INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

In vivo, there is strong evidence that immunoglobulin E (IgE) plays an important role in bronchial hyperresponsiveness (1- 3). In asthmatic patients, symptoms, hyperresponsiveness and IgE concentration in serum are closely related (1, 2). Several immunohistologic studies of bronchial biopsies performed in asthmatic patients have reported an increase in inflammatory cells bearing IgE receptor when compared with normal control subjects (4). Recently, it has been shown that, both in asthmatic and atopic patients, the majority of cells expressing the high-affinity receptor for IgE are mast cells and macrophages (7).

In vitro, passive sensitization of human isolated airways with serum from allergic asthmatic patients provides a good opportunity to study the role of IgE in bronchial reactivity. It has been previously shown that passive sensitization with asthmatic serum containing a high concentration of IgE alters the mechanical response of human isolated airways to electrical field stimulation, nonspecific agonists such as histamine, KCl, tachykinins, or relaxant compounds (8). We have demonstrated that passive sensitization of human airways could be mimicked using monoclonal IgE and an anti-IgE challenge (13). Finally, Watson and coworkers recently provided evidence that hyperresponsiveness of human airways was closely related to the total IgE concentration in the "sensitizing" serum (14). However, target cells bearing IgE during the passive sensitization procedure have not yet been determined. A variety of cell types express specific receptors for IgE. Mast cells, basophils, monocytes, and Langerhan's cells express the high-affinity receptor for IgE named FcRI whereas platelets, macrophages and eosinophils express the low-affinity receptor for IgE, CD23/FcRII (15). Several experimental results ascribe to mast cell, a subpopulation of which is closely localized to the smooth muscle layer in bronchi, a potential role in inducing hyperresponsiveness. In spontaneously sensitized tissues, it has recently been demonstrated that mast cell numbers were increased within smooth muscle (16). In addition, mast cell- derived mediators such as histamine and sulfidopeptide leukotrienes play an important role in mediating both intrinsic tone and antigen-induced contraction of human isolated bronchi (17, 18). Finally, it has been demonstrated in dogs that mast cell-derived tryptase enhances the contractile response to histamine (19), whereas in humans, tryptase could potentiate histamine-induced contraction of isolated bronchi obtained from spontaneously sensitized lung tissue (20).

The aims of the present study were to define the phenotype of cells involved in IgE-induced passive sensitization, and to evaluate the proportion of mast cells implicated in this mechanism.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Tissue Donors

Human lung was obtained at thoracotomy as previously described (9, 21) from 10 patients undergoing resection for pulmonary carcinoma. Morphometric data are given in Table 1. Patients were nonatopic (total serum IgE < 100 IU/ml and negative specific IgE for usual inhaled allergens). They did not report any clinical history of asthma, and did not receive any antiasthmatic medication.

Tissue Preparation

After resection, tissues were immediately transferred to the laboratory in ice-cold oxygenated Krebs-Henseleit (KH) solution (composition in mM: 118.4 NaCl, 4.7 KCl, 2.5 CaCl₂ · 2 H₂O, 1.2 MgSO₄ · 7 H₂O, 1.2 KH₂PO₄, 25.0 NaHCO₃, 11.1 D-glucose). From a macroscopically tumor-free part of each of the specimens, segments of human bronchus (third to fourth generation) 30 to 40 mm in length and 3 to 5 mm in internal diameter were carefully dissected from surrounding parenchyma without damaging the airway membrane. After removal of adhering fat and connective tissue, bronchi were cut into rings measuring 4 to 5 mm in length. For each lung specimen, one bronchial segment was dissected out for immunohistochemical analysis.

Sample Processing and Immunochemistry

Bronchial segments were cut into small fragments of 2/2 mm for the histologic analysis. Specimens were placed into ice-cooled acetone containing 2 mM phenylmethylsulfonyl fluoride and 20 mM iodoacetamide to inhibit protease activity, and fixed overnight at 20° C. The following day the samples were transferred into acetone at room temperature for 15 min, and then in methylbenzoyl for a further 15-min period. The tissue was then immersed in glycolmethacrylate (GMA) monomers (Polysciences, Northampton, UK) at 4° C for 7 h. During this period of incubation, the GMA solution was changed three times. Embedding resin was prepared with benzoyl peroxide, GMA monomers, and N,N-dimethylaniline and polymerized overnight at 4° C (22). The GMA blocks were stored in airtight containers at 20° C, until used for immunostaining.

The GMA sections were cut at 2 µm thickness with ultramicrotome and floated onto ammonia water (1/500), picked onto 0.01% poly-L-lysine glass slides, and allowed to dry at room temperature for 1 to 6 h.

Endogenous peroxidase was inhibited using a solution of 0.1% sodium azide and 0.3% hydrogen peroxide for 30 min, followed by two rinses in Tris-buffered saline (TBS) adjusted at pH 7.6. Undiluted culture supernatant consisting of RPMI 1640 (Gibco BRL, Cergy-Pontoise, France), fetal calf serum, and bovine serum albumin was applied for 30 min. Passive sensitization was performed as previously described for bronchial rings (8, 12, 13) but applied to tissue sections. Briefly, sections were incubated for 2 h at 37° C with serum (200 to 500 µl applied to the slide subsequently covered with a coverslip) from a variety of atopic asthmatic patients (23) whose concentration of both total and specific IgE to Dermatophagoides pteronyssinus (D.pter) was above 1,000 IU/ml and 17.5 Phadebas RAST units (PRU)/ml (i.e., RAST 4+ titer), respectively, or TBS supplemented with monoclonal human IgE (Calbiochem, Meudon, France; 1,000 IU/ ml). Control tissues were incubated with either serum from healthy subjects (nonatopic, negative prick tests, total IgE < 10 IU/ml) or TBS.

After washing with TBS, sections were incubated overnight at room temperature with a variety of mouse monoclonal antibodies (mAb) including anti-human IgE (Immunotech, Marseille, France), anti-human tryptase antibody AA1 (25), anti-human chymase (Chemicon, Souffelweyersheim, France), anti-human activated eosinophils (EG2; Pharmacia, Saint-Quentin-en-Yvelines, France), anti-human monocytes CD14 (Dako, Trappes, France), and anti-human T lymphocytes CD3 (Dako). Control slides were treated similarly, omitting the primary mAb. After rinsing in TBS, biotinylated rabbit anti-mouse F(ab')₂ (Dako) was applied to the sections for 2 h, and followed by the streptavidin-biotin horseradish-peroxidase complex (Dako) for a further 2-h period. After rinsing in TBS, aminoethyl carbamazole (AEC) (Sigma, Saint-Quentin-Fallavier, France) in acetate buffer (pH 5.2) and hydrogen peroxide were used as substrate to develop a peroxide-dependent red color reaction at 37° C. The sections were rinsed and counterstained with Mayer's hematoxylin (22). Finally, slides were mounted with two drops of Crystal/mount (Biomeda, Foster City, CA) for 20 min at 80° C.

Quantification of Immunostaining

Light microscopy was performed using an Optiphot microscope (Nikon, Tokyo, Japan). Cells staining positively with each mAb were counted at a magnification of ×200 in all of the specimens excluding alveoli. The total area examined was calculated by delineating the area of the bronchial section on a video interactive display system (×100) and using an appropriate software (Quancoul, Bordeaux, France). A compartmental analysis was also performed delineating bronchial epithelium, submucosa, and the smooth muscle layer. Cell counts were expressed as number of cells/mm² of tissue. Colocalization of cells in adjacent sections was also performed using the computerized system of cell recognition (Quancoul).

Statistical Analysis

Results from immunohistologic studies were analyzed by means of Wilcoxon paired rank test to compare the number of IgE-bearing cells before versus after passive sensitization, since cell counts are not normally distributed.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

A positive immunostaining for IgE-bearing cells was demonstrated in both sensitized and nonsensitized bronchial specimens. Nevertheless, in bronchial rings passively sensitized with either monoclonal IgE or atopic asthmatic serum, the number of IgE-bearing cells was significantly higher than that in nonsensitized rings, i.e., rings incubated in normal serum or with a saline buffer (Figure 1). There was no significant difference in the number of IgE-bearing cells whether specimens were sensitized with monoclonal IgE or atopic asthmatic serum with a high concentration of IgE. The distribution of IgE-bearing cells according to the different compartments in the bronchial wall is presented in Table 2. The increase in IgE-bearing cell number was observed in all of the compartments, i.e., epithelium, submucosa, and smooth muscle layer.

View larger version (15K): [in this window] [in a new window]   Figure 1.   Number of cells positive for anti-human IgE stained with monoclonal antibody d2 in GMA embedded specimens (n = 10). Sections are nonsensitized (open bars) when incubated with TBS (A) or healthy serum (B) or sensitized (filled bars) with monoclonal human IgE (A) or asthmatic serum with a high concentration of IgE (B). Values are means ± SEM of anti-IgE-positive cells per mm2.

Among the population of cells having the ability to bear IgE molecules, the majority of cells present in sections of bronchial rings was represented by mast cells as detected by antitryptase or antichymase antibodies (Figure 2A). We detected 57% as mast cells of TC subtype (MC_TC, i.e., containing both tryptase and chymase) and 43% of T subtype (MC_T, i.e., containing tryptase alone). Other cell types were mainly represented by T cells and monocytes/macrophages whereas the number of eosinophils remained very low. There was no significant difference for any of the abovementioned phenotypes between sensitized and nonsensitized tissues. Compartmental analysis of different cell types is represented in Figure 2B. The majority of mast cells, eosinophils, and lymphocytes were localized to the submucosa whereas a large proportion of CD14-positive cells was present in the epithelium. A proportion of mast cells, the majority of which was of TC subtype (75%), was observed within the smooth muscle layer.

View larger version (21K): [in this window] [in a new window]   Figure 2.   Cell phenotypes detected by monoclonal antibodies directed against specific markers in GMA embedded specimens. Antibodies are anti-human tryptase (AA1), anti-human chymase (antiC), anti-human activated eosinophils (EG2), anti-human monocytes (CD14), and anti-human lymphocytes (CD3). (A) Actual numbers of positive cells given as means ± SEM of cells per mm2 in the whole tissue section (n = 7). (B) Distribution of each phenotype expressed as a percentage, within epithelium (filled area), submucosa (open area), and smooth muscle layer (hatched area). Total number of each phenotype of cells has been normalized to 100% (n = 5).

To determine which types of cells were involved in passive sensitization, adjacent sections stained with anti-IgE and antibodies directed against specific markers were compared using a computerized system of cell recognition (Figure 3). We detected 58 to 77% of IgE-bearing cells as AA1-positive mast cells and 33 to 55% as antichymase-positive mast cells (i.e., 41.6% of MC_TC and 23.4 of MC_T). From 16 to 30% of IgE-bearing cells were positive for anti-CD14 antibody whereas a smaller proportion was detected as T lymphocytes (0-11%) or EG2-positive eosinophils (0-4%). Within bronchial epithelium, the majority of IgE-bearing cells was represented by CD14-positive cells whereas in submucosal and smooth muscle layers the prominent cell type with a positive staining for anti-IgE was represented by mast cells (Figure 4A). Regarding the distribution of the different phenotypes bearing IgE molecules, we found that the majority of IgE-bearing mast cells was present in the smooth muscle layer and in the submucosa whereas the majority of IgE-bearing CD14-positive cells remained within the epithelium (Figure 4B). After passive sensitization, the proportion of MC_TC and MC_T increased from 55 to 68% and from 26 to 53%, in submucosal and smooth muscle layers, respectively.

View larger version (82K): [in this window] [in a new window]   View larger version (90K): [in this window] [in a new window]   View larger version (80K): [in this window] [in a new window]   Figure 3.   Optic microscopic views (original magnification: ×400) of a representative specimen sensitized with monoclonal IgE, and stained with anti-human IgE d2 (A), anti-human tryptase AA1 (B), or anti-human chymase (C ), on adjacent sections. Arrows indicate IgE-positive cells in (A) that are AA1-positive in (B) and antichymase positive in (C ).

View larger version (23K): [in this window] [in a new window]   Figure 4.   Distribution of IgE-bearing cells following passive sensitization of human bronchi (n = 5). (A) Distribution of IgE-bearing cells according to the different compartments in the airway wall. Antibodies used are AA1, antichymase (anti-C), and CD14. Values are expressed as percentage ± SEM of total anti-IgE-positive cells for each compartment. (B ) Distribution of IgE-bearing cells in the different compartments according to the cell phenotypes. Values are expressed as percentage ± SEM of total anti-IgE-positive cells for each phenotype within epithelium (filled bars), submucosa (open bars), and smooth muscle layer (hatched bars).

Within the smooth muscle layer, we identified the majority of IgE-bearing cells as mast cells. Other phenotypes such as eosinophils or CD14-positive cells were occasionally found within the smooth muscle and did not contribute significantly to IgE binding. T lymphocytes were not present in this compartment of the bronchial wall. We did not detect any specific IgE binding to smooth muscle cells (Figure 5).

View larger version (87K): [in this window] [in a new window]   Figure 5.   Optical microscopic view (original magnification: ×400) of smooth muscle layer from a representative specimen sensitized with monoclonal IgE, and stained with anti-human IgE d2. Arrows indicate IgE-positive cells.

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

The results obtained in the present work indicate that mast cells represent the prominent cell type involved in passive IgE sensitization in human bronchi and that IgE-bearing mast cells are mainly localized to submucosal and smooth muscle layers in the bronchial wall.

Passive sensitization of human isolated bronchi with serum from asthmatic patients provides a good opportunity to study bronchial responsiveness in vitro with special reference to allergic factors. We and others have demonstrated that incubating isolated bronchi with serum containing a high concentration of both specific and total IgE alters the mechanical response to a variety of nonspecific agonists (8). These observations combined with epidemiologic results demonstrating a close relationship between serum IgE and both asthma symptoms and in vivo hyperresponsiveness led us to hypothesize that IgE was the main factor in serum responsible for hyperresponsiveness. However, the mechanism by which IgE may alter the mechanical response of isolated bronchi remains unknown. In vivo, bronchial biopsies performed in asthmatic patients have demonstrated the presence, in the mucosa, of several cell types bearing IgE (4). When examining the expression of IgE receptor, it has been found that the number of cells expressing the high-affinity IgE receptor (FcRI) was higher in asthmatic bronchi than in normal ones and that mast cell was the prominent cell type responsible for this expression (7). In addition, very recent data obtained in bronchi from spontaneously sensitized lung specimens have demonstrated that mast cell numbers are increased within human airway smooth muscle, suggesting an important role for this cell in hyperresponsiveness (16).

In the present study, lung specimens were not previously sensitized because tissue donors had normal concentrations of both total and specific IgE and did not report clinical history of asthma or allergy. In the absence of passive sensitization, IgE-bearing cells were, however, detected in bronchial mucosa and mainly corresponded to mast cells and monocytes-macrophages as it has been previously reported in normal donors using per-fiberoptic biopsies (7). After incubation with monoclonal IgE or asthmatic serum containing a high concentration of IgE, we observed an increase in IgE-bearing cells although the distribution of cell types did not differ from that in nonsensitized specimens. We also found a similar number of IgE-bearing cells using an alternative method of passive sensitization, i.e., incubating human isolated bronchi in atopic asthmatic serum prior to GMA embedding (results not shown). These results are consistent with those reported by Humbert and coworkers demonstrating that, in vivo, atopic, nonasthmatic subjects who are characterized by both specific IgE against one or more allergens and a high concentration of total IgE in serum had increased numbers of FcRI-positive cells when compared with nonatopic, nonasthmatic controls (7).

Regarding the distribution of IgE-bearing cells according to the different compartments in the bronchial wall, we observed that the majority of IgE binding was localized to submucosa although a third of the total population of IgE-bearing cells was detected within the smooth muscle layer. The proportion of IgE-bearing cells in epithelium was difficult to evaluate from our data. The epithelial layer could not be examined in 4 of 10 bronchial specimens because this compartment was either damaged or not involved in these tissue sections. Nevertheless, the proportion of cells bearing IgE and localized to the epithelium may contribute to a very limited extent to total IgE binding. After sensitization, we found a significant increase in anti-IgE-positive cells localized to all of the bronchial compartments although the distribution of IgE-bearing cells did not change significantly. In the present study, the increase in anti-IgE-positive cells cannot be related to a cell recruitment phenomenon in this ex vivo model. It is also unlikely that this increase corresponds to an overexpression of IgE receptors although this hypothesis should be directly examined. The most likely hypothesis is that passive sensitization reduces the number of cells bearing unoccupied IgE receptors, the distribution of which, according to the compartments in the bronchial wall, can vary from one subject to another. Comparing adjacent sections stained with anti-IgE and specific markers for IgE-bearing cells, we found that mainly mast cells were involved in IgE sensitization in human bronchi. In contrast to results previously reported (26) we found that the number of MC_TC was higher than that of MC_T and represented from 30 to 50% of IgE-bearing cells. The majority of mast cells was present in the submucosa although an important population, dominated by MC_TC, was present within the smooth muscle cell layer. The other cell type representing a large proportion of IgE-bearing cells was monocyte-macrophage as detected by anti-CD14 antibody (27). This result has to be compared with the large population of macrophages found to express the high-affinity receptor for IgE in atopic patients (7) although in our study, a small proportion of CD14-positive cells could be represented by dendritic cells that have been demonstrated to express both this marker (28) and the -chain of the FcRI receptor (6).

The present morphometric results indicating that mast cells are involved in passive IgE sensitization of human isolated bronchi further support the hypothesis that mast cell-derived proteases may be involved in the mechanism of IgE-induced hyperresponsiveness. Watson and coworkers have very recently investigated the role of IgE in hyperresponsiveness induced by passive sensitization (14). They concluded that although specific IgE determines allergen responsiveness, contractile response to histamine is not directly related to IgE but to some other factor associated with serum containing high concentrations of total IgE. Data obtained in dogs by Sekizawa and coworkers (19) combined with that recently reported in humans by Johnson and associates (20) provide some evidence that tryptase could represent an important factor inducing hyperresponsiveness in IgE-sensitized tissues. One could consider that active sensitization of lung tissues (i.e., in vivo IgE sensitization) may represent a more marked pathophysiologic status when compared with passive sensitization of tissues as examined in the present study. On the one hand, Ammit and colleagues have observed an increase in the number of mast cells in the smooth muscle of actively sensitized bronchi (16). On the other hand, actively sensitized tissues may be more sensitive to tryptase (i.e., requiring a lower tryptase concentration to exhibit histamine hyperresponsiveness) than passively sensitized tissues as observed by Johnson and associates (20). Additional studies are required to further examine the key role of mast cells in sensitization-induced bronchial hyperresponsiveness. These include examination of the effect of different mast cell-derived proteases on the mechanical response of human airways as well as IgE-mediated protease release in human lung isolated mast cells.