Department of Paediatrics, Ruijin Hospital Affiliated to Shanghai Jiao Tong University School of Medicine, Shanghai, China
Correspondence: Dr Zhenwei Xia, Department of Paediatrics, Ruijin Hospital Affiliated to Shanghai Jiao Tong University School of Medicine, 197 Ruijin Er Road, Shanghai 200025, China. Email: firstname.lastname@example.org
Antigen-induced allergic airway inflammation is mediated by T helper type 2 (Th2) cells and their cytokines, but the mechanism that initiates the Th2 immunity is not fully understood. Recent studies show that basophils play important roles in initiating Th2 immunity in some inflammatory models. Here we explored the role of basophils in ovalbumin (OVA) -induced airway allergic inflammation in BALB/c mice. We found that OVA sensitization and challenge resulted in a significant increase in the amount of basophils in blood and lung, along with the up-regulation of activation marker of CD200R. However, depletion of basophils with MAR-1 or Ba103 antibody attenuated airway inflammation, represented by the significantly decreased amount of the Th2 subset in spleen and draining lymph nodes, interlukin-4 level in lung and OVA-special immunoglobulin E (sIgE) levels in serum. On the other hand, adoptive transfer of basophils from OVA-challenged lung tissue to naive BALB/c mice provoked the Th2 immune response. In addition, pulmonary basophils from OVA-challenged mice were able to uptake DQ-OVA and express MHC class II molecules and CD40 in vivo, as well as to release interleukin-4 following stimulation by IgE–antigen complexes and promote Th2 polarization in vitro. These findings demonstrate that basophils may participate in Th2 immune responses in antigen-induced allergic airway inflammation and that they do so through facilitating antigen presentation and providing interleukin-4.
Asthma is a chronic allergic airway inflammatory disease characterized by intense eosinophil and CD4+ T-cell infiltration in the airway, mucus hypersecretion, airway remodelling and airway hyper-reactivity. Many different cell types and cytokines contribute to allergic airway inflammation.[1-4] It is well documented that antigen-induced allergic airway inflammation is mediated by T helper type 2 (Th2) cells and the cytokines they release.[5-7] Therefore, the initiation of the Th2 immune response is a critical process for the development of allergic asthma.
Although the mechanisms that initiate the Th2 immune responses have been extensively studied, the controversy remains concerning the specific antigen-presenting cells (APCs) that provide ‘early interleukin-4 (IL-4)’ to initiate Th2 differentiation. It has been demonstrated that FceRI+ CD49b+ c-Kit− basophils act as APCs by taking up and processing antigens after helminth infection or papain injection.[9-11] In addition, basophils express MHC class II and co-stimulatory molecules and secrete IL-4 and thymic stromal lymphopoietin,[12-14] which are critical checkpoints in the development of Th2 cell immunity. Moreover, depletion of circulating basophils can significantly inhibit the Th2 immune response in animal models of parasite infection and allergic dermatitis.[12, 15-18] Hence, basophils alone are able to induce Th2 skewing without dendritic cells (DCs) in certain conditions. In contrast, another group has found that IL-4-producing basophils could not present antigens or express the chaperones required for antigen presentation. These authors proposed that DCs are necessary and sufficient for the induction of Th2 immunity to house dust mites in the lung without the participation of basophils.
The roles of basophils in the Th2 immune response differ from model to model or between murine and human.[20-25] Therefore, it is of great importance to elucidate whether or not, and under which conditions, basophils induce a Th2 immune response to foreign antigen exposure. In this study, we sought to define the role of basophils in ovalbumin (OVA) -induced allergic airway inflammation. We found that OVA challenge significantly increased the amount of basophils in blood and lung. Temporary depletion of basophils with MAR-1 or Ba103 antibody in vivo significantly decreased the Th2-mediated airway inflammation. On the other hand, adoptive transfer of basophils from OVA-challenged lung tissue to naive BALB/c mice provoked the Th2 immune response. In addition, pulmonary basophils from OVA-challenged mice were able to uptake DQ-OVA and express MHC class II molecules and CD40 in vivo, as well as to release IL-4 following stimulation by IgE–antigen complexes and promote Th2 polarization in vitro. These findings demonstrate that basophils participate in Th2 immune responses in antigen-induced allergic airway inflammation and that they do so through facilitating antigen presentation and providing IL-4.
Materials and methods
Healthy female BALB/c mice and DO11.10 mice (with BALB/c genetic background) of age 6–8 weeks were purchased from the SLAC Laboratory Animal Co., Ltd (Shanghai, China), and housed at the Laboratory Animal Centre of Ruijin Hospital. The animal study protocol was approved by the Animal Ethics Committee of Ruijin Hospital, Shanghai Jiao Tong University School of Medicine, China.
OVA sensitization, challenge protocol and MAR-1 or Ba103 antibody intervention
In the OVA group, mice were sensitized with 500 μg/ml OVA (Sigma-Aldrich, St Louis, MO) dissolved in 10% aluminium potassium sulphate at day 0 and day 14. Each mouse received 0·2 ml OVA through intraperitoneal injection. On day 14, 50 μl normal saline containing 100 μg OVA was administered intranasally under isoflurane inhalation anaesthesia. On days 23, 24 and 25, the mice were challenged with the same dose of OVA again by intranasal administration to initiate allergic airway inflammation. The control group was sensitized with aluminium hydroxide and challenged with normal saline without OVA. Both groups of mice were killed on day 26.
MAR-1 or Ba103 antibody was used before OVA sensitization and challenge to deplete basophils as reported previously.[14, 16] Briefly, basophils were depleted at days −2, −1, 10, 11, 20 and 21 by tail vein injection of 100 μg MAR-1 antibody (anti-FcεR1α antibody, eBioscience, San Diego, CA; MAR-1 group) or 30 μg Ba103 antibody on days −1, 13 and 23 through the tail vein (HyCult, Uden, the Netherlands; Ba103 group). Meanwhile, basophils were depleted in OVA-sensitization early stage (MAR-1 at days −2, −1, 10 and 11 and Ba103 at days −1 and 13) to observe the alteration of allergic airway inflammation.
Sample harvest and examination
On day 26, 24 hr after the last OVA challenge, mice were killed. Blood (anti-coagulated with heparin) and lung were examined for changes in basophil number. Draining mediastinal lymph nodes and spleen were used to analyse T helper subsets. Serum IL-4 and OVA-special immunoglobulin E (sIgE) concentrations were also examined.
After blood collection, mice were fixed in a supine position, the neck trachea was exposed and ligated distally, and a 22-gauge catheter needle was inserted. The lungs were lavaged with 0·3 ml ice-cold saline three times before the bronchial alveolar lavage fluid (BALF) was retrieved. The total cell count in the BALF was determined with a haemocytometer. Differential counts of eosinophils were determined on smears of BALF samples from individual mice stained with Wright–Giemsa solution and identified by standard morphological criteria after counting 200 cells.
The lower right lung lobe was fixed in 10% formalin, embedded in paraffin, and cut into 5-μm sections. These sections were stained with haematoxylin & eosin and examined under light microscopy (Olympus AX70; Olympus, Shinjuku, Japan). Unfixed lung tissue was used to detect the IL-4 level or isolate basophils for analysis of their surface markers.
To examine basophil antigen uptake, OVA-immunized mice were anaesthetized with isoflurane and intranasally administered with 50 μg DQ-OVA (Invitrogen, Carlsbad, CA) in 50 μl PBS on day 23, and killed 24 hr later. The lung tissues were harvested to examine the uptake of DQ-OVA by basophils.
Serial lung tissue sections (5 μm) were stained with haematoxylin & eosin. Twenty fields of each section were randomly selected to determine inflammation scores. Peribronchial inflammation was graded in a blinded fashion on a subjective scale of 0, 1, 2, 3 and 4 corresponding to minimal, mild, moderate, marked and severe inflammation, respectively.
Preparation of single-cell suspensions
For lymph nodes or spleen, tissues were ground against a 70-μm cell strainer to prepare single-cell suspensions. The lung tissue was minced and digested by collagenase IV in an incubator at 37° with 5% CO2 for 45 min, and then strained to obtain a single-cell suspension. To remove red blood cells, both single-cell suspensions and heparin anti-coagulated blood were treated with erythrocyte lysis buffer before cell purification.
Flow cytometry and cell purification
To purify CD11c− basophils from lung tissue, single-cell suspensions were blocked with 10 μg/ml anti-FcγRII/III (eBioscience) for 30 min at 4°, then quadruple labelled with phycoerythrin-Cy7 (PE-Cy7) -anti-FcεR1, allophycocyanin-anti-c-kit, FITC-anti-CD49b and PE-anti-CD11c antibodies for 1 hr at 4°, and sorted by FACSAria (BD Biosciences, San Jose, CA). FcεR1+ CD49b+ c-kit− CD11c− cells were isolated as CD11c− basophils to exclude ‘inflammatory’ FcεR1+ DCs, which were reported as key cells for the Th2 immune response. Cytospin was also performed, followed by Wright–Giemsa staining. Sorted cells were identified by standard morphological criteria.
To determine the expression of co-stimulatory molecules on basophils and their activation, the single-cell suspensions derived from peripheral blood or lung tissue were surface-labelled with PE-Cy7-anti-FcεR1, allophycocyanin-anti-c-kit, FITC-anti-CD49b and PE-labelled CD40, CD80, CD86 and CD1Ad (Biolegend, San Diego, CA), or CD200R antibody cocktail (eBioscience) after blocking with 10 μg/ml anti-FcγRII/III.
The pulmonary single-cell suspensions were prepared and surface stained with PE-Cy7-anti-FcεR1, allophycocyanin-anti-c-kit and PE-anti-CD49b antibodies to examine the uptake of DQ-OVA.
To detect T helper subsets, cells from draining mediastinal lymph nodes and spleen were resuspended in RPMI-1640 complete medium containing 10% fetal bovine serum (FBS), 100 U/ml penicillin and streptomycin, 55 μm β-mercaptoethanol (Life Technologies, San Diego, CA), and inoculated into 24-well plates at a density of 5 × 106 cells/ml. Cells were stimulated with 500 μg/ml OVA (0·22-μm filtered) for 5 days, followed by 20 ng/ml PMA (Sigma-Aldrich) and 1 μg/ml ionomycin (Sigma-Aldrich) treatment for another 6 hr. Two hours before collection, Brefeldin A (eBioscience) was added into the culture medium at a ratio of 1 : 1000. The collected cells were surface stained with FITC-anti-CD4 antibody (BD Pharmingen, San Jose, CA). After washing, fixing and permeabilizing, cells were intracellularly stained with PE-anti-IL-4 and allophycocyanin-anti-IFN-γ antibodies (eBioscience), and examined with flow cytometry to detect the Th1/Th2 subsets.
For quantification of serum OVA-specific IgE levels, 100 μl of 1 : 200 anti-mouse IgE (Serotec, Kidlington, UK) was incubated in 96-well flat-bottom plates overnight at 4°. The plates were washed with PBS containing 0·05% Tween-20 (PBST) and blocked with PBS containing 10% heat-inactivated FBS for 1 hr. One hundred microlitres of serum (diluted at 1 : 20) or standard mouse OVA-sIgE (Serotec) was added, and the plates were incubated for 2 hr at room temperature, then washed three times with PBST for 5 min each time, treated with 100 μl of 1 : 100 horseradish peroxidase-conjugated OVA (Serotec), incubated at room temperature for 2 hr, washed again three times for 5 min, and treated with 100 μl 3,3′,5,5′-tetra-methylbenzidine reagent for 20 min (Jingmei, Shanghai, China). Reactions were stopped by adding 50 μl 1 m sulphuric acid. Signal was detected with a plate-reader at 450 nm and OVA-sIgE concentration was calculated according to their optical density values against the standard curve.
The lung tissue was lysed in lysis buffer containing 0·15 m NaCl, 5 mm EDTA (pH 8·0), 1% Triton × 100, and 10 mm Tris–HCl (pH 7·4), supplemented with one tablet of Complete Mini Protease Inhibitor for every 7 ml of lysis buffer. The lysate was ultrasonicated, placed in a still at 4° for 1 hr, and centrifuged at 20 000 g for 30 min. The supernatant was used to determine the IL-4 levels using an IL-4 ELISA kit (Biolegend).
Interleukin-4 releasing detection in vitro
Purified pulmonary basophils (0·5 × 106) sorted from OVA-challenged mice were cultured in RPMI-1640 complete medium in 96-well plates and further treated with 100 μg/ml of 2,4-dinitrophenylated ovalbumin (DNP-OVA) (Biosearch Technologies, Novato, CA), 10 μg/ml anti-DNP IgE (Sigma-Aldrich) and 10 ng/ml IL-3 for 16 hr. Supernatants were collected and the concentration of IL-4 was measured by ELISA.
Th2 cell differentiation assay in vitro
Splenic naive T cells from DO11.10 mice were purified using MagCellect (R&D Systems, Minneapolis, MN) according to the manufacturer's instructions. Briefly, cell suspensions were first incubated with the MagCellect Antibody Cocktail which targets unwanted cells. MagCellect Streptavidin Ferrofluid was added to the reaction to allow streptavidin-coated nanoparticles to interact with biotinylated antibody on tagged cells. Naive T cells were isolated by removal of magnetically tagged unwanted cells using the MagCellect magnet.
To polarize Th2 cells, 1 × 106 naive T cells were co-cultured with 0·5 × 106 sorted lung tissue-derived basophils in 48-well plates at 20 U/ml IL-2 (R&D Systems), 10 ng/ml IL-3, 100 μg/ml DNP-OVA and 10 μg/ml anti-DNP IgE (Sigma-Aldrich) with or without 5 μg/ml anti-MHC class II neutralizing antibody (M5/114; eBioscience). Fresh complete RPMI-1640 medium with 10 U/ml IL-2 was added at day 3. After 5 days of culture, intracellular staining of Th1/Th2 was performed according to the above methods and detected by flow cytometry.
Adoptive transfer of basophils
Twenty-four hours after the last OVA challenge, mice were killed on day 26, and lung tissue was collected and digested with collagenase IV. FcεR1+ CD49b+ c-kit− CD11c− basophils were purified by FACSAria as described above and confirmed morphologically by Wright–Giemsa staining. Approximately 1 × 105 basophils were purified from one mouse. Purified basophils were washed four times with PBS and resuspended in PBS at 2·5 × 106/ml. An adoptive transfer model was established as described previously.[14, 16] Briefly, 2·5 × 105 basophils (basophil adoptive transfer group), or 2·5 × 105 FcεR1− CD49b− CD11c− cells (non-basophil group) in 200 μl PBS were delivered to naive BALB/c mice via a tail vein injection on day 0. On days 14, 21, 22 and 23 after adoptive transfer, BALB/c mice were challenged with 100 μg OVA in 50 μl normal saline intranasally under isoflurane inhalation anaesthesia. Mice given a PBS tail vein injection on day 0 and normal saline challenge intranasally were regarded as the control group. Mice were killed on day 24. Serum OVA-IgE and IL-4 were detected by ELISA. The Th2 subset was analysed by flow cytometry. The airway inflammation was histopathologically assessed according to the above methods.
To check the delivery of pulmonary basophils, purified pulmonary basophils from OVA-immunized and -challenged mice were resuspended in PBS at 1 × 106/ml (pre-warmed to room temperature) and were labelled with 1 μm carboxyfluorescein diacetate succinimidyl ester (CFSE; eBioscience). The cells were subsequently washed twice using RPMI-1640 containing 10% FBS followed by two PBS washes. Then, 2·5 × 105 cells in 200 μl PBS were transferred into recipient BALB/c mice by tail vein injection. Mice were killed 24 and 48 hr after injection. Freshly isolated lung tissues were perfused with 1% paraformaldehyde and embedded with a 1 : 1 mixture of Tissue-Tek Optimal Cuttin Temperature (Sakura, Los Angeles, CA) and PBS, followed by snap-freezing. Sections of frozen tissue (7 μm) were then examined by fluorescence microscopy for the presence of CFSE+ cells.
Flow cytometry results were analysed using Cellquest software (BD Pharmingen). Data were analysed with SPSS 16.0 software (SPSS Inc, Chicago, IL). The comparison among groups was performed with an independent-samples t-test or analysis of variance. A P <0·05 was considered statistically significant.
OVA sensitization and challenge promotes basophil expansion and activity
To understand the role of basophils in OVA-induced allergic airway inflammation, we first examined the change and activation of basophils in mice after OVA challenge. The percentage of basophils in peripheral blood and lung from OVA-challenged mice was determined with flow cytometry. As shown in Fig. 1, the OVA-challenged group showed a significantly enhanced basophil population in the tissues examined, accompanied by markedly increased expression of marker CD200R on basophils in peripheral blood and lung (Fig. 1a,b). Morphology of sort-purified basophils was surveyed (Wright–Giemsa staining, oil lens) (Fig. 1c). The changes of basophils in number and activation condition indicated that basophils may involve an OVA-induced Th2 immune response.
Depletion of basophils significantly inhibits the OVA-induced Th2 immune response and allergic airway inflammation
To further confirm the role of basophils in the pathogenesis of allergic airway inflammation, we next investigated whether or not depletion of basophils attenuated OVA-induced airway inflammation. MAR-1 antibody has been reported to deplete basophils effectively and the depletion period could last at least 10 days.[9, 11, 14, 23] In our study, MAR-1 antibody (100 μg) was injected via the tail vein before OVA induction at 10-day intervals according to the method of Sokol et al.. However, MAR-1 antibody targets on FcεR1α, which could lead to non-specific elimination of other cell sources expressing this receptor, including DCs and mast cells. Therefore, Ba103 antibody, a more specific antibody, was also used to deplete basophils in our study. As shown in Fig. 2(a), both antibodies are capable of depleting basophils in peripheral blood effectively and this effect could last for 10 days, which is consistent with the published data. Strikingly, the airway inflammation was effectively blocked when basophils were deleted in early phase (Fig. 2b,c), suggesting that basophils may participate in the initiation of airway inflammation. Nevertheless, MAR-1 or Ba103 antibody was even more effective in OVA-sensitization and challenge periods. Furthermore, the results demonstrated that both MAR-1 and Ba103 antibodies significantly reduced OVA-sIgE in serum (Fig. 3a), as well as total cell number and eosinophil number in BALF (Fig. 3b,c), IL-4 level in lung (Fig. 3d), attenuated airway inflammation and eosinophilic infiltration (Fig. 3e,f). Meanwhile, depletion of basophils markedly decreased the proportion of OVA-specific Th2 cells in draining mediastinal lymph nodes and spleen (Fig. 4a–c). These results clearly showed that temporary depletion of basophils by antibodies could significantly alleviate OVA-induced eosinophilic airway inflammation by inhibiting the Th2 immune response, indicating a crucial role of basophils in OVA-induced allergic airway inflammation.
Adoptive transfer of lung tissue-derived basophils provokes the Th2 immune response in vivo
Although MAR-1 and Ba103 antibodies have been proved to deplete basophils specifically and effectively, there is some debate on the side effects of those antibodies, which might result in depletion or activation of other cell types. To further confirm the role of basophils in OVA-induced allergic airway inflammation, lung-tissue-derived FcεR1+ CD49b+ c-kit− CD11c− basophils were isolated from OVA-challenged BALB/c mice and further identified morphologically. After being washed four times, 2·5 × 105 basophils with or without CFSE label in 200 μl PBS were delivered to naive BALB/c mice via a tail vein injection. As shown in Fig. 5, CFSE-labelled basophils were detected by fluorescence microscopy in lung tissue 24 and 48 hr after transfer (Fig. 5a). Transfusion of basophils significantly invoked airway inflammation characterized by eosinophil infiltration around the airway (Fig. 5b, right), elevated OVA-sIgE and IL-4 levels both in serum and BALF (Fig. 5c–e), and increased the proportion of Th2 cells in spleen (Fig. 5f). Transfer of non-basophils (FcεR1− CD49b− CD11c−) and OVA challenge (non-basophil group) slightly elevated serum OVA-sIgE (Fig. 5c) and induced slight inflammation around the airway; the infiltrated cells were mainly lymphocytes (Fig. 5b, middle). However, IL-4 level in serum and BALF did not change between the control and OVA groups (Fig. 5d,e). These data indicated that in vivo adoptive transfer of lung tissue-derived basophils initiated OVA-specific Th2 immune responses and allergic airway inflammation in naive BALB/c mice, which further proved the important role of basophils in OVA-induced allergic airway inflammation.
Basophil uptake of DQ-OVA and expression of MHC II and co-stimulatory molecules in vivo
The above results have demonstrated the key role of basophils in OVA-induced allergic inflammation. Next, we explored the possible underlying mechanisms. Recently, several studies have shown that basophils function as APCs in vitro during parasitic infection and allergic dermatitis.[12, 15-18] It is unclear whether basophils can play a similar role during the development of allergic airway inflammation. As all the findings in our experiments above indicated an important role for basophils in the mouse model of OVA-induced eosinophilic airway inflammation, we expected to understand whether basophils could act as APCs in this OVA-induced immune response, including the antigen uptake ability and the molecular mechanism for antigen presentation and T-cell activation. Hence, mice with and without OVA sensitization were challenged with 50 μg DQ-OVA on day 23. The results demonstrated that DQ-OVA was taken by pulmonary basophils 24 hr after DQ-OVA challenge (Fig. 6a). It is worth noting that the uptake ability of basophils in the OVA sensitization group significantly increased compared with the control group (Fig. 6a).
We next detected the expression of molecular markers at the surface of basophils derived from lung tissues, namely MHC II, CD40, CD80 and CD86, after OVA sensitization and challenge in vivo. As shown in Fig. 6(a), basophils expressed MHC II and co-stimulatory molecules in vivo, and the expression of MHC II and CD40 was significantly up-regulated, whereas that of CD80 and CD86 was mildly enhanced after OVA sensitization and challenge. These results revealed that basophils could be equipped with molecules required for the interaction with T cells, which suggested that basophils have antigen-presenting features.
Basophils release IL-4 and promote Th2 differentiation in vitro
We next sought to determine whether basophils could produce IL-4, a critical cytokine for promoting Th2 differentiation, and to further investigate whether basophils could directly induce the Th2 lineage. Flow cytometry was used to sort basophils from lung tissue as described in adoptive transfer experiments. Then the purified cells were stimulated in vitro with 10 μg/ml DNP-IgE and 100 μg/ml DNP-OVA for 16 hr, mimicking IgE–antigen complexes to promote degranulation according to the method of Yoshimoto et al.. The IL-4 level in supernatants was assessed by ELISA. We found that basophils release large quantities of IL-4 upon DNP-IgE/DNP-OVA stimulation (Fig. 6b), suggesting that activated basophils could secret IL-4 and might provide the cytokine environment for Th2 polarization.
Finally, we simulated Th2 differentiation in vitro. Purified pulmonary basophils were co-cultured with MACS-isolated naive T cells derived from spleens of DO11.10 mice for 5 days. DNP-IgE/DNP-OVA, IL-3 and IL-2 were added to the media. The proportions of Th1/Th2 cells were analysed by flow cytometry. The results showed a significant increase in the Th2 cells with no change in the Th1 cells (Fig. 6c). Moreover, the Th2 differentiation was suppressed after addition of anti-MHC II antibody (Fig. 6c). These indicate that basophils from OVA-challenged lung tissue are capable of promoting Th2 differentiation in an MHC class II-dependent manner.
The immune mechanism of the Th2 response has been largely revised since basophils have gained new respect as a neglected minority.[26-32] Several independent studies showed that basophils can function as professional APCs and initiate Th2 polarization.[28, 30-32] Those studies established that basophils may participate in the initiation of Th2 polarization. However, the role of basophils in the development and progression of Th2 immune responses against different allergens is still debated. In this study, we explored the roles of basophils in a murine model of OVA-induced allergic airway inflammation. Here, we found that basophils were increased and recruited to the lung after OVA sensitization and challenge, whereas temporary depletion of basophils with MAR-1 or Ba103 antibody in vivo significantly attenuated the Th2-mediated airway inflammation. On the other hand, adoptive transfer of pulmonary basophils to naive BALB/c mice provoked the Th2 immune response, leading to airway inflammation. In this process, further studies confirmed that lung-derived basophils were able to uptake DQ-OVA and express MHC class II molecules and CD40 in vivo, as well as to release IL-4 following stimulation by IgE–antigen complexes and promote Th2 polarization in vitro.
We first successfully established the animal model of airway allergic inflammation. Consistent with previous reports,[33-35] our findings in vivo suggested that OVA challenge resulted in a significant increase of the OVA-specific Th2 subset, elevation of OVA-sIgE levels, and intensive eosinophil infiltration in the airway walls, which mimics the typical symptoms of human allergic airway inflammation. We then investigated the changes in basophil number and activity following OVA challenge. CD200R is used as an activation marker due to the close relationship between CD200R expression and basophil activity. We found that OVA sensitization significantly increased basophil numbers in peripheral blood and lung tissue. As basophils generally circulate in the peripheral blood and barely migrate into the peripheral tissues under physiological conditions, our results suggest that OVA-immunization probably promoted the differentiation of basophils. The basophils were recruited into lungs, where they contribute to the Th2 immune response and the consequent allergic airway inflammation.
To confirm this hypothesis, we depleted basophils before and during OVA challenge using MAR-1 or Ba103 antibody. MAR-1 antibody specifically targets the non-signalling α-chain of the FcεR1 complex. It has been reported that FcεR1α is present on basophils and mast cells during steady-state conditions in mice, and intravenously or intraperitoneally administered MAR-1 antibody can effectively deplete circulating basophils.[9, 11, 14, 23] A recent report showed that MAR-1 decreased numbers of both basophils and mast cells in the lung in a murine asthma model; however, other studies showed that this antibody preferentially depleted basophils and had a minimal effect on tissue-resident mast cells.[9, 14, 37] Therefore, Ba103 antibody was used as an effective way to deplete basophils.[16, 26] This antibody directly neutralizes CD200R3, which is regarded as a more specific basophil-depleting antibody. We found that basophil depletion with both MAR-1 and Ba103 antibodies significantly reduced circulating basophils, decreased the OVA-specific Th2 subset in murine draining mediastinal lymph nodes and spleen, which are the main sites of Th2 cell differentiation and development, lowered lung and serum IL-4 levels as well as the serum OVA-sIgE level, inhibited the Th2 immune response, and attenuated allergic airway inflammation characterized by eosinophil infiltration.
Depletion of basophils with MAR-1 may affect mast cells. To further confirm the key role of basophils in the Th2 immune response, adoptive transfer of basophils to naive BALB/c mice was performed. We purified pulmonary FcεR1+ CD49b+ c-kit− CD11c− basophils by flow cytometry sorting from lungs of OVA-challenged mice to exclude mast cells and FcεR1+ ‘inflammatory DCs’ because they may be involved in allergic inflammation, then we transferred FcεR1+ CD49b+ c-kit− CD11c− basophils to naive BALB/c mice via tail vein injection followed by OVA intranasal challenge. The results showed that transferred basophils invoked a significant Th2 immune response and eosinophil infiltration around the airway. In vitro, those lung-tissue-derived FcεR1+ CD49b+ c-kit− CD11c− basophils from OVA-challenged mice were co-cultured with naive T cells derived from DO11.10 mice. We found that lung tissue derived CD11c− basophils could effectively promote Th2 cell differentiation from naive T cells under stimulation with the DNP-IgE/DNP-OVA complex without exogenous IL-4. This promotion of Th2 differentiation was MHC class II-dependent because the differentiation was suppressed by anti-MHC II antibody. This result is consistent with the finding of Perrigoue et al. These data clearly demonstrate that lung-tissue-derived basophils were able to initiate Th2 polarization in vitro and participate in OVA-induced Th2 immune response and allergic airway inflammation.
Next, we further explored the possible mechanisms by which basophils were involved in allergic airway inflammation. Some researchers considered basophils as professional APCs in initiating a Th2 immune response in vivo and in vitro.[9-11] Another study reported that basophils did not function as APCs for inhaled house dust mite allergens. Recent findings found that basophils alone were able to induce Th2 skewing with haptens and peptide antigens but DCs were required for the induction of Th2 for protein antigens upon epicutaneous immunization. Therefore, it is unclear whether basophils can act as a primary inducer to participate in Th2 immunity in allergic airway inflammation. In the current study, we showed that basophils were able to effectively uptake DQ-OVA delivered to the airway intranasally. DQ-OVA is a self-quenched conjugate of OVA that exhibits green fluorescence on proteolytic degradation to single dye-labelled peptides in the cells. It is regarded as evidence of antigen processing.[38, 39] In addition, we found that OVA sensitization significantly enhanced antigen uptake by basophils in vivo, indicating that a Th2-dominant bias may enhance the function of basophils. This finding indicates that basophils are able to take up soluble antigens, such as OVA, and initiate the Th2 immune response.
Successful antigen presentation requires proper participation of MHC II and co-stimulatory molecules between APCs and T cells. Hence, we examined the expression of surface MHC II, CD40, CD80 and CD86 on basophils freshly isolated from lung tissue, and found that the expression of MHC II and CD40 increased significantly, whereas CD80 and CD86 were enhanced slightly on pulmonary basophils. Hence, basophils are equipped with molecules associated with antigen presentation. Ovalbumin sensitization up-regulates the expression of co-stimulatory molecules on basophils and augments antigen uptake, which suggests that basophils are actively engaged in antigen presentation.
The cytokine milieu during antigen presentation is an important factor for determining T-cell differentiation. Interleukin-4 is believed to be the critical cytokine involved in Th2 cell differentiation. However, the ‘early’ source of IL-4 is still controversial. In the current study, we further investigated the IL-4 releasing ability of basophils. We purified basophils from lung tissue by flow cytometry, and stimulated them with DNP-IgE/DNP-OVA in vitro, which can strengthen the immune reaction. The results showed that 16 hr after IgE–antigen complex stimulation, basophils were capable of degranulating rapidly and releasing large quantities of IL-4 (required to initiate Th2 differentiation). These findings indicate that basophils might be the source of ‘early’ IL-4.
However, the results of our study are different from those of Hammad et al. Their studies showed that basophils in draining lymph nodes expressed a lower level of MHC class II molecules, barely took up fluorescent OVA protein when mice were sensitized with OVA intranasally together with house dust mite extracts, and failed to induce T-cell differentiation or secret effector cytokines when they were incubated with naive OVA-specific T cells. Their study concluded that basophils played a less important role in the development of allergic inflammation, in which house dust mite was used as an antigen. In our study, transfusion of purified pulmonary CD11c− basophils to naive BALB/c mice significantly elevated the IL-4 and OVA-sIgE level in serum and BALF, increased Th2 cells in spleen, and finally invoked airway inflammation. These findings further confirmed the role of basophils in OVA-induced allergic airway inflammation. The main reasons for the difference in observations may depend on the type of immune responses or of antigens, routes of antigen administration, and different experimental models, which play dominant roles in determining basophil dependency. In addition, DNP-OVA/anti-DNP-IgE is important to activate basophils, but was excluded in the Th2 polarization and IL-4 releasing studies in the report by Hammad et al. Our data also showed that basophils released very limited IL-4 without the complex stimulation in vitro.
Our current study noted that MAR-1 antibody inhibited allergic inflammation more effectively than Ba103 antibody did. This was because other FcεRI+ immune cells, especially DCs, could be involved in OVA-induced allergic airway inflammation under certain conditions. It was reported that depletion of DCs did attenuate allergic inflammation. However, DCs are not capable of releasing IL-4 and initiating Th2 cell polarization without exogenous IL-4 in vitro.[9-11, 14] These studies indicate that basophils might serve as an early ‘IL-4’ provider to favour DCs establishing a Th2 immune response.[41, 42] So basophils may collaborate with those immune cells in initiating and maintaining allergic inflammation.
In conclusion, our results strongly indicate that basophils may act as a primary inducer, rather than an effector. They may participate in Th2 immune responses in OVA-induced allergic airway inflammation and they do so through facilitating antigen presentation and providing IL-4.
This work was supported by grants from the National Natural Science Foundation of China (Grants 81070022, 81128001, 81270084, and 81270085), Shanghai Municipal Science and Technology Commission Foundation (13XD1402800, 10410701000, and 12ZR1419100).
Wenwei Zhong designed and performed the experiments, analysed the results, made the figures and drafted the manuscript. Wen Su participated in experimental studies. Yanjie Zhang participated in experimental studies and helped to make the figures. Qi Liu performed the statistical analysis. Jinhong Wu participated in the experimental studies. Caixia Di participated in the experimental studies. Zili Zhang designed the experiments. Zhenwei Xia conceived the study, designed the experiments, analysed the data and helped to draft the manuscript. All authors have read and approved the final manuscript.
The authors declare no competing financial interests.