Mast cell adhesion to bronchial smooth muscle in asthma specifically depends on CD51 and CD44 variant 6

Authors


  • Edited by: Marc Humbert

Pierre-Olivier Girodet, MD, PhD, INSERM U885, Université de Bordeaux, CHU de Bordeaux-INSERM, Centre d’Investigation Clinique, 146 rue Léo Saignat, F-33076 Bordeaux cedex, France.
Tel.: + 33 (0) 5 57 57 15 60
Fax: + 33 (0) 5 57 57 46 71
E-mail: pierre-olivier.girodet@pharmaco.u-bordeaux2.fr

Abstract

To cite this article: Girodet P-O, Ozier A, Trian T, Begueret H, Ousova O, Vernejoux J-M, Chanez P, Marthan R, Berger P, Tunon de Lara JM. Mast cell adhesion to bronchial smooth muscle in asthma specifically depends on CD51 and CD44 variant 6. Allergy 2010; 65: 1004–1012.

Abstract

Background:  Mast cells infiltrate the bronchial smooth muscle (BSM) in asthmatic patients, but the mechanism of mast cell adhesion is still unknown. The adhesion molecules CD44 (i.e. hyaluronate receptor) and CD51 (i.e. vitronectin receptor) are widely expressed and bind to many extracellular matrix (ECM) proteins. The aims of the study are (i) to identify the role of ECM in mast cell adhesion to BSM and (ii) to examine the role of CD51 and CD44 in this adhesion.

Methods:  Human lung mast cells, human mast cell line (HMC-1), and BSM cells from control donors or asthmatic patients were cultured in the presence/absence of various cytokines. Mast cell–BSM interaction was assessed using 3H-thymidine-pulsed mast cells, confocal immunofluorescence, or electron microscopy. Adhesion molecules expression and collagen production on both cell types were evaluated by quantitative RT-PCR, western blot, and flow cytometry.

Results:  Mast cell adhesion to BSM cells mostly involved type I collagen of the ECM. Such an adhesion was increased in normal BSM cells under inflammatory condition, whereas it was maximal in asthmatic BSM cells. Blockade of either CD51 or CD44 significantly decreased mast cell adhesion to BSM. At the molecular level, protein and the transcriptional expression of type I collagen, CD51 or CD44 remained unchanged in asthmatic BSM cells or in mast cells/BSM cells under inflammatory conditions, whereas that of CD44 variant isoform 6 (v6) was increased.

Conclusions:  Mast cell–BSM cell adhesion involved collagen, CD44, and CD51, particularly under inflammatory conditions. CD44v6 expression is increased in asthmatic BSM cells.

Abbreviations
BSA

bovine serum albumine

BSM

bronchial smooth muscle

CD44v

CD44 variants isoform

ECM

extracellular matrix

FN

fibronectin

HLMC

human lung mast cell

PMA

phorbol 12-myristate 13-acetate

RANTES

regulated upon activation normal T cell expressed and secreted

SCF

stem cell factor

TGF-ß1

transforming growth factor-beta 1

Asthma is a very frequent disease characterized by bronchial hyperresponsiveness, chronic bronchial inflammation, and remodeling (1). Mast cells have been shown to specifically infiltrate the asthmatic bronchial smooth muscle (BSM) (2–8). Such an infiltration involves the direct chemotactic activity of BSM cells for mast cells through the production of transforming growth factor-beta 1 (TGF-β1), stem cell factor (SCF) (4), CXCL10 (5), and CX3CL1 (6). Collectively, these findings suggest the presence of an auto-activation loop involving mast cell and BSM cell in asthma (4). The adhesion between both cell types may contribute to long-term effects leading to hyperresponsiveness and airway remodeling but its mechanism(s) remain poorly understood.

A direct cell–cell contact between mast cell and BSM cell has been initially suggested as an enhancing factor of mast cell degranulation, although evidence of a real adhesion between these cell types was not provided (9). Mast cells can really adhere to BSM cells, as demonstrated later by Yang et al. (10), implicating tumor suppressor in lung cancer-1 (TSLC-1) but for less than 30% of the adhesion (10). Alternative mechanisms may thus be implicated in mast cell adhesion of the BSM. Such an adhesion can promote the survival and the proliferation of mast cells (11), leading to a prolonged BSM activation by mast cell mediators as found in asthmatic patients.

In this respect, we focused our attention on cell interaction with the extracellular matrix (ECM) through the CD51/CD61 or CD44 molecules for the following reasons. The heterodimer CD51/CD61 or αV/β3 also called vitronectin receptor belongs to the integrin family and has been associated with asthma phenotypes (12). CD51 is expressed in inflammatory cells including human mast cells (13, 14). In association with CD61, CD51 binds many ECM proteins such as vitronectin, fibronectin (FN), or collagen (15). The adhesion molecule CD44 has also been associated with asthma (16). Moreover, in a murine model of asthma, Katoh et al. have inhibited both bronchial hyperresponsiveness and lung inflammation using a blocking anti-CD44 Ab (17, 18). Finally, CD44 is a class I transmembrane glycoprotein originally described in leukocytes and expressed in most cell types including BSM cells and mast cells (19, 20). For instance, adhesion of T lymphocytes to the BSM implicates CD44 (21). Various ECM compounds including hyaluronic acid, collagen, laminin, and FN are ligands for CD44 (22). However, several CD44 variant isoforms (CD44v) are produced upon alternative splicing (23). Previous studies have shown that different variants are implicated in inflammatory diseases (v6) or allergic hypersensitivity (v7) (24), but their role in asthma remains to be elucidated, particularly in mast cell–BSM cell interaction.

The aims of the present study were thus (i) to identify the role of ECM in mast cell adhesion to BSM cell and (ii) to examine the role of CD51, CD44 and particularly that of CD44 variant isoforms in this adhesion.

Methods

A complete description of all Methods information is available in the Online Repository.

Study population and human tissue collection

A total of 11 patients with severe persistent asthma and 35 nonasthmatic controls were prospectively recruited from the ‘Centre Hospitalier Universitaire de Bordeaux’ according to the Global Initiative for Asthma guidelines (25) (Table 1). All subjects gave their written informed consent to participate in the study after the nature of the procedure had been fully explained. This study followed recommendations outlined in the Helsinki Declaration and received approval from the local ethics committee. Bronchial specimens from all subjects were obtained by either fiberoptic bronchoscopy or lobectomy, as previously described (4, 6, 26).

Table 1.   Clinical and functional characteristics of subjects
CharacteristicsPatients with asthmaControls
  1. Data are the mean ± the SEM. BMI, body mass index; LABA, long-acting ß-2 agonist; ICS, inhaled corticosteroid; OCS, oral corticosteroid; FEV1, forced expiratory volume in one-second; FVC, forced vital capacity; FEF 25-75, forced expiratory flow between 25 and 75% of FVC.

Number of patients  11  35
Age (yr)52.6 ± 4.363.4 ± 1.8
BMI (kg/m2)28.9 ± 1.124.3 ± 0.8
Smoking history
 Pack years0 ± 040.5 ± 6.0
 Current (number of patients)   0  14
 Former (number of patients)   0  21
 Years since quitting   –13.1 ± 6.0
Treatments
 LABA (number of patients)   7   3
 ICS (number of patients)  11   3
 OCS (number of patients)   5   0
FEV1
 Liters1.7 ± 0.42.5 ± 0.1
 Percentage of predicted value66.1 ± 10.882.5 ± 3.7
FEV1: FVC ratio (percentage of FVC)69.9 ± 6.669.0 ± 2.6
FEF 25-75
 Liters per second1.5 ± 0.42.1 ± 0.3
 Percentage of predicted value44.1 ± 12.262.2 ± 7.9

Cell culture and challenge

BSM cells were dissected as described previously (27). Human lung mast cells (HLMC) were dispersed from macroscopically normal lung and then purified using anti-CD117-coated immunomagnetic beads (Dynal collection kit; Dynal, Oslo, Norway) as previously described (28, 29). BSM cells or the human mast cell line HMC-1 (30) was cultured in culture medium DMEM (Invitrogen, Cergy Pontoise, France) containing 10% fetal calf serum. Isolated HLMC were cultured in the same medium containing 100 ng/ml SCF, 50 ng/ml IL-6, and 10 ng/ml IL-10 (all from R&D Systems, Abingdon Oxon, UK) as described previously (28).

Cell purity was assessed by both immunocytochemistry and flow cytometry (Beckton Dickinson, Pont de Claix, France). Adhesion molecules expression on mast cells or BSM cells was also assessed by flow cytometry. Briefly, cells were fixed with 4% paraformaldehyde (Thermo Fisher Scientific, Illkirch, France) and incubated for 30 min on ice with mouse Abs against human CD51 (Beckman Coulter, Roissy CDG, France), panCD44 (Dako, Trappes, France) or CD44v4, v6 or v9 (20) or isotype control Abs. Cells were analyzed for their fluorescence intensity using FACS Calibur (Beckton Dickinson). Results are expressed as mean fluorescence intensity (MFI) or percentage of positive cells.

BSM cells were stimulated with optimal concentrations of IL-1β, TNF-α, IL-6 (all from R&D Systems), or SLIGKV-NH2 (Sigma-Aldrich, Saint Quentin-Fallavier, France), the agonist peptide of protease activated receptor 2 (PAR-2), for 72 h. Mast cells were challenged with optimal concentrations of TGF-β1 (transforming growth factor-beta 1), SCF (stem cell factor), RANTES (regulated upon activation normal T cell expressed and secreted), fractalkine (all from R&D Systems), PMA (phorbol 12-myristate 13-acetate) or calcium ionophore A23187 (both from Sigma-Aldrich) for 1 h at 37°C.

Mast cell–BSM cell adhesion

HMC-1 or HLMC adhesion to BSM cells was initially assessed using radiolabeled mast cells as adapted from Lazaar et al. (21) using pulsed mast cells with 10 μCi/ml [3H]-thymidine (Amersham Biosciences, Orsay, France).

In another set of experiments, mast cells were allowed to adhere to a monolayer of BSM cells for 1 h at 37°C in 8-well chamber slides or cell culture inserts (VWR international, Strasbourg, France). Mast cells and BSM cells were stained respectively with Texas Red®-conjugated anti-tryptase (Dako) and FITC-conjugated anti-α-smooth muscle actin Abs (Molecular Probes, Eugene, OR, USA). Confocal images were obtained using Fluoview laser scanning microscope (Nikon, Paris, France) and reconstituted in 3D images using Imaris Software (Bitplane, Zürich, Switzerland) as described previously (6). Electronic images from inserts were obtained using Philips 301 electron microscope (Philips Electron Optics, Eindhoven, The Netherlands) (7).

RNA extraction, reverse transcription, and real-time quantitative PCR

Total RNA was extracted from BSM cells and mast cells as described previously (31). After reverse transcription, real-time quantitative PCR was performed on a Rotor-GeneTM 2000 (Corbett Research, Mortlake, Sydney, Australia), using appropriate primers (Sigma-Aldrich) designed to target CD51, CD44 or collagen type I α1. Results were normalized by the geometric mean of three internal housekeeping genes (32).

Immunoblotting

Western blot was performed on cell protein extracts (26) using primary antibody directed against collagen type I α monoclonal Ab (Santa Cruz, Heidelberg, Germany). Blot images were acquired using biocaptMW software (Thermo Fisher Scientific). Band densities were quantified using ImageJ software and normalized to that of ß-actin.

Statistical analysis

Values are presented as the mean ± the SEM. Statistical analysis was performed using ncss 2001 software. Statistical significance was analyzed by anova, Kruskal–Wallis and paired Student’s t-test. P-value <0.05 was taken as the threshold for statistical significance.

Results

Human mast cells adhere to human BSM

We have observed that mast cells adhered to a monolayer of BSM cells (Fig. 1A). This adhesion was significantly higher than that of mast cells to either bovine serum albumine (BSA) or FN. Among a variety of cytokines and chemokines that can stimulate these twp cell types, the concomitant stimulation of mast cells by TGF-β1 and BSM cells by TNF-α was the main factor able to significantly increase this adhesion (Table 2). Analysis of time course and concentration effect indicated that adhesion was maximal following stimulation of BSM cells with 100 ng/ml TNF-α for 72 h and HMC-1 cells with 1 ng/ml TGF-β1 for 1 h. Similar results were obtained using either HLMC or HMC-1 for the adhesion on BSM cells (Fig. 1A). In addition, in the absence of any stimulation, mast cell adhesion was significantly increased using BSM cells from asthmatic patients when compared to that using nonasthmatic BSM cells (Fig. 1A).

Figure 1.

 Mast cell–bronchial smooth muscle (BSM) cell adhesion. (A) [3H]-mast cells adhere to proteins, nonasthmatic or asthmatic BSM cells. Adhesion was performed using HMC-1 (white bars; n = 10) or human lung mast cell (grey bars; n = 6). Mast cells were not stimulated (−) or stimulated (+) with 1 ng/ml transforming growth factor-beta 1 (TGF-β1) for 1 h; BSM cell were not stimulated (−) or stimulated (+) with 100 ng/ml TNF-α for 72 h. Data are means ± SEM. *, P < 0.05. (B and C) Representative confocal images presented as three axis slices demonstrating tryptase-positive mast cells (red) adhesion to α-smooth muscle actin-positive BSM cells (green). Bars represent 20 μm. (D and E) Representative electronic microscopic images of HMC-1 adhesion to nonasthmatic BSM cells. Bars represent 1 μm. Both mast cells and BSM cells were not stimulated (B and D) or stimulated with 1 ng/ml TGF-β1 for 1 h and 100 ng/ml TNF-α for 72 h, respectively (C and E).

Table 2.   Modulation of human mast cell adhesion to BSM cell
Mast cell challenge (1 h)BSM cell challenge (72 h)
NonstimulatedTNF-α (100 ng/ml)IL-1β (100 ng/ml)IL-6 (100 ng/ml)SLIGKV (10−4 M)
  1. PMA, phorbol 12-myristate 13-acetate; RANTES, regulated upon activation normal T cell expressed and secreted; SCF, stem cell factor; TGF-β1, transforming growth factor-beta 1.

  2. Mast cells labelled with [3H] thymidine (10 μCi/ml) adhered for 1 h to culture plates coated with growth-arrested bronchial smooth muscle (BSM) cells. Values are expressed relative to maximal adhesion. Data are means ± SEM of six independent experiments. *P < 0.05 vs nonstimulated conditions, paired Student’s t-test.

Nonstimulated34.7 ± 2.739.1 ± 6.840.5 ± 10.942.6 ± 12.426.6 ± 6.3
TGF-β1 (1 ng/ml)36.8 ± 5.549.1 ± 4.5*35.9 ± 6.034.6 ± 4.322.7 ± 5.1
SCF (30 ng/ml)29.7 ± 9.037.5 ± 8.134.7 ± 11.632.4 ± 9.434.1 ± 12.4
RANTES (100 ng/ml)30.1 ± 3.231.2 ± 4.727.4 ± 3.832.5 ± 6.034.1 ± 6.6
Fractalkine (50 ng/ml)30.3 ± 5.434.1 ± 3.837.7 ± 3.231.1 ± 3.229.3 ± 5.1
A23187 (200 nM)31.4 ± 8.336.4 ± 5.730.6 ± 8.032.4 ± 5.134.7 ± 2.9
PMA (40 nM)17.3 ± 2.933.2 ± 8.821.0 ± 4.020.1 ± 6.722.1 ± 6.0

Confocal microscopy and double staining immunofluorescence confirmed that mast cells adhered to BSM cells (Fig. 1B) and that, upon stimulation, the number of adherent mast cells increased (Fig. 1C). Using electron microscopy performed under identical experimental conditions, a close interaction between mast cells and BSM cells was also visualized (Fig. 1D, E). However, we were unable to observe either direct cell–cell contact or tight junctions, whatever the stimulation state.

Human mast cells adhere to BSM through CD51 and CD44

To investigate the mechanism of mast cell adhesion to BSM, we first evaluated the effect of blocking Abs against intercellular adhesion molecules. Blocking ICAM-1 (CD54), VCAM-1 (CD106), CD40L (CD154), c-kit (CD117), or fractalkine (CX3CL1) had no effect on spontaneous or stimulated mast cell adhesion to BSM cell (Table S1 in the Online Repository).

We, then, investigated the role of mast cell and BSM cell receptors for ECM proteins. Blocking the mast cell alpha-v subunit of vitronectin receptor (CD51) slightly decreased mast cell adhesion to stimulated BSM cells, whereas blocking BSM CD51 had no effect (Fig. 2A). Similar results were obtained using either HMC-1-stimulated or HLMC-stimulated cells. However, in unstimulated cells, blocking CD51 was efficient only in HMC-1. Moreover, blocking BSM or mast cell transmembrane glycoprotein CD44 significantly decreased the adhesion of stimulated cells indicating that CD44 is implicated in both cell types (Fig. 2B). In addition, blocking CD44 simultaneously in both mast cells and BSM cells had a synergistic effect. Again, similar results were obtained with either HMC-1 or HLMC.

Figure 2.

 Roles of CD51 and CD44 in mast cell–bronchial smooth muscle (BSM) cell adhesion. (A and B) Effect of anti-CD51 (A) or anti-CD44 (B) blocking Ab on mast cell–BSM adhesion. [3H]-mast cells adhere to fibronectin or nonasthmatic BSM cells. Adhesion was performed using HMC-1 (white bars; n = 10) or human lung mast cell (grey bars; n = 6). Mast cells were not stimulated or stimulated with 1 ng/ml transforming growth factor-beta 1 (TGF-β1) for 1 h; BSM cells were not stimulated or stimulated with 100 ng/ml TNF-α for 72 h. Data are means ± SEM. *, P < 0.05 vs irrelevant Ab conditions. (C) Relative CD44 or CD51 mRNA expression obtained from RT-PCR on BSM (n = 5) and mast cells (n = 5). Cells are either nonstimulated (white bars) or stimulated (black bars) with 1 ng/ml TGF-β1 for 1 h (mast cells) or 100 ng/ml TNF-α for 72 h (BSM cells). (D) FACS representative histograms were obtained using nonasthmatic BSM cells, asthmatic BSM cells or mast cells. CD51 or CD44 expression was assessed using irrelevant Ab (dotted lines) or specific Ab (full lines). Again cells are either nonstimulated (fine lines) or stimulated using similar experimental conditions (thick lines).

To investigate this further, we quantified CD51 and CD44 mRNA using real-time RT-PCR (Fig. 2C). Although CD51 mRNA levels remained unchanged in both cell types following the appropriate stimulation, we observed a transient elevation of CD44 mRNA level after 6 h in TNF-α-stimulated BSM cells. However, this transcriptional effect was not associated with an increase in CD44 protein expression assessed by flow cytometry (Fig. 2D). Indeed, CD44 was highly expressed at the surface of nonasthmatic BSM cells (MFI: 235.3 ± 58.8; percentage of positive cells: 97.8 ± 0.7; n = 6), asthmatic BSM cells (MFI: 242.2 ± 31.4; % of positive cells: 98.1 ± 0.9; n = 6), and mast cells (MFI: 60.2 ± 6.6; percentage of positive cells: 91.2 ± 5.0; n = 6) under baseline conditions.

Mast cell adhesion to BSM involves collagen type I α1 but not hyaluronic acid

As blocking CD44 specifically inhibited mast cell adhesion enhanced by TGF-β1 and TNF-α, we analyzed the role of CD44 ligands in mast cell adhesion to BSM. A pretreatment of BSM with hyaluronidase was used to evaluate the role of hyaluronic acid, the main ligand of CD44. Surprisingly, mast cell adhesion was not affected by such a pretreatment (data not shown). By contrast, previous incubation of BSM cells monolayer with type I collagenase abolished mast cell adhesion, whereas the same treatment had no effect on mast cell adhesion to FN (Fig. 3A).

Figure 3.

 Collagen type I mediates mast cell–bronchial smooth muscle (BSM) cell adhesion. (A) [3H]-HMC-1 cells adhere to fibronectin or nonasthmatic BSM cells. Mast cells were not stimulated (−) or stimulated (+) with 1 ng/ml transforming growth factor-beta 1 for 1 h; BSM cells were not stimulated (-) or stimulated (+) with 100 ng/ml TNF-α for 72 h. Adhesion was inhibited by 0.1 μg/ml (grey bars) and 1 μg/ml (black bars) collagenase (n = 8). Data are means ± SEM. *, P < 0.05 vs condition without collagenase. (B–E) Collagen type I α1 level on BSM cells was assessed by quantitative RT-PCR (B and D, n = 5) and Western blot (C and E, n = 5). (B and C) Cells were either nonstimulated (white bars) or stimulated with 100 ng/ml TNF-α for 72 h. (D and E) BSM cells from control donors (white bars) or asthmatic patients (black bars).

It is noteworthy that such inhibitory effect was observed in both stimulated and nonstimulated cells. As adhesion was upregulated by TGF-β1 and TNF-α, we investigated whether these cytokines increase the production of collagen type I. TGF-β1 and TNF-α did not stimulate collagen synthesis by mast cells (data not shown) and BSM cells, respectively, at both transcriptional (Fig. 3B) and protein levels (Fig. 3C). Comparison between mRNA and protein collagen type I production by BSM cells obtained from control donors (n = 5) and asthmatic patients (n = 3) revealed no significant difference (Fig. 3D, E).

Role of CD44 variants isoforms in mast cell–BSM cell adhesion

As, on the one hand, adhesion of mast cells to BSM cells was increased using asthmatic BSM cells compared to nonasthmatic cells or using stimulated cells compared to nonstimulated cells, and, on the other hand, neither collagen production nor CD44 expression was increased under these experimental conditions, we investigated the expression of several CD44 variant isoforms. Using flow cytometry, we analyzed the expression of CD44v4, CD44v6, and CD44v9, which are all expressed in lung tissue (16). Stimulation of BSM cells by TNF-α increased the expression of CD44v6 (MFI: 28.1 ± 1.7 vs 12.3 ± 2.5, P = 0.02; Fig. 4A), decreased that of CD44v4 (13.7 ± 2.6 vs 24.6 ± 1.9, P = 0.03; Fig. 4B), and left unchanged that of CD44v9 (14.2 ± 1.5 vs 14.7 ± 2.3, P = 0.3; Fig. 4C). By contrast, stimulation of asthmatic BSM cells by TNF-α did not alter the expression of these variant isoforms (Fig. 4D–F). However, CD44v6 was constitutively upregulated in asthmatic BSM cells (MFI: 30.0 ± 2.4 vs 12.3 ± 2.5, P = 0.02; Fig. 4A, D). Stimulation of mast cells by TGF-β1 also increased the expression of CD44v6 (MFI: 43.7 ± 4.1 vs 22.6 ± 3.2, P = 0.001; Fig. 4G) but not that of isoforms v4 and v9 (12.4 ± 3.8 vs 14.2 ± 4.4, P = 0.3 and 31.0 ± 2.9 vs 30.5 ± 4.1, P = 0.5, respectively; Fig. 4H, I). Taken together, these results demonstrated that CD44v6 is upregulated in inflammatory experimental conditions in both mast cells and BSM cells, particularly in that of asthmatics. Such findings suggest that CD44v6 may be an important factor for mast cell adhesion to asthmatic BSM.

Figure 4.

 CD44v6 is specifically upregulated in bronchial smooth muscle (BSM) cells and mast cells. FACS representative histograms were obtained using nonasthmatic BSM cells (A, B, C) asthmatic BSM cells (D, E, F) or mast cells (G, H, I). CD44v6 (A, D, G), CD44v4 (B, E, H) or CD44v9 (C, F, I) expression was assessed using irrelevant Ab (dotted lines) or specific Abs (full lines). Cells are either nonstimulated (fine lines) or stimulated (thick lines) with 1 ng/ml transforming growth factor-beta 1 for 1 h (mast cells) or 100 ng/ml TNF-α for 72 h (BSM cells).

Discussion

In the present study, we have demonstrated that mast cells adhere to the BSM mainly via type I collagen of the ECM. This adhesion is reinforced under inflammatory conditions in normal BSM cells, whereas it is maximal in asthmatic BSM cells, and involves both CD51 and CD44. The increased adhesion of mast cell to asthmatic BSM cells was associated with an increased expression of CD44v6. To the best of our knowledge, this is the first description of the molecular mechanism implicated in such adhesion. This mechanism may account for the very specific mast cell infiltration in the smooth muscle layer of bronchi in asthma (2, 4).

Whereas it has been reported that mast cell adhere to BSM (9–11), the role of ECM has not been investigated so far. In the present study, we demonstrate, for the first time, that the main mechanism of mast cell adhesion to BSM involves ECM proteins which are present at the periphery of myocytes. This cell/matrix/cell interaction may generate a co-stimulatory signal for mast cell activation (33) that, in turn, leads to inflammatory mediators release playing a role in smooth muscle contraction and proliferation (34, 35). Such an interaction may implicate different components of the ECM, although type I collagen appears to be the main ECM protein involved. Indeed, adhesion of mast cells to BSM was decreased by type I collagenase to a low level compared to that of spontaneous nonspecific adhesion to BSA.

In addition, we were able to increase the mast cell adhesion to BSM under inflammatory experimental conditions mimicking asthmatic bronchial inflammation. For instance, TNF-α also increased BSM-induced mast cell chemotaxis (4). Interestingly, mast cell adhesion was also increased using asthmatic BSM cells. Indeed, asthmatic BSM cells retain several pathophysiological characteristics in vitro, such as increased proliferation (26, 36, 37) or increased mast cell chemotaxis (5).

We paid special attention to CD51 and CD44, because both receptors recognize type I collagen (38, 39). Blocking either CD51 or CD44 significantly decreased mast cell adhesion to BSM, thus indicating that both receptors were involved. However, CD44 appears to play a key role in mast cells infiltration of BSM. In this context, it should be kept in mind that CD44 also mediates adhesion of lymphocytes to BSM (21). Indeed, using human T lymphocytes and BSM cells in vitro, Lazaar et al. (21) demonstrated that CD44 is necessary to VLA-4- and LFA-1-mediated cell–cell adhesion. By contrast, we found in the present study that human mast cells adhere to BSM cells mainly through CD44 and pericellular matrix components.

In lung inflammation, however, the role of CD44 remains controversial (40). On the one hand, several studies have demonstrated that CD44 is a proinflammatory agent in a variety of murine models such as eosinophilic and T helper type 2 lung inflammation (17, 18). On the other hand, CD44−/− mice develop exaggerated lung inflammatory responses including increased leukocyte infiltration and chemokine expression (41). However, there is no data regarding mast cell localization within the BSM in CD44−/− mice (41). One possible explanation for such a discrepancy may be the association of different immunological phenotypes represented by CD44 variant isoforms which are predominantly expressed on activated leukocytes (22). In both mice and humans, CD44 gene can produce higher molecular mass CD44 isoforms by alternative splicing. These isoforms contain additional peptide sequences within the extracellular domain of CD44. Several studies focused on specific functions for some of these CD44 variant isoforms by using either blocking antibodies (24, 42) or generating specific knock-out mice (43). For example, early thymocyte development requires the expression of CD44 variant isoforms (42), and blocking of variants 6 and 7 prevents delayed-type hypersensitivity (24). Disruption of CD44v6v7 also prevents experimental colitis by increasing cell death in the inflammatory lesions (43), whereas CD44v6 can interfere with CD95 trimerization and hence block apoptosis signaling (44). In the present study performed in humans, we observed an increased expression of CD44v6 in both BSM cells from asthmatic patients when compared with healthy controls and mast cells after TGF-ß1 stimulation. Further studies should analyze the in vivo functions of CD44v isoforms in murine models of experimental asthma.

In conclusion, we have demonstrated the key role of CD51 and CD44v6 in mast cell adhesion to BSM in asthma via the extracellular matrix. This mechanism may account for the very specific mast cell infiltration observed in asthma (2, 4). As it is believed that such infiltration contribute to long-term effect leading to bronchial hyperreactivity and remodeling, CD44v6 may thus represent a potentially attractive new target for asthma therapy.

We thank the staff of ‘Service de Chirurgie Thoracique – CHU de Bordeaux’ for the supply of human lung tissue. We thank Ursula Gunthert from Basel Switzerland for the supply of various antibodies against CD44 variants. We also thank Nadia El Bachachi and Vincent Pitard for their assistance and Derrick Robinson for his expertise in electron microscopy.

Funding

This study was supported by a grant from the Fondation pour la Recherche Médicale, France; the Société Française d’Allergologie et d’Immunologie Clinique, France and the Agence Nationale de la Recherche, France. The authors have no relationships to declare.

Ancillary