Production of chemokines and proinflammatory and antiinflammatory cytokines by human alveolar macrophages activated by IgE receptors

Production of chemokines and proinflammatory and antiinflammatory cytokines by human alveolar macrophages activated by IgE receptors Philippe Gosset, PhD,a Isabelle Tillie-Leblond, MD,a,b Séverine Oudin, MBS,a Odile Parmentier, MBS,a Benoit Wallaert, MD,a,b Michel Joseph, PhD,a and André-Bernard Tonnel, MDa,b Lille, France

Background: The alveolar macrophage (AM) expresses the low affinity IgE receptor and has the ability to produce not only several proinflammatory cytokines (TNF-α, IL-1, IL-6) but also antiinflammatory cytokines (IL-1 receptor antagonist [IL-lra], IL-10), chemokines (IL-8, monocyte chemotactic protein-1 [MCP-1]), and macrophage inflammatory protein-1α (MIP-1α). Objective: The aim of this study was to evaluate the capacity of the AM from patients with allergic asthma and control subjects to produce chemokines and antiinflammatory versus proinflammatory cytokines after activation by IgE receptors and to define the role of CD23 in this activation. Methods: AMs were collected by bronchoalveolar lavage from 13 patients with allergic asthma and 14 healthy subjects. Adherent AMs were activated either by the successive addition of IgE and anti-IgE or by monoclonal mouse IgG anti-CD23 or by a control monoclonal mouse antibody. TNF, IL-1β, IL1ra, IL-10, IL-8, MCP-1, and MIP-1α levels were evaluated in supernatants of AMs incubated for 18 hours and in some cases after 4 hours of incubation. Results: Activation by IgE and anti-IgE antibodies significantly increased the production of TNF, IL-1β, IL-8, MCP-1, MIPlα, and IL-10 in both control subjects and patients with asthma, whereas the increase for IL-1ra was only significant for the control subjects. Whereas F(ab) fragments of anti-CD23 antibodies inhibited IgE plus anti-IgE–induced cytokine production, activation by monoclonal IgG anti-CD23 antibodies reproduced the effect of IgE immune complexes. At 4 hours, the secretion of proinflammatory cytokines was increased by activation by IgE receptors, in contrast to antiinflammatory cytokines. In addition, analysis of the balance between proinflammatory and antiinflammatory cytokines showed that IgEdependent activation largely favored the proinflammatory cytokines, particularly in patients with asthma. Conclusion: IgE-dependent activation by the FcεRII receptor upregulates the synthesis of both chemokines and antiinflammatory cytokines in addition to proinflammatory cytokines. However, AMs from patients with allergic asthma may promote airway inflammation after activation by IgE receptors through its preferential effect on proinflammatory cytokines. (J Allergy Clin Immunol 1999;103:289-97.) Key words: Chemokine, cytokine, IgE receptor, alveolar macrophage, allergic asthma

From aUnité INSERM U416, Institut Pasteur, and bClinique des maladies respiratoires, Hopital Calmette, Lille, France. Received for publication Dec 4, 1997; revised July 17, 1998; accepted Sept 16, 1998. Reprint requests: Philippe Gosset, Unité INSERM U416, Institut Pasteur, BP245, 59019 Lille, France. Copyright © 1999 by Mosby, Inc. 0091-6749/99 $8.00 + 0 1/1/94632

Abbreviations used AM: Alveolar macrophage BAL: Bronchoalveolar lavage BM: Blood monocyte FcεRI: High affinity IgE receptor FcεRII: Low affinity IgE receptor (CD23) GAPDH: 6-Glyceraldehyde phosphodeshydrogenase IL-1ra: IL-1 receptor antagonist MIP-1α: Macrophage inflammatory protein-1α MCP-1α: Monocyte chemotactic protein-1

In airways, macrophages constitute the majority of cells recovered by bronchoalveolar lavage (BAL). Mononuclear phagocytes express IgE receptors and IgEmediated ligation of these moieties triggers cell activation, suggesting that IgE-dependent stimulation of mononuclear phagocytes is involved during the development of the allergic reaction. The IgE binding sites present on monocytes/macrophages are the low affinity IgE receptor CD23 (FcεRII)1-3 and the IgE-binding protein (εBP).4 Recently, Maurer et al5 demonstrated that blood monocytes (BMs) can also bind monomeric IgE by means of the high affinity IgE receptor (FcεRI) and that FcεRI receptor expression on BMs was increased in patients with atopy. The expression by the alveolar macrophage (AM) of the α-chain of FcεRI was recently shown by immunohistochemistry.6 IgE-dependent activation of mononuclear phagocytes induces the release of mediators promoting the inflammatory reaction. These bioactive products include enzymes,3 lipid mediators,7,8 radical oxygen species, and proinflammatory cytokines.9 Indeed, treatment with IgE immune complexes enhances the secretion of TNF-α and IL-6 by AMs and BMs.10 In addition, activation by means of the IgE receptor increases secretion of IL-1β by BMs, whereas the release of an IL-1 inhibitor does not allow the detection of IL-1 activity in AM supernatants.11 However, levels of IL-1β in BAL fluid from patients with asthma are increased when compared with healthy subjects, and AMs from patients with asthma express high levels of IL-1β mRNA.12 289

290 Gosset et al

Macrophages have the ability to produce other cytokines, such as chemokines and antiinflammatory cytokines. Chemokines belong to the β-thromboglobulin supergene family and are characterized by the presence of 4 conserved cysteines whose location allows their separation into 2 groups of molecules. IL-8, identified for its neutrophil chemotactic activity, was described as the first member of the C-X-C group.13 Now it has been shown that IL-8 is also chemotactic for lymphocytes, basophils, and activated eosinophils and activates some of these cells. The C-C group (including monocyte chemotactic protein (MCP)-1, macrophage inflammatory protein (MIP)-1α and -β, and RANTES) was cloned on the basis of its chemotactic activity for mononuclear cells. As for IL-8, chemokines of the C-C group modulate the activity of eosinophils and basophils at different degrees.14 The antiinflammatory cytokines include IL-10 and one member of the IL-1 family: the IL-1 receptor antagonist (IL-1ra). Whereas IL-1α and IL-1β express proinflammatory functions largely shared with TNF, IL-1ra is an antagonist of IL-1 binding to the IL-1 receptor type I.15 In opposition to the specific activity of IL-1ra, IL-10 demonstrates the ability to reduce mononuclear phagocyte activity, such as the secretion of proinflammatory cytokines16,17 and the expression of adhesion molecules.18 In addition, endogenous production of IL-10 inhibited the secretion of TNF, IL-1β, and IL-8 by LPSstimulated BMs.16 Whereas the kinetics of mRNA expression for IL-1β and IL-1ra in AMs was similar with a maximum between 3 to 8 hours,19 IL-10 secretion and mRNA expression was relatively late as compared with that of IL-1β, IL-8, and TNF.16 Because IgE-dependent activation plays a key role in the development of allergic inflammation, we evaluated the effect of stimulation by means of IgE receptors on the production of chemokines (IL-8, MCP-1, and MIP-lα) and antiinflammatory cytokines (IL-1ra and IL-10) by AMs. Moreover, we quantified the ratio of proinflammatory/antiinflammatory cytokines secreted by these cells and its modulation by IgE-dependent mechanisms. Because the role of CD23 is not clearly demonstrated in AM activation through the IgE receptor, the consequence of activation by an anti-CD23 mAb was measured in terms of cytokine production. Our results show that IgEdependent activation increases the production of IL-1β, IL-8, MCP-1, MIP-1α, IL-1ra, and IL-10. Moreover, the ratio proinflammatory/antiinflammatory cytokines is augmented by IgE receptor cross-linking. mAb-mediated ligation of FcεRII reproduces the pattern of cytokine secretion obtained with IgE-dependent activation.

METHODS Study population Informed consent for BAL was obtained from 13 patients with allergic asthma and 14 healthy subjects. The study was approved by the ethics committee of Lille (CCPPRB Lille n° 9307). All these subjects were nonsmokers. Patients with asthma responded to the usual criteria of allergic asthma: history of personal or familial atopy, mild asthma at a distance from any acute phase, at least one

J ALLERGY CLIN IMMUNOL FEBRUARY 1999

positive cutaneous prick test, and elevated levels of total serum IgE (mean, 1065 KIU/L) and specific IgE antibodies. Other criteria were used to select patients with asthma: minimal baseline FEV1 higher than 70% of the predicted value and stable pulmonary function tests for 1 month before BAL. Airway reversibility obstruction was evaluated after inhalation of 200 µg of albuterol; all patients exhibited a reversible obstruction as judged by a 15% or more change in FEV1 after bronchodilatation. BAL was performed after inhaled corticosteroids and β2-adrenergic drugs had been discontinued for at least 1 week. Theophyllin and cromolyn sodium were also stopped at least 1 day before the lavage. Normal subjects were selected with the following criteria: normal serum IgE (54 KIU/L), no history of asthma or drug intolerance, absence of atopy, and normal pulmonary function tests. BAL was performed as described previously20 by instillation of saline solution into the bronchoalveolar tree under fiberoptic bronchoscopic observation. Total and differential cell counts were determined. The viability of AMs was assessed by trypan blue exclusion. The respective percentages of macrophages, lymphocytes, neutrophils, and eosinophils were 90.1% ± 2.2%, 7.5% ± 2.1%, 1.7% ± 0.3%, and 0.5% ± 0.2% for the control subjects and 84.2% ± 3.7%, 13.5% ± 3.6%, 1.1% ± 0.3%, and 1.5% ± 0.5% for the patients with asthma; AM viability was 90.7% ± 0.8% and 89.1% ± 2.4%, respectively. Epithelial cells were undetectable in BAL fluid, as the first fraction of the lavage; so-called bronchial wash corresponding to the injection of the first 20 mL of saline solution was systematically discarded.

AM isolation and culture AMs were isolated by adherence, as previously described.10 Briefly, the lavage fluid was filtered through sterile surgical gauze and centrifuged at 400g for 10 minutes at 4°C. After 3 washings, the pellet was resuspended at a cell concentration of 1.5 × 106/mL in RPMI1640 (Gibco, Grand Island, NY) containing 5% heat-inactivated FCS and 2 mmol/L L-glutamine (Gibco). Endotoxin contamination of medium was controlled by limulus amoebocyte test (Boerhinger Ingelheim Bioproducts, Heidelberg, Germany) and was below 50 pg/mL. This quantity was insufficient to trigger TNF mRNA expression and secretion, as previously demonstrated.10 Cells were allowed to adhere to plastic Petri dishes (2 mL in a 35-mm diameter well) for 2 hours at 37°C. The nonadherent cells were removed by 3 washings with RPMI-1640. Adherent cells contained more than 95% AMs and less than 3% lymphocytes.

AM activation Cytokine production was evaluated after in vitro IgE-dependent activation. AMs were passively sensitized by a 30-minute incubation with serum (20% vol/vol) rich in IgE (mean values, 1540 ± 240 KUI/L). After 3 washings, anti-human IgE (sheep IgG fraction, 8 µL/mL; Cappel; Cooper Biomedical, Malvern, Pa) was added to the culture. The addition of anti-IgE antibody to control AMs not sensitized with IgE has no effect on cytokine production compared with unstimulated cells. To determine the nature of the IgE receptor involved in cytokine production, AMs were also incubated with a mouse monoclonal anti-human CD23 antibody (clone 135; a gift from Dr Delespesse, Montreal, Canada) and with an unrelevant mouse mAb (a gift from Dr J. Kalife, Lille, France). Preliminary experiments were performed to determine the optimal concentration for monoclonal anti-CD23 antibody, which was 10 µg/mL. In some cases, AMs were preincubated for 30 minutes with F(ab) fragments of anti-CD23 antibodies (10 µg/mL), then sensitized with IgE and activated by anti-IgE antibodies to inhibit the IgE binding on FcεRII and to block receptor activation by anti-IgE. Because antibody solutions contained detectable amounts of endotoxin, all these preparations were passed through a column of Detoxi-gel (Pierce, Rockford,

Gosset et al 291

J ALLERGY CLIN IMMUNOL VOLUME 103, NUMBER 2, PART 1

Ill). After this step, antibody preparations contained no detectable endotoxin, as determined by limulus amoebocyte test. After stimulation, AMs were incubated for 4 hours or 18 hours. Then the supernatants were collected and stored at –30°C until cytokine assays.

Cytokine assays TNF concentration was evaluated sequentially by L929 cytotoxic test21 and immunoenzymetric assay. Biologic activity was estimated from a standard curve as the amount necessary to kill 50% of actinomycin-treated L929 mouse fibroblasts after 18 hours of culture. Subsequently the TNF level was confirmed with an immunoenzymetric assay (Biosource Int, Camarillo, Calif) after adequate dilution. A highly significant correlation was found between both methods (r = 0.85; P < .001). The TNF concentration was finally expressed in nanograms per milliliter, as defined by the standards provided by the manufacturer. The minimum detectable dose is estimated to be 3 pg/mL. IL-1β, IL-1ra, IL-10, IL-8, MIP-1α, and MCP-1 concentrations were measured by sandwich immunoenzymetric assays (for IL-1β, IL-10, and IL-8: Biosource Int; for IL-1ra, MIP-1α, and MCP-1: R&D Systems, Abingdon, UK). Samples were assayed after adequate dilution. The ELISAs were performed as described by the manufacturers. The results were expressed in nanograms per milliliter for IL-1ra, IL-8, MIP-1α, and MCP-1 and in picograms per milliliter for IL-10 and IL-1β. The minimum detectable dose is estimated to be 2, 5, 1, 14, 5, and 7 pg/mL, respectively, for IL-1β, IL8, IL-10, IL-1ra, MCP-1, and MIP-1α. In some cases, the results were also expressed as the increased percentage calculated as followed: ([activated AMs – unstimulated AMs]/unstimulated AMs) × 100. We also calculated the ratios TNF/IL10 and IL-1ra/IL-1β using the same denominator (the 4 cytokines were expressed in picograms per milliliter) in each AM supernatant.

Expression of mRNA encoding for cytokines AMs from 6 patients with allergic asthma and from 6 control subjects were activated as previously described, and RNA was collected after 4 hours of incubation. Total RNA was isolated with the Trizol solution (Life Technologies, Paisley, Scotland). The RNA was reverse-transcribed to first strand cDNA with Moloney mouse leukemia virus–reverse transcriptase (Life Technologies) and oligo-dT primers (Boehringer, Mannheim, Germany). The first strand cDNA was amplified with primer sets for 6-glyceraldehyde-phosphate-dehydrogenase (GAPDH), TNF, IL-8, IL-10, and MCP-1 (Eurogentec, Seraing, Belgium) in a 25-µL reaction with Taq polymerase (Perkin Elmer, Foster City, Calif), magnesium chloride, deoxyribonucleoside triphosphate, and 10-fold concentrated buffer. Nucleotide sequences for oligonucleotide 5´- and 3´-primers were as follows, respectively: GAPDH, GTCTTCACCACCATGGAGA and CCAAAGTTGTCATGGATGACC; TNF, ACAAGCCTGTAGCCCATGTT and AAAGTAGACCTGCCCAGACT; IL-8, TTGGCAGCCTTCCTGATT and AACTTCTCCA-CAACCCTCTG; IL-10, AAATTTGGTTCTAGGCCGGG and GAGTACAGGGGCATGATATC; MCP1, TCCAGCATGAAAGTCTCTGC and TGGAATCCTGAACCCACTTC. A 21-cycle PCR was performed for GAPDH in a DNA thermal cycler (Mastercycler 5330; Eppendorf, Hamburg, Germany), with a denaturation step at 94°C for 1 minute, annealing at 55°C for 1 minute, and extension at 72°C for 1 minute. For TNF, IL-8, and MCP-1, a 22-cycle PCR was performed, whereas 25 cycles were done for IL-10. The number of cycles was defined by kinetics experiments, giving a linear dose-response curve. After PCR, 10 µL of the PCR products was electrophoresed on a 1.5% agaroseethidium bromide gel. The optical density of each band was measured with a video-camera coupled with a computer equipped with the Bio-Profil Software (Vilber Lourmat). Results for cytokine were expressed as percentage of GAPDH optical density.

A

B FIG 1. Production of proinflammatory cytokines, namely TNF (A) and IL-1β (B), by AMs from healthy subjects and patients with allergic asthma. Cytokine secretion was evaluated in 18-hour supernatants from AMs incubated in medium alone, in the presence of either an unrelevant mouse monoclonal IgG1 antibody (control IgG1) or a mouse monoclonal IgG1 antibody anti-CD23, and after incubation with IgE and anti-IgE. *P < .05 versus AMs from control subjects cultured in medium alone; #P < .01 versus AMs from the same group cultured in medium alone.

Statistical analysis Results were expressed as means ± SEM. To evaluate the effect of IgE-dependent activation on cytokine production, results were also expressed as percentages of change in comparison with the respective control; the calculation was performed as followed: ([activated AMs – unstimulated AMs]/unstimulated AMs) × 100. We also determined the ratios TNF/IL-10 and IL-1ra/IL-1β for which cytokine levels were expressed in picograms per milliliter. Statistical analyses were performed with the Mann-Whitney U test and Wilcoxon test.

RESULTS FcεRII-mediated secretion of tumor necrosis factor-α and interleukin-1β by AMs activated by IgE-dependent mechanism To determine whether the secretion of proinflammatory cytokines was dependent on FcεRII cross-linking, TNF and IL-1β were assayed in supernatants from AMs activated for 18 hours by means of IgE receptors. As previously described,10 secretion of TNF by AMs in medium alone from patients with allergic asthma was significantly increased compared with control subjects (P < .05; Fig l, A), whereas there is no significant difference for

292 Gosset et al

A

J ALLERGY CLIN IMMUNOL FEBRUARY 1999

(P = NS). The involvement of FcεRII in cytokine production induced by IgE immune complexes was demonstrated by preincubation of AMs (n = 3) with F(ab) fragments of anti-FcεRII antibodies that strongly inhibited the production of TNF and IL-1β (90% ± 8.4% and 80.6% ± 12.9% inhibition, respectively). The addition of F(ab) fragments of anti-FcεRII antibodies alone did not modulate cytokine secretion (data not shown). A significant increase in TNF and IL-1β secretion by AMs from patients with asthma and from control subjects was also observed after 4 hours of activation by anti-FcεRII antibodies and by IgE plus anti-IgE (Table I). There is no significant difference between both groups of patients (P = NS).

IgE-dependent activation induces the production of IL-8, MCP-1, and MIP-1α

B

C FIG 2. Production of chemokines, namely IL-8 (A), MCP-1 (B), and MIP-1α (C), by AMs from healthy subjects and patients with allergic asthma. The secretion of chemokines was evaluated in 18hour supernatants from AMs incubated in medium alone, in the presence of either an unrelevant mouse monoclonal IgG1 antibody (control IgG1) or a mouse monoclonal IgG1 antibody antiCD23, and after incubation with IgE and anti-IgE. *P < .05 versus AMs from control subjects cultured in medium alone; #P < .01 versus AMs from the same group cultured in medium alone.

IL-1β (Fig l, B). IgE-dependent activation significantly upregulated the synthesis of TNF and IL-1β in both groups of subjects (P < .01). Anti-FcεRII antibodies significantly enhanced (P < .01) the production of TNF and IL-1β at a level similar to that detected in cells stimulated by IgE and anti-IgE in both groups of subjects. In contrast, control IgG1 antibodies did not. Secretion of TNF and IL-1β by AMs stimulated by IgE receptors did not differ between control subjects and patients with asthma

IL-8, MCP-1, and MIP-1α concentrations were measured in supernatants from AMs cultured for 18 hours. Spontaneous secretion of IL-8 by AMs was significantly higher in patients with allergic asthma compared with control subjects (Fig 2, A), whereas there was no significant difference for MCP-1 and MIP-1α (Fig 2, B and C, respectively). Activation by IgE plus anti-IgE and by anti-FcεRII antibodies markedly enhanced the secretion of IL-8, MCP-1, and MIP-1α by AMs from both patients with allergic asthma and control subjects (P < .05). There was no significant difference between the response to IgE-dependent stimulation in both groups of subjects. In contrast, unrelevant antibodies did not modulate chemokine production. AM sensitization by IgE before the addition of anti-IgE antibody significantly enhanced chemokine production compared with AMs activated by anti-IgE alone (P < .05; data not shown), demonstrating the specificity of IgE-dependent activation. The addition of F(ab) fragments of anti-FcεRII antibodies alone did not modulate cytokine secretion. AM preincubation with F(ab) fragments of anti-FcεRII antibodies inhibited the secretion of IL-8, MCP-1, and MIP-1α induced by IgE immune complexes (93.8% ± 1.6%, 82% ± 6.1%, and 85.6% ± 11.1% inhibition, respectively), showing the involvement of FcεRII in the production of chemokines induced by IgE-dependent activation. In some cases, synthesis of chemokines was also evaluated at 4 hours. As mentioned earlier, the addition of IgE plus anti-IgE and of anti-FcεRII antibodies increased the secretion of IL-8, MCP-1, and MIP-1α by AM from patients with asthma (Table I). With AMs from control subjects, activation by IgE receptors only significantly upregulated the production of IL-8 and MIP-1α (P < .05). In AM supernatants collected at 4 hours, levels of MCP-1 are clearly lower than those of the other chemokines; however, their levels were similar at 18 hours.

IgE-dependent activation induces the production of IL-1ra and IL-10 In parallel with the assays of chemokines and proinflammatory cytokines, we measured the levels of antiin-

Gosset et al 293

J ALLERGY CLIN IMMUNOL VOLUME 103, NUMBER 2, PART 1

TABLE I. Production of chemokines and proinflammatory and antiinflammatory cytokines by AMs activated for 4 hours by anti-CD23 antibodies and IgE plus anti-IgE antibodies AMs from control subjects Medium alone

TNF-α (ng/mL) IL-1β (pg/mL) IL-8 (ng/mL) MCP-1 (ng/mL) MIP-1α (ng/mL) IL-1ra (ng/mL) IL-10 (pg/mL)

Anti-CD23

0.3 ± 0.2 15.2 ± 3.5 20.2 ± 3.7 0.7 ± 0.3 1.8 ± 0.4 3.7 ± 1.6 4.1 ± 1.5

AMs from patients with asthma

IgE + anti-IgE

1.3 ± 0.4 58.7 ± 28.5* 28.8 ± 4.5* 0.9 ± 0.4 13.2 ± 3.6* 3.9 ± 1.3 16.2 ± 5.1

1.5 ± 0.5* 78.9 ± 47.3* 33.6 ± 4.5* 3.1 ± 1.6 12.4 ± 3.2* 4.5 ± 1.5 16.7 ± 5.2

Medium alone

Anti-CD23

0.2 ± 0.2 13.4 ± 9.5 22.6 ± 10.2 0.2 ± 0.1 6.8 ± 4.2 3 ± 0.4 3.7 ± 1.8

1 ± 0.3* 74.4 ± 45.4* 28.3 ± 9.9* 0.5 ± 0.1* 13.6 ± 7.6* 3.3 ± 0.5 8.5 ± 3.2

IgE + anti-IgE

1.2 ± 0.4* 116 ± 67* 33.2 ± 12.6* 2.2 ± 1.5* 17.9 ± 10.8* 4 ± 0.6 11 ± 5

AMs were isolated from BAL fluid obtained from normal subjects (n = 6) and patients with allergic asthma (n = 6). *P < .05 versus AMs in medium alone.

TABLE II. Percentage of increase in cytokine production by AMs from control subjects and patients with allergic asthma measured at 20 hours Control subjects

TNF IL-10 IL-1β IL-1ra

Anti-CD23

IgE plus anti-IgE

1166 ± 349 633 ± 272 1147 ± 738 225 ± 70

1985 ± 615 724 ± 364* 1358 ± 884 272 ± 91*

Patients with allergic asthma Anti-CD23

1045 ± 294 414 ± 123† 1022 ± 290 49 ± 43*†

IgE plus anti-IgE

1187 ± 302 497 ± 163* 878 ± 206 61 ± 39*†

The percentage was measured for macrophages activated by anti-CD23 antibodies and IgE plus anti-IgE antibodies in comparison with levels obtained in supernatants from AMs maintained in medium alone. The values are given as mean ± SEM of percentages obtained for individual patients. *P < .05 versus the percentages of increase obtained for TNF with the same stimulus. †P < .05 versus the percentages of increase obtained for IL-1β with the same stimulus.

flammatory cytokines in AM supernatants collected at 18 hours. Concentrations of IL-1ra (Fig 3, A) and IL-10 (Fig 3, B) in adherent AMs were not significantly different between healthy subjects and patients with asthma. Activation by either anti-IgE or anti-CD23 antibodies increased significantly the secretion of IL-10 by AMs from patients with asthma and from control subjects (P < .01). Concerning IL-1ra, cytokine production by AMs from healthy subjects was significantly higher with both types of stimulation by IgE receptors compared with unstimulated cells (P < .01), whereas IL-1ra secretion by AMs from patients with asthma was not increased (P = NS). As previously shown for proinflammatory cytokines and chemokines, AM preincubation with F(ab) fragments of anti- FcεRII antibodies blocked the secretion of IL-10 and IL-1ra induced by IgE immune complexes (89.5% ± 5% and 95.6% ± 10% inhibition, respectively), whereas F(ab) fragments of anti-FcεRII antibodies alone had no effect on cytokine production (data not shown). The secretion of IL-1ra and IL-10 was also evaluated after 4 hours of stimulation. IgE-dependent activation for 4 hours did not significantly modulate synthesis of both cytokines by AM from control subjects and from patients with asthma (Table I). The amount of IL-1ra and IL-10 increased lately in comparison with the proinflammatory cytokines.

IgE-dependent activation upregulates mRNA expression for cytokines mRNA expression for TNF, IL-8, MCP-1, and IL-10 was evaluated after 4 hours of incubation by RT-PCR in

AMs from 6 healthy subjects and 6 patients with asthma. The results, expressed as a ratio of GAPDH mRNA expression, showed that the addition of IgE plus anti-IgE or anti-CD23 antibodies clearly increased mRNA expression of TNF, IL-8, and IL-10 (an increase of about 2 times; Fig 4). IgE-dependent stimulation upregulated mRNA expression at a lower level for MCP-1 than for the other cytokines, the difference being statistically significant for anti-CD23 antibody-stimulated AMs from both groups and for anti-IgE-activated AMs from patients with asthma (P < .05).

Modulation of the balance between proinflammatory and antiinflammatory cytokines by IgE-dependent activation in AMs From the results, one can suggest that on IgE-dependent activation, the proinflammatory cytokines were upregulated earlier and at a higher level than the secretion of IL-1ra and IL-10. To confirm this hypothesis, we determined the increased percentage in cytokine synthesis induced by IgE-dependent activation for 18 hours as the ratio between levels obtained in AMs activated by IgE-dependent mechanisms and those detected for AMs incubated in medium alone. These data showed that the lowest percentages were obtained for IL-1ra and the highest for TNF. In patients with asthma, the percentages of increase for TNF and IL-1β obtained for AMs were higher than those for IL-1ra and IL-10 (Table II). In control subjects, only the percentage of increase obtained for AMs stimulated with anti-IgE was significantly higher

294 Gosset et al

J ALLERGY CLIN IMMUNOL FEBRUARY 1999

DISCUSSION

A

B FIG 3. Production of antiinflammatory cytokines, namely IL-1ra (A) and IL-10 (B), by AMs from healthy subjects and patients with allergic asthma. Secretion of IL-1ra and IL-10 was evaluated in 18hour supernatants from AMs incubated in medium alone, in the presence of either an unrelevant mouse monoclonal IgG1 antibody (control IgG1) or a mouse monoclonal IgG1 antibody antiCD23, and after incubation with IgE and anti-IgE. #P < .01 versus AMs from the same group cultured in medium alone.

for TNF than for IL-1ra and IL-10. It is noteworthy for TNF that the increased percentage was slightly higher for control subjects than for patients with asthma (P = NS); this can be explained by the significant difference in the baseline level of TNF secretion between both groups. Another way to define such a balance was to evaluate the ratio between proinflammatory and antiinflammatory cytokines detected in each AM supernatant collected at 18 hours. The ratio TNF/IL-10 was significantly higher in unstimulated AMs from patients with asthma as compared with AMs from control subjects (727 ± 282 and 55 ± 12 AU, respectively), whereas the ratio IL-1ra/IL-1β was not (3797 ± 840 and 1706 ± 562, respectively). Activation by anti-CD23 and by anti-IgE slightly increased TNF/IL-10 ratio (at least a 2-fold increase); however, the difference was not statistically significant for control subjects (312 ± 125 and 1119 ± 606, respectively) and for patients with asthma (1177 ± 388 and 1461 ± 394, respectively) compared with unstimulated cells. Stimulation by anti-CD23 or IgE plus anti-IgE significantly decreased the ratio IL1ra/IL-1β in AMs from patients with asthma (899 ± 312 and 888 ± 299, respectively) compared with unstimulated cells, whereas it did not in the AMs from control subjects (773 ± 197 and 1113 ± 425, respectively).

In this study, we demonstrated that IgE-dependent stimulation of AMs increased the production of chemokines (IL-8, MCP-1, and MIP-1α) and antiinflammatory cytokines (IL-1ra and IL-10) in addition to its effect on the secretion of proinflammatory cytokines. However, the IgE-induced production of antiinflammatory cytokines appeared later and was amplified in a lesser extent than proinflammatory cytokines. Moreover, we showed that the receptor of low affinity (FcεRII or CD23) was involved in the secretion of these cytokines, as shown by the inhibitory effect of F(ab) fragments of anti-CD23 antibodies on the activation process induced via IgE receptor and by activation in the presence of native anti-CD23 antibodies. The expression of both FcεRI and FcεRII receptors were detected on AMs; however, the level of expression of FcεRI remained very low compared with FcεRII as shown by immunofluorescence (unpublished data). In our work, the response of AMs from patients with asthma and control subjects is similar after CD23 cross-linking, demonstrating that the main difference between both groups is the presence of allergen-specific IgE and not the level of CD23 expression by AMs. Nevertheless, we observed some differences between control subjects and patients with allergic asthma, mainly for the secretion of TNF and of IL-1ra. The lack of differences in CD23 expression suggest that these variations are probably due to modifications of the cytokine releasability of AMs. In BMs, the triggering of CD23 antigen by anti-CD23 antibodies and by IgE immune complexes induced the production of TNF and IL-10.22,23 These authors also reported that the production of TNF was dependent on the synthesis of nitric oxide and of cGMP,22 whereas the transduction signals involved in AMderived cytokine secretion are unknown. These data suggest that maturation of BMs in AMs does not modify the response to IgE-dependent activation at least concerning the synthesis of TNF and IL-10. Concerning IL-1β and IL-1ra, the synthesis of IL-1β was increased by IgE-dependent activation in both groups of subjects, whereas stimulation by IgE receptors only enhanced IL-1ra secretion by AMs from control subjects. IL-1β and TNF and IL-6, which are produced in the same conditions, are multifunctional cytokines involved in the regulation of the immune response, hematopoiesis, and inflammation. IL-1β and TNF upregulate endothelial cell function, such as adhesion molecule expression and cytokine production.15 Moreover, IL-1β modulates recruited leukocyte effector function; it increases oxygen metabolite generation and leukotriene production. Our data suggested that AMs activated by IgE receptors may be one cell source of IL-1β observed in BAL fluid from patients with asthma either during symptomatic episodes24 or after segmental allergen challenge.25 The implication of IL-1ra in asthma is not well documented, but its function is probably to damper IL1–mediated inflammation in vivo as suggested in models of endotoxic shock, sepsis, graft-versus-host disease, and

J ALLERGY CLIN IMMUNOL VOLUME 103, NUMBER 2, PART 1

Gosset et al 295

FIG 4. Expression of mRNA encoding for TNF, IL-10, IL-8, and MCP-1 was measured by RT-PCR in AMs from control subjects (n = 6) and patients with allergic asthma (n = 6). AMs were incubated for 4 hours in medium alone, in the presence of either an unrelevant mouse monoclonal IgG1 antibody (control IgG1) or a mouse monoclonal IgG 1 antibody anti-CD23, and after incubation with IgE and anti-IgE. The expression of mRNA was quantified by densitometry, and the results were expressed as the ratio between cytokine and GAPDH mRNA (×100). *P < .05 versus AMs from the same group cultured in medium alone.

rheumatoid arthritis.26 IL-1ra competes with IL-1β for type I IL-1 receptor without inducing signal transduction or agonist activities.27 We have also found that in control subjects the secretion of IL-1ra by AMs activated by IgE receptors appears later (after 4 hours) than that of IL-1β, suggesting that IL-1ra might only control the very late effect of IL-1β. The levels of IL-1ra in AMs activated by IgE receptors were about 1000-fold higher than that of IL-1β. This could explain that we previously found no detectable IL-1 activity in AM supernatants obtained at baseline and after activation by IgE receptors.11 Whereas IL-1ra/IL-1β ratios were not modified after IgE-dependent stimulation in AMs from control subjects, the same activation significantly increased these ratios in patients with asthma, showing a specific modification of this bal-

ance in patients with asthma. Moreover, in terms of systemic treatment in humans and animals, it seems that the ratio IL-1ra/IL-1 needs to be elevated (until 100,000fold) to limit the severity of the disease and to represent a therapeutic approach, compared with the 100- to 1000fold excess observed naturally.26 An important point is that IgE-dependent activation of AMs increased the production of chemokines, such as IL-8, MCP-1, and MIP-1α. Such a process could be involved in the IL-8 secretion observed in lavage fluid from patients with atopy after a segmental25 or a nasal allergen challenge.28 Moreover, an increased expression of β-chemokines was also detected in atopic asthma.29 In asthma, these chemokines are thought to participate in the recruitment and the activation of leukocytes, particu-

296 Gosset et al

larly of eosinophils and basophils. IL-8 induces activation and transendothelial migration of neutrophils, as well as being chemoattractant for lymphocytes,30 basophils,31 and primed eosinophils.32 MCP-1 and MIP1α are chemotactic for monocytes,33 whereas MIP-1α is also chemotactic for eosinophils and naive T cells.34 MCP-1, with MCP-3, induces high levels of histamine release in basophils35,36 similar to those obtained with the complement component C5a. MIP-1α37 and high concentrations of IL-838 have a similar effect on basophils with a lower activity than MCP-1. Kinetics of chemokine production by AMs activated by IgE receptors seem different between MCP-1 on the one hand and MIP-1α and IL-8 on the other hand. As previously shown in another model, the secretion of MCP-1 was delayed compared with that of IL-8.39 This suggests a particular role of MCP-1 in the late allergic reaction probably in mononuclear cell recruitment and basophil activation. In this study, we demonstrated that IgE-dependent activation induces IL-10 production by AMs. Impairment of IL-10 production has been observed in AMs from patients with asthma when activated by LPS and TNF.40 However, IL-10 generation by phytohemagglutinin-stimulated BAL cells was higher in patients with asthma compared with control subjects.41 Our data showed an impaired TNF/IL-10 ratio in unstimulated AMs from patients with allergic asthma compared with those from control subjects. On IgE-dependent stimulation, AMs from both groups release significant amounts of IL-10. This is in agreement with Robinson et al42 who demonstrated an increase of IL-10 production in patients with mild asthma challenged by allergen. All these data suggest that IL-10 production at baseline is impaired in patients with asthma compared with control subjects; however, stimulation by IgE receptor (in vitro and in vivo) or by mitogen could induce IL-10 synthesis in patients with asthma. IL-10 was originally discovered as a cytokine synthesis inhibitory factor produced by TH2 cells that could inhibit IFN-γ production by TH2 cells responding to antigen. The inhibitory effect of IL-10 on cytokine synthesis by T cells was due, in part, to the inhibition of the expression of membrane bound costimulators required for T-cell activation,43 adhesion molecules,18 and the production of proinflammatory cytokines.16 In these studies, high concentrations of rhIL-10 (1 to 10 ng/mL) were used.16,17 However, it has been demonstrated that the autologous IL-10 production by AMs and monocytes may limit the production of proinflammatory cytokines induced by viruses and LPS, respectively.16,44 In these cases, the concentrations present in the cell environment are certainly close to those observed in AM cultures. Nevertheless, analysis of the TNF/IL-10 ratio and of the percentages of increase showed that IgE-dependent stimulation favored the production of TNF compared with IL-10, suggesting that the inhibitory effect of IL-10 on mononuclear cell function was limited in this situation. As previously shown, we

J ALLERGY CLIN IMMUNOL FEBRUARY 1999

also reported that IL-10 was produced later than TNF and IL-1β, which suggests that IL-10 might shorten the duration of the secretion of proinflammatory cytokines and, in consequence, that of the inflammatory reaction. Nevertheless, mRNA expression for IL-10 was clearly increased by IgE-dependent activation, whereas protein secretion was not obvious. Additional studies are required to elucidate the mechanism controlling IL-10 synthesis in this model. Although the release of proinflammatory cytokines and chemokines is significantly increased by IgE-dependent stimulation at 4 hours, the secretion of IL-10 and IL1ra is not modulated at that time. So, during the first hours after activation, the balance between proinflammatory and antiinflammatory cytokines is largely in favor of the former. The secretion of antiinflammatory cytokines may contribute to limiting the chronic inflammatory reaction at distance from the allergen exposure period. The fact that the ratio between proinflammatory and antiinflammatory cytokines is increased after stimulation by IgE receptors compared with cells at baseline and that the kinetics of secretion are different suggest that, in the lungs of patients with asthma, AMs exposed to allergen will favor the development of inflammation, at least during the first hours. Because macrophages represent most of the cells colonizing alveolar spaces and bronchial lumen,45 our data and the in vivo activation of AMs after allergen challenge20,46 suggest that AMs are involved in the development of airway inflammation at least by the secretion of cytokines. REFERENCES 1. Melewicz FM, Zieger RS, Mellon MH, O’Connor RD, Spiegelberg HL. Increased peripheral blood monocytes with Fc receptors for IgE in patients with severe allergic disorders. J Immunol 1981;126:1592-5. 2. Williams J, Johnson S, Mascali JJ, Smith H, Rosenwasser JL, Borish L. Regulation of low affinity IgE receptor (CD23) expression on mononuclear phagocytes in normal and asthmatic subjects. J Immunol 1992;149:2823-8. 3. Joseph M, Tonnel AB, Torpier G, Capron A, Arnoux B, Benveniste J. Involvement of IgE in the secretory process of alveolar macrophages from asthmatic patients. J Clin Invest 1983;221-30. 4. Liu F-T, Albrandt K, Mendel E, Kulzycki A Jr, Orida NK. Identification of an IgE-binding protein by molecular cloning. Proc Natl Acad Sci USA 1985;82:4100-4. 5. Maurer D, Fiebiger E, Reininger B, Wolff-Winiski B, Jouvin MH, Kilgus O, et al. Expression of functional high affinity immunoglobulin E receptors (FcεRI) on monocytes of atopic individuals. J Exp Med 1994;179:745-50. 6. Humbert M, Grant A, Taborda-Barata L, Durham SR, Pfister R, Menz G, et al. High-affinity IgE receptor (FcεRI)-bearing cells in bronchial biopsies from atopic and non-atopic asthma. Am J Respir Crit Care Med 1996;153:1931-7. 7. Fuller RW, Morris PK, Sykes RD, Varndell IM, Kemeny DM, Cole PJ, et al. Immunoglobulin E-dependent stimulation of human alveolar macrophages: significance in type 1 hypersensitivity. Clin Exp Immunol 1986;65:416-26. 8. Gosset P, Tonnel AB, Joseph M, Prin L, Mallart A, Charon J, et al. Secretion of a chemotactic factor for neutrophils and eosinophils by alveolar macrophages from asthmatic patients. J Allergy Clin Immunol 1984;74:827-34. 9. Borish L, Mascali JJ, Rosenwasser LJ. IgE-dependent cytokine produc-

J ALLERGY CLIN IMMUNOL VOLUME 103, NUMBER 2, PART 1

10.

11.

12.

13. 14. 15. 16.

17.

18.

19.

20.

21. 22.

23.

24.

25.

26.

27. 28.

29.

tion by human peripheral blood mononuclear phagocytes. J Immunol 1991;146:63-7. Gosset P, Tsicopoulos A, Wallaert W, Joseph M, Capron A, Tonnel AB. Tumor necrosis factor alpha and interleukin-6 production by human mononuclear phagocytes from allergic asthmatics after IgE-dependent stimulation. Am Rev Respir Dis 1992;146:768-74. Gosset P, Lassalle P, Tonnel AB, Dessaint JP, Wallaert B, Prin L, et al. Production of an IL 1 inhibitory factor by human alveolar macrophages from normals and allergic asthmatic patients. Am Rev Respir Dis 1988;138:40-6. Borish L, Mascali JJ, Dishuck J, Beam WR, Martin RJ, Rosenwasser RJ. Detection of alveolar macrophage-derived IL-1β in asthma: inhibition with corticosteroids. J Immunol 1992;149:3078-82. Baggiolini M, Loetscher P, Moser B. Interleukin-8 and the chemokine family. Int J Immunopharmacol 1995;17:103-8. Baggiolini M, Dahinden CA. CC chemokines in allergic inflammation. Immunol Today 1994;15:127-33. Dinarello CA. Interleukin-1 and interleukin-1 antagonism. Blood 1991;77:1627-52. de Waal Malefyt R, Abams J, Benett B, Fidgor CG, de Vries J. Interleukin-10 (IL-10) inhibits cytokine synthesis by human synthesis: an autoregulatory role of IL-10 produced by monocytes. J Exp Med 1991;174:1209-20. Armstrong L, Jordan N, Millar A. IL-10 regulation of tumour necrosis factor alpha (TNF-alpha) from human alveolar macrophages and peripheral blood monocytes. Thorax 1996;51:143-9. Willems F, Marchand A, Delville JP, Gerard C, Delvaux A, Velu T, et al. Interleukin-10 inhibits B7 and intercellular adhesion molecule-1 expression on human monocytes. Eur J Immunol 1994;24:1007-9. Galve de Rochemontaix B, Nicod LP, Chicheportiche R, Lacraz S, Baumberger C, Dayer JM. Regulation of IL-1ra, IL-lα and IL-lβ production by human alveolar macrophages with phorbol myristate acetate, lipopolysaccharide and IL-4. Am J Respir Cell Mol Biol 1993;8:160-8. Tonnel AB, Joseph M, Gosset P, Fournier E, Capron A. Stimulation of alveolar macrophages in asthmatic patients after local provocation test. Lancet 1983;1:1406-8. Ruff MR, Oifford GE. Tumor necrosis factor. Lymphokine Res 1981;2:35-40. Mossalayi MD, Paul-Eugène N, Ouaaz F, Arock M, Kolb JP, Kilchherr E, et al. Involvement of FcεRII/CD23 and L-arginine dependant pathway in IgE-mediated stimulation of human monocyte functions. Int Immunol 1994;6:931-4. Dugas N, Vouldoukis I, Bécherel P, Arock M, Debrè P, Tardieu M, et al. Triggering of CD23b antigen by anti-CD23 monoclonal antibodies induces interleukin-10 production by human macrophages. Eur J Immunol 1996;26:1394-8. Broide DH, Lotz M, Cuomo AJ, Coburn DA, Federman EC, Wasserman SI. Cytokines in symptomatic asthma airways. J Allergy Clin Immunol 1992;89:958-67. Virchow J-C, Walker C, Hafner D, Kortsik C, Werner P, Matthys H, et al. T cells and cytokines in bronchoalveolar lavage fluid after segmental allergen provocation in allergic asthma. Am J Respir Crit Care Med 1995;151:960-8. Fisher E, Van Zee KJ, Marano MA, Rock CS, Kenney JS, Poutsiaka DD, et al. Interleukin-1 receptor antagonist circulates in experimental inflammation and in human disease. Blood 1992;79:2196-200. Arend WP. Interleukin-1 receptor antagonist: a new member of interleukin-1 family. J Clin Invest 1991;88:1445-51. Gosset P, Tillie-Leblond I, Malaquin F, Durieu J, Wallaert B, Tonnel AB. Interleukin-8 secretion in patients with allergic rhinitis after an allergen challenge: interleukin-8 is not the main chemotactic factor present in nasal lavages. Clin Exp Allergy 1997;27:379-88. Powell N, Humbert M, Durham SR, Assoufi B, Kay AB, Corrigan CJ. Increased expression of mRNA encoding RANTES and MCP-3 in the bronchial mucosa in atopic asthma. Eur Respir J 1996;9:2454-60.

Gosset et al 297

30. Larsen CG, Anderson AO, Appella E, Oppenheim JJ, Matsushima K. The neutrophil-activating protein (NAP-1) is also chemotactic for T lymphocytes. Science 1989;243:1464-8. 31. Leonard EJ, Skeel A, Yoshimura T, Noer K, Kutvirt S, Van Epps D. Leukocyte specificity and binding of human neutrophil attractant/activator protein-1. J Immunol 1991;144:1323-30. 32. Warringa RAJ, Koenderman L, Kok PTM, Kreukniet J, Bruijnzeel LB. Modulation and induction of eosinophil chemotaxis by granulocytemacrophage colony-stimulating factor and interleukin-3. Blood 1991;77:2694-700. 33. Yoshimura T, Robinson EA, Tanaka S, Appella E, Leonard EJ. Purification and amino-acid analysis of two monocyte chemoattractants produced by phytohemagglutinin-stimulated human blood mononuclear leukocytes. J Immunol 1989;142:1956-62. 34. Rot A, Krieger M, Brunner T, Bishoff SC, Schall TJ, Dahinden CA. RANTES and macrophage inflammatory protein-lα induce the migration and activation of normal human eosinophil granulocytes. J Exp Med 1992;176:1489-95. 35. Kuna P, Reddigari S, Rucinki D, Oppenheim JJ, Kaplan AP. Monocyte chemotactic and activating factor is a potent histamine-releasing factor for human basophils. J Exp Med 1992;175:489-93. 36. Alam R, Lett-Brown MA, Forsythe PA, Anderson-Walters DJ, Kenamore C, Kormos C, et al. Monocyte chemotactic and activating factor is a potent histamine-releasing factor for basophils. J Clin Invest 1992;89:723-8. 37. Bischoff SC, Krieger M, Brunner T, Rot A, Tscharner VV, Baggiolini M, et al. RANTES and related chemokines activate human basophil granulocytes through different G protein-coupled receptors. Eur J Immunol 1993;23:761-70. 38. White MV, Yoshumira T, Hook W, Kaliner MA, Leonard EJ. Neutrophil attractant/activation protein-1 (NAP-1) causes human basophil histamine release. Immunol Lett 1989;22:151-6. 39. Liebler JM, Kunkel SL, Burdick MD, Standiford TJ, Rolfe MW, Strieter RM. Production of IL-8 and monocyte chemotactic peptide-1 by peripheral blood monocytes: disparate responses to phytohemagglutinin and lipopolysaccharide. J Immunol 1994;152:241-9. 40. Calhoun WJ, Hinton KL, Brick JJ, Sharma AJ, Rosen S. Spontaneous and stimulated IL-10 release by alveolar macrophages (AM) but not blood monocytes (BM) is reduced in allergic asthmatics (AA) [abstract]. Am J Respir Crit Care Med 1996;153:A881. 41. Jarjour NN, Kelly EAB, Rodriguez R, Swenson CA, Schroth M, Busse WW. Enhanced generation of airways cytokines in asthma patients is associated with increased airway obstruction [abstract]. Am J Respir Crit Care Med 1996;153:A878. 42. Robinson DS, Tsicopoulos A, Meng Q, Durham S, Kay AB, Hamid Q. Increased interleukin-10 messenger mRNA expression in atopic allergy and asthma. Am J Respir Cell Mol Biol 1996;14:113-7. 43. de Waal Malefyt R, Haanen J, Spits H, Roncarolo M-G, te Velde A, Figdor C, et al. Interleukin-10 (IL-10) and viral IL-10 strongly reduce antigen-specific human T cell proliferation by diminishing the antigen-presenting capacity of monocytes via downregulation of class II major histocompatibility complex expression. J Exp Med 1991;174:915-24. 44. Panuska JR, Merolla R, Rebert NA, Hoffmann SP, Tsivitse, Cirino NM, et al. Respiratory syncitial virus induces IL-10 by human alveolar macrophages: suppression of early cytokine production and implications for incomplete immunity. J Clin Invest 1995;96:2245-53. 45. Rankin JA, Marcy T, Rochester CL, Sussman J, Smith S, Buckley P, et al. Human airway macrophages: a technique for their retrieval and a descriptive comparison with alveolar macrophages. Am Rev Respir Dis 1992;145:928-33. 46. Gosset P, Tsicopoulos A, Wallaert B, Vannimenus C, Joseph M, Tonnel AB, et al. Increased secretion of tumor necrosis factor α and interleukin 6 by alveolar macrophages consecutively to the late asthmatic reaction after bronchial allergen challenge. J Allergy Clin Immunol 1991;88:56171.