Journal of Immunological Methods. 127 (1990) 263-269
263
Elsevier JIM 05490
A monoclonal antibody which blocks the function of factor D of human complement M a n u e l Pascual ~, E m a n u e l l e C a t a n a 1, F r a n c o i s Spertini 2 K e v i n M a c o n 3, J o h n E. V o l a n a k i s 3 a n d Jiirg A. Schifferli 1 Departments of t Medicine and 2 Pathology, H~pital Cantonal Universitaire. 1211 Geneva 4, .Switzerland, and 3Department of Medicine, University of Alabama, Birmingham. AL, U.S.A.
(Received 24 August 1989, revised received 7 November 1989, accepted 13 November 1989)
Factor D is an essential enzyme for activation of complement by the alternative pathway (AP). It has been difficult to obtain mouse monoclonal antibodies (Mabs) which block the function of factor D. We have developed a strategy to obtain such Mabs using a double screening procedure of the initial clones. We selected the clone whose supernatant had the lowest level of anti-factor D A b by ELISA and abolished factor D haemolytic activity. Addition of this Mab to human serum was shown to abolish conversion of C3 by cobra venom factor, haemolysis of rabbit erythrocytes, and activation of C3 and C5 by cuprophane dialysis membranes. Key words." Complement inhibitor; Factor D; Alternative pathway (of complement): Monoclonal antibody
Introduction Inhibition of the AP of complement might reduce inflammation in many pathological situations, such as myocardial infarction (Maroko et al., 1978) and immune complex diseases (Couser et al., 1985). In vivo, the function of the AP can be abrogated by various techniques: cobra venom factor (CVF) activates the AP and leads to C3 depletion (Inoue, 1987). A similar result can be obtained with anti-factor H antibody (Ab), which deregulates the control of the AP (Veerhuis et al., 1985). However, these methods are not equivalent
Correspondence to: J. Schifferli, Clinique M6dicale, H6pital Cantonal Universitaire, 1211 Geneva 4, Switzerland. Abbreviations: Ab, antibody; AP, alternative pathway; CVF, cobra venom factor; Mabs. monoclonal antibodies; RD, ~rum depleted in factor D.
to blockade of activation since fragments with biological activities are released. Factor D is a highly specific serine protease essential for AP activation (Pangburn and Miiller-Eberhard, 1978). It cleaves factor B bound to C3b, generating the C3bBb enzyme which is the AP C3 convertase. Factor D might be a good target for inhibition, since its plasma concentration is very low (1.8 /~g/ml), and it has been shown to be the limiting enzyme of the AP activation sequence (Lesavre and MiJller-Eberhard, 1978; Volanakis et al., 1985). The aim of this work was to develop a mouse monoclonal anti-factor D A b (Mab) which would abolish the function of human factor D. Previous Mabs have been shown to produce partial inhibition (Niemann et al., 1984). The strategy adopted was to perform a double screening procedure, so as to select the clone producing the Ab with the most favorable profile, i.e. blocking factor D function at the lowest possible concentration.
0022-1759/90/$03.50 © 1990 Elsevier .~ience Publishers B.V. (Biomedical Division)
264 Materials and methods
Purified human factor D, polyclonal Abs and factor D measurements Factor D was purified as described (Volanakis and Macon, 1987). Rabbits were immunized with three injections of 100 ~g of factor D at 2 week intervals using complete (first immunization) and incomplete Freund's adjuvant. The IgG fraction of the serum, harvested 2 weeks after the last immunization, was purified by sodium sulfate precipitation (Heide and Schwick, 1978) followed by ion exchange chromatography on DE-32 cellulose (Whatman, Maidstone U.K.) (Fahey and Terry, 1978). Fab fragments were obtained by hydrolysis with papain followed by chromatography on CM32 cellulose (Whatman) (Stanworth and Turner, 1986). Purity was assessed by SDS-PAGE under reducing and non-reducing conditions (Laemmli, 1970). Normal human serum was depleted of factor D (RD) by immunoaffinity using rabbit IgG anti-factor D coupled to cyanogen bromideactivated Sepharose 4B (Pharmacia, Uppsala, Sweden). Hemolytic factor D was measured by a plate assay using guinea-pig erythrocytes and RD (Harrison and Lachmann, 1986) (limit of sensitivity of the assay: 2.5% of the normal serum concentration of factor D).
Monoclonal antibodies (Mabs) and the screening procedures B A L B / c mice were twice immunized subcutaneously with 20-30 ~g of factor D in complete Freund's adjuvant, with a 3 week interval. Six weeks later, they were boosted intraperitoneally with 90 /xg of factor D. The mice were killed 4 days later and their spleen cells fused with X63Ag8.653 mouse myeloma cells at a ratio of 5.3 : 1 (KiShler and Milstein, 1976; Kearney et al., 1979). Hybridomas were distributed in culture plates on a feeder layer of B A L B / c peritoneal macrophages, and were maintained in Dulbecco's modified Eagle's medium (DMEM) supplemented with 10% fetal calf serum, glutamine, sodium pyruvate, Hepes buffer and penicillin-streptomycin (Gyotoku et al., 1987). Positive wells were cloned by limiting dilution. The first method of screening the initial hybridoma supernatants and subsequent clones was
an ELISA. Purified factor D was diluted in borate buffered saline pH 8.4 (BBS) to a concentration of 5 # g / m l , dispensed into the wells of microtiter plates, incubated for 3 h at 37°C, and washed three times with BBS. After blockade with BBS containing 1 % bovine serum albumin (BSA), 50 #1 of undiluted hybridoma supernatants were added to the wells. The plates were incubated for 2 h at 37°C, and then washed five times with BBS. Anti-factor D A b s were revealed with alkaline phosphatase-conjugated goat anti-mouse (~, chain) Ab (Cappel Lab, Cochranville, PA), in order to select IgG Abs. The second screening method was functional: 40 ~1 of undiluted hybridoma supernatants were mixed with 40 p.1 of phosphatebuffered saline (PBS, pH 7.4, Oxoid, Basinkstoke, U.K.) containing 0.1% BSA and 2 ng of human factor D. The mixtures were incubated for 20 rain at room temperature before residual factor D activity was measured by the haemolytic assay. Pristane primed. B A L B / c mice were injected intraperitoneaily with 7.5 x 10 6 cloned hybridoma cells (Gyotoku et al., 1987); the ascitic fluid was tested for Ab activity, and the IgG was purified by protein A-Sepharose CL-4B (Pharmacia). The purity of the Mabs was assessed by SDS-PAGE and isoelectric focusing. Alkaline phosphatase-conjugated rabbit anti-mouse lgG subclass antibodies (Litton Bionetics, Kensington, MD) were used to determine the subclass of the Mabs.
Specificity of the Mabs (a) Purified factor D (100 ng) was run on SDSPAGE (Laemmli, 1970) followed by transfert to nitrocellulose according to standard procedure. The nitrocellulose was then incubated overnight in a 500 mi solution containing 0.5 mg of the antifactor D Mab. Biotinylated rabbit lgG anti-mouse IgG was used as the second antibody, avidin and biotinylated-horseradish peroxidase were added in sequence, and revealed with DAB (diaminobenzidine) following the manufacturer's instructions (Vector Laboratories, Burlingame, CA 94010). (b) Radiolabelled factor D (320 pg) (Pascual et al., 1988) was mixed with 660 ng of the anti-factor D Mab, in a volume of 50 ~1. After an incubation of 30 rain at 37°C, 5 pl of a 1/100 dilution of anti-mouse IgG1 (kind gift from Prof. S. Izui) were added and the mixture incubated for 24 h at
265 of rabbit erythrocytes was added at time 0. The decrease of turbidity due to cell lysis was recorded over 10 min. (c) The concentration of C3a and C5a in serum incubated with cuprophane dialysis fibers was measured using radioimmunoassays (Amersham, U.K.). Mixtures containing 50 gl of serum and 150 #1 of buffer were incubated with 10 mg of cuprophane fibers at 3 7 ° C for 30 min. The cuprophane fibers were prepared by cutting short fragments ( < 3 mm) of fibers (11.5 #m) of a Gambro hemodialysis filter (Lund, Sweden).
4 ° C. 500 #1 of 1% polyethylene glycol (PEG 6000, Merck Schuchardt, 8011 Hohenbrunn bei Miinchen) were then added. After centrifugation, the radioactivity in the pellet was measured. Controls include an unrelated IgG1 Mab.
Functional assays for the inhibition of A P function in human serum These assays were performed in the presence of the various anti-factor D A b s and relevant controls (normal rabbit IgG or Fab, and an unrelated control Mab anti-TNP IgG1) using as buffer PBS supplemented with 2 mM Mg 2÷ and 10 mM EGTA (final concentration). (a) C3 actioation by CVF (Cordis, Miami, FL) was measured by C3 conversion to C3c using crossed immunoelectrophoresis (Laurell, 1965). Human serum (30 ~1) was incubated at 3 7 ° C for 30 min with 10 #1 of CVF (0.3 mg/ml, 100 inhibitory U / m l ) in buffer (final volume: 100 #1). This assay was capable of detecting > 0.1% of residual factor D activity in serum, as determined from RD supplemented with purified factor D. (b) Lysis of rabbit erythrocytes was assessed in a kinetic assay performed at 3 7 ° C (Polhill et al., 1979). Mixtures containing 100 #1 of normal human serum and 100 #1 of buffer were incubated at 37 ° C in the cell of a Bio/data PAP 4 aggregometer (Hatboro, PA, USA). 5 #1 of a 50% suspension
Results
42 hybridomas were obtained which produced Abs reactive with solid-phase bound human factor D, i.e., optical density values by ELISA above background ( > 0 . 7 5 ) (Fig. 1). These 42 supernatants were tested in the hemolytic assay to define which ones were capable of blocking factor D function. In 30 supernatants there was either no inhibition or partial inhibition. Only 12 supernatants abolished factor D function (Fig. 1). Of these 12 blocking hybridomas, two were chosen for recloning and further analysis (clone 72, low ELISA value and clone 71, high ELISA value) (see Fig. 1). After two cloning procedures (first 0.5
1.50
~t
1.25
8. 0
0.75
0.50
. . . . . . . . . . . . . . .
.........
i
~
l
l
l
l
w
•
|
•
Supernatants
Fig. I. Of the 42 hybridomasselectedby ELISA (optical density > 0.75), 12 inhibited factor D hemolyticactivity (in black). Clone: 72 *, clone 71: **
266
94 Kd
67 Kd
43 Kd d
30 Kd
20 Kd
14 Kd
A
B
Fig. 2. Binding of the Mab 72-96-25 to purified factor D by immunoblot. Lane A: molecular weight markers. Lane B: factor D, identified by Mab 72-96-25. cell/well, second 0.3 cell/well) using the same double screening as described above, two different clones (72-96-25 and 71-22-10, from clone 72 and 71 respectively) were used for the production of ascites fluid in B A L B / c mice. Both Mabs were lgG of subclass 1. The Mab 7 2 - 9 6 - 2 5 was shown to recognize purified factor D by immunoblot (Fig. 2), and to precipitate radiolabelled factor D in a fluid phase immunoassay (42% precipitation with Mab 7 2 - 9 6 - 2 5 and 5% with unrelated control Mab). Both Mabs were then tested and compared to the rabbit Fab anti-factor D in various assays.
Fig. 3 illustrates the effect of the various Ab on the C3 conversion obtained by CVF. Only rabbit Fab anti-factor D (30 p.g of Fab for 30 /zl of serum) and the Mab 72-96-25 at a molar ratio of 80:1 ( M a b / f a c t o r D) inhibited completely the cleavage of C3 to C3c. The second Mab (71-22-10) produced only partial inhibition, and control Abs had no effect. When R D was supplemented with factor D at 0.1% normal concentration some C3 cleavage could be detected. Thus Mab 7 2 - 9 6 - 2 5 blocked >99.9% of factor D function. The lysis of rabbit erythrocytes was abolished by rabbit Fab anti-factor D (50 /~g of Fab for 100 ttl of serum) and by Mab 7 2 - 9 6 - 2 5 at a molar ratio of 40 : 1 ( M a b / f a c t o r D) (Fig. 4). At the same ratio, Mab 7 1 - 2 2 - 1 0 produced only partial inhibition. Lysis was unaffected by the control Abs. The formation of C3a in human serum incubated with cuprophane dialysis fibers was abolished by Mab 7 2 - 9 6 - 2 5 (molar ratio: 40: 1) (Fig. 5a). Even the spontaneous C3a formation seen in serum incubation at 3 7 ° C without cuprophane was abolished, indicating that the "tick-over" of the AP was blocked (Pangburn and Miiller-Eberhard, 1978; Pascuai et al., 1988). Significant CSa production was seen only after 30 min incubation with cuprophane, and this could be substantially decreased by incubation with the Mab 72-96-25, down to control values seen on incubation of normal human serum with buffer (Fig. 5b). The rabbit Fab had similar effects (not shown). In the experiments with the cuprophane dialysis fibers the control Mab caused some inhibition of C3a and C5a productions. Although this was considered to be non specific, there was no convincing explanation for it.
Discussion Complete blockade of human factor D was obtained with a Mab which was selected by a double screening procedure. Specific anti-factor D Abs were detected in m a n y supernatants by ELISA, but only a few were capable of blocking factor D function in the hemolytic assay. Indeed ELISA screening is not ideal for the production of blocking Mabs since the epitopes recognized may
267
C3 C3c
A
NHS + CVF
49
A
h,
NHS + R a b b i t
Fab anti-D
÷ CVF
NXS + M o 7 2 - 9 6 - 2 5
anti-D
+ CVF
NHS + M o 7 1 - 2 2 - I 0
anti-D
+ CVF
A
jx.
Fig. 3. The C3 conversion (C3c formation) induced by CVF in human serum was inhibited by anti-factor D (anti-D) rabbit Fab and Mab (Mo 72-96-25). 0%
NHS + Rabbit Fab a n t l - D NHS * M 0 7 2 - 9 6 - 2 5 a n t l - D
0 vt vl m
E vs
50%
NHS* M071-22-I0
antl-D
NHS
Ioo%p 0
I
I
I
I
I
I
2
4
6
8
I0
12
t i m e in m i n u t e s
Fig. 4. The lysis of rabbit erythrocytes was blocked by the anti-factor D rabbit Fab and Mab 72-96-25 but not by Mab 71-22-10.
268 (pt-ml)
C3a
30-
a
• [] r~l
Ou Cu*72-96-25 Cu+MabCo
[]
r,s-ts
CSa (ng/ml) 300 -
b
•
Cu
[]
Cu+72-96.25
[]
Cu+ M a b C o
O~-~s
20
200 "
10
100 -
0 0
5
5
15 time in rnlnut¢~;
30 time in minute~
Fig. 5. Formation of C3a (a) and C5a (b) by cuprophane dialysis fibers was blocked by Mab 72-96-25. The control Mab: Mab Co.
not be involved in the expression of the enzymatic activity. Furthermore, the antigen bound to the plastic may be in a different conformation than its active fluid phase counterpart. Thus we considered that the most efficient Mab would be selected by choosing the clone whose supernatant gave the lowest ELISA signal but was capable of blocking factor D function. To test for AP blockade by the Mab, we selected an assay that would require minimal factor D activity, i.e., CVF treatment of serum. Once the initial CVFBb enzyme is formed, slow C3 conversion occurs despite blockade of the amplification loop (Pangburn and Miiller-Eberhard, 1978). Indeed, our assay was very sensitive, detecting 0.1% of normal factor D. Lack of C3 conversion in the presence of the Mab, meant almost total blockade of the AP. Activation of the AP was also blocked on cell surfaces (rabbit erythrocytes) which are 'AP activators', and on cuprophane dialysis fibers. This latter assay was chosen because, during hemodialysis, blood is exposed to bio-incompatible membranes, such as cuprophane, which produce complement activation by the AP (Hakim, 1984). Other enzymes have been shown to cleave factor B in vitro (reviewed in Sundsmo and Fair, 1983). Using the polyclonal Fab and monoclonal anti factor D antibodies in our assays, we could not demonstrate the presence of factor D-independent activation of the alternative pathway. This suggests that, in the macromolecular environment
of serum, other enzymes cannot produce detectable AP activation. Another explanation would be that the anti-factor D antibodies cross-reacted with the reactive site of these enzymes. The anti-factor D Mab developed here is certainly not yet the best reagent. To obtain complete blockade of human factor D required a molar ratio of Mab to factor D of 80 : 1. More efficient Mabs are needed. The immunization protocol could be modified, or multiple Mabs might be used simultaneously. However, refining the selection process may be the best option.
Acknowledgements This work was supported by grants from the Fonds National Suisse de la Recherche Scientifique (32-25606.88), and the Swiss Federal Department of Public Health. M.P. is supported by the Sir Jules Thorn Charitable Trust, and J.A.S. by the Max Clo~tta Foundation.
References Couser, W.G., Baker, P.J. and Adler, S. (1985) Mechanisms of glomerular injury in immune complex disease. Kidney Int. 28, 569. Fahey, J.L. and Terry, E.W. (1978) Ion exchange chromatography and gel filtration. In: D.M. Weir (Ed.), Handbook of Experimental Immunology, 3rd edn.. Blackwell, Oxford, Vol. 1, chapter 8.
269 Gyotoku, Y., Abdelmoula, M., Spertini, F., lzui, S. and Lambert, P.H. (1987) Cryoglobulinemia induced by monoclonal immunoglobulin G rheumatoid factors derived from autoimmune MRL/MpJ-Ipr/Ipr mice. J. lmmunol. 138, 3785. Hakim, R.M., Breillatt, J., Lazarus, J.M. and Port, F.K. (1984) Complement activation and hypersensitivity reactions to dialysis membranes. New Engl. J. Med. 311,878. Harrison, R.A. and Lachmann, P.J. (1986) Complement technology, in: D.M. Weir (Ed.), Handbook of Experimental Immunology, 4th edn. Blackwell, Oxford, Vol. 1, chapter 39. Heide, K. and Schwick, H.G. (1978) Salt fractionation of immunoglobulins. In: D.M. Weir (Ed.), Handbook of Experimental Immunology, 3rd edn. Blackwell, Oxford, Vol. 1, chapter 7. Inoue, K. (1987) In vivo manipulation of the complement system. In: K. Rother and G.O. Till (Eds.), The Complement System. Springer-Verlag, Berlin, p. 520. Kearney, J.F., Radbruch, A., Liesegang, B. and Rajewsky, K. (1979) A new mouse myeloma cell line that has lost immunoglobulin expression but permits the construction of antibody-secreting hybrid cell lines. J. lmmunol. 123, 1548. K~hler, G. and Milstein, C. (1976) Derivation of specific antibody-producing tissue culture and tumor lines by cell fusion. Eur. J. Immunol. 6, 511. Laemmli, U.K. (1970) Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature 227, 680. Laurell, C.B. (1965) Quantitative estimation of proteins by electrophoresis in agarose gel containing antibodies. Anal. Biochem. 10, 358. Lesavre, P.H. and Milller-Eberhard, H.J. (1978) Mechanism of action of factor D of the alternative complement pathway. J. Exp. Med. 148, 1498. Maroko, P.R., Carpenter, C.B.. Chiarello, M., Fishbein, M.C.,
Radvany, P., Knostman, J.K. and Hale, S.L. (1978) Reduction by cobra venom factor of myocardial necrosis after coronary artery occlusion. J. Clin. Invest. 61,661. Niemann, M.A., Kearney, J.F. and Volanakis, J.E. (1984) The use of monoclonal antibodies as probes of the three-dimensional structure of human complement factor D. J. lmmunol. 132, 809. Pangburn, M.K. and Miiller-Eberhard, H.J. (1984) The alternative pathway of complement. Springer Sem. lmmunopathol. 7, 163. Pascual, M., Steiger, G., Estreicher, J., Macon, K., Volanakis, J. and Schifferli, J.A. (1988) Metabolism of complement factor D in renal failure. Kidney Int. 34, 529. Polhill, R.B., Pruitt, K.M. and Johnston, R.B. (1979) Kinetic assessment of alternative complement pathway activity in a hemolytic system. I. Experimental and mathematical analyses. J. Immunol. 121,363. Stanworth, D.R. and Turner, M.W. (1986) Immunochemical analysis of human and rabbit immunoglobulins and their subunits. In: D.M. Weir (Ed.), Handbook of Experimental Immunology, 4th edn. Blackwell, Oxford, Vol. 1, Chapter 12. Sundsmo, J.S. and Fair, D.S. (1983) Relationships among the complement, kinin, coagulation, and fibrinolytic systems. Springer Semin. lmmunopathol. 6, 231-258. Veerhuis, R., Van Es, L.A. and Daha, M.R. (1985) In vivo modulation of rat complement activities by infusion of anti-H antibodies. Immunobiology 170, 133. Volanakis, J.E., Barnum, S.R., Giddens, M. and Galla, J.H. (1985) Renal filtration and catabolism of complement protein D. New Engl. J. Med. 312, 395. Volanakis, J.E. and Macon, K.J. (1987) Isolation of complement protein D from urine of patients with Fanconi's syndrome. Anal. Biochem. 163, 242.
263
Elsevier JIM 05490
A monoclonal antibody which blocks the function of factor D of human complement M a n u e l Pascual ~, E m a n u e l l e C a t a n a 1, F r a n c o i s Spertini 2 K e v i n M a c o n 3, J o h n E. V o l a n a k i s 3 a n d Jiirg A. Schifferli 1 Departments of t Medicine and 2 Pathology, H~pital Cantonal Universitaire. 1211 Geneva 4, .Switzerland, and 3Department of Medicine, University of Alabama, Birmingham. AL, U.S.A.
(Received 24 August 1989, revised received 7 November 1989, accepted 13 November 1989)
Factor D is an essential enzyme for activation of complement by the alternative pathway (AP). It has been difficult to obtain mouse monoclonal antibodies (Mabs) which block the function of factor D. We have developed a strategy to obtain such Mabs using a double screening procedure of the initial clones. We selected the clone whose supernatant had the lowest level of anti-factor D A b by ELISA and abolished factor D haemolytic activity. Addition of this Mab to human serum was shown to abolish conversion of C3 by cobra venom factor, haemolysis of rabbit erythrocytes, and activation of C3 and C5 by cuprophane dialysis membranes. Key words." Complement inhibitor; Factor D; Alternative pathway (of complement): Monoclonal antibody
Introduction Inhibition of the AP of complement might reduce inflammation in many pathological situations, such as myocardial infarction (Maroko et al., 1978) and immune complex diseases (Couser et al., 1985). In vivo, the function of the AP can be abrogated by various techniques: cobra venom factor (CVF) activates the AP and leads to C3 depletion (Inoue, 1987). A similar result can be obtained with anti-factor H antibody (Ab), which deregulates the control of the AP (Veerhuis et al., 1985). However, these methods are not equivalent
Correspondence to: J. Schifferli, Clinique M6dicale, H6pital Cantonal Universitaire, 1211 Geneva 4, Switzerland. Abbreviations: Ab, antibody; AP, alternative pathway; CVF, cobra venom factor; Mabs. monoclonal antibodies; RD, ~rum depleted in factor D.
to blockade of activation since fragments with biological activities are released. Factor D is a highly specific serine protease essential for AP activation (Pangburn and Miiller-Eberhard, 1978). It cleaves factor B bound to C3b, generating the C3bBb enzyme which is the AP C3 convertase. Factor D might be a good target for inhibition, since its plasma concentration is very low (1.8 /~g/ml), and it has been shown to be the limiting enzyme of the AP activation sequence (Lesavre and MiJller-Eberhard, 1978; Volanakis et al., 1985). The aim of this work was to develop a mouse monoclonal anti-factor D A b (Mab) which would abolish the function of human factor D. Previous Mabs have been shown to produce partial inhibition (Niemann et al., 1984). The strategy adopted was to perform a double screening procedure, so as to select the clone producing the Ab with the most favorable profile, i.e. blocking factor D function at the lowest possible concentration.
0022-1759/90/$03.50 © 1990 Elsevier .~ience Publishers B.V. (Biomedical Division)
264 Materials and methods
Purified human factor D, polyclonal Abs and factor D measurements Factor D was purified as described (Volanakis and Macon, 1987). Rabbits were immunized with three injections of 100 ~g of factor D at 2 week intervals using complete (first immunization) and incomplete Freund's adjuvant. The IgG fraction of the serum, harvested 2 weeks after the last immunization, was purified by sodium sulfate precipitation (Heide and Schwick, 1978) followed by ion exchange chromatography on DE-32 cellulose (Whatman, Maidstone U.K.) (Fahey and Terry, 1978). Fab fragments were obtained by hydrolysis with papain followed by chromatography on CM32 cellulose (Whatman) (Stanworth and Turner, 1986). Purity was assessed by SDS-PAGE under reducing and non-reducing conditions (Laemmli, 1970). Normal human serum was depleted of factor D (RD) by immunoaffinity using rabbit IgG anti-factor D coupled to cyanogen bromideactivated Sepharose 4B (Pharmacia, Uppsala, Sweden). Hemolytic factor D was measured by a plate assay using guinea-pig erythrocytes and RD (Harrison and Lachmann, 1986) (limit of sensitivity of the assay: 2.5% of the normal serum concentration of factor D).
Monoclonal antibodies (Mabs) and the screening procedures B A L B / c mice were twice immunized subcutaneously with 20-30 ~g of factor D in complete Freund's adjuvant, with a 3 week interval. Six weeks later, they were boosted intraperitoneally with 90 /xg of factor D. The mice were killed 4 days later and their spleen cells fused with X63Ag8.653 mouse myeloma cells at a ratio of 5.3 : 1 (KiShler and Milstein, 1976; Kearney et al., 1979). Hybridomas were distributed in culture plates on a feeder layer of B A L B / c peritoneal macrophages, and were maintained in Dulbecco's modified Eagle's medium (DMEM) supplemented with 10% fetal calf serum, glutamine, sodium pyruvate, Hepes buffer and penicillin-streptomycin (Gyotoku et al., 1987). Positive wells were cloned by limiting dilution. The first method of screening the initial hybridoma supernatants and subsequent clones was
an ELISA. Purified factor D was diluted in borate buffered saline pH 8.4 (BBS) to a concentration of 5 # g / m l , dispensed into the wells of microtiter plates, incubated for 3 h at 37°C, and washed three times with BBS. After blockade with BBS containing 1 % bovine serum albumin (BSA), 50 #1 of undiluted hybridoma supernatants were added to the wells. The plates were incubated for 2 h at 37°C, and then washed five times with BBS. Anti-factor D A b s were revealed with alkaline phosphatase-conjugated goat anti-mouse (~, chain) Ab (Cappel Lab, Cochranville, PA), in order to select IgG Abs. The second screening method was functional: 40 ~1 of undiluted hybridoma supernatants were mixed with 40 p.1 of phosphatebuffered saline (PBS, pH 7.4, Oxoid, Basinkstoke, U.K.) containing 0.1% BSA and 2 ng of human factor D. The mixtures were incubated for 20 rain at room temperature before residual factor D activity was measured by the haemolytic assay. Pristane primed. B A L B / c mice were injected intraperitoneaily with 7.5 x 10 6 cloned hybridoma cells (Gyotoku et al., 1987); the ascitic fluid was tested for Ab activity, and the IgG was purified by protein A-Sepharose CL-4B (Pharmacia). The purity of the Mabs was assessed by SDS-PAGE and isoelectric focusing. Alkaline phosphatase-conjugated rabbit anti-mouse lgG subclass antibodies (Litton Bionetics, Kensington, MD) were used to determine the subclass of the Mabs.
Specificity of the Mabs (a) Purified factor D (100 ng) was run on SDSPAGE (Laemmli, 1970) followed by transfert to nitrocellulose according to standard procedure. The nitrocellulose was then incubated overnight in a 500 mi solution containing 0.5 mg of the antifactor D Mab. Biotinylated rabbit lgG anti-mouse IgG was used as the second antibody, avidin and biotinylated-horseradish peroxidase were added in sequence, and revealed with DAB (diaminobenzidine) following the manufacturer's instructions (Vector Laboratories, Burlingame, CA 94010). (b) Radiolabelled factor D (320 pg) (Pascual et al., 1988) was mixed with 660 ng of the anti-factor D Mab, in a volume of 50 ~1. After an incubation of 30 rain at 37°C, 5 pl of a 1/100 dilution of anti-mouse IgG1 (kind gift from Prof. S. Izui) were added and the mixture incubated for 24 h at
265 of rabbit erythrocytes was added at time 0. The decrease of turbidity due to cell lysis was recorded over 10 min. (c) The concentration of C3a and C5a in serum incubated with cuprophane dialysis fibers was measured using radioimmunoassays (Amersham, U.K.). Mixtures containing 50 gl of serum and 150 #1 of buffer were incubated with 10 mg of cuprophane fibers at 3 7 ° C for 30 min. The cuprophane fibers were prepared by cutting short fragments ( < 3 mm) of fibers (11.5 #m) of a Gambro hemodialysis filter (Lund, Sweden).
4 ° C. 500 #1 of 1% polyethylene glycol (PEG 6000, Merck Schuchardt, 8011 Hohenbrunn bei Miinchen) were then added. After centrifugation, the radioactivity in the pellet was measured. Controls include an unrelated IgG1 Mab.
Functional assays for the inhibition of A P function in human serum These assays were performed in the presence of the various anti-factor D A b s and relevant controls (normal rabbit IgG or Fab, and an unrelated control Mab anti-TNP IgG1) using as buffer PBS supplemented with 2 mM Mg 2÷ and 10 mM EGTA (final concentration). (a) C3 actioation by CVF (Cordis, Miami, FL) was measured by C3 conversion to C3c using crossed immunoelectrophoresis (Laurell, 1965). Human serum (30 ~1) was incubated at 3 7 ° C for 30 min with 10 #1 of CVF (0.3 mg/ml, 100 inhibitory U / m l ) in buffer (final volume: 100 #1). This assay was capable of detecting > 0.1% of residual factor D activity in serum, as determined from RD supplemented with purified factor D. (b) Lysis of rabbit erythrocytes was assessed in a kinetic assay performed at 3 7 ° C (Polhill et al., 1979). Mixtures containing 100 #1 of normal human serum and 100 #1 of buffer were incubated at 37 ° C in the cell of a Bio/data PAP 4 aggregometer (Hatboro, PA, USA). 5 #1 of a 50% suspension
Results
42 hybridomas were obtained which produced Abs reactive with solid-phase bound human factor D, i.e., optical density values by ELISA above background ( > 0 . 7 5 ) (Fig. 1). These 42 supernatants were tested in the hemolytic assay to define which ones were capable of blocking factor D function. In 30 supernatants there was either no inhibition or partial inhibition. Only 12 supernatants abolished factor D function (Fig. 1). Of these 12 blocking hybridomas, two were chosen for recloning and further analysis (clone 72, low ELISA value and clone 71, high ELISA value) (see Fig. 1). After two cloning procedures (first 0.5
1.50
~t
1.25
8. 0
0.75
0.50
. . . . . . . . . . . . . . .
.........
i
~
l
l
l
l
w
•
|
•
Supernatants
Fig. I. Of the 42 hybridomasselectedby ELISA (optical density > 0.75), 12 inhibited factor D hemolyticactivity (in black). Clone: 72 *, clone 71: **
266
94 Kd
67 Kd
43 Kd d
30 Kd
20 Kd
14 Kd
A
B
Fig. 2. Binding of the Mab 72-96-25 to purified factor D by immunoblot. Lane A: molecular weight markers. Lane B: factor D, identified by Mab 72-96-25. cell/well, second 0.3 cell/well) using the same double screening as described above, two different clones (72-96-25 and 71-22-10, from clone 72 and 71 respectively) were used for the production of ascites fluid in B A L B / c mice. Both Mabs were lgG of subclass 1. The Mab 7 2 - 9 6 - 2 5 was shown to recognize purified factor D by immunoblot (Fig. 2), and to precipitate radiolabelled factor D in a fluid phase immunoassay (42% precipitation with Mab 7 2 - 9 6 - 2 5 and 5% with unrelated control Mab). Both Mabs were then tested and compared to the rabbit Fab anti-factor D in various assays.
Fig. 3 illustrates the effect of the various Ab on the C3 conversion obtained by CVF. Only rabbit Fab anti-factor D (30 p.g of Fab for 30 /zl of serum) and the Mab 72-96-25 at a molar ratio of 80:1 ( M a b / f a c t o r D) inhibited completely the cleavage of C3 to C3c. The second Mab (71-22-10) produced only partial inhibition, and control Abs had no effect. When R D was supplemented with factor D at 0.1% normal concentration some C3 cleavage could be detected. Thus Mab 7 2 - 9 6 - 2 5 blocked >99.9% of factor D function. The lysis of rabbit erythrocytes was abolished by rabbit Fab anti-factor D (50 /~g of Fab for 100 ttl of serum) and by Mab 7 2 - 9 6 - 2 5 at a molar ratio of 40 : 1 ( M a b / f a c t o r D) (Fig. 4). At the same ratio, Mab 7 1 - 2 2 - 1 0 produced only partial inhibition. Lysis was unaffected by the control Abs. The formation of C3a in human serum incubated with cuprophane dialysis fibers was abolished by Mab 7 2 - 9 6 - 2 5 (molar ratio: 40: 1) (Fig. 5a). Even the spontaneous C3a formation seen in serum incubation at 3 7 ° C without cuprophane was abolished, indicating that the "tick-over" of the AP was blocked (Pangburn and Miiller-Eberhard, 1978; Pascuai et al., 1988). Significant CSa production was seen only after 30 min incubation with cuprophane, and this could be substantially decreased by incubation with the Mab 72-96-25, down to control values seen on incubation of normal human serum with buffer (Fig. 5b). The rabbit Fab had similar effects (not shown). In the experiments with the cuprophane dialysis fibers the control Mab caused some inhibition of C3a and C5a productions. Although this was considered to be non specific, there was no convincing explanation for it.
Discussion Complete blockade of human factor D was obtained with a Mab which was selected by a double screening procedure. Specific anti-factor D Abs were detected in m a n y supernatants by ELISA, but only a few were capable of blocking factor D function in the hemolytic assay. Indeed ELISA screening is not ideal for the production of blocking Mabs since the epitopes recognized may
267
C3 C3c
A
NHS + CVF
49
A
h,
NHS + R a b b i t
Fab anti-D
÷ CVF
NXS + M o 7 2 - 9 6 - 2 5
anti-D
+ CVF
NHS + M o 7 1 - 2 2 - I 0
anti-D
+ CVF
A
jx.
Fig. 3. The C3 conversion (C3c formation) induced by CVF in human serum was inhibited by anti-factor D (anti-D) rabbit Fab and Mab (Mo 72-96-25). 0%
NHS + Rabbit Fab a n t l - D NHS * M 0 7 2 - 9 6 - 2 5 a n t l - D
0 vt vl m
E vs
50%
NHS* M071-22-I0
antl-D
NHS
Ioo%p 0
I
I
I
I
I
I
2
4
6
8
I0
12
t i m e in m i n u t e s
Fig. 4. The lysis of rabbit erythrocytes was blocked by the anti-factor D rabbit Fab and Mab 72-96-25 but not by Mab 71-22-10.
268 (pt-ml)
C3a
30-
a
• [] r~l
Ou Cu*72-96-25 Cu+MabCo
[]
r,s-ts
CSa (ng/ml) 300 -
b
•
Cu
[]
Cu+72-96.25
[]
Cu+ M a b C o
O~-~s
20
200 "
10
100 -
0 0
5
5
15 time in rnlnut¢~;
30 time in minute~
Fig. 5. Formation of C3a (a) and C5a (b) by cuprophane dialysis fibers was blocked by Mab 72-96-25. The control Mab: Mab Co.
not be involved in the expression of the enzymatic activity. Furthermore, the antigen bound to the plastic may be in a different conformation than its active fluid phase counterpart. Thus we considered that the most efficient Mab would be selected by choosing the clone whose supernatant gave the lowest ELISA signal but was capable of blocking factor D function. To test for AP blockade by the Mab, we selected an assay that would require minimal factor D activity, i.e., CVF treatment of serum. Once the initial CVFBb enzyme is formed, slow C3 conversion occurs despite blockade of the amplification loop (Pangburn and Miiller-Eberhard, 1978). Indeed, our assay was very sensitive, detecting 0.1% of normal factor D. Lack of C3 conversion in the presence of the Mab, meant almost total blockade of the AP. Activation of the AP was also blocked on cell surfaces (rabbit erythrocytes) which are 'AP activators', and on cuprophane dialysis fibers. This latter assay was chosen because, during hemodialysis, blood is exposed to bio-incompatible membranes, such as cuprophane, which produce complement activation by the AP (Hakim, 1984). Other enzymes have been shown to cleave factor B in vitro (reviewed in Sundsmo and Fair, 1983). Using the polyclonal Fab and monoclonal anti factor D antibodies in our assays, we could not demonstrate the presence of factor D-independent activation of the alternative pathway. This suggests that, in the macromolecular environment
of serum, other enzymes cannot produce detectable AP activation. Another explanation would be that the anti-factor D antibodies cross-reacted with the reactive site of these enzymes. The anti-factor D Mab developed here is certainly not yet the best reagent. To obtain complete blockade of human factor D required a molar ratio of Mab to factor D of 80 : 1. More efficient Mabs are needed. The immunization protocol could be modified, or multiple Mabs might be used simultaneously. However, refining the selection process may be the best option.
Acknowledgements This work was supported by grants from the Fonds National Suisse de la Recherche Scientifique (32-25606.88), and the Swiss Federal Department of Public Health. M.P. is supported by the Sir Jules Thorn Charitable Trust, and J.A.S. by the Max Clo~tta Foundation.
References Couser, W.G., Baker, P.J. and Adler, S. (1985) Mechanisms of glomerular injury in immune complex disease. Kidney Int. 28, 569. Fahey, J.L. and Terry, E.W. (1978) Ion exchange chromatography and gel filtration. In: D.M. Weir (Ed.), Handbook of Experimental Immunology, 3rd edn.. Blackwell, Oxford, Vol. 1, chapter 8.
269 Gyotoku, Y., Abdelmoula, M., Spertini, F., lzui, S. and Lambert, P.H. (1987) Cryoglobulinemia induced by monoclonal immunoglobulin G rheumatoid factors derived from autoimmune MRL/MpJ-Ipr/Ipr mice. J. lmmunol. 138, 3785. Hakim, R.M., Breillatt, J., Lazarus, J.M. and Port, F.K. (1984) Complement activation and hypersensitivity reactions to dialysis membranes. New Engl. J. Med. 311,878. Harrison, R.A. and Lachmann, P.J. (1986) Complement technology, in: D.M. Weir (Ed.), Handbook of Experimental Immunology, 4th edn. Blackwell, Oxford, Vol. 1, chapter 39. Heide, K. and Schwick, H.G. (1978) Salt fractionation of immunoglobulins. In: D.M. Weir (Ed.), Handbook of Experimental Immunology, 3rd edn. Blackwell, Oxford, Vol. 1, chapter 7. Inoue, K. (1987) In vivo manipulation of the complement system. In: K. Rother and G.O. Till (Eds.), The Complement System. Springer-Verlag, Berlin, p. 520. Kearney, J.F., Radbruch, A., Liesegang, B. and Rajewsky, K. (1979) A new mouse myeloma cell line that has lost immunoglobulin expression but permits the construction of antibody-secreting hybrid cell lines. J. lmmunol. 123, 1548. K~hler, G. and Milstein, C. (1976) Derivation of specific antibody-producing tissue culture and tumor lines by cell fusion. Eur. J. Immunol. 6, 511. Laemmli, U.K. (1970) Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature 227, 680. Laurell, C.B. (1965) Quantitative estimation of proteins by electrophoresis in agarose gel containing antibodies. Anal. Biochem. 10, 358. Lesavre, P.H. and Milller-Eberhard, H.J. (1978) Mechanism of action of factor D of the alternative complement pathway. J. Exp. Med. 148, 1498. Maroko, P.R., Carpenter, C.B.. Chiarello, M., Fishbein, M.C.,
Radvany, P., Knostman, J.K. and Hale, S.L. (1978) Reduction by cobra venom factor of myocardial necrosis after coronary artery occlusion. J. Clin. Invest. 61,661. Niemann, M.A., Kearney, J.F. and Volanakis, J.E. (1984) The use of monoclonal antibodies as probes of the three-dimensional structure of human complement factor D. J. lmmunol. 132, 809. Pangburn, M.K. and Miiller-Eberhard, H.J. (1984) The alternative pathway of complement. Springer Sem. lmmunopathol. 7, 163. Pascual, M., Steiger, G., Estreicher, J., Macon, K., Volanakis, J. and Schifferli, J.A. (1988) Metabolism of complement factor D in renal failure. Kidney Int. 34, 529. Polhill, R.B., Pruitt, K.M. and Johnston, R.B. (1979) Kinetic assessment of alternative complement pathway activity in a hemolytic system. I. Experimental and mathematical analyses. J. Immunol. 121,363. Stanworth, D.R. and Turner, M.W. (1986) Immunochemical analysis of human and rabbit immunoglobulins and their subunits. In: D.M. Weir (Ed.), Handbook of Experimental Immunology, 4th edn. Blackwell, Oxford, Vol. 1, Chapter 12. Sundsmo, J.S. and Fair, D.S. (1983) Relationships among the complement, kinin, coagulation, and fibrinolytic systems. Springer Semin. lmmunopathol. 6, 231-258. Veerhuis, R., Van Es, L.A. and Daha, M.R. (1985) In vivo modulation of rat complement activities by infusion of anti-H antibodies. Immunobiology 170, 133. Volanakis, J.E., Barnum, S.R., Giddens, M. and Galla, J.H. (1985) Renal filtration and catabolism of complement protein D. New Engl. J. Med. 312, 395. Volanakis, J.E. and Macon, K.J. (1987) Isolation of complement protein D from urine of patients with Fanconi's syndrome. Anal. Biochem. 163, 242.
