HYBRIDOMA Volume 9, Number 4, 1990 Mary Ann Liebert, Inc., Publishers
Monoclonal Antibodies to Complement Components Without the Need of Their Prior Purification. II. Antibodies to Mouse C3 And C4 ELISABETH KREMMER,1 STEFAN THIERFELDER,1 ERICH FELBER,2 GERTRUD HOFFMANN-FEZER,1 and MICHAEL WASILIU1 'Institute of Immunology, GSF, Marchioninistr. 25, 8000 München 70, FRG 'Institute of Immunology, LMU München, Goelhestr. 31, 8000 München 2, FRG
ABSTRACT Rat monoclonal antibodies (MAbs) to mouse complement components C3 and C4 were produced by immunizing rats with cell-bound C3 and C4. This principle involves: a) using mouse thymocytes coated with syngeneic rat antibody isotypes that show high affinity to Clq, b) the intercalation of Clq from serum and c) the subsequent activation of the classical complement pathway leading to deposition of cell-bound complement components. Screening for anti-complement antibodies was performed on antibody coating microtiter plates with mouse serum as source of complement. The reactivity of the MAbs was determined by variations of the ELISA screening system using EDTAserum to inhibit complement activation by Cl dissociation, serum rendered deficient of functionally active C3 by treatment with cobra venom factor (CVF) or serum of genetically C5-deficient mice. The specificity of the MAbs was confirmed by affinity chromatography followed by SDS-PAGE and immunoblotting. We were able to establish a panel of anti-C3 and anti-C4 MAbs of various isotypes.
INTRODUCTION Monoclonal and
polyclonal antibodies are widely used as powerful tools for suppressing transplant immunity (1,2,3,4). Prerequisites for efficient target cell elimination by monoclonal antibodies, other than the MAb specifity, are the antigen density of the target cell and the MAb isotype, leading to Clquptake on antibody-coated cells. However, some controversy exists about the antibody-dependent
elimination mechanism and the function of distinct complement components or receptors for target cell elimination. Besides the damage and subsequent lysis of target cells by the membrane-attackcomplex (MAC) or by antibody-dependent cell-mediated cytotoxicity (ADCC), Clq-receptor bearing effector cells or opsonisation of the target cell by deposition of C3b during complement activation have been discussed (5,6,7,8). In order to elucidate the mode of cell elimination in a mouse model, we raised MAbs to antigenic determinants of cell-bound complement components; these MAbs could be useful for the study of in vivo complement activation by anti-T cell antibodies injected in mice in an experimental transplantation model. We avoided the necessity to purify complement components from mouse serum by immunizing rats with murine complement components deposited on cells during complement activation by rat antibodies. A convenient way of fixing complement to mouse cells is the intercalation of Clq to rat IgG2b or mouse IgG2a antibodies bound to mouse thymocytes via Thy-1 (9). 309
of surface Thy-1 antigen and the intrinsic affinity of mouse Clq to rat IgG2b activation of the classical pathway of complement with ravalent binding of C4 and C3 promotes and non-covalent attachment of C5b-9, the membrane-attack-complex (12).
High density
(10,11) to
cells
MATERIAL AND METHODS
Animals
C57BL/6 (H2b), AKR/J mice (H2k), Balb/c (H2d) and CBA (H2k) mice, originally from The Jackson Laboratory (Bar Harbor, ME), and Lou/c rats were raised and maintained in our breeding facilities. Initial breeding pairs of inbred strain Lou/c rats were a generous gift from Prof.H.Bazin (University of Louvain, Bruxelles). Immunization schedule
2x10* C57BL/6 thymocytes were incubated at room temperature with saturating amounts of a rat anti-
Thy-1 monoclonal antibody of IgG2b subclass (RmTl; 13) and 20% freshly prepared C57BL/6 serum. After 20 min the cells were washed once with PBS, and 5xl07 cells injected intraperitoneaUy and subcutaneously into female Lou/c rats. 4 weeks later the rats were boosted intraperitoneaUy and intravenously with 5xl07 thymocytes, coated with antibodies and cell-bound complement components as described before. 4 days later the animals were sacrificed and the spleens were collected asepticaUy for cell fusion. Cell Fusion Fusion of the
myeloma P3X63 Ag 8-653 and rat immune spleen cells was performed according to the general procedure described by Köhler and Milstein (14). Screening Assay Hybridoma supernatants were tested by solid-phase immunoassay. Polystyrene microtiter plates (Greiner, Niirtingen) were coated with 5 pg/nd of a mouse monoclonal antibody of subclass IgG2a (MmTl, 15) in carbonate-bicarbonate buffer (50 mM, pH 9.5). Before use, the wells were blocked with PBS containing 1% powdered milk (Fink, Herrenberg, FRG). Freshly prepared C57BL/6 serum in a final concentration of 20% was taken as source of complement and allowed to activate for 30 min at 37°C. C57BL/6 EDTA-semm was used as negative control. After washing three times with PBS, the culture supernatants were added and incubated for another 60 min. Wells were washed again and were then incubated with 100 pi of appropriately diluted peroxidase-labeled mouse F(ab)2 anti-rat IgG (Jackson Lab., Westgroves) for one hour at room temperature. The coloring reaction was performed by orthophenylenediamine (OPD; Sigma, Deisenhofen, FRG) and the optical density at 405 nm was measured by an immuno-reader (SLT, Salzburg, Austria). Serological tests Antibody producing hybridomas reacting positively in the described screening system were further analyzed using the following variations of the initial ELISA system. C57BL/6 serum used as a complement source was replaced by: a) C57BL/6 serum containing 5 mM EDTA leading to dissociation of Cl (16), b) C57BL/6 serum treated with 10 U/ml cobra venom factor (CVF, Diamedix, Miami, PA) for 2h at 4°C resulting in complete consumption of functionaUy active C3 (17), c) serum of CBA mice with low C4 levels (18) and d) serum of AKR/J mice known to be C5-deficient (19) and showing low levels of C4 as do all H2k strains.
Cloning and isotype analysis Hybridomas were cloned at
least twice by limiting dUution in microtiter wells. The immunoglobulin determined in a solid-phase ELISA with mouse anti-rat antibodies as capture and biotinylated monoclonal mouse anti-rat kappa, -lambda or mouse anti-IgM antibodies (Zymed, CA) or anti-IgG subclasses (20) as indicator. class
was
310
Purification of MAbs MAbs were purified from culture supernatant by affinity chromatography on a column of Protein G Sepharose 4 Fast Flow (Pharmacia, Freiburg, FRG) using Fast Performance Liquid Chromatography
(FPLC).
Purification of mouse C3 and C4 The purified antibodies were coupled to CNBr-activated Sepharose 4B as described by the manufacturer (Pharmacia, Freiburg, FRG). EDTA-plasma from heparinized C57BL/6 mice was appUed to the column in the presence of protease inhibitor (PMSF, 10 pM). Bound complement components were eluted with 50 mM diethylamine (pH 11.5) and immediately neutralized with 1/10 volume of 500 mM NaH2P04. Contamination by rat MAbs were removed from the eluates by Protein G Sepharose 4 Fast Flow in a batch procedure.
Immunoblotting AUquots of affinity-purified mouse C3 or C4 were reduced with dithiothreitol and subjected to SDSpolyacrylamide gel electro-phoresis (10%) according to LaemmU (21, 22). The separated proteins were transferred to nitrocellulose in a transfer chamber (25 mM Tris; 192 mM Glycin pH 7.2) for 1.5h at 60V. The nitrocellulose membrane was sequentially incubated with 1% powdered milk containing 0.05% Tween 20 to block unreacted sites, followed by monoclonal antibodies from culture supernatants or unspecific MAb as a control, peroxidase-labeled goat anti-human C3 (Cooper Biomédical, Malvern, PA) or goat anti-human C4 (ATAB, Scarborough, ME), followed by peroxidase-labeled rabbit antigoat IgG (Dakopatts, Hamburg, FRG). The MAbs were finaUy incubated with peroxidase-conjugated mouse anti-rat IgG antibodies (Jackson, Westgrove). The blotting membrane was then developed with 4-chloro-naphtol as substrate. Immunhistochemistry Spleens from C57BL/6 mice were removed and snap frozen in Uquid nitrogen. Cryostat sections (5 pm thick) were airdried and fixed with acetone for 10 min. Incubations with antibodies lasted for 60 min and were stopped by washing in PBS. C3- or C4-binding was demonstrated by the following incubations: a) mouse anti-Thy-1 antibodies (MmTl) together with an equal part of serum from C57BL/6 mice or CVF-treated serum, b) supernatants of RmC3-llH9 or RmC4-16D2, c) peroxidaselabeled mouse anti-rat IgG. Peroxidase activity was revealed with amino-ethyl-carbazol (23). The tissue sections were counterstained with hematoxiline.
RESULTS
Production of monoclonal anti-mouse complement antibodies of P3X63-Ag8.653 myeloma cells with spleen ceils of immunized Lou/c rats were screened for specific antibody production by ELISA using MmTl coating microtiter plates and C57BL/6 serum as described above. 39 hybridomas reacted positively with mouse serum activated on microtiter plates, but negatively with EDTA-serum, and were, therefore, further investigated.
Supernatants of hybridomas
Differentiation of MAbs by serological tests
Due to our immunization approach, we expected antibodies to cell-bound complement components. To evaluate the specifity of the MAbs,we screened with mouse serum lacking certain complement components. No anti-Clq specificity was found with mouse serum after addition of EDTA (5 mM) in order to dissociate Clq from the Clr2-Cls2 tetramer. Antibody to complement components prior to C3 were screened with mouse serum pretreated with cobra venom factor (CVF, 10 U/ml). CVF stabilizes the alternative C3 convelíase C3bBb, causing the complete consumption of C3. MAbs recognizing components prior to C3 should, therefore, react positively. 11 MAbs reacted with CVF-treated C57BL/6-serum but weakly with serum of CBA mice with low C4 levels. Serum of Balb/c mice (H2 ) showed the same reactivity pattern as C57BL/6 serum (data not shown). 28 MAbs reacted negatively with CVF-treated C57BL/6-serum but positively with C5-deficient serum of AKR/J mice, indicating specifity for C3 (RmC3). Reactivity with proteins of the alternative pathway (amplification loop) was
311
Group
Group 2
1
FIGURE 1: The columns represent the average reactivity of the MAbs with different pretreated sera in relation to untreated C57BL/6 serum. Two groups of antibodies were distinguished by serological tests: 1. MAbs reacting positively with serum of all mouse strains, best with C57BL/6 serum (100%), but negatively in CVF-treated serum were classified as anti-C3 MAbs. 2. MAbs reacting positively in serum (100%) and CVF-treated serum of C57BL/6 mice, but showing little reactivity in mouse serum with low C4 levels (AKR, CBA) revealed their anti-C4 specificity in
immunoblotting.
as unlikely since cells were washed stringently during immunization differentiation of the MAbs by serological tests is shown in figure 1.
regarded
procedure.
The
Immunoblotting of mouse C3 and mouse C4 RmC3 and RmC4 antibodies were investigated by immunoblotting. One antibody, each of presumed C3 or C4 specifity, was coupled to CNBr-activated Sepharose 4B, and EDTA-plasma from heparinized C57BL/6 mice was applied. The eluted protein was subjected to SDS gel electrophoresis under
reduced conditions and transferred to nitrocellulose. Detection of mouse C3 in the eluate of RmC311H9 using crossreacting goat anti-human C3 showed 2 distinct bands, corresponding to the a- and ßchain of the C3 molecule, whereas the RmC3-llH9 eluate reacted negatively with goat anti-human C4 and the serologicaUy defined anti-mouse C4 MAbs. Anti-C3 reactivity of 22 MAbs out of 28 serologicaUy defined anti-C3 MAbs could be demonstrated by immunoblotting (Fig. 2a). Mouse C4 in the eluate of RmC4-16D2 showed two bands with goat anti-human C4, corresponding to the a- and ß-chain of C4 (the gamma-chain could not be detected), but none with goat anti-human C3 nor with the anti-mouse C3 MAbs. Of 11 MAbs against mouse C4, as demonstrated by serological tests, 7 MAbs were confirmed by immunoblotting (Fig. 2b). Reactivity of the MAbs are summarized in table 1.
Immunohistochemical studies In the absence of fresh mouse serum, labeled RmC3-llH9 was positive in a netlike framework in germinal centers of mouse spleens and furthermore stained single cells as well as aggregations of a few cells in the red pulp. RmC3-llH9 also distinctly labeled the fiber network surrounding central arterioles of white pulp and stained more faintly the white pulp fibers bordering on the marginal zone. RmC3-llH9 bound to T-cell region foUowing a simultaneous incubation of anti-Thy-1 MAb MmTl with fresh C57BL/6 serum as source of complement (Fig.3a). RmC3-llH9 was negative following incubation of anti-Thy-1 MAb MmTl with CVF-treated C57BL/6 serum (Fig.3b). In the absence of fresh mouse serum, RmC4-16D2 stained a coarse network in germinal centers and a
312
B KD •180
KD 180
116
-116
84
-84
ß-i
dg58
-58
-48,5
-48,5
36,5
-36,5
26,6
-26,6
12 3 4 5
6 7 8 9 10
FIGURE 2A: Immunoblot of the eluate of mouse EDTA-plasma applied to immobilized RmC3-llH9. Detection of mouse C3 by: 1) goat anti-human C3, 2) goat anti-human C4, 3) RmC3-llH9, 4) RmC3-
5A5, 5) RmC3-17D5; dg: degradation products.
FIGURE 2B: Immunoblot of the eluate of mouse EDTA-plasma applied to immobilized RmC416D2.Detection of mouse C4 by: 6) goat anti-human C4, 7) goat anti-human C3, 8) RmC4-16D2, 9) RmC4-17Cl, 10) RmC4-17D9.
faint network along the border between white pulp and marginal zone in tissue sections of mouse spleen. Fibers around arterioles were negative. In red pulp, single positive cells were observed. RmC416D2 was bound to T-cell regions following incubation with anti-Thy-1 MAb MmTl and fresh or CVFtreated C57BL/6 serum as source of complement (Fig.3c). DISCUSSION So
far, MAbs to human complement components have been raised by immunizing mice with purified
factors (24,25,26). Monoclonal antibodies to mouse complement have been obtained in the case of C5 by immunizing C5 deficient mice with serum from normal mice (27). Recently we described an immunization scheme for the production of anti-mouse Clq MAbs that avoids the necessity to purify serum-derived Clq and requires only smaU amounts of EDTA-serum as a source of mouse Clq (28). In the present study we used the same principle of immunization, namely the high density of surface Thy-1 antigen on C57BL/6 thymocytes and the intrinsic affinity of mouse Clq to rat IgG2b antibodies bound to mouse thymocytes via Thy-1. The use of freshly prepared serum as a source of complement components activates the classical complement cascade, thus leading to the deposition of cell-bound complement components. In order to circumvent cellular test systems, we established a solid phase ELISA system for detection of anti-mouse complement antibodies. The requirement for such a test system is sufficient adsorption of Clq-affine antibodies (rat IgG2b, mouse IgG2a) to polystyrene microtiter plates in such a way that Clq molecules are able to intercalate between adjacent pairs of Fc portions of antibodies (30-40 nm, 29)
complement
313
TABLE 1:
Reactivity Pattern of Rat anti-Mouse Complement Immunoblotting Techniques MAb
subclass1
C57BL/6 (C4high)
C57BL/6
+ EDTA
+CVF
(C4high)
Antibodies
AKR
CBA
(C41ow, C5 def.)
(C4 low)
as
Revealed
2a 2a Gl
+
+
+
+
+
+
+
(+) (+)
2b
+ + + + +
+
+++
6C9
2b
+
15H1 15H12
M 2b
+ +
16D2 17C1 17D9 31C3
2a M 2a Gl
+
M 2a Gl Gl M
+ +
+
+
(+)
+
+
+
+
+
+
+
+
++ +
+
+
+
+
+
+++
+
+
+ + + + + + +
+
(+) (+)
+
+
+ +
+
+
(+) (+)
(+) (+) (+) (+) (+)
+
+
+
(+)
+ + + + +
and
Immunoblot reactivity2 to C3 to C4
2H9 4A3 4A7 5A5 5E8 5F5 6C11 6E8 6G12 9A11 11B1 11H9 12B6 17C6 17D5 17F5 20G10 23G4 31E8 31D6 31F1 32G2
2a Gl Gl 2a 2a 2a 2a 2a 2a Gl Gl 2a 2a
by Serological
+ +
(+) +
(+)
(+) (+) (+) (+) (+) (+) (+)
+ + + + + + +
1
all MAbs have kappa light chains gradiations in response in immunoblotting: reactivity 2
no, ( + ) weak, +
intermediate,
++
good,
+++
strong
and to induce the classical complement activation. A mouse IgG2a MAb (MmTl) was, therefore, coated to microtiter plates in an optimal concentration of 5 pg/rrA. A 20% dilution of C57BL/6 serum was used to screen the fusion products. The negative control was performed with C57BL/6 EDTAserum. A similar ELISA system was recently applied to detect deficiencies in the classical complement pathway in humans (30). We expected rat MAbs to mouse complement components C3 and C4 because of their ravalent attachment to proteins on the cell surface used for immunization and also deposited on microtiter plates in the screening system during activation. No anti-C2 MAbs were identified. C2 (and also C5b-9, MAC) is only bound non-covalently to the ceU surface and may have been released from the mouse thymocytes before immunization could take place. MAbs to complement
314
¡3*v«
/ ,
V,"' -fi
^m;^.,-7,£*.
Monoclonal Antibodies to Complement Components Without the Need of Their Prior Purification. II. Antibodies to Mouse C3 And C4 ELISABETH KREMMER,1 STEFAN THIERFELDER,1 ERICH FELBER,2 GERTRUD HOFFMANN-FEZER,1 and MICHAEL WASILIU1 'Institute of Immunology, GSF, Marchioninistr. 25, 8000 München 70, FRG 'Institute of Immunology, LMU München, Goelhestr. 31, 8000 München 2, FRG
ABSTRACT Rat monoclonal antibodies (MAbs) to mouse complement components C3 and C4 were produced by immunizing rats with cell-bound C3 and C4. This principle involves: a) using mouse thymocytes coated with syngeneic rat antibody isotypes that show high affinity to Clq, b) the intercalation of Clq from serum and c) the subsequent activation of the classical complement pathway leading to deposition of cell-bound complement components. Screening for anti-complement antibodies was performed on antibody coating microtiter plates with mouse serum as source of complement. The reactivity of the MAbs was determined by variations of the ELISA screening system using EDTAserum to inhibit complement activation by Cl dissociation, serum rendered deficient of functionally active C3 by treatment with cobra venom factor (CVF) or serum of genetically C5-deficient mice. The specificity of the MAbs was confirmed by affinity chromatography followed by SDS-PAGE and immunoblotting. We were able to establish a panel of anti-C3 and anti-C4 MAbs of various isotypes.
INTRODUCTION Monoclonal and
polyclonal antibodies are widely used as powerful tools for suppressing transplant immunity (1,2,3,4). Prerequisites for efficient target cell elimination by monoclonal antibodies, other than the MAb specifity, are the antigen density of the target cell and the MAb isotype, leading to Clquptake on antibody-coated cells. However, some controversy exists about the antibody-dependent
elimination mechanism and the function of distinct complement components or receptors for target cell elimination. Besides the damage and subsequent lysis of target cells by the membrane-attackcomplex (MAC) or by antibody-dependent cell-mediated cytotoxicity (ADCC), Clq-receptor bearing effector cells or opsonisation of the target cell by deposition of C3b during complement activation have been discussed (5,6,7,8). In order to elucidate the mode of cell elimination in a mouse model, we raised MAbs to antigenic determinants of cell-bound complement components; these MAbs could be useful for the study of in vivo complement activation by anti-T cell antibodies injected in mice in an experimental transplantation model. We avoided the necessity to purify complement components from mouse serum by immunizing rats with murine complement components deposited on cells during complement activation by rat antibodies. A convenient way of fixing complement to mouse cells is the intercalation of Clq to rat IgG2b or mouse IgG2a antibodies bound to mouse thymocytes via Thy-1 (9). 309
of surface Thy-1 antigen and the intrinsic affinity of mouse Clq to rat IgG2b activation of the classical pathway of complement with ravalent binding of C4 and C3 promotes and non-covalent attachment of C5b-9, the membrane-attack-complex (12).
High density
(10,11) to
cells
MATERIAL AND METHODS
Animals
C57BL/6 (H2b), AKR/J mice (H2k), Balb/c (H2d) and CBA (H2k) mice, originally from The Jackson Laboratory (Bar Harbor, ME), and Lou/c rats were raised and maintained in our breeding facilities. Initial breeding pairs of inbred strain Lou/c rats were a generous gift from Prof.H.Bazin (University of Louvain, Bruxelles). Immunization schedule
2x10* C57BL/6 thymocytes were incubated at room temperature with saturating amounts of a rat anti-
Thy-1 monoclonal antibody of IgG2b subclass (RmTl; 13) and 20% freshly prepared C57BL/6 serum. After 20 min the cells were washed once with PBS, and 5xl07 cells injected intraperitoneaUy and subcutaneously into female Lou/c rats. 4 weeks later the rats were boosted intraperitoneaUy and intravenously with 5xl07 thymocytes, coated with antibodies and cell-bound complement components as described before. 4 days later the animals were sacrificed and the spleens were collected asepticaUy for cell fusion. Cell Fusion Fusion of the
myeloma P3X63 Ag 8-653 and rat immune spleen cells was performed according to the general procedure described by Köhler and Milstein (14). Screening Assay Hybridoma supernatants were tested by solid-phase immunoassay. Polystyrene microtiter plates (Greiner, Niirtingen) were coated with 5 pg/nd of a mouse monoclonal antibody of subclass IgG2a (MmTl, 15) in carbonate-bicarbonate buffer (50 mM, pH 9.5). Before use, the wells were blocked with PBS containing 1% powdered milk (Fink, Herrenberg, FRG). Freshly prepared C57BL/6 serum in a final concentration of 20% was taken as source of complement and allowed to activate for 30 min at 37°C. C57BL/6 EDTA-semm was used as negative control. After washing three times with PBS, the culture supernatants were added and incubated for another 60 min. Wells were washed again and were then incubated with 100 pi of appropriately diluted peroxidase-labeled mouse F(ab)2 anti-rat IgG (Jackson Lab., Westgroves) for one hour at room temperature. The coloring reaction was performed by orthophenylenediamine (OPD; Sigma, Deisenhofen, FRG) and the optical density at 405 nm was measured by an immuno-reader (SLT, Salzburg, Austria). Serological tests Antibody producing hybridomas reacting positively in the described screening system were further analyzed using the following variations of the initial ELISA system. C57BL/6 serum used as a complement source was replaced by: a) C57BL/6 serum containing 5 mM EDTA leading to dissociation of Cl (16), b) C57BL/6 serum treated with 10 U/ml cobra venom factor (CVF, Diamedix, Miami, PA) for 2h at 4°C resulting in complete consumption of functionaUy active C3 (17), c) serum of CBA mice with low C4 levels (18) and d) serum of AKR/J mice known to be C5-deficient (19) and showing low levels of C4 as do all H2k strains.
Cloning and isotype analysis Hybridomas were cloned at
least twice by limiting dUution in microtiter wells. The immunoglobulin determined in a solid-phase ELISA with mouse anti-rat antibodies as capture and biotinylated monoclonal mouse anti-rat kappa, -lambda or mouse anti-IgM antibodies (Zymed, CA) or anti-IgG subclasses (20) as indicator. class
was
310
Purification of MAbs MAbs were purified from culture supernatant by affinity chromatography on a column of Protein G Sepharose 4 Fast Flow (Pharmacia, Freiburg, FRG) using Fast Performance Liquid Chromatography
(FPLC).
Purification of mouse C3 and C4 The purified antibodies were coupled to CNBr-activated Sepharose 4B as described by the manufacturer (Pharmacia, Freiburg, FRG). EDTA-plasma from heparinized C57BL/6 mice was appUed to the column in the presence of protease inhibitor (PMSF, 10 pM). Bound complement components were eluted with 50 mM diethylamine (pH 11.5) and immediately neutralized with 1/10 volume of 500 mM NaH2P04. Contamination by rat MAbs were removed from the eluates by Protein G Sepharose 4 Fast Flow in a batch procedure.
Immunoblotting AUquots of affinity-purified mouse C3 or C4 were reduced with dithiothreitol and subjected to SDSpolyacrylamide gel electro-phoresis (10%) according to LaemmU (21, 22). The separated proteins were transferred to nitrocellulose in a transfer chamber (25 mM Tris; 192 mM Glycin pH 7.2) for 1.5h at 60V. The nitrocellulose membrane was sequentially incubated with 1% powdered milk containing 0.05% Tween 20 to block unreacted sites, followed by monoclonal antibodies from culture supernatants or unspecific MAb as a control, peroxidase-labeled goat anti-human C3 (Cooper Biomédical, Malvern, PA) or goat anti-human C4 (ATAB, Scarborough, ME), followed by peroxidase-labeled rabbit antigoat IgG (Dakopatts, Hamburg, FRG). The MAbs were finaUy incubated with peroxidase-conjugated mouse anti-rat IgG antibodies (Jackson, Westgrove). The blotting membrane was then developed with 4-chloro-naphtol as substrate. Immunhistochemistry Spleens from C57BL/6 mice were removed and snap frozen in Uquid nitrogen. Cryostat sections (5 pm thick) were airdried and fixed with acetone for 10 min. Incubations with antibodies lasted for 60 min and were stopped by washing in PBS. C3- or C4-binding was demonstrated by the following incubations: a) mouse anti-Thy-1 antibodies (MmTl) together with an equal part of serum from C57BL/6 mice or CVF-treated serum, b) supernatants of RmC3-llH9 or RmC4-16D2, c) peroxidaselabeled mouse anti-rat IgG. Peroxidase activity was revealed with amino-ethyl-carbazol (23). The tissue sections were counterstained with hematoxiline.
RESULTS
Production of monoclonal anti-mouse complement antibodies of P3X63-Ag8.653 myeloma cells with spleen ceils of immunized Lou/c rats were screened for specific antibody production by ELISA using MmTl coating microtiter plates and C57BL/6 serum as described above. 39 hybridomas reacted positively with mouse serum activated on microtiter plates, but negatively with EDTA-serum, and were, therefore, further investigated.
Supernatants of hybridomas
Differentiation of MAbs by serological tests
Due to our immunization approach, we expected antibodies to cell-bound complement components. To evaluate the specifity of the MAbs,we screened with mouse serum lacking certain complement components. No anti-Clq specificity was found with mouse serum after addition of EDTA (5 mM) in order to dissociate Clq from the Clr2-Cls2 tetramer. Antibody to complement components prior to C3 were screened with mouse serum pretreated with cobra venom factor (CVF, 10 U/ml). CVF stabilizes the alternative C3 convelíase C3bBb, causing the complete consumption of C3. MAbs recognizing components prior to C3 should, therefore, react positively. 11 MAbs reacted with CVF-treated C57BL/6-serum but weakly with serum of CBA mice with low C4 levels. Serum of Balb/c mice (H2 ) showed the same reactivity pattern as C57BL/6 serum (data not shown). 28 MAbs reacted negatively with CVF-treated C57BL/6-serum but positively with C5-deficient serum of AKR/J mice, indicating specifity for C3 (RmC3). Reactivity with proteins of the alternative pathway (amplification loop) was
311
Group
Group 2
1
FIGURE 1: The columns represent the average reactivity of the MAbs with different pretreated sera in relation to untreated C57BL/6 serum. Two groups of antibodies were distinguished by serological tests: 1. MAbs reacting positively with serum of all mouse strains, best with C57BL/6 serum (100%), but negatively in CVF-treated serum were classified as anti-C3 MAbs. 2. MAbs reacting positively in serum (100%) and CVF-treated serum of C57BL/6 mice, but showing little reactivity in mouse serum with low C4 levels (AKR, CBA) revealed their anti-C4 specificity in
immunoblotting.
as unlikely since cells were washed stringently during immunization differentiation of the MAbs by serological tests is shown in figure 1.
regarded
procedure.
The
Immunoblotting of mouse C3 and mouse C4 RmC3 and RmC4 antibodies were investigated by immunoblotting. One antibody, each of presumed C3 or C4 specifity, was coupled to CNBr-activated Sepharose 4B, and EDTA-plasma from heparinized C57BL/6 mice was applied. The eluted protein was subjected to SDS gel electrophoresis under
reduced conditions and transferred to nitrocellulose. Detection of mouse C3 in the eluate of RmC311H9 using crossreacting goat anti-human C3 showed 2 distinct bands, corresponding to the a- and ßchain of the C3 molecule, whereas the RmC3-llH9 eluate reacted negatively with goat anti-human C4 and the serologicaUy defined anti-mouse C4 MAbs. Anti-C3 reactivity of 22 MAbs out of 28 serologicaUy defined anti-C3 MAbs could be demonstrated by immunoblotting (Fig. 2a). Mouse C4 in the eluate of RmC4-16D2 showed two bands with goat anti-human C4, corresponding to the a- and ß-chain of C4 (the gamma-chain could not be detected), but none with goat anti-human C3 nor with the anti-mouse C3 MAbs. Of 11 MAbs against mouse C4, as demonstrated by serological tests, 7 MAbs were confirmed by immunoblotting (Fig. 2b). Reactivity of the MAbs are summarized in table 1.
Immunohistochemical studies In the absence of fresh mouse serum, labeled RmC3-llH9 was positive in a netlike framework in germinal centers of mouse spleens and furthermore stained single cells as well as aggregations of a few cells in the red pulp. RmC3-llH9 also distinctly labeled the fiber network surrounding central arterioles of white pulp and stained more faintly the white pulp fibers bordering on the marginal zone. RmC3-llH9 bound to T-cell region foUowing a simultaneous incubation of anti-Thy-1 MAb MmTl with fresh C57BL/6 serum as source of complement (Fig.3a). RmC3-llH9 was negative following incubation of anti-Thy-1 MAb MmTl with CVF-treated C57BL/6 serum (Fig.3b). In the absence of fresh mouse serum, RmC4-16D2 stained a coarse network in germinal centers and a
312
B KD •180
KD 180
116
-116
84
-84
ß-i
dg58
-58
-48,5
-48,5
36,5
-36,5
26,6
-26,6
12 3 4 5
6 7 8 9 10
FIGURE 2A: Immunoblot of the eluate of mouse EDTA-plasma applied to immobilized RmC3-llH9. Detection of mouse C3 by: 1) goat anti-human C3, 2) goat anti-human C4, 3) RmC3-llH9, 4) RmC3-
5A5, 5) RmC3-17D5; dg: degradation products.
FIGURE 2B: Immunoblot of the eluate of mouse EDTA-plasma applied to immobilized RmC416D2.Detection of mouse C4 by: 6) goat anti-human C4, 7) goat anti-human C3, 8) RmC4-16D2, 9) RmC4-17Cl, 10) RmC4-17D9.
faint network along the border between white pulp and marginal zone in tissue sections of mouse spleen. Fibers around arterioles were negative. In red pulp, single positive cells were observed. RmC416D2 was bound to T-cell regions following incubation with anti-Thy-1 MAb MmTl and fresh or CVFtreated C57BL/6 serum as source of complement (Fig.3c). DISCUSSION So
far, MAbs to human complement components have been raised by immunizing mice with purified
factors (24,25,26). Monoclonal antibodies to mouse complement have been obtained in the case of C5 by immunizing C5 deficient mice with serum from normal mice (27). Recently we described an immunization scheme for the production of anti-mouse Clq MAbs that avoids the necessity to purify serum-derived Clq and requires only smaU amounts of EDTA-serum as a source of mouse Clq (28). In the present study we used the same principle of immunization, namely the high density of surface Thy-1 antigen on C57BL/6 thymocytes and the intrinsic affinity of mouse Clq to rat IgG2b antibodies bound to mouse thymocytes via Thy-1. The use of freshly prepared serum as a source of complement components activates the classical complement cascade, thus leading to the deposition of cell-bound complement components. In order to circumvent cellular test systems, we established a solid phase ELISA system for detection of anti-mouse complement antibodies. The requirement for such a test system is sufficient adsorption of Clq-affine antibodies (rat IgG2b, mouse IgG2a) to polystyrene microtiter plates in such a way that Clq molecules are able to intercalate between adjacent pairs of Fc portions of antibodies (30-40 nm, 29)
complement
313
TABLE 1:
Reactivity Pattern of Rat anti-Mouse Complement Immunoblotting Techniques MAb
subclass1
C57BL/6 (C4high)
C57BL/6
+ EDTA
+CVF
(C4high)
Antibodies
AKR
CBA
(C41ow, C5 def.)
(C4 low)
as
Revealed
2a 2a Gl
+
+
+
+
+
+
+
(+) (+)
2b
+ + + + +
+
+++
6C9
2b
+
15H1 15H12
M 2b
+ +
16D2 17C1 17D9 31C3
2a M 2a Gl
+
M 2a Gl Gl M
+ +
+
+
(+)
+
+
+
+
+
+
+
+
++ +
+
+
+
+
+
+++
+
+
+ + + + + + +
+
(+) (+)
+
+
+ +
+
+
(+) (+)
(+) (+) (+) (+) (+)
+
+
+
(+)
+ + + + +
and
Immunoblot reactivity2 to C3 to C4
2H9 4A3 4A7 5A5 5E8 5F5 6C11 6E8 6G12 9A11 11B1 11H9 12B6 17C6 17D5 17F5 20G10 23G4 31E8 31D6 31F1 32G2
2a Gl Gl 2a 2a 2a 2a 2a 2a Gl Gl 2a 2a
by Serological
+ +
(+) +
(+)
(+) (+) (+) (+) (+) (+) (+)
+ + + + + + +
1
all MAbs have kappa light chains gradiations in response in immunoblotting: reactivity 2
no, ( + ) weak, +
intermediate,
++
good,
+++
strong
and to induce the classical complement activation. A mouse IgG2a MAb (MmTl) was, therefore, coated to microtiter plates in an optimal concentration of 5 pg/rrA. A 20% dilution of C57BL/6 serum was used to screen the fusion products. The negative control was performed with C57BL/6 EDTAserum. A similar ELISA system was recently applied to detect deficiencies in the classical complement pathway in humans (30). We expected rat MAbs to mouse complement components C3 and C4 because of their ravalent attachment to proteins on the cell surface used for immunization and also deposited on microtiter plates in the screening system during activation. No anti-C2 MAbs were identified. C2 (and also C5b-9, MAC) is only bound non-covalently to the ceU surface and may have been released from the mouse thymocytes before immunization could take place. MAbs to complement
314
¡3*v«
/ ,
V,"' -fi
^m;^.,-7,£*.
