Transcription Factor GATA-3 Is Differentially Expressed in Murine Th1 and Th2 Cells and Controls Th2-specific Expression of the Interleukin-5 Gene*

(Received for publication, May 2, 1997, and in revised form, June 5, 1997)

Dong-Hong Zhang Dagger , Lauren Cohn Dagger §, Prabir Ray Dagger , Kim Bottomly § and Anuradha Ray Dagger par

From the Dagger  Department of Internal Medicine, Pulmonary and Critical Care Section, the § Section of Immunobiology, and the  Howard Hughes Medical Institute, Yale University School of Medicine, New Haven, Connecticut 06520

ABSTRACT
INTRODUCTION
EXPERIMENTAL PROCEDURES
RESULTS
DISCUSSION
FOOTNOTES
ACKNOWLEDGEMENTS
REFERENCES


ABSTRACT

Interleukin-5 (IL-5), which is produced by CD4+ T helper 2 (Th2) cells, but not by Th1 cells, plays a key role in the development of eosinophilia in asthma. Despite increasing evidence that the outcome of many diseases is determined by the ratio of the two subsets of CD4+ T helper cells, Th1 and Th2, the molecular basis for Th1- and Th2-specific gene expression remains to be elucidated. We previously established a critical role for the transcription factor GATA-3 in IL-5 promoter activation in EL-4 cells, which express both Th1- and Th2-type cytokines. Our studies reported here demonstrate that GATA-3 is critical for expression of the IL-5 gene in bona fide Th2 cells. Whereas mutations in the GATA-3 site abolished antigen- or cAMP-stimulated IL-5 promoter activation in Th2 cells, ectopic expression of GATA-3 in Th1 cells or in a non-lymphoid, non-IL-5-producing cell line activated the IL-5 promoter. During the differentiation of naive CD4+ T cells isolated from T cell receptor transgenic mice, GATA-3 gene expression was up-regulated in developing Th2 cells, but was down-regulated in Th1 cells, and antigen- or cAMP-activated Th2 cells (but not Th1 cells) expressed the GATA-3 protein. Thus, GATA-3 may play an important role in the balance between Th1 and Th2 subsets in immune responses. Inhibition of GATA-3 activity has therapeutic potential in the treatment of asthma and other hypereosinophilic diseases.


INTRODUCTION

Activated CD4+ T cells are subdivided into two subsets, T helper 1 (Th1) and Th2, based on their biological functions, which, in turn, depend on the cytokines they produce (1-4). Th1 cells produce interleukin-2 (IL-2)1 and interferon-gamma (IFN-gamma ) and stimulate microbicidal activity in macrophages and promote cell-mediated immunity (1-4). Th2 cells, on the other hand, produce IL-4 and IL-5, which stimulate IgE production and eosinophilic inflammation, respectively (1-4). There is good evidence that in atopic asthmatics, a Th2-type response occurs in the airways (5-8). Although it appears that the outcome of many diseases such as asthma is determined by the ratio of Th1 to Th2 cells (4), the molecular basis for Th1- and Th2-specific gene expression remains to be elucidated.

Asthma is a chronic obstructive disease of the small airways. Evolving evidence indicates that asthma is the result of an inflammatory process interacting with a susceptible airway best defined at present as airway hyperresponsiveness (9). In asthma, the most striking and consistent pathophysiology is damage to the bronchial epithelium caused by cytotoxic cationic proteins released by infiltrating eosinophils (10, 11). Various lines of evidence indicate that secreted products of activated T cells, such as the cytokine IL-5, play a central role in orchestrating the unique inflammatory response seen in asthma. Since it was isolated and cloned in the mid-1980s, the intimate relationship between IL-5, eosinophils, and asthma has been extensively documented (12).

IL-5 has multiple effects on the biology of eosinophils not limited to differentiation, proliferation, recruitment, and activation (12). Increasing evidence places IL-5 in a key role in the development of eosinophilia in asthma (5, 6). IL-5 mRNA was significantly enhanced in bronchoalveolar lavage cells obtained from asthmatics challenged with ragweed antigen (8). Again, peripheral blood T cells from asthmatics were found to secrete IL-5 in response to the common house dust mite (Dermatophagoides farinae) antigen (13). Most striking, in ovalbumin-sensitized guinea pigs and mice, monoclonal antibody to IL-5 decreased pulmonary eosinophilia and prevented the development of airway hyperresponsiveness (14, 15). Also, in a mouse model of asthma, IL-5-deficient mice were found to lack eosinophilia, lung pathology (16, 17), and airway hyperresponsiveness upon allergen challenge (16).

In both humans and mice, the production of IL-5 is restricted to a few cell types, which include T cells (7), mast cells (18), and eosinophils (19), the predominant source being T cells of the Th2-type (7). In general, IL-5 is not produced constitutively by Th2 cells. IL-5 gene expression has been shown to be stimulated by antigen, mitogens (concanavalin A), eicosanoid compounds (leukotriene B4 and prostaglandins), and cytokines (20, 21). Intracellular cAMP-increasing agents, such as IL-1alpha , prostaglandin E2, and the cAMP analogue dibutyryl cyclic AMP (Bt2cAMP), have been shown to differentially regulate cytokine production by Th1 and Th2 cells. Whereas the production of the Th1 response-inducing cytokine IL-12 and that of the Th1 cytokines IL-2 and IFN-gamma are inhibited by cAMP-increasing agents, the production of IL-5 is strongly induced by the same agents, suggesting a possible immunoregulatory role for this second messenger (21-24).

The molecular mechanisms underlying Th2 cell-specific IL-5 gene expression are unclear. In our previous studies of IL-5 promoter activation by cAMP in the murine cell line EL-4, which expresses both Th2- and Th1-type cytokines, we showed that deletion of the IL-5 promoter to -66, which disrupted a GATA site located between -70 and -60, abolished activation of the promoter (25). Furthermore, in electrophoretic mobility shift assays (EMSAs), we demonstrated that the transcription factor GATA-3, but not GATA-4, binds to this GATA site (25). This was the first description of the involvement of GATA-3 in the transcription of any cytokine gene (25). Yamagata et al. (26), in their studies of transcription of the human IL-5 gene in the ATL-16T cell line, also demonstrated the importance of the GATA site in expression of the human IL-5 gene. However, two important points merit consideration in comparing these two studies. First, IL-5 gene expression in ATL-16T cells is largely constitutive (27).2 However, in both humans and mice, the IL-5 gene is expressed in an inducible fashion, and therefore, the ATL-16T cells do not reflect the typical expression characteristics of the IL-5 gene either in humans or in mice. Second, GATA-4 is predominantly expressed in the heart, intestines, epithelium, and reproductive organs, and its expression is low or undetectable in both human and murine T cells (28, 29). Therefore, the atypical high level of GATA-4 activity in ATL-16T cells may contribute to the atypical (constitutive) nature of IL-5 gene expression in these cells. In another study, Prieschl et al. (30) showed that the GATA site located between -70 and -60 in the IL-5 promoter is also important for IL-5 gene expression in mast cells. In our previous study, we additionally demonstrated that activation of the IL-5 promoter also requires an intact AP-1 site within the CLE0 element (consensus lymphokine element 0) located between -53 and -39 in the promoter; mutation of this site in the context of an ~550-bp promoter totally abrogated promoter activity.

In this report, we show that the transcription factor GATA-3 is crucial for IL-5 gene expression in bona fide Th2 cells and that ectopic expression of GATA-3 alone results in IL-5 promoter activation in a non-IL-5-producing cell line. We also show that GATA-3 activity is present only in Th2 cells and is undetectable in Th1 cells. Inhibition of GATA-3 activity may therefore be effective in the treatment of asthma and other hypereosinophilic diseases.


EXPERIMENTAL PROCEDURES

Generation and Maintenance of Th1 and Th2 Cells

Both D10 and C19 clones were maintained in Click's medium supplemented with 10% fetal bovine serum, 5 units/ml murine recombinant IL-2 (Boehringer Mannheim), 50 µM beta -mercaptoethanol, 2 mM L-glutamine, and 50 µg/ml gentamycin at 37 °C with 5% CO2 (31, 32). The cells were stimulated every 2 weeks with the specific antigen (conalbumin for D10 cells, used at 100 µg/ml) and peptide AC1-16 (ASQKRPSQRHGSKYL; derived from myelin basic protein, used at 5 µg/ml) and mitomycin C-treated splenocytes from syngeneic mice (I-Ak for D10 cells and I-Au for C19 cells). Prior to use in experiments, dead cells were removed by density gradient fractionation using lymphocyte separation medium (Organon Teknika). DO11.10 mice, which are transgenic for the TCR recognizing the ovalbumin peptide 323-339 (pOVA323-339), were provided on BALB/c background by Dr. Ken Murphy (Washington University, St. Louis, MO). To generate Th1 or Th2 cells from DO11.10 mice, naive CD4+ T cells were first isolated from the spleens by negative selection using monoclonal antibodies to CD8, class II MHC I-Ad, and anti-Ig-coated magnetic beads (Collaborative Research). Cultures were set up in flasks containing equal numbers of CD4+ T cells and T cell-depleted APCs at a concentration of 2 × 106 cells/ml. To generate Th1 cells, cultures contained pOVA323-339 at 5 µg/ml, IL-12 at 5 ng/ml, IL-2 at 10 units/ml, and anti-IL-4 at inhibitory concentrations. To generate Th2 cells, cultures contained pOVA323-339 at 5 µg/ml, IL-4 at 200 units/ml, IL-2 at 10 units/ml, and anti-IFN-gamma antibody. Cells were maintained in culture for 3 days, and at the end of this period, cells were further stimulated with fresh mitomycin C-treated and T cell-depleted APCs and Ag for 8 or 24 h (for making nuclear extracts) or for 48 h for cytokine assays. Culture supernatants were assayed for the presence of cytokines by ELISA using kits from Endogen, Inc. (sensitivity: IL-4, 6 pg/ml; IL-5, 0.1 ng/ml; and IFN-gamma , 2 ng/ml).

RNA Isolation and Northern Analysis

Total cellular RNA was prepared by using Trizol (Life Technologies, Inc.) according to the instructions of the manufacturer. 10 µg of total RNA from each sample was fractionated on a formaldehyde-agarose gel and transferred to a nylon membrane. DNA fragments derived from murine GATA-3 cDNA (~60-bp BglI-ClaI fragment not containing any part of the zinc finger domain) were labeled with [alpha -32P]dCTP using a random primer DNA labeling kit (Boehringer Mannheim). Hybridization was performed using QuikHyb (Stratagene) according to the instructions of the manufacturer.

Transfection Assays

Rested D10 or A.E7 cells were washed once in serum-free RPMI 1640 medium and resuspended in the same medium. Cells (5 × 106) were incubated with 15 µg of DNA (5 µg of reporter plasmid, 2 µg of cytomegalovirus-beta -galactosidase plasmid as a monitor for transfection efficiency, and carrier plasmid pGEM7Z to make up to 15 µg of total DNA) for 10 min at room temperature, and electroporation was carried out using a GenePulser (Bio-Rad) at 0.27 kV and 960 microfarads. The cells were left on ice for 10-30 min, diluted to 5 ml with fresh medium, and incubated at 37 °C with or without Bt2cAMP + PMA. For antigen stimulation, rested cells were first stimulated with conalbumin and mitomycin C-treated and T cell-depleted APCs in complete medium containing 5 units/ml IL-2 for 72 h and then subjected to electroporation. Cells were harvested for reporter gene assays as described previously (25). HeLa cells were transfected as described previously (33, 34).

Electrophoretic Mobility Shift Assays

Cells were left unstimulated or were stimulated as described above. All APCs were mitomycin C-treated and depleted of T cells. Nuclear extracts were prepared as described previously (25). The probes in the EMSAs were two double-stranded oligonucleotides containing sequences between -57 and -34 (containing the CLE0 element) and between -73 and -54 (containing the GATA element) in the IL-5 gene, and 22-bp oligonucleotides containing the consensus CREB element (from Stratagene). The oligonucleotides for the mutant CREB element were purchased from Santa Cruz Biotechnology. The sequences of the oligonucleotides used in the EMSAs were as follows: -73CCTCTATCTGATTGTTAGCA-54 (wild-type GATA), CCTCgcgaTGATTGTTAGCA (GATA mutant 1), CCTCTATCTGAaaccTAGCA (GATA mutant 2), CCTCTATCcttTTGTTAGCA (GATA mutant 3), and -57AGCAATTATTCATTTCCTCAGAGA-34 (CLE0). Complementary oligonucleotides were annealed before use in EMSAs. The antibodies to the c-Jun, JunB, JunD, and GATA-3 proteins were purchased from Santa Cruz Biotechnology. The anti-GATA-3 antibody was a mouse monoclonal IgG1 that does not cross-react with GATA-1, GATA-2, or GATA-4. The anti-Fos antibody was purchased from Oncogene Science Inc. The anti-GATA-4 antibody was kindly provided by Dr. David Wilson. The competitor oligonucleotides were added at a 100-fold molar excess. The binding reactions were analyzed by electrophoresis on 6% native polyacrylamide gels (acrylamide/bisacrylamide = 30:1). Electrophoresis was carried out at 200 V in 0.5 × TBE (1 × TBE = 0.05 M Tris base, 0.05 M boric acid, and 1.0 mM EDTA) at 4 °C. Gels were dried and subjected to autoradiography.


RESULTS

We have used the established nontransformed murine T cell clones D10.G4.1 (Th2) (31) and C19 (Th1) (32) and Th1 and Th2 cells obtained by differentiation of naive CD4+ T cells from DO11.10 TCR transgenic mice to gain insight into mechanisms that permit IL-5 gene expression in Th2 cells but limit its expression in Th1 cells.

The GATA Site in the 5'-Flanking Region of the IL-5 Gene Is Critical for IL-5 Promoter Activation in Th2 Cells

In our previous studies of cAMP-induced IL-5 promoter activation using an ~550-bp promoter fragment and the murine T cell line EL-4, we had identified two regions in the IL-5 5'-flanking region that were critical for induction of the IL-5 promoter: one was the AP-1-binding site within the CLE0 element, whereas the other was a region between -70 and -60 containing two overlapping GATA sites, deletion of which abrogated activation of the promoter (25). To elucidate the molecular mechanisms underlying transcriptional activation of the IL-5 gene by Ag in Th2 cells, we transfected the murine Th2 clone D10.G4.1 (31) with a reporter gene (firefly luciferase) construct containing a 1.7-kb promoter fragment from the 5'-flanking region immediately upstream of the transcriptional start site of the IL-5 gene. Both Ag and Bt2cAMP + PMA caused a 10-20-fold activation of the wild-type IL-5 promoter, and mutations in the AP-1 site or the GATA site in the context of the 1.7-kb promoter completely abolished activation of the IL-5 promoter (Fig. 1A). Mutations in the NF-AT site, on the other hand, had no effect on IL-5 promoter activity.


Fig. 1. Transcriptional activation of the murine IL-5 promoter in D10 cells requires the GATA-3 and AP-1 sites (within the CLE0 element), but not the NF-AT site. A, rested cells were transfected with the indicated plasmids by electroporation and either left unstimulated or stimulated with Bt2cAMP (1 mM) + PMA (25 ng/ml). The wild-type (wt) promoter contained sequences between approximately -1700 and +24 of the IL-5 promoter, and the individual mutations were made in the context of this fragment. For antigen stimulation, rested cells were first stimulated with conalbumin and mitomycin C-treated and T cell-depleted APCs in complete medium containing 5 units/ml IL-2 for 72 h and then subjected to electroporation. After 18-20 h, cells were harvested, and luciferase and beta -galactosidase assays were performed as described previously (25). The luciferase activity (arbitrary units) for each of the reporter plasmids is shown in the bar graph, with results representing the average of multiple experiments and normalized for beta -galactosidase activity. The deviations were no more than 15% between experiments. B, the location and sequence of the three site-directed mutations in the IL-5 promoter are identified. Base pair changes are identified with lower-case letters. mut., mutant.
[View Larger Version of this Image (20K GIF file)]

GATA-3 DNA Binding Activity Is Present in Activated Th2 Cells, but Is Absent in Activated Th1 Cells

We next investigated the binding of nuclear proteins to the GATA element and the CLE0 element using nuclear extracts from Th2 (D10) cells. As shown in Fig. 2A (lane 1), two complexes were detected using nuclear extracts from unstimulated D10 cells. The binding intensity of both complexes was augmented upon stimulation of the cells with Bt2cAMP + PMA (lane 2). Although both complexes were competed for by an excess of the unlabeled wild-type oligonucleotide, competition for complex I formation was incomplete even with a 100-fold molar excess of the unlabeled competitor, suggesting that complex I binds with a lower affinity to the GATA site than complex II (lane 3). Oligonucleotides containing specific mutations in three different regions of the double GATA site were also used as competitors. Mutant 1 contained mutations in the distal GATA sequence, and mutant 2 in the proximal sequence, whereas mutant 3 was mutated in both sequences. None of these mutants was able to compete for formation of the complexes, suggesting the involvement of the entire sequence between -70 and -60 in the formation of the complexes (lanes 4-6). The complexes (especially complex II) were supershifted by an anti-GATA-3 antibody (Ab) (lane 7), but not by an anti-GATA-4 antiserum (lane 9). To compare GATA-3 DNA binding activity between Th1 and Th2 cells, we performed a similar analysis with nuclear extracts from both D10 and C19 (Th1) cells prepared in the same experiment under identical conditions. As shown in Fig. 2B (lower panel), nuclear proteins from induced D10 cells generated two complexes that were supershifted by the anti-GATA-3 Ab (lane 3), but not by the anti-GATA-4 Ab (lane 4). Using identical protein amounts of nuclear extracts prepared from C19 cells, we detected a very low level of binding activity in resting cells (Fig. 2B, upper panel, lane 1). Upon treatment of the C19 cells with Bt2cAMP and PMA, whether alone (data not shown) or in combination, the intensity of the complexes did not increase, but consistently diminished (Fig. 2B, upper panel, compare lane 2 with lane 1).


Fig. 2. Th1 and Th2 clones differentially express GATA-3 DNA binding activity. A, shown is the GATA-3 DNA binding activity in unstimulated and stimulated Th2 (D10) cells. B, GATA-3 DNA binding activity increases upon stimulation of Th2 cells with Bt2cAMP+ PMA, but decreases in similarly stimulated Th1 cells, whereas Th1 and Th2 cells contain equivalent CLE0 binding activity. Nuclear extracts were prepared as described previously (25). The probes in the EMSAs were two double-stranded oligonucleotides containing sequences between (i) -57 and -34 (containing the CLE0 element) and (ii) -73 and -54 (containing the GATA element) in the IL-5 gene. The EMSAs were performed essentially as reported previously (25). 2 and 1 µg of protein were used in each lane for GATA and CLE0 binding assays, respectively. 2 µg of the indicated antibodies was added per 20-µl reaction volume. The competitor oligonucleotides were added at a 100-fold molar excess. The binding reactions were analyzed by electrophoresis on 6% native polyacrylamide gels. Gels were dried and subjected to autoradiography. mut., mutant.
[View Larger Version of this Image (88K GIF file)]

We then tested the same extracts for binding to the AP-1 site within the CLE0 element. As shown in Fig. 2B, used at only half the amounts used in the GATA-3 binding assays, robust inducible binding activity was detected with both D10 and C19 nuclear extracts. The anti-c-Jun and anti-c-Fos (reactive to all Fos family proteins) Abs did not affect the formation of the complex with either extract (lanes 7 and 10, respectively), whereas the anti-JunB and anti-JunD Abs supershifted/inhibited complex formation (lanes 8 and 9, respectively). Taken together, these results demonstrated that D10 and C19 cells contain similar levels of AP-1 binding activity. However, D10 cells constitutively contain some GATA-3 DNA binding activity that is augmented upon stimulation of the cells. In contrast, unstimulated C19 cells contain very little GATA-3 DNA binding activity that decreases upon stimulation of the cells.

Th1 and Th2 Cells Obtained by Differentiation of Naive CD4+ T Cells Display Differential GATA-3 DNA Binding Activity

We investigated whether the observed difference in GATA-3 binding activity between the Th1 (C19) and Th2 (D10) clones was also true in Th1 and Th2 cells obtained by differentiation of naive splenic CD4+ T cells from DO11.10 TCR transgenic mice. Nuclear extracts were prepared from the Th1 and Th2 cells stimulated with Ag for 8 or 24 h, and GATA and AP-1 binding activities were determined with the extracts. The Th1 cell extract from both time points generated two complexes with the GATA probe (Fig. 3A, lanes 1-6), whereas the Th2 extract generated three complexes (lanes 7-12). Complexes I and II were of the same mobility and reactivity to antisera as complexes I and II illustrated in Fig. 2. Complex II was not formed with the Th1 extract. A new complex (III), specific for Ag stimulation and of mobility intermediate between complexes I and II, was formed with both Th1 and Th2 extracts. Complex III was more distinct with the 24-h extracts (Fig. 3A, left panel, compare lane 4 with lane 1 and lane 10 with lane 7). Complex III, formed with both the Th1 extracts (lanes 3 and 6) and the Th2 extracts (lanes 9 and 12), was inhibited by the anti-GATA-4 antiserum. However, the anti-GATA-3 Ab did not react with any of the complexes formed with the Th1 extracts (lanes 2 and 5). By contrast, in our EMSAs with the Th2 extracts, complex I was slightly inhibited by the anti-GATA-3 Ab, and complex II, only formed with Th2 nuclear proteins, was supershifted by the anti-GATA-3 Ab (lanes 8 and 11). Essentially identical data were obtained in EMSAs with nuclear extracts from Ag-stimulated D10 and C19 cells (data not shown). We have also explored GATA-3 binding activity in another Th1-type clone, A.E7 (35). No IL-5 mRNA was detected in Northern blot analyses of RNA prepared from A.E7 cells that were stimulated for 24 h with Ag (data not shown). Also, nuclear extracts prepared from Ag-induced A.E7 cells had no detectable GATA-3 binding activity (data not shown). We do not know the exact composition of the different complexes that are formed with Th2 nuclear extracts and the IL-5 GATA site. Complex II could represent a higher order form (dimer or tetramer) of complex I that may contain a monomer or dimer of GATA-3. The oligonucleotide competition experiments suggest that complex I binds to the GATA site with a lower affinity than complex II. Also, formation of complex II was consistently more sensitive to the anti-GATA-3 antibody than formation of complex I, suggesting that the epitope recognized by the anti-GATA-3 monoclonal antibody in the GATA-3 protein is more accessible in complex II. Taken together, it appears that activated Th1 cells lack GATA-3 DNA binding activity. The significance of the reactivity of complex III to the anti-GATA-4 antiserum is unclear at the present time since the expression of GATA-4 has only been described in the heart, intestines, and gonads (28, 29).


Fig. 3. Murine CD4+ Th2 cells, but not Th1 cells, express GATA-3 DNA binding activity. Th1 and Th2 cells were generated by differentiation of naive CD4+ T cells isolated from TCR transgenic mice as described under "Experimental Procedures." Cultures were maintained for 3 days. The cells were further stimulated at the end of this period with fresh mitomycin C-treated and T cell-depleted APCs and Ag for 8 or 24 h, and nuclear extracts were prepared. The extracts were used in EMSAs as described in the legend to Fig. 2. mut., mutant.
[View Larger Version of this Image (67K GIF file)]

We also investigated AP-1 binding activity with the extracts of both cells. In both cases, a major complex (Fig. 3A, right panel, shown by an arrow) and a minor complex were obtained. Nuclear extracts from Th1 cells that were stimulated for 8 h contained less AP-1 binding activity than extracts from similarly treated Th2 cells (Fig. 3A, right panel, compare lanes 1 and 7). However, the AP-1 binding activities of the 24-h extracts from the two cell types were equivalent (lanes 2 and 8). The supershift/inhibition studies again revealed the presence of JunB and JunD in the complexes formed with both Th1 (lanes 4 and 5) and Th2 (lanes 10 and 11) extracts. The anti-c-Jun Ab also partially inhibited formation of the Th1-specific complex (lane 3). However, since there was residual DNA binding activity in the complexes generated with both Th1 and Th2 extracts that was not abolished by any of these antisera, we cannot rule out the possibility that related proteins are also present in the complexes. The residual binding appears not be due to any binding to the Elf-1-like site within the CLE0 element since an oligonucleotide containing a mutation in the Elf-1 site was able to efficiently compete for binding to the CLE0 element using nuclear extracts of either clones or cultured cells (data not shown). To ensure that the lower intensity of the 8-h Th1 complex was not related to technique, we tested the extracts for binding to a probe containing a binding site for CREB. As illustrated in Fig. 3B, the CREB binding activity was equivalent in the two nuclear extracts, ruling out a general deficiency of proteins in the Th1 cell extracts. Immunoprecipitation experiments performed with cell extracts prepared from metabolically labeled cells have shown that the lack of GATA-3 DNA binding activity in Th1 cells is due to the absence of GATA-3 protein in these cells (data not shown).

Collectively, our data indicate that Th2 cells contain GATA-3 protein in a constitutive fashion, which increases upon stimulation of the cells by Ag or cAMP. In contrast, Th1 cells express very little/no GATA-3 at the basal level, which reproducibly decreases upon stimulation of the cells.

Transactivation of the IL-5 Promoter by Ectopic Expression of GATA-3

We reasoned that if GATA-3 controls tissue-specific expression of IL-5, then ectopic expression of GATA-3 in non-GATA-3-expressing cells would allow IL-5 gene expression in the cells. Upon stimulation, neither Th1 clones nor the cervical carcinoma cell line HeLa expresses IL-5 or GATA-3 protein, but do express AP-1 proteins. To test whether expression of GATA-3 RNA would make these cells permissive for IL-5 promoter activation, Th1 cells (A.E7) (Fig. 4A) or HeLa cells (Fig. 4B) were transfected with the 1.7-kb IL-5 promoter-reporter construct together with a vector for either sense or antisense (negative control) GATA-3 RNA expression. Although it has been difficult to sustain high levels of GATA-3 RNA expression in non-Th2 cells (especially in Th1 cells) by transient transfection methods, in both A.E7 cells and HeLa cells, expression of GATA-3 sense RNA, but not antisense RNA, resulted in 4-6-fold activation of the 1.7-kb IL-5 promoter upon stimulation with Bt2cAMP and PMA. The absence of induction of the promoter without stimulation of the cells suggests the need for binding of inducible proteins to other DNA sites, most likely the AP-1 site within the CLE0 element, for the induction of the promoter. However, stimulation of the cells alone without GATA-3 expression did not activate the IL-5 promoter. A minimal promoter containing IL-5 DNA sequences between -76 and +24, which include the GATA site between -70 and -60 and the CLE0 element, but no other identifiable transcription factor-binding sites, was also similarly activated by coexpression of the sense (but not the antisense) GATA-3 expression vector in both cell types (data not shown). Also, as shown in Fig. 4B, overexpression of the p65 subunit of NF-kappa B, a potent transactivator, failed to up-regulate IL-5 promoter activity in HeLa cells, demonstrating the specificity of GATA-3 in this experimental setting. Thus, although Th1 cells and HeLa cells express several transcription factors, the DNA-binding sites for a few of which can be identified in the 1.7-kb IL-5 promoter, these cells can only activate the promoter in the presence of GATA-3. This strongly suggests that GATA-3 serves the role of a tissue-specific regulator of IL-5 gene expression. Our studies do not rule out the involvement of additional control elements that might be involved in the overall enhancement of transcription of the IL-5 gene in Th2 cells.


Fig. 4. Activation of the IL-5 promoter in A.E7 cells and HeLa cells by ectopic expression of GATA-3. A, A.E7 cells were transfected by electroporation with 15 µg of DNA (5 µg of reporter plasmid, 2 µg of cytomegalovirus-beta -galactosidase plasmid, and carrier plasmid with or without 5 µg of murine GATA-3 sense or antisense expression plasmid). Cells were either left uninduced or were induced with Bt2cAMP + PMA as described in the legend to Fig. 1. Cells were harvested and assayed for luciferase and beta -galactosidase activity. In each case, the results represent an average of two independent experiments. B, monolayer cultures of HeLa cells in 60-mm plates were transfected by the calcium phosphate coprecipitation procedure as described previously (33, 34) with 10 µg of DNA (2 µg of reporter plasmid, 2 µg of an expression plasmid for murine GATA-3 sense or antisense RNA expression or p65 RNA expression, 1 µg of cytomegalovirus-beta -galactosidase, and 5 µg of carrier plasmid). 16 h after transfection, the cells were washed and were either left unstimulated in serum-free medium or stimulated with Bt2cAMP + PMA as described in the legend to Fig. 1. Cells were harvested 6 h after stimulation and assayed for luciferase and beta -galactosidase activity. Shown is a representative experiment of three with <3% variation between experiments. S, sense; AS, antisense; wt, wild-type.
[View Larger Version of this Image (21K GIF file)]

Kinetics of GATA-3 RNA Expression during Development of Th1 and Th2 Cells from Naive Spleen Thp Cells

To determine whether the difference in GATA-3 activity between Th1 and Th2 cells reflected a preferential up-regulation of GATA-3 gene expression in Th2 cells in the course of primary stimulation, naive spleen Thp cells were allowed to differentiate along a Th1 or Th2 pathway by treatment with Ag and the appropriate cytokines and anti-cytokine antibodies. Northern blot analysis of GATA-3 mRNA expression was carried out in differentiating cells harvested at different time points after stimulation. Identification of differentiated cells as predominantly Th1 or Th2 populations was carried out by ELISA for IL-4, IL-5, and IFN-gamma protein in culture supernatants after secondary stimulation with Ag. As shown in Fig. 5A, naive Thp cells were found to express low levels of GATA-3 mRNA. In cells differentiated along a Th2 pathway, there was a substantial increase in GATA-3 mRNA at 24 h after stimulation, which continued to increase until 48 h and reached a plateau thereafter. In contrast, no such induction was seen in Th1 cells, and there was actually a decrease in the level of GATA-3 message at 24 h after stimulation, which reached a minimum at 48 h post-stimulation, by which time a commitment of the developing cells along the Th1/Th2 lineage has already occurred (36). The loading of RNA was equivalent for the two sets of cells as evident from hybridization of the same blot with glyceraldehyde-3-phosphate dehydrogenase (data not shown). The cells were restimulated after day 5 with fresh T cell-depleted and mitomycin C-treated APCs and Ag for 48 h, and cytokine levels in the culture supernatants were evaluated by ELISA (Fig. 5B). Taken together, our experiments show that differentiation along the Th2 pathway results in a substantial increase in GATA-3 gene expression, whereas that along the Th1 pathway leads to a decrease in GATA-3 gene expression.


Fig. 5. Up-regulation of GATA-3 gene expression in developing Th2 cells and down-regulation in developing Th1 cells. Naive CD4+ T cells isolated from the spleens of DO11.10 mice were allowed to differentiate along the Th1 or Th2 pathway as described under "Experimental Procedures." A, expression of GATA-3 mRNA in developing Th1 and Th2 cells. Total RNA was isolated from cells harvested at the indicated time points and analyzed by Northern blotting techniques. B, cytokine production by Th1 and Th2 cell populations. After 5 days in culture, cells were washed and restimulated with fresh T cell-depleted and mitomycin C-treated APCs and Ag for 48 h. Culture supernatants were assayed for cytokine production by ELISA. mGATA-3, murine GATA-3.
[View Larger Version of this Image (29K GIF file)]


DISCUSSION

This study establishes a new role for the transcription factor GATA-3 as a determinative factor in Th2-specific IL-5 gene expression. The DNA sequence of the GATA-3 double site is identical in the human and murine IL-5 genes. GATA-3 belongs to the GATA family of transcription factors that bind to the WGATAR (W = A/T; R = A/G) DNA sequence through a highly conserved C4 zinc finger domain. Six members (GATA-1 to GATA-6) of this family have been identified in avians, with homologues in mammals and amphibians (29). Based on their expression profile, the GATA proteins may be classified functionally as hemopoietic (GATA-1 to GATA-3) or non-hemopoietic (GATA-4 to GATA-6), and this classification is also valid from structural considerations (29). GATA-3 is expressed primarily in T lymphocytes and in the embryonic brain. Functionally important GATA-3-binding sites have been identified in T cell receptor genes and the CD8 gene (37-41). However, in the case of most of these genes, a mutation of the GATA site in the context of a large promoter fragment fails to inhibit the activity of the respective promoters, suggesting redundancy in their enhancers (42). Knock out of the GATA-3 gene in mice results in embryonic death on day 12, with a failure of fetal hematopoiesis and defects in the central nervous system (43). Recently, GATA-3 was shown to be an essential component in the earliest steps of T cell development in the thymus using antisense oligonucleotides for GATA-3 in fetal thymus organ cultures (44) and using embryonic stem cells containing homozygous mutations in the GATA-3 gene and the RAG-2 gene (45). Collectively, these studies indicate that GATA-3 is not a functionally redundant GATA family protein.

Activation of IL-5 gene expression by both Ag and a combination of Bt2cAMP and PMA requires the GATA site and the AP-1 site, but not the NF-AT site, in the IL-5 promoter, and both stimuli trigger similar binding activities at the corresponding sites. This suggests that activation of the TCR·CD3 complex might involve stimulation of cAMP-dependent signaling pathways in the cell. Indeed, engagement of the TCR·CD3 complex by foreign Ag or by anti-CD3 monoclonal antibodies has been shown to result in elevation of intracellular cAMP levels in T cells that is associated with the onset of DNA synthesis in the cells (46). Whereas several studies have triggered a heightened appreciation for the role of cAMP as an immunomodulator (47), there is a need to identify a molecular basis for the differential effects of cAMP in the regulation of gene expression at specific stages of activation or of differentiation of T cells following stimulation via the TCR·CD3 complex. Our studies suggest that specific transcription factors such as GATA-3 may play a role in the differential effects of cAMP on immune responses during T cell activation and/or differentiation.

There are two possible explanations for why the IL-5 gene is not expressed in uninduced Th2 cells despite high basal levels of GATA-3. First, the activation by GATA-3 requires post-translational modification of the protein; indeed, we have identified several potential protein kinase A and protein kinase C phosphorylation sites in the GATA-3 protein. Second, IL-5 promoter activation also requires binding of proteins to the AP-1 site within the CLE0 element (also identical in the murine and human genes), which is only achieved once the cells are activated. Among the AP-1 family of proteins, we have detected only JunB and JunD in the complex that forms with the CLE0 element using both Th1 and Th2 extracts. Induction of JunB transcription by cAMP has been previously described (48, 49). Similar to our findings, in one of these studies (45), anti-Jun or anti-Fos antibodies did not completely supershift/inhibit cAMP-induced complexes formed with an AP-1 site. It is possible that Ag and Bt2cAMP induce the formation of heterodimers of JunB/JunD with an as yet unidentified Jun-related protein, and the resulting complex exhibits a lower affinity for anti-JunB and anti-JunD antibodies. Alternatively, cAMP induces the formation of heteromeric complexes between JunB/JunD and other related bZIP proteins. Although the AP-1 proteins recognize the TPA response element, which differs from the cAMP response element by only 1 base pair, the distinction between the TPA response element/cAMP response element and AP-1/CREB is, however, not absolute. TPA response element- and cAMP response element-related sequences have now been identified that are recognized by both groups of proteins, and heterodimers between some members of the CREB/ATF family and JunB, JunD, and c-Jun have been reported (50). The complex formed with the AP-1 site in the IL-5 promoter might therefore be composed of heteromeric complexes between JunB/JunD and CREB/ATF proteins.

Recently, the proximal region of the IL-4 promoter was shown to interact with c-Maf, the product of the proto-oncogene c-maf, which is expressed in Th2 cells, but not in Th1 cells (51). The results of multiple studies by different investigators indicate that several transcription factors binding to the promoter proximal region or to regions outside of the 800-bp IL-4 promoter may coordinately regulate overall IL-4 promoter activity, the Th2 specificity being conferred by proteins such as c-Maf and probably also NF-IL-6, which, like c-Maf, is expressed in Th2 cells, but not in Th1 cells (52-54). The IL-4 promoter also contains a double GATA site between -274 and -264 (which is very similar to the one in the IL-5 promoter located between -70 and -60) and a single GATA motif between -112 and -107. It will be interesting to examine the activity of a large promoter fragment of the IL-4 gene containing mutations in the GATA sites and the effect of ectopically expressed GATA-3 on wild-type promoter activity.

At the RNA level, the expression of GATA-3 in Th1 cells generated by in vitro differentiation of naive CD4+ T cells is at one-twentieth to one-thirtieth the level detected in Th2 cells. In EMSAs, we have consistently detected a very low level of GATA-3 binding activity in the resting Th1 clones C19 and A.E7. Also, this activity reproducibly decreases upon stimulation of the cells, whether by Ag or by a combination of Bt2cAMP and PMA. This decrease in binding activity may result from a higher rate of turnover of GATA-3 RNA in activated Th1 cells as seen in developing Th1 cells during a primary stimulation of naive primary CD4+ T cells. This effect also appears to be specifically targeted at GATA-3 since the same extracts used at half the amounts displayed significant inducible binding activity at the AP-1 site. Our data suggest that the decrease in GATA-3 levels in stimulated Th1 cells is another mechanism that ensures the specific profile of cytokine production by Th1 and Th2 cells.

Studies of asthma in both human and animal models are consistent with the concept that airway inflammation, a characteristic feature of asthma, requires the presence of activated Th2 cells. Among the Th2 cytokines, IL-5 is key to the eosinophilia typically associated with the disease, being intimately involved with eosinophil differentiation, proliferation, and survival. Eosinophilia is also a feature of a number of other pathological conditions such as idiopathic hypereosinophilic syndrome, parasitic infections, and allergies (12). If asthma results from an abnormal skewing of the immune response in the lung toward the generation of Th2 cells, then understanding the reversal of this process is critical. One important target for the development of anti-asthma therapeutics is therefore inhibition of IL-5 gene expression in T cells. GATA-3 is an especially attractive target in this regard since its expression is limited to T cells in the adult, and therefore, any GATA-3-specific antagonist can be expected to have minimum side effects. Another important question that arises from this study is whether forced expression of GATA-3 in Th1 cells would aid in altering Th1/Th2 ratios. This is important in controlling Th1-driven pathologies, which include autoimmune diseases such as insulin-dependent diabetes mellitus. In a recent study investigating the role of CD4+ Th1 and Th2 cells in the induction of diabetes mellitus, Th1 cells promoted disease, whereas Th2 cells did not (55).

Thus, this study identifies differential expression of the GATA-3 gene in Th1 and Th2 cells and relates it to the production of a cytokine of significant biomedical importance. This raises intriguing questions about the role of GATA-3 in determining the balance between Th1 and Th2 subsets in immune responses and disease states. Additionally, identification of GATA-3 as a critical component in IL-5 gene expression raises possibilities for the therapy of asthma and allergic diseases and idiopathic hypereosinophilic syndrome via blockade of GATA-3 activation.


FOOTNOTES

*   This work was supported by National Institutes of Health Grant AI31137 and Specialized Center for Research in Asthma Grant P50 HL56389 (to A. R.) and National Institutes of Health Grant HL52014 (to P. R.).The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
par    To whom correspondence should be addressed: Dept. of Internal Medicine, Pulmonary and Critical Care Section, Yale University School of Medicine, P. O. Box 208057, 333 Cedar St., New Haven, CT 06520. Tel.: 203-737-2705; Fax: 203-785-3826; E-mail: Anuradha.Ray{at}qm.yale.edu.
1   The abbreviations used are: IL, interleukin; IFN-gamma , interferon-gamma ; Bt2cAMP, dibutyryl cyclic AMP; EMSAs, electrophoretic mobility shift assays; bp, base pair; kb, kilobase; TCR, T cell receptor; APCs, antigen-presenting cells; Ag, antigen; ELISA, enzyme-linked immunosorbent assay; PMA, phorbol 12-myristate 13-acetate; Ab, antibody; TPA, 12-O-tetradecanoylphorbol-13-acetate.
2   D.-H. Zhang and A. Ray, unpublished observations.

ACKNOWLEDGEMENTS

We thank T. Honjo for the gift of the plasmid p4k-pUC18 containing IL-5 promoter sequences, J. D. Engel for the plasmid containing the murine GATA-3 cDNA and GATA-3 expression vectors, K. M. Murphy for the DO11.10 TCR transgenic mice, D. Wilson for the anti-GATA-4 antiserum, and A. Marinov for excellent technical assistance with ELISAs.


REFERENCES

  1. Bottomly, K. (1988) Immunol. Today 9, 268-273 [CrossRef][Medline] [Order article via Infotrieve]
  2. Mossman, T. R., and Coffman, R. L. (1989) Annu. Rev. Immunol. 7, 145-173 [CrossRef][Medline] [Order article via Infotrieve]
  3. Janeway, C. A., Jr., and Bottomly, K. (1994) Cell 76, 275-285 [Medline] [Order article via Infotrieve]
  4. Abbas, A. K., Murphy, K. M., and Sher, A. (1996) Nature 383, 787-793 [CrossRef][Medline] [Order article via Infotrieve]
  5. Walker, C., Virchow, J., Bruijnzeel, P. L. B., and Blaser, K. (1991) J. Immunol. 1990, 1829-1835
  6. Hamid, Q., Azzawi, M., Ying, S., Moqbel, R., Wardlaw, A. J., Corrigan, C. J., Bradley, B., Durham, S. R., Collins, J. V., Jeffery, P. K., Quint, D. J., and Kay, A. B. (1991) J. Clin. Invest. 87, 1541-1546 [Medline] [Order article via Infotrieve]
  7. Robinson, D. S., Hamid, Q., Ying, S., Tsicopoulos, A., Barkans, J., Bentley, A. M., Corrigan, C., Durham, S. R., and Kay, A. B. (1992) N. Engl. J. Med. 326, 298-304 [Abstract]
  8. Krishnaswamy, G., Liu, M. C., Su, S., Kumai, M., Xiao, H., Marsh, D. G., and Huang, S. (1993) Am. J. Respir. Cell Mol. Biol. 9, 279-286 [Medline] [Order article via Infotrieve]
  9. Martin, L. B., Kita, H., Leiferman, K. M., and Gleich, G. J. (1996) Int. Arch. Allergy Immunol. 109, 207-215 [Medline] [Order article via Infotrieve]
  10. Frick, W. E., Sedgwick, J. B., and Busse, W. W. (1989) Am. Rev. Respir. Dis. 139, 1401-1406 [Medline] [Order article via Infotrieve]
  11. Bousquet, J., Chanez, P., Lacoste, J. Y., Barneon, G., Ghavanian, N., Enander, I., Venge, P., Ahlstedt, S., Simony-Lafontaine, J., Godard, P., and Michel, F. B. (1990) N. Engl. J. Med. 323, 1033-1039 [Abstract]
  12. Sanderson, C. J. (1992) Blood 79, 3101-3109 [Medline] [Order article via Infotrieve]
  13. Kamei, T., Ozaki, T., Kawaji, K., Banno, K., Sano, T., Azuma, M., and Ogura, T. (1993) Am. J. Respir. Cell Mol. Biol. 9, 378-385 [Medline] [Order article via Infotrieve]
  14. Gulbenkian, A. R., Egan, R. W., Fernandez, X., Jones, H., Kreutner, W., Kung, T., Payvandi, F., Sullivan, L., Zurcher, J. A., and Watnick, A. S. (1991) Am. Rev. Respir. Dis. 146, 263-265
  15. Nakajima, H., Iwamoto, I., Tomoe, S., Matsumura, R., Tomioka, H., Takatsu, K., and Yoshida, S. (1992) Am. Rev. Respir. Dis. 146, 374-377 [Medline] [Order article via Infotrieve]
  16. Foster, P. S., Hogan, S. P., Ramsay, A. J., Matthaei, K. I., and Young, I. G. (1996) J. Exp. Med. 183, 195-201 [Abstract]
  17. Kopf, M., Brombacher, F., Hodgkin, P. D., Ramsay, A. J., Milbourne, E. A., Dai, W. J., Ovington, K. S., Behm, C. A., Kohler, G., Young, I. G., and Matthaei, K. I. (1996) Immunity 4, 15-24 [Medline] [Order article via Infotrieve]
  18. Plaut, M., Pierce, J. H., Watson, C. J., Hanley-Hyde, J., Nordan, R. P., and Paul, W. E. (1989) Nature 339, 64-67 [CrossRef][Medline] [Order article via Infotrieve]
  19. Broide, D. H., Paine, M. M., and Firestein, G. S. (1992) J. Clin. Invest. 90, 1414-1424 [Medline] [Order article via Infotrieve]
  20. Bohjanen, P. R., Okajima, M., and Hodes, R. J. (1990) Proc. Natl. Acad. Sci. U. S. A. 87, 5283-5287 [Abstract]
  21. Lee, H. J., Koyano-Nakagawa, N., Naito, Y., Nishida, J., Arai, N., Arai, K., and Yokota, T. (1993) J. Immunol. 151, 6135-6142 [Abstract/Free Full Text]
  22. van der Pouw Kraan, T. C. T. M., Boeije, L. C. M., Smeenk, R. J. T., Wijdenes, J., and Aarden, L. A. (1995) J. Exp. Med. 181, 775-779 [Abstract]
  23. Munoz, E., Zubiaga, A. M., Merrow, M., Sauter, N. P., and Huber, B. T. (1990) J. Exp. Med. 1990, 95-103
  24. Snijdewint, F. G. M., Kalinski, P., Wierenga, E. A., Bos, J. D., and Kapsenberg, M. L. (1993) J. Immunol. 150, 5321-5329 [Abstract/Free Full Text]
  25. Siegel, M. D., Zhang, D.-H., Ray, P., and Ray, A. (1995) J. Biol. Chem. 270, 24548-24555 [Abstract/Free Full Text]
  26. Yamagata, T., Nishida, J., Sakai, R., Tanaka, T., Honda, H., Hirano, N., Mano, H., Yazaki, Y., and Hirai, H. (1995) Mol. Cell. Biol. 15, 3830-3839 [Abstract]
  27. Noma, T., Nakakubo, H., Sugita, M., Kumagai, S., Maeda, M., Shimizu, A., and Honjo, T. (1989) J. Exp. Med. 169, 1853-1858 [Abstract]
  28. Arceci, R. J., King, A. A. J., Simon, M. C., Orkin, S. H., and Wilson, D. B. (1993) Mol. Cell. Biol. 13, 2235-2246 [Abstract]
  29. Laverriere, A. C., MacNeill, C., Mueller, C., Poelmann, R. E., Burch, J. B. E., and Evans, T. (1994) J. Biol. Chem. 269, 23177-23184 [Abstract/Free Full Text]
  30. Prieschl, E. E., Goouilleux-Gruart, V., Walker, C., Harrer, N. E., and Baumruker, T. (1995) J. Immunol. 154, 6112-6119 [Abstract/Free Full Text]
  31. Kaye, J., Porcelli, S., Tite, J., Jones, B., and Janeway, J. C. A. (1983) J. Exp. Med. 158, 836-856 [Abstract]
  32. Baron, J. L., Madri, J. A., Ruddle, N. H., Hashim, G., and Janeway, J. C. A. (1993) J. Exp. Med. 177, 57-68 [Abstract]
  33. Ray, A., LaForge, K. S., and Sehgal, P. B. (1991) Proc. Natl. Acad. Sci. U. S. A. 88, 7086-7090 [Abstract]
  34. Ray, A., and Prefontaine, K. E. (1994) Proc. Natl. Acad. Sci. U. S. A. 86, 752-756
  35. Beverly, B., Kang, S. M., Lenardo, M. J., and Schwartz, R. H. (1992) Int. Immunol. 4, 661-671 [Abstract]
  36. Nakamura, T., Kamogawa, Y., Bottomly, K., and Flavell, R. A. (1997) J. Immunol. 158, 1085-1094 [Abstract]
  37. Ho, I.-C., Vorhees, P., Marin, N., Oakley, B. K., Tsai, S.-F., Orkin, S. H., and Leiden, J. M. (1991) EMBO J. 10, 1187-1192 [Abstract]
  38. Ko, L. J., Yamamoto, M., Leonard, M. W., George, K. M., and Engel, J. D. (1991) Mol. Cell. Biol. 11, 2778-2784 [Medline] [Order article via Infotrieve]
  39. Marine, J., and Winoto, A. (1991) Proc. Natl. Acad. Sci. U. S. A. 88, 7284-7288 [Abstract]
  40. Hambor, J. E., Mennone, J., Coon, M. E., Hanke, J. H., and Kavathas, P. (1993) Mol. Cell. Biol. 13, 7056-7070 [Abstract]
  41. Henderson, A. J., McDougall, S., Leiden, J., and Calame, K. L. (1994) Mol. Cell. Biol. 14, 4286-4294 [Abstract]
  42. Smith, V. M., Lee, P. P., Szychowski, S., and Winoto, A. (1995) J. Biol. Chem. 270, 1515-1520 [Abstract/Free Full Text]
  43. Pandolfi, P. P., Roth, M. E., Karis, A., Leonard, M. W., Dzierzak, E., Grosveld, F. G., Engel, J. D., and Lindenbaum, M. H. (1995) Nat. Genet. 11, 40-44 [Medline] [Order article via Infotrieve]
  44. Hattori, N., Kawamoto, H., Fujimoto, S., Kuno, K., and Katsura, Y. (1996) J. Exp. Med. 184, 1137-1147 [Abstract]
  45. Ting, C.-N., Olson, M. C., Barton, K. P., and Leiden, J. M. (1996) Nature 384, 474-478 [CrossRef][Medline] [Order article via Infotrieve]
  46. Feuerstein, N., Firestein, R., Aiyar, N., He, X., Murasko, D., and Cristofalo, V. (1996) J. Immunol. 156, 4582-4593 [Abstract/Free Full Text]
  47. Haraguchi, S., Good, R. A., and Day, N. K. (1995) Immunol. Today 16, 595-603 [CrossRef][Medline] [Order article via Infotrieve]
  48. Tamir, A., and Isakov, N. (1994) J. Immunol. 152, 3391-3399 [Abstract/Free Full Text]
  49. Amato, S. F., Nakajima, K., Hirano, T., and Chiles, T. C. (1996) J. Immunol. 157, 146-155 [Abstract]
  50. Sassone-Corsi, P., Ransone, L. J., and Verma, I. (1990) Oncogene 5, 427-431 [Medline] [Order article via Infotrieve]
  51. Ho, I.-C., Hodge, M. R., Rooney, J. W., and Glimcher, L. H. (1996) Cell 85, 973-983 [Medline] [Order article via Infotrieve]
  52. Davydov, I. V., Krammer, P. H., and Li-Weber, M. (1995) J. Immunol. 155, 5273-5279 [Abstract]
  53. Li-Weber, M., Salgame, M., Hu, C., Davydov, I. V., and Krammer, P. H. (1997) J. Immunol. 158, 1194-1200 [Abstract]
  54. Wenner, C. A., Szabo, S. J., and Murphy, K. M. (1997) J. Immunol. 158, 765-773 [Abstract]
  55. Katz, J. D., Benoist, C., and Mathis, D. (1995) Science 268, 1185-1188 [Medline] [Order article via Infotrieve]

©1997 by The American Society for Biochemistry and Molecular Biology, Inc.