this model accounts for the
activation of IL-2 gene transcription in Th1 cells, it fails to explain
the loss of IL-2 gene expression observed in Th2 cells.
Various hypotheses have been proposed to explain the silencing of IL-2 expression in fully differentiated Th2 cells. The transient activation of IL-2 transcription and restriction of IL-2 expression to the Th1 subset suggest the presence of an inducible repressor in Th2 cells. We have shown previously that cycloheximide (CHX) treatment of activated Th2 cells results in transcription of the IL-2 gene (16). Furthermore, the same CHX treatment protocol, when applied to activated IL-2-producing EL-4 thymoma cells and Th1 cells, led to a superinduction of IL-2 transcription (17). These results provide functional evidence that a labile transcriptional repressor protein controls IL-2 expression in Th cells. A negative regulatory element, designated NRE-A, has been mapped 3' of the IL-2 promoter NFIL-2A site by mutational analysis, and a candidate binding factor, Nil-2-a, was isolated from a Jurkat cDNA library (18). However, it was later found that this is only a partial cDNA clone, since a protein with an identical carboxyl-terminal sequence, and an additional amino-terminal fragment, was cloned from a HeLa tumor cell cDNA library (19). The full-length protein, ZEB, contains two potential zinc finger E box-binding domains and silences transcription of the Ig heavy chain (IgH) gene (19).
The data presented in this work show that ZEB fulfills the criteria of a CHX-sensitive, transcriptional repressor of IL-2, involved in silencing of the IL-2 gene in activated Th2 cells and down-regulation of IL-2 expression in Th1 cells.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The EL.4IL-2 (EL-4) line is a Th1-like lymphoma that produces IL-2 after PMA stimulation (20, 21). The Th1 clone BK2.43 is specific for keyhole limpet hemocyanin, presented by I-Ab (kind gift of Dr. A. Infante, University of Texas, San Antonio, TX) (22). The Th2 clone D10G4.1, specific for conalbumin in the context of I-Ak, was obtained from American Type Culture Collection (Rockville, MD). D10G4.1 was characterized originally as a Th2 cell by its ability to produce IL-4, but not IL-2, upon activation (23). All Th cell clones were stimulated twice monthly with irradiated syngeneic APC and Ag. Briefly, splenocytes were prepared from I-Ak or I-Ab mice, irradiated with a dose of 2000 rad, and added to Th cells at a ratio of 10 to 1. APC and Th cells were resuspended in complete RPMI at 106 cells/ml. The mixture of Th cells and APC was then added to 24-well plates at 1 ml/well. Twenty-four hours after plating, the culture media were supplemented with either 10 U/ml rIL-2 or Con A-stimulated rat spleen cell supernatant to 10% by volume.
Polarized effector Th cells were generated by culturing splenocytes
from transgenic mice bearing a Vß3/V
11 TCR specific for pigeon
cytochrome c peptide 88–104 with APCs and cytokines (24, 25). Briefly, splenocytes from 3- to 6-mo-old mice were prepared and
passed over nylon wool columns. The nonadherent cells were treated with
anti-CD8 and anti-J11D Abs plus complement. Cells purified by
this protocol are 90% naive, resting CD4+ cells (26). The
naive CD4+ cells were cultured with mitomycin-treated
DCEK-ICAM cells, a fibroblast line that expresses surface
I-Ek, ICAM, and B7.1 molecules (27) as APC and pigeon
cytochrome c at 5 µg/ml. Th1 effectors were generated by
culturing CD4+ cells and APC in 50 U/ml IL-2, 1000 U/ml and
10 µg/ml anti-IL-4 Ab for 4 days, while Th2 effectors were
generated by culturing the cells in 200 U/ml IL-4 and 10 µg/ml
anti-IFN- Ab (24).
Electrophoretic mobility shift assays (EMSA)
The cultured Th cells were harvested by centrifugation and
resuspended at 5 x 105 cells/ml. Cells were
stimulated by the addition of PMA (10 ng/ml) plus ionomycin (500 ng/ml)
and returned to culture. Three hours after stimulation, CHX was added
to the appropriate cultures at a final concentration of 10 µg/ml.
Nine hours poststimulation, all cultures were washed three times with
cold PBS/0.1% BSA. Nuclear extracts were prepared from the cells,
using a modified Dingam protocol (28). Protein concentration for each
extract was determined with a Micro BCA Protein Assay Reagent Kit
(Pierce, Rockford, IL). Nuclear extracts were stored at -80°C. For
binding assays, 5 µg of total nuclear extract were incubated with
20,000 cpm of -32P-labeled oligonucleotide in the
presence of 0.5 µg of poly(dI-dC)·poly(dI-dC) on ice for 30 min.
The oligos used in binding and competition assays are as follows:
murine NRE-A sequence, 5'-CTGCCACACAGGTAAAGTCTT-3'; mutant NRE-A
sequence, 5'-CTGCCACACATTTAAAGTCTT-3' (mutation
indicated in bold); reporter mutant sequence,
5'-CTGCCACCTCTAGAGAGTCTT-3' (mutation indicated in bold);
and consensus Oct-1, 5'-TGTCGAATGCAAATCACTAGAA-3'. The NRE-A, NRE-A
mutant, and reporter mutant sequences were synthesized as separate,
single-stranded complimentary oligos, and then annealed by standard
protocol. The Oct-1 double-stranded oligo was purchased from Santa Cruz
Biotechnology (Santa Cruz, CA). All competitor sequences were used at
100-fold molar excess.
In certain EMSA-binding reactions, rZEB protein, anti-ZEB Ab, and
control antiserum were used. rZEB protein was produced by expressing
the pGEX-2X vector in bacteria and purifying the products on
glutathione agarose beads (19). The pGEX-2X vector contains carboxy
domain sequences of ZEB subcloned into a glutathione
S-transferase (GST) construct (19). Anti-ZEB antiserum
obtained from Funahashi et al. was purified from rabbits immunized with
EF1, the chicken homologue of ZEB (29). Anti-NF-AT antiserum, a gift
of McCaffrey et al. (30), was purified from rabbits immunized with
rNF-AT protein.
The binding reactions were loaded onto a 6% nondenaturing acrylamide gel in Tris borate EDTA (TBE) buffer and electrophoresed for 90 min at 4°C at a constant 150 V. The gel was dried onto Whatman paper and exposed to a Phospho Screen (Molecular Dynamics, Sunnyvale, CA) for 8 h. The exposed screen was read on a PhosphorImager (Molecular Dynamics) and printed on a Laserjet III (Hewlett Packard, Santa Clara, CA).
Transfection of reporter gene constructs and reporter gene assays
All constructs used in this study have been described previously: pIL-2 (-110/-101)/Luc, pIL-2 (-548)/Luc (18), and the ZEB expression and antisense plasmids (19). EL-4 and D10G4.1 Th cells were transfected by electroporation. Cells were removed from long term culture, washed twice, and resuspended at a concentration of 5 x 107 cells/ml in RPMI. A quantity amounting to O.5 ml of cell suspension was added to 30 µg reporter plasmid DNA in a 0.4-cm electroporation cuvette (Bio-Rad, Hercules, CA). To control for transfection efficiency, 10 µg of MSV-ß-Gal plasmid was cotransfected in each cuvette. For antisense experiments, 30 µg of each antisense vector, pCMV4x and pCMV6x, or 30 µg of each expression vector, pBSK4x and pSV2A2X, was cotransfected with reporter construct and MSV-ß-gal plasmid. Cells and plasmids were mixed by pipetting, and incubated at room temperature for 15 min. Cells were electroporated with 450 V, 125 µF in a Gene Pulsar (Bio-Rad). Immediately following electroporation, the cuvettes were put on ice. After 10 min, the cells were resuspended in RPMI at a concentration of 5 x 105 cells/ml. Electroporated Th cells were incubated overnight and then stimulated by the addition of 10 ng/ml PMA plus 500 ng/ml ionomycin. Transfected cells were harvested 20 h after stimulation by centrifugation, washed twice with PBS, and lysed in 0.5 ml cell culture lysis buffer (Promega, Madison, WI) by pipetting. Lysates were incubated for 15 min at room temperature, centrifuged to pellet debris, and frozen at -80°C until assayed.
For luciferase assays, Th cell lysates were removed from -80°C and thawed on ice. Forty microliters of lysate were mixed with 200 µl of luciferase assay substrate (Promega) and read in a luminometer (Tropix, Bedford, MA) after a 10-s delay. To control for transfection efficiency, ß-gal assays were performed on each sample separately, and the luciferase readings were normalized to the ß-gal activity. This was done with a Lumigal Detection Kit (Clontech Laboratories, Palo Alto, CA).
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The NRE-A site has been defined in the IL-2 promoter as a
potential binding site for a negative transcription factor that dampens
IL-2 production in the human Jurkat tumor (18). Furthermore, in vivo
treatment of stimulated murine Th2 and Th1 cells with CHX suggests the
activity of an IL-2 transcriptional repressor (16, 17). To determine
whether a connection exists between these two described activities,
nuclear extracts of murine Th cells were assayed for proteins that
could bind to the NRE-A site. As can be seen in Figure 1
, nuclear extracts from PMA plus
ionomycin-stimulated IL-2-nonproducing Th2, and IL-2-producing Th1 and
EL-4 thymoma cells contain an activity that binds specifically to the
IL-2 promoter NRE-A element in vitro. Nuclear proteins from D10G4.1
(Th2 clone), EL-4 (T cell tumor line), and BK2.43 (Th1 clone) cells
were incubated with a -32P-labeled double-stranded
NRE-A site oligo in EMSA. The reaction mixtures were then
electrophoresed on a nondenaturing acrylamide gel, and two retarded
DNA/protein complexes were observed (Fig. 1A, lane
4, and Fig. 1, B and C, respectively,
lane 1). The addition of a 100-fold molar excess of
unlabeled NRE-A sequence to the binding reaction completely interfered
with the formation of both the upper and lower complexes (Fig. 1A, lane 3, and Fig. 1, B and
C, respectively, lane 2), while an equal
amount of unlabeled mutant NRE-A (Fig. 1A, lane
2, and Fig. 1, B and C, respectively,
lane 3) or an unrelated DNA sequence containing the
Oct-1 binding site (Fig. 1A, lane 1; Fig. 1B, lane 4; and Fig. 1C, lane
5) was unable to block the formation of the upper complex,
indicating that protein from the nucleus of the Th cells binds
specifically to the NRE-A sequence. As an additional control, a second
NRE-A mutant sequence that is present in a mutant IL-2 promoter
construct described later was unable to compete for the binding of
radioactively labeled NRE-A sequence to the Th cell nuclear protein
(Fig. 1C, lane 4). These results
demonstrate that the nuclei of activated EL-4, Th1, and Th2 cells
contain DNA-binding proteins specific for the NRE-A site.
|
To establish that the DNA-binding specificity of the T cell
nuclear protein is consistent with that of ZEB, the Th cell nuclear
extracts were replaced in the binding mixture by rZEB, expressed in
bacteria as a GST fusion protein (19). As shown in Figure 1
D, the binding pattern with rZEB correlates with that
obtained using Th cell nuclear proteins. Mobility of the radioactively
32P-labeled NRE-A site was retarded by the binding of ZEB
protein (lane 1), and a 100-fold molar excess
of unlabeled NRE-A sequence competed away this binding
(lane 2), while an equal amount of unlabeled
mutant NRE-A sequence (lane 3) or an unrelated
DNA sequence (lane 4) could not. These data
demonstrate that both rZEB and the NRE-A-binding protein from Th cells
display an identical NRE-A sequence-specific binding in EMSA,
suggesting that the nuclear protein in the T cell extracts may in fact
be ZEB or a highly related protein.
The NRE-A-binding activity is ZEB or a highly related protein
To confirm that the nuclear NRE-A-binding protein in Th cells is
indeed ZEB, we used ZEB-specific Abs in the EMSA experiments to see if
they would react with the Th cell protein and block its interaction
with the NRE-A element. As expected, addition of polyclonal Abs raised
against the chicken homologue of ZEB, EF1 (29), to the reaction
mixture containing GST-ZEB completely prevented the formation of the
GST-ZEB-labeled NRE-A site complex (Fig. 2, lane 9). In
contrast, an equal amount of control polyclonal anti-NF-AT Ab did
not affect the formation of the GST-ZEB-labeled NRE-A site complex
(lane 10). Most importantly, the addition of
anti-ZEB Ab to the reaction mixtures containing Th cell nuclear
protein similarly abolished the formation of the specific
protein-labeled NRE-A site complex (note absence of upper complex in
Fig. 2, lane 4), while addition of control
anti-NF-AT Ab to the reaction mixture had no effect on the
formation of the protein/NRE-A site complex (Fig. 2, lane
5). Lanes 1 to 3 and 6 to
8 demonstrate the sequence-specific binding of the Th cell
nuclear protein and GST-ZEB, respectively, to the NRE-A site. These
results clearly demonstrate that ZEB or a highly related molecule is
the NRE-A-binding protein present in Th cell nuclear extracts.
|
The results presented to date indicate that nuclear extracts from
activated Th2, EL-4, and Th1 cells contain the NRE-A-binding protein,
ZEB. Figure 3 demonstrates that the
binding of nuclear ZEB to the NRE-A site is reduced by CHX treatment of
the Th cells. We had shown previously that treatment of stimulated
cells with CHX 3 h postactivation led to IL-2 transcription in the
Th2 clone D10G4.1 (16), and enhanced IL-2 transcription in EL-4 cells
(17). Thus, it was of interest to test whether CHX treatment had any
influence on the level of nuclear NRE-A-binding activity. As can be
seen in Figure 3, ZEB binding to NRE-A is almost totally abolished in
the nuclei of activated EL-4 cells that have been treated with CHX
(lane 2), compared with nuclear extracts from
activated EL-4 cells without CHX treatment (lane
4). CHX treatment of resting EL-4 cells also sharply
reduces the binding activity of ZEB for NRE-A (Fig. 3, lane
1 compared with lane 3). These results establish
that CHX-sensitive ZEB binding to the NRE-A site is correlated with
IL-2 transcriptional silencing. As CHX is known to inhibit protein
synthesis, we conclude that this compound is interfering with the
production of ZEB after cellular activation. This could account for the
derepression of IL-2 gene transcription observed in D10G4.1 Th2 cells
(16) and an increase in transcription observed in EL-4 cells
(17).
|
To confirm our findings in long-term Th cell clones, we examined
ZEB NRE-A binding in Th cells that are directly derived from a common
naive precursor. Purified naive CD4+ cells isolated
from AND TCR transgenic mice were polarized into effector Th1 and Th2
cells by culture with IFN- and IL-12 or IL-4, respectively (24, 25).
As can be seen in Figure 4A,
unstimulated polarized effector Th1 and Th2 cells (lanes
1 and 3, respectively) contain NRE-A-binding activity.
Upon treatment with PMA plus ionomycin for 6 h, NRE-A-binding
activity is reduced in IL-2-producing Th1 cells by 40%
(lane 2), compared with resting Th1 cells, as
determined by densitometry reading. On the other hand, NRE-A-binding
activity remains elevated in Th2 cells compared with resting Th2 cells
(23% reduction) (lane 4). Figure 4B depicts the NRE-A-binding specificity of the nuclear
factor in these cells. As previously shown in Th clones (Fig. 1), the
binding of 32P-labeled NRE-A sequence to Th nuclear protein
(lane 1) is fully competed away by a 100-fold
molar excess of unlabeled NRE-A (lane 2), but
not by the same amount of a mutant NRE-A sequence (lane
3). This EMSA analysis of NRE-A binding shows that activity
is reduced in polarized Th1 cells that produce large amounts of IL-2
and maintained in Th2 cells that make little IL-2. These results are
significant for two reasons: 1) They are consistent with the role of
ZEB as an IL-2 transcriptional repressor shown in Th clones; and 2)
they suggest that Th1 and Th2 cells derived from the same precursors
have different levels of ZEB activity following activation.
|
Once we had established that ZEB binds to the IL-2 NRE-A element
in activated Th cells, we wanted to determine a time course of
ZEB-binding activity in Th2 cells. If ZEB is a transcriptional
repressor of IL-2, then one would expect that ZEB NRE-A-binding
activity would be highest when IL-2 transcription is initiated.
Therefore, NRE-A-binding activity in D10G4.1 Th2 cells was examined at
multiple times after PMA plus ionomycin treatment. As can be seen in
Figure 5, this activity increases shortly
after activation (15% increase) (lane 2) and
remains at a high level for up to 6 h (lane
4). By 9 h after activation (lane
5), ZEB binding to the NRE-A site falls below levels
observed in resting D10G4.1 cells (62% reduction) (lane
1). This result, by demonstrating that ZEB-binding activity
in Th cells is highest at a time when IL-2 transcription is initiated
(7), suggests that ZEB can repress IL-2 transcription by binding to the
NRE-A element.
|
When the wild-type IL-2 promoter construct, pIL-2(-548)Luc, was
transiently transfected into the non-IL-2-transcribing Th2 clone,
D10G4.1, an insignificant level of reporter activity was observed after
activation with PMA plus ionomycin (Fig. 6A, top bar
compared with the bar directly below). This result is consistent with
the findings reported by others using the D10G4.1 clone (14, 15). Thus,
the wild-type IL-2 promoter construct accurately reflects the activity
of the endogenous IL-2 gene in Th2 cells, which is transcriptionally
silent upon activation. On the other hand, an IL-2 promoter construct,
pIL-2(-110/-101)Luc, lacking a functional ZEB binding NRE-A site,
produced a significant amount of luciferase activity in PMA plus
ionomycin-stimulated D10G4.1 Th2 cells (Fig. 6A, third
row), compared with the wild-type construct
(top row). As expected, neither construct
exhibited significant luciferase activity in resting D10G4.1 Th2 cells
(Fig. 6A, second and fourth
rows). This pattern of promoter activity was also observed
in another Th2 clone, CDC25 (data not shown), indicating that binding
of ZEB to the NRE-A site is a potential mechanism by which activated
Th2 cells repress IL-2 production in vivo.
|
B, third row), compared with the
wild-type IL-2 promoter (Fig. 6B, top
row). The key difference between these results and those
observed in the Th2 cells is the significant inducibility of the
wild-type IL-2 promoter construct in activated EL-4 cells
(top row) (note that the scale of units used
in Fig. 6B is 10 times higher than in Fig. 6A). This indicates that ZEB binding to the IL-2
promoter NRE-A site can block IL-2 transcription in both Th2 and Th1
cells. Interestingly, similar results were obtained in the B cell line
BalI7 (data not shown), and it is known that B cells express
high levels of ZEB (19). The observation that the wild-type IL-2
promoter construct in EL-4 cells produces fivefold less luciferase
activity than the mutant construct, along with data obtained in the
EMSA, suggests that the NRE-A-binding factor is a negative
transcriptional regulator of the IL-2 gene in Th1 cells. Cotransfection of ZEB antisense constructs increases the inducible transcription of the IL-2 promoter
To confirm that ZEB binding to NRE-A represses IL-2
transcription, ZEB antisense constructs were used to disrupt the
activity of the putative repressor (Fig. 7, A and B).
Vectors containing antisense sequences to the carboxyl- and
amino-terminal halves of ZEB were cotransfected with the IL-2 reporter
constructs into EL-4 cells. These antisense vectors are under control
of the CMV promoter, which yields strong constitutive expression in
lymphocytes (31). As controls, vectors containing noncoding, partial
ZEB sequences in the sense orientation or pBluescript plasmid
DNA lacking ZEB sequences were cotransfected with the IL-2 promoter
constructs. When the ZEB antisense vectors were cotransfected with the
wild-type IL-2 promoter construct, a threefold increase in the activity
of the reporter gene was seen in response to PMA plus ionomycin (Fig. 7A, row 1), compared with the
cotransfection of the ZEB sense vectors or control plasmid DNA (Fig. 7A, rows 3 and 5). This result
specifically demonstrates that it is the binding of ZEB to the IL-2
promoter NRE-A site that represses IL-2 transcription in these
cells.
|
B, row
1), compared with cotransfection with control vectors
(rows 3 and 5). Thus, as predicted
by our model, the induction of the NRE-A mutant IL-2 promoter is
insensitive to ZEB disruption. While the reporter constructs alone
provide evidence for the repressive effect of ZEB binding to the NRE-A
site, the cotransfection experiments clearly implicate ZEB as one of
the factors that mediate the repression. |
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
We had reported earlier that repression of IL-2 transcription in
activated Th2 cells is controlled by a CHX-sensitive element (16).
These studies extend initial observations by Efrat and Kaempfer (32),
who had also concluded that IL-2 transcription is controlled by a
CHX-sensitive transcriptional repressor. Furthermore, we had
demonstrated that IL-2 transcription is increased in Th1 cells by CHX
treatment after activation, and that this enhanced response was not
solely due to increased NF-B activity, resulting from the block in
inhibitory IB production by CHX (17). These observations led us to
propose that the activity of an IL-2-specific transcriptional silencer
was responsible for the silencing of IL-2 expression in activated Th2
cells and dampening of the IL-2 response in Th1 cells.
The present results identify ZEB as the repressor factor. We show by EMSA that nuclear extracts from Th cells contain specific NRE-A-binding activity. The binding specificity of rZEB for the NRE-A site is identical to that of the Th cell nuclear factor. Furthermore, the ability of anti-ZEB Abs to ablate the NRE-A-binding complex in the gel retardation studies demonstrates that at least one component in this complex shares an epitope with ZEB. Therefore, we conclude that the NRE-A-binding complex in the Th cell clones contains ZEB or a highly related protein. It was also found that a CHX treatment protocol that induces IL-2 transcription in activated Th2 cells and increases transcription in EL-4 cells, decreases the binding activity of ZEB to the NRE-A in the nucleus of these cells. In addition, EMSA analyses of NRE-A-binding proteins in polarized Th cells demonstrate that activated Th1 cells have reduced levels of IL-2 promoter NRE-A binding, while in Th2 cells NRE-A-binding activity remains high. We believe that activated Th2 cells are unable to reduce ZEB binding to the IL-2 promoter NRE-A site, despite induction of IL-2 transcriptional activators.
These in vitro data correlate well with the results obtained in functional assays, involving transient transfection of wild-type and NRE-A mutant IL-2 promoter constructs. We show that a construct lacking a ZEB binding NRE-A site produced significant reporter gene expression in activated Th2 cells, compared with a wild-type promoter construct that showed only negligible activity. These results indicate that the interaction of ZEB with the NRE-A site can repress IL-2 transcription in a cell type that normally does not produce IL-2 upon activation. This is especially interesting in light of the fact that nuclear extracts from D10G4.1 Th2 cells have been shown to contain NF-AT proteins (14, 15, 33) that regulate the transcriptional activation of multiple cytokine genes, including IL-2 (7) and IL-4 (14). In addition, we have observed that in stimulated Th1-like EL-4 cells, the NRE-A mutant IL-2 construct is at least five times more active than the wild-type IL-2 construct. These results are consistent with our observation that CHX treatment 3 h postactivation enhances transcriptional inducibility of the wild-type, but not the NRE-A mutant IL-2 promoter construct (data not shown).
The cotransfection of EL-4 cells with ZEB antisense vectors, along with wild-type or mutant IL-2 reporter constructs, provides further evidence that identifies ZEB as the IL-2 silencer in Th cells. The presumed disruption of ZEB mRNA translation by these vectors led to a significant increase in the inducibility of the IL-2 wild-type promoter construct, while the activity of an IL-2 NRE-A mutant promoter was much less affected. Furthermore, we have evidence that the ZEB antisense treatment that we use does in fact reduce the nuclear concentration of ZEB in EL-4 cells (data not shown). These independently derived experimental results indicate that ZEB is an inducible, CHX-sensitive repressor of IL-2 gene transcription in Th2 and, to a lesser extent, Th1 cells. Again, based on our results, it appears that ZEB binding to the promoter NRE-A site disrupts the formation of a stable transcription complex, thereby preventing IL-2 expression. These findings are all the more remarkable in that ZEB binding to the NRE-A site appears to overcome the effect of multiple activators that are known to bind to IL-2 promoter elements (7).
Analysis of transcription factor binding sites in the IL-2 promoter by other investigators also suggests the existence of a repressor element in a sequence that overlaps the NRE-A site (34). Furthermore, our studies agree with results obtained in anergized Th cells, showing that NRE-A-binding factors mediate the IL-2 transcriptional silencing observed in these cells (35). In addition, it has been reported that the NRE-A site can confer repression on a heterologous AP-1 promoter construct in an in vitro transcription assay (35).
A key difference between Th1 and Th2 subsets is the extinction of IL-2 production in Th2 cells. While there are few differences in IL-2 transcriptional activators between Th1 and Th2 cells (9, 15, 36, 37), elevated cAMP activity (38, 39, 40) has been suggested to cause IL-2 transcriptional silencing in Th2 cells. In this regard, it should be noted that not only do Th2 cells tolerate a significantly higher level of cAMP (41), they contain more cAMP activity than Th1 cells (42). Other studies indicate that activation-induced cAMP inhibition of IL-2 transcription in Th cells occurs by blocking tyrosine phosphorylation of a 100-kDa protein (38). This phosphoprotein might then affect ZEB activity in Th2 cells.
More recently, it was shown that fused Th1/Th2 cell hybrids produce both IL-2 and IL-4, leading the authors to conclude that a Th2-specific repressor of IL-2 transcription is unlikely (43). However, these results can also be explained by an increase in transcriptional activators provided by the Th1 cell, overriding the effect of an IL-2 repressor in the fused cells. In addition, Th cells from NF-AT1 knockout mice produce normal amounts of IL-2 protein (44, 45), contradicting the hypothesis that combined NF-AT/AP-1 activity fully accounts for the regulation of IL-2 transcription. However, targeted disruption of the other known NF-AT gene family members may yet prove the necessity of this factor in IL-2 transcription. Together, these preliminary findings illustrate the complexity of IL-2 transcriptional regulation and demonstrate that a mechanism that fully defines its expression in Th cells has yet to be described.
Our data extend the findings of these studies and show that ZEB is
present in the nuclei of resting Th2 and Th1 cells. In Th1 cells,
ligation of the TCR/CD3 complex and CD28 by Ag/APC interaction induces
the production of AP-1 and translocation of NF-AT and NF-B proteins
to the nucleus. We propose that the binding of these and other factors
to the IL-2 promoter overrides ZEB-mediated repression and results in
the transcription of IL-2 in Th1 cells, while in Th2 cells these
activating factors are unable to overcome the repressive activity of
ZEB binding to the IL-2 promoter NRE-A site.
There are many potential mechanisms that could link activation of Th2 cells to transcriptional repression of IL-2 by ZEB. Our results indicate that activation increases ZEB activity in these cells. It is important to note that ZEB and its homologues contain at least one potential protein-protein interaction site in a proline-rich domain (19, 46). Thus, it is possible that ZEB NRE-A-binding activity can be regulated in Th cells by protein-protein interactions. Experiments are underway to further define the regulation of IL-2 by ZEB in Th cells and to determine how the expression of ZEB is controlled.
ZEB has been shown to be a transcriptional repressor in diverse
systems. In vitro studies suggest that ZEB binding to the µE5 E box
element in the IgH promoter silences transcription of this gene (19).
This conclusion is supported by data obtained in transgenic mice
bearing a mutant IgH µE5 site. These mice produce inappropriate IgH
mRNA transcripts in muscle, heart, and lung (47). In addition, it has
been reported that EF1, the chicken homologue of ZEB, represses
transcription of -crystallin, a component of the eye, and
muscle-specific genes in the developing chicken embryo (48). ZEB has
also been suggested as a potential transcriptional silencer of the CD4
gene (49). Thus, ZEB, which is highly conserved in evolution (46, 50),
may act as a repressor of multiple genes, including IL-2.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
2 Address correspondence and reprint requests to Tufts University School of Medicine, Department of Pathology, 136 Harrison Avenue, Boston, MA 02111-1800. E-mail address:
3 Abbreviations used in this paper: NF-AT, nuclear factor-activated T cells; ß-gal, ß-galactosidase; CHX, cycloheximide; EMSA, electrophoretic mobility shift assay; GST, glutathione S-transferase; NF-B, nuclear factor-B; NRE, negative regulatory element.
Received for publication August 1, 1997. Accepted for publication January 15, 1998.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
B. J. Immunol. 147:416.[Abstract]
B activation in T helper 1 and T helper 2 cells. J. Immunol. 156:56.[Abstract]
gene expression at multiple levels. J. Immunol. 148:2845.[Abstract]
-crystalline enhancer-binding protein delta EF1 is a repressor of E2-box mediated gene activation. Mol. Cell. Biol. 14:5692.EF1 and cross-species conservation of its domains. Gene 173:227.[Medline]
This article has been cited by other articles:
|
|
 |
|
  T. Kowase, H. E. Walsh, D. S. Darling, and M. A. Shupnik Estrogen Enhances Gonadotropin-Releasing Hormone-Stimulated Transcription of the Luteinizing Hormone Subunit Promoters via Altered Expression of Stimulatory and Suppressive Transcription Factors Endocrinology, December 1, 2007; 148(12): 6083 - 6091. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 W.-h. Feng, R. J. Kraus, S. J. Dickerson, H. J. Lim, R. J. Jones, X. Yu, J. E. Mertz, and S. C. Kenney ZEB1 and c-Jun Levels Contribute to the Establishment of Highly Lytic Epstein-Barr Virus Infection in Gastric AGS Cells J. Virol., September 15, 2007; 81(18): 10113 - 10122. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. Chen, I. Yusuf, H.-M. Andersen, and D. A. Fruman FOXO Transcription Factors Cooperate with {delta}EF1 to Activate Growth Suppressive Genes in B Lymphocytes. J. Immunol., March 1, 2006; 176(5): 2711 - 2721. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 E. S. Hwang, J.-H. Hong, and L. H. Glimcher IL-2 production in developing Th1 cells is regulated by heterodimerization of RelA and T-bet and requires T-bet serine residue 508 J. Exp. Med., November 7, 2005; 202(9): 1289 - 1300. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C Nielsen, H Laustrup, A Voss, P Junker, S Husby, and S T Lillevang A putative regulatory polymorphism in PD-1 is associated with nephropathy in a population-based cohort of systemic lupus erythematosus patients Lupus, July 1, 2004; 13(7): 510 - 516. [Abstract] [PDF]
|
|
|
|
 |
|
 P. Christmas, N. Carlesso, H. Shang, S.-M. Cheng, B. M. Weber, F. I. Preffer, D. T. Scadden, and R. J. Soberman Myeloid Expression of Cytochrome P450 4F3 Is Determined by a Lineage-specific Alternative Promoter J. Biol. Chem., June 27, 2003; 278(27): 25133 - 25142. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. Rao, S. Gerondakis, D. Woltring, and M. F. Shannon c-Rel Is Required for Chromatin Remodeling Across the IL-2 Gene Promoter J. Immunol., April 1, 2003; 170(7): 3724 - 3731. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
R. J. Kraus, J. G. Perrigoue, and J. E. Mertz ZEB Negatively Regulates the Lytic-Switch BZLF1 Gene Promoter of Epstein-Barr Virus J. Virol., December 6, 2002; 77(1): 199 - 207. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 G. Fontemaggi, A. Gurtner, S. Strano, Y. Higashi, A. Sacchi, G. Piaggio, and G. Blandino The Transcriptional Repressor ZEB Regulates p73 Expression at the Crossroad between Proliferation and Differentiation Mol. Cell. Biol., December 15, 2001; 21(24): 8461 - 8470. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. B. Han, S. H. Park, Y. J. Jeon, Y. K. Kim, H. M. Kim, and K. H. Yang Prodigiosin Blocks T Cell Activation by Inhibiting Interleukin-2Ralpha Expression and Delays Progression of Autoimmune Diabetes and Collagen-Induced Arthritis J. Pharmacol. Exp. Ther., November 1, 2001; 299(2): 415 - 425. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 D. L. Lazarova, M. Bordonaro, and A. C. Sartorelli Transcriptional Regulation of the Vitamin D3 Receptor Gene by ZEB Cell Growth Differ., June 1, 2001; 12(6): 319 - 326. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. Cianferoni, J. T. Schroeder, J. Kim, J. W. Schmidt, L. M. Lichtenstein, S. N. Georas, and V. Casolaro Selective inhibition of interleukin-4 gene expression in human T cells by aspirin Blood, March 15, 2001; 97(6): 1742 - 1749. [Abstract] [Full Text] [PDF]
|
|
