CD4⁺CD25⁺ T Cells Regulate Airway Eosinophilic Inflammation by Modulating the Th2 Cell Phenotype¹

In recent years, allergic asthma has considerably increased in prevalence worldwide. The disease is characterized by airway hyperreactivity (AHR)³and chronic mucosal inflammation mediated by CD4⁺ Th2 cells (1). Such events are associated with pulmonary eosinophilia, mucus hypersecretion, and airway remodeling (2). It has been suggested by several laboratories that the chronic inflammation evident in asthma arises as a consequence of a defect in immune regulation. However, the specific events that serve to down-regulate airway Th2-mediated inflammation are poorly understood. Certainly, the production of anti-inflammatory cytokines TGF-β or IL-10 and specific prostanoids at sites of inflammation serve to limit mucosal immune responses (3, 4). More notably, regulatory T cells producing high levels of IL-10 have recently been shown to modulate allergen-induced airway responses (5).

Regulatory T cells have been identified in mice and humans as a distinct population of CD4⁺ T cells that constitutively express the IL-2R α-chain (CD25) (6, 7). CD4⁺CD25⁺ T cells play an essential role in the maintenance of peripheral self-tolerance (8, 9) by preventing the activation and proliferation of autoreactive T cells that have escaped thymic deletion (10). The seminal finding that prompted speculation that these cells played a critical role in preventing autoimmunity was the observation that depletion of CD4⁺CD25⁺ T cells resulted in the development of organ-specific autoimmune disorders which could be prevented by the adoptive transfer of CD4⁺CD25⁺ T cells (6).

Following TCR engagement, CD4⁺CD25⁺ cells can suppress the activation and proliferation of other CD4⁺ and CD8⁺ T cells in an Ag-nonspecific manner (11, 12, 13). Murine CD4⁺CD25⁺ T cells mediate the suppression of effector T cell function both in vitro and in vivo via several mechanisms requiring either cell-cell contact (14) or the production of immunosuppressive cytokines such as IL-10 (15) and TGF-β (16). A role for glucocorticoid-induced TNFR in abrogating CD4⁺CD25⁺ T cell-mediated suppression has been proposed (17, 18). More recently, it was reported that the forkhead transcription factor Foxp3 is specifically expressed in CD4⁺CD25⁺ T cells and is required for their development. An important observation was that Foxp3 mutant scurfy and Foxp3-null mice developed a lethal autoimmune syndrome as a consequence of a deficiency in CD4⁺CD25⁺ regulatory T cells. Transfer of CD4⁺CD25⁺ T cells into neonatal Foxp3-deficient mice prevented the development of disease (19, 20).

CD4⁺CD25⁺ T cells have been demonstrated to prevent the onset of colitis, a Th1-mediated disease (21); however, their precise role in Th2-driven disease is unclear. In this study, we investigated whether Th2-mediated pulmonary inflammation is influenced by Ag-specific CD4⁺CD25⁺ T cells. In DO11.10 mice, 4–6% of CD4⁺ T cells constitutively expressed CD25 and coexpressed the transgenic TCR. Surprisingly, depletion of CD4⁺CD25⁺ T cells before Th2 differentiation profoundly reduced the production of Th2 cytokines. However, the CD4⁺CD25⁻-derived cells did not secrete IFN-γ, suggesting that they were indeed Th2 polarized. Unexpectedly, transfer of CD4⁺CD25⁻-derived Th2 cells into BALB/c mice resulted in a heightened pulmonary eosinophilic inflammation following OVA inhalation in these animals compared with recipients of total CD4⁺ Th2 cells. This eosinophilia was associated with increased levels of eotaxin but reduced amounts of IL-10 in the bronchoalveolar lavage (BAL). Our data demonstrate that the suppressive properties of CD4⁺CD25⁺ T cells do extend to Th2 responses and that these cells play a crucial role in modulating allergic lung inflammation.

DO11.10 TCR-transgenic mice (originally developed by Dr. D. Y. Loh, Howard Hughes Medical Institute, St. Louis, MO) were provided by Dr. E. Shevach (National Institutes of Health, Bethesda, MD). These animals were housed in microisolator cages under pathogen-free conditions at Southampton University (Southhampton, U.K.). BALB/c mice were obtained from Harlan (Loughborough, U.K.). All experiments were performed according to the Home Office guidelines.

To prepare CD4⁺CD25⁻ and CD4⁺CD25⁺ T cells, peripheral lymph node (PLN) cells from DO11.10 mice were first depleted of B cells by panning using anti-Ig Abs. Cells were then further depleted of both class II⁺ and CD8⁺ cells by a second round of panning using both M5/114 and YTS169.4 Abs (Serotec, Oxford, U.K.). The resultant CD4⁺ T cells were either left unfractionated or separated into CD25⁺ and CD25⁻ T cells by magnetic microbeads (Miltenyi Biotec, Bisley, Surrey, U.K.) using both biotinylated anti-CD25 Abs 3C7 and 7D4, labeled with avidin-FITC (BD PharMingen, San Diego, CA), and anti-FITC beads (Miltenyi Biotec). Flow cytometry was used to determine the purity of CD4⁺ T cells (>98%) and the fractionated CD4⁺CD25⁻ (>98%) or CD4⁺CD25⁺ T cells (>82%).

To drive T cell differentiation into Th2 effector phenotype, purified CD4⁺CD25⁻ or unfractionated (total) CD4⁺ T cells (5 × 10⁵/ml) were incubated for 4 days in the presence of OVA_323–339 peptide (1 μg/ml) and murine IL-4 (2 ng/ml; R&D Systems, Abingdon, U.K.) plus anti-IFN-γ Ab (5 μg/ml, R4-6A2; American Type Culture Collection, Manassas, VA). Irradiated (3000 rad) splenic APCs (depleted of CD4⁺ and CD8⁺ cells using complement) were also added to these cultures (5 × 10⁵/ml). After 4 days of incubation, cells were restimulated as before for another 4 days, but this time also in the presence of IL-2 (100 U/ml; Cetus, Emeryville, CA). Following culture for a total of 8 days, polarized effector Th2 cells prepared from either total CD4⁺ T cells or CD4⁺CD25⁻ T cells were adoptively transferred into BALB/c mice as described below. To drive T cell differentiation into a Th1 phenotype, PLN cells were incubated (5 × 10⁵/ml) in the presence of OVA_323–339 peptide (1 μg/ml), and mouse IL-12 (1 ng/ml; R&D Systems) plus anti-IL-4 Ab (5 μg/ml, 11B11; American Type Culture Collection). After 4 days of culture, cells were restimulated as before for another 4 days but this time also in the presence of IL-2 (100 U/ml; Cetus). After 8 days, Th1 cells were analyzed for cytokine expression by real-time RT-PCR.

The limited numbers of CD4⁺CD25⁺ T cells in DO11.10 mice made a complete analysis of the properties of these cells difficult. Consequently, purified CD4⁺CD25⁺ T cells (5 × 10⁵/ml) were expanded for 8 days as described above in the presence of APCs, OVA peptide, anti-IFN-γ Ab and either IL-2 alone (added throughout the 8-day culture) or IL-2 plus IL-4. The expansion of CD4⁺CD25⁺ T cells in the presence of exogenous IL-2 was limited (30-fold increase in cell numbers over 8 days) compared with that observed in the presence of IL-2 plus IL-4 (80-fold increase over 8 days). The CD4⁺CD25⁺ T cells expanded in IL-2 plus IL-4 were used for experiments to examine suppressor function. We were able to maintain these regulatory T cells for up to 10 days in culture.

To monitor the effects of CD4⁺CD25⁺ cells on T cell proliferative responses, either freshly isolated (day 0) or expanded (day 8) CD4⁺CD25⁺ T cells were added at various concentrations to DO11.10 PLN cells (2 × 10⁵) and the proliferation in response to immobilized anti-CD3 (2 μg/ml) or OVA peptide (1 μg/ml) was determined after 48 h by [³H]thymidine incorporation.

To examine cytokine production, 8-day polarized cells (5 × 10⁵/ml) were stimulated with immobilized anti-CD3 (2 μg/ml) for 24 h, and the supernatants were harvested for measurement of IL-4, IL-5, IL-13, and IFN-γ by ELISA as described previously (4, 22). For IL-10 measurement by ELISA, JES5-2A5 (BD PharMingen) was used as capture Ab and biotinylated polyclonal anti-IL-10 Ab for detection (PeproTech, Rocky Hill, NJ).

DO11.10 Th2 cells derived from either CD4⁺CD25⁻ or total CD4⁺ T cells were injected i.v. into BALB/c mice (10⁷cells/mouse). Mice (four to six per group) were then intranasally challenged by exposure to aerosolized solutions of OVA (0.5%, Grade V; Sigma-Aldrich, Poole, U.K.) for 20 min a day over 7 consecutive days using a Wright’s nebulizer (Buxco Europe, Petersfield, U.K.). Control mice were exposed to OVA aerosols but did not receive DO11.10 Th2 cells. AHR was measured on day 7 in response to methacholine inhalation by whole-body plethysmography (Buxco Europe, Petersfield, U.K.). Animals were placed in chambers and exposed to nebulized PBS (baseline) followed by increasing concentrations of methacholine. Enhanced pause (Penh) was measured after each 3-min exposure. Mice were killed on day 8, and BAL fluid was collected for analysis. Lung tissue was dispersed by collagenase (Sigma-Aldrich) and the resultant lung mononuclear cells (LMCs) were stimulated with OVA peptide or anti-CD3 for 24 h. Cytokine production by the lung cells was measured by ELISA (as described above) and eosinophil peroxide (EPO) levels present in the LMCs were determined by colorimetric analysis as previously described (22). Macrophages were not depleted from LMC preparations since we have previously shown in this model that interstitial macrophages suppress IL-2 production and the associated T cell proliferation but do not affect Th2 cytokine expression (22).

Nonlavaged lungs were obtained and one part of the lung tissue was fixed in 10% Formalin, embedded in paraffin, and then stained with H&E. Small samples (3 mm²) from another part of the lung were fixed in acetone (containing protease inhibitors), embedded in glycol methacrylate, and then sections (2 μm) were stained with biotinylated anti-clonotypic Ab KJ1-26 and avidin-peroxidase.

BAL was performed by cannulating the trachea of each animal and washing the airways with 3 × 0.5 ml of PBS to collect BAL fluid. BAL fluid of four animals was pooled and EPO levels present in BAL cells were determined by colorimetric analysis as described before (22). Cell differential percentages were determined by light microscopic evaluation of stained cytospin preparations and expressed as absolute cell numbers. Levels of cytokines IL-4, IL-5, IL-10, IL-13, IFN-γ, and the chemokine eotaxin in the BAL were measured using sensitive commercially available ELISA kits (all from BioSource International, Camarillo, CA, except IL-4 and eotaxin which are from R&D Systems), according to the manufacturers’ instructions.

In certain experiments, 8-day expanded DO11.10 CD4⁺CD25⁺ T cells (5 × 10⁶ cells/mouse) were cotransferred into BALB/c recipients simultaneously with CD4⁺CD25⁻-derived Th2 cells (10⁷cells/mouse). The level of airway eosinophilia was assessed as described above.

Cells were stained and analyzed either on a FACSAria (BD Biosciences, San Diego, CA) using FACSDiVa software for performing three-color analysis to enumerate CD4⁺ T cells (using GK1.5-APC-Cy7), OVA-specific T cells (using KJ1-26-FITC) and CD25⁺ T cells (using 7D4-PE), or on a FACSCalibur (BD Biosciences) using CellQuest software to enumerate CD4⁺ T cells (GK1.5-PE; BD PharMingen), clonotypic T cells (KJ1-26-FITC), and CD25⁺ T cells (anti-7D4-biotin and avidin-FITC).

RNA was extracted from Th1- and Th2-polarized cells or purified CD4⁺CD25⁺ T cells (following stimulation with immobilized anti-CD3) using TRIzol reagent (Life Technologies, Renfrewshire, U.K.). Total RNA (2 μg) was then reverse transcribed using Omniscript II (Qiagen, Crawley, U.K.) at 37°C for 1 h using oligo(dT)₁₅ as a primer and the cDNA was then PCR amplified and quantified using the TaqMan technique (Applied Biosystems, Warrington, U.K.). Real-time PCR was performed using the Perkin-Elmer AB1 Prism 7700 Sequence Detection System (Applied Biosystems). The expression of GAPDH (housekeeping gene), IL-4, IL-5, IL-10, IL-13, and IFN-γ was determined (4, 23). Equal amounts of cDNA were used in triplicate and amplified with the TaqMan master mix according to manufacturer’s instructions (Applied Biosystems). Thermal cycling conditions were 2 min at 50°C and 10 min at 95°C, followed by 40 cycles of two-step PCR consisting of 15 s at 95°C and 1 min at 60°C. Threshold cycle was measured as the cycle number at which the reporter fluorescent emission increased above a threshold level. The amount of mRNA was expressed as fold difference relative to the amount obtained from unstimulated control cells. Amplification efficiencies were validated and normalized against GAPDH. For all samples, total RNA that was not reverse transcribed was also analyzed to determine genomic DNA contamination, which was negligible.

Data are summarized as means ± SEM. Data obtained from adoptive transfer experiments were analyzed using the Mann-Whitney U test, and differences were considered statistically significant with p < 0.05.

CD4⁺CD25⁺ regulatory T cells have been demonstrated to play a critical role in preventing organ-specific autoimmune diseases (6). We used DO11.10 mice to model allergic lung inflammation and investigated whether a Th2 inflammatory response is influenced by Ag-specific CD4⁺CD25⁺ T cells. We first assessed the characteristics of CD4⁺CD25⁺ T cells in nonimmunized DO11.10 mice. Consistent with previous reports, FACS analysis revealed that 4–6% of CD4⁺ T cells from DO11.10 mice constitutively expressed CD25 (Fig. 1,A), whereas in BALB/c mice 10–15% of CD4⁺ T cells express CD25 (11, 24, 25). It is likely that the expression of the nonautoreactive transgenic TCR contributes to the reduced number of CD4⁺CD25⁺ regulatory T cells present in DO11.10 mice (26). The limited numbers of DO11.10 CD4⁺CD25⁺ cells made a complete analysis of the properties of these cells difficult and necessitated the expansion of these cells in vitro over 8 days. The majority of freshly isolated (day 0) and expanded (day 8) CD4⁺CD25⁺ or CD4⁺CD25⁻ T cells were OVA specific since they expressed the transgenic TCR i.e., KJ1-26⁺ (Fig. 1, A and B). Interestingly, both newly isolated and 8-day expanded CD4⁺CD25⁺ T cells exhibited regulatory function by suppressing proliferative responses of effector T cells to anti-CD3 or OVA_323–339 peptide in a dose-dependent manner (Fig. 1 C). The inhibition elicited by OVA peptide was less potent than that observed with anti-CD3 stimulation. This may arise because anti-CD3 is a more effective stimulus, being less dependent on APCs. Second, the anti-CD3 used was immobilized on a plate and may thus exert a cross-linking function between regulatory and nonregulatory CD4⁺ T cells. Such a bridging effect could potentially augment suppressor function.

To evaluate whether regulatory T cells suppressed Th2 responses, we initially determined the effect of depleting CD4⁺CD25⁺ T cells on Th2 cell polarization in vitro. To this end, DO11.10 CD4⁺CD25⁻ or unfractionated (total) CD4⁺ T cells were differentiated for 8 days in the presence of IL-2 and IL-4 and the expression of cytokines in response to anti-CD3 stimulation was then determined. The level of proliferation of CD4⁺CD25⁻-derived and total CD4⁺ T cells were similar (80-fold increase in cell numbers over 8 days). Unexpectedly, 8-day polarized CD4⁺CD25⁻-derived Th2 effector cells secreted significantly less IL-4, IL-5, IL-10, and IL-13 than total CD4⁺ Th2 cells in response to immobilized anti-CD3 (2 μg/ml) stimulation for 24 h (Fig. 2,A). Moreover, the difference in Th2 cytokine production was maintained using different concentrations of immobilized anti-CD3, and the addition of accessory cells had no effect on the level of cytokines produced by these polarized cells (data not shown). Interestingly, no IFN-γ was secreted by either type of Th2-differentiated cells (Fig. 2,A). The lack of IFN-γ production by CD4⁺CD25⁻-derived Th2 cells implies that their reduced level of Th2 cytokine production did not arise as a consequence of incomplete Th2 polarization. The expression of mRNA transcripts for IL-4, IL-5, IL-10, and IL-13 were also markedly reduced in CD4⁺CD25⁻-derived Th2 cells compared with total CD4⁺ Th2 cells (Fig. 2,B). This suggests that the difference in the level of Th2 cytokine production is a consequence of reduced transcription rather than consumption of the cytokine. In comparison, Th1 cells and CD4⁺CD25⁺ regulatory T cells expressed small or negligible amounts of IL-4, IL-5, and IL-13 mRNA (regulatory T cells did express IL-10 transcripts). Neither CD4⁺CD25⁻-derived, or total CD4⁺ Th2 cells expressed any IFN-γ mRNA (Fig. 2 B). Thus, depletion of CD4⁺CD25⁺ T cells before Th2 differentiation resulted in markedly reduced cytokine expression by the polarized cells.

We next studied allergic pulmonary inflammation by adoptively transferring DO11.10 Th2 cells, generated from unfractionated CD4⁺ T cells, into BALB/c mice (10⁷ cells/animal) which were then exposed to OVA aerosols for 7 consecutive days. To investigate how regulatory T cells influence airway inflammation, CD4⁺CD25⁺ T cells were removed before Th2 polarization, and the Th2 cells prepared from CD4⁺CD25⁻ T cells were also transferred into BALB/c mice. Following exposure to OVA aerosols, a pronounced peribronchial and perivascular eosinophilic inflammation and an increase in KJ1-26⁺ T cells was observed in the lung parenchyma of recipients of Th2 cells generated from either CD4⁺CD25⁻ or total CD4⁺ T cells compared with animals that did not receive any cells (Fig. 3,A). We next examined the influx of inflammatory cells into the BAL. Paradoxically, recipients of CD4⁺CD25⁻-derived Th2 cells displayed a striking increase in the number of eosinophils compared with recipients of total Th2 cells (Fig. 3,B). Consistently, there was a significant rise in the level of EPO activity in the BAL of animals injected with CD4⁺CD25⁻-derived Th2 cells (Fig. 3,C) and in LMCs of these animals (14.8 ± 2.6 ng/ml EPO for CD4⁺CD25⁻ Th2 recipients vs 7.4 ± 1.7 ng/ml for total CD4⁺ Th2 recipients). However, the AHR was not significantly elevated in recipients of CD4⁺CD25⁻-derived Th2 cells compared with recipients of total CD4⁺ T cells (Fig. 3 D).

Interestingly, when monitoring pulmonary inflammation at earlier time points, a lag in the onset of pulmonary eosinophilia mediated by Th2 cells derived from CD4⁺CD25⁻ T cells was observed. However, invariably the eosinophilic inflammation observed in mice that have received CD4⁺CD25⁻-derived Th2 cells was markedly more intense following 7 days of aerosol exposure (Table I). The delayed kinetics in the onset of eosinophilia by CD4⁺CD25⁻ Th2 cells is likely to reflect decreased IL-5 production by these cells and the consequent slower maturation of eosinophil precursors. The cotransfer of expanded CD4⁺CD25⁺ regulatory T cells along with CD4⁺CD25⁻-derived cells did not reduce the airway inflammatory response (data not shown). Thus, although removal of CD4⁺CD25⁺ T cells before Th2 polarization resulted in a Th2 phenotype with reduced cytokine expression, these cells were effective at inducing a heightened airway eosinophilic inflammation following OVA inhalation for 7 days.

Since DO11.10 Th2 cells prepared from CD4⁺CD25⁻ T cells secreted lower amounts of IL-4, IL-5, and IL-13 in vitro, yet caused an elevated airway eosinophilia, we examined levels of Th2 cytokines expressed in vivo in the BAL and lung tissue. The levels of IL-13 were significantly lower in the BAL of recipients of CD4⁺CD25⁻-derived Th2 cells compared with recipients of total CD4⁺ Th2 cells, whereas the amounts of IL-4 and IL-5 were only marginally reduced (Fig. 4,A). Moreover, BAL levels of the anti-inflammatory cytokine IL-10 were also decreased in recipients of CD4⁺CD25⁻-derived Th2 cells (Fig. 4,B). No IFN-γ was present in the BAL of these two experimental groups (data not shown). Interestingly, although there was a reduction in cytokine production in mice that have received Th2 cells prepared from CD4⁺CD25⁻ T cells, the amount of the chemokine eotaxin present in the BAL was elevated in these animals (Fig. 4 C).

To determine the cytokine production by DO11.10 T cells recruited to the lung, the lung tissue was enzymatically digested and the production of cytokines by LMCs was assessed after restimulation with OVA_323–339 peptide or immobilized anti-CD3 for 24 h. Following OVA inhalation, there was a reduced production of both IL-4 and IL-5 by stimulated LMCs from animals injected with CD4⁺CD25⁻-derived Th2 cells compared with recipients of total CD4⁺ Th2 cells (Fig. 5). However, the frequency of KJ1-26⁺ T cells in the lungs of these mice, as assessed by FACS, was comparable (2.5% in recipients of total CD4⁺ cells, vs 2.6% in recipients of CD4⁺CD25⁻ cells). Negligible IFN-γ was secreted by LMCs from these animals in response to OVA peptide stimulation. These results suggest that the OVA-specific CD4⁺CD25⁻-derived Th2 cells retained their Th2 phenotype in vivo and their reduced cytokine production. In contrast, anti-CD3 stimulation did elicit an increase in IFN-γ production by LMCs from recipients of CD4⁺CD25⁻-derived Th2 cells, presumably reflecting increased secretion of the Th1 cytokine by the host T cells as a consequence of reduced IL-4 secretion by DO11.10 T cells present (i.e., cross-regulation).

Allergic asthma is characterized by airway hyperresponsiveness and mucosal inflammation mediated by CD4⁺ Th2 cells. That these events arise as a consequence of a defect in immune regulation is implied from the observation that lung mucosal immune responses are normally tightly regulated (22). However, the mechanisms that transpire to resolve, or possibly prevent, airway mucosal Th2-mediated inflammation remain poorly defined. In this context, both TGF-β and IL-10 have been shown to regulate mucosal inflammation (3). In addition, PGI₂ generated during allergic pulmonary inflammation serves to selectively limit the progression of Th2, but not Th1, responses (4).

Recent work has shown that following Ag inhalation, CD4⁺CD25⁺ regulatory T cells play a key immunomodulatory role (5). Such suppressor T cells have been shown to prevent autoimmunity (6, 27) and the development of colitis induced by Helicobacter hepaticus (3, 28). CD4⁺CD25⁺ T cells regulate both CD4⁺ and CD8⁺ T cell responses, partly by inhibiting IL-2 production, the subsequent clonal expansion of T cells, and the development of memory (11, 12). In addition, it has been proposed that both IL-10 and TGF-β can mediate suppression in vivo. However, regulation can take place in the absence of either of these cytokines (29). In this instance, cell-cell contact forms a prerequisite for mediating immune regulation (11, 14) which can be reversed by Abs to glucocorticoid-induced TNFR (17).

It has been reported that Th1 responses are more prone to regulation by CD4⁺CD25⁺ T cells than Th2 responses (30). Nevertheless, CD4⁺CD25⁺ T cells can suppress Th2 maturation (31), possibly by inhibiting IL-4 production (32) and the development of pulmonary eosinophilic inflammation (33). The mechanism by which regulatory T cells mediate these effects remains unresolved. This is, in part, due to the difficulty in obtaining Ag-specific CD4⁺CD25⁺ T cells.

In our study, we used the DO11.10 TCR-transgenic mouse to examine the role of Ag-specific regulatory T cells in a model of allergic pulmonary inflammation. In DO11.10 mice, 4–6% of CD4⁺ T cells constitutively express CD25, which was consistent with previous findings (25, 34). Our analysis revealed that the majority of CD4⁺CD25⁺ T cells in these animals express the transgenic TCR as defined by staining with the KJ1-26 Ab. The low frequency of CD4⁺CD25⁺ T cells in DO11.10 mice made it necessary to expand these cells over 8 days in culture before they could be characterized in detail. Although these cells are anergic, they could be rapidly expanded by stimulation with OVA_323–339 peptide in the presence of exogenous IL-2 and IL-4. Our results showed that freshly isolated and 8-day expanded CD4⁺CD25⁺ T cells exhibited regulatory function by suppressing proliferative responses of DO11.10 T cells to OVA_323–339 peptide or anti-CD3 stimulation. Consequently, the regulatory activity of these cells was not notably affected by expansion in the presence of Ag or IL-4.

To investigate whether CD4⁺CD25⁺ T cells regulated Th2-mediated pulmonary inflammation, it was important to evaluate how these cells influenced either the initial Th2 polarization stage or the subsequent development of allergic inflammation in vivo. The role of CD4⁺CD25⁺ T cells in modulating these processes was examined separately by monitoring the effect of depleting these cells on both Th2 differentiation in vitro and the inflammation elicited in vivo. CD4⁺CD25⁺ T cells had to be removed before polarization of DO11.10 T cells for 8 days, since activated T cells express CD25. CD4⁺CD25⁺ T cells, both freshly isolated or expanded in the presence of IL-2 plus IL-4, expressed only low levels of IL-4, IL-5, and IL-13 mRNA, implying that these cells do not differentiate into Th2 cells. Unexpectedly, depletion of regulatory T cells markedly reduced the cytokine production by Th2 cells generated in vitro. Most notably, the expression of mRNA and protein for the cytokines IL-4, IL-5, IL-10, and IL-13 by Th2 cells derived from CD4⁺CD25⁻ T cells was markedly lower than that generated from total CD4⁺ T cells. This effect was observed irrespective of the level of anti-CD3 stimulation used to induce cytokine secretion and the addition of accessory cells had no effect. Such differences did not seem to reflect different levels of proliferation or the amount of IL-2 present given that exogenous IL-2 was added and the level of proliferation was similar. Conceivably, CD4⁺CD25⁺ T cells may influence either the maturation of Th2 cells directly or act by inhibiting Th1 development (i.e., cross-regulation). Using the DO11.10 model of airway inflammation, it has been suggested that regulatory T cells favor Th2 polarization by inhibiting Th1 development (35). However, in our experiments, the Th2 cells prepared from CD4⁺CD25⁻ T cells did not produce any detectable IFN-γ, even after adoptive transfer into mice that were then repeatedly exposed to OVA aerosols. This implies that the reduced cytokine expression by CD4⁺CD25⁻-derived Th2 cells was not simply a consequence of incomplete Th2 polarization or a shift toward Th1 maturation

We next examined whether the inflammatory response mediated by CD4⁺CD25⁻-derived Th2 cells in vivo was similarly compromised. Surprisingly, when transferred into BALB/c hosts, the Th2 cells prepared from CD4⁺CD25⁻ T cells were capable of eliciting a pronounced increase in the number of eosinophils present in the BAL following OVA inhalation for 7 days compared with that observed following injection of total CD4⁺ Th2 cells. However, AHR was not significantly different between the two groups. This observation is consistent with the recent findings of Hadeiba and Locksley (36) that CD4⁺CD25⁺ regulatory T cells suppress type 2 immune responses but not AHR. It is possible that CD4⁺CD25⁺ T cells are effective at regulating effector Th2 responses but are unable to curtail certain events in vivo that involve other cell types that influence AHR. The increased inflammation did not appear to be a consequence of more effective expansion or recruitment of DO11.10 T cells in recipients of CD4⁺CD25⁻-derived Th2 cells, since the number of KJ1-26⁺ T cells in the airways was not notably different from that observed following transfer of total CD4⁺ Th2 cells. Stimulation of LMCs with OVA_323–339 peptide revealed that the lung cells from recipients of CD4⁺CD25⁻ Th2 cells consistently produced lower amounts of IL-4 and IL-5 when compared with recipients of total CD4⁺ Th2 cells. However, the frequency of CD4⁺KJ1-26⁺ T cells in the lungs of these two groups of mice was comparable. Thus, the Th2 phenotype characterized by reduced cytokine production appeared to remain stable in vivo even after repeated exposure of the animals to OVA aerosols. Anti-CD3 stimulation, however, did elicit an increase in IFN-γ production by LMCs from recipients of CD4⁺CD25⁻-derived cells, presumably reflecting increased secretion of this cytokine by the host T cells as a consequence of reduced IL-4 secretion by DO11.10 T cells present (i.e., cross-regulation). Consistent with the reduced cytokine expression by the CD4⁺CD25⁻ Th2 cells, there was a marked decrease in IL-13 levels in the BAL of recipients of these cells. The amounts of IL-4 or IL-5 were only marginally lower, possibly reflecting the different cellular provenance of the cytokines or production by host T cells.

The observation that CD4⁺ helper T cells producing low levels of cytokine are particularly effective in vivo has been previously reported (37). These authors demonstrated that Th1 cells could be categorized with respect to the level of IFN-γ they produce, with cells deficient in secreting IFN-γ surviving for long periods in vivo. Such IFN-γ⁻ cells retained their Th1 phenotype in vivo and were effective Th1 memory cells (37). Conceivably a similar classification may exist for CD4⁺ Th2 cells, with such memory-type responses being suppressed by CD4⁺CD25⁺ regulatory T cells. Alternative explanations for the reduced cytokine expression could be that removal of CD4⁺CD25⁺ T cells resulted in an increase in the proportion of anergic T cells, or that the regulatory T cells themselves may have a positive influence on cytokine production by Th2 cells. The effectiveness of CD4⁺CD25⁻-derived Th2 cells in vivo thus suggests that these cells possess additional proinflammatory characteristics that compensate for an apparent deficiency in their Th2 cytokine production. In this context, an increased level of eotaxin, a potent chemoattractant of eosinophils in allergic inflammation (38), was observed in the BAL of mice that have received CD4⁺CD25⁻-derived Th2 cells. Moreover, this was associated with a decrease in IL-10 production in the BAL of these mice. Whether reduced IL-10 levels were directly responsible for elevated eotaxin release in the airways is unclear. Certainly, IL-10 has been suggested to inhibit the production of a range of chemokines (39). The cotransfer of cultured CD4⁺CD25⁺ T cells was found to be unable to reverse the severe eosinophilic inflammation elicited by CD4⁺CD25⁻-derived Th2 cells. Consequently, it would seem unlikely that heightened inflammation elicited by the CD4⁺CD25⁻-derived Th2 cells arises because this response is not modulated by Ag-specific CD4⁺CD25⁺ T cells in vivo.

In summary, CD4⁺CD25⁺ T cells play a key role in regulating airway eosinophilic inflammation. Our results demonstrate that the immunomodulatory properties of CD4⁺CD25⁺ T cells do extend to Th2 responses, most notably by limiting the development of a proinflammatory CD4⁺ Th2 phenotype characterized by reduced cytokine production. An understanding of the regulation of Th2 responses in vivo could provide better insight into the design of novel approaches to modulate the chronic airway inflammatory reaction evident in bronchial asthma.

We thank Drs. Ethan Shevach and Peter Kilshaw for valuable discussions and gratefully acknowledge the Biomedical Imaging and Histology Research Units (Southampton General Hospital) for their technical assistance.