Eosinophil as antigen-presenting cell
Eur. J. Immunol. 1992. 22: 1919-1925
Victoria Del Pozo+, Belen De AndrQs+, Elena MartinA, Blanca Cardaba., Julio Cesar Fernandez, Soledad Gallardo, Paloma Tramon, Francisco Leyva-Cobian., Pilar Palomino and Carlos Lahoz Department of Immunology, Fundacion Jimknez Diaz, Madrid and Department of Immunologym, Hospital Universitario “MarquCs de Valdecilla”, Santander
Eosinophil as antigen-presenting cell: activation of T cell clones and T cell hybridoma by eosinophils after antigen processing* We have studied the role of murine eosinophils as antigen-presenting cells (APC). Eosinophils have several characteristics that support the hypothesis of its function as potential APC: they have phagocytic capacity, express adhesion molecules and major histocompatibility complex (MHC) class I1 antigens and can produce and release interleukin-1 (IL-1). We have obtained several T cell clones specific for Mesocestoides corti antigens and used T cell hybridoma specific for ovalbumin (OVA) to test this hypothesis. Granulocyte-macrophage colony-stimulating factor-activated pure eosinophils (99.9 %), express class I1 antigens and are able to present M . corti antigens to specific T cell clones or OVA to T cell hybridoma 3 D 0 11.10, inducing the proliferation of Tcell clones and IL-2 release by the T cell hybridoma. Proliferation of T cells clones is dependent on the number of eosinophils used as APC. We have compared the efficiency of the same number of macrophages and eosinophils as APC, and have found that macrophages are more efficient than eosinophils. Lysosomotropic agents, such as chloroquine and ammonium chloride, that inhibit antigen processing, impaired eosinophil presentation. This presentation is restricted by MHC class I1 and inhibited by anti-I-Ad monoclonal antibody. The present study provides clear evidence of APC function for eosinophils. Our investigation points to a new role for eosinophils in the immune response.
1 Introduction MHC class I1 expression is essential for antigen presentation by APC. Other properties such as cytokine production [l,21, the capacity to process antigen [3] and expression of adhesion molecules [4] can enhance the efficiency of antigen presentation. The prototype APC in the immune system is the macrophage. However, many other cellular types that express class I1 determinants are able to function as APC: Langerhans cells, dendritic cells, enterocytes, endothelial cells, B cells, and T cells [5-71. There is evidence supporting the conclusion that although Ia expression is necessary, it is not sufficient to ensure APC function [8, 91. Recent experiments show that liposomes bearing class I1 and membrane IL-1 are able to present antigen [lo, 111. Eosinophils are classically described as effector cells in hypersensitivity disorders [ 121 and parasitic infections [13]. Although some of their functions remain unclear, interesting characteristics have been recently described suggesting
I1 98921
*
A
This work has been supported by grants from “Fond0 de Investigacioncs Sanitarias de la Seguridad Social 8811742 and 9010238”. Recipients of thc “Conchita RBbago Fellowship” Supported by “Fondo de Investigaciones Sanitarias de la Seguridad Social”. Fellow from “Comunidad Autonoma de Madrid”
Correspondence: Carlos Lahoz, Dcpt. Immunology, Fundacion Jimtncz Diaz, Avd. Reyes Catcilicos 2, E-28040 Madrid, Spain Abbreviation:
1919
GM-CSF Granulocytc-macrophage CSF
0 VCH Verlagsgescllschaft mbH, D-6940 Weinheim, 1992
an important role of these cells in the immune response: eosinophils are phagocytic cells [14], express adhesion molecules that facilitate intercellular adhesion [151, can express class I1 MHC after treatment with granulocytemacrophage (GM)-CSF [16], are able to produce and release 1L-1 [17], and express the IL-2R pS.5 subunit [18]. All the characteristics described above suggest a possible role of the eosinophils as APC. Investigating this possibility, we confirmed class I1 expression by murine eosinophils and demonstrated their ability to interact with CD4 lymphocytes and present antigens in the context of MHC class I1 molecules.
2 Materials and methods 2.1 Animals, induction of peritoneal eosinophilia and eosinophil purification Four- to five-week-old BALB/c (H-2d) and CBAN (H-2k) female mice, were purchased from Iffa Credo EspaAa, (El Prat de Llobregat, Barcelona, Spain).
To induce peritoneal eosinophilia, BALB/c or CBA/J mice were given a single intraperitoneal injection of 70-100 Mesocestoides corti larvae as previously described [ 191. After peritoneal wash, cells were washed in PBS-0.1 % BSA-0.1 M EDTA, pH 7 (PBS-BSA-EDTA). Cells were stained with eosin dye and counted in a Fuchs-Rosenthal chamber. Contaminating cells (30 %-50 %) were mostly macrophages. Mouse eosinophils were purified by centrifugation in a discontinuous density gradient of Percoll [MI. Percoll was adjusted to isotonicity by adding 1 part of 0.5 M NaCl to
+
0014-2980/92/0707-1919$3.50 .25/0
1020
V. Del Pozo, B. de Andres. E. Martin et al.
9 parts Percoll. Four layers with different concentrations of Percoll in sterile PBS (1.5 ml of 68 % and 3 ml. of 66 %, 59 YOand 50 YO)were placed in 16 x 125 mm-plastic tubes (Costar 205, Cambridge, MA). A total of 3 x lo7- 4 x lo7 cells in 2 ml of PBS-BSA-EDTA, were placcd on top of the gradient. After centrifugation at 1800 x g for 18 min at room temperature, eosinophils were found between the layers of59 % and 66 % and the majority of macrophages on top of the gradient. Cell purity was evaluated by cytofluorymetry.
2.2 Tissue culture medium reagents and antigens Eosinophils were maintained in RPMI 1640 supplemented with 10 % FCS, 2 mM L-glutamine, 50 U/ml penicillin, M 2-ME (Flow LaboSO pg/ml streptomycin and 5 X ratories, Irvine, Scotland). Proliferation assays were done in the same medium. Chloroquine and paraformaldehyde were purchased from Sigma (St. Louis,-MO). For antigen extraction of M . corti, larvae from long-time infected-mice were washed with sterile PBS, freezed and thawed several times and sonicated for 10 min. The resulting extract was centrifuged at 12 000 x g for 30 min at 4 "C. Finally, supernatant was dialyzed against PBS overnight in the cold. Protein content was measured by Bradford's Method (Bio-Rad, Munchen, FRG). Chicken ovalbumin (OVA) was obtained from Sigma.
2.3 T cell clones and T cell hybridoma T cell clones were obtained by the method of Kimoto et al. [20] with slight modifications. Primed lymphocytes from lymph nodes (2 x lo6) of BALB/c mice infected with M . corti were cultured in 2 ml/well culture plates (Costar 3524) with 2 x lo6 irradiated syngeneic spleen cells and 100 pg/ml of M. coyti extract. After 4 to 6 days of stimulation, dead cells were removed by density gradient centrifugation over Lymphocyte Separation Medium (Flow Laboratories), and viable cells were plated without antigen or feeder cells for 7-8 days. T cells were then stimulated with antigen and irradiated syngeneic spleen cells. The cells were maintained in this cycle of alternative stimulation and rest. After the first cycle of stimulation, cells were cloned three times by limiting dilution (0.5 cells/well) in cultures containing: antigen, irradiated syngeneic spleen cells as a source of APC and 5 % mouse IL-2 prepared by stimulating EL4 thymoma cells with phorbol myristate acetate. T cell clone proliferation was assayed by culturing 1 X loJ T cells and 4 x 105 irradiated spleen cells with different antigen doses. Cultures were maintained in a total volume ot 200 p1 in flat-bottom wells (Costar 3596) for 4 days at 37 "C in 5 Y CO? humidified air, and ["H] thymidine (New England Nuclear, Boston, MA) was added for the last 18 h. Radiactivity incorporated was measured in a scintillation counter (Beckman LS 2800, Palo Alto, CA). The cloned, OVA-specific, H-2d-restricted T lymphocyte hybridoma 3 D 0 11.10 used in these experiments has described [21] and was kindly provided by Dr. A. Szabo, HBpital Cochin, Paris.
Eur. J. Irnmunol. 1992. 22: 1019-1925
2.4 Flow cytometric analysis 2.4.1 Eosinophils purity Eosinophils purity was evaluated incubating 3 x lo5 cells for 30 min at 4°C with IgGz rat anti-mouse macrophages mAb MAS034 (Sera-Lab, Ltd., Sussex, GB) and washed twice in cold PBS. Pellets were incubated for an additional 30 min with fluoresceinated anti-rat IgG (Sera-Lab). Finally, cells were washed and analyzed in a flow cytometer (Epics-C, Coulter Electronics, Hialeah, FL).
2.4.2 Class I1 expression Eosinophils (5 x lo5) were stained with fluorescein isothiocyanate-conjugated monoclonal antibody to I-Ad MHC (Beckton Dickinson, Mountainview, CA) and flow cytometry was performed.
2.4.3 Analysis of clones Tcells (1 x lo5) were incubated 30min on ice, with a biotin-conjugated monoclonal antibody against CD4 and FITC-conjugated monoclonal antibody against CD8 (SeraLab). After incubation, the cells were washed three times and incubated with 5 % mouse serum in PBS. After three washes, cells were incubated with phycoerythrin-avidin and analyzed on an Epics C (Coulter Electronics) equipped with a 488 nm argon laser (green) and with a short pass of 520 nm (red). Prior to two-color fluorescence analysis, negative as well as positive samples labeled with a single conjugated antibody (red or green) were analyzed. Fluorescein-conjugated anti-mouse B cells and anti-mouse macrophages antibody (Sera-Lab) were used for B cell and macrophage detection.
2.5 Antigen presentation to T cell clones Purified eosinophils (2 x lo5) in 100 pl of complete RPMI supplemented with 50 U/ml of recombinant murine GMCSF (Genzyme, Boston, MA) were plated for 120 min at 37°C in flat-bottom 96-well plates. In the last 90 min of incubation 100 pglml of the corresponding antigen were added. After three washes cells were fixed with 100 pl of 2 % paraformaldehyde (final concentration 1%) for 20 min at room temperature to avoid degranulation of the eosinophils. After three washes with PBS, 100 p1 of complete medium was added and fixed cells were incubated overnight at 4 "C for possible paraforinaldehyde leaching. Cells were washed three more times before the addition of 3 x lo4 - 4 x lo4 T cells to each well. Cell proliferation was measured as described previously.
2.6 Antigen presentation to T hybridoma 3 D 0 11.10 and IL-2 assays The assay for the stimulation of IL-2 production by antigen-specific hybridoma was performed as described [22].Briefly, 2 x 105APC, 1 x lo5 hybridomaT cells, and antigen at the appropriate concentrations were cultured in 0.3 ml complete medium in microculture plates for 24 h at
37°C. The presence of IL-2 in the culture supernatant was then assayed using the IL-Zdependent CTLL-2 cell line. 2.7 Treatment of eosinophils with lysosomotropic agents chloroquine and NH&l, and APC inhibition using anti-I-Aa
Chloroquine and NH4CI were freshly prepared in RPMI and used at the concentration of 100kM and 10 pM, respectively for 30 min at 37 "C. Treated cells were washed in medium three times, and processed as indicated previously for use as APC. Eosinophils (2 x 1O6/ml) pre-treated with GM-CSF were incubated with an appropriate concentration (62.5 pg antibody/ml) of dialyzed anti-mouse I-Ad antisera (Becton Dickinson) for 30 min at 37 "C. Alternatively anti-murine B cells antibody, under the same experimental conditions, was used a5 a negative control. Thereafter eosinophils were treated as previously described to be used as antigenpresenting cells.The viability after treatment with antibody was measured by the trypan blue exclusion and was > 95%.
the expression of MHC class I1 was analyzed. As shown in Table 1an increase of MHC class IT surface expression was found during the culture. In some experiments we detected a certain degree of class I1 expression at the beginning of culture which increased with GM-CSF treatment. Similar results have been found with human eosinophils where in one case 23.6 % of eosinophils expressed HLA-DR molecules without previous activation [l5]. 3.4 Proliferation of T cell clones using eosinophils as APC M . corti-specificT cell clones proliferate when added to the plates with eosinophils previously treated with GM-CSF and pulsed with antigen (100 pg/ml) as described in
Table 1. I-Ad expressiona) on eosinophils treated with GMCSFb) Timc
3.1 Eosinophil purification
Experiments were usually done with 99.9 % pure eosinophi1 preparation as tested by cytofluorimetry and cytocentrifuge preparations. In some instances, preparations with a purity ranging between 99 % to 95 % were employed, but in all theses cases the contaminating cells were not typical activated macrophages as determined by morphological and fluorocytometric analysis. Fig. 1 shows the cytotluorimetric analysk of a typical eosinophil preparation stained with anti-mouse macrophages antibody before (panel C) and after cell separation by Percoll gradient (panel B). Macrophages are absent in the final eosinophil preparation (panel B). Panel A shows the control background fluorescence using the second fluoresceinated antibody.
ESP. I
Ilxp.2 Enp.3 Iixp.4 Iixp.5 1ixp.h Eup.7 1:xp.X
(11)
(I
3 Results
1921
Eosinophil as antigen-presenting cell
Eur. J. lmmunol. 1992. 22: 1919-1925
2 24
0 5CJ 9.66 70.2
ND
28.7 38.3
I4 ND
28.1
SD
99.5
0.4
11.9
32.2 N1)
11.3 11.7
NI)
(1.0 KD 78.1
0.7 KIl 40
a) Eosinophils were stained with fluoresceinated monoclonal antibody anti I-Ad for I-A detection and analyzed by flow cytometry. b) Purified eosinophils were cultured for the indicated periods of time in the presence of 50 U/ml of recombinant GM-CSE c) Results are expressed as the percentage of positive cells.
3.2 Characterization of T cell clones
All three clones (2B1,4B I , and 2B2) used in this study were analyzed by flow cytometry. The phenotype of these cells was CD3+, CD4+, CDX-. These cells did not express any macrophage or B cell surface marker. These clones respond to M . corti antigens when presented on H-2d irradiated spleen cells. Fig. 2 shows the proliferation of these three T cell clones to different doses of antigen ranging from 1 to 1000 pg/ml of M . c o r k I n contrast, no proliferation was detected when an unrelated antigen (OVA) was used. 3.3 Expression Expression of of class class I1 I1 antigens antigens on on eosinophils eosinophils 3.3 In humans Lucey et al. [lS] have found that eosinophils maintained in culture with recombinant human GM-CSF express HLA-DR antigens. Eosinophils (2 X 106/ml)were cultured with SO U/ml of recombinant mouse GM-CSF, and
Figure 1. 1. Cytofluorometric Cytofluorometric analysis analysis of of eosinophil eosinophil preparations. preparations. In In Figure panel A A (background (background control), control),iiss shown shown the the cell cell population populationstained stained panel with fluoresceinated anti-rat Ig. Panel B: analysis of eosinophil preparation after Percoll gradient using mAb against mouse macrophages MAS034, as describedin Sect. 2.4.1. Panel C: Profile of the whole cell population before purification stained with anti-macrophage antibody. Five thousand cells were analyzed.
V. Del Pozo, B. de Andres, E. Martin et al.
1922
Sect. 2.5. Fig. 3 shows the specificT cell clone proliferation in response to antigen presented by eosinophils from syngeneic mice. Only background proliferation was detected using T cells alone (516 k 43 dpm), T cells with GM-CSF (495 k 224 dpm). eosinophils without GM-CSF (471 f 91 dpm) and T cells with soluble antigen (307 f 16 dpm) as controls. The differences between controls without antigen and the proliferating clones were statistically significant 0, < 0.05). If eosinophils were fixed with paraformaldehyde prior to pulsing with antigen. no proliferation was found (516 dpm k 162). We wanted to compare the efficiency of different number of macrophages and eosinophils, similarly treated with GMCSF, for antigen presentation. Fig. 4A shows T cell proliferation induced by different amounts of APC. The proliferation of Tcells (4B1) is dependent on the number of eosinophils and macrophages used as APC. The optimum number for antigen presentation is 2.5 x los - 3 X 105/well. When 1.5 x 10" macrophages were used as APC, no ('H) thymidine incorporation (dpm x10
1
-3
Eur. J. Immunol. 1992. 22: 1919-1925
proliferation was found (325 k 127 dpm) which was close to the value obtained without APC (435 t 124dpm). In Fig. 4B, 2 x lo5 macrophages obtained from the top of the gradient and the same number of eosinophils were used as APC to stimulate three different T cell clones.The index of stimulation obtained with macrophages was higher than that obtained with eosinophils (index of stimulation: proliferation in the presence of APC/proliferation in the absence of APC).
3.5 1L-2 production by 3 D 0 11-10 hybridoma using eosinophils as APC Eosinophils activated with GM-CSF and pulsed with OVA may act as APC and stimulate IL-2 production on 3 D 0 11.10. IL-2 production by 3 D 0 11.10 cells with different doses of OVA using pure preparations of either eosinophils or macrophages as APC is shown in Table 2. Maximum IL-2 production using pure eosinophils was obtained with 500 pg/ml of OVA.When the same number of macrophages were used as APC, IL-2 production was higher and the optimal dose was 10 pg/ml. Using 500 pg of
-3
(3H)thymidine incorporation (dpm x 10 )
I
12,
4A
lOi, -Mo
I
21 /
$--
~-
OLP
A P
0
1000
100
10
1
2-
-
Antigen I u g l m l )
02.5
Figure 2 Specificitv of T cell clones. T cell clonc proliferation was m a y e d by culturing 1 x lo4 T cells and 4 x lo5 irradiated 3pleen cclls with different concentrations of M Lorti antigen or with OVA (1 to 1000 Fgirnl) during 4 days The amount of ['HI thymidine incorpoiated during the List 18 h was measured. Three diffeient T cell clones were tcsted (---,2B2, ,2B1,-.4Bl) Results are expresscd the mean dpm ot triplicate cultures
0.6
1.25
x105 cells
Index of Stimulation
6
40
MO
T cell clone
M.corti clones
-
+
261
+
461
+
282
Figurc 3. T cell clones recognize M. corti antigens presented by eosinophils. Eosinophils were preincubated with GM-CSF and with (+). and without (-) M. corti extract (100 pg/ml) as describcd in Sect. 2.5.They were fixed then and 3 x lo4 - 4 x lo4 cloned Tcells added to the p1ates.T cell proliferation was measured by incorporation of [?HIthymidine and data shown as mean of triplicate cultures I SD.
0.15
0.3
282
EOS
481
EOS
MO
261
Figure 4. Comparative antigen presentation abilities of macrophages and eosinophils treated in similar experimental conditions. (A) APC function was assayed with eosinophils and macrophages from the same donors using T cell clones specific for M . corti antigen (4B1). Data shows proliferation of Tcell clone using different number of APC (eosinophils or macrophages). (B) Using the same number of APC (2.5 x lo5 cells), antigen presentation was tested using different T cells clones. Data represent the mean of triplicate cultures. Standard deviations are not indicated since they are lower than 5 76 of the mean. Stimulation index (S. 1.) was determined as noted in Sect. 3.4. Maximal [3H]thymidine incorporation (17478 k 2930 dpm), was obtained when macrophages presented antigen to clone 4B1. MO: macrophages; EOS: eosinophils.
Eur. J. lmmunol. 1992. 22: 1919-1925
Eosinophil as antigen-presenting cell
OVA as optimal antigen dose, four different preparations of pure eosinophils treated for 2 h with GM-CSF were used for antigen presentation to 3 D 0 11.10 hybridoma. We obtained a significant (pO.l). In contrast,T cell proliferation was statistically significant compared to the controls (p
Eur. J. Immunol. 1992. 22: 1919-1925
Victoria Del Pozo+, Belen De AndrQs+, Elena MartinA, Blanca Cardaba., Julio Cesar Fernandez, Soledad Gallardo, Paloma Tramon, Francisco Leyva-Cobian., Pilar Palomino and Carlos Lahoz Department of Immunology, Fundacion Jimknez Diaz, Madrid and Department of Immunologym, Hospital Universitario “MarquCs de Valdecilla”, Santander
Eosinophil as antigen-presenting cell: activation of T cell clones and T cell hybridoma by eosinophils after antigen processing* We have studied the role of murine eosinophils as antigen-presenting cells (APC). Eosinophils have several characteristics that support the hypothesis of its function as potential APC: they have phagocytic capacity, express adhesion molecules and major histocompatibility complex (MHC) class I1 antigens and can produce and release interleukin-1 (IL-1). We have obtained several T cell clones specific for Mesocestoides corti antigens and used T cell hybridoma specific for ovalbumin (OVA) to test this hypothesis. Granulocyte-macrophage colony-stimulating factor-activated pure eosinophils (99.9 %), express class I1 antigens and are able to present M . corti antigens to specific T cell clones or OVA to T cell hybridoma 3 D 0 11.10, inducing the proliferation of Tcell clones and IL-2 release by the T cell hybridoma. Proliferation of T cells clones is dependent on the number of eosinophils used as APC. We have compared the efficiency of the same number of macrophages and eosinophils as APC, and have found that macrophages are more efficient than eosinophils. Lysosomotropic agents, such as chloroquine and ammonium chloride, that inhibit antigen processing, impaired eosinophil presentation. This presentation is restricted by MHC class I1 and inhibited by anti-I-Ad monoclonal antibody. The present study provides clear evidence of APC function for eosinophils. Our investigation points to a new role for eosinophils in the immune response.
1 Introduction MHC class I1 expression is essential for antigen presentation by APC. Other properties such as cytokine production [l,21, the capacity to process antigen [3] and expression of adhesion molecules [4] can enhance the efficiency of antigen presentation. The prototype APC in the immune system is the macrophage. However, many other cellular types that express class I1 determinants are able to function as APC: Langerhans cells, dendritic cells, enterocytes, endothelial cells, B cells, and T cells [5-71. There is evidence supporting the conclusion that although Ia expression is necessary, it is not sufficient to ensure APC function [8, 91. Recent experiments show that liposomes bearing class I1 and membrane IL-1 are able to present antigen [lo, 111. Eosinophils are classically described as effector cells in hypersensitivity disorders [ 121 and parasitic infections [13]. Although some of their functions remain unclear, interesting characteristics have been recently described suggesting
I1 98921
*
A
This work has been supported by grants from “Fond0 de Investigacioncs Sanitarias de la Seguridad Social 8811742 and 9010238”. Recipients of thc “Conchita RBbago Fellowship” Supported by “Fondo de Investigaciones Sanitarias de la Seguridad Social”. Fellow from “Comunidad Autonoma de Madrid”
Correspondence: Carlos Lahoz, Dcpt. Immunology, Fundacion Jimtncz Diaz, Avd. Reyes Catcilicos 2, E-28040 Madrid, Spain Abbreviation:
1919
GM-CSF Granulocytc-macrophage CSF
0 VCH Verlagsgescllschaft mbH, D-6940 Weinheim, 1992
an important role of these cells in the immune response: eosinophils are phagocytic cells [14], express adhesion molecules that facilitate intercellular adhesion [151, can express class I1 MHC after treatment with granulocytemacrophage (GM)-CSF [16], are able to produce and release 1L-1 [17], and express the IL-2R pS.5 subunit [18]. All the characteristics described above suggest a possible role of the eosinophils as APC. Investigating this possibility, we confirmed class I1 expression by murine eosinophils and demonstrated their ability to interact with CD4 lymphocytes and present antigens in the context of MHC class I1 molecules.
2 Materials and methods 2.1 Animals, induction of peritoneal eosinophilia and eosinophil purification Four- to five-week-old BALB/c (H-2d) and CBAN (H-2k) female mice, were purchased from Iffa Credo EspaAa, (El Prat de Llobregat, Barcelona, Spain).
To induce peritoneal eosinophilia, BALB/c or CBA/J mice were given a single intraperitoneal injection of 70-100 Mesocestoides corti larvae as previously described [ 191. After peritoneal wash, cells were washed in PBS-0.1 % BSA-0.1 M EDTA, pH 7 (PBS-BSA-EDTA). Cells were stained with eosin dye and counted in a Fuchs-Rosenthal chamber. Contaminating cells (30 %-50 %) were mostly macrophages. Mouse eosinophils were purified by centrifugation in a discontinuous density gradient of Percoll [MI. Percoll was adjusted to isotonicity by adding 1 part of 0.5 M NaCl to
+
0014-2980/92/0707-1919$3.50 .25/0
1020
V. Del Pozo, B. de Andres. E. Martin et al.
9 parts Percoll. Four layers with different concentrations of Percoll in sterile PBS (1.5 ml of 68 % and 3 ml. of 66 %, 59 YOand 50 YO)were placed in 16 x 125 mm-plastic tubes (Costar 205, Cambridge, MA). A total of 3 x lo7- 4 x lo7 cells in 2 ml of PBS-BSA-EDTA, were placcd on top of the gradient. After centrifugation at 1800 x g for 18 min at room temperature, eosinophils were found between the layers of59 % and 66 % and the majority of macrophages on top of the gradient. Cell purity was evaluated by cytofluorymetry.
2.2 Tissue culture medium reagents and antigens Eosinophils were maintained in RPMI 1640 supplemented with 10 % FCS, 2 mM L-glutamine, 50 U/ml penicillin, M 2-ME (Flow LaboSO pg/ml streptomycin and 5 X ratories, Irvine, Scotland). Proliferation assays were done in the same medium. Chloroquine and paraformaldehyde were purchased from Sigma (St. Louis,-MO). For antigen extraction of M . corti, larvae from long-time infected-mice were washed with sterile PBS, freezed and thawed several times and sonicated for 10 min. The resulting extract was centrifuged at 12 000 x g for 30 min at 4 "C. Finally, supernatant was dialyzed against PBS overnight in the cold. Protein content was measured by Bradford's Method (Bio-Rad, Munchen, FRG). Chicken ovalbumin (OVA) was obtained from Sigma.
2.3 T cell clones and T cell hybridoma T cell clones were obtained by the method of Kimoto et al. [20] with slight modifications. Primed lymphocytes from lymph nodes (2 x lo6) of BALB/c mice infected with M . corti were cultured in 2 ml/well culture plates (Costar 3524) with 2 x lo6 irradiated syngeneic spleen cells and 100 pg/ml of M. coyti extract. After 4 to 6 days of stimulation, dead cells were removed by density gradient centrifugation over Lymphocyte Separation Medium (Flow Laboratories), and viable cells were plated without antigen or feeder cells for 7-8 days. T cells were then stimulated with antigen and irradiated syngeneic spleen cells. The cells were maintained in this cycle of alternative stimulation and rest. After the first cycle of stimulation, cells were cloned three times by limiting dilution (0.5 cells/well) in cultures containing: antigen, irradiated syngeneic spleen cells as a source of APC and 5 % mouse IL-2 prepared by stimulating EL4 thymoma cells with phorbol myristate acetate. T cell clone proliferation was assayed by culturing 1 X loJ T cells and 4 x 105 irradiated spleen cells with different antigen doses. Cultures were maintained in a total volume ot 200 p1 in flat-bottom wells (Costar 3596) for 4 days at 37 "C in 5 Y CO? humidified air, and ["H] thymidine (New England Nuclear, Boston, MA) was added for the last 18 h. Radiactivity incorporated was measured in a scintillation counter (Beckman LS 2800, Palo Alto, CA). The cloned, OVA-specific, H-2d-restricted T lymphocyte hybridoma 3 D 0 11.10 used in these experiments has described [21] and was kindly provided by Dr. A. Szabo, HBpital Cochin, Paris.
Eur. J. Irnmunol. 1992. 22: 1019-1925
2.4 Flow cytometric analysis 2.4.1 Eosinophils purity Eosinophils purity was evaluated incubating 3 x lo5 cells for 30 min at 4°C with IgGz rat anti-mouse macrophages mAb MAS034 (Sera-Lab, Ltd., Sussex, GB) and washed twice in cold PBS. Pellets were incubated for an additional 30 min with fluoresceinated anti-rat IgG (Sera-Lab). Finally, cells were washed and analyzed in a flow cytometer (Epics-C, Coulter Electronics, Hialeah, FL).
2.4.2 Class I1 expression Eosinophils (5 x lo5) were stained with fluorescein isothiocyanate-conjugated monoclonal antibody to I-Ad MHC (Beckton Dickinson, Mountainview, CA) and flow cytometry was performed.
2.4.3 Analysis of clones Tcells (1 x lo5) were incubated 30min on ice, with a biotin-conjugated monoclonal antibody against CD4 and FITC-conjugated monoclonal antibody against CD8 (SeraLab). After incubation, the cells were washed three times and incubated with 5 % mouse serum in PBS. After three washes, cells were incubated with phycoerythrin-avidin and analyzed on an Epics C (Coulter Electronics) equipped with a 488 nm argon laser (green) and with a short pass of 520 nm (red). Prior to two-color fluorescence analysis, negative as well as positive samples labeled with a single conjugated antibody (red or green) were analyzed. Fluorescein-conjugated anti-mouse B cells and anti-mouse macrophages antibody (Sera-Lab) were used for B cell and macrophage detection.
2.5 Antigen presentation to T cell clones Purified eosinophils (2 x lo5) in 100 pl of complete RPMI supplemented with 50 U/ml of recombinant murine GMCSF (Genzyme, Boston, MA) were plated for 120 min at 37°C in flat-bottom 96-well plates. In the last 90 min of incubation 100 pglml of the corresponding antigen were added. After three washes cells were fixed with 100 pl of 2 % paraformaldehyde (final concentration 1%) for 20 min at room temperature to avoid degranulation of the eosinophils. After three washes with PBS, 100 p1 of complete medium was added and fixed cells were incubated overnight at 4 "C for possible paraforinaldehyde leaching. Cells were washed three more times before the addition of 3 x lo4 - 4 x lo4 T cells to each well. Cell proliferation was measured as described previously.
2.6 Antigen presentation to T hybridoma 3 D 0 11.10 and IL-2 assays The assay for the stimulation of IL-2 production by antigen-specific hybridoma was performed as described [22].Briefly, 2 x 105APC, 1 x lo5 hybridomaT cells, and antigen at the appropriate concentrations were cultured in 0.3 ml complete medium in microculture plates for 24 h at
37°C. The presence of IL-2 in the culture supernatant was then assayed using the IL-Zdependent CTLL-2 cell line. 2.7 Treatment of eosinophils with lysosomotropic agents chloroquine and NH&l, and APC inhibition using anti-I-Aa
Chloroquine and NH4CI were freshly prepared in RPMI and used at the concentration of 100kM and 10 pM, respectively for 30 min at 37 "C. Treated cells were washed in medium three times, and processed as indicated previously for use as APC. Eosinophils (2 x 1O6/ml) pre-treated with GM-CSF were incubated with an appropriate concentration (62.5 pg antibody/ml) of dialyzed anti-mouse I-Ad antisera (Becton Dickinson) for 30 min at 37 "C. Alternatively anti-murine B cells antibody, under the same experimental conditions, was used a5 a negative control. Thereafter eosinophils were treated as previously described to be used as antigenpresenting cells.The viability after treatment with antibody was measured by the trypan blue exclusion and was > 95%.
the expression of MHC class I1 was analyzed. As shown in Table 1an increase of MHC class IT surface expression was found during the culture. In some experiments we detected a certain degree of class I1 expression at the beginning of culture which increased with GM-CSF treatment. Similar results have been found with human eosinophils where in one case 23.6 % of eosinophils expressed HLA-DR molecules without previous activation [l5]. 3.4 Proliferation of T cell clones using eosinophils as APC M . corti-specificT cell clones proliferate when added to the plates with eosinophils previously treated with GM-CSF and pulsed with antigen (100 pg/ml) as described in
Table 1. I-Ad expressiona) on eosinophils treated with GMCSFb) Timc
3.1 Eosinophil purification
Experiments were usually done with 99.9 % pure eosinophi1 preparation as tested by cytofluorimetry and cytocentrifuge preparations. In some instances, preparations with a purity ranging between 99 % to 95 % were employed, but in all theses cases the contaminating cells were not typical activated macrophages as determined by morphological and fluorocytometric analysis. Fig. 1 shows the cytotluorimetric analysk of a typical eosinophil preparation stained with anti-mouse macrophages antibody before (panel C) and after cell separation by Percoll gradient (panel B). Macrophages are absent in the final eosinophil preparation (panel B). Panel A shows the control background fluorescence using the second fluoresceinated antibody.
ESP. I
Ilxp.2 Enp.3 Iixp.4 Iixp.5 1ixp.h Eup.7 1:xp.X
(11)
(I
3 Results
1921
Eosinophil as antigen-presenting cell
Eur. J. lmmunol. 1992. 22: 1919-1925
2 24
0 5CJ 9.66 70.2
ND
28.7 38.3
I4 ND
28.1
SD
99.5
0.4
11.9
32.2 N1)
11.3 11.7
NI)
(1.0 KD 78.1
0.7 KIl 40
a) Eosinophils were stained with fluoresceinated monoclonal antibody anti I-Ad for I-A detection and analyzed by flow cytometry. b) Purified eosinophils were cultured for the indicated periods of time in the presence of 50 U/ml of recombinant GM-CSE c) Results are expressed as the percentage of positive cells.
3.2 Characterization of T cell clones
All three clones (2B1,4B I , and 2B2) used in this study were analyzed by flow cytometry. The phenotype of these cells was CD3+, CD4+, CDX-. These cells did not express any macrophage or B cell surface marker. These clones respond to M . corti antigens when presented on H-2d irradiated spleen cells. Fig. 2 shows the proliferation of these three T cell clones to different doses of antigen ranging from 1 to 1000 pg/ml of M . c o r k I n contrast, no proliferation was detected when an unrelated antigen (OVA) was used. 3.3 Expression Expression of of class class I1 I1 antigens antigens on on eosinophils eosinophils 3.3 In humans Lucey et al. [lS] have found that eosinophils maintained in culture with recombinant human GM-CSF express HLA-DR antigens. Eosinophils (2 X 106/ml)were cultured with SO U/ml of recombinant mouse GM-CSF, and
Figure 1. 1. Cytofluorometric Cytofluorometric analysis analysis of of eosinophil eosinophil preparations. preparations. In In Figure panel A A (background (background control), control),iiss shown shown the the cell cell population populationstained stained panel with fluoresceinated anti-rat Ig. Panel B: analysis of eosinophil preparation after Percoll gradient using mAb against mouse macrophages MAS034, as describedin Sect. 2.4.1. Panel C: Profile of the whole cell population before purification stained with anti-macrophage antibody. Five thousand cells were analyzed.
V. Del Pozo, B. de Andres, E. Martin et al.
1922
Sect. 2.5. Fig. 3 shows the specificT cell clone proliferation in response to antigen presented by eosinophils from syngeneic mice. Only background proliferation was detected using T cells alone (516 k 43 dpm), T cells with GM-CSF (495 k 224 dpm). eosinophils without GM-CSF (471 f 91 dpm) and T cells with soluble antigen (307 f 16 dpm) as controls. The differences between controls without antigen and the proliferating clones were statistically significant 0, < 0.05). If eosinophils were fixed with paraformaldehyde prior to pulsing with antigen. no proliferation was found (516 dpm k 162). We wanted to compare the efficiency of different number of macrophages and eosinophils, similarly treated with GMCSF, for antigen presentation. Fig. 4A shows T cell proliferation induced by different amounts of APC. The proliferation of Tcells (4B1) is dependent on the number of eosinophils and macrophages used as APC. The optimum number for antigen presentation is 2.5 x los - 3 X 105/well. When 1.5 x 10" macrophages were used as APC, no ('H) thymidine incorporation (dpm x10
1
-3
Eur. J. Immunol. 1992. 22: 1919-1925
proliferation was found (325 k 127 dpm) which was close to the value obtained without APC (435 t 124dpm). In Fig. 4B, 2 x lo5 macrophages obtained from the top of the gradient and the same number of eosinophils were used as APC to stimulate three different T cell clones.The index of stimulation obtained with macrophages was higher than that obtained with eosinophils (index of stimulation: proliferation in the presence of APC/proliferation in the absence of APC).
3.5 1L-2 production by 3 D 0 11-10 hybridoma using eosinophils as APC Eosinophils activated with GM-CSF and pulsed with OVA may act as APC and stimulate IL-2 production on 3 D 0 11.10. IL-2 production by 3 D 0 11.10 cells with different doses of OVA using pure preparations of either eosinophils or macrophages as APC is shown in Table 2. Maximum IL-2 production using pure eosinophils was obtained with 500 pg/ml of OVA.When the same number of macrophages were used as APC, IL-2 production was higher and the optimal dose was 10 pg/ml. Using 500 pg of
-3
(3H)thymidine incorporation (dpm x 10 )
I
12,
4A
lOi, -Mo
I
21 /
$--
~-
OLP
A P
0
1000
100
10
1
2-
-
Antigen I u g l m l )
02.5
Figure 2 Specificitv of T cell clones. T cell clonc proliferation was m a y e d by culturing 1 x lo4 T cells and 4 x lo5 irradiated 3pleen cclls with different concentrations of M Lorti antigen or with OVA (1 to 1000 Fgirnl) during 4 days The amount of ['HI thymidine incorpoiated during the List 18 h was measured. Three diffeient T cell clones were tcsted (---,2B2, ,2B1,-.4Bl) Results are expresscd the mean dpm ot triplicate cultures
0.6
1.25
x105 cells
Index of Stimulation
6
40
MO
T cell clone
M.corti clones
-
+
261
+
461
+
282
Figurc 3. T cell clones recognize M. corti antigens presented by eosinophils. Eosinophils were preincubated with GM-CSF and with (+). and without (-) M. corti extract (100 pg/ml) as describcd in Sect. 2.5.They were fixed then and 3 x lo4 - 4 x lo4 cloned Tcells added to the p1ates.T cell proliferation was measured by incorporation of [?HIthymidine and data shown as mean of triplicate cultures I SD.
0.15
0.3
282
EOS
481
EOS
MO
261
Figure 4. Comparative antigen presentation abilities of macrophages and eosinophils treated in similar experimental conditions. (A) APC function was assayed with eosinophils and macrophages from the same donors using T cell clones specific for M . corti antigen (4B1). Data shows proliferation of Tcell clone using different number of APC (eosinophils or macrophages). (B) Using the same number of APC (2.5 x lo5 cells), antigen presentation was tested using different T cells clones. Data represent the mean of triplicate cultures. Standard deviations are not indicated since they are lower than 5 76 of the mean. Stimulation index (S. 1.) was determined as noted in Sect. 3.4. Maximal [3H]thymidine incorporation (17478 k 2930 dpm), was obtained when macrophages presented antigen to clone 4B1. MO: macrophages; EOS: eosinophils.
Eur. J. lmmunol. 1992. 22: 1919-1925
Eosinophil as antigen-presenting cell
OVA as optimal antigen dose, four different preparations of pure eosinophils treated for 2 h with GM-CSF were used for antigen presentation to 3 D 0 11.10 hybridoma. We obtained a significant (pO.l). In contrast,T cell proliferation was statistically significant compared to the controls (p
