Heterogeneity of Expression of IgA Receptors by Human, Mouse, and Rat Eosinophils
http://www.100md.com
免疫学杂志 2005年第2期
Abstract
IgA is the most abundant class of Abs at mucosal surfaces where eosinophils carry out many of their effector functions. Most of the known IgA-mediated functions require interactions with IgA receptors, six of which have been identified in humans. These include the IgA FcR FcRI/CD89 and the receptor for the secretory component, already identified on human eosinophils, the polymeric IgR, the Fc/μR, asialoglycoprotein (ASGP)-R, and transferrin (Tf)R/CD71. In rodents, the existence of IgA receptors on mouse and rat eosinophils remains unclear. We have compared the expression and function of IgA receptors by human, rat, and mouse eosinophils. Our results show that human eosinophils express functional polymeric IgR, ASGP-R, and TfR, in addition to CD89 and the receptor for the secretory component, and that IgA receptors are expressed by rodent eosinophils. Indeed, mouse eosinophils expressed only TfR, whereas rat eosinophils expressed ASGP-R and CD89 mRNA. These results provide a molecular basis for the differences observed between human, rat, and mouse regarding IgA-mediated immunity.
Introduction
Eosinophils are known to play a role in protection against helminth parasites as well as in the physiopathology of allergic diseases, such as bronchial asthma and allergic rhinitis. These functions are linked to the eosinophil ability to release biologically active mediators upon activation through IgRs (1), complement (2), or adhesion molecules.
IgA is the predominant Ab isotype in humans and is present in the circulation as well as at the mucosal surfaces. Secretory (S)3 IgA and eosinophils are two main components of the mucosal immunity (3). Most of IgA-mediated functions require an interaction with IgA receptors. Recently, many reports have demonstrated the existence of different IgA-binding structures, which differ between species regarding their cellular distribution and function (4).
FcRI/CD89 is the receptor for the Fc portion of IgA, and this receptor is the most widely distributed in humans. Indeed, not only eosinophils (5, 6) but also neutrophils (7), monocytes (8), macrophages subsets, Kupffer cells, and dendritic cells (9) express CD89 at their surface. In rats, a CD89 homolog has recently been identified (10). By contrast, CD89 is not present in the mouse genome (11). The polymeric (p)IgR recognizes the J chain joining polymeric IgA and IgM and is responsible for transporting them through mucosal epithelial cells. A specific protease then cleaves the extracellular portion of pIgR, also known as secretory component (SC), and releases SIgA and IgM into mucosal secretions (12). In rodents, pIgR is expressed in the liver where it is involved in IgA transport into the bile (13). Fc/μR was recently characterized on human and mouse B cells and macrophages and on different tissues like liver, spleen, and intestine (14). It binds monomeric IgA and IgM and is involved in endocytosis of IgA/M-opsonized bacteria. TfR/CD71 was described as an IgA1-binding molecule involved in human renal IgA deposition on mesangial cells (15). Recently, two isoforms of this receptor were identified (16). TfR1 is expressed on a wide range of tissues, whereas TfR2 was found only in liver and on erythroid cells. Asialoglycoprotein (ASGP)-R binds IgA via its carbohydrate moieties and is mainly expressed on human and rat liver cells (17, 18) where it plays a crucial role in IgA metabolism. It has also been found at the surface of immature dendritic cells (19). The SCR is the receptor for the SIgA and is expressed on human eosinophils (20) and basophils (21).
In mouse, previous studies have described the physiological role of IgA (22), J chain (23), and pIgR (24). Early studies have also shown functional interactions between IgA and rat eosinophils in Ab-dependent cellular cytotoxicity reactions toward parasite larvae (25) or in mucosal immunity (26), thus suggesting that rat eosinophils could express functional receptors for IgA. However, the molecular characterization of IgA receptors expressed by mouse and rat eosinophils remains unknown.
Due to this lack of data about rodent, which are widely used as experimental models for human parasitic and allergic diseases, and due to major differences among IgRs already identified between humans and the two rodent species (1), we have undertaken a comparative analysis of IgA receptors on human, rat, and mouse eosinophils.
Materials and Methods
Eosinophil donors
Normal donors and hypereosinophilic patients with drug hypersensitivity were selected for this study, after informed consent.
Animals
Animals were housed in a specific pathogen-free facility from the Institut Pasteur de Lille, with continuous access to food and water. Eight- to 10-wk-old Brown Norway rats and 6- to 12-wk-old BALB/c mice or IL-5 transgenic mice expressing IL-5 under the control of the human CD2 promoter (27) were used.
Reagents
Anti-human CD16-, streptavidin-coated magnetic beads, and the MACS system were from Miltenyi Biotec. Percoll was from Amersham Biosciences. RPMI 1640, glutamine, penicillin, streptomycin, FCS, and ultroser were obtained from Invitrogen Life Technologies. Paraformaldehyde, saponin, HEPES, human SIgA, human IgM, human IgG, biotinylated human transferrin (Tf), biotinylated anti-rabbit IgG, hydrogen peroxide, and luminol were from Sigma-Aldrich. BSA was from Eurobio. Luciferin was purchased from Roche Diagnostics. Human serum IgA and mouse anti-human SC was obtained from ICN. Biotinylated anti-human IgA, anti-human IgM, anti-mouse CD19, anti-CD8 (Ly-2), anti-CD90 (Thy1.2), and mouse anti-rat IgA; rat anti-mouse IgA; purified mouse IgA; purified rat IgA; and purified rat IgM were from BD Pharmingen. For stimulation experiments, purified mouse anti-human IgM (Serotec), rat anti-DNP IgA (Technopharm), and DNP-keyhole limpet hemocyanin (KLH) (Calbiochem) were used. PE-conjugated streptavidin (SA-PE) was obtained from Molecular Probes. FITC-conjugated goat anti-rat, donkey anti-rabbit IgG (H+L) F(ab')2 and purified rat IgG were obtained from Jackson ImmunoResearch Laboratories. Goat anti-mouse IgG1 was purchased from Southern Biotechnology. Human SC purified from colostrum was a kind gift from Dr. A. Pierce (University Lille I, Lille, France). Rabbit anti-human and anti-rat ASGP-R Ab were kindly provided by Dr. D. Alpers (Washington University School of Medicine, St. Louis, MO) and Dr. R. Stockert (Albert Einstein College of Medicine, Bronx, NY).
Cell purification
Human eosinophils were isolated from the venous blood of patients and healthy donors, using immunomagnetic beads and the MACS system as previously described (28). Diluted whole blood (1/1) was layered onto a Percoll gradient (density, 1.082 g/L) and centrifuged at 1800 rpm for 20 min. After centrifugation, PBMC were collected at the interface and the granulocytes in the pellet. After hypotonic saline lysis, granulocytes were incubated with anti-CD16 beads to remove neutrophils. Eosinophils were eluted following passage through the field of a permanent magnet. Eosinophil purity, assessed on cytospin preparation following staining with May-Grünwald Giemsa (RAL 555), was >98%.
Mouse splenocytes were obtained by gentle dissociation of spleens in ice-cold PBS. Aggregates were removed by passage through a nylon filter and erythrocytes were lysed using hypotonic saline. Splenic eosinophils were purified from IL-5 transgenic mice by negative selection as previously described (29). Briefly, cells were incubated 30 min on ice with biotinylated anti-mouse CD19, anti-CD8 (Ly-2), and anti-CD90 (Thy1.2). After washing, cells were incubated 15 min with streptavidin-coated beads at 4°C. Purified eosinophils were obtained following passage through the field of a permanent magnet. Purity was >90%.
Splenic B lymphocytes were purified from BALB/c mice by positive selection. Briefly, cells were incubated 30 min with anti-mouse CD19, washed twice, and incubated with streptavidin-coated beads at 4°C. Lymphocytes were recovered from the column after detachment of the cells out of the magnetic field. Purity was >95%.
Mouse peritoneal cells were obtained by flushing the peritoneal cavity of BALB/c mice with 10 ml of PBS. After filtration on a nylon filter, cells were cultured overnight in complete medium, and adherent macrophages were recovered by scraping. Macrophage purity was >95%.
Rat peritoneal cells were obtained by flushing the peritoneal cavity with 50 ml of ice-cold PBS. Aggregates were removed from cell suspensions by filtration on a nylon filter, and erythrocytes were lysed using hypotonic saline. For stimulation experiments, purified rat eosinophils were obtained by sorting peritoneal cells according to their typical forward and side scatter (30), using an ELITE cell sorter (Coulter). Eosinophil purity after sorting was ranging between 90 and 98%. Contaminating cells were macrophages.
Rat splenocytes were obtained by gentle dissociation of spleens in ice-cold PBS. Aggregates were removed by nylon filtration, and erythrocytes were lysed using hypotonic saline.
Cell culture
For TfR up-regulation experiments, purified human eosinophils were cultured for 1 day in RPMI 1640 supplemented with 10% heat-inactivated FCS, 25 mM HEPES, 2 mM L-glutamine, 100 IU/ml penicillin, and 100 μg/ml streptomycin (complete medium) in the absence or in the presence of 25 μg/ml Tf.
After cell sorting, rat peritoneal eosinophils were kept for 18 h in phenol red-free RPMI 1640 supplemented with 2% Ultroser before stimulation.
The various cell lines used as control, HT29, CaCo2, and HEPG2 were cultured in complete medium.
RT-PCR
Total RNA was extracted from 10 x 106 highly purified (98%) eosinophils, PBMC, cell lines, mouse lymphocytes, mouse macrophages, or rat spleen cells using RNAplus extraction reagent (Q Biogene). Mouse and rat liver mRNA were extracted by CsCl centrifugation. Reverse transcription was performed using SuperScript RT (Invitrogen Life Technologies). Amplification by PCR using Taq polymerase (Bioprobe) was performed with the primers (Proligo) and under the conditions described in Table I. Amplified products were then normalized according to the signal intensity provided by the -actin amplification, loaded on a 1% agarose gel, stained with ethidium bromide, and photographed under UV light.
Table I. Sequences of primers used for PCR amplification of cDNA, amplicon sizes, annealing temperatures, and PCR cycle numbersa
Flow cytometry
Unless specified otherwise, all incubations were performed at 4°C for 30 min with cells resuspended at 4 x 106/ml in PBS containing 1% BSA and 0.05% sodium azide (FACS buffer) with 10 μg/ml Ab, and cells were washed twice with FACS buffer between each step of staining.
Fifty microliters of the cell suspensions (2 x 105 cells) was incubated with 100 μg/ml human IgM or with rat or mouse IgA for 45 min in round-bottom 96-well plates. After washing, the cells were then incubated with biotinylated anti-human IgM for human cells, biotinylated mouse anti-rat IgA for rat cells, and purified rat anti-mouse IgA for mouse cells. Biotinylated or purified anti-Ig were incubated followed by an incubation with SA-PE or with FITC goat anti-rat Igs.
For human pIgR, competition experiments were performed between IgM and the other isotypes, and cells were thus preincubated with 500 μg/ml SIgA, serum (monomeric) IgA or IgG before incubation with human IgM.
For the competition assay between SC and IgM, a concentration of 500 μg/ml SC was chosen, as it was previously determined by flow cytometry to saturate cell surface receptors. SC was incubated 45 min and its binding was revealed by a purified mouse anti-SC and a PE-conjugated goat anti-mouse IgG.
Rabbit Abs (anti-human pIgR and anti-human and rat ASGP-R) and normal rabbit serum were used at 1:400, and their binding was revealed by a donkey FITC-conjugated anti-rabbit IgG. Competition with the anti-ASGP-R Ab was performed by incubating the cells with 400 μg/ml rat IgA, human serum IgA, human SIgA, IgG, or IgM, a concentration that saturates surface receptors, before adding the anti-ASGP-R Ab.
Biotinylated human Tf (100 μg/ml) was incubated with mouse and human eosinophils, and binding was detected after incubation with SA-PE for 20 min. For the competition assay, a concentration of 500 μg/ml mouse, human serum IgA, SIgA, IgM, or IgG was chosen because it was previously determined to saturate the receptor. Competitions were performed by preincubation of the various Ig 45 min before adding Tf.
For intracellular detection of Tf binding, cells were fixed with 2% paraformaldehyde in PBS for 10 min. After washing in PBS, cells were resuspended at 4 x 106/ml in PBS containing 1% BSA and 0.1% saponin for 10 min at room temperature. The cells were then incubated 45 min with biotinylated human Tf followed by an incubation with SA-PE. Samples were washed twice in PBS-saponin buffer, once in PBS, and resuspended in PBS-0.5% BSA for analysis.
Cells were analyzed by flow cytometry using a FACSCalibur equipped with the CellQuest software (BD Biosciences). A total of 104 events was acquired for each sample.
IgA-anti-IgA immune complexes (IC)
Mouse or rat IgA-anti-IgA IC were prepared by incubating concentrations ranging from 2 to 18 μg/ml IgA with 18 to 2 μg/ml anti-IgA for 30 min at room temperature to obtain an IgA-anti-IgA ratio ranging from 1:9 to 9:1. IC were then incubated with cells for 30 min at 4°C. For rat eosinophils, competition of IgA IC binding to eosinophils was performed using 100 μg/ml rat IgM or IgG.
Reactive oxygen species (ROS) production
ROS production by highly purified eosinophils was examined using luminol-dependent chemiluminescence as previously described (31). For IgM activation, 50 μl of purified eosinophils (5 x 105 cells) was incubated alone or with IgM (15 μg/ml) for 1 h at 37°C. The inhibition of IgM-induced ROS release was performed by preincubating the cells with 500 μg/ml SC to saturate the SCR, and IgM was incubated with 15 μg/ml SC at 37°C to block its J chain for 1 h. The complex was then incubated for 1 h with the cells, followed, after washing, by the addition of anti-IgM (20 μg/ml) and luminol (25 μg/ml). Kinetics was performed at 37°C over a 40-min period.
For Tf activation, 96-well flat-bottom tissue culture plates (Falcon) were coated with 100 μg/ml human Tf in PBS at 37°C for 2 h. Coated wells were washed twice with PBS before use, and cells (5 x 105 cells) were added to the wells. For IgA and Tf competition assay, cells were incubated for 1 h at 37°C with human serum IgA and washed to remove the IgA excess before the Tf stimulation. Chemiluminescence was measured for 5 s with a luminometer (Victor2 Wallac; PerkinElmer). Kinetics was performed at 37°C over a 40-min period. Results were expressed in counts per second.
Measurement of eosinophil neurotoxin (EDN) and eosinophil peroxidase (EPO) for assessment of degranulation
For EDN measurement, highly purified human eosinophils (2 x 106/ml) were first incubated in culture medium either alone or with human IgM (15 μg/ml). After 1 h, cells were stimulated with 20 μg/ml purified mouse anti-human IgM at 37°C.
For Tf stimulation, wells were coated with 100 μg/ml human Tf for 2 h. After washing twice in PBS, cells were added to the wells at the concentration of 2 x 106/ml. IgA and Tf competition was performed as described for the chemiluminescence assay.
After 18 h, supernatants were collected and stored at –20°C until measurement of EDN concentration using a specific ELISA (MBL International) according to the manufacturer’s instructions. Lower detection limit of the assay was 0.62 ng/ml.
For EPO measurement, rat peritoneal eosinophils (2 x 105) were incubated either alone or with 10 μg/ml rat anti-DNP IgA for 1 h. IgA was then cross-linked by incubation of the cells with 20 μg/ml DNP-KLH for 2 h. After stimulation, cells were centrifuged, and 50 μl of supernatant was collected. Fifty microliters of luciferin (6 x 10–5 M), hydrogen peroxide (0.15%), and luminol (200 μg/ml) were added to the supernatants before the EPO measurement with a luminometer (Victor2 Wallac).
Statistical analysis
Statistical analysis was performed using Student’s paired t test. Statistical significance was set at p < 0.05.
Results
Human eosinophils express pIgR, ASGP-R, and TfR
Because CD89 and SCR expression has been described on human eosinophils, we determined whether other IgA-binding structures might also be expressed by these cells. RT-PCR was performed on RNA extracted from highly purified eosinophils from a healthy donor and a patient with drug hypersensitivity. An amplicon was detected for pIgR in both eosinophil preparations (Fig. 1A), as well as in neutrophils and PBMC (data not shown). By contrast, no Fc/μR transcripts were detected on eosinophils, whereas a signal was detected on PBMC (Fig. 1B). The expression of pIgR was confirmed by flow cytometry using an anti-pIgR-specific Ab (Fig. 1C) or IgM, one of the pIgR ligand (D). J chain-containing IgM binding was not inhibited by an excess of SIgA (with its J chain blocked by SC) or serum (monomeric) IgA or IgG (Fig. 1D, inset), confirming that this receptor actually binds polymeric Ig through its free J chain as previously reported (12).
FIGURE 1. Expression of pIgR by human eosinophils. A, RT-PCR analysis of pIgR mRNA expression by human eosinophils. RNA extracted from highly purified eosinophils (ND, normal donor; DH, drug hypersensitivity) was subjected to reverse transcription and amplified by PCR using primers specific for pIgR and -actin. Total mRNA from HT29 was used as positive control. B, RT-PCR analysis of Fc/μR mRNA expression in human PBMC and eosinophils (ND; DH). C, Flow cytometric analysis of pIgR expression by human eosinophils. Eosinophils were incubated with rabbit anti-pIgR Ab (black line) followed by a FITC-conjugated anti-rabbit IgG (dashed line). Normal rabbit serum was used as negative control (filled histogram). D, IgM binding on human eosinophils. Eosinophils (ND) were incubated with IgM (black line). Biotinylated anti-human IgM was used as control (filled histogram). Inset, Competition of IgM binding by preincubation with a 5-fold excess of serum IgA (IgA), SIgA, and IgG. E, Representative time course experiment of IgM-induced ROS release by human eosinophils (ND). Human eosinophils were incubated for 1 h with () or without IgM followed by the addition, after washing, of anti-IgM (). For inhibition of IgM binding, cell surface was saturated by human SC and then incubated with IgM/SC complexes (?). Results are expressed in counts per second (CPS). F, IgM-induced EDN release by eosinophils. Eosinophils (ND) were incubated or not with IgM for 1 h. Anti-human IgM was then added to cross-link receptors, and supernatants were collected after 18-h stimulation (n = 4). MFI, Mean fluorescence intensity.
Furthermore, activation with IgM followed by cross-linking with anti-IgM Ab induced ROS production (Fig. 1E). To confirm that eosinophil activation was mediated by pIgR, an inhibition assay between SC and IgM was performed. ROS release was inhibited when IgM J chain binding to pIgR was prevented by preincubation of both IgM and eosinophils with SC. Finally, IgM cross-linked with anti-IgM was able to induce EDN release by eosinophils (Fig. 1F). These results thus demonstrate that human eosinophils could express a functional pIgR.
ASGP-R expression was first investigated by RT-PCR. Primers were chosen in a highly conserved region sharing 99% homology with mouse and rat sequences, thus allowing the use of rat liver as positive control. Human eosinophils from a normal donor and a patient presenting a drug hypersensitivity expressed ASGP-R mRNA (Fig. 2A). Surface expression of the receptor was then confirmed by flow cytometry using an anti-ASGP-R Ab (Fig. 2B). To confirm that eosinophils ASGP-R also bind IgA, a competition between serum IgA, secretory IgA, IgM, or IgG, and the anti-ASGP-R Ab was performed. Ab binding inhibition was ranging between 17.5 and 47.5% upon preincubation with SIgA (dimeric) but not with monomeric serum IgA, IgM, or IgG, suggesting that ASGP-R expressed by human eosinophils might act as an IgA receptor (Fig. 2B, inset). ASGP-R-mediated eosinophil activation was then confirmed by the release of ROS following stimulation with the anti-ASGP-R Ab (Fig. 2C).
FIGURE 2. ASGP-R expression by human eosinophils. A, RT-PCR analysis of ASGP-R mRNA expression by human eosinophils (ND, normal donor; DH, drug hypersensitivity). Primers were chosen in a region showing a high degree of homology between rat and human. Rat liver was used as positive control. B, Flow cytometric analysis of ASGP-R expression by eosinophils (ND). Binding of anti-ASGP-R Ab (black line) was detected using an FITC-conjugated anti-rabbit IgG. Normal rabbit serum (filled histogram) was used as control. Inhibition of anti-ASGP-R binding was obtained by cell preincubation with 400 μg/ml SIgA (dotted line). Inset, Competition of anti-ASGP-R binding by preincubation with 400 μg/ml IgM, IgG, and (serum) IgA. C, Representative time course experiment showing ROS release by normal donor eosinophils. Cells were incubated for 1 h at 37°C with a rabbit anti-ASGP-R Ab (). Measurement was initiated upon addition of the anti-rabbit Ig. Normal rabbit serum was used as control () (n = 3). MFI, Mean fluorescence intensity.
Expression of TfR mRNA by human eosinophils was first assessed by RT-PCR. Fig. 3A shows that TfR1 but not TfR2 was detected. Flow cytometric analysis of TfR expression showed a weak expression at the surface of freshly isolated eosinophils (Fig. 3B). However, staining of permeabilized cells revealed the presence of an intracellular pool of receptors, which might be available for subsequent mobilization (Fig. 3B, inset). Indeed, following overnight culture in the presence of Tf, free and occupied TfR were detected at the cell surface (Fig. 3C). To demonstrate that TfR acts as an IgA receptor on eosinophils, we thus tested whether preincubation with IgA was able to inhibit Tf binding. An excess of serum IgA was inhibiting Tf binding to its receptor, whereas no inhibition was obtained with an excess of IgM or IgG. Interestingly, an excess of SIgA was not able to inhibit Tf binding to its receptor (Fig. 3D, inset).
FIGURE 3. TfR expression by human eosinophils. A, TfR1 and -2 mRNA expression by eosinophils revealed by PCR. Eosinophils (ND, normal donor; DH, drug hypersensitivity), HT29, and CaCo2 were positive for TfR1 expression. HEPG2 was used as positive control for TfR2 expression. B, Flow cytometric analysis of the Tf binding to normal donor eosinophils. Freshly purified eosinophils were incubated with biotinylated Tf (black line) followed by an incubation with SA-PE (negative control, filled histogram). Inset, Intracellular detection of Tf in human eosinophils. C, Tf binding on eosinophils (ND) cultured for 1 day with 25 μg/ml biotinylated Tf. Cells were incubated (total receptors, black line) or not (occupied receptors, dotted line) with 100 μg/ml Tf. SA-PE, in the absence of Tf (filled histogram) was used as negative control. D, Competition of Tf binding by (serum) IgA. Inset, Competition by preincubation with 500 μg/ml IgM, IgG, and SIgA. E, Representative time course experiment of ROS release following stimulation of eosinophils with immobilized Tf. Eosinophils (ND) were incubated in tissue culture wells coated () or not (?) for 2 h with 100 μg/ml Tf. To inhibit Tf stimulation, eosinophils were incubated with human serum IgA, washed twice with PBS before incubation in Tf-coated wells () (n = 5). F, EDN release by eosinophils (ND) following Tf stimulation. Eosinophils were stimulated for 18 h in Tf-coated wells. Inhibition of stimulation was obtained by preincubating the cells with serum IgA as described above (n = 5). MFI, Mean fluorescence intensity.
Furthermore, immobilized Tf was able to stimulate ROS production from freshly isolated cells, suggesting that TfR surface expression levels were below the detection threshold of our cytometric method. As might have been expected from flow cytometry data, serum IgA was also able to efficiently inhibit Tf-induced ROS release (inhibition range, 59–70%), thus indicating that TfR expressed on eosinophils was also a functional receptor for serum IgA (Fig. 3E). This result was confirmed by a significant inhibition (94.5%) of the Tf-induced EDN release by serum IgA (Fig. 3F).
IgA receptors expressed on mouse eosinophils
Because CD89 is not found in mice, we examined which molecular structure would bind IgA on mouse eosinophils. Because naive wild-type mice harbor only very few eosinophils, we used purified splenic eosinophils from IL-5 transgenic mice.
RT-PCR was performed on freshly purified eosinophils to detect the presence of specific amplicons for pIgR, Fc/μR, ASGP-R, as well as TfR1 and -2. According to the receptor investigated, RNA from tissues or cells known to express the receptor of interest were chosen as positive controls. Amplicons corresponding to Fc/μR, TfR1, and -2 were detected in eosinophil samples. By contrast, no bands corresponding to pIgR or ASGP-R were amplified, whereas the corresponding positive controls displayed amplification of the fragments with the expected size (Fig. 4A).
FIGURE 4. IgA receptors expressed by mouse eosinophils. A, RT-PCR analysis of Fc/μR, pIgR, ASGP-R, and TfR expression by mouse eosinophils. Mouse liver was used as positive control and mouse -actin was used as housekeeping gene. B, Monomeric IgA (black line) was detected using rat anti-mouse IgA and FITC-conjugated goat anti-rat F(ab')2 (filled histogram). C, Binding curve of IgA-anti-IgA IC on mouse eosinophils. D, Tf binding on mouse eosinophils and inhibition by mouse IgA (n = 3). Results are expressed as mean fluorescence intensity (MFI) after subtraction of the control.
Expression of IgA receptors by mouse eosinophils was next confirmed by flow cytometry. A weak binding of monomeric IgA was detected on freshly isolated eosinophils (Fig. 4B). Because the expression of a low affinity receptor could account for this weak binding, IgA-anti-IgA IC were then used, and significant increase in binding to eosinophils, compared with monomeric IgA, was observed (Fig. 4C). Maximal binding was achieved with a 1:1 ratio between IgA and anti-IgA. A very weak IgM binding to the same eosinophil preparation was also observed, but no competition between IgA and IgM was detected, suggesting the absence not only of pIgR but also of Fc/μR at the surface of mouse eosinophils (data not shown).
Because the two isoforms of TfR mRNA were found in eosinophils, we investigated whether this receptor was expressed at their surface. Human Tf was used because the Tf binding site is highly conserved between species (32), and binding to cell surface was detected. To assess that TfR also binds IgA, competition experiments were performed. For this assay, a high dose of mouse IgA was used to induce IgA aggregation, because aggregated IgA binds more efficiently to the receptor than monomeric IgA. An inhibition of Tf binding ranging between 17 and 58% was observed in the presence of IgA (Fig. 4D), demonstrating that TfR is an IgA receptor in mouse eosinophils.
IgA receptors expressed by rat eosinophils
To determine which receptor is involved in IgA binding to rat eosinophils, we analyzed the expression of potential IgA-binding structures (CD89, pIgR, Fc/μR, ASGP-R, and TfR) by peritoneal eosinophils by RT-PCR.
No expression of pIgR, Fc/μR, TfR1, or -2 was detected on rat eosinophils, despite the presence of a strong signal on liver tissue chosen as positive control. By contrast, a specific signal was detected for ASGP-R and for a homolog of human CD89 (Fig. 5A).
FIGURE 5. IgA receptors expressed by rat eosinophils. A, RT-PCR analysis of IgA receptor expression by rat eosinophils. RNA from cell-sorted rat eosinophils, total liver and spleen cells (80% lymphocytes, 20% macrophages) was subjected to reverse transcription and amplified by PCR using primers specific for ASGP-R, TfR1 and -2, pIgR, CD89, and Fc/μR. B, Flow cytometric analysis of monomeric IgA binding on rat eosinophils. Purified rat IgA (black line) was incubated for 45 min at 4°C. Binding was detected using biotinylated anti-rat IgA followed by an incubation with SA-PE (filled histogram). C, Binding curve of rat IgA-anti-IgA IC. D, Flow cytometric analysis of ASGP-R expression. Peritoneal rat eosinophils were incubated with a rabbit anti-human ASGP-R Ab (black line) followed by incubation with FITC-conjugated donkey anti-rabbit IgG. Rabbit serum was used as negative control (filled histogram). Inhibition of the anti-ASGP-R Ab binding was obtained by preincubating with 400 μg/ml rat IgA (dotted line). Inset, Competition by preincubation with 400 μg/ml rat IgG and IgM. E, EPO release by purified peritoneal rat eosinophils. Eosinophils were incubated or not with 10 μg/ml anti-DNP IgA for 1 h at 37°C, followed by the addition of 20 μg/ml DNP-KLH. After 2-h stimulation, EPO was measured in the supernatants (n = 3). MFI, Mean fluorescence intensity.
We next analyzed rat IgA binding on eosinophils. No binding of monomeric IgA was observed on rat eosinophils (Fig. 5B). However, IgA-anti-IgA IC displayed a weak binding (Fig. 5C). No binding inhibition was observed by preincubation with a 10-fold excess of either IgG or IgM (data not shown). To confirm at the protein level, the results obtained by PCR evidencing an amplicon for ASGP-R, a flow cytometric approach was used. ASGP-R was detected using a specific Ab raised against the intracellular portion of the receptor (data not shown). This result was confirmed by using an anti-human ASGP-R Ab, which allowed us to detect the receptor at the cell surface (Fig. 5D). Indeed, rat and human ASGP-R shared 70% homology in their extracellular domains. To assess that ASGP-R expressed on rat eosinophils also binds IgA, a competition was performed using rat IgA, IgM, or IgG. An inhibition of Ab binding ranging between 30 and 35% was observed when cells were incubated with IgA (Fig. 5D) but not with IgM or IgG (D, inset). This result confirmed that ASGP-R can be considered as an IgA receptor on the rat eosinophils. Furthermore, rat eosinophils were activated by IgA, because EPO was released 2 h following stimulation, thus demonstrating the functionality of at least one IgA receptor (Fig. 5E).
Discussion
IgA and eosinophils are two major players in mucosal immunity, but their interaction is not restricted to mucosal tissues and can also occur in blood in the course of parasitic or allergic diseases (33). Several convergent studies have shown that IgA is one of the most potent stimuli for eosinophils (3, 34). In the present work, we investigated whether human eosinophils could express IgA receptors other than CD89 and the SCR, and we determined for the first time which receptors were involved in IgA binding to mouse and rat eosinophils.
Our data indicate that human eosinophils expressed a functional pIgR and demonstrate that this receptor can be found on cells other than secretory epithelial cells. Furthermore, these results suggest that eosinophils themselves might be a potential source of SC, which may act on eosinophils as an autocrine mediator after binding to SCR. Such a mechanism was demonstrated for many other eosinophil mediators, such as IL-5 (35), platelet-activating factor (36), or eotaxin (37, 38). Our data suggest that eosinophils are able to be activated by pIg following lamina propria infiltration and thus play an active role in host defense against pathogens that have crossed the epithelial barrier. Such a result was already demonstrated for polymeric IgA activation (39), but we here demonstrate for the first time that this activation leads to ROS and EDN release following IgM stimulation of eosinophils. Furthermore, we have shown that EDN secretion was observed upon incubation with anti-IgM alone suggesting the presence of IgM, presumably bound to pIgR at the cell surface.
No pIgR expression was detected, even at the mRNA level, in mouse or rat eosinophils, thus demonstrating that receptor expression is restricted to human eosinophils. In rodents, IgA is mainly produced as polymers (4), and the absence of pIgR on circulating cells may prevent IgA-induced, pIgR-mediated activation.
Fc/μR was not detected on human and rat eosinophils, even at the mRNA level, but mouse eosinophils were found weakly positive for the expression of Fc/μR transcripts. However, because no competition was observed between IgA and IgM, the existence of the receptor at the cell surface was not confirmed (data not shown). These results thus support the absence of Fc/μR on Gr-1+ granulocytes (likely neutrophils) observed by Shibuya et al. (14).
Some studies have described the important role of the ASGP-R in IgA metabolism (40). We determined whether this receptor was expressed by cells other than hepatocytes and whether this receptor could play a role in IgA binding to eosinophils. Indeed, ASGP-R was detected on eosinophils, thus extending the demonstration of ASGP-R mRNA in human granulocytes (likely neutrophils) derived from CD34+ precursors in the absence of IL-5 (19). Furthermore, SIgA but not serum IgA was able to compete with the anti-ASGP-R Ab, a result explained by the fact that ASGP-R binds IgA2 and not IgA1, the main isotype of IgA found in serum. The partial inhibition of Ab binding to human eosinophils by SIgA is consistent with the expression of CD89 and SCR, which also bind SIgA.
Only human and rat eosinophils were found positive for ASGP-R expression. ASGP-R mediates the release of ROS by human eosinophils and might also account for EPO release by rat eosinophils, suggesting that the receptor is involved not only in endocytosis of glycoproteins but also in the development of cytotoxic responses. Recent studies have also demonstrated that this receptor can be used by viruses for attachment and entry into cells. In particular, Dotzauer et al. (41) have shown that hepatitis A virus complexed with specific IgA can use IgA-specific receptors for entry into hepatocytes, and some viruses, such as respiratory syncytial virus, have been shown to infect eosinophils (42). Our results suggest that such phenomenon may occur via ASGP-R.
We next studied the possibility that human and rodent eosinophils could express TfR, recently identified as an IgA1 receptor (15). The two specific TfR have already been described on myeloid cells, and Sakamoto et al. (43) have reported the existence of TfR on neutrophils. No TfR was detected on rat eosinophils, whereas human eosinophils were found positive for the TfR1 mRNA expression and mouse eosinophils expressed mRNA for the two isoforms. Our results demonstrate that TfR is expressed by eosinophils and is an IgA receptor, because serum IgA, mainly containing IgA1, inhibits Tf-induced ROS and EDN release. This suggests that IgA may play an anti-inflammatory role in blood by blocking the release of cytotoxic mediators induced by circulating Tf. Some studies have demonstrated that TfR is an entry receptor for some enveloped viruses (32). Because we showed that IgA binds to this receptor, we suggest that TfR might mediate the entry of IgA-coated viruses and thus represent a way for viruses to infect eosinophils.
CD89 mRNA was found in rat eosinophils, confirming that rat, like human eosinophils, express FcR (10). Due to the lack of available Ab, we were unable to confirm CD89 expression at the protein level, but the present study following the recent publication of the rat genome will hopefully lead to a renewed interest for this animal model.
Taking into account our own data as well as those from the available literature, we would speculate that CD89 is probably the major receptor for eosinophil IgA-mediated activation. Indeed, no other known ligand is able to compete for binding and CD89 binds all forms of IgA: monomeric, dimeric, and secretory. pIgR would be the second most effective receptor for IgA-induced eosinophil activation. Indeed, this receptor binds the J chain present in both IgA and IgM polymers, and J chain blocking by SC led to an inhibition of ROS production ranging between 70 and 95%. TfR would next be involved in IgA binding on eosinophils. TfR only binds monomeric IgA, and inhibition of Tf-induced production by IgA was ranging between 59 and 70%. ASGP-R would be the least effective at mediating IgA effects on eosinophils. ASGP-R binds preferentially IgA2 isoforms, mainly found in SIgA, which also binds to CD89. Furthermore, anti-ASGP-R-induced ROS production was inhibited by <50% by IgA. The respective importance of the various receptors would be conditioned by eosinophil localization, according to the type of IgA encountered in the various body compartments. Tissue eosinophils within mucosae would be stimulated by SIgA through activation of CD89, pIgR, SCR, and ASGP-R, whereas blood eosinophils would be activated by serum IgA only through CD89 and TfR.
In conclusion, we have examined for the first time the expression of five known IgA receptors on human, mouse, and rat eosinophils. As it was shown for IgE receptors (1, 30) and now summarized in Table II, we have demonstrated that IgA receptor distribution is strikingly different in the three species, and that rat and mouse are thus two complementary animal models to study IgA biology.
Table II. Summary table of the IgA receptors expressed by human, mouse, and rat eosinophils
Footnotes
The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
1 This work was supported by Institut National de la Santé et de la Recherche Médicale and Institut Pasteur de Lille.
2 Address correspondence and reprint requests to Dr. David Dombrowicz, Institut National de la Santé et de la Recherche Médicale Unité 547, Institut Pasteur de Lille, 1, rue du Professeur Calmette, BP 245, 59019 Lille cedex, France. E-mail address: david.dombrowicz@pasteur-lille.fr
3 Abbreviations used in this paper: S, secretory; p, polymeric; SC, secretory component; ASGP, asialoglycoprotein; KLH, keyhole limpet hemocyanin; SA-PE, PE-conjugated streptavidin; Tf, transferrin; IC, immune complex; ROS, reactive oxygen species; EDN, eosinophil-derived neurotoxin; EPO, eosinophil peroxidase.
Received for publication July 14, 2004. Accepted for publication October 14, 2004.
References
Kayaba, H., D. Dombrowicz, G. Woerly, J. P. Papin, S. Loiseau, M. Capron. 2001. Human eosinophils and human high affinity IgE receptor transgenic mouse eosinophils express low levels of high affinity IgE receptor, but release IL-10 upon receptor activation. J. Immunol. 167:995.
Monteiro, R. C., R. W. Hostoffer, M. D. Cooper, J. R. Bonner, G. L. Gartland, H. Kubagawa. 1993. Definition of immunoglobulin A receptors on eosinophils and their enhanced expression in allergic individuals. J. Clin. Invest. 92:1681.
Honorio-Franca, A. C., P. Launay, M. M. Carneiro-Sampaio, R. C. Monteiro. 2001. Colostral neutrophils express Fc receptors (CD89) lacking chain association and mediate noninflammatory properties of secretory IgA. J. Leukocyte Biol. 69:289.
Patry, C., Y. Sibille, A. Lehuen, R. C. Monteiro. 1996. Identification of Fc receptor (CD89) isoforms generated by alternative splicing that are differentially expressed between blood monocytes and alveolar macrophages. J. Immunol. 156:4442.
Geissmann, F., P. Launay, B. Pasquier, Y. Lepelletier, M. Leborgne, A. Lehuen, N. Brousse, R. C. Monteiro. 2001. A subset of human dendritic cells expresses IgA Fc receptor (CD89), which mediates internalization and activation upon cross-linking by IgA complexes. J. Immunol. 166:346.
Maruoka, T., T. Nagata, M. Kasahara. 2004. Identification of the rat IgA Fc receptor encoded in the leukocyte receptor complex. Immunogenetics 55:712.
van Egmond, M., A. J. van Vuuren, H. C. Morton, A. B. van Spriel, L. Shen, F. M. Hofhuis, T. Saito, T. N. Mayadas, J. S. Verbeek, J. G. van de Winkel. 1999. Human immunoglobulin A receptor (FcRI, CD89) function in transgenic mice requires both FcR chain and CR3 (CD11b/CD18). Blood 93:4387.
Norderhaug, I. N., F. E. Johansen, H. Schjerven, P. Brandtzaeg. 1999. Regulation of the formation and external transport of secretory immunoglobulins. Crit. Rev. Immunol. 19:481
Robinson, J. K., T. G. Blanchard, A. D. Levine, S. N. Emancipator, M. E. Lamm. 2001. A mucosal IgA-mediated excretory immune system in vivo. J. Immunol. 166:3688
Shibuya, A., N. Sakamoto, Y. Shimizu, K. Shibuya, M. Osawa, T. Hiroyama, H. J. Eyre, G. R. Sutherland, Y. Endo, T. Fujita, et al 2000. Fc/μ receptor mediates endocytosis of IgM-coated microbes. Nat. Immunol. 1:441.
Moura, I. C., M. N. Centelles, M. Arcos-Fajardo, D. M. Malheiros, J. F. Collawn, M. D. Cooper, R. C. Monteiro. 2001. Identification of the transferrin receptor as a novel immunoglobulin (Ig)A1 receptor and its enhanced expression on mesangial cells in IgA nephropathy. J. Exp. Med. 194:417.
Vogt, T. M., A. D. Blackwell, A. M. Giannetti, P. J. Bjorkman, C. A. Enns. 2003. Heterotypic interactions between transferrin receptor and transferrin receptor 2. Blood 101:2008.
Hu, C. B., E. Y. Lee, J. E. Hewitt, J. U. Baenziger, J. Z. Mu, K. DeSchryver-Kecskemeti, D. H. Alpers. 1991. The minor components of the rat asialoglycoprotein receptor are apically located in neonatal enterocytes. Gastroenterology 101:1477
Haynes, P. A., J. D. Medh, P. H. Weigel. 1994. Inhibition of tyrosine phosphorylation in the rat hepatic lectin 1 subunit of the rat asialoglycoprotein receptor prevents ATP-dependent receptor inactivation in permeabilized hepatocytes. J. Biol. Chem. 269:33152.
Valladeau, J., V. Duvert-Frances, J. J. Pin, M. J. Kleijmeer, S. Ait-Yahia, O. Ravel, C. Vincent, F. Vega, Jr, A. Helms, D. Gorman, et al 2001. Immature human dendritic cells express asialoglycoprotein receptor isoforms for efficient receptor-mediated endocytosis. J. Immunol. 167:5767.
Lamkhioued, B., A. S. Gounni, V. Gruart, A. Pierce, A. Capron, M. Capron. 1995. Human eosinophils express a receptor for secretory component: role in secretory IgA-dependent activation. Eur. J. Immunol. 25:117.
Iikura, M., M. Yamaguchi, T. Fujisawa, M. Miyamasu, T. Takaishi, Y. Morita, T. Iwase, I. Moro, K. Yamamoto, K. Hirai. 1998. Secretory IgA induces degranulation of IL-3-primed basophils. J. Immunol. 161:1510.
Mbawuike, I. N., S. Pacheco, C. L. Acuna, K. C. Switzer, Y. Zhang, G. R. Harriman. 1999. Mucosal immunity to influenza without IgA: an IgA knockout mouse model. J. Immunol. 162:2530.
Hendrickson, B. A., L. Rindisbacher, B. Corthesy, D. Kendall, D. A. Waltz, M. R. Neutra, J. G. Seidman. 1996. Lack of association of secretory component with IgA in J chain-deficient mice. J. Immunol. 157:750.
Shimada, S., M. Kawaguchi-Miyashita, A. Kushiro, T. Sato, M. Nanno, T. Sako, Y. Matsuoka, K. Sudo, Y. Tagawa, Y. Iwakura, M. Ohwaki. 1999. Generation of polymeric immunoglobulin receptor-deficient mouse with marked reduction of secretory IgA. J. Immunol. 163:5367.
Grezel, D., M. Capron, J. M. Grzych, J. Fontaine, J. P. Lecocq, A. Capron. 1993. Protective immunity induced in rat schistosomiasis by a single dose of the Sm28GST recombinant antigen: effector mechanisms involving IgE and IgA antibodies. Eur. J. Immunol. 23:454.
Giffroy, D., A. Langendries, M. Maurice, F. Daniel, B. Lardeux, P. J. Courtoy, J. P. Vaerman. 1998. In vivo stimulation of polymeric Ig receptor transcytosis by circulating polymeric IgA in rat liver. Int. Immunol. 10:347
Dent, L. A., M. Strath, A. L. Mellor, C. J. Sanderson. 1990. Eosinophilia in transgenic mice expressing interleukin 5. J. Exp. Med. 172:1425
Woerly, G., N. Roger, S. Loiseau, D. Dombrowicz, A. Capron, M. Capron. 1999. Expression of CD28 and CD86 by human eosinophils and role in the secretion of type 1 cytokines (interleukin 2 and interferon ): inhibition by immunoglobulin A complexes. J. Exp. Med. 190:487.
Huang, W. W., E. A. Garcia-Zepeda, A. Sauty, H. C. Oettgen, M. E. Rothenberg, A. D. Luster. 1998. Molecular and biological characterization of the murine leukotriene B4 receptor expressed on eosinophils. J. Exp. Med. 188:1063.
Dombrowicz, D., B. Quatannens, J. P. Papin, A. Capron, M. Capron. 2000. Expression of a functional FcRI on rat eosinophils and macrophages. J. Immunol. 165:1266.
Woerly, G., S. Loiseau, M. Loyens, C. Schoch, M. Capron. 2003. Inhibitory effects of ketotifen on eotaxin-dependent activation of eosinophils: consequences for allergic eye diseases. Allergy 58:397.
Ross, S. R., J. J. Schofield, C. J. Farr, M. Bucan. 2002. Mouse transferrin receptor 1 is the cell entry receptor for mouse mammary tumor virus. Proc. Natl. Acad. Sci. USA 99:12386.
Dombrowicz, D., M. Capron. 2001. Eosinophils, allergy and parasites. Curr. Opin. Immunol. 13:716.
Tomassini, M., A. Tsicopoulos, P. C. Tai, V. Gruart, A. B. Tonnel, L. Prin, A. Capron, M. Capron. 1991. Release of granule proteins by eosinophils from allergic and nonallergic patients with eosinophilia on immunoglobulin-dependent activation. J. Allergy Clin. Immunol. 88:365.
Gregory, B., A. Kirchem, S. Phipps, P. Gevaert, C. Pridgeon, S. M. Rankin, D. S. Robinson. 2003. Differential regulation of human eosinophil IL-3, IL-5, and GM-CSF receptor -chain expression by cytokines: IL-3, IL-5, and GM-CSF down-regulate IL-5 receptor expression with loss of IL-5 responsiveness, but up-regulate IL-3 receptor expression. J. Immunol. 170:5359.
Bartemes, K. R., S. McKinney, G. J. Gleich, H. Kita. 1999. Endogenous platelet-activating factor is critically involved in effector functions of eosinophils stimulated with IL-5 or IgG. J. Immunol. 162:2982.
Dunzendorfer, S., N. C. Kaneider, A. Kaser, E. Woell, J. M. Frade, M. Mellado, C. Martinez-Alonso, C. J. Wiedermann. 2001. Functional expression of chemokine receptor 2 by normal human eosinophils. J. Allergy Clin. Immunol. 108:581.
Nakajima, T., H. Yamada, M. Iikura, M. Miyamasu, S. Izumi, H. Shida, K. Ohta, T. Imai, O. Yoshie, M. Mochizuki, et al 1998. Intracellular localization and release of eotaxin from normal eosinophils. FEBS Lett. 434:226.
Abu-Ghazaleh, R. I., T. Fujisawa, J. Mestecky, R. A. Kyle, G. J. Gleich. 1989. IgA-induced eosinophil degranulation. J. Immunol. 142:2393.
Rifai, A., K. Fadden, S. L. Morrison, K. R. Chintalacharuvu. 2000. The N-glycans determine the differential blood clearance and hepatic uptake of human immunoglobulin (Ig)A1 and IgA2 isotypes. J. Exp. Med. 191:2171.
Dotzauer, A., U. Gebhardt, K. Bieback, U. Gottke, A. Kracke, J. Mages, S. M. Lemon, A. Vallbracht. 2000. Hepatitis A virus-specific immunoglobulin A mediates infection of hepatocytes with hepatitis A virus via the asialoglycoprotein receptor. J. Virol. 74:10950.
Olszewska-Pazdrak, B., K. Pazdrak, P. L. Ogra, R. P. Garofalo. 1998. Respiratory syncytial virus-infected pulmonary epithelial cells induce eosinophil degranulation by a CD18-mediated mechanism. J. Immunol. 160:4889.(Véronique Decot, Ga?tane )
IgA is the most abundant class of Abs at mucosal surfaces where eosinophils carry out many of their effector functions. Most of the known IgA-mediated functions require interactions with IgA receptors, six of which have been identified in humans. These include the IgA FcR FcRI/CD89 and the receptor for the secretory component, already identified on human eosinophils, the polymeric IgR, the Fc/μR, asialoglycoprotein (ASGP)-R, and transferrin (Tf)R/CD71. In rodents, the existence of IgA receptors on mouse and rat eosinophils remains unclear. We have compared the expression and function of IgA receptors by human, rat, and mouse eosinophils. Our results show that human eosinophils express functional polymeric IgR, ASGP-R, and TfR, in addition to CD89 and the receptor for the secretory component, and that IgA receptors are expressed by rodent eosinophils. Indeed, mouse eosinophils expressed only TfR, whereas rat eosinophils expressed ASGP-R and CD89 mRNA. These results provide a molecular basis for the differences observed between human, rat, and mouse regarding IgA-mediated immunity.
Introduction
Eosinophils are known to play a role in protection against helminth parasites as well as in the physiopathology of allergic diseases, such as bronchial asthma and allergic rhinitis. These functions are linked to the eosinophil ability to release biologically active mediators upon activation through IgRs (1), complement (2), or adhesion molecules.
IgA is the predominant Ab isotype in humans and is present in the circulation as well as at the mucosal surfaces. Secretory (S)3 IgA and eosinophils are two main components of the mucosal immunity (3). Most of IgA-mediated functions require an interaction with IgA receptors. Recently, many reports have demonstrated the existence of different IgA-binding structures, which differ between species regarding their cellular distribution and function (4).
FcRI/CD89 is the receptor for the Fc portion of IgA, and this receptor is the most widely distributed in humans. Indeed, not only eosinophils (5, 6) but also neutrophils (7), monocytes (8), macrophages subsets, Kupffer cells, and dendritic cells (9) express CD89 at their surface. In rats, a CD89 homolog has recently been identified (10). By contrast, CD89 is not present in the mouse genome (11). The polymeric (p)IgR recognizes the J chain joining polymeric IgA and IgM and is responsible for transporting them through mucosal epithelial cells. A specific protease then cleaves the extracellular portion of pIgR, also known as secretory component (SC), and releases SIgA and IgM into mucosal secretions (12). In rodents, pIgR is expressed in the liver where it is involved in IgA transport into the bile (13). Fc/μR was recently characterized on human and mouse B cells and macrophages and on different tissues like liver, spleen, and intestine (14). It binds monomeric IgA and IgM and is involved in endocytosis of IgA/M-opsonized bacteria. TfR/CD71 was described as an IgA1-binding molecule involved in human renal IgA deposition on mesangial cells (15). Recently, two isoforms of this receptor were identified (16). TfR1 is expressed on a wide range of tissues, whereas TfR2 was found only in liver and on erythroid cells. Asialoglycoprotein (ASGP)-R binds IgA via its carbohydrate moieties and is mainly expressed on human and rat liver cells (17, 18) where it plays a crucial role in IgA metabolism. It has also been found at the surface of immature dendritic cells (19). The SCR is the receptor for the SIgA and is expressed on human eosinophils (20) and basophils (21).
In mouse, previous studies have described the physiological role of IgA (22), J chain (23), and pIgR (24). Early studies have also shown functional interactions between IgA and rat eosinophils in Ab-dependent cellular cytotoxicity reactions toward parasite larvae (25) or in mucosal immunity (26), thus suggesting that rat eosinophils could express functional receptors for IgA. However, the molecular characterization of IgA receptors expressed by mouse and rat eosinophils remains unknown.
Due to this lack of data about rodent, which are widely used as experimental models for human parasitic and allergic diseases, and due to major differences among IgRs already identified between humans and the two rodent species (1), we have undertaken a comparative analysis of IgA receptors on human, rat, and mouse eosinophils.
Materials and Methods
Eosinophil donors
Normal donors and hypereosinophilic patients with drug hypersensitivity were selected for this study, after informed consent.
Animals
Animals were housed in a specific pathogen-free facility from the Institut Pasteur de Lille, with continuous access to food and water. Eight- to 10-wk-old Brown Norway rats and 6- to 12-wk-old BALB/c mice or IL-5 transgenic mice expressing IL-5 under the control of the human CD2 promoter (27) were used.
Reagents
Anti-human CD16-, streptavidin-coated magnetic beads, and the MACS system were from Miltenyi Biotec. Percoll was from Amersham Biosciences. RPMI 1640, glutamine, penicillin, streptomycin, FCS, and ultroser were obtained from Invitrogen Life Technologies. Paraformaldehyde, saponin, HEPES, human SIgA, human IgM, human IgG, biotinylated human transferrin (Tf), biotinylated anti-rabbit IgG, hydrogen peroxide, and luminol were from Sigma-Aldrich. BSA was from Eurobio. Luciferin was purchased from Roche Diagnostics. Human serum IgA and mouse anti-human SC was obtained from ICN. Biotinylated anti-human IgA, anti-human IgM, anti-mouse CD19, anti-CD8 (Ly-2), anti-CD90 (Thy1.2), and mouse anti-rat IgA; rat anti-mouse IgA; purified mouse IgA; purified rat IgA; and purified rat IgM were from BD Pharmingen. For stimulation experiments, purified mouse anti-human IgM (Serotec), rat anti-DNP IgA (Technopharm), and DNP-keyhole limpet hemocyanin (KLH) (Calbiochem) were used. PE-conjugated streptavidin (SA-PE) was obtained from Molecular Probes. FITC-conjugated goat anti-rat, donkey anti-rabbit IgG (H+L) F(ab')2 and purified rat IgG were obtained from Jackson ImmunoResearch Laboratories. Goat anti-mouse IgG1 was purchased from Southern Biotechnology. Human SC purified from colostrum was a kind gift from Dr. A. Pierce (University Lille I, Lille, France). Rabbit anti-human and anti-rat ASGP-R Ab were kindly provided by Dr. D. Alpers (Washington University School of Medicine, St. Louis, MO) and Dr. R. Stockert (Albert Einstein College of Medicine, Bronx, NY).
Cell purification
Human eosinophils were isolated from the venous blood of patients and healthy donors, using immunomagnetic beads and the MACS system as previously described (28). Diluted whole blood (1/1) was layered onto a Percoll gradient (density, 1.082 g/L) and centrifuged at 1800 rpm for 20 min. After centrifugation, PBMC were collected at the interface and the granulocytes in the pellet. After hypotonic saline lysis, granulocytes were incubated with anti-CD16 beads to remove neutrophils. Eosinophils were eluted following passage through the field of a permanent magnet. Eosinophil purity, assessed on cytospin preparation following staining with May-Grünwald Giemsa (RAL 555), was >98%.
Mouse splenocytes were obtained by gentle dissociation of spleens in ice-cold PBS. Aggregates were removed by passage through a nylon filter and erythrocytes were lysed using hypotonic saline. Splenic eosinophils were purified from IL-5 transgenic mice by negative selection as previously described (29). Briefly, cells were incubated 30 min on ice with biotinylated anti-mouse CD19, anti-CD8 (Ly-2), and anti-CD90 (Thy1.2). After washing, cells were incubated 15 min with streptavidin-coated beads at 4°C. Purified eosinophils were obtained following passage through the field of a permanent magnet. Purity was >90%.
Splenic B lymphocytes were purified from BALB/c mice by positive selection. Briefly, cells were incubated 30 min with anti-mouse CD19, washed twice, and incubated with streptavidin-coated beads at 4°C. Lymphocytes were recovered from the column after detachment of the cells out of the magnetic field. Purity was >95%.
Mouse peritoneal cells were obtained by flushing the peritoneal cavity of BALB/c mice with 10 ml of PBS. After filtration on a nylon filter, cells were cultured overnight in complete medium, and adherent macrophages were recovered by scraping. Macrophage purity was >95%.
Rat peritoneal cells were obtained by flushing the peritoneal cavity with 50 ml of ice-cold PBS. Aggregates were removed from cell suspensions by filtration on a nylon filter, and erythrocytes were lysed using hypotonic saline. For stimulation experiments, purified rat eosinophils were obtained by sorting peritoneal cells according to their typical forward and side scatter (30), using an ELITE cell sorter (Coulter). Eosinophil purity after sorting was ranging between 90 and 98%. Contaminating cells were macrophages.
Rat splenocytes were obtained by gentle dissociation of spleens in ice-cold PBS. Aggregates were removed by nylon filtration, and erythrocytes were lysed using hypotonic saline.
Cell culture
For TfR up-regulation experiments, purified human eosinophils were cultured for 1 day in RPMI 1640 supplemented with 10% heat-inactivated FCS, 25 mM HEPES, 2 mM L-glutamine, 100 IU/ml penicillin, and 100 μg/ml streptomycin (complete medium) in the absence or in the presence of 25 μg/ml Tf.
After cell sorting, rat peritoneal eosinophils were kept for 18 h in phenol red-free RPMI 1640 supplemented with 2% Ultroser before stimulation.
The various cell lines used as control, HT29, CaCo2, and HEPG2 were cultured in complete medium.
RT-PCR
Total RNA was extracted from 10 x 106 highly purified (98%) eosinophils, PBMC, cell lines, mouse lymphocytes, mouse macrophages, or rat spleen cells using RNAplus extraction reagent (Q Biogene). Mouse and rat liver mRNA were extracted by CsCl centrifugation. Reverse transcription was performed using SuperScript RT (Invitrogen Life Technologies). Amplification by PCR using Taq polymerase (Bioprobe) was performed with the primers (Proligo) and under the conditions described in Table I. Amplified products were then normalized according to the signal intensity provided by the -actin amplification, loaded on a 1% agarose gel, stained with ethidium bromide, and photographed under UV light.
Table I. Sequences of primers used for PCR amplification of cDNA, amplicon sizes, annealing temperatures, and PCR cycle numbersa
Flow cytometry
Unless specified otherwise, all incubations were performed at 4°C for 30 min with cells resuspended at 4 x 106/ml in PBS containing 1% BSA and 0.05% sodium azide (FACS buffer) with 10 μg/ml Ab, and cells were washed twice with FACS buffer between each step of staining.
Fifty microliters of the cell suspensions (2 x 105 cells) was incubated with 100 μg/ml human IgM or with rat or mouse IgA for 45 min in round-bottom 96-well plates. After washing, the cells were then incubated with biotinylated anti-human IgM for human cells, biotinylated mouse anti-rat IgA for rat cells, and purified rat anti-mouse IgA for mouse cells. Biotinylated or purified anti-Ig were incubated followed by an incubation with SA-PE or with FITC goat anti-rat Igs.
For human pIgR, competition experiments were performed between IgM and the other isotypes, and cells were thus preincubated with 500 μg/ml SIgA, serum (monomeric) IgA or IgG before incubation with human IgM.
For the competition assay between SC and IgM, a concentration of 500 μg/ml SC was chosen, as it was previously determined by flow cytometry to saturate cell surface receptors. SC was incubated 45 min and its binding was revealed by a purified mouse anti-SC and a PE-conjugated goat anti-mouse IgG.
Rabbit Abs (anti-human pIgR and anti-human and rat ASGP-R) and normal rabbit serum were used at 1:400, and their binding was revealed by a donkey FITC-conjugated anti-rabbit IgG. Competition with the anti-ASGP-R Ab was performed by incubating the cells with 400 μg/ml rat IgA, human serum IgA, human SIgA, IgG, or IgM, a concentration that saturates surface receptors, before adding the anti-ASGP-R Ab.
Biotinylated human Tf (100 μg/ml) was incubated with mouse and human eosinophils, and binding was detected after incubation with SA-PE for 20 min. For the competition assay, a concentration of 500 μg/ml mouse, human serum IgA, SIgA, IgM, or IgG was chosen because it was previously determined to saturate the receptor. Competitions were performed by preincubation of the various Ig 45 min before adding Tf.
For intracellular detection of Tf binding, cells were fixed with 2% paraformaldehyde in PBS for 10 min. After washing in PBS, cells were resuspended at 4 x 106/ml in PBS containing 1% BSA and 0.1% saponin for 10 min at room temperature. The cells were then incubated 45 min with biotinylated human Tf followed by an incubation with SA-PE. Samples were washed twice in PBS-saponin buffer, once in PBS, and resuspended in PBS-0.5% BSA for analysis.
Cells were analyzed by flow cytometry using a FACSCalibur equipped with the CellQuest software (BD Biosciences). A total of 104 events was acquired for each sample.
IgA-anti-IgA immune complexes (IC)
Mouse or rat IgA-anti-IgA IC were prepared by incubating concentrations ranging from 2 to 18 μg/ml IgA with 18 to 2 μg/ml anti-IgA for 30 min at room temperature to obtain an IgA-anti-IgA ratio ranging from 1:9 to 9:1. IC were then incubated with cells for 30 min at 4°C. For rat eosinophils, competition of IgA IC binding to eosinophils was performed using 100 μg/ml rat IgM or IgG.
Reactive oxygen species (ROS) production
ROS production by highly purified eosinophils was examined using luminol-dependent chemiluminescence as previously described (31). For IgM activation, 50 μl of purified eosinophils (5 x 105 cells) was incubated alone or with IgM (15 μg/ml) for 1 h at 37°C. The inhibition of IgM-induced ROS release was performed by preincubating the cells with 500 μg/ml SC to saturate the SCR, and IgM was incubated with 15 μg/ml SC at 37°C to block its J chain for 1 h. The complex was then incubated for 1 h with the cells, followed, after washing, by the addition of anti-IgM (20 μg/ml) and luminol (25 μg/ml). Kinetics was performed at 37°C over a 40-min period.
For Tf activation, 96-well flat-bottom tissue culture plates (Falcon) were coated with 100 μg/ml human Tf in PBS at 37°C for 2 h. Coated wells were washed twice with PBS before use, and cells (5 x 105 cells) were added to the wells. For IgA and Tf competition assay, cells were incubated for 1 h at 37°C with human serum IgA and washed to remove the IgA excess before the Tf stimulation. Chemiluminescence was measured for 5 s with a luminometer (Victor2 Wallac; PerkinElmer). Kinetics was performed at 37°C over a 40-min period. Results were expressed in counts per second.
Measurement of eosinophil neurotoxin (EDN) and eosinophil peroxidase (EPO) for assessment of degranulation
For EDN measurement, highly purified human eosinophils (2 x 106/ml) were first incubated in culture medium either alone or with human IgM (15 μg/ml). After 1 h, cells were stimulated with 20 μg/ml purified mouse anti-human IgM at 37°C.
For Tf stimulation, wells were coated with 100 μg/ml human Tf for 2 h. After washing twice in PBS, cells were added to the wells at the concentration of 2 x 106/ml. IgA and Tf competition was performed as described for the chemiluminescence assay.
After 18 h, supernatants were collected and stored at –20°C until measurement of EDN concentration using a specific ELISA (MBL International) according to the manufacturer’s instructions. Lower detection limit of the assay was 0.62 ng/ml.
For EPO measurement, rat peritoneal eosinophils (2 x 105) were incubated either alone or with 10 μg/ml rat anti-DNP IgA for 1 h. IgA was then cross-linked by incubation of the cells with 20 μg/ml DNP-KLH for 2 h. After stimulation, cells were centrifuged, and 50 μl of supernatant was collected. Fifty microliters of luciferin (6 x 10–5 M), hydrogen peroxide (0.15%), and luminol (200 μg/ml) were added to the supernatants before the EPO measurement with a luminometer (Victor2 Wallac).
Statistical analysis
Statistical analysis was performed using Student’s paired t test. Statistical significance was set at p < 0.05.
Results
Human eosinophils express pIgR, ASGP-R, and TfR
Because CD89 and SCR expression has been described on human eosinophils, we determined whether other IgA-binding structures might also be expressed by these cells. RT-PCR was performed on RNA extracted from highly purified eosinophils from a healthy donor and a patient with drug hypersensitivity. An amplicon was detected for pIgR in both eosinophil preparations (Fig. 1A), as well as in neutrophils and PBMC (data not shown). By contrast, no Fc/μR transcripts were detected on eosinophils, whereas a signal was detected on PBMC (Fig. 1B). The expression of pIgR was confirmed by flow cytometry using an anti-pIgR-specific Ab (Fig. 1C) or IgM, one of the pIgR ligand (D). J chain-containing IgM binding was not inhibited by an excess of SIgA (with its J chain blocked by SC) or serum (monomeric) IgA or IgG (Fig. 1D, inset), confirming that this receptor actually binds polymeric Ig through its free J chain as previously reported (12).
FIGURE 1. Expression of pIgR by human eosinophils. A, RT-PCR analysis of pIgR mRNA expression by human eosinophils. RNA extracted from highly purified eosinophils (ND, normal donor; DH, drug hypersensitivity) was subjected to reverse transcription and amplified by PCR using primers specific for pIgR and -actin. Total mRNA from HT29 was used as positive control. B, RT-PCR analysis of Fc/μR mRNA expression in human PBMC and eosinophils (ND; DH). C, Flow cytometric analysis of pIgR expression by human eosinophils. Eosinophils were incubated with rabbit anti-pIgR Ab (black line) followed by a FITC-conjugated anti-rabbit IgG (dashed line). Normal rabbit serum was used as negative control (filled histogram). D, IgM binding on human eosinophils. Eosinophils (ND) were incubated with IgM (black line). Biotinylated anti-human IgM was used as control (filled histogram). Inset, Competition of IgM binding by preincubation with a 5-fold excess of serum IgA (IgA), SIgA, and IgG. E, Representative time course experiment of IgM-induced ROS release by human eosinophils (ND). Human eosinophils were incubated for 1 h with () or without IgM followed by the addition, after washing, of anti-IgM (). For inhibition of IgM binding, cell surface was saturated by human SC and then incubated with IgM/SC complexes (?). Results are expressed in counts per second (CPS). F, IgM-induced EDN release by eosinophils. Eosinophils (ND) were incubated or not with IgM for 1 h. Anti-human IgM was then added to cross-link receptors, and supernatants were collected after 18-h stimulation (n = 4). MFI, Mean fluorescence intensity.
Furthermore, activation with IgM followed by cross-linking with anti-IgM Ab induced ROS production (Fig. 1E). To confirm that eosinophil activation was mediated by pIgR, an inhibition assay between SC and IgM was performed. ROS release was inhibited when IgM J chain binding to pIgR was prevented by preincubation of both IgM and eosinophils with SC. Finally, IgM cross-linked with anti-IgM was able to induce EDN release by eosinophils (Fig. 1F). These results thus demonstrate that human eosinophils could express a functional pIgR.
ASGP-R expression was first investigated by RT-PCR. Primers were chosen in a highly conserved region sharing 99% homology with mouse and rat sequences, thus allowing the use of rat liver as positive control. Human eosinophils from a normal donor and a patient presenting a drug hypersensitivity expressed ASGP-R mRNA (Fig. 2A). Surface expression of the receptor was then confirmed by flow cytometry using an anti-ASGP-R Ab (Fig. 2B). To confirm that eosinophils ASGP-R also bind IgA, a competition between serum IgA, secretory IgA, IgM, or IgG, and the anti-ASGP-R Ab was performed. Ab binding inhibition was ranging between 17.5 and 47.5% upon preincubation with SIgA (dimeric) but not with monomeric serum IgA, IgM, or IgG, suggesting that ASGP-R expressed by human eosinophils might act as an IgA receptor (Fig. 2B, inset). ASGP-R-mediated eosinophil activation was then confirmed by the release of ROS following stimulation with the anti-ASGP-R Ab (Fig. 2C).
FIGURE 2. ASGP-R expression by human eosinophils. A, RT-PCR analysis of ASGP-R mRNA expression by human eosinophils (ND, normal donor; DH, drug hypersensitivity). Primers were chosen in a region showing a high degree of homology between rat and human. Rat liver was used as positive control. B, Flow cytometric analysis of ASGP-R expression by eosinophils (ND). Binding of anti-ASGP-R Ab (black line) was detected using an FITC-conjugated anti-rabbit IgG. Normal rabbit serum (filled histogram) was used as control. Inhibition of anti-ASGP-R binding was obtained by cell preincubation with 400 μg/ml SIgA (dotted line). Inset, Competition of anti-ASGP-R binding by preincubation with 400 μg/ml IgM, IgG, and (serum) IgA. C, Representative time course experiment showing ROS release by normal donor eosinophils. Cells were incubated for 1 h at 37°C with a rabbit anti-ASGP-R Ab (). Measurement was initiated upon addition of the anti-rabbit Ig. Normal rabbit serum was used as control () (n = 3). MFI, Mean fluorescence intensity.
Expression of TfR mRNA by human eosinophils was first assessed by RT-PCR. Fig. 3A shows that TfR1 but not TfR2 was detected. Flow cytometric analysis of TfR expression showed a weak expression at the surface of freshly isolated eosinophils (Fig. 3B). However, staining of permeabilized cells revealed the presence of an intracellular pool of receptors, which might be available for subsequent mobilization (Fig. 3B, inset). Indeed, following overnight culture in the presence of Tf, free and occupied TfR were detected at the cell surface (Fig. 3C). To demonstrate that TfR acts as an IgA receptor on eosinophils, we thus tested whether preincubation with IgA was able to inhibit Tf binding. An excess of serum IgA was inhibiting Tf binding to its receptor, whereas no inhibition was obtained with an excess of IgM or IgG. Interestingly, an excess of SIgA was not able to inhibit Tf binding to its receptor (Fig. 3D, inset).
FIGURE 3. TfR expression by human eosinophils. A, TfR1 and -2 mRNA expression by eosinophils revealed by PCR. Eosinophils (ND, normal donor; DH, drug hypersensitivity), HT29, and CaCo2 were positive for TfR1 expression. HEPG2 was used as positive control for TfR2 expression. B, Flow cytometric analysis of the Tf binding to normal donor eosinophils. Freshly purified eosinophils were incubated with biotinylated Tf (black line) followed by an incubation with SA-PE (negative control, filled histogram). Inset, Intracellular detection of Tf in human eosinophils. C, Tf binding on eosinophils (ND) cultured for 1 day with 25 μg/ml biotinylated Tf. Cells were incubated (total receptors, black line) or not (occupied receptors, dotted line) with 100 μg/ml Tf. SA-PE, in the absence of Tf (filled histogram) was used as negative control. D, Competition of Tf binding by (serum) IgA. Inset, Competition by preincubation with 500 μg/ml IgM, IgG, and SIgA. E, Representative time course experiment of ROS release following stimulation of eosinophils with immobilized Tf. Eosinophils (ND) were incubated in tissue culture wells coated () or not (?) for 2 h with 100 μg/ml Tf. To inhibit Tf stimulation, eosinophils were incubated with human serum IgA, washed twice with PBS before incubation in Tf-coated wells () (n = 5). F, EDN release by eosinophils (ND) following Tf stimulation. Eosinophils were stimulated for 18 h in Tf-coated wells. Inhibition of stimulation was obtained by preincubating the cells with serum IgA as described above (n = 5). MFI, Mean fluorescence intensity.
Furthermore, immobilized Tf was able to stimulate ROS production from freshly isolated cells, suggesting that TfR surface expression levels were below the detection threshold of our cytometric method. As might have been expected from flow cytometry data, serum IgA was also able to efficiently inhibit Tf-induced ROS release (inhibition range, 59–70%), thus indicating that TfR expressed on eosinophils was also a functional receptor for serum IgA (Fig. 3E). This result was confirmed by a significant inhibition (94.5%) of the Tf-induced EDN release by serum IgA (Fig. 3F).
IgA receptors expressed on mouse eosinophils
Because CD89 is not found in mice, we examined which molecular structure would bind IgA on mouse eosinophils. Because naive wild-type mice harbor only very few eosinophils, we used purified splenic eosinophils from IL-5 transgenic mice.
RT-PCR was performed on freshly purified eosinophils to detect the presence of specific amplicons for pIgR, Fc/μR, ASGP-R, as well as TfR1 and -2. According to the receptor investigated, RNA from tissues or cells known to express the receptor of interest were chosen as positive controls. Amplicons corresponding to Fc/μR, TfR1, and -2 were detected in eosinophil samples. By contrast, no bands corresponding to pIgR or ASGP-R were amplified, whereas the corresponding positive controls displayed amplification of the fragments with the expected size (Fig. 4A).
FIGURE 4. IgA receptors expressed by mouse eosinophils. A, RT-PCR analysis of Fc/μR, pIgR, ASGP-R, and TfR expression by mouse eosinophils. Mouse liver was used as positive control and mouse -actin was used as housekeeping gene. B, Monomeric IgA (black line) was detected using rat anti-mouse IgA and FITC-conjugated goat anti-rat F(ab')2 (filled histogram). C, Binding curve of IgA-anti-IgA IC on mouse eosinophils. D, Tf binding on mouse eosinophils and inhibition by mouse IgA (n = 3). Results are expressed as mean fluorescence intensity (MFI) after subtraction of the control.
Expression of IgA receptors by mouse eosinophils was next confirmed by flow cytometry. A weak binding of monomeric IgA was detected on freshly isolated eosinophils (Fig. 4B). Because the expression of a low affinity receptor could account for this weak binding, IgA-anti-IgA IC were then used, and significant increase in binding to eosinophils, compared with monomeric IgA, was observed (Fig. 4C). Maximal binding was achieved with a 1:1 ratio between IgA and anti-IgA. A very weak IgM binding to the same eosinophil preparation was also observed, but no competition between IgA and IgM was detected, suggesting the absence not only of pIgR but also of Fc/μR at the surface of mouse eosinophils (data not shown).
Because the two isoforms of TfR mRNA were found in eosinophils, we investigated whether this receptor was expressed at their surface. Human Tf was used because the Tf binding site is highly conserved between species (32), and binding to cell surface was detected. To assess that TfR also binds IgA, competition experiments were performed. For this assay, a high dose of mouse IgA was used to induce IgA aggregation, because aggregated IgA binds more efficiently to the receptor than monomeric IgA. An inhibition of Tf binding ranging between 17 and 58% was observed in the presence of IgA (Fig. 4D), demonstrating that TfR is an IgA receptor in mouse eosinophils.
IgA receptors expressed by rat eosinophils
To determine which receptor is involved in IgA binding to rat eosinophils, we analyzed the expression of potential IgA-binding structures (CD89, pIgR, Fc/μR, ASGP-R, and TfR) by peritoneal eosinophils by RT-PCR.
No expression of pIgR, Fc/μR, TfR1, or -2 was detected on rat eosinophils, despite the presence of a strong signal on liver tissue chosen as positive control. By contrast, a specific signal was detected for ASGP-R and for a homolog of human CD89 (Fig. 5A).
FIGURE 5. IgA receptors expressed by rat eosinophils. A, RT-PCR analysis of IgA receptor expression by rat eosinophils. RNA from cell-sorted rat eosinophils, total liver and spleen cells (80% lymphocytes, 20% macrophages) was subjected to reverse transcription and amplified by PCR using primers specific for ASGP-R, TfR1 and -2, pIgR, CD89, and Fc/μR. B, Flow cytometric analysis of monomeric IgA binding on rat eosinophils. Purified rat IgA (black line) was incubated for 45 min at 4°C. Binding was detected using biotinylated anti-rat IgA followed by an incubation with SA-PE (filled histogram). C, Binding curve of rat IgA-anti-IgA IC. D, Flow cytometric analysis of ASGP-R expression. Peritoneal rat eosinophils were incubated with a rabbit anti-human ASGP-R Ab (black line) followed by incubation with FITC-conjugated donkey anti-rabbit IgG. Rabbit serum was used as negative control (filled histogram). Inhibition of the anti-ASGP-R Ab binding was obtained by preincubating with 400 μg/ml rat IgA (dotted line). Inset, Competition by preincubation with 400 μg/ml rat IgG and IgM. E, EPO release by purified peritoneal rat eosinophils. Eosinophils were incubated or not with 10 μg/ml anti-DNP IgA for 1 h at 37°C, followed by the addition of 20 μg/ml DNP-KLH. After 2-h stimulation, EPO was measured in the supernatants (n = 3). MFI, Mean fluorescence intensity.
We next analyzed rat IgA binding on eosinophils. No binding of monomeric IgA was observed on rat eosinophils (Fig. 5B). However, IgA-anti-IgA IC displayed a weak binding (Fig. 5C). No binding inhibition was observed by preincubation with a 10-fold excess of either IgG or IgM (data not shown). To confirm at the protein level, the results obtained by PCR evidencing an amplicon for ASGP-R, a flow cytometric approach was used. ASGP-R was detected using a specific Ab raised against the intracellular portion of the receptor (data not shown). This result was confirmed by using an anti-human ASGP-R Ab, which allowed us to detect the receptor at the cell surface (Fig. 5D). Indeed, rat and human ASGP-R shared 70% homology in their extracellular domains. To assess that ASGP-R expressed on rat eosinophils also binds IgA, a competition was performed using rat IgA, IgM, or IgG. An inhibition of Ab binding ranging between 30 and 35% was observed when cells were incubated with IgA (Fig. 5D) but not with IgM or IgG (D, inset). This result confirmed that ASGP-R can be considered as an IgA receptor on the rat eosinophils. Furthermore, rat eosinophils were activated by IgA, because EPO was released 2 h following stimulation, thus demonstrating the functionality of at least one IgA receptor (Fig. 5E).
Discussion
IgA and eosinophils are two major players in mucosal immunity, but their interaction is not restricted to mucosal tissues and can also occur in blood in the course of parasitic or allergic diseases (33). Several convergent studies have shown that IgA is one of the most potent stimuli for eosinophils (3, 34). In the present work, we investigated whether human eosinophils could express IgA receptors other than CD89 and the SCR, and we determined for the first time which receptors were involved in IgA binding to mouse and rat eosinophils.
Our data indicate that human eosinophils expressed a functional pIgR and demonstrate that this receptor can be found on cells other than secretory epithelial cells. Furthermore, these results suggest that eosinophils themselves might be a potential source of SC, which may act on eosinophils as an autocrine mediator after binding to SCR. Such a mechanism was demonstrated for many other eosinophil mediators, such as IL-5 (35), platelet-activating factor (36), or eotaxin (37, 38). Our data suggest that eosinophils are able to be activated by pIg following lamina propria infiltration and thus play an active role in host defense against pathogens that have crossed the epithelial barrier. Such a result was already demonstrated for polymeric IgA activation (39), but we here demonstrate for the first time that this activation leads to ROS and EDN release following IgM stimulation of eosinophils. Furthermore, we have shown that EDN secretion was observed upon incubation with anti-IgM alone suggesting the presence of IgM, presumably bound to pIgR at the cell surface.
No pIgR expression was detected, even at the mRNA level, in mouse or rat eosinophils, thus demonstrating that receptor expression is restricted to human eosinophils. In rodents, IgA is mainly produced as polymers (4), and the absence of pIgR on circulating cells may prevent IgA-induced, pIgR-mediated activation.
Fc/μR was not detected on human and rat eosinophils, even at the mRNA level, but mouse eosinophils were found weakly positive for the expression of Fc/μR transcripts. However, because no competition was observed between IgA and IgM, the existence of the receptor at the cell surface was not confirmed (data not shown). These results thus support the absence of Fc/μR on Gr-1+ granulocytes (likely neutrophils) observed by Shibuya et al. (14).
Some studies have described the important role of the ASGP-R in IgA metabolism (40). We determined whether this receptor was expressed by cells other than hepatocytes and whether this receptor could play a role in IgA binding to eosinophils. Indeed, ASGP-R was detected on eosinophils, thus extending the demonstration of ASGP-R mRNA in human granulocytes (likely neutrophils) derived from CD34+ precursors in the absence of IL-5 (19). Furthermore, SIgA but not serum IgA was able to compete with the anti-ASGP-R Ab, a result explained by the fact that ASGP-R binds IgA2 and not IgA1, the main isotype of IgA found in serum. The partial inhibition of Ab binding to human eosinophils by SIgA is consistent with the expression of CD89 and SCR, which also bind SIgA.
Only human and rat eosinophils were found positive for ASGP-R expression. ASGP-R mediates the release of ROS by human eosinophils and might also account for EPO release by rat eosinophils, suggesting that the receptor is involved not only in endocytosis of glycoproteins but also in the development of cytotoxic responses. Recent studies have also demonstrated that this receptor can be used by viruses for attachment and entry into cells. In particular, Dotzauer et al. (41) have shown that hepatitis A virus complexed with specific IgA can use IgA-specific receptors for entry into hepatocytes, and some viruses, such as respiratory syncytial virus, have been shown to infect eosinophils (42). Our results suggest that such phenomenon may occur via ASGP-R.
We next studied the possibility that human and rodent eosinophils could express TfR, recently identified as an IgA1 receptor (15). The two specific TfR have already been described on myeloid cells, and Sakamoto et al. (43) have reported the existence of TfR on neutrophils. No TfR was detected on rat eosinophils, whereas human eosinophils were found positive for the TfR1 mRNA expression and mouse eosinophils expressed mRNA for the two isoforms. Our results demonstrate that TfR is expressed by eosinophils and is an IgA receptor, because serum IgA, mainly containing IgA1, inhibits Tf-induced ROS and EDN release. This suggests that IgA may play an anti-inflammatory role in blood by blocking the release of cytotoxic mediators induced by circulating Tf. Some studies have demonstrated that TfR is an entry receptor for some enveloped viruses (32). Because we showed that IgA binds to this receptor, we suggest that TfR might mediate the entry of IgA-coated viruses and thus represent a way for viruses to infect eosinophils.
CD89 mRNA was found in rat eosinophils, confirming that rat, like human eosinophils, express FcR (10). Due to the lack of available Ab, we were unable to confirm CD89 expression at the protein level, but the present study following the recent publication of the rat genome will hopefully lead to a renewed interest for this animal model.
Taking into account our own data as well as those from the available literature, we would speculate that CD89 is probably the major receptor for eosinophil IgA-mediated activation. Indeed, no other known ligand is able to compete for binding and CD89 binds all forms of IgA: monomeric, dimeric, and secretory. pIgR would be the second most effective receptor for IgA-induced eosinophil activation. Indeed, this receptor binds the J chain present in both IgA and IgM polymers, and J chain blocking by SC led to an inhibition of ROS production ranging between 70 and 95%. TfR would next be involved in IgA binding on eosinophils. TfR only binds monomeric IgA, and inhibition of Tf-induced production by IgA was ranging between 59 and 70%. ASGP-R would be the least effective at mediating IgA effects on eosinophils. ASGP-R binds preferentially IgA2 isoforms, mainly found in SIgA, which also binds to CD89. Furthermore, anti-ASGP-R-induced ROS production was inhibited by <50% by IgA. The respective importance of the various receptors would be conditioned by eosinophil localization, according to the type of IgA encountered in the various body compartments. Tissue eosinophils within mucosae would be stimulated by SIgA through activation of CD89, pIgR, SCR, and ASGP-R, whereas blood eosinophils would be activated by serum IgA only through CD89 and TfR.
In conclusion, we have examined for the first time the expression of five known IgA receptors on human, mouse, and rat eosinophils. As it was shown for IgE receptors (1, 30) and now summarized in Table II, we have demonstrated that IgA receptor distribution is strikingly different in the three species, and that rat and mouse are thus two complementary animal models to study IgA biology.
Table II. Summary table of the IgA receptors expressed by human, mouse, and rat eosinophils
Footnotes
The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
1 This work was supported by Institut National de la Santé et de la Recherche Médicale and Institut Pasteur de Lille.
2 Address correspondence and reprint requests to Dr. David Dombrowicz, Institut National de la Santé et de la Recherche Médicale Unité 547, Institut Pasteur de Lille, 1, rue du Professeur Calmette, BP 245, 59019 Lille cedex, France. E-mail address: david.dombrowicz@pasteur-lille.fr
3 Abbreviations used in this paper: S, secretory; p, polymeric; SC, secretory component; ASGP, asialoglycoprotein; KLH, keyhole limpet hemocyanin; SA-PE, PE-conjugated streptavidin; Tf, transferrin; IC, immune complex; ROS, reactive oxygen species; EDN, eosinophil-derived neurotoxin; EPO, eosinophil peroxidase.
Received for publication July 14, 2004. Accepted for publication October 14, 2004.
References
Kayaba, H., D. Dombrowicz, G. Woerly, J. P. Papin, S. Loiseau, M. Capron. 2001. Human eosinophils and human high affinity IgE receptor transgenic mouse eosinophils express low levels of high affinity IgE receptor, but release IL-10 upon receptor activation. J. Immunol. 167:995.
Monteiro, R. C., R. W. Hostoffer, M. D. Cooper, J. R. Bonner, G. L. Gartland, H. Kubagawa. 1993. Definition of immunoglobulin A receptors on eosinophils and their enhanced expression in allergic individuals. J. Clin. Invest. 92:1681.
Honorio-Franca, A. C., P. Launay, M. M. Carneiro-Sampaio, R. C. Monteiro. 2001. Colostral neutrophils express Fc receptors (CD89) lacking chain association and mediate noninflammatory properties of secretory IgA. J. Leukocyte Biol. 69:289.
Patry, C., Y. Sibille, A. Lehuen, R. C. Monteiro. 1996. Identification of Fc receptor (CD89) isoforms generated by alternative splicing that are differentially expressed between blood monocytes and alveolar macrophages. J. Immunol. 156:4442.
Geissmann, F., P. Launay, B. Pasquier, Y. Lepelletier, M. Leborgne, A. Lehuen, N. Brousse, R. C. Monteiro. 2001. A subset of human dendritic cells expresses IgA Fc receptor (CD89), which mediates internalization and activation upon cross-linking by IgA complexes. J. Immunol. 166:346.
Maruoka, T., T. Nagata, M. Kasahara. 2004. Identification of the rat IgA Fc receptor encoded in the leukocyte receptor complex. Immunogenetics 55:712.
van Egmond, M., A. J. van Vuuren, H. C. Morton, A. B. van Spriel, L. Shen, F. M. Hofhuis, T. Saito, T. N. Mayadas, J. S. Verbeek, J. G. van de Winkel. 1999. Human immunoglobulin A receptor (FcRI, CD89) function in transgenic mice requires both FcR chain and CR3 (CD11b/CD18). Blood 93:4387.
Norderhaug, I. N., F. E. Johansen, H. Schjerven, P. Brandtzaeg. 1999. Regulation of the formation and external transport of secretory immunoglobulins. Crit. Rev. Immunol. 19:481
Robinson, J. K., T. G. Blanchard, A. D. Levine, S. N. Emancipator, M. E. Lamm. 2001. A mucosal IgA-mediated excretory immune system in vivo. J. Immunol. 166:3688
Shibuya, A., N. Sakamoto, Y. Shimizu, K. Shibuya, M. Osawa, T. Hiroyama, H. J. Eyre, G. R. Sutherland, Y. Endo, T. Fujita, et al 2000. Fc/μ receptor mediates endocytosis of IgM-coated microbes. Nat. Immunol. 1:441.
Moura, I. C., M. N. Centelles, M. Arcos-Fajardo, D. M. Malheiros, J. F. Collawn, M. D. Cooper, R. C. Monteiro. 2001. Identification of the transferrin receptor as a novel immunoglobulin (Ig)A1 receptor and its enhanced expression on mesangial cells in IgA nephropathy. J. Exp. Med. 194:417.
Vogt, T. M., A. D. Blackwell, A. M. Giannetti, P. J. Bjorkman, C. A. Enns. 2003. Heterotypic interactions between transferrin receptor and transferrin receptor 2. Blood 101:2008.
Hu, C. B., E. Y. Lee, J. E. Hewitt, J. U. Baenziger, J. Z. Mu, K. DeSchryver-Kecskemeti, D. H. Alpers. 1991. The minor components of the rat asialoglycoprotein receptor are apically located in neonatal enterocytes. Gastroenterology 101:1477
Haynes, P. A., J. D. Medh, P. H. Weigel. 1994. Inhibition of tyrosine phosphorylation in the rat hepatic lectin 1 subunit of the rat asialoglycoprotein receptor prevents ATP-dependent receptor inactivation in permeabilized hepatocytes. J. Biol. Chem. 269:33152.
Valladeau, J., V. Duvert-Frances, J. J. Pin, M. J. Kleijmeer, S. Ait-Yahia, O. Ravel, C. Vincent, F. Vega, Jr, A. Helms, D. Gorman, et al 2001. Immature human dendritic cells express asialoglycoprotein receptor isoforms for efficient receptor-mediated endocytosis. J. Immunol. 167:5767.
Lamkhioued, B., A. S. Gounni, V. Gruart, A. Pierce, A. Capron, M. Capron. 1995. Human eosinophils express a receptor for secretory component: role in secretory IgA-dependent activation. Eur. J. Immunol. 25:117.
Iikura, M., M. Yamaguchi, T. Fujisawa, M. Miyamasu, T. Takaishi, Y. Morita, T. Iwase, I. Moro, K. Yamamoto, K. Hirai. 1998. Secretory IgA induces degranulation of IL-3-primed basophils. J. Immunol. 161:1510.
Mbawuike, I. N., S. Pacheco, C. L. Acuna, K. C. Switzer, Y. Zhang, G. R. Harriman. 1999. Mucosal immunity to influenza without IgA: an IgA knockout mouse model. J. Immunol. 162:2530.
Hendrickson, B. A., L. Rindisbacher, B. Corthesy, D. Kendall, D. A. Waltz, M. R. Neutra, J. G. Seidman. 1996. Lack of association of secretory component with IgA in J chain-deficient mice. J. Immunol. 157:750.
Shimada, S., M. Kawaguchi-Miyashita, A. Kushiro, T. Sato, M. Nanno, T. Sako, Y. Matsuoka, K. Sudo, Y. Tagawa, Y. Iwakura, M. Ohwaki. 1999. Generation of polymeric immunoglobulin receptor-deficient mouse with marked reduction of secretory IgA. J. Immunol. 163:5367.
Grezel, D., M. Capron, J. M. Grzych, J. Fontaine, J. P. Lecocq, A. Capron. 1993. Protective immunity induced in rat schistosomiasis by a single dose of the Sm28GST recombinant antigen: effector mechanisms involving IgE and IgA antibodies. Eur. J. Immunol. 23:454.
Giffroy, D., A. Langendries, M. Maurice, F. Daniel, B. Lardeux, P. J. Courtoy, J. P. Vaerman. 1998. In vivo stimulation of polymeric Ig receptor transcytosis by circulating polymeric IgA in rat liver. Int. Immunol. 10:347
Dent, L. A., M. Strath, A. L. Mellor, C. J. Sanderson. 1990. Eosinophilia in transgenic mice expressing interleukin 5. J. Exp. Med. 172:1425
Woerly, G., N. Roger, S. Loiseau, D. Dombrowicz, A. Capron, M. Capron. 1999. Expression of CD28 and CD86 by human eosinophils and role in the secretion of type 1 cytokines (interleukin 2 and interferon ): inhibition by immunoglobulin A complexes. J. Exp. Med. 190:487.
Huang, W. W., E. A. Garcia-Zepeda, A. Sauty, H. C. Oettgen, M. E. Rothenberg, A. D. Luster. 1998. Molecular and biological characterization of the murine leukotriene B4 receptor expressed on eosinophils. J. Exp. Med. 188:1063.
Dombrowicz, D., B. Quatannens, J. P. Papin, A. Capron, M. Capron. 2000. Expression of a functional FcRI on rat eosinophils and macrophages. J. Immunol. 165:1266.
Woerly, G., S. Loiseau, M. Loyens, C. Schoch, M. Capron. 2003. Inhibitory effects of ketotifen on eotaxin-dependent activation of eosinophils: consequences for allergic eye diseases. Allergy 58:397.
Ross, S. R., J. J. Schofield, C. J. Farr, M. Bucan. 2002. Mouse transferrin receptor 1 is the cell entry receptor for mouse mammary tumor virus. Proc. Natl. Acad. Sci. USA 99:12386.
Dombrowicz, D., M. Capron. 2001. Eosinophils, allergy and parasites. Curr. Opin. Immunol. 13:716.
Tomassini, M., A. Tsicopoulos, P. C. Tai, V. Gruart, A. B. Tonnel, L. Prin, A. Capron, M. Capron. 1991. Release of granule proteins by eosinophils from allergic and nonallergic patients with eosinophilia on immunoglobulin-dependent activation. J. Allergy Clin. Immunol. 88:365.
Gregory, B., A. Kirchem, S. Phipps, P. Gevaert, C. Pridgeon, S. M. Rankin, D. S. Robinson. 2003. Differential regulation of human eosinophil IL-3, IL-5, and GM-CSF receptor -chain expression by cytokines: IL-3, IL-5, and GM-CSF down-regulate IL-5 receptor expression with loss of IL-5 responsiveness, but up-regulate IL-3 receptor expression. J. Immunol. 170:5359.
Bartemes, K. R., S. McKinney, G. J. Gleich, H. Kita. 1999. Endogenous platelet-activating factor is critically involved in effector functions of eosinophils stimulated with IL-5 or IgG. J. Immunol. 162:2982.
Dunzendorfer, S., N. C. Kaneider, A. Kaser, E. Woell, J. M. Frade, M. Mellado, C. Martinez-Alonso, C. J. Wiedermann. 2001. Functional expression of chemokine receptor 2 by normal human eosinophils. J. Allergy Clin. Immunol. 108:581.
Nakajima, T., H. Yamada, M. Iikura, M. Miyamasu, S. Izumi, H. Shida, K. Ohta, T. Imai, O. Yoshie, M. Mochizuki, et al 1998. Intracellular localization and release of eotaxin from normal eosinophils. FEBS Lett. 434:226.
Abu-Ghazaleh, R. I., T. Fujisawa, J. Mestecky, R. A. Kyle, G. J. Gleich. 1989. IgA-induced eosinophil degranulation. J. Immunol. 142:2393.
Rifai, A., K. Fadden, S. L. Morrison, K. R. Chintalacharuvu. 2000. The N-glycans determine the differential blood clearance and hepatic uptake of human immunoglobulin (Ig)A1 and IgA2 isotypes. J. Exp. Med. 191:2171.
Dotzauer, A., U. Gebhardt, K. Bieback, U. Gottke, A. Kracke, J. Mages, S. M. Lemon, A. Vallbracht. 2000. Hepatitis A virus-specific immunoglobulin A mediates infection of hepatocytes with hepatitis A virus via the asialoglycoprotein receptor. J. Virol. 74:10950.
Olszewska-Pazdrak, B., K. Pazdrak, P. L. Ogra, R. P. Garofalo. 1998. Respiratory syncytial virus-infected pulmonary epithelial cells induce eosinophil degranulation by a CD18-mediated mechanism. J. Immunol. 160:4889.(Véronique Decot, Ga?tane )