Treatment of Allergic Airway Inflammation and Hyperresponsiveness by Antisense-Induced Local Blockade of Gata-3 Expression

Allergic asthma is a chronic pulmonary disease associated with bronchoconstriction and chronic inflammation of the airways 1,2,3. In allergic asthma, there is a polarization of T lymphocyte responses and enhanced secretion of cytokines involved in regulation of IgE, mast cells, basophils, and eosinophils, ultimately leading to inflammation and disease 4. The proinflammatory cytokines produced by T cells contribute to the initiation and perpetuation of allergic asthma 1,2,3. In particular, it has been shown that allergic airway inflammation in patients with asthma and animal models of allergic airway inflammation is associated with increased Th2 (IL-4, IL-5, IL-13) cytokine production 5,6,7,8. In this regard, IL-4 has been postulated to be critical for the development of Th2 T cells in asthmatic airways and the reduced expression of the IL-12 receptor β2 chain on lung Th1 cells 6,9,10. Furthermore, it is believed to induce IgE isotype switching in B cells in asthma and promotes goblet cell metaplasia, mucus hypersecretion, and the recruitment of eosinophils by upregulating vascular cell adhesion molecule 1 expression in pulmonary endothelial cells 10. The importance of IL-4 is also underlined by the observation that inhalation of IL-4 causes the development of sputum eosinophilia and increased responsiveness of the airways and the finding that neutralizing antibodies to IL-4 are therapeutically effective for asthma 1,11,12.

GATA-3 is a pleiotropic transcription factor of the C4 zinc finger family expressed in T cells, mast cells, eosinophils, basophils, and embryonic brain and kidney that binds to a 5′-WGATAR-3′ consensus DNA sequence 13,14,15,16. GATA-3 has been shown to be essential for the development of the earliest T cell progenitor cells 17. Furthermore, GATA-3 was found to be selectively expressed in Th2 but not in Th1 cells and to play an important role in cytokine gene expression in T cells 18,19,20,21. In particular, GATA-3 is important for the expression of IL-5 in T cells by transactivation of the IL-5 promoter together with Ets-1 and Ets-2 proteins 20,21. Furthermore, GATA-3 weakly transactivates the IL-4 promoter in T cells 22. These functions of GATA-3 on Th2 cytokine gene promoters can be suppressed by repressor of GATA (ROG), a recently cloned lymphoid-specific repressor of GATA-3–induced transactivation 23. Finally, ectopic expression of GATA-3 in developing Th1 cells leads to upregulation of IL-4 and IL-5 and downregulation of IFN-γ. The latter effect appears to be partly due to downregulation of the IL-12 receptor β2 chain 18,19,24.

Activation studies in retroviral infected T cells have shown that the activation of GATA-3 in T cells occurs upon activation of the IL-4/signal transducer and activator of transcription (STAT)-6 signaling pathway 25, suggesting that the exposure of naive T cells to IL-4 may be a very early event that induces GATA-3 activation and Th2 cell differentiation. However, GATA-3 can fully reconstitute Th2 development in STAT-6–deficient T cells, suggesting that it is a master switch both in STAT-6–dependent and –independent Th2 development 26. Finally, GATA-3 has been shown to exert STAT-6–independent autoactivation, creating a feedback pathway stabilizing Th2 commitment 26.

Based on the above data, it was of particular interest to analyze the expression and functional role of GATA-3 in patients with atopic asthma. Indeed, in a recent study Nakamura and co-workers 27 showed an increased expression of GATA-3 mRNA in asthmatic airways, suggesting that GATA-3 may be involved in the regulation of Th2 cytokine responses in asthma. Based on this observation, we analyzed in the present study the role of GATA-3 in a mouse model of asthma 28 using an antisense DNA strategy. We observed increased local production of GATA-3 in this model and demonstrate that local treatment with GATA-3 antisense oligonucleotides could be successfully used to downregulate Th2-induced pulmonary inflammation and airway hyperresponsiveness.

The following reagents were used: water-soluble dexamethasone (DEX) and aluminum potassium sulfate (alum; Sigma-Aldrich); crystalline OVA (Pierce Chemical Co.); biotinylated monoclonal anti–mouse IL-4 (BD PharMingen), monoclonal and polyclonal goat antibody anti–GATA-3, (Santa Cruz Biotechnology, Inc.) and monoclonal antiactin (Santa Cruz Biotechnology, Inc.); horseradish peroxidase–conjugated anti–rabbit Ig and anti–mouse Ig (Amersham Pharmacia Biotech), the ECL Western blotting system (Amersham Pharmacia Biotech); biotinylated anti–goat IgG (Vector Laboratories); Cy3/AMCA Streptavidin conjugated (Dianova); and phosphorothioate oligonucleotides (Roth GmbH and GIBCO BRL).

Oligonucleotides were synthesized with a phosphorothioate backbone to improve resistance to endonucleases. The antisense oligonucleotides consisted of 18-mer analogues to the 5′ end of the murine GATA-3 sequence which spans the translation initiation site. This sequence had no C plus G dinucleotide or a quadruple G sequence element that are known to cause unspecific effects. In addition, several control (mismatched and nonsense) oligonucleotides were prepared. The nonsense oligonucleotides contained the same nucleotide composition as the antisense oligonucleotides.

The sequences were as follows (mismatches are underlined): GATA-3 antisense DNA (with or without FITC), 5′-AGT CAC CTC CAT GTC CTC-3′; GATA-3 nonsense DNA, 5′-CTA TGT CAT CCG CTC CAC-3′; GATA-3 mismatched DNA, 5′-AGC CAC CTA CAT TTC CTA-3′; and GATA-3 mismatched DNA 2, 5′-AGC CAC CTA GGC ATC CTC-3′.

Naive splenic CD4⁺ T cells (purity > 97%) were obtained using monoclonal antibodies to CD4 coupled to magnetic beads (Miltenyi Biotec) according to the protocol provided by the manufacturer. Cells were cultured on 10 μg/ml anti-CD3–coated wells in the presence of 2 μg/ml anti-CD28, 100 U/ml rmIL-2, and 1,000 U/ml rmIL-4 for 3 d. Cells were then cultured for an additional 2 d in the presence of rmIL-2 and IL-4 only. During the last 12 h cells were challenged with 50 ng/ml PMA and 1 μg/ml ionomycin. For coincubation experiments phosphorothioate oligonucleotides (12–15 μM) were preincubated for 45 min with Lipofectamine (Life Technologies/GIBCO BRL). At days 2 and 4 the oligonucleotides were added to the cell cultures. Cells treated with FITC-conjugated antisense DNA were washed, fixed in 4% paraformaldehyde, and analyzed by FACS^® analysis for uptake efficiency. Cell viability was assessed by trypan blue staining and supernatants were analyzed for IL-4 and IL-9 production by ELISA.

To enrich for antisense-transfected T cells, T cells were cotransfected on day 4 with the pMACS K^k.II vector (Miltenyi Biotec) expressing a truncated H2-K^k molecule (that is not expressed in BALB/c mice) and FITC-labeled or unlabeled phosphorothioate oligonucleotides. Cells were harvested 48 h after cotransfection, resuspended in PBE (PBS with 5 mM EDTA) buffer, and incubated with MACSelect K^k microbeads (Miltenyi Biotec) followed by magnetic separation. FITC-positive cells were visualized by fluorescence microscopy or confocal laser microscopy using a Bio-Rad Laboratories MRC-1024/MP microscope. To assess enrichment of antisense-transfected cells, T cell nuclei were counterstained with DAPI (Vector Laboratories) and the percentage of FITC-positive cells was determined by counting 12 randomly selected high power fields (HPFs) in three samples per condition by fluorescence microscopy. Cytokine production in magnetic cell sorting (MACS)-selected T cells was assessed after 18 h stimulation with 50 ng/ml PMA plus 1 μg/ml ionomycin by ELISA.

Female BALB/c mice (6–8 wk of age) received an intraperitoneal injection of 100 μg/ml OVA (0.2 ml of 500 μg/ml in normal saline) complexed with alum on days 0 and 14. On days 25, 26, and 27, the mice underwent anesthesia with 0.2 ml of avertine intraperitoneally (1 mg/ml of tryplomethan in amylalcohol; 2.5% in PBS) before receiving 50 μg OVA in saline intranasally. Control animals received saline intraperitoneally with alum on days 0 and 14 and 50 μl of saline intravenously on days 25, 26, and 27.

To assess the effect of antisense GATA-3 treatment in airway inflammation, antisense or control oligonucleotides were given by intranasal administration (200 μg/treatment) at days 24, 25, 26, and 27 (30 min before intravenous injection of OVA) to OVA-sensitized mice. To localize antisense oligonucleotides in vivo, we performed the last intranasal administration with FITC-labeled antisense oligonucleotides.

To compare the efficiency of antisense GATA-3 treatment to the classical treatment with glucocorticoids, a group of OVA-treated mice was given 1 mg (50 μl vol) of water soluble DEX intravenously in saline on days 24, 25, 26, and 27.

24 h after the last intranasal challenge with either OVA or saline at day 27, bronchioalveolar lavage (BAL) of the right lung was performed after tying off the left lung at the mainstem bronchus. Total BAL fluid (BALF; 1 ml of saline three times) was collected and cells were counted using a 100-μl aliquot. Samples were centrifuged at 1,200 rpm for 5 min at 4°C, and cell pellets were resuspended in PBS. Cytospins were made by centrifugation at 450 rpm for 5 min at room temperature. Eosinophils were detected by staining according to May-Grünwald Giemsa and subsequently analyzed with a ZEISS microscope at 400×.

To determine cytokine concentrations in BALF, samples were centrifuged at 1,200 rpm for 5 min at 4°C. Supernatants were analyzed by specific ELISA for content of IFN-γ and IL-4 using commercially available kit systems (BD PharMingen).

24 h after the last intranasal challenge with either OVA or saline, lungs were analyzed by histology (day 28). The right lung was frozen immediately and stored in liquid nitrogen until use. The left lung was fixed in 10% formaline, dehydrated, mounted in paraffin, sectioned, and stained with hematoxylin/eosin. Eosinophils in the lungs were quantified blindly by the same pathologist (H.A. Lehr) using a Leitz D33 microscope, a JVC1 chip camera, the program Photoshop^®, and an Image Processing Tool Kit (v2.1). Three HPFs in the area of peripheral bronchi were randomly selected for quantification. Counts are given as eosinophils per square millimeter.

Airway mucus was identified by the alcian blue/periodic acid-Schiff (PAS) reaction by using a standard protocol in use in our Department of Pathology. PAS-positive mucosubstances in the small and large airways were also assessed by semiquantitative analysis in four to five mice per group using the following grading scale: −, no mucosubstances; +, mucosubstances detectable; ++, high amount of mucosubstances.

Immunohistochemistry was performed as described previously 29,30,31. In brief, frozen sections were fixed in 4% paraformaldehyde and immunostained for GATA-3 and IL-4. In brief, sections were preincubated with 2% normal rabbit serum in PBS/1% BSA/0.4% saponin for 45 min followed by an incubation with anti–GATA-3 goat polyclonal antibodies (5 μg/ml) in PBS/1% BSA/0.4% saponion/2% normal rabbit serum overnight at 4°C. The next day sections were incubated with biotinylated rabbit anti–goat IgG (1:100 dilution) followed by Streptavidin-Cy3 (1:1,000 dilution).

For IL-4 staining studies, sections were then incubated overnight with an anti–IL-4 biotinylated antibody (10 μg/ml; BD PharMingen) in PBS/1% BSA/0.4% saponin followed by the Streptavidin–conjugated Cy3 complex (1:1,000 in PBS). Sections were analyzed with an Olympus fluorescence microscope.

GATA-3– and IL-4–positive cells in the lungs were quantified blindly by the same pathologist using an Olympus microscope and an Image Processing Tool Kit (v2.1). 10 HPFs in the area of peripheral bronchi were randomly selected in 4–5 mice per group for quantification. Counts are given as positive cells per HPF.

Lung tissues were homogenized by using an electrical homogeneizator, and proteins were extracted in PBS in the presence of protease inhibitors (6.75% aprotinin, 312 μg/ml trypsin inhibitor) and 0.62% NP-40. Protein concentrations were determined by a protein assay according to the manufacturer's instructions (Bio-Rad Laboratories). The standard curve was performed by using 0, 4, 8, and 16 μg of BSA per milliliter. Under our experimental conditions ∼4 mg of proteins per lung could be recovered.

50 μg of total proteins isolated from the whole lung was separated by 15% SDS-PAGE and blotted onto a nitrocellulose membrane overnight 31. Equal loading was assessed with Ponceau's solution (Sigma-Aldrich). The membrane was then incubated in blocking solution (5% dry milk in PBS/0.05% Tween 20) for 1 h at room temperature, and subsequently exposed to 0.8 μg/ml of monoclonal antibodies anti–GATA-3 or anti–β-actin overnight at 4°C. The day after, the membrane was incubated with peroxidase-conjugated anti–mouse Igs (1:2,500) for 1 h at room temperature. Finally, enzyme activity was detected with the ECL-Plus Western blotting detection system according to the manufacturer's instructions (Amersham Pharmacia Biotech).

Electrophoretic mobility shift assay (EMSA) was carried out by the addition of 25,000 cpm radiolabeled GATA-3, OCT-1, and SP1 consensus DNA probes (Santa Cruz Biotechnology, Inc.) to the binding reaction that also contained 1 μg synthetic DNA duplex of poly(dIdC), 30 μg of total lung proteins, and binding buffer 32. Complex formation was allowed to proceed for 30 min. Finally, the complexes were separated from unbound DNA by native PAGE on 5% gels. The GATA-3 bands on the EMSA gels were analyzed by densitometry as specified in Results. For supershift analysis, 1 μg GATA-3–specific antibodies and control antibodies (obtained from Santa Cruz Biotechnology, Inc.) were used as specified in Results.

Airway reactivity was assessed in 12 mice per group by head-out plethysmography as described previously 33. In brief, mice were placed in four body plethysmographs attached to an exposure chamber (Crown Glass). Airflow was measured with a PTM 378/1.2 pneumotachograph (Hugo Sachs Electronics) and an 8-T2 differential pressure transducer (Gaeltec). In addition, airflow was analyzed in response to various concentrations of methacholine (MCh) (25, 50, 75, 100, 150 mg/ml for 1 min) delivered by a jet nebulizer (Pari-Boy; Pariwerke). Finally, the concentration of MCh that caused a 50% reduction in expiratory airflow (MCh₅₀) was determined.

In an additional series of experiments, airway responsiveness was assessed by analyzing enhanced pause (PenH) responses of saline- or OVA-treated mice in a body plethysmograph (model PLY 3211; Buxco Electronics, Inc.) 34. Measurements of MCh responsiveness were obtained by exposing mice for 5 min to 200 mg/ml of aerosolized MCh and monitoring PenH. Results were expressed as the peak of fold-increase of PenH within 5 min after MCh treatment.

In addition, pulmonary resistance (RL) was measured in the same cohort of untreated and treated OVA-sensitized mice as described previously 35,36. In brief, dose–response curves to MCh were obtained in anesthetized mice after administering increasing doses of intravenous MCh (33–1,000 μg/kg). Data were expressed as mean values of RL ± SEM.

Differences were evaluated for significance (P < 0.05) by the Student's two-tailed t test for independent events (Statworks). Data are given as mean values ± SEM.

For the analysis of RL data the multivariate analysis MANOVA (Wilks Lambda) was used.

In an attempt to specifically downregulate expression of GATA-3 in T lymphocytes in vitro and in vivo, we used an antisense DNA strategy. Accordingly, we designed specific CpG-lacking antisense phosphorothioate oligonucleotides targeting the translation start site of murine GATA-3 (see Materials and Methods). Then we tested the ability of such oligonucleotides to specifically downregulate GATA-3 expression in CD4⁺ cells isolated from the spleen during Th2 cell differentiation in vitro. We observed in coincubation studies an efficient uptake of FITC-labeled antisense oligonucleotides in cultured and already differentiated Th2 cells (day 6: 50–70% of cells) and Jurkat cell lines (>65% as assessed by FACS^® analysis and immunofluorescence; Fig. 1A and Fig. B) 24 h after the treatment with antisense oligonucleotides to GATA-3. This finding was associated with a downregulation of GATA-3 expression and a significant reduction of IL-4 release 12 h after incubation with PMA/ionomycin (Fig. 1C and Fig. D, respectively). However, IL-9 production by spleen T cells remained unchanged after antisense GATA-3 treatment (Fig. 1 E). It should be noted that the observed 50% reduction of PMA plus ionomycin-inducible IL-4 production upon antisense treatment is remarkable, as only 50–70% of the spleen cells showed DNA uptake. Indeed, cotransfection- based enrichment of antisense-transfected T cells led to a more pronounced downregulation of IL-4 production by Th2 cells (Fig. 1).

Treatment with control nonsense oligonucleotides did not affect GATA-3 expression or IL-4 and IL-9 release under the same experimental conditions. Treatment with mismatched DNA also led to a reduction of the average number of GATA-3–expressing cells and IL-4 release, possibly due to a remaining weak hybridization of the mismatched DNA to the GATA-3 target mRNA (Fig. 1). However, this effect was less pronounced compared with the antisense DNA effect, and the mismatched DNA–induced changes were not statistically significant. Furthermore, this effect was not seen using a second mismatched control oligonucleotide (Fig. 1 H), suggesting that antisense oligonucleotides to GATA-3 cause a specific suppression of GATA-3 expression and IL-4 production by cultured T cells.

Based on previous studies showing a key role for GATA-3 in Th2 cytokine production and increased Th2 cytokine production in asthma 2,18, we analyzed in a consecutive series of studies the expression of GATA-3 in a murine model of asthma 37 associated with late phase allergic pulmonary inflammation in OVA-sensitized mice. Accordingly, we immunostained lung sections from OVA-treated and control mice with anti–GATA-3 antibodies. As shown in Fig. 2, OVA immunization caused a striking increase in lung-infiltrating cells expressing GATA-3 compared with saline treatment.

In further studies, we wanted to directly test the effects of GATA-3 antisense phosphorothioate oligonucleotides on the expression of GATA-3 during the late phase allergic asthmatic reaction in vivo. Intranasal delivery of FITC-labeled GATA-3 antisense oligonucleotides led to a DNA uptake by GATA-3–expressing lung cells in vivo (Fig. 2C and Fig. F). This finding was accompanied by a significant decrease of GATA-3 expression in the lung of OVA-sensitized mice (Fig. 2B and Fig. C, and bottom panel), while control nonsense oligonucleotides had no such effect (Fig. 2 E, and bottom panel). Mismatched DNA led to a reduction of the number of GATA-3- positive cells in the lung (Fig. 2 D, and bottom panel), although this effect was not statistically significant.

In additional studies, we performed Western blot analysis of total proteins derived from the lung to assess local GATA-3 expression. As shown in Fig. 3, these studies confirmed the downregulation of GATA-3 protein expression in antisense GATA-3–treated lungs but not in lungs treated with control nonsense oligonucleotides (Fig. 3 C). Similarly, analysis of lung proteins for GATA-3 expression by gel retardation assays (EMSA) using a GATA-3 consensus oligonucleotide confirmed the downregulation of GATA-3 upon antisense treatment (Fig. 3 A). Densitometric analysis of the EMSA bands showed a significant downregulation of GATA-3 upon antisense DNA treatment (Fig. 3 B). In contrast, the transcription factors OCT-1 and SP-1 were not affected by GATA-3 antisense DNA (Fig. 3 B). Furthermore, nonsense control oligonucleotides were virtually inactive with regard to GATA-3 expression. Taken together, these results indicated that local administration of specific antisense oligonucleotides leads to specific downregulation of pulmonary GATA-3 expression in an asthma model in vivo.

To assess the histologic effects of GATA-3 antisense phosphorothioate oligonucleotides on allergen-induced airway inflammation, we analyzed lung tissues 24 h after the last of three sequential OVA challenges (days 25–27) on day 28. In untreated OVA-sensitized mice, an infiltration of the bronchial interstitium with eosinophils was observed (Fig. 4C and Fig. D). Treatment of OVA-sensitized mice with nonsense GATA-3 oligonucleotides had no detectable effects (Fig. 4G and Fig. H). In contrast, no inflammatory signs were seen in OVA-sensitized mice upon treatment with corticosteroids or GATA-3 antisense oligonucleotides (Fig. 4I, Fig. K, Fig. E, and Fig. F). Control mice were treated with saline complexed with alum and rechallenged intranasally with saline; they show an intact airway epithelium and no sign of inflammation (Fig. 4A and Fig. B).

In further studies on the potential mechanisms of antisense-induced suppression of lung inflammation, we assessed allergen-induced infiltration of eosinophils in the lung. By morphometric analysis, the influx of eosinophils into the lung interstitium was significantly reduced after administration of GATA-3 antisense oligonucleotides compared with untreated mice (Fig. 5). Antisense-induced suppression of the number of eosinophils in the BALF was comparable to that induced by intranasal DEX treatment (BAL cell viability 77.5 ± 9.8%). However, no difference was seen in cell viability in the BALF after antisense treatment in OVA-sensitized lungs compared with saline treatment (84.4 ± 3.5% and 91.15 ± 6.3% viability, respectively, as assessed by trypan blue exclusion). In contrast, treatment with control mismatched (BAL cell viability 87.27 ± 3.8%) and nonsense oligonucleotides (BAL cell viability 79.15 ± 5.45%) had no significant effect on the OVA-induced influx of eosinophils in the airways (Fig. 5).

Since T cell cytokines are known to regulate airway inflammation in vivo, we next determined Th1 and Th2 cytokine levels in the BALF of OVA-sensitized mice (Fig. 6). Interestingly, neither OVA nor antisense treatment altered IFN-γ levels in the BALF compared with saline-treated mice. However, IL-4 concentration in the BAL was significantly suppressed (P < 0.05) by GATA-3 antisense DNA treatment, suggesting that blockade of GATA-3 expression selectively inhibits Th2 cytokine levels in the BAL. Nonsense control DNA did not cause significant changes in IL-4 production, whereas mismatched oligonucleotides led to downregulation of the average IL-4 production. However, this effect was not statistically significant.

Many mucus-producing cells (acidic sulfated and neutral mucosubstances) were seen in the airways of mice sensitized to OVA (Fig. 7) compared with saline-treated mice (saline), as assessed by the alcian blue/PAS reaction. These inflammatory changes were absent in mice that were treated with GATA-3 antisense DNA (AS-GATA-3). DEX treatment reduced PAS-positive cells in the airways. In contrast, no changes were seen after treatment with nonsense control oligonucleotides.

In a final series of experiments, we analyzed whether antisense oligonucleotides to GATA-3 would affect airway hyperreactivity in OVA-sensitized mice. This point was of particular importance since it was unclear whether modulation of the function of GATA-3 would not only affect airway inflammation but also airway hyperresponsiveness.

In initial experiments, we subjected 12 OVA-sensitized mice or control mice per group to treatment with antisense oligonucleotides, control nonsense oligonucleotides, or saline and assessed airway hyperreactivity by noninvasive head-out body plethysmography after MCh challenge at day 28. As shown in Fig. 8A and Fig. B, OVA-sensitized mice showed an increase in airway hyperreactivity compared with saline-treated control mice after challenge. Furthermore, it was found that antisense oligonucleotides to GATA-3 lead to a significant reduction of airway hyperreactivity to levels comparable to saline-treated control mice. Finally, we observed that control nonsense oligonucleotides had no effect on airway hyperreactivity in OVA-sensitized mice.

To validate our findings on a significant downregulation of airway hyperreactivity upon antisense treatment, we assessed in subsequent studies PenH responses of OVA-sensitized mice in a body plethysmograph (Fig. 8 C). It was found that administration of antisense oligonucleotides to GATA-3 results in a significant downregulation of MCh responsiveness, whereas nonsense control oligonucleotides had no such effect. Finally, to assess the MCh response in the lower airways, we measured RL and dose–response curves to intravenous MCh in anesthetized OVA-treated mice by body plethysmography (Fig. 8 D). These experiments demonstrated a significant reduction of RL in OVA-sensitized mice upon treatment with antisense but not nonsense control oligonucleotides. Thus, using three independent methods to assess airway hyperreactivity, these data showed that airway hyperreactivity was significantly suppressed by administration of antisense oligonucleotides to GATA-3.

Recent studies in transgenic mice have revealed that expression of a dominant negative form of GATA-3 can prevent allergic airway inflammation 38, indicating that GATA-3 plays a crucial role in the development of Th2-mediated lung inflammation 39. However, it remains unclear whether GATA-3 plays a role in the effector phase of allergic airway inflammation and whether antagonizing the expression and/or function of GATA-3 can be used for the therapy of allergic airway inflammation and hyperresponsiveness. Furthermore, the potential mechanisms of action of such therapy remain to be elucidated. In this study, we demonstrate a key role of the transcription factor GATA-3 for cytokine production of T cells in the effector phase of a murine asthma model induced by OVA. We found that specific blockade of GATA-3 expression in the lung by antisense oligonucleotides leads to suppression of airway inflammation and Th2 cytokine production both in the BAL and in the interstitium of the lung. Antisense-induced blockade of GATA-3 was at least as effective to suppress lung inflammation as the administration of corticosteroids. Furthermore, using three independent methods to assess airway hyperresponsiveness to MCh, we demonstrate that GATA-3 antisense oligonucleotides significantly reduce airway hyperresponsiveness in a model of allergic asthma, suggesting a novel function of GATA-3 as a key regulator of airway hyperresponsiveness in allergic asthma. Taken together with the recently described increased expression of GATA-3 mRNA in lung CD3⁺ T cells of patients with allergic asthma 27, these data suggest the potential therapeutic utility of GATA-3 antisense oligonucleotides as a novel molecular approach for the treatment of patients with allergic asthma in humans. This approach has the potential advantage of blocking the production of several Th2 cytokines simultaneously rather than suppressing the activity of a single cytokine.

The GATA family of zinc finger transcription factors has pleiotropic roles in cell development and differentiation. While GATA-1 and GATA-2 are involved in the development of the erythroid and megakaryocytic lineages, GATA-3 is highly expressed in early T lymphocytes and controls T cell receptor gene expression and thymocyte development 17,14. Furthermore, within mature T cells GATA-3 was found to be selectively expressed in naive T cells and to control Th2 T cell development 18,21,40,41. These data prompted us to determine the expression and functional role of GATA-3 in a murine model of asthma associated with Th2 cytokine production. We observed that OVA sensitization of mice caused a strong upregulation of pulmonary GATA-3 expression. This finding led us to further investigate the functional role of this factor using an antisense DNA strategy. We found that exposure of cultured spleen Th2 cells to an antisense GATA oligonucleotide suppressed IL-4 expression. This effect on IL-4 production could be augmented by MACS-based selection of antisense-transfected T cells using a cotransfection protocol with a plasmid expressing a truncated H2-K^k molecule. In the supernatant of antisense-treated T cells, we could not see any decrease in IL-9 production after treatment with antisense to GATA-3, possibly because the IL-9 promoter has no binding sites for GATA-3 42. Using antisense oligonucleotides we then tested the potential pathogenic role of GATA-3 in the OVA-dependent asthma model in vivo. We observed that intranasal administration of FITC-labeled GATA-3 antisense phosphorothioate oligonucleotides causes an efficient uptake in interstitial lung cells that coexpress large amounts of GATA-3 and IL-4 upon OVA sensitization. This uptake was followed by downregulation of lung GATA-3 expression (compared with OVA-sensitized mice without treatment) within 24 h after the last antisense administration. The specificity of the antisense effect on GATA-3 expression was confirmed according to several previously described criteria 43,44: (a) control nonsense oligonucleotides did not have an effect on GATA-3 expression; (b) the antisense oligonucleotide had no effect on a nontargeted protein (SP-1, OCT-1); and (c) direct toxic effects of the GATA-3 antisense oligonucleotide were excluded by cell viability tests. Interestingly, mismatched control DNA led to a reduction of GATA-3 expression and IL-4 production in T cells, probably due to a remaining weak hybridization of the mismatched DNA to the GATA-3 target mRNA. However, these changes were less pronounced compared with the antisense DNA effect and did not reach statistical significance.

The antisense-induced suppression of GATA-3 expression was accompanied by a suppression of Th2 cytokine synthesis in vitro and in vivo and the accumulation of eosinophils in the airways, whereas treatment with control nonsense oligonucleotides had no such effects. Mucus occlusion of the airway lumen in the OVA-treated mice was present 24 h after the last intravenous OVA administration on day 28. Furthermore, in vivo noninvasive pulmonary function tests using body plethysmography showed that the OVA-immunized challenged mice reproducibly developed increased airway responsiveness to MCh that could be completely reversed by treatment with antisense DNA to GATA-3 but not by control nonsense DNA. This is an important novel information that highlights GATA-3 as a key regulator of both T cell effector function and airway hyperresponsiveness in allergic airway inflammation. It should be noted that the above finding on antisense-induced suppression of airway hyperresponsiveness was validated by two additional independent methods using assessment of PenH responses and of RL in anesthetized mice. All three methods demonstrated a significant downregulation of OVA-induced airway hyperresponsiveness upon administration of antisense oligonucleotides to GATA-3, definitively demonstrating for the first time a critical regulatory role of GATA-3 in airway hyperresponsiveness in allergic asthma.

The potential relevance of this GATA-3 antisense-based approach to inhibit Th2 cytokine production is underlined by various functional studies showing a key regulatory role for both IL-4 and IL-5 in asthma. For instance, studies in transgenic and knockout mice showed that the Th2 cytokine IL-4 is important for IgE mRNA production in allergic asthmatic reactions 45,46. Furthermore, studies in a knockout model for the IL-4–dependent STAT-6 have shown that suppression of IL-4 signal transduction profoundly inhibits influx of eosinophils into the lung 47. Finally, suppression of IL-5 has been shown to prevent allergic inflammation in an animal model 48, possibly due to suppression of IL-5–dependent activation of eosinophils 49.

Recent in vitro and in vivo studies have shown that various transcription factors such as c-maf, nuclear factor of activated T cells c1-4, NIP 45, STAT-6, and GATA-3 are important for Th2-type cytokine production in T cells 14,50. Here, we show that local GATA-3 antisense DNA treatment directly affects Th2 cytokine production and lung inflammation in asthma, although the oligonucleotides lack CpG motifs. These data support and significantly extend data from a very recent report 38 showing that expression of a dominant negative mutant of GATA-3 in T cells from transgenic mice leads to profound attenuation of OVA-induced eosinophilia and Th2 cytokine production. However, we demonstrate in this study for the first time that GATA-3 plays a role in the effector phase of allergic airway inflammation using mice with a normal immune system. Furthermore, we show that blockade of GATA-3 expression can be used as a therapeutic approach to suppress lung inflammation and mucus production in a murine model of asthma in vivo. Finally, we have found that GATA-3 is a critical factor for the maintenance of airway hyperresponsiveness in allergic airway inflammation.

Taken together, these data suggest that antisense DNA treatment to GATA-3 is a novel selective way to simultaneously suppress the production of several Th2 type cytokines in asthmatic reactions in vivo. Furthermore, we have demonstrated here that GATA-3 antisense DNA treatment of asthma is at least as effective as corticosteroid treatment, suggesting that antisense oligonucleotides to GATA-3 may be a novel molecular approach for the treatment of inflammatory disorders such as asthma in humans.

The authors would like to thank ChongQing Liu for his technical help with the body plethysmography. In addition, the authors greatly acknowledge Professor R. Buhl (University of Mainz) for his support in this study.

The research of M.F. Neurath and S. Finotto was supported by grants from the Innovationsstiftung Rheinland-Pfalz, the MAIFOR program, the Deutsche Forschungsgemeinschaft (Ne 490/1-1, Ne 490/2-1) and the Gerhard Hess program of the Deutsche Forschungsgemeinschaft (Ne 490/3-1).