Differences in allergen-induced T cell activation between allergic asthma and rhinitis: Role of CD28, ICOS and CTLA-4

Four groups of patients were recruited: allergic rhinitics (R), allergic rhinitics and asthmatics (AR), non allergic asthmatics (A), and controls (C). All allergic patients were selected to display house dust mite (HDM) allergy. As rBetv1 birch pollen allergen was used as control antigen for in vitro stimulation of T cells, patients were selected to be not sensitized to birch pollen. The diagnosis of HDM allergy was determined by positive skin prick test to Dermatophagoides pteronyssinus extract (Stallergenes, France). Allergic rhinitis was defined by the presence of perennial nasal symptoms out of viral infection such as nasal obstruction, sneezing, rhinorrhea and nasal pruritus. The diagnosis of asthma was done on the basis of a history of dyspnea and wheezes with a reversible obstructive ventilatory defect or a positive methacholine challenge. The distinction between mild and moderate asthma was done according to GINA classification [19]. In patients, any inhaled corticosteroids and anti-histamines were discontinued 15 days before sampling. As controls, healthy non smoker individuals with normal lung function, negative methacholine challenge and negative skin prick test were included. In controls, absence of allergy was established by the negativity of 35 skin prick tests to common environmental aeroallergens, and absence of asthma was stated on negative methacholine challenge and induced sputum eosinophil count below 3% (see additional file 1: Skin testing, methacholine challenge and induced sputum procedures). The positive methacholine test was defined by a drop of at least 20% of FEV1(forced expiratory volume in 1 second) in response to 200 μg or less of metacholine. This project was approved by the local Ethic Committee and written informed consent was obtained from each patient.

Peripheral blood mononuclear cells (PBMC) were isolated from peripheral venous blood by Ficoll-Hypaque plus (GE Healthcare, Uppsala, Sweden) density gradient centrifugation. Cells were then washed three times and resuspended in complete medium RPMI-1640 supplemented with 10% (v/v) foetal calf serum (FCS), 2 mM L-glutamine, 1 mM sodium pyruvate, 1 μM 2-mercapto-ethanol (Sigma Chemical, Saint-Louis, Missouri), 1000 U/ml Penicillin and Streptomycin. All culture reagents, except 2-mercapto-ethanol, were purchased from GIBCO^®.

Recombinant (r) Betv 1 of birch pollen (Betula verrucosa) and purified (p) Derp 1 of house dust mite (Dermatophagoides pteronyssinus) were provided by Stallergènes (Antony, France). None of the allergens contained detectable amounts of LPS.

Optimal dose of stimulatory pDerp1 and kinetics of T cell cytokine secretion and proliferation were determined in an independent pilot study on 5 house dust mite allergics and 5 healthy volunteers.

PBMC (5 × 10⁵) were cultured in 96 wells plates (Falcon) in 100 μl medium containing 1 μg/ml pDerp1 at 37°C in 5%CO₂ and cells were harvested after 8 days culture. 50 μl of fresh complete medium was added every 2 days in each well. rBetv1 was used as control antigen at a concentration of 1 μg/ml.

After 8 days of culture with pDerp 1, PBMC (5 × 10⁵) were harvested and stained with anti-CD4-PE-Cy5, anti-CD25-FITC, (Beckman Coulter, Marseille, France); anti-CD3-FITC (Dako, Trappes, France), anti-CD3-PE-Cy5 (Immunotools, Friesoythe, Germany), anti-CD28-FITC, anti-ICOS-PE, or anti-CTLA-4-PE-Cy5 (BD Pharmingen, le Pont de Claix, France) mAbs at recommended concentrations. To detect Foxp3 intracellular transcription factor, T cells were then fixed, permeabilized, and stained with anti-Foxp3-PE mAb (eBiosciences, San Diego, California). The Treg population was identified as CD4+CD25^Hi+Fox p 3+ cells.

Fluorescence was detected with a 15 mW argon ion laser on a three colors FACSCan^® (Becton Dickinson, Franklin Lakes, NJ, USA). Standard acquisition and analysis software were obtained through Cellquest^® Software (Becton Dickinson).

PBMC (5 × 10⁵) were cultured for 8 days with pDerp 1. PMA (Sigma Chemical, Saint-Louis, Missouri, 50 ng/ml), Ionomycin (Euromedex, 2 μg/ml) and Monensin (Sigma Chemical, 2 μM) were added during the last 6 hours of culture. These culture conditions allow the detection of cytokines already engaged in a synthesis process in vivo [20]. Cells were harvested and stained with CD3-PE-Cy5 (Immunotools, Friesoythe, Germany). Cells were then fixed, permeabilized, and stained with antibodies to detect intracellular cytokines (anti-IFNγ-FITC, anti-IL-4-FITC, BD Pharmingen, le Pont de Claix, France; anti-IL13-PE, anti-IL-10-PE, R&D system, Lille, France). IL-4+ and IL-13+ cells were considered as Th2 cells, IFN-γ + cells as Th1 cells. IL-10+ cells were considered as belonging to Treg cell population.

To determine the role of co-receptors in T cell activation, PBMC cultures were performed with or without anti-CTLA-4 (clone 14D3, 12 μg/ml), anti-ICOS (clone ISA-3, 12 μg/ml) or anti-CD28 (clone CD28.6, 3 μg/ml) monoclonal antibodies (mAb). These mAb were purchased from eBioscience.

Analysis was performed using the Statview^® Software. Normal distributions of the variables were checked with a Kolmogorov-Smirnof's test. Average percentages of positive cells and cytokine concentrations were then compared between groups (controls, non allergic asthmatics, allergic rhinitics and allergic asthmatics) using the analysis of variance (ANOVA). When the ANOVA showed statistical difference between groups, a multiple linear regression analysis was done to identify if allergy, asthma, or both could explain the variable studied. Between-groups comparisons were performed using a Student's t-test. A paired t test was used to compare differences between paired groups. A p value < 0.05 was considered as statistically significant for all statistical tests. Results are expressed as mean ± standard error (SE).

Sixty-nine subjects (33 males, 36 females, mean age 37.20 ± 1.90) were included. Blood samples from 20 healthy individuals with no history of allergy or asthma, 18 allergic asthmatics (AR), 18 allergic rhinitics (R), and 13 non allergic asthmatics (A) were collected. Characteristics of the patients are shown in table 1.

None of the subjects was a smoker. Patients interrupted their local or systemic steroids or antihistamines 15 days before sampling. Asthmatics were mild asthmatics for one half and moderate asthmatics for the other half. All allergic patients displayed symptoms compatible with allergic rhinitis. All non allergic asthmatics also complained from nasal symptoms. Healthy volunteers did not report any symptom.

Sputum eosinophil counts were significantly higher in asthmatics than in control subjects or allergic rhinitis, with no significant difference between allergic and non allergic asthmatics. None of the subjects was sensitized to birch. The age difference between the A+R group and other groups (A, R and C) was not significant statistically.

Treg cells proportion, Th1 and Th2 cytokines production and co-receptors expression (CTLA-4, ICOS, CD28) in each group were first assessed by flow cytometry, prior to any specific stimulation.

In non-stimulated conditions, CTLA-4+ T cells were decreased in asthmatics (p < 0.05 vs controls, figure 1A), whatever their allergic status. In keeping with this result, a reduced Treg population (p < 0.025, figure 1B) was found in these patients. Relevantly, Treg cell proportions were higher in mild asthmatics than in moderate counterparts (p < 0.012, figure 1C). IFN-γ + cells were increased (p < 0.022 vs controls, figure 1D) in asthmatics. No significant difference in Th2 cytokines or IL-10 production was found (table 2) between groups.

ICOS expression was higher in R compared to controls (p = 0.029, figure 1E), but similar in AR and controls. No significant variation was found at the level of CD28 expression between groups (table 2).

The multiple linear regression analysis showed that asthma (A + AR) was associated to lower ICOS and CTLA-4 expression and Treg cell proportions, but to higher IFN-γ+ T cells (table 3). By contrast, allergic rhinitis (with or without asthma) was positively linked to ICOS expression.

PBMCs were cultured in the presence or not of pDerp1 during 8 days. T cell activation and co-receptors expression were then studied by flow cytometry.

In AR, Der p 1 up-regulated CD28 (89.78 ± 1.33 vs 91.01 ± 1.48; p = 0.0016) and ICOS expression, and decreased CTLA-4 (figure 2A). Furthermore, Derp1 stimulation induced an increase in IL-4+ and IL-13+ cells (figure 2A), without significant variation in IFN-γ+ cells (not shown). This increase in Th2 cells was associated to a decrease in IL-10+ cells and Treg cells (figure 2A).

In R, Derp1 also increased CD28 (89.03 ± 1.71 vs 91.00 ± 1.48; p = 0.0025) but not ICOS expression (figure 2B). It decreased CTLA-4+ cell proportions. Allergen stimulation induced an increase in Th2 cells without variation of IFN-γ + cells (not shown), and a decrease in IL10+ and Treg cells (figure 2B).

Therefore at the exception of ICOS, that was already increased at baseline in R and thus could not increase upon stimulation, the profile of T cell activation and co-receptor expression induced by Derp1 was similar in AR and R subjects.

After specific stimulation (figure 2C-D), T cells from asthmatic and non asthmatic allergics displayed higher expression of ICOS (p < 0.02) and lower expression of CTLA-4 compared to controls (p < 0.007). In addition Th2 cell proportions were higher in allergics whereas Treg cells were decreased (IL-4, p < 0.0022; IL-13, p < 0.0001; Treg, p < 0.008). CD28+ cell percentages were not different between groups after allergen-specific stimulation (not shown). In non allergic subjects (figure 2C-D) no significant variation was found in any of the parameters studied

The multiple linear regression analysis showed that after Derp 1 specific stimulation, allergy (R + AR) correlated positively with percentages of ICOS, IL-13 and IL-4-expressing T cells and negatively with CTLA-4 and IL-10-expressing T cells (table 3).

No variation was found in any subject for any co-receptor or cytokine expression after stimulation with irrelevant rBetv1 (not shown).

In order to study the respective role of CD28, ICOS and CTLA-4 in T cell activation patterns in the context of allergen presentation, PBMC were stimulated with Derp1 in the absence or presence of anti-ICOS, anti-CTLA-4 or anti-CD28 mAb.

In allergics, whatever the asthmatic status (R + AR), anti-ICOS and anti-CD28 mAb specifically decreased IL-4+ and IL-13+ cells (figure 3A and table 4), but had no influence on IFN-γ+ cells (table 4). Anti-CTLA-4 mAb had no effect on IL-4+ cells, but unexpectedly decreased IL-13+ cell proportions (table 4).

In non allergic subjects (A + controls), anti-co-receptor antibodies did not affect Th1 or Th2 cytokine production (figure 3 and table 4).