Human eosinophils express CR1 and CR3 complement receptors for cleavage fragments of C3

CELLULAR

IMMUNOLOGY

97,291-306 (1986)

Human Eosinophils Express CR1 and CR3 Complement Receptors for Cleavage Fragments of C3’ ELIZABETHFISCHER,MONIQUE CAPRON,LIONELPRIN, JEANPIERREKUSNIERZ,AND MICHEL D. KAZATCHKINE INSERM U28, Hopital Broussais, Paris, and Centre d’lmmunologie et de Biologie Parasitaire, Unite mixte INSERM U167-CNRS 624. Lille, France Received July 25, 1985; acceptedSeptember 29, I985 The functional and antigenic characteristics of C3 receptors expressedon human eosinophils were investigated using rosette assayswith sheeperythrocytes coated with C3 fragments and flow cytometric analysis of cells stained with anti-receptor antibodies. Purified peripheral blood eosinophils from 13 patients with hypereosinophilia expressedCR 1 antigens. In 8 patients, a mean of 14 + 9.5% eosinophils formed C3bdependent rosettes that were inhibited by F(ab’)r anti-CR1 antibodies. This number increased to 33% following stimulation with leukotriene B4 (LTB4) (IO-’ M). Similar numbers of C3b rosetteswere formed by hypodenseand normodenseeosinophils. Eosinophils from 2 patients from this group expressed20,000 ‘ZSI-labeledmonoclonal anti-CR 1 antibody binding sites/cell. In another group of patients, 55 -t 9% eosinophils spontaneously formed C3bdependent rosettesthat could not be enhanced by LTB4. In all patients, a mean of 16 + 9% eosinophils formed cation-dependent rosetteswith C3bibearing intermediates that were inhibited by anti-CR3 antibody OKM 1. All eosinophils stained with monoclonal antibodies against the (Ychain of CR3. There was no C3ddependent rosette formation with eosinophils and no eosinophils stained with monoclonal anti-CR2 antibody. Thus, human eosinophils express CR1 and CR3. SinceCR3 is required for the adhesionofgranulocytes to surfacesand antibody-dependent cellular cytotoxicity of neutrophils, the interaction of C3 fragments with CR3 and CR1 on eosinophils may be of importance in eosinophil-mediated damage of opsonized targets. o 1986 Academic

Pm,

Inc.

INTRODUCTION Human leukocytes express three types of C3 receptors, CRl’, CR2, and CR3, for the C3b, C3bi, and C3dg/C3d fragments of C3 that are covalently attached to complement activators ( 1). The C3b receptor, CR1 , is a polymorphic glycoprotein of ’ This work was supported by Institut National de la Sante et de la Recherche Medicale (INSERM), France. * Abbreviations used: CR 1, C3b receptor; CR2, C3d receptor; CR3, C3bi receptor; E’, sheeperythrocytes; Er3b, C3b-bearing E*; E’C3bi, C3bibearing E’; ET3dg, C3dg-bearing E’; E”CM, C3dbearing Es; E’, Es treated with pyruvic aldehyde; BSA, bovine serum albumin; FITC, fluorescein isothiocyanate conjugate; HBSS, Hanks’ balanced salt solution; HBSS++,HBSS containing 0.6 W Ca*’ and 0.9 m&f Me; LTB4, leukotriene B4; PBS, phosphate-buffered saline; PBMC, peripheral blood mononuclear cells; PMN, polymorphonuclear leukocytes; VBS, Veronal-buffered saline; GVB, VBS containing 0.1% gelatin; DGVB++, half isotonic GVB with 2.5% dextrose containing 0. I5 mMcalcium and 0.5 mM magnesium; GVB-EDTA, GVB containing 0.04 M ethylenediaminetetraacetate. 291 0008-8749186$3.00 Copyright 0 I986 by Academic Press, Inc. All rigbu of reproduction in any form resxved.

298

FISCHER

ET AL.

160,000-250,000 it4, present on human erythrocytes, neutrophils, monocytes, B lymphocytes, a subpopulation of T lymphocytes, follicular dendritic cells, and glomerular podocytes (2-6). CR1 strongly binds surface-bound C3b and may also bind C3bi and C4b with lesser affinity. CR2, the C3d receptor, is a glycoprotein of M, 140,000145,000 expressedon B lymphocytes, human B-cell lines, and follicular dendritic cells that mediates the binding of particles bearing C3dg/C3d and, more weakly, that of surface-fixed C3bi and C3b (6-9). CR3 consists of two polypeptide chains, an (Ychain of M, 165,000 involved in leukocyte adhesion to surfaces(10, 11) and a @subunit of M, 95,000 that is common to the other leukocyte surface molecules LFA 1 and p 150,95 (12). CR3 binds surface-bound C3bi in a cation-dependent reaction; it is present on neutrophils, monocytes, follicular dendritic cells, and null cells (6, 10, 13, 14). Expression of CR1 and CR3 antigens on neutrophils is upregulated at 37°C and following stimulation of the cells with chemotactic factors (15, 16). Human eosinophils expressFc receptorsspecific for IgG and for IgE that are involved in the ability of the cells to kill antibody-coated parasitic targets( 17, 18). &E-dependent cytotoxicity of eosinophils is restricted to cells of lower density (“hypodense” eosinophils) present in the blood of highly hypereosinophilic patients ( 18). Eosinophils also express C3 receptors that have first been characterized by the capacity of the cells to form rosetteswith sheeperythrocytes (E”) coated with C3b, C4b, and C3d ( 19-2 1). In the present study, human eosinophils were characterized as only expressing CR1 and CR3 by functional and immunohistochemical criteria using rosette assayswith E” coated with defined C3 fragments and monoclonal antibodies against antigenic determinants of CRl, CR2, and CR3. MATERIALS

AND METHODS

Bufers and reagents.Veronal-buffered saline (VBS), VBS containing 0.1% gelatin (GVB), half isotonic GVB with 2.5% dextrose containing 0.15 mM Ca” and 0.5 mM Mg2+ (DGVB++), and GVB containing 0.04 M EDTA (GVB-EDTA) were prepared as described (22). Hanks’ balanced salt solution lacking Ca2+and Mg2+ (HBSS; Gibco BioCult Ltd, Paisley, Scotland) and HBSS that was made 0.6 m&f Ca2’ and 0.9 mM Mg+ (HBSS++)were used in rosette experiments. Leukotriene B4 (LTB4) was obtained from Dr J. Rokach (Merck Frost Laboratories, Quebec, Canada). Isolation of human peripheral blood leukocytes.Human eosinophils were purified from whole blood of 13 patients with hypereosinophilia of various etiologies by dextran sedimentation of erythrocytes and centrifugation of leukocytes on discontinuous metrizamide gradients (23). Cell fractions from the top of the 25 and 24% metrizamide layers containing 80 to 95% pure eosinophils (“normodense” eosinophils) were pooled, washed in HBSS, and adjusted to 5 X 106/ml in this buffer. In some experiments, “hypodense” eosinophils were collected from low-density layers (i.e., less than 23% me&amide) from highly hypereosinophilic patients (24). Polymorphonuclear leucocytes (PMN) and peripheral blood mononuclear cells (PBMC) were separatedfrom heparinized blood of normal donors by dextran sedimentation of erythrocytes and centrifugation on Ficoll-Hypaque cushions (Pharmacia Fine Chemicals, Uppsala, Sweden) (25) at room temperature. PMN and PBMC were washed in HBSS and suspended to 10 X 106/ml for rosette assays. Cellular intermediates bearing C3 fragments. Human C3 (26), B (27), D (28), P (29), and I (31) were purified to homogeneity from normal plasma as described.

HUMAN

EOSINOPHILS EXPRESS CR1 AND CR3

299

C3b receptor (CRl) was purified from human erythrocyte membranes (32). E”bearing C3b (EsC3b) were prepared by sequential deposition of C3b by fluid phase and cellbound amplification convertases using purified C3, B, and D as described (33). E”C3b carried approximately 40,000 C3b molecules/cell as assessedin binding experiments using radiolabeled monoclonal anti-C3 antibody (Bethesda Research Laboratories, Inc., Gaithesburg, Md.). EsC3bi were prepared by treating 1 X lo8 E”C3b with 50 pg H and 3 pg I in 1.Oml GVB++ for 1 hr at 37°C. ESC3dgwere obtained by incubating 1 X lo8 EsC3b with 0.8 gg CR1 and 4 gg I in 1.0 ml DGVB++ for 30 min at 37°C. E”C3d were obtained by treating EsC3bi with trypsin (2 &lo* cells; type III from bovine pancreas,Sigma Chem&al Co., St Louis, MO.) asdescribed(34). EsC3bi,EsC3dg, and ESC3dcould not form C3b,Bb,P amplification C3 convertase sitesupon incubation with B, D, and P, whereas C3b,Bb,P sites could be formed on E”C3b. E”C3b, E”C3bi, ESC3dg,and ESC3dwere agglutinated with rabbit antiC3d antibodies (Dutch Red Cross, Amsterdam, The Netherlands), whereas only E”C3b and EsC3bi were agglutinated with polyclonal anti-C3 antibodies (Behringwerke, Marburglahn, FRG). Cellular intermediates were stored in GVB-EDTA at 4°C and used within 2 weeks. Rosette assaysfir complement receptors. One volume (30- 100 ~1) of C3-bearing intermediates at 1 X lO’/ml in HBSS or HBSS++ was added to an equal volume of HBSS or HBSS++containing eosinophils, PMN, or PBMC (5 X 106/ml) in a 10 X 75 mm plastic tube. After centrifugation (125g, 5 min at 2O”C), the sedimented cells were incubated for 30 min at 37’C. The percentage of 100-300 leukocytes to which three or more red cells were attached was then assessedmicroscopically. Rosette assaysfor IgG and IgE receptors on eosinophils. A rosette assay using aldehyde-fixed sheep erythrocytes (Es) coated with human IgG (P.S. myeloma protein kindly donated by Dr H. L. Spiegelberg,La Jolla, Calif.) normal IgG, or bovine serum albumin (BSA; Sigma) was performed according to previous papers (17, 18). The percentage of rosette-forming eosinophils was evaluated on cell suspensions stained with Discombe’s diluent (7). Antibodies against C3 receptors.Rabbit polyclonal anti-CR1 antibodies and mouse monoclonal anti-CR1 antibody J3D3 were prepared asdescribed(2 1). The IgG fraction from rabbit anti-CR1 antiserum was prepared by ammonium sulfate precipitation and chromatography on DEAE cellulose. F(ab’)* fragments were obtained by pepsin digestion of the IgG fraction and chromatography on Sephadex G- 150 (Pharmacia) (35). IgG was purified from J3D3 ascites by ammonium sulfate precipitation and sequential chromatography on DEAE Trisacryl M (IBF, Villeneuve la Garenne, France) and CM Trisacryl (IBF). OKMl, a mouse monoclonal antibody that recognizes the (Ychain of CR3 (13) was purchased from Ortho Pharmaceuticals, (Raritan, N.J.). AntiMO 1, a monoclonal antibody also directed against CR3 (IO), was a kind gift from Dr. A. Amaout (Boston, Mass.). Anti-B2, a mouse monoclonal IgM against CR2 (8) was obtained from Coultronics, Margency, France. Flow cytofluorographic analysis of C3 receptorson eosinophils. Eosinophils (5 X lo5 in 100 ~1HBSS) were incubated with J3D3, OKM 1, anti-MO 1, or anti-B2 antibodies at saturating concentrations in 100 ~1HBSS for 30 min at 4°C. The cells were washed two times in cold HBSS and resuspended in 100 ~1 of the same buffer containing appropriate amounts of fluoresce&conjugated (FITC) rabbit F(ab’)* anti-mouse IgG antibodies (Cappel, Cochranville, Pa.) or FITC goat anti-mouse p chain antibodies (Cappel). After incubation for 30 min at 4’C, the cells were washed two times in HBSS and resuspended in 50 ~1 cold HBSS containing 1% pamformaldehyde. Cell

300

FISCHER ET AL.

surface fluorescence was analyzed by flow cytometry using an EPIC S.C. cytofluorograph (Coultronics). Eosinophils that were incubated with FITC F(ab’)* anti-mouse IgG or anti-mouse p chains alone were used as controls to determine background staining. Binding of radiolabeled monoclonal anti-CR1 antibody to CR1 on eosinophils. Monoclonal anti-CR1 antibody J3D3 was radiolabeled with 125I(Amersham, Les Ulis, France) to a specific activity of 250,000-450,000 cpm/pg using iodogen (Pierce Chemicals, Rockford, Ill.). For equilibrium binding studies, lOO-~1aliquots of phosphatebuffered saline ,(PBS), pH 7.2, containing 4 X lo6 eosinophils were incubated with 100 ~1PBS containing 0.005-0.45 pg ‘*‘I-J3D3 for 60 min at 4°C. Duplicate samples of cells (75 ~1) were layered on 0.25 ml of dibutylphthalate (Merck AG, Schuchardt, FRG) and centrifuged at 8OOOgfor 1 min. The tubes were cut and radioactivity in the pellets was assessedin an LKB minigamma counter (LKB, Orsay, France). Nonspecific binding of ‘251-J3D3to eosinophils was determined by incubating the cells with I*% J3D3 in the presence of a 50-fold molar excessof unlabeled antibody. The binding data were analyzed according to Scatchard (36). RESULTS Rosette Formation by Eosinophils The capacity of eosinophils from hypereosinophilic patients (with peripheral blood eosinophil counts ranging from 0.4 to 8 X log/liter) to form rosettes with E”C3b, EsC3bi, ESC3dg,and ESC3dwas examined at physiological ionic strength (Table 1). The capacity of PMN and PBMC from a normal individual to form rosettes with the same C3bearing intermediates was asses& in parallel control experiments. The mean percentageof eosinophils forming C3b rosetteswas 30 +- 23. Patients could be separated into two distinct groups in which averagesof 14 + 9.5 and 55 f 9% eosinophils formed rosettes with E”C3b. Group I included eight patients among which were two patients with drug hypersensitivity, three patients with hypereosinophilic syndrome, one patient TABLE I Rosette Formation between C3-Bearing E and Human Eosinophils from Hypereosinophilic Patients” % Rosettes Cellular intermediate

Buffer

Eosinophils* (‘7%rosettes + 1 SD)

PMN

PBMC

E’C3b

HBSS

14 + 9.5 (8) 55 f 9.0 (5)

70

45

ET3bi

HBSS++ HBSS

16 + 9.0 (7) 1.8 -I 0.7 (3)

61 0

42 16

E”c3dg

HBSS++ HBSS

14 + 5.0 (3) 0.3 + 0.5 (3)

23 0

36 10

E’CM

HBSS

2.0 + 0.6 (3)

0

10

0 Rosette formation with polymorphonuclear leukocytes (PMN) and peripheral blood mononuclear cells (PBMC) from a normal individual was amessedin parallel experiments. * Parenthesesindicate the number of patients tested.

HUMAN

EOSINOPHILS EXPRESS CR1 AND CR3

301

with intrinsic asthma, one patient with neoplasia, and one patient with alcoholic cirrhosis. Group II included five patients with onchocerciacis (2) and eosinophilic pleural effusion (3). Eosinophils from five patients from group I and from four patients from group II were incubated in HBSS alone or containing lo-’ or 10d8M LTB4 for 1 hr at 37”C, washedtwice in HBSS, and examined for their capacity to form E”C3b rosettes. The mean percentage of eosinophils forming C3b rosettes from patients in group I was enhanced from 16 to 33% with a relative increase of 145 + 73% (1 SD) following incubation with lo-’ M LTB4 whereas it decreasedfrom 55 to 29% with a relative decreaseby 49 k 2 1%(1 SD) for eosinophils from patients in group II (Fig. 1). Optimal enhancement of C3b-dependent rosette formation in group I was observed with LTB4 at lo-’ M; no LTBCmediated increase in C3b rosettes was seen in group II. In two highly hypereosinophilic patients from group I, normodense eosinophils were separated from hypodense eosinophils. No difference was observed in the mean percentage of C3b-rosette-forming cells between the two eosinophil subpopulations. Seventy percent PMN and forty-five percent PBMC from the control individual formed rosettes with E”C3b. Preincubation of eosinophils (5 X 10’ cells) from three patients with 0.5-2.0 pg of F(ab’)zpolyclonal anti-CR1 antibody for 30 min at 37°C suppressedthe capacity of the cells to form rosettes with ESC3bin a dose-dependent manner (Fig. 2). Eosinophils from sevenpatients were tested for their ability to form FC3bi rosettes. The mean percentage of cells forming EsC3bi rosettes in the presence of Ca” and Mg2+ was 16 + 9 and was independent, for each patient, of the relative C3b-dependent rosette-forming capacity of the cells. No C3bi-dependent rosettes with eosinophils were formed in the absenceof divalent cations. Sixty-one percent PMN and forty-two percent PBMC from the control individual formed EsC3birosettesin HBSS++whereas no PMN and 16% PBMC formed E”C3bi rosettes in HBSS (Table 1). Preincubation of eosinophils with OKMl suppressed the capacity of eosinophils to form cationdependent EsC3bi rosettes (Fig. 2). A mean of 14 -t 5% eosinophils formed rosettes with E”C3dg in HBSW. No significant binding of ESC3dgand of E”C3d to eosinophils was observed in the absence of cations. Twenty-three percent PMN and thirty-six percent PBMC formed rosettes

FIG. 1. Effect of LTB4 (IO-’ M) on the capacity of eosinophils from hypemosinophilic patients to form rosettes with E’C3b. The C3b rosette-forming capacity of the cells of each patient in the absenceof LTB4 was considered as being 100. Each bar represents the relative C3b rosette-forming capacity of eosinophils from each patient following incubation of the cells with LTB4 for 60 min at 37°C. Open bars: patients from group I (mean spontaneous C3b rosette formation 16 + 11%). Hatched bars: patient from group II (mean spontaneous C3b rosette formation 55 + 14%).

302

FISCHER ET AL. EsC3b

0 Fhb’)p

pC3bi

rosettes

roIMnes

L b

0.5 0.7

1

2

anti -CRl(ClO/assay)

01

OKM

5

1 ~pl/easy)

FIG. 2. Inhibition of rosette formation between E’C3b and E’C3bi and human eosinophils from an hypcreosinopbilic patient by F(ab’k anti-CR1 rabbit antibodies and mouse monoclonal antibody to the chain of CR3, OKM 1.

with ESC3dgin HBSS++;no PMN and 10%PBMC formed EsC3dgand ESC3drosettes in HBSS (Table 1). The capacity of eosinophils to form rosettes with erythrocytes coated with human IgG or IgE was assessedin four of the patients whose cells had been examined for their rosette-forming capacity with all Cfbearing intermediates. The percentagesof rosette-forming eosinophils were 24 + 2 for IgE-bearing Es, 31 + 13 for IgG-bearing E”, and 8 f 3 for EScoated with BSA, the latter cells being used as controls.

Flow CytojluorographicAnalysis of EosinophilsStained with Anti-ReceptorAntibodies The expression of C3 receptor antigens by eosinophils from two hypereosinophilic patients was investigated by staining the cells with mouse monoclonal anti-receptor antibodies and subjecting the cells to flow cytometric analysis. Ninety-five percent of the cells in a purified eosinophil preparation stained with monoclonal anti-CR1 antibody J3D3; the histogram of immunofluorescence demonstrated a major peak containing approximately 80% of the cells and a minor peak of lower fluorescenceintensity that contained approximately 15%of the cells (Fig. 3). Ninety-five percent of the cells in the eosinophil preparations stained homogeneously with monoclonal anti-CR3 antibodies OKMl and anti-MO1 (Fig. 3). No staining of eosinophils was observed with monoclonal anti-CR2 antibody anti-B2.

Enumeration of C3b Receptorson Eosinophil The number of C3b receptors expressedby eosinophils was assessedby examining the binding of radiolabeled monoclonal anti-CR 1 antibody J3D3 to cells purified from two hypereosinophilic patients from group I. Scatchard analysis of the binding data obtained using cells that had been prepared at room temperature demonstrated that eosinophils expressed 18,000-2 1,000 monoclonal anti-CR 1 binding sites/cell and that the antibody bound to the receptor with a mean affinity of 6.3 X 10’ M-’ (Fig. 4).

HUMAN

303

EOSINOPHILS EXPRESS CR1 AND CR3

FIG. 3. Expression of CR1 antigen (A) and Mol antigen (B) by eosinophils as assessedby indirect immunofluorescenceand flow cytometry. Cells were stained with monoclonal anti-CR I or anti-MO1antibodies and were then reacted with FITC (F(ab’h anti-mouse IgCi. Histograms demonstrate relative fluorescence intensity (log) on the x axis and cell number on the y axis. Panel C depicts staining with FITC (Fab’h antimouse IgG.

Normodense and hypodense eosinophils were isolated in one patient and examined separately for their capacity to bind ‘251-J3D3.No difference in the calculated number of CR1 antigenic sites was found between the two types of cells (17,600 and 18,000).

4

10 ‘251-J3D3

input

(ng/4x106ceIIS)

1251-J3D3

molecules

20 boundxIO~%cell

FIG. 4. Enumeration of C3b receptors on eosinophils. Eosinophils from a hypereosinophilic patient were incubated with increasing amounts of radiolabeled monoclonal anti-CR1 antibody J3D3 for 1 hr at 4°C. Nonspecific binding was assessedin parallel experiments performed in the presenceof a 50-fold molar excess of cold antibody. Specific binding of ‘251-J3D3to eosinophils is shown in the left panel. Scatchard analysis of the binding data obtained in this experiment revealed the presenceof 2 1,000 sites/cell and an affinity of the antibody for CR1 of 5.5 X 10’ M-’ (right panel).

304

FISCHER

ET AL.

When cells from one patient were preincubated for 1 hr at 37°C before incubation with I*?-J3D3, the mean number of J3D3 binding sites was found to be 37,OOO/cell. DISCUSSION Using rosette assayswith sheep erythrocytes (ES)bearing C3 fragments and flow cytometric analysis of cells stained with monoclonal antibodies directed against human C3 receptors, human eosinophils were found only to expressCR1 and CR3 receptors for cleavage fragments of C3. Eosinophils from normal and hypereosinophilic individuals have previously been demonstrated to form rosettes with ES coated with rabbit IgM and human C3b (EAC4b3b) and ESbearing large amounts of C4b (19,20). The rosette-forming capacity of eosinophils from hypereosinophilic patients with EAC4b3b was found to be enhanced by ECF-A tetrapeptides (37) and LTB4 (38). In the present study, eosinophils purified from whole blood of hypereosinophilic patients bound Esto which C3b molecules had been covalently fixed in the absenceof antibody using purified alternative pathway proteins. Patients were found to be heterogeneous with regard to the C3brosette-forming capacity of eosinophils and susceptibility to LTBCinduced enhancement of eosinophil C3b rosette formation. Thus, in some patients (Group I), an average of 16% peripheral blood eosinophils spontaneously formed rosettes with ESbearing 40,000 C3b/cell and an average of 33% eosinophils formed C3b rosettes following stimulation with lo-’ M LTB4. This could be due to an increase in the number of CR1 molecules expressed per cell, an increased number of cells bearing sufficient amounts of CR1 in order to form rosettes, or an increase in affinity of the receptor for C3b. Eosinophils from another group of patients that included patients with parasitic diseases(group II) appeared as being upregulated in vivo with regard to C3b receptor expression, since the cells spontaneously formed 55% C3b rosettes. Eosinophils from hypereosinophilic patients and eosinophils from Schistosoma mansoni-infected rats were previously shown to form significantly higher numbers of Fc t receptor-dependent rosettes (17). No enhancement in C3bdependent rosettes but a decreasein rosette formation was found with eosinophils from patients of group II in the presence of LTB4. It is possible that LTB4 causes ligand-independent inhibition of membrane C3b receptor expression on cells expressinglarge numbers of receptors in an analogous manner to l&and-independent regulation of CR 1 expression on neutrophils by phorbol my&ate acetate (39). By indirect immunofluorescence, 95% cells in an eosinophil preparation containing more than 80% eosinophils stained for CR1 antigen. The histogram of fluorescence revealed a major peak containing 80% of the cells expressing CR1 and a peak of cells exhibiting less fluorescence intensity. It is unclear whether the latter peak represents contaminating PMN in the eosinophil preparation or a subpopulation of eosinophils expressinglessCR 1 antigen. Preincubation of eosinophils with F(ab’)* anti-CR1 antibody suppressedthe C3bdependent rosette-forming capacity of the cells. The mean number of CR1 antigenic sites on eosinophils was calculated to be approximately 20,00O/cell in binding experiments using ‘251-labeledmonoclonal anti-CR1 antibody. This antibody recognizes a mean of 700 sites on human erythrocytes which indicates that one molecule of monoclonal antibody binds to one CR1 functional site (32). Hypodense and normodense eosinophils expressedsimilar numbers of CR1 sites/cell and had a similar capacity to form rosettes with ESC3b.Thus, cells which differ in their number of Fc receptors for IgE and in their capacity to mediate

HUMAN

EOSINOPHILS

EXPRESS CR1 AND

CR3

305

&E-dependent killing of 5’. munsoni (17) expresssimilar numbers of CR1 molecules. CR1 had previously been shown not to be required for &E-dependent cytotoxicity of schistosomesby hypodense human eosinophils ( 18). Human eosinophils also express C3bi (CR3) receptors since the cells were found to bind particles bearing C3bi in a cation-dependent manner (40) and express the Mol and OKMl antigens of the a! chain of CR3. Using ESbearing approximately 40,000 C3bi molecules/cell, an average of 16% eosinophils formed rosettes with EsC3bi in the presence of cations at physiological ionic strength, whereas no rosetting occurred in the absenceof cations. Cationdependent binding of ESC3dgto eosinophils was likely to be mediated by CR3 or by a distinct receptor similar to that recently identified on neutrophils (41), rather than by CR2, since no binding of C3dg was observed in the absence of cations, no C3d rosettes were formed with eosinophils, and no expression of CR2 antigen was found on the cells by indirect immunofluorescence, Eosinophils behaved in a similar fashion to PMN with regard to their capacity to bind C3bi and C3dg. Rosette formation between PBMC and ESC3dgin the absence of cations (10%) was mediated by CR2 since a similar number of PMBC formed rosetteswith E”C3d. C3bidependent rosettes of eosinophils were inhibited by OKMl, as previously reported for CR3-dependent rosette formation of U937 cells with E IgM coated with mouse complement (11). By indirect immunofluorescence, all cells in cell preparations containing more than 80% eosinophils stained homogeneously for MO 1 and OKM 1. The fact that a small percentage of eosinophils formed CR3- and CRl-dependent rosettes whereas almost all the cells expressedCR3 and CR1 antigens may be due to the relatively low number of ligand molecules/indicator cells (42) used in rosette assays. Becausethe MO 1-Mac 1-OKM 1 molecule is essential for the adhesive properties of granulocytes to surface (10) and becausebound C3bi increasesantibody-mediated cytotoxicity of lymphocytes and PMN for targets bearing IgG (43,44), the interaction of CR3 with C3bi could be of importance in facilitating adherence and killing of opsonized parasites by eosinophils in the presenceof specific antibody. Furthermore, since expression of C3bi receptors is upregulated on PMN by the samestimuli as CR 1 expression ( 16), enhancement of CR3 expression may probably occur on eosinophils which may prove to be essential in mediating eosinophil-dependent damageto targets in the inflammatory process. ACKNOWLEDGMENTS The contribution of P. Debre to flow cytometric experiments, the technical assistanceof J. P. Papin, J. L. Neyrinck, and C. Blanc, and the secretarial assistanceof C. Sauvion are gratefully acknowledged.

REFERENCES 1. 2. 3. 4. 5. 6. 7. 8. 9.

Fearon, D. T., and Wong, W. W., Annu. Rev. Immunol. 1, 243, 1983. Fearon, D. T., J. Exp. Med. 152,20, 1980. Tedder, T. F., Fearon, D. T., Gartland, G. L., and Cooper, M. D., J. Immunol. 130, 1668, 1983. Wilson, J. G., Tedder, T. F., and Fearon, D. T., J. Immunol. 131,684, 1983. Kazatchkine, M. D., Fearon, D. T., Appay, M. D., Mandet, C., and Bariety J., J. Clin. Invest. 69,900, 1982. Reynes, M., Aubert, J. P., Cohen, J. H. M., Audouin, J., Tricottet, V., Diebold, J., and Kazatchkine, M. D., J. Immunol. 135,2687, 1985. Eden, A., Miller, G. W., and Nussenzweig, V., J. C/in. Invest. 52, 3239, 1973. Iida, K., Nadler, G. W., and Nuzzenzweig, V., J. Exp. Med. 158, 1021, 1983. Weis, J. J., Tedder, T. F., and Fearon, D. T., Proc. Natl. Acad. Sci. USA 1,243, 1983.

306

FISCHER ET AL.

10. Amaout, M. A., Todd, R. F., III, Dana, N., Melamed, J., Schlossman,S. F., and Colten, H. R., .I. Clin. Invest. 72, 171, 1983. 11. Sanchez-Madrid, F., Nagi, J. A., Robbins, E., Simon, P., and Springer, T. A., J. Exp. Med. 158, 1785, 1983. 12. Ross, G. D., Cain, J. A., and Lachmann, P. J., J. Immunol. 134, 3307, 1985. 13. Wright, S. D., Rao, P. E., Van Voorhis, W. C., Craigmyle, L. S., Iida, K., Talle, M. A., Westberg, E. F., Goldstein, G., and Silverstein, S. C., Proc. Natl. Acad. Sci. USA 80, 5699, 1983. 14. Ault, K. A., and Springer, T. A., J. Immunol. 126, 359, 1981. 15. Fearon, D. T., and Collins, L. A., J. Immunol. 130,370, 1983. 16. Berger, M., O’Shea, J., Cross, A. S., Folks, T. M., Chused, T. M., Brown, E. J., and Frank, M. M., J. Clin. Invest. 74, 1566, 1984. 17. Capron, M., Capron, A., Dessaint, J. P., Torpier, G., Johansson, S. G. O., and Prin, L., J. Immunol. 126,2087,

1981.

18. Capron, M., Spiegelberg, H. S., Prin, L., Bennich, H., Butterworth, A. E., Pierce, R. J., Ouassi, M. A., and Capron, A., J. Immunol. 132,462, 1984. 19. Gupta, S., Ross, G. D., Good, R. A., and Siegal, F. P., Blood. 48, 755, 1976. 20. Anwar, A. R. E., and Kay, A. B., J Immunol. 119,976, 1977. 21. Winquist, I., Olofsson, T., Olsson, I., Persson,A. M., and Hallberg, T., Immunology 47, 531, 1982. 22. Nelson, R. A., Jensen,J., Gigli, I., and Tamura, N., Immunochemistry 3, 1I, 1966. 23. Vadas, M. A., David, J. R., Butterworth, A. E., Pisani, N. T., and Siongok, A., J. Immunol. 122, 1228, 1979.

24. Prin, L., Capron, M., Tonnel, A. B., Bletry, O., and Capron, A., Int. Arch. Allergy Appl. Immunol. 72, 336, 1983. 25. Boyum, A., &and. J. Clin. Lab. Invest. Zl(Supp1. 91), 31, 1968. 26. Tack, B. F., and Prahl, J. W., Biochemistry 15,4513, 1976. 27. Hunsicker, L. G., Ruddy, S., and Austen, K. F., J. Immunol. 110, 113, 1979. 28. Fearon, D. T., and Austen, K. F., J. Exp. Med. 142, 856, 1975. 29, Fearon, D. T., and Austen, K. F., Proc. Natl. Acad. Sci. USA 74, 1683, 1677. 30. Jouvin, M. H., Kazatchkine, M. D., Cahour, A., and Bernard, N., J. Immunol. 133, 3250, 1984. 31. Fearon,D. T., J. Immunol. 119, 1248, 1977. 32. Cook, J., Fischer, E., Boucheix, C., Mirsrahi, M., Jouvin, M. H., Weiss,L., Jack, R. M., and Kazatchkine, M. D., Mol. Immunol. 22, 531, 1985. 33. Fischer, E., and Kazatchkine, M. D., J. Immunol 130,2821, 1983. 34. Lachmann, P. J., Pangbum, M. K., and Olroyd, R. G., J. Exp. Med. 156,205, 1982. 35. Nisonoff, A., Methods. Med. Res. 10, 134, 1964. 36. Scatchard, G., Ann. N. Y. Acad. Sci. 51,660, 1949. 37. Anwar, A. R. E., and Kay, A. B., J. Immunol. 121, 1245, 1978. 38. Nagy, L., Lee, T. H., Goetzl, E. J., Picket& W. C., and Kay, A. B., Clin. Exp. Immunoi. 47,541, 1982. 39. Jack, R. M., Changelian, P. S., and Fearon, D. T., Complement 1, 172 (Abstract). 40. Ross,G. D., Newman, S. L., Lambris, J. D., Devery-Pocius, J. E., Cain, J. A., and Lachmann, P. J., J. Exp. Med. 158,334, 1983. 41. Vik, D. P., and Fearon, D. T., J. Immunol. 134,2571, 1985. 42. Gaither, T. A., Magrath, I. T., Berger, M., Hammer, C. H., Novikovs, L., Santaella, M., and Frank, M. M., J. Immunol. 131, 899, 1983. 43. Perlmann, H., Perlmann, P., Schreiber, R. D., and Muller-Eberhard, H. J., J. Exp. Med. 153, 1592, 1981. 44. Kohl, S., Springer, T. A., Schmalstieg, F. C., Loo, L. S., and Anderson, D. C., J. Immunol. 133,2972, 1984.