The levels of CD4+CD25+ regulatory T cells in paediatric patients with allergic rhinitis and bronchial asthma

Authors


Dr Bor-Luen Chiang, Department of Pediatrics, National Taiwan University Hospital, 7 Chung-Shan South Road, Taipei, Taiwan, China.
E-mail: gicmbor@ha.mc.ntu.edu.tw

Summary

Our purpose was to determine whether numbers of CD4+CD25+ T [T regulatory (Treg)] cells and mRNA expression of functional molecules of Treg are related to airway allergy and disease severity in 51 paediatric patients with allergic rhinitis or bronchial asthma and 47 healthy controls. Surface markers were evaluated with flow cytometry, and mRNA was determined with real-time polymerase chain reaction. Children with allergic disease had fewer CD4+CD25+ T cells (8·49% ± 2·41% versus 9·58% ± 2·43%, P < 0·05) and CD4+CD25hi T cells (1·32% ± 0·68% versus 1·70% ± 0·68%, P < 0·01) than control subjects. Numbers of CD4+CD25+ and CD4+CD25hi T lymphocytes were higher in children with persistent allergic rhinitis and/or moderate–severe bronchial asthma than in those with respective milder disease. The number of Treg cells was correlated positively with total immunoglobulin E level. The mRNA expression of forkhead box P3 (FoxP3) was increased in moderate–severe versus mild asthma (2·93 ± 0·38 versus 1·60 ± 0·31, P < 0·01). Patients with moderate–severe bronchial asthma also had increased mRNA expression of interleukin (IL)-10 compared with patients with mild asthma (15·24 ± 4·07 versus 3·77 ± 2·18, P < 0·01). The suppressive function of Treg cells from patients with more severe asthma was competent in vitro. On average, decreased numbers of Treg cells in children with allergic airway disease might represent a defect of the Treg population. With increased expression of FoxP3 and IL-10 in Treg from patients with relatively severe allergic disease, adaptive and functional Treg might be generated in response to aggravated atopy and disease severity.

Introduction

Allergic diseases are chronic inflammatory disorders characterized by an aberrant immune response to harmless environmental antigens. They are defined by excessive production of immunoglobulin E (IgE) against specific allergens (allergen sensitization), such as house dust mites, and can lead to diseases including bronchial asthma (BA), allergic rhinitis (AR) and atopic dermatitis (AD) [1]. The differentiation of CD4+ T cells to effectors that produce a T helper 2 (Th2) profile of cytokines, such as interleukin (IL)-4, IL-5, IL-9 and IL-13, is the initial event and the key factor for the development of allergic responses [1,2]. Triggering an immune response by means of allergen-specific IgE on the surface of mast cells and basophils can lead to immediate symptoms due to the release of histamine and other mediators, while eosinophilic airway inflammation contributes to nasal symptoms and/or airway hyperresponsiveness in patients with allergic rhinitis or bronchial asthma [1]. Persistent airway inflammation and remodelling are believed to underlie the chronic functional and pathologic abnormalities, as well as the intermittent and episodic clinical manifestations, of allergic rhinitis and bronchial asthma [3].

As a minor fraction of approximately 5–10% of CD4+ T cells in unmanipulated animals and healthy humans, CD4+CD25+ regulatory T cells (Treg) play a critical role in preventing organ-specific autoimmunity and allograft rejection and in maintaining self-tolerance [4]. These cells express constitutively high levels of the α chain of the interleukin (IL)-2 receptor (CD25) and are produced in the thymus as a functionally mature T cell subpopulation (naturally occurring Treg, or nTreg). Treg cells express exclusively the forkhead-winged helix transcription factor gene, FoxP3 (forkhead box P3), which appears to be a master control gene for the development of these cells [5–7]. The acquisition of regulatory phenotype and suppressive function depends on the expression of FoxP3 [7]. Ectopic expression of FoxP3 in non-regulatory CD4+CD25 T cells converts them into Treg cells that acquire a regulatory function [5–9]. Other subsets of Treg, called adaptive or induced Treg, have also been described. These are largely T cell populations induced by in vitro or in vivo manipulation [10,11]. Adaptive or induced Treg in the setting of allergic response are represented by CD4+CD25+FoxP3+ cells generated from naive conventional CD4+ T cells (CD4+CD25FoxP3 cells) after polyclonal or antigen-specific stimulation [12,13]. By all criteria measured, these adaptive or induced Treg are indistinguishable from natural Treg[8,13].

The role of Treg in the pathogenesis of allergic disease and atopy was not defined until recently. Patients who specifically lack CD4+CD25+ Treg (e.g. those with immune dysregulation, polyendocrinopathy, enteropathy or X-linked syndrome, a syndrome caused by mutations in FoxP3) develop severe eczema, elevated IgE levels, eosinophilia and food allergy [14]. Treg may block the transition from the early activation stage to the differentiated T helper 2 (Th2) state, limit airway allergic inflammation and act to prevent inappropriate Th2 responses to environmental allergens [15,16]. Several independent lines of evidence indicate that the number or function of Treg is impaired or altered in patients with allergies compared with healthy individuals [17–22]. However, data concerning the role of Treg in the pathogenesis of paediatric allergic disease are rare.

The mechanism of immunological regulation by which Treg cells mediate their effect on other activated T cells is an area of active investigation and ongoing controversy and may involve a spectrum of regulatory roles in several disease states. Suggested mechanisms include an essential role of cytokines, cell–cell interaction, modifications of antigen-presenting cells by Treg cells and competition with naive T cells for specificity to the same antigens in adhesion to antigen-presenting cells [23,24]. Candidate molecules responsible for the immunosuppressive function of Treg cells include inhibitory cytokines IL-10 and transforming growth factor (TGF)-β, those of the B7 family interacting with cytotoxic T lymphocyte-associated antigen 4 (CTLA-4, a negative co-stimulatory molecule) and the glucocorticoid-induced tumour necrosis factor family-related receptor (GITR) and its related signal transduction. Although activated CD4+ effector T cells also express GITR and CTLA-4, the highly constitutive levels of these two molecules on Treg make them functionally related to Treg[25,26].

We aimed to investigate the role of Treg in the pathogenesis of paediatric allergic disease by exploring relationships among FoxP3 expression, CD25 expression and the Treg population during allergic inflammation. We sought to determine whether numbers of Treg cells and mRNA expression of molecules functionally related to CD4+ Treg cells are associated with atopic status and/or disease severity in paediatric patients with allergic rhinitis or bronchial asthma.

Materials and methods

Study groups

We enrolled 51 children with allergic airway disease, including 27 with allergic rhinitis and 24 with bronchial asthma. We also enrolled 47 age-matched, healthy control subjects without allergy. Table 1 shows the subjects' demographic features and medication histories.

Table 1.  Demographic and clinical characteristics of study participants.
 ControlsAR interAR persistBA mildBA mod-sev
  1. AR inter: intermittent AR; AR, persist: persistent AR; BA, mild: mild intermittent + mild persistent BA; BA, mod-sev: moderate persistent + severe persistent BA; FEV1: forced expiratory volime in 1 second. Data are expressed as mean ± standard deviation. P < 0·01 compared with intermittent AR. *P < 0·01 compared with mild BA. P < 0·05 compared with mild BA.

Case number4714131212
Age (years) 6·4 ± 2·3 6·1 ± 2·4 5·9 ± 2·7 6·6 ± 2·6 6·8 ± 3·0
Sex (M/F) 25/22 7/7 6/7 7/5 6/6
Inhaled/oral steroid (no.)1439
Inhaled/oral α2-agonist (no.)611
%FEV1100·1 ± 12·5 96·7 ± 15·0 92·3 ± 13·8 95·3 ± 17·5 78·7 ± 15·6
Total IgE (kU/l) 68·8 ± 28·6107·5 ± 98·8192·4 ± 168·9282·0 ± 222·0560·9 ± 390·2*

Allergic rhinitis was diagnosed if an IgE-mediated response induced nasal itching, sneezing, watery rhinorrhoea and/or nasal stiffness after allergen sensitization, as confirmed by the presence of IgE antibodies to specific allergens in the patient's serum. Intermittent allergic rhinitis was defined as symptoms occurring on fewer than 4 days per week and for less than 4 weeks. If symptoms were more frequent than this, allergic rhinitis was classified as persistent [27]. Patients with rhinitis of infectious and/or inflammatory, occupational, drug-induced, hormonal or other non-allergic causes were excluded. Patients with comorbidities of asthma, sinusitis, otitis media or structural abnormalities were also excluded.

Bronchial asthma was diagnosed as IgE-mediated chronic airway inflammation and increased airway hyperresponsiveness that led to recurrent episodes of expiratory wheezing, breathlessness, dry coughing and/or chest tightness, particularly at night or in the early morning [28]. Allergen sensitization was documented as the presence of serum allergen-specific IgE antibodies. Standardized lung function testing was performed in all patients with asthma by using spirometry to measure their forced expiratory volume in 1 second (FEV1). Airway hyperresponsiveness was evaluated by administering bronchoprovocation challenges with inhaled methacholine, as described previously [29]. The severity of asthma was classified as mild intermittent, mild persistent, moderate persistent or severe persistent [28]. We divided our asthmatic patients into two groups: mild group (mild intermittent + mild persistent) and moderate–severe group (moderate persistent + severe persistent). Patients with respiratory conditions similar to asthma, such as bacterial or viral infections, anatomic abnormalities and foreign bodies, were excluded. Patients with chronic rhinitis and/or sinusitis or gastro-oesophageal reflux, both comorbidities associated with allergic rhinitis, were also excluded.

Blood samples were collected from subjects during their visits to the out-patient clinics and/or when they were hospitalized for an exacerbation of symptoms. Informed consent was obtained from all subjects' parents or guardians, and approval was obtained from our institutional review board.

Antibodies

The following monoclonal antibodies to human cell-surface molecules were purchased (Becton-Dickinson Immunocytometry Systems, San Jose, CA, USA): fluorescein isothiocyanate-conjugated anti-CD25, phycoerythrin-conjugated anti-CD69, phycoerythrin-conjugated anti-CD122, peridinin chlorophyll protein-conjugated anti-CD4, fluorescein isothiocyanate-conjugated mouse immunoglobin G1 isotype control, phycoerythrin-conjugated mouse immunoglobin G1 isotype control and peridinin chlorophyll protein-conjugated mouse immunoglobin G1 isotype control.

Flow cytometric analysis

Fluorochrome-conjugated sets of monoclonal antibodies were used to stain T cell surface markers. Optimal conditions for monoclonal-antibody concentrations and incubation periods were established as recommended by the manufacturer. In brief, a 50-µl aliquot of whole blood was incubated with the fluorochrome-conjugated monoclonal antibody at room temperature in the dark for 30 min. It was then incubated with 50 µl of fluorescence-activated cell sorter lysing solution (Becton-Dickinson Immunocytometry Systems) for 15 min, washed twice with 2 ml of phosphate-buffered saline containing 0·1% bovine serum albumin and fixed with 1% paraformaldehyde.

Fluorocytometry was performed with a flow cytometer (FACSCalibur; Becton-Dickinson Immunocytometry Systems) equipped with a 488-nm blue laser and a 635-nm red diode laser for multicolour fluorescence, in addition to forward-scatter and side-scatter measurements. Laser and photomultiplier parameters were kept constant for all experiments. All fluorocytometric data were subsequently analysed and displayed graphically using CellQuest Pro software (Becton Dickinson Immunocytometry Systems). In each analysis, we measured a minimum of 20 000 cells.

Preparation of peripheral blood mononuclear cells

Heparinized venous whole blood from each donor was layered onto a Ficoll-Plaque density gradient (Amersham Pharmacia Biotech, Piscataway, NJ, USA) and centrifuged for 30 min at 500 g at room temperature. The layer of peripheral blood mononuclear cells was collected, washed and resuspended in Hanks's balanced salt solution (Gibco-BRL Life Technologies, Gaithersburg, MD, USA). Cell viability was greater than 95%, as determined by performing a trypan blue exclusion assay under an optical microscope.

Isolation of CD4+ T cells

Human peripheral blood mononuclear cells were centrifuged for 10 min at 500 g at room temperature and resuspended in RPMI-1640 medium (BioWhittaker Inc., Walkersville, MD, USA) containing 10% heat-inactivated fetal calf serum, 40 µmol/l l-glutamine, 100 U/ml penicillin, 100 U/ml streptomycin and 20 mmol/l HEPES buffer solution (Gibco-BRL Life Technologies). CD4+ T cells were purified by means of negative selection with immunomagnetic beads (by depleting CD8+, CD11b+, CD16+, CD19+, CD36+ and CD56+ cells) by using a human CD4+ T cell isolation kit (Miltenyi Biotec, Gladbach, Germany). Starting with 1–2 × 107 peripheral blood mononuclear cells, we usually isolated 4·5–5·0 × 106 CD4+ T cells, with a CD4+ T cell purity of more than 90%. All steps were performed according to the manufacturer's instructions.

Extraction of RNA, first cDNA synthesis and quantitative real-time polymerase chain reaction (PCR)

Total mRNA was extracted by using a mRNA isolation kit (GeneStrips; RNAture, Irvine, CA, USA) according to the manufacturer's protocol. In brief, isolated subsets of T cells were placed into 40 µl of lysis buffer, transferred to hybridization tubes, shaken for 60–90 min and finally washed with 100 µl of wash buffer. Next, 50–500 ng of mRNA was reverse-transcribed in a 50-µl reaction volume containingprimers (Random Primers; Promega, Madison, WI, USA), reverse transcriptase (SuperScript II; Invitrogen Corp., Carlsbad, CA, USA) and ribonuclease inhibitor (RNaseOUT; Invitrogen) in the same hybridization tube. The primers and probes used in the real-time PCR for FoxP3, IL-10, TGF-β, GITR and CTLA-4 were purchased (Assays-on-Demand Gene Expression Products; Applied Biosystems, Foster City, CA, USA) and glyceraldehyde-3-phosphate dehydrogenase (GAPDH) used as an endogenous control.

Each real-time PCR reaction system contained 1–6 µl of cDNA, 10 µl of TaqMan MasterMix (Applied Biosystems), and diethyl pyrocarbonate–H2O was added to make a final desired volume of 20 µl. Real-time PCR reactions for each individual gene were performed in triplicate, and results were analysed by means of the dual-labelled fluorogenic probe method with a sequence detector (ABI Prism 7700; Applied Biosystems). The mRNA expression levels of each gene were calculated by using the 2−ΔΔCT (comparative threshold cycle, or CT) method, as detailed by the manufacturer (Technical Bulletin 2; Applied Biosystems). The efficiencies of the target (FoxP3, IL-10, TGF-β, GITR and CTLA-4) and reference (GAPDH) were approximately equal, as proved in individual validation experiments. Quantities of target gene expression in the test samples were normalized to the corresponding GAPDH mRNA level of a healthy control subject (calibrator).

In vitro proliferation assay

CD4+CD25+ T cells were isolated by using a human CD4+CD25+ regulatory T cells isolation kit (Miltenyi Biotec) from peripheral blood mononuclear cells collected as described previously. Starting with 5 × 107 peripheral blood mononuclear cells we usually isolated 3 × 105 CD4+CD25+ T cells, with a purity of more than 95%.

To compare the suppression of polyclonal antigen-driven cultures between paediatric patients with moderate–severe airway allergy and normal controls, we cultured freshly isolated CD4+ cells alone, CD4+CD25 cells alone or mixed CD4+CD25 and CD4+CD25+ T cells at a ratio of 2 : 1 at 1 × 105 cells per ml in a final volume of 200 µl in the presence of 10 µg/ml soluble anti-CD3 monoclonal antibodies (Immunotech, Marseille, France) plus 10 µg/ml soluble anti-CD28 monoclonal antibodies (Immunotech) in triplicate in RPMI-1640 medium containing 10% heat-inactivated fetal calf serum, 40 µmol/l of l-glutamine, 100 U/ml of penicillin, 100 U/ml of streptomycin and 20 mmol/l of HEPES buffer solution (Gibco brl Life Technologies) in 96-well round-bottomed plates for 72 h at 37°C in a 5% CO2 humidified atmosphere. In all cases, we included triplicate medium control wells as a negative control.

We added [3H]-thymidine (37 KBq/well) (Amersham, Chalfont St Giles, UK) for the final 18 h of culture, after which cells were harvested, and incorporated radioactivity was counted as an index of proliferation. We expressed data as counts per minute (cpm) in stimulated wells with medium control wells subtracted.

Statistical analysis

Data were expressed as the mean ± standard deviation, unless specified otherwise. Flow cytometric data were compared statistically among the groups by using the Mann–Whitney U-test. Expression of different mRNAs among groups, as measured during real-time PCR, was analysed with Student's t-test. The correlation coefficient r was generated by using Spearman's rank correlation. We compared suppression of proliferation between groups with analysis of variance (anova) with Prism software (GraphPad, San Diego, CA, USA). A P-value of less than 0·05 was considered to indicate a statistically significant difference.

Results

Characteristics of study participants

As shown in Table 1, study subjects in all groups were matched for age and sex. Patients with moderate–severe bronchial asthma had a relatively high rate of using inhaled and/or oral steroids and β-agonists. Lung function results as %FEV1 and serum total IgE levels differed significantly among the bronchial asthma groups. Patients with persistent allergic rhinitis had serum total IgE values significantly higher than those of patients with intermittent allergic rhinitis.

Numbers of CD4+CD25+ and CD4+CD25hi T cells in all patients with allergic airway disease

CD4+CD25hi T cells were CD4+ T cell subsets with high or bright CD25 surface expression and were defined accordingly [30].

Children with allergic airway disease had significantly decreased numbers of CD4+CD25+ T lymphocytes compared with control subjects (8·49% ± 2·41% versus 9·58% ± 2·43%, P < 0·05) (Fig. 1a). Numbers of CD4+CD25hi T cells were also decreased significantly in patients compared with control subjects (1·32% ± 0·68% versus 1·70% ± 0·68%, P < 0·01) (Fig. 1b).

Figure 1.

Comparison of numbers of CD4+CD25+ (a) and CD4+CD25hi (b) T cells between control subjects and patients with subgroups of airway allergy. Transverse line marks the median. AR, inter = intermittent allergic rhinitis; AR, persist = persistent allergic rhinitis; BA, mild = mild intermittent or persistent bronchial asthma; and BA, mod-sev = moderate–severe persistent bronchial asthma.

Numbers of Treg cells in airway allergy of different severities

Children with persistent allergic rhinitis had more Treg cells than did patients with intermittent allergic rhinitis (9·52% ± 2·07% versus 6·51% ± 1·11%, P < 0·001) (Fig. 1a). Similarly, children with moderate–severe bronchial asthma had more Treg lymphocytes than patients with mild bronchial asthma (10·76% ± 2·12% versus 7·40% ± 1·64%, P = 0·01) (Fig. 1a).

The number of CD4+CD25hi T cells was elevated in patients with persistent allergic rhinitis compared with patients with intermittent allergic rhinitis (1·48% ± 0·63% versus 0·96% ± 0·40%, P < 0·01) (Fig. 1b). Patients with moderate–severe bronchial asthma had more CD4+CD25hi T lymphocytes than patients with mild bronchial asthma (1·73% ± 0·90% versus 1·18% ± 0·54%, P < 0·05) (Fig. 1b).

Surface expression of CD69 and CD122 on Treg cells

To determine the phenotypic characteristics of our Treg study population, patients with serum total IgE levels greater than 150 kU/l were selected for investigation of the surface expression of CD69 and CD122 on CD4+CD25+ T cells (Fig. 2). Seventeen patients (seven patients with persistent allergic rhinitis and 10 patients with moderate–severe bronchial asthma) and 15 matched control subjects were included in this comparison. Patients with allergic airway disease did not have increased CD69 expression on Treg cells compared with control subjects, with values of 3·50% ± 2·64% versus 2·74% ± 2·46% (P > 0·05). Patients with allergic airway disease had insignificantly higher CD122 expression on Treg cells than did control subjects, with values of 33·3% ± 18·6% versus 24·1% ± 11·8% (P > 0·05).

Figure 2.

The surface expression of CD122 and CD69 on CD4+CD25+ T cells between control subjects (n = 15) and patients with airway allergy (seven patients with persistent allergic rhinitis and 10 patients with moderate–severe bronchial asthma, n = 17). A transverse line marks the median. AR, allergic rhinitis; BA, = bronchial asthma.

Correlation of total serum IgE level with the number of Treg cells

To investigate further the association of atopic status with the number of Treg, serum total IgE levels were obtained from 51 patients. The number of Treg cells was correlated moderately with total serum IgE level (= 0·34, P < 0·05) (Fig. 3a). However, the number of CD4+CD25hi T cells was not correlated significantly with total serum IgE level (= 0·16, P > 0·05) (Fig. 3b).

Figure 3.

Correlation of numbers of CD4+CD25+ T cells and CD4+CD25hi T cells with serum total IgE level (presented in log10 scale). The number of CD4+CD25+ T cells was correlated positively with the total IgE level (= 0·34, P = 0·02).

Expression of mRNA for FoxP3 in airway allergy of different severities

We compared levels of mRNA expression for FoxP3 in eight patients with intermittent allergic rhinitis, eight patients with persistent allergic rhinitis receiving inhaled steroids, eight patients with mild bronchial asthma, eight patients with moderate–severe bronchial asthma receiving oral and/or systemic steroids and eight control subjects.

On average, mRNA expression of FoxP3 was higher in patients with bronchial asthma than in patients with allergic rhinitis. Patients with persistent allergic rhinitis had higher mRNA expression of FoxP3 than that of patients with intermittent allergic rhinitis (1·34 ± 0·34 versus 0·66 ± 0·26, P < 0·05). Patients with moderate–severe bronchial asthma also had higher mRNA expression of FoxP3 than that of patients with mild bronchial asthma (2·93 ± 0·38 versus 1·60 ± 0·31, P < 0·01) (Fig. 4).

Figure 4.

Comparison of mRNA expression of forkhead box P3 (FoxP3) between patients with allergic airway disease with different severities. Expression of mRNA is represented by folds relative to the calibrator. All data (mean ± s.d.) are representative of experiments in triplicate of eight individuals. †P < 0·01 compared with moderate–severe persistent bronchial asthma; *P < 0·05 compared with persistent allergic rhinitis. AR, inter = intermittent allergic rhinitis; AR, persist = persistent allergic rhinitis; BA, mild = mild (intermittent and persistent) bronchial asthma; BA, mod-sev = moderate–severe persistent bronchial asthma.

Of interest, mRNA expression of FoxP3 was similar between patients with intermittent allergic rhinitis and healthy control subjects (0·66 ± 0·26 versus 0·78 ± 0·34, P > 0·05). However, patients with persistent allergic rhinitis, patients with mild bronchial asthma and patients with moderate–severe bronchial asthma all had significantly higher mRNA expression of FoxP3 than that of control subjects (P < 0·05, P = 0·01, and P < 0·01, respectively) (Fig. 4).

Expression of mRNA for IL-10 and TGF-β in asthma of different severities

We compared mRNA expression levels for IL-10 and TGF-β in eight patients with mild bronchial asthma, eight patients with moderate–severe bronchial asthma receiving oral and/or systemic steroids and eight control subjects. Expression of mRNA for IL-10 was higher in patients with mild bronchial asthma (3·77 ± 2·18) and in patients with moderate–severe bronchial asthma (15·24 ± 4·07) than in control subjects (1·55 ± 0·78, P < 0·05 and P < 0·01, respectively). Patients with moderate–severe bronchial asthma also had significantly increased mRNA expression for IL-10 compared with patients with milder disease (P < 0·01) (Fig. 5a).

Figure 5.

Comparison of mRNA expression for interleukin-10 (a) and transforming growth factor-β (b) in patients with bronchial asthma of different severities. Expression of mRNA is represented by folds relative to the calibrator. All data (mean ± s.d.) are representative of experiments in triplicate of eight individuals. †P < 0·01 compared with moderate–severe persistent bronchial asthma; *P < 0·05 compared with moderate–severe persistent bronchial asthma. BA, mild = mild intermittent and persistent bronchial asthma; and BA, mod-sev = moderate–severe persistent bronchial asthma.

Expression of mRNA for TGF-β was lower in patients with mild bronchial asthma (0·62 ± 0·41) compared with control subjects (1·30 ± 0·63, P < 0·05). Patients with moderate–severe bronchial asthma had higher mRNA expression of TGF-β (1·30 ± 0·63) than that of patients with mild bronchial asthma (P < 0·05) (Fig. 5b). Comparison between control subjects and patients with moderate–severe disease was not significantly different (P > 0·05).

Expression of mRNA for GITR and CTLA-4 on CD4+ T cells

We examined CD4+ T cells from eight patients with moderate–severe bronchial asthma and eight control subjects to compare mRNA expression for GITR and CTLA-4 with CD4 T cells. In patients with bronchial asthma and in control subjects, GITR expression was elevated in CD4+ T cells compared with CD4 T cells; respective values were 2·88 ± 0·75 versus 1·02 ± 0·61 in patients with bronchial asthma (P < 0·01) and 2·87 ± 1·02 versus 0·87 ± 0·64 in control subjects (P < 0·01). The mRNA expression for GITR in CD4+ T cells did not differ between patients with bronchial asthma and control subjects (P > 0·05) (Fig. 6a).

Figure 6.

The mRNA expression for glucocorticoid-induced tumour necrosis factor family-related receptor (a) and cytotoxic T lymphocyte-associated antigen 4 (b) between patients with bronchial asthma and normal controls. Expression of mRNA is represented by folds relative to the calibrator. All data (mean ± s.d.) are representative of experiments in triplicate of eight individuals. †P < 0·01 compared with CD4+ T cells; *P < 0·01 compared with moderate–severe persistent bronchial asthma. BA = moderate–severe persistent bronchial asthma.

In patients with bronchial asthma and in control subjects, CTLA-4 expression was greater in CD4+ T cells than in CD4 T cells; respective values were 1·31 ± 0·21 versus 6·34 ± 3·83 × 10−2 in patients with bronchial asthma (P < 0·01) and 0·72 ± 0·22 versus 6·07 ± 5·60 × 10−2 in control subjects (P < 0·01). Levels of mRNA expression for CTLA-4 in CD4+ T cells were significantly higher in patients with bronchial asthma than in control subjects (P < 0·01) (Fig. 6b).

Suppression of proliferation of CD4+CD25 T cells by CD4+CD25+ T cells

We determine the suppressive capacity of CD4+CD25+ T cells between three patients with moderate–severe bronchial asthma and three control subjects. As shown in Fig. 7a, the average suppression of proliferation of CD4+CD25 by CD4+CD25+ T cells were not significantly different between these two groups (P > 0·05). The suppressive responses were dose-related, i.e. the degree of suppression increased as the number of CD4+CD25+ T cells increased. However, we noticed a change point (at ratio of 1/16) in the suppressive curves, indicating that the percentage suppression of proliferation by CD4+CD25+ T cells was significantly greater in patient groups than in control subjects (P < 0·01) at this point. After that point, the dose-related effects of CD4+CD25+ T cells were not prominent in the asthmatic group.

Figure 7.

(a) Dose-related suppression of proliferation of CD4+CD25 T cells by different ratios of CD4+CD25+ T cells from asthmatic patients (diamonds with solid lines; n = 3) and control subjects (squares with dashed lines; n = 3). Proliferation responses were determined by [3H]-thymidine uptake, as represented by counts per minute (cpm). *P < 0·01 compared with normal controls, determined by analysis of variance. BA, mod = moderate–severe persistent bronchial asthma. All data are representative of experiments in triplicate from three individuals. (b) Comparison of proliferative capacity of CD4+ T cells alone, CD4+CD25 T cells alone and CD4+CD25+ T cells alone between asthmatic patients and control groups. *P < 0·01 compared with proliferative response of CD4+CD25 T cells alone.

Figure 7b showed the comparison of proliferative capacities between CD4+ T cells alone, CD4+CD25 T cells alone and CD4+CD25+ T cells alone. For both asthmatic patients and control groups, the proliferative responses of CD4+ T cells and CD4+CD25+ T cells were reduced significantly compared with those of CD4+CD25 T cells (P < 0·01, respectively). The difference between the proliferative responses of CD4+CD25 T cells from patients and those from controls was not significant (P > 0·05).

Discussion

Paediatric patients with airway allergy had, on average, significantly decreased numbers of CD4+CD25+ T cells and CD4+CD25hi T cells. Patients with more severe allergic airway disease had more CD4+CD25+ T cells and CD4+CD25hi T cells than did patients with less severe disease. Numbers of CD4+CD25+ T were correlated positively with serum total IgE levels. Patients with bronchial asthma with more severe disease had higher mRNA expression for FoxP3 and IL-10 than patients with milder disease. The suppression of proliferation of CD4+CD25 T cells by CD4+CD25+ T cells from patients with more severe asthma were functionally competent in vitro.

The frequency of Treg cells in the peripheral CD4+ T cells has become an important parameter for characterizing populations of Treg cells [4]. The reported frequency of Treg cells in peripheral blood from patients with allergy is inconsistent (Table 2) [17–22]. The difference in these observed frequencies might be related to the patients' age and disease status. However, the relationship between the number of Treg cells and disease severity has not been emphasized. To our knowledge, only Shi et al. reported that numbers of Treg cells increased during an exacerbation of asthma [20]. This finding was consistent with our results.

Table 2.  Summary of literatures concerning the role of Treg from patients with allergic airway disease.
StudyAge (years)Patient numberDiseasesPhenotype of Treg measuredFrequency of Treg inPBSuppression of Th2 response*FoxP3 expression
  1. ↑: Indicates increased; ↓: decreased, ≅: equal frequency of CD4+CD25+ T cells, respectively. +: Indicates the occurrence of the corresponding function. †Expressed as a percentage relative to total CD4+ T cells. Suppression of effector T proliferation and suppression of Th2 cytokine production. n.a. data not available. PB: peripheral blood. FoxP3: forkhead box P3; TH2: T helper 2; Treg: T regulatory cells.

Akdis [17]Mean, 3617Pollen or mite allergyIL-10-producing CD4+ (Tr1)+n.a.
Ling [18]20–5823Atopic 12, hayfever 11CD4+CD25++
Karagiannidis [19]n.a.50Asthma moderate–severeCD4+CD25+↑ after steroidn.a.↑ mRNA after steroid
Shi [20]26–3915 acute exacerbation +14 stableAsthmaCD4+CD25+↑ in exacerbation; ≅ in stablen.a.
Bellinghausen [21]n.a.9Allergic rhinoconjuntivitis, asthmaCD4+CD25+n.a.
Grindebacke [22]19–5023Birch-allergic rhinoconjuntivitisCD4+CD25+, CD4+CD25hin.a.

The severity of disease was associated with rising levels of CD25 and CD25hi cells, with a level almost equal to that of healthy subjects (Fig. 1). This observation seemed contradictory to the consensus that defective Treg cells are associated with the pathogenesis of allergy. Our results also revealed that the number of Treg cells was correlated with atopic status, as determined by measuring serum total IgE values, a phenomenon not observed in healthy subjects (Fig. 3). We speculated initially that Treg cells from subjects with relatively severe disease contained fewer Treg and more activated effector T cells during allergic inflammation, as recently activated effector T cells also express CD25. We confirmed the characteristics of Treg cells by finding low levels of CD69 expression and moderate levels of CD122 expression on Treg cells from patients with relatively severe airway allergy. CD122 was expressed by an estimated 85% of CD4+CD25hi T cells, serving as a marker for Treg[30]. Therefore, we speculated that the high levels of CD25 and CD25hi cells from patients with severe disease were composed mainly of Treg. Because of the increased expression of FoxP3 and IL-10 and the unimpaired suppressive function of CD4+CD25+ T cells, these Treg might represent the induced or adaptive Treg generated during an exacerbation of allergic inflammation as a consequence of an immune response [11]. Further research should include identification of these Treg with a specific intracellular anti-FoxP3 monoclonal antibody to understand how many of the CD4+CD25+ or CD4+CD25hi T cells are effective Treg cells [31].

That FoxP3 programmes the development and suppressive function of Treg cells is well established [5–7]. We observed increased FoxP3 expression in patients with relatively severe allergic rhinitis or bronchial asthma compared with patients with milder disease. Again, our results seemed inconsistent with the rationale that severe disease should be associated with profound defects in FoxP3 expression. However, rare studies evaluate the suppressive capacity of Treg in allergic disease in parallel with the expression of FoxP3 in patients with allergy (Table 2). Studies have demonstrated that the suppressive capacity of Treg in human allergic disease is not impaired [18,21,22]. Our in vitro suppressor assay was in agreement with this conclusion. One possible interpretation of our results was that the acquisition of adaptive Treg acts as a form of feedback to regulate conventional antigen-specific immune responses [11,12,32]. In this model, FoxP3 functions not as a Treg-lineage specialization factor but as a transcriptional regulator of an immunosuppressive effecter programme [11,12]. Because we analysed FoxP3 expression in CD4+ T cell populations, the observed increase in FoxP3 expression may have been due to an expansion of a small population of FoxP3+ Treg cells or to a conversion of peripheral non-regulatory CD4+ T cells to Treg. We speculate that persistent antigen stimulation and frequent airway inflammation were associated with the increased FoxP3 expression and subsequent increase in the Treg cell subset observed in our patients with severe disease. However, the phenomenon we observed might have been related to steroid use among patients with relatively severe disease (Table 1). FoxP3 expression and the frequency of occurrence of Treg cells is reported to increase after the administration of glucocorticoids [11]. Although the regulatory/suppressive capacity and FoxP3 expression in Treg from allergic patients were not impaired [21,33], the ‘change point’ in our in vitro suppressive curve might imply that the suppressive capacity was somewhat different between more severe asthmatic children and normal subjects.

In the context of allergic disease, IL-10 is of particular interest. It inhibits proinflammatory cytokine production and both Th1 and Th2 cell activation. It also impairs the activation of mast cells and eosinophils and promotes the synthesis of IgG4 [34]. Resolution of allergen-induced airway inflammation and hyperreactivity related to the transfer of allergen-specific Treg cells in vivo depends on the production of IL-10 [16]. Our mRNA analysis of CD4+ T cells showed high IL-10 expression in patients with severe bronchial asthma; this implied that IL-10 expression is associated with aggravated allergic inflammation. Combined with the data of increased FoxP3 expression in moderate–severe bronchial asthma, we speculate that some IL-10-producing CD4+CD25+FoxP3+ cells are generated during aggravated allergic inflammation. Although most data suggest that Treg cells are not major producers of IL-10, Treg might possibly boost their capacity for IL-10 production.

As for TGF-β, it plays a dual role in allergic disease. It is important in inducing regulatory T cells and participating directly in suppression of effector T cells, and also mediates remodelling of airway tissues [35]. Whether asthma benefits from TGF-β-mediated suppression of specific immune responses or whether TGF-β-mediated tissue remodelling aggravates diseases more than it helps to control immune reactions is unclear [35]. Our real-time PCR results showed lower TGF-α expression in milder bronchial asthma and intact TGF-α expression in moderate–severe bronchial asthma. Such a discrepancy may imply that modification of expression is associated with the dual role of TGF-β in allergic inflammation of asthma. The clinical implications must be determined.

The suppressor capacities of Treg were up-regulated and down-regulated after stimulation by CTLA-4 and GITR, respectively [24,36]. GITR is expressed preferentially at high levels on Treg cells and plays a key role in the peripheral tolerance that Treg cells mediate [25]. The role of GITR in the pathogenesis of allergic disease is unclear. In allergic and non-atopic donors alike, Treg almost all express GITR [37]. The similar expression of GITR between our patients and control subjects implies that GITR expression might not be impaired in paediatric patients with allergy. The constitutive expression of CTLA-4 was also restricted primarily to Treg cells, and the immune-suppressive function of Treg cells in vivo depends on CTLA-4 expression [26]. The role of CTLA-4 on Treg cells in allergic disease has been reported [38]. Our data demonstrated that mRNA levels for CTLA-4 in CD4+ T cells were higher in patients with bronchial asthma than in control subjects. This result indicated that increased CTLA-4 expression is associated with paediatric allergic airway disease.

From a literature review concerning human airway allergy, relationships among FoxP3 expression, CD25 expression and suppressor capacity of the Treg population during allergic inflammation remain inconclusive. Does a subset of T cells up-regulate FoxP3 and then increase the numbers of Treg cells during allergic inflammatory responses? Data from animal and human studies confirm the existence of adaptive Treg. However, few studies have been conducted to explore such possibilities with human studies in vivo.

In conclusion, paediatric patients with airway allergy had, on average, decreased numbers of Treg. This suggests that decreased numbers of Treg cells might represent a defect in the Treg cell population in children with allergic airway disease. However, the high numbers of Treg observed during severe allergy (in terms of total IgE level and disease severity) might represent a subset of adaptive Treg (with increased FoxP3 and IL-10 expression and competent suppressive function) generated during exacerbated allergic inflammation as part of the immune response.

Ancillary