- Email: [email protected]
Development of solid-phase immunoassay using chemiluminescent IgG conjugates
Journal o f Immunological Methods, 48 ( 1982) 159--168 Elsevier Biomedical Press
159
DEVELOPMENT OF SOLID-PHASE IMMUNOASSAY USING CHEMILUMINESCENT IgG CONJUGATES
P.-J. CHENG, I. HEMMIL.~ and T. LC)VGREN 1 Wallac Biochemical Laboratory, P.O. Box, 20101 Turku 10, Finland
(Received 15 April 1981, accepted 10 August 1981)
Solid-phase luminescent immunoassay (LIA) was studied using mainly aminobutylethylisoluminol-IgG conjugates. Different solid-phase supports such as immunobeads, polystyrene balls and tubes gave comparable results although tubes were preferred in most of the assays because of the better linearity and reproducibility obtained and their ease of handling. The properties of the conjugates were tested using both direct and sandwich LIA. Direct LIAs performed using sheep-anti-rabbit IgG coated balls provided information on the antigenicity of the conjugates while the sandwich LIA was the actual working system. A sensitivity of 1 ng was obtained in this system. The stability, easy use, safe handling and low cost of the conjugates coupled with a short assay time make this luminescent system a potential alternative to RIA.
Keywords: chemiluminescence -- immunoassay
INTRODUCTION
Since its introduction, radioimmunoassay (RIA) has been widely adopted for use in both clinical and research laboratories. The large number of RIA tests performed today is convincing evidence of the reliability, accuracy and sensitivity of this assay system. However, RIA has certain drawbacks, e.g., the inevitable radiation hazard, the relatively short half-life of certain isotopes, the long time required for counting, and efforts have been made to find new useful probes (Haimovich et al., 1970; Engvall and Perlmann, 1971; Van Weemen and Schuurs, 1971; Leute and Ullman, 1972). During recent years the enzyme immunoassay has become the most c o m m o n non-isotopic technique (Rubenstein et al., 1972; Schuurs and Van Weemen, 1977). The possibility of replacing radioisotopes by luminescent labels has recently awakened considerable interest (Whitehead et al., 1979). This paper
1 Send correspondence to: Dr. Timo LSvgren, Wallac Oy, P.O. Box 10, SF-20101 Turku 10, Finland. 0022-1759/82/0000--0000/$02.75 © 1982 Elsevier Biomedical Press
160 deals with results obtained with luminescent immunoassays performed with a chemiluminescent rabbit-anti-human IgG conjugate. The assays were made on a solid phase using both direct and sandwich LIA techniques. MATERIALS AND METHODS
Antibodies An IgG fraction of sheep-anti-rabbit IgG (SaR), an IgG fraction of rabbitanti-human IgG (Rail) and human IgG (HIgG) were kindly donated by Mr. G.D. Wu and an IgG fraction of normal rabbit serum (RIgG) was kindly donated by Prof. T. Sun, both from the Institute of Biochemistry, Academica Sinica, Shanghai. All fractions were used after either dialysis or desalting by gel filtration.
Solid phase Rabbit~nti-human IgG (Rail) immunobeads and normal rabbit IgG (RIgG) immunobeads were obtained from Bio-Rad Laboratories. The polystyrene balls were obtained from Precision Ball Co. and LKB polystyrene tubes were used.
Chemicals ABEI {6-{N-4-aminobutyl)-N-ethyl)amino-2,3-dihydrophthalazine-l,4dione) was kindly d o n a t e d b y the Wallac Organic Chemistry Laboratory. EDC (l~thyl-3{3~limethyl-aminopropyl)-carbodiimide-HC1) and microperoxidase were obtained from Sigma. All other chemicals were of analytical reagent grade.
Coating procedure Coating was performed in carbonate buffer pH 9.5, either at 37°C for 1 h followed by incubation at 4°C overnight (or even longer) or at r o o m temperature for 20 h. The coated tubes or balls were stored at 4°C until use. Before incubation with the ligand under investigation, the solid phases were saturated with a solution containing 0.01 or 0.1% BSA in PBS (0.01 M phosphate-buffered saline, pH 7.4). The same solution was used to dilute the IgGs.
Labeling ABEI conjugates. The IgG fraction of rabbit-anti-human IgG was first mixed with ABEI in m o l a r ratios of 10, 50, 100 and 150, respectively. ABEI was made up to 50 mM stock solution b y dissolving in 0.1 N HC1. The pH was adjusted to 4.75 and solid EDC was added in molar ratios (EDC/IgG) of 1 0 0 , 3 0 0 and 500 (Carraway et al., 1969; Fersht and Sperling, 1973). The coupling t o o k place at room temperature for 1 or 4 h. Excess EDC was blocked by either acetic acid or glycine or removed by gel filtration. All conjugates were freed from excess reagent by gel filtration through a Sephadex
161 G-50 column equilibrated with and eluted b y 0.001 N HC1. The molar ratio of ABEI/IgG was calculated using a molar extinction coefficient of 17,000 for ABEI at 323 nm and the IgG added was considered 100% recovered. Diazo coupling. The diazotization of luminol was performed as follows. A 20 mM luminol solution was prepared in 0.1 N Na2CO3, pH 11.4 (Schroeder and Yeager, 1978). Concentrated HC1 was added to make the final concentration of HC1 2.4 N (Simpson et al., 1979) just prior to addition of NaNO2. The molar ratio of NaNO2 to luminol or isoluminol was approximately 1.5. Diazotization t o o k place at 4°C overnight. Excess nitrous acid was destroyed by urea (Pratt et al., 1978). The degree of diazotization was checked by using resorcinol (Koltun, 1957). The diazotized material was stored at 4°C. Isoluminol was dissolved in 25% fluoroboric acid before addition of stoichiometric amounts of NaNO2. The reaction was allowed to proceed at 0°C for 2 h after which precipitated diazoisoluminol fluoroborate was collected by filtration and washed with water. The coupling of the label to IgG t o o k place at 4°C overnight at pH 8.0. The excess label was removed by gel filtration. Any precipitate occurring during conjugation was removed b y centrifugation before the gel filtration step. An estimation of conjugation yield was obtained using absorption ratios at 3 3 5 / 2 7 8 nm for luminol and 330/275 nm for isoluminol conjugates.
Luminescent immunoassay (LIA ) In the direct assay SaR coated balls were incubated with 200 pl labeled Rail or RIgG from different batches at 37°C for 1 h. The balls were washed 3 times, 10 min per wash, with 300 /21 0.1% Tween-PBS before measuring luminescence. In the sandwich assay 200 pl of each different dilutions of human IgG were incubated with Rail coated balls or tubes for 1 h at 37°C. The coated surfaces were washed 3 times, 10 min per wash, with 300/~l 0.1% TweenPBS followed by incubation with 200 pl of different batches o f conjugates at 37°C for 1 h. The washing procedure was repeated 3 times before measurement of the luminescence. The sandwich assay using immunobeads was carried o u t with a 40/~l bead suspension (5 mg/ml) which was washed once with phosphate buffer and saturated with 100 pl BSA-PBS solution as the other solid-phase systems. Dilutions of human IgG were added in 100/~l aliquots followed by different Rail conjugates. Incubations and washings were performed as stated for the other assays. All solutions were aspirated carefully by gentle suction after centrifugation. Luminescence measurement A microperoxidase-H202 system in phosphate buffer (0.05 M, pH 8.6) was used (Roswell and White, 1978; Schroeder and Yeager, 1978). The reaction mixture had the following composition: 0.05 M phosphate buffer, pH 8.6; 0.5 ml or 0.2 ml 5 pM microperoxidase; 50 pl or 25 pl
162
0.44 mM H202; 0.45 ml or 0.3 ml sample (omitted in solid-phase assays); 10 pl The chemiluminescent reaction was initiated by the injection of H202. For the standard luminescence assay a 1 ml reaction volume was used but for solid-phase LIA 0.5 ml was used. The luminescence was measured with the 1250 luminometer (LKB-Wallac) connected to the 1223 Luminescence Analyzer (LKB-Wallac). A time integral of 20 sec was used for ordinary luminescence and an integral of 60 sec for LIA. The longer time intervals used for integration using LIA was due to heterogenic phase factor encountered in solid-phase LIA which changed the light kinetics and decay rate of the reaction.
Control experiments RIgG coated balls or tubes or RIgG immunobeads were used as the controls in solid-phase assays and labeled RIgG as the control for conjugates. R E S U L T S AND DISCUSSION
Properties of the conjugates Conjugation using diazoreaction proceeded well and resulted in luminescent conjugates with a typical absorption at 330 nm. High conjugation yield caused, however, decreased solubility and loss of immunoreactivity. The luminescence intensity was n o t linear with respect to the extent of labeling because of an obvious quenching at a high conjugation yield. Table 1 shows some examples of diazoconjugates and their properties. A water-soluble carbodiimide (EDC) was used to conjugate an isoluminol derivative (ABEI) to IgG. The properties of some ABEI-conjugates, their luminescence values and limits of detection are given in Table 1.
Comparison of different solid phases Several investigations have been carried out concerning the usefulness of the solid phases in immunoassays (Salonen and Vaheri, 1979; Lehtonen and Viljanen, 1980). The supports most c o m m o n l y used have been polystyrene or polyvinyl in the form of balls, tubes or microtiter plates. In the present study on LIA we were able to compare beads, balls and tubes with regard to efficiency, practicability and reproducibility of the results. Comparable results were obtained in all 3 systems. The beads worked more efficiently than the others probably owing to the larger surface area available. However, the inconvenience of centrifugation after each step to sediment the beads ruled t h e m out as the system of choice for routine work. The balls were the easiest to work with but the reproducibility was n o t as good as with the tubes, probably because of the inaccessibility of the surface of the ball which was facing downwards. Though balls gave higher luminescence than tubes and beads, the non-specific binding was also higher, because the IgGs were
Conjug. react.
EDC EDC EDC EDC Diazoniumchloride Diazoniumchloride Diazonium fluoroborate Diazonium fluoroborate
Luminol deriv.
ABEI ABEI ABEI ABEI Diazoluminol Diazoluminol Diazoisoluminol Diazoisoluminol
50 100 150 100 40 100 10 $4
Lumin. def.
Reagents
323 300 310 100 ---
EDC
A-1 A-2 A-3 A-4 DL-I DL-2 DIL-1 DIL-2
Code
0.39 0.39 0.51 0.33 0.29 0.50 0.76 0.70
Conjugate 330 nm/280 nm
950 1 585 1 580 1 211 112 737 2 757 161
Luminescence (mV sec)
1.2 0.6 0.3 1.0 ND 3.5 0.35 ND
Det. lira. (ng/ml)
Properties of the luminescent conjugates of RIgG. The reagent concentration is given as the molar ratio in respect of RIgG. Luminescence intensity was measured for 50 ng labeled RIgG.
TABLE 1
¢o
164
also bound to the surface of the tubes that contained the balls. Removing the balls to clean tubes before measuring luminescence would certainly decrease the background. Tubes were easy to handle and the reproducibility was acceptable.
Coating conditions The different concentrations of IgG used for coating were compared. In tubes and balls, no great differences were found with concentrations above 1 pg per tube. However, a concentration below 0.5 pg was not sufficient to bind the antigen effectively in the samples (Fig. 1). This observation is consistent with the results o f Cantarero et al. (1979). The coating of the balls seemed to improve up to 5 pg IgG per ball, though 2 pg per ball could be used. Coating at room temperature for at least 20 h worked as effectively as coating at 37°C for 1 h followed by storage at 4°C overnight, but the reproducibility seemed better in the latter case.
Non-specific binding Non-specific binding in the sandwich assay was determined by carrying
600-
500-
U >.
400-
300-
.J
200-
100-
I - log HIgG. (g~ Fig. 1. Comparison of coating concentrations in polystyrene tubes by LIA (sandwich) of human IgG. The tubes were coated with different amounts of rabbit-anti-human IgG. (e), 2 p g / t u b e ; (A), 1 p g / t u b e ; (o), 0.5 pg/tube; and (x), 0.2 pg/tube.
165 TABLE 2 Determination o f non-specific binding for two conjugates using Rail coated tubes in the assay and RIgG coated tubes in the controls. Experiment
Assay Control Assay Control
Tubes
Rail RIgG Rail RIgG
HIgG (g)
10 -7 10 -7 10 -7 10 -7
Conjugate Batch
pg
Rail Rail Rail Rail
1.1 1.1 1.0 1.0
A-2 A-2 A-4 A-4
Luminescence (mV see)
Non-specific (%)
503 32 334 13
6.4 4
800
700 -
600-
500-
400-
=1 300-
200-
100-
I 6
I 7
I 6
I 9
- log HIgG, (g)
Fig. 2. LIA (sandwich) using different batches o f labeled rabbit-anti-human IgG (Rail): A-1 ($), --2 (X), --3 (o) and --4 (a) in polystyrene tubes coated with Rail. - - - - - , nonspecific luminescence with RIgG tube and labeled Rail; . . . . . . , non-specific luminescence with Rail tube and labeled RIgG.
50 50 50
Rail DL-1 Rail A-1 Rail DIL-1
737 2412 2 757
mV see
ng
Batch
<3.5 <0.55 ~0.35
Sensitivity (ng)
Luminescence
Conjugate
1 1.1 1
74 732 58
10 1.1 10
Rail conj. (pg/tube)
Rail conj. (pg/tube)
Luminescence (mV sec)
Sandwich
Direct
LIA
Comparison of luminescence intensity in ordinary luminescence and direct and sandwich LIA.
TABLE 3
114 503 79
100 ng HIgG
Luminescence (mV see)
-~10 <1 ND
Sensitivity
167 out controls using either RIgG or Rail as the solid phase together with HIgG and labeled Rail or RIgG, respectively. The 0.1% BSA-PBS used for dilution in each incubation and the 0.1% Tween-PBS used in washing were f o u n d to be effective in reducing non-specific luminescence. The Rail solid phase combined with labeled RIgG always gave less non-specific luminescence than the RIgG solid phase combined with Rail conjugates. With good conjugates, the non-specific binding was less than 5% of the specific binding (Table 2).
Sensitivity and linearity The luminescence intensity observed in the 3 assay systems, i.e., standard luminescence, direct LIA and sandwich LIA, usually correlated well (Table 3). The detection limit of the assay was defined as the concentration at which the luminescent intensity in the sandwich LIA was twice the intensity obtained when non-specific binding was measured. Many of the conjugates gave a sensitivity below 1 ng in the sandwich assay (Fig. 2). In addition, the linearity of luminescence intensity at different conjugate concentrations per 700
600--
500-
r-
400--
= ,.d
300--
200--
100--
I
I
I
I
7
8
9
10
log HIgG, (g)
Fig. 3. LIA (sandwich) using labeled rabbit-anti-human IgG (batch A-l)in polystyrene bails coated with either 10 pg (o), 5 pg (x) or 2 gg (o) Rail per ball. ---- --, non-specific luminescence with RIgG ball and A-l; . . . . . . , non-specific luminescence with Rail bail and labeled RIgG.
168
se and in the LIA {sandwich) was excellent (Fig. 2). The LIA covered a fairly wide dynamic range extending from 1 ng to 1 pg although the increase in the luminescent intensity observed in LIA of HIgG is low when compared with the standard luminescence obtained with increasing amounts o f conjugated antibody. For some unexplained reason the curve tended to level o f f at HIgG concentrations below 1 ng. The effect was most obvious when balls were used as the solid phase (Fig. 3).
Reproducibility Because the chemiluminescent assay using the microperoxidase-H202 oxidation system is quite sensitive to interference from impurities, even though it gave the best sensitivity of the systems tested, it proved advisable to check the system with a blank (without luminescent label) and a standard luminescent solution. Under optimized conditions the variation coefficient was about 5%. In assays with background interference only a few values had to be rejected when a coefficient of variation of 10% was regarded as acceptable. The results might improve further on standardization of the coating procedure, a question that recently has resulted in several investigations (Cantarero et al., 1979; Salonen and Vaheri, 1979; Mariotti et al., 1980). REFERENCES Cantarero, L.A., J.E. Butler and J.W. Osborne, 1979, Fed. Proc. 38, 1012. Carraway, K.L., P. Spoerl and D.E. Koshland, Jr., 1969, J. Mol. Biol. 4 2 , 1 3 3 . Engvall, E. and P. Perlmann, 1971, Immunochemistry 8 , 8 7 1 . Fersht, A.R. and J. Sperling, 1973, J. Mol. Biol. 74,137. Haimovich, J., E. Hurvitz, N. Norvik and M. Sela, 1970, Biochim. Biophys. Acta 207, 125. Koltun, W.L., 1957, J. Am. Chem. Soc. 79, 5681. Lehtonen, O.P. and M.K. Viljanen, 1980, J. Immunol. Methods 34, 61. Leute, R.K. and E.F. Ullman, 1972, Nature 236, 93. Mariotti, S., J.J.-F. Oger, P. Fragur, J.P. Antel, H.-H. Kuo and L.J. DeGroot, 1980, J. Immunol. Methods 3 5 , 1 8 9 . Pratt, J.J., M.G. Woldring and L. Villerius, 1978, J. Immunol. Methods 21,179. Roswell, D.F. and E.H. White, 1978, in: Methods in Enzymology, Vol. LVII, Bioluminescence and Chemiluminescence, ed. M. DeLuca (Academic Press, New York) p. 409. Rubenstein, K.E., R.S. Schneider and E.F. Ullman, 1972, Biochem. Biophys. Res. Commun. 4 7 , 8 4 6 . Saloner~, E.-M. and A. Vaheri, 1979, J. Immunol. Methods 30, 209. Schroeder, H.R. and F.M. Yeager, 1978, Anal. Chem. 50, 1114. Schuurs, A.H.W.M. and B.K. Van Weemen, 1977, Clin. Chim. Acta 81, 1. Simpson, J.S.A., A.K. Campbell, M.E.T. Ryall and J.S. Woodhead, 1979, Nature 279, 646. Van Weemen, B.K. and A.H.W.M. Schuurs, 1971, FEBS Lett. 15,232. Whitehead, T.P., L.J. Kricka, T.J.N. Carter and G.H.G. Thorpe, 1979, Clin. Chem. 25, 1531.
159
DEVELOPMENT OF SOLID-PHASE IMMUNOASSAY USING CHEMILUMINESCENT IgG CONJUGATES
P.-J. CHENG, I. HEMMIL.~ and T. LC)VGREN 1 Wallac Biochemical Laboratory, P.O. Box, 20101 Turku 10, Finland
(Received 15 April 1981, accepted 10 August 1981)
Solid-phase luminescent immunoassay (LIA) was studied using mainly aminobutylethylisoluminol-IgG conjugates. Different solid-phase supports such as immunobeads, polystyrene balls and tubes gave comparable results although tubes were preferred in most of the assays because of the better linearity and reproducibility obtained and their ease of handling. The properties of the conjugates were tested using both direct and sandwich LIA. Direct LIAs performed using sheep-anti-rabbit IgG coated balls provided information on the antigenicity of the conjugates while the sandwich LIA was the actual working system. A sensitivity of 1 ng was obtained in this system. The stability, easy use, safe handling and low cost of the conjugates coupled with a short assay time make this luminescent system a potential alternative to RIA.
Keywords: chemiluminescence -- immunoassay
INTRODUCTION
Since its introduction, radioimmunoassay (RIA) has been widely adopted for use in both clinical and research laboratories. The large number of RIA tests performed today is convincing evidence of the reliability, accuracy and sensitivity of this assay system. However, RIA has certain drawbacks, e.g., the inevitable radiation hazard, the relatively short half-life of certain isotopes, the long time required for counting, and efforts have been made to find new useful probes (Haimovich et al., 1970; Engvall and Perlmann, 1971; Van Weemen and Schuurs, 1971; Leute and Ullman, 1972). During recent years the enzyme immunoassay has become the most c o m m o n non-isotopic technique (Rubenstein et al., 1972; Schuurs and Van Weemen, 1977). The possibility of replacing radioisotopes by luminescent labels has recently awakened considerable interest (Whitehead et al., 1979). This paper
1 Send correspondence to: Dr. Timo LSvgren, Wallac Oy, P.O. Box 10, SF-20101 Turku 10, Finland. 0022-1759/82/0000--0000/$02.75 © 1982 Elsevier Biomedical Press
160 deals with results obtained with luminescent immunoassays performed with a chemiluminescent rabbit-anti-human IgG conjugate. The assays were made on a solid phase using both direct and sandwich LIA techniques. MATERIALS AND METHODS
Antibodies An IgG fraction of sheep-anti-rabbit IgG (SaR), an IgG fraction of rabbitanti-human IgG (Rail) and human IgG (HIgG) were kindly donated by Mr. G.D. Wu and an IgG fraction of normal rabbit serum (RIgG) was kindly donated by Prof. T. Sun, both from the Institute of Biochemistry, Academica Sinica, Shanghai. All fractions were used after either dialysis or desalting by gel filtration.
Solid phase Rabbit~nti-human IgG (Rail) immunobeads and normal rabbit IgG (RIgG) immunobeads were obtained from Bio-Rad Laboratories. The polystyrene balls were obtained from Precision Ball Co. and LKB polystyrene tubes were used.
Chemicals ABEI {6-{N-4-aminobutyl)-N-ethyl)amino-2,3-dihydrophthalazine-l,4dione) was kindly d o n a t e d b y the Wallac Organic Chemistry Laboratory. EDC (l~thyl-3{3~limethyl-aminopropyl)-carbodiimide-HC1) and microperoxidase were obtained from Sigma. All other chemicals were of analytical reagent grade.
Coating procedure Coating was performed in carbonate buffer pH 9.5, either at 37°C for 1 h followed by incubation at 4°C overnight (or even longer) or at r o o m temperature for 20 h. The coated tubes or balls were stored at 4°C until use. Before incubation with the ligand under investigation, the solid phases were saturated with a solution containing 0.01 or 0.1% BSA in PBS (0.01 M phosphate-buffered saline, pH 7.4). The same solution was used to dilute the IgGs.
Labeling ABEI conjugates. The IgG fraction of rabbit-anti-human IgG was first mixed with ABEI in m o l a r ratios of 10, 50, 100 and 150, respectively. ABEI was made up to 50 mM stock solution b y dissolving in 0.1 N HC1. The pH was adjusted to 4.75 and solid EDC was added in molar ratios (EDC/IgG) of 1 0 0 , 3 0 0 and 500 (Carraway et al., 1969; Fersht and Sperling, 1973). The coupling t o o k place at room temperature for 1 or 4 h. Excess EDC was blocked by either acetic acid or glycine or removed by gel filtration. All conjugates were freed from excess reagent by gel filtration through a Sephadex
161 G-50 column equilibrated with and eluted b y 0.001 N HC1. The molar ratio of ABEI/IgG was calculated using a molar extinction coefficient of 17,000 for ABEI at 323 nm and the IgG added was considered 100% recovered. Diazo coupling. The diazotization of luminol was performed as follows. A 20 mM luminol solution was prepared in 0.1 N Na2CO3, pH 11.4 (Schroeder and Yeager, 1978). Concentrated HC1 was added to make the final concentration of HC1 2.4 N (Simpson et al., 1979) just prior to addition of NaNO2. The molar ratio of NaNO2 to luminol or isoluminol was approximately 1.5. Diazotization t o o k place at 4°C overnight. Excess nitrous acid was destroyed by urea (Pratt et al., 1978). The degree of diazotization was checked by using resorcinol (Koltun, 1957). The diazotized material was stored at 4°C. Isoluminol was dissolved in 25% fluoroboric acid before addition of stoichiometric amounts of NaNO2. The reaction was allowed to proceed at 0°C for 2 h after which precipitated diazoisoluminol fluoroborate was collected by filtration and washed with water. The coupling of the label to IgG t o o k place at 4°C overnight at pH 8.0. The excess label was removed by gel filtration. Any precipitate occurring during conjugation was removed b y centrifugation before the gel filtration step. An estimation of conjugation yield was obtained using absorption ratios at 3 3 5 / 2 7 8 nm for luminol and 330/275 nm for isoluminol conjugates.
Luminescent immunoassay (LIA ) In the direct assay SaR coated balls were incubated with 200 pl labeled Rail or RIgG from different batches at 37°C for 1 h. The balls were washed 3 times, 10 min per wash, with 300 /21 0.1% Tween-PBS before measuring luminescence. In the sandwich assay 200 pl of each different dilutions of human IgG were incubated with Rail coated balls or tubes for 1 h at 37°C. The coated surfaces were washed 3 times, 10 min per wash, with 300/~l 0.1% TweenPBS followed by incubation with 200 pl of different batches o f conjugates at 37°C for 1 h. The washing procedure was repeated 3 times before measurement of the luminescence. The sandwich assay using immunobeads was carried o u t with a 40/~l bead suspension (5 mg/ml) which was washed once with phosphate buffer and saturated with 100 pl BSA-PBS solution as the other solid-phase systems. Dilutions of human IgG were added in 100/~l aliquots followed by different Rail conjugates. Incubations and washings were performed as stated for the other assays. All solutions were aspirated carefully by gentle suction after centrifugation. Luminescence measurement A microperoxidase-H202 system in phosphate buffer (0.05 M, pH 8.6) was used (Roswell and White, 1978; Schroeder and Yeager, 1978). The reaction mixture had the following composition: 0.05 M phosphate buffer, pH 8.6; 0.5 ml or 0.2 ml 5 pM microperoxidase; 50 pl or 25 pl
162
0.44 mM H202; 0.45 ml or 0.3 ml sample (omitted in solid-phase assays); 10 pl The chemiluminescent reaction was initiated by the injection of H202. For the standard luminescence assay a 1 ml reaction volume was used but for solid-phase LIA 0.5 ml was used. The luminescence was measured with the 1250 luminometer (LKB-Wallac) connected to the 1223 Luminescence Analyzer (LKB-Wallac). A time integral of 20 sec was used for ordinary luminescence and an integral of 60 sec for LIA. The longer time intervals used for integration using LIA was due to heterogenic phase factor encountered in solid-phase LIA which changed the light kinetics and decay rate of the reaction.
Control experiments RIgG coated balls or tubes or RIgG immunobeads were used as the controls in solid-phase assays and labeled RIgG as the control for conjugates. R E S U L T S AND DISCUSSION
Properties of the conjugates Conjugation using diazoreaction proceeded well and resulted in luminescent conjugates with a typical absorption at 330 nm. High conjugation yield caused, however, decreased solubility and loss of immunoreactivity. The luminescence intensity was n o t linear with respect to the extent of labeling because of an obvious quenching at a high conjugation yield. Table 1 shows some examples of diazoconjugates and their properties. A water-soluble carbodiimide (EDC) was used to conjugate an isoluminol derivative (ABEI) to IgG. The properties of some ABEI-conjugates, their luminescence values and limits of detection are given in Table 1.
Comparison of different solid phases Several investigations have been carried out concerning the usefulness of the solid phases in immunoassays (Salonen and Vaheri, 1979; Lehtonen and Viljanen, 1980). The supports most c o m m o n l y used have been polystyrene or polyvinyl in the form of balls, tubes or microtiter plates. In the present study on LIA we were able to compare beads, balls and tubes with regard to efficiency, practicability and reproducibility of the results. Comparable results were obtained in all 3 systems. The beads worked more efficiently than the others probably owing to the larger surface area available. However, the inconvenience of centrifugation after each step to sediment the beads ruled t h e m out as the system of choice for routine work. The balls were the easiest to work with but the reproducibility was n o t as good as with the tubes, probably because of the inaccessibility of the surface of the ball which was facing downwards. Though balls gave higher luminescence than tubes and beads, the non-specific binding was also higher, because the IgGs were
Conjug. react.
EDC EDC EDC EDC Diazoniumchloride Diazoniumchloride Diazonium fluoroborate Diazonium fluoroborate
Luminol deriv.
ABEI ABEI ABEI ABEI Diazoluminol Diazoluminol Diazoisoluminol Diazoisoluminol
50 100 150 100 40 100 10 $4
Lumin. def.
Reagents
323 300 310 100 ---
EDC
A-1 A-2 A-3 A-4 DL-I DL-2 DIL-1 DIL-2
Code
0.39 0.39 0.51 0.33 0.29 0.50 0.76 0.70
Conjugate 330 nm/280 nm
950 1 585 1 580 1 211 112 737 2 757 161
Luminescence (mV sec)
1.2 0.6 0.3 1.0 ND 3.5 0.35 ND
Det. lira. (ng/ml)
Properties of the luminescent conjugates of RIgG. The reagent concentration is given as the molar ratio in respect of RIgG. Luminescence intensity was measured for 50 ng labeled RIgG.
TABLE 1
¢o
164
also bound to the surface of the tubes that contained the balls. Removing the balls to clean tubes before measuring luminescence would certainly decrease the background. Tubes were easy to handle and the reproducibility was acceptable.
Coating conditions The different concentrations of IgG used for coating were compared. In tubes and balls, no great differences were found with concentrations above 1 pg per tube. However, a concentration below 0.5 pg was not sufficient to bind the antigen effectively in the samples (Fig. 1). This observation is consistent with the results o f Cantarero et al. (1979). The coating of the balls seemed to improve up to 5 pg IgG per ball, though 2 pg per ball could be used. Coating at room temperature for at least 20 h worked as effectively as coating at 37°C for 1 h followed by storage at 4°C overnight, but the reproducibility seemed better in the latter case.
Non-specific binding Non-specific binding in the sandwich assay was determined by carrying
600-
500-
U >.
400-
300-
.J
200-
100-
I - log HIgG. (g~ Fig. 1. Comparison of coating concentrations in polystyrene tubes by LIA (sandwich) of human IgG. The tubes were coated with different amounts of rabbit-anti-human IgG. (e), 2 p g / t u b e ; (A), 1 p g / t u b e ; (o), 0.5 pg/tube; and (x), 0.2 pg/tube.
165 TABLE 2 Determination o f non-specific binding for two conjugates using Rail coated tubes in the assay and RIgG coated tubes in the controls. Experiment
Assay Control Assay Control
Tubes
Rail RIgG Rail RIgG
HIgG (g)
10 -7 10 -7 10 -7 10 -7
Conjugate Batch
pg
Rail Rail Rail Rail
1.1 1.1 1.0 1.0
A-2 A-2 A-4 A-4
Luminescence (mV see)
Non-specific (%)
503 32 334 13
6.4 4
800
700 -
600-
500-
400-
=1 300-
200-
100-
I 6
I 7
I 6
I 9
- log HIgG, (g)
Fig. 2. LIA (sandwich) using different batches o f labeled rabbit-anti-human IgG (Rail): A-1 ($), --2 (X), --3 (o) and --4 (a) in polystyrene tubes coated with Rail. - - - - - , nonspecific luminescence with RIgG tube and labeled Rail; . . . . . . , non-specific luminescence with Rail tube and labeled RIgG.
50 50 50
Rail DL-1 Rail A-1 Rail DIL-1
737 2412 2 757
mV see
ng
Batch
<3.5 <0.55 ~0.35
Sensitivity (ng)
Luminescence
Conjugate
1 1.1 1
74 732 58
10 1.1 10
Rail conj. (pg/tube)
Rail conj. (pg/tube)
Luminescence (mV sec)
Sandwich
Direct
LIA
Comparison of luminescence intensity in ordinary luminescence and direct and sandwich LIA.
TABLE 3
114 503 79
100 ng HIgG
Luminescence (mV see)
-~10 <1 ND
Sensitivity
167 out controls using either RIgG or Rail as the solid phase together with HIgG and labeled Rail or RIgG, respectively. The 0.1% BSA-PBS used for dilution in each incubation and the 0.1% Tween-PBS used in washing were f o u n d to be effective in reducing non-specific luminescence. The Rail solid phase combined with labeled RIgG always gave less non-specific luminescence than the RIgG solid phase combined with Rail conjugates. With good conjugates, the non-specific binding was less than 5% of the specific binding (Table 2).
Sensitivity and linearity The luminescence intensity observed in the 3 assay systems, i.e., standard luminescence, direct LIA and sandwich LIA, usually correlated well (Table 3). The detection limit of the assay was defined as the concentration at which the luminescent intensity in the sandwich LIA was twice the intensity obtained when non-specific binding was measured. Many of the conjugates gave a sensitivity below 1 ng in the sandwich assay (Fig. 2). In addition, the linearity of luminescence intensity at different conjugate concentrations per 700
600--
500-
r-
400--
= ,.d
300--
200--
100--
I
I
I
I
7
8
9
10
log HIgG, (g)
Fig. 3. LIA (sandwich) using labeled rabbit-anti-human IgG (batch A-l)in polystyrene bails coated with either 10 pg (o), 5 pg (x) or 2 gg (o) Rail per ball. ---- --, non-specific luminescence with RIgG ball and A-l; . . . . . . , non-specific luminescence with Rail bail and labeled RIgG.
168
se and in the LIA {sandwich) was excellent (Fig. 2). The LIA covered a fairly wide dynamic range extending from 1 ng to 1 pg although the increase in the luminescent intensity observed in LIA of HIgG is low when compared with the standard luminescence obtained with increasing amounts o f conjugated antibody. For some unexplained reason the curve tended to level o f f at HIgG concentrations below 1 ng. The effect was most obvious when balls were used as the solid phase (Fig. 3).
Reproducibility Because the chemiluminescent assay using the microperoxidase-H202 oxidation system is quite sensitive to interference from impurities, even though it gave the best sensitivity of the systems tested, it proved advisable to check the system with a blank (without luminescent label) and a standard luminescent solution. Under optimized conditions the variation coefficient was about 5%. In assays with background interference only a few values had to be rejected when a coefficient of variation of 10% was regarded as acceptable. The results might improve further on standardization of the coating procedure, a question that recently has resulted in several investigations (Cantarero et al., 1979; Salonen and Vaheri, 1979; Mariotti et al., 1980). REFERENCES Cantarero, L.A., J.E. Butler and J.W. Osborne, 1979, Fed. Proc. 38, 1012. Carraway, K.L., P. Spoerl and D.E. Koshland, Jr., 1969, J. Mol. Biol. 4 2 , 1 3 3 . Engvall, E. and P. Perlmann, 1971, Immunochemistry 8 , 8 7 1 . Fersht, A.R. and J. Sperling, 1973, J. Mol. Biol. 74,137. Haimovich, J., E. Hurvitz, N. Norvik and M. Sela, 1970, Biochim. Biophys. Acta 207, 125. Koltun, W.L., 1957, J. Am. Chem. Soc. 79, 5681. Lehtonen, O.P. and M.K. Viljanen, 1980, J. Immunol. Methods 34, 61. Leute, R.K. and E.F. Ullman, 1972, Nature 236, 93. Mariotti, S., J.J.-F. Oger, P. Fragur, J.P. Antel, H.-H. Kuo and L.J. DeGroot, 1980, J. Immunol. Methods 3 5 , 1 8 9 . Pratt, J.J., M.G. Woldring and L. Villerius, 1978, J. Immunol. Methods 21,179. Roswell, D.F. and E.H. White, 1978, in: Methods in Enzymology, Vol. LVII, Bioluminescence and Chemiluminescence, ed. M. DeLuca (Academic Press, New York) p. 409. Rubenstein, K.E., R.S. Schneider and E.F. Ullman, 1972, Biochem. Biophys. Res. Commun. 4 7 , 8 4 6 . Saloner~, E.-M. and A. Vaheri, 1979, J. Immunol. Methods 30, 209. Schroeder, H.R. and F.M. Yeager, 1978, Anal. Chem. 50, 1114. Schuurs, A.H.W.M. and B.K. Van Weemen, 1977, Clin. Chim. Acta 81, 1. Simpson, J.S.A., A.K. Campbell, M.E.T. Ryall and J.S. Woodhead, 1979, Nature 279, 646. Van Weemen, B.K. and A.H.W.M. Schuurs, 1971, FEBS Lett. 15,232. Whitehead, T.P., L.J. Kricka, T.J.N. Carter and G.H.G. Thorpe, 1979, Clin. Chem. 25, 1531.