- Email: [email protected]
Enzyme-linked immunosorbent assay for C1q in human serum by use of monoclonal antibodies
Journal of Immunological Methods, 74 (1984) 299- 306
299
Elsevier
JIM 03268
Enzyme-Linked Immunosorbent Assay for Clq in Human Serum by Use of Monoclonal Antibodies Ursula Antes, Hans-Peter Heinz and Michael Loos * Institute of Medical Microbiologr', Johannes Gutenberg University, Hochhaus am A ugustusplatz, 6500 Mainz, F.R.G.
(Received 12 June 1984, accepted 24 July 1984)
A sandwich ELISA system has been developed for the detection of C l q in h u m a n serum. It is specific, uses monoclonal antibodies, is sensitive into the nanogram range and is rapidly performed. Therefore, it may be a helpful tool for clinical routine diagnosis, e.g., detecting abnormal C l q levels in patients with rheumatic disorders. Various combinations of poly- and monoclonal antibodies were tested in a sandwich assay. One of these combinations, in particular, resulted in a highly reproducible standard curve: C l q bound to solid-phase polyclonal anti-Clq was detected by the monoclonal antibody 242 G3. In this assay, the Clq concentration in sera of normal individuals was found to be 160 # g / m l (mean value of 70 normal h u m a n sera). This ELISA detected nanogram levels of C l q and gave results comparable to those obtained by haemolytic C l q titration. One nanogram of C l q corresponded to ca. 2.6 x 101° effective C l q molecules. With this technique, selective C l q deficient sera as well as sera from patients with rheumatoid diseases were analysed. Key words: CI q- E L I S A - monoclonal antibody - Cl q in N H S and patients' sera
Introduction
Clq, a subcomponent of the first complement component initiates activation of the complement cascade via the classical pathway by binding to the Fc region of antibodies (reviewed by Loos, 1982). Therefore, this component plays an important role in complement activation by immune complexes. Determination of Clq content is of clinical interest (e.g., in patients with rheumatic disorders), in diagnostics and as a guide for therapy. To date, exact amounts of Clq have been measured in a functional haemolytic assay. Compared to the haemolytic assay, radial immunoassays for Clq are very insensitive. Therefore, in order to measure Clq immunochemically, an enzyme-lin-
* Mailing address: Dr. Michael Loos, Institut ftir Med. Mikrobiologie, Johannes Gutenberg-Universit~t, Hochhaus Augustusplatz, D-6500 Mainz, F.R.G. 0022-1759/84/$03.00 © 1984 Elsevier Science Publishers B.V.
300 ked immunosorbent assay (ELISA) was developed for the detection of C l q antigen in human serum. The original ELISA system, first described by Engvall and Perlmann (1972) was modified using different solid-phase antibodies in a sandwich technique. Of several mouse monoclonal antibodies to human C l q available in our laboratory (Golan et al., 1982), only one was found to be suitable for an ELISA sandwich technique. This antibody recognizes only the C chain and not the A and B chains of C l q (Heinz et al., 1984). The aim of this study was to develop a rapid, specific, sensitive C l q assay which would be economical and technically simple and correlate well with the haemolytic assay for Clq. This report describes the method in detail, shows the effect of altering various test parameters and gives the results of C l q in normal and in patients' sera.
Materials and methods
Antiserum tO human C l q was obtained from rabbits (Behringwerke, Marburg) or goats (Atlantic Antibodies, Scarborough). Peroxidase-conjugated rabbit immunoglobulins to mouse immunoglobulins and peroxidase-conjugated swine immunoglobulins to rabbit immunoglobulins were obtained from Dako, Copenhagen. Monoclonal anti-human C l q antibodies were raised in B A L B / c mice as recently described by Golan et al. (1982). The IgG fraction of the monoclonal antibodies was obtained by ammonium-sulphate precipitation and subsequent chromatography on a D E A E 52 column (Heinz et al., 1984). Serum samples were obtained from normal healthy donors and from patients with rheumatoid diseases and selective C l q deficiency (Berkel et al., 1979; Loos et al., 1980). Protein concentrations were determined by the method of Lowry et al. (1951) with bovine serum albumin acting as standard. The haemolytic activity of C l q was determined as described by Golan et al. (1981).
C/q Two different C l q preparations were used as C l q standards. The first preparation was purified from normal human serum according to the procedure of Volanakis and Stroud (1972). The second purification was carried out according to Stemmer and Loos (1984), using fast protein liquid chromatography (FPLC).
ELISA-procedure The following ELISA procedure was found to be the most suitable. All reagents and test samples were diluted in 10 mM phosphate-buffered saline (PBS), p H 7.5, of ionic strength 15 mS, unless stated otherwise. The assays were performed at room temperature unless otherwise stated. Wells of polystyrol microtitration plates ( N U N C - I M M U N O , 96 F-Plates No. 439454, Wiesbaden) were coated with rabbit antiserum to human Clq (Behring-
301 werke, Marburg) by addition of 20 ~tl of a 1 : 100 dilution, protein concentration 40 mg/ml. The plates were incubated for 1 h, washed in PBS pH 7.5 containing 0.5% BSA (bovine serum albumin, No. A 2153, Sigma, Munich) and left in the same buffer for 1 h to block remaining binding sites. The plates were used immediately or stored at 4°C. Purified human C l q (protein concentration 1.6 m g / m l ) at different dilutions or serum samples were added (10 t~l per well) and incubated for 1 h. After washing twice with 0.5% PBS-BSA, the plates were incubated with a monoclonal mouse antibody to human C l q (20/~l/well of a 1 : 200 dilution, protein concentration 1.7 m g / m l ) for 1 h and washed twice as described above. Bound human C l q antibodies were detected with peroxidase-conjugated rabbit anti-mouse IgG (10/~1 of a 1 : 200 dilution added to each well, protein concentration 5 m g / m l ) (Dako, Copenhagen). After 1 h, the plates were washed twice and 20/~1 50 mM ABTS substrate was added, diluted in PBS pH 6.0, containing 25 btl/ml of a 0.1% hydrogen peroxide solution. ABTS (2,2'-azino-di(ethyl-benz-thiazoline-6sulphonic acid) diammonium salt) was obtained from Serva, No. 14364, Heidelberg. The colour reaction was stopped after 30 min at 37°C by addition of 100 ~tl of a 0.0125% sodium azide solution in PBS (pH 6.0). Absorbance values were measured with a Titertek Multiscan MC (Flow Laboratories, Meckenheim) at 414 nm against a substrate solution blank.
Results
Combinations of antibodies The following combinations of antibodies were tested to establish the optimum conditions for the CIq-ELISA. (1) Solid-phase polyclonal goat anti-human C l q as the first and polyclonal rabbit anti-human C l q as the second step antibody. (2) Solid-phase monoclonal mouse anti-human C l q as the first and polyclonal rabbit anti-human C l q as the second step. (3) Solid-phase polyclonal rabbit anti-human C l q and monoclonai mouse antihum'an C l q for detection. The results in Fig. 1 show that the third combination was the most effective for the detection of C l q in serum, and it remained the method of choice for subsequent tests. Different monoclonal antibodies Comparing different monoclonal antibodies against purified human C l q (Golan et al., 1982), it was found that one antibody in particular (242 G3) was very useful in this ELISA system. Fig. 2 demonstrates the reaction of 5 monoclonal antibodies against human-Clq. Monoclonal antibodies 244 G8 and 244 H4 showed very high background values due to a high non-specific binding. With two other monoclonal antibodies (239 F10 and 241 F l l ) , differences in absorbance values were not high enough at the concentration used to standardize the C l q content of serum.
302
Optimum assay conditions Once the antibodies, the best sandwich combination and the Clq standard had been established, optimal conditions of the following test parameters were investigated: concentrations and volume of antibodies, test samples and reagents, incubation time. Figs. 3 5 show the effect of a number of variations. The amount of polyclonal antibody for coating the plates (Fig. 3) was varied from 2 to 800 /~g, the optimal concentration was found to be 8-20 /~g per well. Fig. 4 shows the same for the amount of monoclonal antibody 242 G3. It was used as a second step antibody in the sandwich in a dilution of ] : 200 which resulted in about 8 /~g/well, obviously enough to detect antibody-bound Clq. The effects of changing the incubation time of the polyclonal antibody used for coating and the BSA buffer (A), Clq antigen (B), monoclonal antibody from the mouse (C) and anti-mouse peroxidase-conjugated antibody (D) are shown in Fig. 5. An incubation time of 60 min was found to be sufficient for all of the antibodies and antigens used.
1600
EI.,~
1400
1400
J
1200
1200
lO0O
1000
800
800
600
600
400
/-,00
E~
/ /
~
I)2/,2G3
/
2MFI1 239 FIO
200
200
0
0'1
o's
1'0
,s
concentration of purified human Clq (pg/m[)
0
0'1
10
106
concentration of human Clq (~g/m[)
Fig. l. Combinations of antibodies for a sandwich Clq-ELISA. Microtitration plates were coated with antibody I, followed by saturation with BSA buffer. Bound C | q was detected by antibody lI. Antibody II was detected by an excess of peroxidase-conjugated anti-mouse antibody or by peroxidase-conjugated anti-rabbit antibody respectively. Working dilutions for each individual system were optimized in a test system using Clq-coated plates, antibody II and peroxidase-conjugated antibodies, zx zx, antibody I: solid-phase polyclonal goat anti-human Clq; antibody II: polyclonal rabbit anti-human Clq. • •, antibody I: solid-phase monoclonal mouse anti-human Clq; antibody I|: polyclonal rabbit anti-human Clq. O O, antibody I: solid-phase polyclonal rabbit anti-human Clq; antibody lI: monoclonal mouse anti-human Clq. Fig. 2. Comparison of 5 monoclonal anti-human C l q antibodies used in ELISA. Plates were coated with polyclonal rabbit anti-human C l q antibody. Bound C l q was detected by 5 different monoclonal mouse anti-human C l q antibodies.
303
800
E ~,~
700 E,,,
1100
600
1000
5OO
900
400
800
300/
/
200
700
~o
100
600 i
i
0
)8o
~o
1'o
~-Clq monoclonal(~g)
ct Clq polyclonal (pg)
Fig. 3. Determination of an optimal amount of polyclonal rabbit anti-human C l q for coating the plates. Fig. 4. Determination of an optimal amount of monoclonal mouse anti-human C l q (242 G3) used as second antibody in ELISA. . E ,,,
B
1 200 -
1000"
800-
t,O0
r ~
/
600-
"~
-
~oo 200-
20O 0
0 I ~,110
2; ,0 ~o E,,
tl.,~°l C
0
20
i~.oo" E,),
t.O
60
t (rain) D
1200.
121111
1000-
t 000
800.
600-
61111
/,00"
/
200'
21111 j
0
0
0
Fig. 5. Effect of incubation time on A: BSA buffer used for blocking non-specific binding sites (Lx zx) after incubation of the plates with polyclonal rabbit anti-human C l q (D D) and before exposure to purified Clq. B: binding of C l q to solid-phase rabbit antihuman Clq. C: binding of monoclonal mouse anti-human C l q to detect bound Clq. D: detection of mouse monoclonal anti-Clq antibodies with peroxidase-conjugated anti-mouse IgG.
304
Cl q standard Using these conditions, 2 preparations of purified C l q were assayed with equivalent results to obtain a standard curve. A linear relationship was found between absorbance values and C l q concentration on a semi-log scale, ranging from 0.1 to 5.0/~g C l q per ml. Therefore test sample values of C l q can easily be extrapolated and calculated. The ELISA for Clq in serum is linear at least between serum dilutions of 1 : 100 and 1 : 4000. The addition of EDTA to the sera for dissociation of the macromolecular C] complex was unnecessary. In controls omitting one part of the sandwich it was shown that non-specific binding of Clq to the antibodies did not occur. Fig. 6 demonstrates the Clq standard curve and the calculation of Clq content in serum test samples. C1 q was detected down into the nanogram range and in normal human sera the mean value of Clq concentration was found to be 160 /~g/ml.
Reproducibifity The ELISA system was found to be reproducible and in order to make it both more practical and more rapid, tests were carried out with precoated plates. The plates covered with lids were stored in a refrigerator. Before the tests were performed the volume was adjusted to 1 0 0 / H / w e l l with PBS, pH 7.5. As shown in Fig. 7, the antibody/BSA-coated plates could be stored at 4°C for several weeks without loss of antibody activity. Storing at room temperature resulted in an increase of background absorbance values.
1400
1200
1200
E~
o//
1000
o /~/.Hs //
1000
800
/
800 /
600
600 ~00
/ /5
200
/
,/
/
4OO 200
0
0 0
01
10
100
concentration of purified human Clq (#g/ml)
0
0.1
10
100
concentration of purified human C|q (~Jg/m[)
Fig. 6. Standard curve of highly purified human Clq (protein concentration 1.6 m g / m l ) and calculation of C l q content in serum test samples diluted 1 : 100. Purified human C l q in different dilutions or serum samples were added to polyclonal anti-human Clq-coated plates (10 p,1/well) and incubated for 1 h. After washing twice, the plates were incubated with a monoclonal mouse antibody to human Clq for 1 h and washed twice. Bound human C l q antibodies were detected with peroxidase-conjugated rabbit anti-mouse IgG. Fig. 7. Reproducibility of the C l q standard. Microtitration plates were precoated with polyclonal rabbit anti-human C l q antiserum, washed in 0.5% BSA-PBS and kept in the same buffer. The plates were stored at 4°C, determinations were performed on day 1 (O), day 8 (O), and day 22 (z~). Before testing, the volume was adjusted to 100/11 per well with PBS, pH 7.5.
305 I tOO
E,,,
1200 &
1000 °•
800
•.
,.
•
°°°o
;. ..
600
".
•'~.-;
. ,~
_*•
t.O0 200 0
,'o,°
16.
,6., eff
Clq
,6.,
,0,,'
motecules
Fig. 8. Correlation between Clq-ELISA and the haemolytic C l q assay performed in normal h u m a n sera, in sera with decreased CH50 activity and in sera with C l q deficiency.
Comparison of ELISA with haemolytic assay To compare the results of the ELISA with the functional haemolytic assay, 70 sera from healthy persons and patients with lower haemolytic CH50 titres or C l q deficient sera were tested. Fig. 8 demonstrates the correlation between the haemolytic assay system and C l q measured in the Clq ELISA. In normal human sera the haemolytic titres ranged from 1.4 x 1012 to 3.7 x 10 t3 effective molecules per ml. Considering the corresponding C l q absorption in the same sera diluted 1 : 100, it was calculated that one nanogram C l q protein corresponded to 2.6 × 10 l° effective C l q molecules per ml.
Discussion In this report the development of a sandwich ELISA for the measurement of C l q in human serum is described. Three combinations of antibodies were used. It was not possible to establish an ELISA for C l q using only polyclonal anti-Clq antibodies. The assay which finally resulted is both specific and sensitive and makes use of both a monoclonal and a polyclonal anti-Clq antibody. Optimum results were obtained using plates coated with polyclonal antiserum to human Clq. Serum C l q bound to this antibody was then detected by the use of a monoclonal anti-human Clq. Of 5 different mouse monoclonal anti-human C l q antibodies, only one was found to be suitable in this technique. This monoclonal antibody 242G3 recognizes epitopes on the C l q molecule which are not occupied by the polyclonal antibody. This may be the explanation why a second polyclonal anti-Clq antibody failed to be useful in an ELISA technique. The monoclonal antibody 242G3 which was found to be suitable recognizes only the C-chain and not the A- and B-chains of the molecule. In addition, this antibody detects predominantly C l q bound by the globular portions to the Fc regions of antibodies in immune complexes, indicating that epitopes of the C l q molecules are recognized which are not involved in the binding of C l q to the Fc portion.
306 In kinetic e x p e r i m e n t s the o p t i m a l c o n d i t i o n s for p e r f o r m i n g the C l q - E L 1 S A were d e t e r m i n e d . O p t i m a l b i n d i n g of either purified or serum C l q to the s o l i d - p h a s e p o l y c l o n a l a n t i b o d y a n d of m o n o c l o n a l a n d p e r o x i d a s e - c o n j u g a t e d a n t i b o d i e s occurred within 60 min at r o o m temperature. A d o s e - r e s p o n s e curve was established with highly p u r i f i e d C l q . A linear d e p e n d e n c y was o b s e r v e d ranging from 0.1 to 5.0 /~g C l q p e r ml. U s i n g this s t a n d a r d curve the C l q c o n t e n t in serum could easily be calculated. T h e m e a n value in 70 different n o r m a l h u m a n sera was f o u n d to be 160 / ~ g / m l . C o m p a r i s o n of the h a e m o l y t i c assay with the E L I S A technique showed that one n a n o g r a m C l q p r o t e i n c o r r e s p o n d e d to 2.6 × 101° effective C l q molecules. The C l q E L I S A c o r r e l a t e d closely with the C l q titre in normal, p a r t i a l l y deficient or c o m p l e t e l y deficient h u m a n sera. Very small quantities of all reagents are required a n d using p r e c o a t e d plates which can be stored for several weeks at 4°C, a large n u m b e r of test samples can be screened in a few hours. Since the Clq c o m p o n e n t p l a y s an integral role in c o m p l e m e n t activation by i m m u n e complexes, the d e t e r m i n a t i o n of C l q c o n c e n t r a t i o n s in the serum of p a t i e n t s with r h e u m a t i c disorders, SLE, etc. m a y be useful in diagnostics a n d as a guide for therapy.
References Berkel, A.I., M. Loos, O. Sanal, G. Mauff, Y. Gi~ngen, U. Ors, F. Ersoy and O. Yegin, 1979, Clin. Exp. Immunol. 38, 5'). Engvall, E. and P. Perlmann, 1972, J. Immunol. 109, 129. Golan, M.D., T. Hitschold and M. Loos, 1981, FEBS Lett. 128, 281. Golan, M.D., R.. Burger and M. Loos, 1982, J. Immunol. 129, 445. Heinz, H.P., R. Burger, M.D. Golan and M. Loos, 1984, J. Immunol. 132, 804. Loos, M., 1982, Prog. Allergy 3l, 135. Loos, M., A.B. Laurell, A. Sj6holm, U. Martensson and J. Berke|, 1980, J. Immunol. 124, 59. Lowry, O.H., N.J. Rosebrough, A.L. Farr and R.J. Randall, 1951, J. Biol. Chem. 193, 265. Stemmer, I. and M. Loos, 1984, submitted. Volanakis, J.E. and R.M. Stroud, 1972, J. Immunol. Methods 2, 25.
299
Elsevier
JIM 03268
Enzyme-Linked Immunosorbent Assay for Clq in Human Serum by Use of Monoclonal Antibodies Ursula Antes, Hans-Peter Heinz and Michael Loos * Institute of Medical Microbiologr', Johannes Gutenberg University, Hochhaus am A ugustusplatz, 6500 Mainz, F.R.G.
(Received 12 June 1984, accepted 24 July 1984)
A sandwich ELISA system has been developed for the detection of C l q in h u m a n serum. It is specific, uses monoclonal antibodies, is sensitive into the nanogram range and is rapidly performed. Therefore, it may be a helpful tool for clinical routine diagnosis, e.g., detecting abnormal C l q levels in patients with rheumatic disorders. Various combinations of poly- and monoclonal antibodies were tested in a sandwich assay. One of these combinations, in particular, resulted in a highly reproducible standard curve: C l q bound to solid-phase polyclonal anti-Clq was detected by the monoclonal antibody 242 G3. In this assay, the Clq concentration in sera of normal individuals was found to be 160 # g / m l (mean value of 70 normal h u m a n sera). This ELISA detected nanogram levels of C l q and gave results comparable to those obtained by haemolytic C l q titration. One nanogram of C l q corresponded to ca. 2.6 x 101° effective C l q molecules. With this technique, selective C l q deficient sera as well as sera from patients with rheumatoid diseases were analysed. Key words: CI q- E L I S A - monoclonal antibody - Cl q in N H S and patients' sera
Introduction
Clq, a subcomponent of the first complement component initiates activation of the complement cascade via the classical pathway by binding to the Fc region of antibodies (reviewed by Loos, 1982). Therefore, this component plays an important role in complement activation by immune complexes. Determination of Clq content is of clinical interest (e.g., in patients with rheumatic disorders), in diagnostics and as a guide for therapy. To date, exact amounts of Clq have been measured in a functional haemolytic assay. Compared to the haemolytic assay, radial immunoassays for Clq are very insensitive. Therefore, in order to measure Clq immunochemically, an enzyme-lin-
* Mailing address: Dr. Michael Loos, Institut ftir Med. Mikrobiologie, Johannes Gutenberg-Universit~t, Hochhaus Augustusplatz, D-6500 Mainz, F.R.G. 0022-1759/84/$03.00 © 1984 Elsevier Science Publishers B.V.
300 ked immunosorbent assay (ELISA) was developed for the detection of C l q antigen in human serum. The original ELISA system, first described by Engvall and Perlmann (1972) was modified using different solid-phase antibodies in a sandwich technique. Of several mouse monoclonal antibodies to human C l q available in our laboratory (Golan et al., 1982), only one was found to be suitable for an ELISA sandwich technique. This antibody recognizes only the C chain and not the A and B chains of C l q (Heinz et al., 1984). The aim of this study was to develop a rapid, specific, sensitive C l q assay which would be economical and technically simple and correlate well with the haemolytic assay for Clq. This report describes the method in detail, shows the effect of altering various test parameters and gives the results of C l q in normal and in patients' sera.
Materials and methods
Antiserum tO human C l q was obtained from rabbits (Behringwerke, Marburg) or goats (Atlantic Antibodies, Scarborough). Peroxidase-conjugated rabbit immunoglobulins to mouse immunoglobulins and peroxidase-conjugated swine immunoglobulins to rabbit immunoglobulins were obtained from Dako, Copenhagen. Monoclonal anti-human C l q antibodies were raised in B A L B / c mice as recently described by Golan et al. (1982). The IgG fraction of the monoclonal antibodies was obtained by ammonium-sulphate precipitation and subsequent chromatography on a D E A E 52 column (Heinz et al., 1984). Serum samples were obtained from normal healthy donors and from patients with rheumatoid diseases and selective C l q deficiency (Berkel et al., 1979; Loos et al., 1980). Protein concentrations were determined by the method of Lowry et al. (1951) with bovine serum albumin acting as standard. The haemolytic activity of C l q was determined as described by Golan et al. (1981).
C/q Two different C l q preparations were used as C l q standards. The first preparation was purified from normal human serum according to the procedure of Volanakis and Stroud (1972). The second purification was carried out according to Stemmer and Loos (1984), using fast protein liquid chromatography (FPLC).
ELISA-procedure The following ELISA procedure was found to be the most suitable. All reagents and test samples were diluted in 10 mM phosphate-buffered saline (PBS), p H 7.5, of ionic strength 15 mS, unless stated otherwise. The assays were performed at room temperature unless otherwise stated. Wells of polystyrol microtitration plates ( N U N C - I M M U N O , 96 F-Plates No. 439454, Wiesbaden) were coated with rabbit antiserum to human Clq (Behring-
301 werke, Marburg) by addition of 20 ~tl of a 1 : 100 dilution, protein concentration 40 mg/ml. The plates were incubated for 1 h, washed in PBS pH 7.5 containing 0.5% BSA (bovine serum albumin, No. A 2153, Sigma, Munich) and left in the same buffer for 1 h to block remaining binding sites. The plates were used immediately or stored at 4°C. Purified human C l q (protein concentration 1.6 m g / m l ) at different dilutions or serum samples were added (10 t~l per well) and incubated for 1 h. After washing twice with 0.5% PBS-BSA, the plates were incubated with a monoclonal mouse antibody to human C l q (20/~l/well of a 1 : 200 dilution, protein concentration 1.7 m g / m l ) for 1 h and washed twice as described above. Bound human C l q antibodies were detected with peroxidase-conjugated rabbit anti-mouse IgG (10/~1 of a 1 : 200 dilution added to each well, protein concentration 5 m g / m l ) (Dako, Copenhagen). After 1 h, the plates were washed twice and 20/~1 50 mM ABTS substrate was added, diluted in PBS pH 6.0, containing 25 btl/ml of a 0.1% hydrogen peroxide solution. ABTS (2,2'-azino-di(ethyl-benz-thiazoline-6sulphonic acid) diammonium salt) was obtained from Serva, No. 14364, Heidelberg. The colour reaction was stopped after 30 min at 37°C by addition of 100 ~tl of a 0.0125% sodium azide solution in PBS (pH 6.0). Absorbance values were measured with a Titertek Multiscan MC (Flow Laboratories, Meckenheim) at 414 nm against a substrate solution blank.
Results
Combinations of antibodies The following combinations of antibodies were tested to establish the optimum conditions for the CIq-ELISA. (1) Solid-phase polyclonal goat anti-human C l q as the first and polyclonal rabbit anti-human C l q as the second step antibody. (2) Solid-phase monoclonal mouse anti-human C l q as the first and polyclonal rabbit anti-human C l q as the second step. (3) Solid-phase polyclonal rabbit anti-human C l q and monoclonai mouse antihum'an C l q for detection. The results in Fig. 1 show that the third combination was the most effective for the detection of C l q in serum, and it remained the method of choice for subsequent tests. Different monoclonal antibodies Comparing different monoclonal antibodies against purified human C l q (Golan et al., 1982), it was found that one antibody in particular (242 G3) was very useful in this ELISA system. Fig. 2 demonstrates the reaction of 5 monoclonal antibodies against human-Clq. Monoclonal antibodies 244 G8 and 244 H4 showed very high background values due to a high non-specific binding. With two other monoclonal antibodies (239 F10 and 241 F l l ) , differences in absorbance values were not high enough at the concentration used to standardize the C l q content of serum.
302
Optimum assay conditions Once the antibodies, the best sandwich combination and the Clq standard had been established, optimal conditions of the following test parameters were investigated: concentrations and volume of antibodies, test samples and reagents, incubation time. Figs. 3 5 show the effect of a number of variations. The amount of polyclonal antibody for coating the plates (Fig. 3) was varied from 2 to 800 /~g, the optimal concentration was found to be 8-20 /~g per well. Fig. 4 shows the same for the amount of monoclonal antibody 242 G3. It was used as a second step antibody in the sandwich in a dilution of ] : 200 which resulted in about 8 /~g/well, obviously enough to detect antibody-bound Clq. The effects of changing the incubation time of the polyclonal antibody used for coating and the BSA buffer (A), Clq antigen (B), monoclonal antibody from the mouse (C) and anti-mouse peroxidase-conjugated antibody (D) are shown in Fig. 5. An incubation time of 60 min was found to be sufficient for all of the antibodies and antigens used.
1600
EI.,~
1400
1400
J
1200
1200
lO0O
1000
800
800
600
600
400
/-,00
E~
/ /
~
I)2/,2G3
/
2MFI1 239 FIO
200
200
0
0'1
o's
1'0
,s
concentration of purified human Clq (pg/m[)
0
0'1
10
106
concentration of human Clq (~g/m[)
Fig. l. Combinations of antibodies for a sandwich Clq-ELISA. Microtitration plates were coated with antibody I, followed by saturation with BSA buffer. Bound C | q was detected by antibody lI. Antibody II was detected by an excess of peroxidase-conjugated anti-mouse antibody or by peroxidase-conjugated anti-rabbit antibody respectively. Working dilutions for each individual system were optimized in a test system using Clq-coated plates, antibody II and peroxidase-conjugated antibodies, zx zx, antibody I: solid-phase polyclonal goat anti-human Clq; antibody II: polyclonal rabbit anti-human Clq. • •, antibody I: solid-phase monoclonal mouse anti-human Clq; antibody I|: polyclonal rabbit anti-human Clq. O O, antibody I: solid-phase polyclonal rabbit anti-human Clq; antibody lI: monoclonal mouse anti-human Clq. Fig. 2. Comparison of 5 monoclonal anti-human C l q antibodies used in ELISA. Plates were coated with polyclonal rabbit anti-human C l q antibody. Bound C l q was detected by 5 different monoclonal mouse anti-human C l q antibodies.
303
800
E ~,~
700 E,,,
1100
600
1000
5OO
900
400
800
300/
/
200
700
~o
100
600 i
i
0
)8o
~o
1'o
~-Clq monoclonal(~g)
ct Clq polyclonal (pg)
Fig. 3. Determination of an optimal amount of polyclonal rabbit anti-human C l q for coating the plates. Fig. 4. Determination of an optimal amount of monoclonal mouse anti-human C l q (242 G3) used as second antibody in ELISA. . E ,,,
B
1 200 -
1000"
800-
t,O0
r ~
/
600-
"~
-
~oo 200-
20O 0
0 I ~,110
2; ,0 ~o E,,
tl.,~°l C
0
20
i~.oo" E,),
t.O
60
t (rain) D
1200.
121111
1000-
t 000
800.
600-
61111
/,00"
/
200'
21111 j
0
0
0
Fig. 5. Effect of incubation time on A: BSA buffer used for blocking non-specific binding sites (Lx zx) after incubation of the plates with polyclonal rabbit anti-human C l q (D D) and before exposure to purified Clq. B: binding of C l q to solid-phase rabbit antihuman Clq. C: binding of monoclonal mouse anti-human C l q to detect bound Clq. D: detection of mouse monoclonal anti-Clq antibodies with peroxidase-conjugated anti-mouse IgG.
304
Cl q standard Using these conditions, 2 preparations of purified C l q were assayed with equivalent results to obtain a standard curve. A linear relationship was found between absorbance values and C l q concentration on a semi-log scale, ranging from 0.1 to 5.0/~g C l q per ml. Therefore test sample values of C l q can easily be extrapolated and calculated. The ELISA for Clq in serum is linear at least between serum dilutions of 1 : 100 and 1 : 4000. The addition of EDTA to the sera for dissociation of the macromolecular C] complex was unnecessary. In controls omitting one part of the sandwich it was shown that non-specific binding of Clq to the antibodies did not occur. Fig. 6 demonstrates the Clq standard curve and the calculation of Clq content in serum test samples. C1 q was detected down into the nanogram range and in normal human sera the mean value of Clq concentration was found to be 160 /~g/ml.
Reproducibifity The ELISA system was found to be reproducible and in order to make it both more practical and more rapid, tests were carried out with precoated plates. The plates covered with lids were stored in a refrigerator. Before the tests were performed the volume was adjusted to 1 0 0 / H / w e l l with PBS, pH 7.5. As shown in Fig. 7, the antibody/BSA-coated plates could be stored at 4°C for several weeks without loss of antibody activity. Storing at room temperature resulted in an increase of background absorbance values.
1400
1200
1200
E~
o//
1000
o /~/.Hs //
1000
800
/
800 /
600
600 ~00
/ /5
200
/
,/
/
4OO 200
0
0 0
01
10
100
concentration of purified human Clq (#g/ml)
0
0.1
10
100
concentration of purified human C|q (~Jg/m[)
Fig. 6. Standard curve of highly purified human Clq (protein concentration 1.6 m g / m l ) and calculation of C l q content in serum test samples diluted 1 : 100. Purified human C l q in different dilutions or serum samples were added to polyclonal anti-human Clq-coated plates (10 p,1/well) and incubated for 1 h. After washing twice, the plates were incubated with a monoclonal mouse antibody to human Clq for 1 h and washed twice. Bound human C l q antibodies were detected with peroxidase-conjugated rabbit anti-mouse IgG. Fig. 7. Reproducibility of the C l q standard. Microtitration plates were precoated with polyclonal rabbit anti-human C l q antiserum, washed in 0.5% BSA-PBS and kept in the same buffer. The plates were stored at 4°C, determinations were performed on day 1 (O), day 8 (O), and day 22 (z~). Before testing, the volume was adjusted to 100/11 per well with PBS, pH 7.5.
305 I tOO
E,,,
1200 &
1000 °•
800
•.
,.
•
°°°o
;. ..
600
".
•'~.-;
. ,~
_*•
t.O0 200 0
,'o,°
16.
,6., eff
Clq
,6.,
,0,,'
motecules
Fig. 8. Correlation between Clq-ELISA and the haemolytic C l q assay performed in normal h u m a n sera, in sera with decreased CH50 activity and in sera with C l q deficiency.
Comparison of ELISA with haemolytic assay To compare the results of the ELISA with the functional haemolytic assay, 70 sera from healthy persons and patients with lower haemolytic CH50 titres or C l q deficient sera were tested. Fig. 8 demonstrates the correlation between the haemolytic assay system and C l q measured in the Clq ELISA. In normal human sera the haemolytic titres ranged from 1.4 x 1012 to 3.7 x 10 t3 effective molecules per ml. Considering the corresponding C l q absorption in the same sera diluted 1 : 100, it was calculated that one nanogram C l q protein corresponded to 2.6 × 10 l° effective C l q molecules per ml.
Discussion In this report the development of a sandwich ELISA for the measurement of C l q in human serum is described. Three combinations of antibodies were used. It was not possible to establish an ELISA for C l q using only polyclonal anti-Clq antibodies. The assay which finally resulted is both specific and sensitive and makes use of both a monoclonal and a polyclonal anti-Clq antibody. Optimum results were obtained using plates coated with polyclonal antiserum to human Clq. Serum C l q bound to this antibody was then detected by the use of a monoclonal anti-human Clq. Of 5 different mouse monoclonal anti-human C l q antibodies, only one was found to be suitable in this technique. This monoclonal antibody 242G3 recognizes epitopes on the C l q molecule which are not occupied by the polyclonal antibody. This may be the explanation why a second polyclonal anti-Clq antibody failed to be useful in an ELISA technique. The monoclonal antibody 242G3 which was found to be suitable recognizes only the C-chain and not the A- and B-chains of the molecule. In addition, this antibody detects predominantly C l q bound by the globular portions to the Fc regions of antibodies in immune complexes, indicating that epitopes of the C l q molecules are recognized which are not involved in the binding of C l q to the Fc portion.
306 In kinetic e x p e r i m e n t s the o p t i m a l c o n d i t i o n s for p e r f o r m i n g the C l q - E L 1 S A were d e t e r m i n e d . O p t i m a l b i n d i n g of either purified or serum C l q to the s o l i d - p h a s e p o l y c l o n a l a n t i b o d y a n d of m o n o c l o n a l a n d p e r o x i d a s e - c o n j u g a t e d a n t i b o d i e s occurred within 60 min at r o o m temperature. A d o s e - r e s p o n s e curve was established with highly p u r i f i e d C l q . A linear d e p e n d e n c y was o b s e r v e d ranging from 0.1 to 5.0 /~g C l q p e r ml. U s i n g this s t a n d a r d curve the C l q c o n t e n t in serum could easily be calculated. T h e m e a n value in 70 different n o r m a l h u m a n sera was f o u n d to be 160 / ~ g / m l . C o m p a r i s o n of the h a e m o l y t i c assay with the E L I S A technique showed that one n a n o g r a m C l q p r o t e i n c o r r e s p o n d e d to 2.6 × 101° effective C l q molecules. The C l q E L I S A c o r r e l a t e d closely with the C l q titre in normal, p a r t i a l l y deficient or c o m p l e t e l y deficient h u m a n sera. Very small quantities of all reagents are required a n d using p r e c o a t e d plates which can be stored for several weeks at 4°C, a large n u m b e r of test samples can be screened in a few hours. Since the Clq c o m p o n e n t p l a y s an integral role in c o m p l e m e n t activation by i m m u n e complexes, the d e t e r m i n a t i o n of C l q c o n c e n t r a t i o n s in the serum of p a t i e n t s with r h e u m a t i c disorders, SLE, etc. m a y be useful in diagnostics a n d as a guide for therapy.
References Berkel, A.I., M. Loos, O. Sanal, G. Mauff, Y. Gi~ngen, U. Ors, F. Ersoy and O. Yegin, 1979, Clin. Exp. Immunol. 38, 5'). Engvall, E. and P. Perlmann, 1972, J. Immunol. 109, 129. Golan, M.D., T. Hitschold and M. Loos, 1981, FEBS Lett. 128, 281. Golan, M.D., R.. Burger and M. Loos, 1982, J. Immunol. 129, 445. Heinz, H.P., R. Burger, M.D. Golan and M. Loos, 1984, J. Immunol. 132, 804. Loos, M., 1982, Prog. Allergy 3l, 135. Loos, M., A.B. Laurell, A. Sj6holm, U. Martensson and J. Berke|, 1980, J. Immunol. 124, 59. Lowry, O.H., N.J. Rosebrough, A.L. Farr and R.J. Randall, 1951, J. Biol. Chem. 193, 265. Stemmer, I. and M. Loos, 1984, submitted. Volanakis, J.E. and R.M. Stroud, 1972, J. Immunol. Methods 2, 25.