Cytotoxic T-lymphocyte antigen-4 inhibits GATA-3 but not T-bet mRNA expression during T helper cell differentiation

Abstract

Naive CD4⁺ T-cell differentiation to T helper 1 (Th1) and Th2 cells is dependent on T-bet and GATA-3 factors, respectively. T-bet and GATA-3, indeed, through chromatin remodelling allow transcriptional activation of Ifnγ and Th2 cytokine (Il4, Il5, Il13) genes, respectively. We investigated the effects of the negative costimulatory receptor cytotoxic T-lymphocyte antigen-4 (CTLA-4) on GATA-3 and T-bet mRNA expression and Th cell differentiation in mouse naive CD4⁺ T cells. Our results show that CTLA-4 inhibits GATA-3 mRNA expression and Th2 cell differentiation. At variance, CTLA-4 does not affect T-bet mRNA expression and Th1 cell differentiation. GATA-3 mRNA expression is inhibited when CD4⁺ cells are stimulated under both neutral (i.e. absence of cytokines) and Th2-polarizing (i.e. presence of interleukin (IL)-4) conditions, the effect being larger under the latter condition. Hence CTLA-4 might affect the IL-4/signal transducer and activator of transcription-6 (STAT6) pathway leading to GATA-3 mRNA up-regulation. We found, indeed, that CTLA-4 engagement inhibits STAT6 activation leaving unaffected the STAT6 protein level. Moreover, CTLA-4 engagement drastically inhibits IL-4Rα mRNA and protein up-regulation under Th2-polarizing conditions. Thus, CTLA-4 exerts a tight control on Th2 cell differentiation by negatively regulating both the CD3/CD28 and the IL-4/STAT6 pathways.

Introduction

Upon activation, naive CD4⁺ T cells begin a differentiation process that commits these cells to produce a specific pattern of cytokines. T helper 1 (Th1) cells produce interferon-γ (IFN-γ) (and not interleukin (IL)-4) and promote cell-mediated immune responses against intracellular pathogens. Th2 cells produce IL-4, IL-5 and IL-13 (and not IFN-γ), sustain humoral immune responses against parasites such as helminths, and favour the onset of allergies. Protective immunity depends on the proper balance between Th1 and Th2 cell responses. Many factors influence the process leading naive CD4⁺ T cells to differentiate to either Th1 or Th2 cells.¹ IL-12 and IL-4 play a major role in Th cell polarization by promoting, through the activation of signal transducer and activator of transcription-4 (STAT4) and STAT6, Th1 and Th2 cell differentiation, respectively.^2,3 Besides STAT6, GATA-3 has been described to play a relevant role in Th2 cell commitment.^4,5 GATA-3 is expressed at low level in naive CD4⁺ T cells, is increased by T-cell receptor (TCR)/CD28 costimulation⁶ and up-regulated by IL-4 through STAT6.⁷ GATA-3 can transactivate the IL-5 promoter,⁸ but its major role is to establish transcriptional competence for the Th2 cytokine gene cluster (Il-13/Il-4/Il-5) through chromatin remodelling. ^9–11 GATA-3 has also a transcriptional auto-activating loop as GATA-3 protein drastically increases GATA-3 mRNA expression.⁷ The T-box transcription factor T-bet plays a crucial role in Th1 cell differentiation¹² by promoting chromatin remodelling of the IFN-γ locus.¹³ T-bet induces IL-12Rβ2 subunit expression and sustains its own expression through the autocrine effects of IFN-γ/STAT1 signalling.^13,14

Many studies showed that beside the leading cytokines other factors are involved in Th cell differentiation. Antigen dose and type, strength of TCR signals, costimulatory receptors, cytokines other than IL-12 and IL-4, and genetic background are relevant variables in this process.^1,15 Cytotoxic T lymphocyte antigen-4 (CTLA-4) is a costimulatory receptor that inhibits T-cell proliferation and cytokine production. ^16–18 CTLA-4 knockout (CTLA-4^−/−) mice show lymphoproliferative disorders^19,20 characterized by massive CD4⁺ T-cell infiltrates with Th2-skewed cytokine secretion pattern.²¹ In mice CTLA-4 blockade enhances allergic sensitization and eosinophilic airway inflammation.²² In humans, CTLA-4 polymorphisms have been associated with increased immunoglobulin E (IgE) serum levels and allergies.^23,24 Lack of CTLA-4 promotes antigen-specific Th2 responses in vitro²⁵ and renders Th2 differentiation independent from IL-4/STAT6 pathway in vivo.²⁶ Moreover, CTLA-4 engagement has been shown to inhibit Th2^25,27 and promote Th1 cell differentiation.²⁸

Altogether, these findings induced us to investigate the effects of CTLA-4 engagement on GATA-3 and T-bet mRNA expression during Th cell differentiation. Results show that CTLA-4 exerts a double control on GATA-3 mRNA expression by dampening the effects of TCR/CD28 costimulation as well as the activation of the IL-4R/STAT6 pathway. Conversely, T-bet mRNA expression was not affected by CTLA-4.

Materials and methods

Cell purification and cultures

Specific pathogen-free C57Bl/6 mice were purchased from Charles River Laboratories Italia SpA (Milan, Italy). Splenic naive CD4⁺ T cells were purified by immuno-magnetic cell sorting according to the manufacturer's instructions (Miltenyi Biotec, Bergish Gladbach, Germany), catalogue numbers 130-058-701 and 130-049-701). Collected cells were found to be almost exclusively (>95%) CD4⁺ CD44^lo CD45RB^hi by flow cytometry analysis. To induce Th cell polarization, naive CD4⁺ cells were cultured in 24-well plates precoated with anti-CD3ε monoclonal antibody (mAb, clone 145-2C11; 10 µg/ml). Anti-CD28 mAb (clone 37.51) was added in soluble form (1 µg/ml). In Th1 polarizing cultures, IL-12 (10 ng/ml) and an anti-IL-4 blocking mAb were also added. In Th2 polarizing cultures, IL-4 (5 ng/ml) and an anti-IFN-γ blocking mAb were added. CD4⁺ T cells were also stimulated with anti-CD3 and CD28 mAbs in the presence of both anti-IFN-γ and anti-IL-4 blocking antibodies (neutral conditions). IL-2 (10 ng/ml; BD-Pharmingen, San Diego, CA, USA) was added under all conditions. Anti-CTLA-4 mAb (clone UC10-4F10-11) or anti-trinitrophenol (TNP, clone A19-3) mAb was bound to culture plates at the concentration of 10 µg/ml. We previously found^29,30 that this anti-CTLA-4 mAb when immobilized to plastic induces inhibition of IL-2 production as well as of nuclear factor (NF)-κB and NF-AT activation. At variance, the anti-TNP mAb does not affect these parameters. All antibodies used in culture were sodium azide- and endotoxin-free as certified by the producer (BD-Pharmingen).

Cytokine titration

IFN-γ and IL-4 were titrated in culture supernatants by sandwich enzyme-linked immunosorbent assay (ELISA) as already described.²⁷ IL-5 was titrated by using a commercial kit (Endogen, Woburn MA; cat. KM-IL-5). Absorbance was measured at 405 nm. The reference straight line obtained by plotting the absorbance vs. the standard cytokine concentrations was used to calculate the cytokine concentrations in the culture supernatants.

Flow cytometry analysis

Cells were preincubated with Fc Block (anti-CD16/32, clone 2.4G2, BD-PharMingen) to prevent cytophilic binding of labelled antibodies. Cytokine-producing cells were revealed by staining cells with fluoroscein isothiocyanate (FITC)-conjugated anti-IFN-γ mAb (clone XMG1.2), phycoerythrin (PE)-conjugated anti-IL-4 mAb (clone BVD4-1D11) and antigen-presenting cell (APC)-conjugated anti-IL-5 mAb (clone TRFK5). Before staining, cells were permeabilized with saponin (0·5%) and fixed with paraformaldehyde (2%). Isotype-matched antibodies were used as controls. The optimal concentrations of all the antibodies were assessed in preliminary experiments. Samples of 2 × 10⁵ cells were analysed, and fluorescence signals collected in log mode using a FACScalibur (Becton Dickinson, San Jose, CA).

Transcript analysis

Reverse transcription–polymerase chain reaction (RT–PCR) was performed using real-time fluorogenic 5′-nuclease PCR with an ABI prism 7700 (Applied Biosystems, Foster City, CA) according to the manufacturer's instructions. Total RNA samples were extracted by TRIzol Reagent (Life Technologies, Gaithersburg, MD), treated with DNase I (Sigma-Aldrich, St Louis, MO) and reverse-transcribed using oligo(dT) and reverse-transcriptase (Perkin-Elmer GeneAmp RNA PCR kit; Perkin-Elmer, Wellsley, MA, USA). Primers and probes for Pgk1, GATA-3, T-bet and IL-4Rα chain cDNA amplification were from ‘assays on demand’ kits (Applied Biosystems; catalogue numbers Mm00435617_m1; Mm00484683_m1; Mm00450960_m1; Mm00439634_m1). Cycling conditions were according to the producer's instructions. The absence of PCR products from genomic DNA contamination was confirmed by PCR on sham-reverse-transcribed samples (samples that underwent reverse transcription in the absence of RT). Linearity between threshold cycle number and sample concentration has been verified by amplifying serially diluted relevant samples for each transcript. Pgk1, GATA-3 and T-bet mRNA were analysed simultaneously on the same plate. Pgk1 and IL-4Rα chain were analysed simultaneously in another plate. GATA-3 and T-bet as well as IL-4Rα chain transcripts were normalized to Pgk1 level. To evaluate mRNA stability for GATA-3, after 24 hr stimulation, mRNA synthesis was blocked by actinomycin D (10 µg/ml, Sigma-Aldrich). After 30, 60 and 120 min, RNA samples were collected and processed as described above. The amount of the residual specific mRNA was expressed as fraction of the initial amount.

Electrophoretic mobility shift assay (EMSA)

CD4⁺ T cells were stimulated for 1 and 3 days under Th2-polarizing conditions and, after 4 hr starvation, restimulated by IL-4 (5 ng/ml) for 15 min and lysed to obtain protein extracts as previously described.³¹ For binding reaction protein extracts were incubated with a ³²P-end-labelled double-stranded oligonucleotide corresponding toSTAT6 binding site. The resulting DNA–protein complexes were resolved by electrophoresis on non-denaturating polyacrylamide gel. After drying, gels were exposed on a phosphor screen which was subsequently analysed by phosphor/fluorescence imager Typhoon 9210 (Amersham Biosciences, Piscataway, NJ, USA). The intensity of the bands was directly quantified by Image QuaNT software (Amersham Biosciences) which gives rise to a volume report by integrating the area of the band and its density.

Cell-based ELISA for Pho-STAT6

STAT6 phosphorylation was evaluated by using a FACE™ STAT ELISA Kit (Active Motif, Carlsbad, CA, USA) according to the producer's protocol. Briefly, after 1 or 3 days stimulation under Th2 conditions, cells (2 × 10⁵/well) were seeded in poly L-lysine-coated (Sigma P4707, 20 µg/ml, o/n, 4°) 96-well plates (Costar 3596) and re-stimulated with rIL-4 (10 ng/ml; RM-IL4-10; Endogen, Woburn, MA, USA) for 0, 5 or 15 min. Cells were then fixed (8% formaldehyde, 20 min) and washed with phosphate-buffered saline (PBS) containing 0·1% Triton-X-100. Endogenous peroxidase was quenched with 1% H₂O₂ and 0,1% sodium azide. Cytophilic non-specific binding of antibodies was prevented with a blocking buffer provided by the producer. Plates were then incubated overnight with the anti-Pho-STAT6 or the anti-total-STAT6 antibody. A peroxidase-conjugated goat anti-rabbit IgG was used to detect Pho-STAT6- or total-STAT6-bound primary antibody. Then, substrate was added and absorbance measured at 450 nm (dual wavelength, reference wavelength 630 nm). For each sample, data are presented as ratio between the Pho-STAT6 OD value and the total-STAT6 OD value. As values in this assay may be affected by cell number, particular attention was used during cell counts, cell dilutions and plating to reduce errors. Nevertheless, a cell quantification assay was performed after the Pho/total-STAT6 ELISA using a crystal violet solution according to the producer. OD values at 595 nm were used to normalize Pho-STAT6 and total-STAT6 sample values. All tests were performed in triplicate.

Results

CTLA-4 engagement inhibits Th2 but not Th1 cell polarization

Purified naive CD4⁺ T cells were stimulated with anti-CD3, anti-CD28 and blocking anti-IL-4 mAbs in the presence of IL-12 (hereafter Th1 polarizing conditions) to induce Th1 cell differentiation. Th2 cell differentiation was induced by anti-CD3, anti-CD28, and blocking anti-IFN-γ mAbs in the presence of IL-4 (hereafter Th2 polarizing conditions). CTLA-4 stimulation was induced by plate-bound mAb while a isotype-matched mAb was used as control.

One week-polarized Th1 and Th2 cells were stimulated with phorbol 12-myriste 13-acetate (PMA) and ionomycin and stained intracellularly for IFN-γ, IL-4 and IL-5. Under Th1-polarizing conditions 27% of the cell population produces IFN-γ whereas almost no cells produce IL-4 (Fig. 1a) and IL-5 (not shown). In parallel cultures Th1-polarized cells were stimulated with anti-CD3 and anti-CD28 mAbs or PMA and ionomycin for 48 hr. Results show that CTLA-4 engagement does not change the frequency of IFN-γ-producing cells and the amount of IFN-γ produced (Fig. 1a, b).

Figure 1.

CTLA-4 engagement inhibits differentiation of naive CD4⁺ T cells to Th2 but not to Th1 cells. Naive CD4⁺ T cells were stimulated under Th1 (a and b) or Th2 (c–f) polarizing conditions in the presence of either anti-CTLA-4 (αCTLA-4, black columns) or an isotype-matched control (iso ctrl, white columns) mAb. (a and c) After 1 week polarization, cells were restimulated with PMA and ionomycin in the presence of brefeldin A for 5 hr and intracellularly stained with FITC-conjugated anti-IFN-γ, PE-conjugated anti-IL-4 and APC-conjugated anti-IL-5 mAbs and analysed by flow cytometry. Isotype-matched control mAb were used to set quadrant positions. (b, d and e) After 1 week polarization, cells were restimulated with anti-CD3 and anti-CD28 mAbs or PMA and ionomycin for 48 hr. Culture supernatants were analysed by ELISA. (f) Shows percentage of IL-4⁺ and IL-5⁺ cells in cultures stimulated as in (c). Each panel shows data from one experiment out of three. Values in (f) represent means of three independent experiments ± SE. *P < 0·05.

As shown in Fig. 1(c), after 1 week stimulation under Th2 polarizing conditions one fourth of the cell population expresses IL-4 whereas very few cells are IFN-γ⁺. CTLA-4 engagement under Th2 polarization reduces the frequency of IL-4 and IL-5-producing cells from 28·3 ± 4·7% and 24·6 ± 5·8 to 14·2 ± 3·9% and 9·4 ± 5·2%, respectively (Fig. 1c, f). These reduced frequencies results in a lower IL-4 and IL-5 production as induced by re-stimulation with anti-CD3 and anti-CD28 mAbs (Fig. 1d, e). Inhibition of IL-4 production is also shown upon PMA/ionomycin re-stimulation (Fig. 1d). Thus, these data, extending previous observations from our²⁷ and other²⁵ groups, show that CTLA-4 inhibits Th2 but not Th1 cell polarization.

CTLA-4 engagement inhibits GATA-3 mRNA expression

Naive CD4⁺ cells were stimulated under either neutral or Th1/Th2-polarizing conditions for 24 and 72 hr. GATA-3 mRNA was evaluated by retro-transcription and real time PCR. Results, expressed as arbitrary units, are referred to the GATA-3 mRNA level of the un-stimulated cells (Fig. 2). Figure 2(a) shows that upon CD3/CD28 costimulation GATA-3 mRNA expression is readily up-regulated. CTLA-4 engagement reduces GATA-3 mRNA expression, the inhibition being 55% and 50% after 24 and 72 hr stimulation, respectively. Under Th1-polarizing conditions after an initial weak up-regulation, GATA-3 mRNA expression is drastically (85%) inhibited (Fig. 2b). Under these conditions CTLA-4 engagement slightly affects GATA-3 mRNA levels. Figure 2(c) shows GATA-3 mRNA expression under Th2-polarizing conditions. After 24 hr, the presence of IL-4 up-regulates GATA-3 mRNA expression as compared to CD3/CD28 costimulation alone (Fig. 2c versus Fig. 2a). CTLA-4 engagement dampens GATA-3 mRNA expression, the inhibition (75–77%) under Th2 polarizing conditions being higher than under neutral conditions. Altogether these results show that CTLA-4 negatively affects GATA-3 expression and indicates that this control can be exerted not only on the CD3/CD28 pathway but also on the IL-4-induced up-regulation of GATA-3 mRNA expression.

Figure 2.

CTLA-4 inhibits GATA-3 mRNA expression. Naive CD4⁺ T cells were stimulated under neutral (a), Th1 (b) or Th2 (c and d) polarizing conditions in the presence of either anti-CTLA-4 mAb (black columns) or an isotype-matched control mAb (white columns). At the indicated time, total RNA was isolated and, after DNAse treatment, relative GATA-3 and Pgk1 mRNA expression was analysed by RT and real time PCR. Results, normalised on the house keeping gene (Pgk1) mRNA level, are referred to the GATA-3 mRNA level in naive cells and expressed as arbitrary units. (d) After 24 hr stimulation under Th2-polarizing conditions, mRNA expression was blocked by actinomycin D (Act. D). GATA-3 mRNA level was analysed after further 30, 60 and 120 min and expressed as residual amount of the GATA-3 mRNA level before actinomycin D addition. Results were confirmed in two other independent experiments. (a and c) *P < 0·05 and **P < 0·01 for cells stimulated in the presence of anti-CTLA-4 mAb versus cells stimulated in the presence of the isotype-matched control mAb; (b) §P < 0·05 for cells stimulated versus unstimulated cells.

CTLA-4 engagement does not affect GATA-3 mRNA stability

One of the major regulatory mechanisms in gene expression involves mRNA turn-over. Naive CD4⁺ T cells were stimulated under Th2 polarizing conditions to test whether CTLA-4 engagement affects GATA-3 mRNA stability. After 24 hr stimulation, mRNA transcription was blocked by actinomycin D and mRNA expression analysed by RT and real time PCR after 30, 60 and 120 min. Results, expressed as fraction of the initial amount of GATA-3 mRNA, show that GATA-3 mRNA turn-over is not affected by CTLA-4 engagement (Fig. 2d).

CTLA-4 engagement does not affect T-bet mRNA expression

Naive CD4⁺ cells were stimulated under either neutral or Th1/Th2-polarizing conditions for 24 and 72 hr. Results, expressed as arbitrary units, are referred to the T-bet mRNA level of the un-stimulated cells (Fig. 3). Upon CD3/CD28 costimulation T-bet mRNA expression is up-regulated (Fig. 3a) and further increased under Th1-polarizing (Fig. 3b) conditions. CTLA-4 engagement slightly affects T-bet mRNA levels under all conditions tested. Figure 3(c) shows T-bet mRNA expression under Th2-polarizing conditions. As expected, IL-4 drastically down-regulates T-bet mRNA expression. Altogether these data show that CTLA-4 engagement does not affect T-bet mRNA expression. GATA-3 versus T-bet mRNA levels have been correlated with Th2 versus Th1 cytokine profiles.³² Figure 3(d) shows GATA-3 mRNA level/T-bet mRNA level ratio from data of panels A and C from Figs 2 and 3. It appears that CTLA-4 engagement reduces GATA-3/T-bet ratio under both neutral and Th2-polarizing conditions.

Figure 3.

CTLA-4 does not affect T-bet mRNA expression. Naive CD4⁺ cells were stimulated under neutral (a), Th1 (b) or Th2 (c) polarizing conditions and mRNA level analysed as in Fig. 2. Results are referred to T-bet mRNA level in naive cells and expressed as arbitrary units. (d) GATA-3/T-bet mRNA ratio. Values represent the ratio between data from (a and c) from Figs 2 and 3. Results were confirmed in two other independent experiments. §P < 0·05 for cells stimulated versus unstimulated cells

CTLA-4 engagement inhibits STAT6 activation

The inhibitory effect of CTLA-4 on GATA-3 mRNA expression occurs both under Th2-polarizing conditions and during CD3/CD28 costimulation alone. The link between IL-4 and Th2 development through a STAT6-dependent pathway is well known.³³ To examine whether CTLA-4 engagement affects the IL-4/STAT6 pathway naive CD4⁺ T cells were stimulated under Th2-polarizing conditions in the presence of either anti-CTLA-4 or a isotype-matched control mAb. After stimulation for 1 and 3 days cells were collected and re-stimulated with IL-4 to induce STAT6 activation. Protein extracts were then analysed by EMSA for binding to the STAT6 consensus sequence. Results show that STAT6 activation increases during cell polarization, the binding activity being evident after 1 day and increased after 3 days. CTLA-4 engagement during polarization inhibits STAT6 activation after stimulation for 1 and 3 days (Fig. 4a). Inducible shifted DNA–protein complexes were identified as STAT6–probe complexes by super shift analyses.

Figure 4.

CTLA-4 engagement inhibits STAT6 activation. Naive CD4⁺ T cells were stimulated under Th2 polarizing conditions for 1 or 3 days in the presence of either a CTLA-4 engaging mAb (+, black columns) or an isotype-matched control (−, white columns) mAb. (a) After 15 min re-stimulation with IL-4, cells were lysed and cellular extracts analysed by EMSA. Inducible shifted DNA-protein complexes were identified as STAT6-probe complexes by super shift analyses (control, isotype-matched control antibody; anti-STAT6, anti-STAT6 antibody). Columns represent volume report as calculated by the ImageQuaNT software integrating the area of each band and its optical density. Values are means (± SE) from three independent experiments. (b and c) After 5 or 15 min re-stimulation with IL-4, Pho-STAT6 and total-STAT6 levels were analysed by cell-based ELISA. Columns in (b) represent means (± SE) of the ratios between OD value for the Pho-STAT6 and OD value for the corresponding total-STAT6 sample from two independent experiments. *P < 0·05 for cells stimulated in the presence of anti-CTLA-4 mAb versus cells stimulated in the presence of the isotype-matched control mAb. (c) Columns represent means (± SE) of the OD value for the total-STAT6 samples from two independent experiments.

STAT6 activation (dimerization and nuclear transfer) depends on tyrosin phosphorylation. Naive CD4⁺ T cells were stimulated under Th2-polarizing conditions in the presence of either anti-CTLA-4 or a isotype-matched control mAb. Total STAT6 and Pho-STAT6 levels were analysed by cell-based ELISA after 5 and 15 min re-stimulation with IL-4. Results (Fig. 4b) show that CTLA-4 engagement during polarization inhibits STAT6 phosphorylation, indicating that cells are rendered less responsive to IL-4 stimulation. The effect is not caused by a different kinetics in the response as STAT6 phosphorylation is similarly inhibited after 5 and 15 min re-stimulation. STAT6 protein levels are not affected by CTLA-4 engagement (Fig. 4c) as confirmed also by Western blot (not shown).

CTLA-4 engagement inhibits IL-4Rα chain expression

STAT6 activation requires IL-4R engagement by IL-4 which leads to IL-4Rα chain and common γ chain heterodimerization. IL-4Rα chain expression is regulated by different signalling pathways, including those activated by TCR and IL-4 itself, that lead to its up-regulation during Th2 cell differentiation.³⁴ IL-4Rα chain expression was examined by Western blot in CD4⁺ T cells stimulated under Th2 polarizing conditions for 1–5 days. Results show that IL-4Rα chain expression is low in naive CD4⁺ T cells and increases after stimulation. CTLA-4 engagement during polarization decreases IL-4Rα chain expression, the effect being evident after 1 day and maintained for all the polarization period (Fig. 5). This negative effect is due to a reduction in IL4Rα chain mRNA expression. Data obtained by RT and real time PCR, indeed, show that after 24 h stimulation, CTLA-4 engagement reduces IL-4Rα chain mRNA expression by 60%.

Figure 5.

CTLA-4 inhibits IL-4Rα chain expression. Lysates from cells stimulated under Th2 polarizing conditions in the presence of either a CTLA-4 engaging mAb or an isotype-matched control mAb for 1–5 days were separated by sodium dodecyl sulphate–polyacrylamide gel electrophoresis and immunostained with anti-IL-4Rα and anti-β-actin mAbs. Columns represent mean (± SE) of the ratios between volume report value for the band of the IL-4Rα chain blot and value for the corresponding β-actin control from three independent experiments.

Discussion

Upon priming, naive CD4⁺ T cells undergo differentiation and acquire effector functions. Th1 and Th2 cells, the two main effector cell types, promote cell-mediated and humoral immunity, respectively. As both cell subsets may be involved in pathological processes, Th cell differentiation is under the control of several factors to ensure that a protective not harmful immune response is developed. CTLA-4 has been described to affect the fate of naive CD4⁺ cells by promoting differentiation to Th1 cells²⁸ and inhibiting differentiation to Th2 cells.²⁵ The present findings, confirming and extending previous results from our²⁷ and other²⁵ groups, show that CTLA-4 engagement during cell polarization inhibits Th2 cell differentiation. This result is also supported by the observation that CTLA-4^−/− mice have CD4⁺ T cells with a predominant Th2 type cytokine pattern²¹ and high levels of serum IgE.¹⁹ Although the mechanisms through which CTLA-4 controls T-cell activation have been extensively investigated (reviewed in ^16–18), the effects of CTLA-4 engagement on transcription factors involved in Th2 cell differentiation have not been completely defined.

TCR engagement, costimulation and cytokines activate signalling pathways and transcription factors that synergize by promoting specific patterns of (cytokine) gene expression. Our results show that under neutral conditions both GATA-3 and T-bet mRNA expression are readily up-regulated in CD4⁺ T cells. Noteworthy, CTLA-4 engagement inhibits GATA-3 but not T-bet mRNA expression. GATA-3 is a transcription factor necessary for Th2 cell differentiation. By allowing histone acetylation GATA-3 induces changes in the chromatin structure of the Il-4 locus that lead to derepression, acquisition of DNase hypersensitivity and transcription of Th2 cytokine genes.^10,11 At variance, epigenetic modifications in the Ifn-γ gene are promoted by T-bet,¹³ which plays a crucial role in Th1 cell differentiation.¹² Lack of effects of CTLA-4 engagement on T-bet mRNA expression under Th1 polarizing conditions cannot be the result of a reduced, absent or delayed CTLA-4 expression as under these conditions CD4⁺ T cells express CTLA-4 at high levels as we have previously published.²⁷ During the initial (1–2 days) phase both GATA-3 and T-bet mRNA are up-regulated, increases in histone acetylation in both the Il4 locus and the Ifnγ promoter are induced^35,36 and both IL-4 and IFN-γ mRNA are expressed.³⁷ CTLA-4 by inhibiting GATA-3 but not T-bet mRNA expression might affect the induction of the epigenetic modifications which lead to up-regulation of Th2 and silencing of Th1 cytokine genes. Recently, it has been demonstrated that T-bet negatively regulates Th2 cell differentiation by a kinase-dependent binding to GATA-3. Phosphorylated T-bet sequesters GATA-3 away from its binding sites.³⁸ By reducing GATA-3/T-bet ratio, CTLA-4 might favour the sequestering effect of T-bet. The early phenomena in Th cell differentiation require TCR engagement but not stimulation by skewing cytokines as they are independent of STAT4 and STAT6.³⁹ The control exerted by CTLA-4 on TCR/CD28-induced GATA-3 activation has been recently shown also in vivo. CTLA-4^−/− STAT6^−/− double knock-out mice express GATA-3 mRNA at higher levels as compared to CTLA-4^+/+ STAT6^−/− mice indicating that lack of CTLA-4 control on TCR signals renders Th2 differentiation independent from the IL-4/STAT6 pathway.²⁶ However, CTLA-4^−/− STAT6^−/− mice show reduced Th2 responses as compared to CTLA-4^−/− STAT6^+/+ mice, pointing out the importance of both TCR/CD28 and IL-4/STAT6 pathways in Th2 cell differentiation.

The cytokine milieu is important to sustain prolonged histone acetylation of the relevant loci. STAT6^−/− mice indeed show compromised epigenetic modifications in the Il4 locus. Epigenetic modifications sustained by the cytokine milieu lead to stabilisation of the polarized phenotype and confer cytokine memory to cell progeny.^1,39,40 We found that under Th2-polarizing conditions CTLA-4 engagement inhibits GATA-3 mRNA expression as well as STAT6 phosphorylation and DNA binding. Under this condition, indeed, the inhibition of GATA-3 expression is even higher than under neutral condition. Our results suggest that post-transcriptional effects of CTLA-4 engagement on GATA-3 mRNA expression seem not to play a relevant role as GATA-3 mRNA turnover is not affected. In naive CD4⁺ T cells GATA-3 is expressed at a level insufficient to promote differentiation (stable epigenetic modifications). Upon stimulation with IL-4, GATA-3 expression is up-regulated and its level becomes sufficient to sustain its own expression.⁷ CTLA-4, by limiting STAT6 activation, might compromise the IL-4-induced GATA-3 auto-activation. STAT6 activation depends on the heterodimerization of IL-4Rα chain and common γ chain as induced by IL-4 binding. IL-4Rα chain expression is regulated by different signalling pathways, including those activated by TCR and IL-4.³⁴ Upon activation, naive CD4⁺ T cells up-regulate IL-4Rα chain expression. Our data show that under Th2 polarizing conditions IL-4Rα chain is readily up-regulated and kept at high level during the stimulation period. CTLA-4 engagement results in a compromised IL-4Rα chain expression. Our data, obtained by Western blot and by RT-PCR on cell lysates, show that it is the level of the IL-4Rα to be reduced, suggesting that CTLA-4 exerts a control at transcriptional level. To our knowledge this is the first evidence for a negative regulation of IL-4Rα chain expression by CTLA-4 and indicates a possible mechanism for the inhibition of STAT6 activation and GATA-3 mRNA up-regulation. In the presence of the anti-CTLA-4 mAb (when IL-4Rα expression is reduced), T-bet mRNA expression is lower under Th2 polarizing conditions than under neutral conditions. This finding suggests that GATA-3 and T-bet expression may be differently sensitive to the effects of IL-4 and/or that CTLA-4 might affect also other mechanisms involved in STAT6-dependent signals. It is known, indeed, that the IL-4/STAT6 responsiveness may be modulated positively⁴¹ and negatively⁴² by other receptors.

TCR/CD28 costimulation induces GATA-3 expression through the activation of NF-κB,^6,31 a transcription factor that is expressed at high levels upon TCR/CD28 costimulation⁴³ and synergizes with STAT6 to promote GATA-3 mRNA expression.³¹ We have previously shown that CTLA-4 engagement inhibits NF-κB activation,²⁹ a finding more recently confirmed in in vivo models.⁴⁴ The present findings show that CTLA-4 inhibits STAT6 activation. Thus, it is conceivable that CTLA-4 negatively affects GATA-3 mRNA by limiting the activation of these two relevant transcription factors.

In conclusion, our results show that CTLA-4 exerts a relevant control on Th2 cell differentiation by modulating the effects of both CD3/CD28 and IL-4R/STAT6 pathways on GATA-3 mRNA expression.