Journal of Immunological Methods, 131 (1990) 83-89

83

Elsevier JIM 05624

A fast and efficient method for quantification of monoclonal antibodies in an ELISA using a novel incubation system R.E. M u s h e n s a n d M.L. S c o t t International Blood Group Reference Laboratory, Southmead Hospital, Bristol, U.K.

(Received 10 January 1990, revisedreceived14 March 1990, accepted 3 April 1990)

We have developed a sensitive, reliable, optimized ELISA to measure human IgM monoclonal antibodies using a novel shaking incubator system with short incubation periods of 15 min at 37°C for all stages. The shaking incubator is compared with a static incubator over a range of incubation times and temperatures. For each stage using static incubation conditions the system does not reach a saturation level and the results are inconsistent, unlike the shaking incubator. N o 'edge effects' are observed in the shaking system due to even heating from beneath and across the plate. The orbital shaking ensures optimal mixing of reagents which eliminates a diffusion limited reaction rate caused by a depletion of reactants at the solid phase as observed in the static system. The optimized shaking system permits economical use of reagents since the coating antibody can be used at high dilutions. Key words: ELISA; Rapid shaking incubator system; Incubation condition; Shaking incubator; Static incubator; Optimization of

antibody-antigen interaction

Introduction

Enzyme-linked immunosorbent assay (ELISA) systems (Voller et al., 1979) are used widely in laboratories which produce monoclonal antibodies. The ELISAs are used in cloning to screen large numbers of samples for the presence of desired antibodies and to quantify the amount of antibody present in those samples. These applications require that ELISA systems must be simple to perform, fast, safe, sensitive, and capable of processing large numbers of samples.

Correspondence to: R.E. Mushens, International Blood Group Reference Laboratory, Southmead Hospital, Southmead Road, Bristol BS10 5ND, U.K. Abbreoiations: Ig, immunoglobulin; HPR, horseradish peroxidase; GAH, goat anti-human; ELISA, enzyme-linkedimmunosorbent assay; Ag, antigen.

Many aspects of the solid-phase ELISA have been extensively studied, including an assessment of coating efficiency (Sorensen, 1986), the binding characteristics of immunoglobulins to capture antibodies adsorbed on plastic and their detection by antibody-enzyme conjugates (Butler et al., 1986) and the kinetics of antigen-antibody reactions at solid-liquid interfaces (Stenberg and Nygren, 1988). However, there appears to be a lack of information and standardisation of ELISA protocols with reference to incubation temperatures and times. These factors may influence the results of quantitative assays and may account for inter and intra laboratory variation. In this paper, we describe a microELISA 'sandwich' technique (Vos et al., 1982) to measure human IgM monoclonal antibodies using either a purified human IgM monoclonal antibody or a calibrated human IgM serum sample as a standard. The antibody-antigen interaction was opti-

0022-1759/90/$03.50 © 1990 ElsevierScience Publishers B.V. (BiomedicalDivision)

84 mised using a novel shaking incubator system with short incubation periods for all stages in the assay. This was compared with a conventional static incubator system and incubation periods. Both systems were evaluated over a range of different incubation periods. Once the rapid shaking incubator system had been established, we were able to obtain results quickly, that were consistent from day to day. Materials and methods Incubators Two different 37°C incubators were used. (1) A shaking incubator, Wellwarm I (Denley Instruments, Sussex). This instrument has an incubation chamber mounted on a shaking platform which is allowed to orbit freely (at 1300 orbits/min), without rotation, giving optimal mixing of reagents. The chamber is heated by a continual circulation of warm air from beneath and from the sides of the plate. There is a feedback thermostatic temperature control system giving an accuracy of + 0.5°C. (2) A conventional static upright incubator (Charles Hearson, London). Antibodies An anti-Rh(D) human monoclonal IgM antibody used as a reagent for blood grouping in our laboratory (MAD-2) was used as the unknown test sample. Goat anti-human (GAH) IgM (g chain specific) (Sigma, 1-8633) was used as the coating capture antibody. G A H - H P R IgM (g chain specific) peroxidase conjugate (Sigma, A-8650) was used as an enzyme-labelled antibody. Two IgM standards were used; one was a calibrated human serum sample (Serotec, H P R 005L), and the other was a chromatographically purified human monoclonal antibody (MAD-2). MAD-2 (1 m g / m l by ELISA) was purified using a sodium chloride gradient (5 m M sodium phosphate, 80 m M sodium chloride p H 7 to 10 mM sodium phosphate, 1 M sodium chloride, p H 7) on an LKB ion-exchange column (TSK-SP 5PW 8 × 75 mm) at a flow rate of 0.3 ml- min -a. The IgM eluted at 0.2 M NaC1 and the protein concentration was measured by absorbance at 280 nm. The purified IgM migrated as a single band in non-reducing SDS polyacrylamide gel electrophoresis.

ELISA Polystyrene flat-bottomed microtitre plates (Immulon 2, Dynatech) were coated with 75 /~1 G A H IgM (diluted 1 / 5 0 0 to 1/8000 in carbonate buffer, p H 9.6 to optimize coating concentration and 1/1000 in carbonate buffer, p H 9.6 for use in optimizing other stages in the assay). Incubation times and conditions were as described below. The wells were washed three times, allowing a 3 min delay between each wash, with phosphate-buffered saline (PBS) p H 7.2 containing 0.055% w / v Tween 20 (PBS-T) in order to block unsaturated binding sites. 75 gl of dilutions of MAD-2 in PBS-T and the standard IgM (either the calibrated serum sample or the purified MAD-2) also diluted in PBS-T to give a concentration of 1-80 n g / m l , were added in duplicate to the appropriate wells together with PBS-T as a negative control. Incubation times and conditions were as described below. The wells were washed three times using PBS-T with no delay time and 75/~1 of G A H - H P R IgM (diluted 1/2000 in PBS-T) were added to every well. Incubation times and conditions were as described below. The plates were washed three times with PBS-T and 75 gl of substrate were added (1 m g / m l o-phenylenediamine in citrate phosphate buffer, p H 5.0 containing 0.4 g l / m l of hydrogen peroxide). After a 15 min incubation at room temperature, the reaction was stopped by adding 75 gl of 1 M hydrochloric acid and the plates were read on a Titertek Multiskan ELISA reader (Flow Laboratories), at 492 nm. A standard curve relating absorbance to the concentration of IgM in the standard dilutions was constructed and used to calculate IgM concentrations in the MAD2 samples by determining the mean absorbance readings on the linear part of the standard curve. Optimization of conditions of incubation with the unknown test monoclonal antibodies and the calibrated human serum standard The coating antibody ( G A H IgM), diluted 1 in 1000, was incubated for 16 h at 4°C in static conditions. MAD-2 and the human IgM serum sample standard were incubated at 37°C for 15, 45, 90 and 180 min in a static and a shaking incubator. The peroxidase conjugate ( G A H - H P R IgM), diluted 1 in 2000, was incubated at 37°C in a static incubator for 60 min.

85

Optim&ation of conditions of incubation with the peroxidase-conjugated antibody (GAH-HPR IgM) G A H IgM, diluted 1 in 1000, was incubated for 16 h at 4°C under static conditions. MAD-2 and the human IgM serum sample standard were incubated at 3 7 ° C for 15 min in a shaking incubator. G A H - H P R IgM, diluted 1 in 2000, was incubated at 3 7 ° C for 15, 45, 90 and 180 min in a static and a shaking incubator.

TABLE I COMPARISON OF DIFFERENT SLOPES OF HUMAN IgM STANDARD a CURVES AND QUANTITATION OF HUMAN IgM TEST ANTIBODY b USING STANDARD CURVES OBTAINEDAFTER INCUBATIONAT 37 ° C OF TEST ANTIBODY AND PEROXIDASE CONJUGATE STAGES IN A STATIC AND SHAKING INCUBATOR FOR DIFFERENT TIMES Stage

Incubator Time(min) 15

Optimization of conditions of incubation with the coating capture antibody (GAH IgM) G A H IgM, diluted 1 in 1000, was incubated for 16 h at 4°C, f o r 15, 30, 45 and 60 rain at 37°C in a shaking incubator and for 60 min at 37°C in a static incubator. G A H IgM was also diluted 1 in 500, 1000, 2000, 4000 and 8000 to optimize the coating concentration in the shaking incubator for 15 min at 37°C. MAD-2 and the human IgM purified monoclonal antibody (MAD-2) standard were incubated at 37°C for 15 min in a shaking incubator. G A H - H P R IgM, diluted 1 in 2000, was incubated at 3 7 ° C for 15 min in a shaking incubator.

Results

Optimization of conditions of incubation with test antibodies and standard The standard and unknown antibody incubation times and conditions were varied whilst keeping all other incubation stages in the ELISA constant. Table I compares the different slopes of standard curves obtained after incubation for different times at 37°C in the static and shaking incubators. In the static incubator, the slope increased steadily with time, while in the shaking incubator, the slope decreased steadily after 45 min incubation. In each incubation system the MAD-2 concentration measured decreased with an increase in incubation time. In the static incubator the test antibody did not behave in a similar manner to the standard, as demonstrated by the slope of the standard curve increasing with time, whilst the quantitation of the sample decreased with time. The antigen-antibody interaction for the standard in the static incubator appeared not to have reached a saturation level

45

90

180

Test antibody Standard slope Shaking (Absorbance/ng. ml - 1) ×10-3 Static

21.2 20.3 14.9 11.5

Sample quantitation (/~g.m1-1)

52.8 49.7 43.3 40.6 57.5 57.1 46.2 35.4

Shaking Static

8.2

9.1 11.2 18.7

Peroxidase Standard slope Shaking (Absorbance/ng. ml - 1) Xl0 -3 Static

12.3 15.2 17.2 18.6

Sample quantitation (/~g.m1-1)

71.4 68.5 72.6 70.1 65.3 66.6 58.2 60.8

Shaking Static

11.8 15.8 24.0 25.1

a Calibrated human IgM serum sample standai'd (Serotec). b MAD-2 test antibody.

because the slope was still increasing at 180 min and did not reach the slope achieved after 15 min in the shaking incubator. In the shaking incubator, the sample test antibody was behaving in a similar manner to the standard as demonstrated by the direct correlation between the slope of the standard curve and the quantitation of the sample - they all decreased with incubation time. From these results, it would appear that after 15 min the antibody-antigen interaction had reached a saturation level and only as the time increased did the absorbance values decrease, possibly due to evaporation from the plate. A loss in volume of the test antibodies and standard samples would prevent all sites of the coated antibody from being saturated.

Optimization of conditions of incubation with peroxidase-conjugated antibody (GAH-HPR lgM) Once the above conditions were optimized at 37°C for 15 min in the shaking incubator, they were kept constant for the optimization of conditions of incubation with the peroxidase conjugated

86

antibody. Table I compares the different slopes of standard curves obtained after different times at 37°C in the static and shaking incubator. In both systems the standard slope increased with time. There is no significant difference with the two different methods, i.e., shaking and static, up to 45 min, but after this time the slope of the standard curve in the static incubator increased. In the shaking incubator, the quantitation of MAD-2 remains relatively constant with incubation time. This indicated that in the shaking incubator the sample and standard had reached a saturation level and were behaving similarly, i.e., both increased their absorbance values proportionally with time. However, in the static incubator, the quantitation of MAD-2 remained relatively constant between 15 and 45 rain but there was variation after this time indicating that this was a less reliable method than the shaking incubator.

Optimization of conditions of incubation with coating antibody Once the above conditions were optimized for 15 rain at 37°C in the shaking incubator, the coating antibody stage was optimized. Since a serum sample may not always behave in a similar manner as a test antibody a purified MAD-2 standard was used instead of the serum sample in order to relate the monoclonal antibody samples performance to a similar antibody sample in the ELISA assay. Three different incubation conditions were used for the coating antibody; a static 16 h incubation at 4°C (which is used by many laboratories), a 15 rain incubation at 37°C in the shaking incubator and a 60 min incubation at 37°C in the static incubator. The results (Fig. 1) indicated that a 15 rain incubation in the shaking incubator gave the optimum result. The sample and standard should behave similarly as they are the same antibody, but a difference in concentration was observed between the three different incubations (Table II) possibly due to the static systems not coating sufficient antibody evenly on to the plate. This experiment was extended to see whether 15 rain in the shaking incubator at 37°C was the optimum time of incubation for this system. The slope of the standard curves in the shaking incubator did increase with time but the quantita-

~,

14

1.2 ~,

T

1

c

o.a 06 0.4

02

0

20

40

60

80

ng/ml Human IgM standard

Fig. 1. A comparison of purified h u m a n IgM standard (MAD-2) curves after incubation at A = 1 5 min at 37°C in a shaking incubator, B = 60 min at 37°C in a static incubator and C = 16 h static incubation at 4°C of an anti-human IgM coating antibody ( G A H IgM).

tion of MAD-2 remained constant with time (Table III) indicating the sample and standard were behaving in a similar manner. However, the static incubation at 4 ° C for 16 h showed a significant increase in concentration of antibody detected due to the lower slope of the standard. This indicated that the capture antibody was coated on to the plate at a lower concentration in the static system than in the shaking incubator. T A B L E II C O M P A R I S O N O F D I F F E R E N T SLOPES OF H U M A N IgM S T A N D A R D a C U R V E S A N D Q U A N T I T A T I O N O F H U M A N IgM TEST A N T I B O D Y b U S I N G S T A N D A R D C U R V E S O B T A I N E D A F T E R I N C U B A T I O N O F COATING ANTIBODY STAGE UNDER DIFFERENT CONDITIONS Incubation condition Shaking 37°C 15 min

Static 4°C 16 h

Static 37°C 60 min

Standard slope ( A b s o r b a n c e / n g - m l - 1) x 10 -3

23.3

16.4

18.6

Sample quantitation (/xg.m1-1)

47.1

50.9

55.5

aPurified M A D - 2 h u m a n IgM antibody standard. b Several samples of M A D - 2 at a range of concentrations were tested (results not shown here) that showed the same trend of increasing concentration with the above three methods.

87 TABLE III COMPARISON OF DIFFERENT SLOPES OF HUMAN IgM STANDARDa CURVES AND QUANTITATION OF HUMAN IgM TEST ANTIBODYb USING STANDARD CURVES OBTAINED AFTER INCUBATION WITH COATING ANTIBODY STAGE UNDER DIFFERENT CONDITIONS Incubation condition Static Shakingincubator at 37 o C 4°C 15 min 30 min 45 min 60 min 16 h Standard slope (Absorbance/ng. ml- 1) X 10-3 14.0 16.0

18.0

20.0 25.0

Sample quantitation (/~g.m1-1)

39.5

39.1

Discussion 47.5 39.5

40.2

a Purified MAD-2 human IgM standard. b MAD-2 test antibody.

A further experiment was carried out in the shaking incubator with various concentration of the capture antibody. The results (Table IV) show a decrease in the standard curve slopes with increasing dilution of coating antibody, but the quantitation of MAD-2 remained constant. This shows a direct correlation between the absorbance of the sample and the standard, indicating even coating on to the plate with the capture antibody which can be used at high dilutions in the shaking system.

TABLE IV COMPARISON OF DIFFERENT SLOPES OF HUMAN IgM STANDARDa CURVES AND QUANTITATION OF HUMAN IgM TEST MAD-2 ANTIBODY USING STANDARD CURVES OBTAINEDAFTER INCUBATIONIN A SHAKING INCUBATOR AT 37°C FOR 15 MIN WITH DIFFERENT CONCENTRATIONSOF COATING ANTIBODY Dilution of coating antibody

500

1000 2000 4000 8000

Standard slope (absorbance/ng. ml- 1) >(10-3 18.0 16.4 12.4 10.4 Sample quantitation (ttg-m1-1)

The background (PBS-T) absorbance in all assays was low indicating an absence of any cross reactivity between antibody stages. Also there was an absence of 'edge effects' with the shaking incubator, i.e., where outer coated wells of the solid phase give higher absorbance readings than those of inner wells. However, in the static incubator, 'edge effects' were observed. From the experiments, 15 rain in the shaking incubator at 37°C was the optimum incubation for all the stages in our ELISA. Similar results were obtained throughout with two other IgM monoclonal antibodies.

6.0

22.8 23.0 23.2 22.4 22.2

" Purified MAD-2 human IgM standards.

The optimal time and temperature for incubation steps in an ELISA have to be established for each system. Therefore a systematic investigation of the various experimental parameters must reveal the minimum time required for reproducible results. Coating of the solid phase is generally achieved much faster at 37°C than under commonly used conditions of overnight at 4°C and reaction times can be cut depending on the partitular assay requirements (Mason, 1980). This is because increasing the temperature increases the rate of the reaction. This reflects the higher diffusion rate and therefore the higher collision rate of particles with the solid phase. Our results confirm this as 4°C was not as efficient as 37°C for binding of the capture antibody to the solid phase. ELISA assays are performed using microtitre plates in which the samples to be analysed are placed in cylindrical wells. A limited diffusion reaction scheme has been applied to this particular geometry and condition of a non-stirred solution (Stenberg et al., 1988). These authors found that reactions at the solid-liquid interface could be diffusion limited due to depletion of reactants close to the solid-phase surface which could in turn depend on the surface concentration of the receptor molecules. In the shaking incubator, however, the shaking ensures the mixing of reagents to give a stirred solution which eliminate problems encountered with low diffusion rates in the nonstirred solution. The stirring may follow more closely in vivo antibody-antigen reactions by mimicking naturally occurring cell surface anti-

88 body-antigen interactions w h i c h are not normally diffusion limited. For the unstirred solution, the binding rate of the initial antibody and subsequent antibody binding stages seemed to become diffusion rate limited at the plate surface and did not reach a saturation level. Mattila (1985) has estimated an error of between 10-70-fold when using myeloma proteins bound to plastic as standards in assaying Ig bound to a specific Ag. However Our system uses anti-lg antibodies to capture Ig to be quantified. The use of high quality purified antibodies in both the capture and indicator steps ( G A H - I g M and G A H H P R IgM) is very importan! in maximising assay sensitivity and reliability (Fleming, 1988). Pruslin (1986) provides data to support our findings that a reference serum or purified antibody of known lg content should be included inlorder to compensate for plate to plate variability. The background values in our ELISA were extremely low, indicating a sensitive assay and no cross reactivity between the antibody stages. Coating of microtitre plates with antibody has been shown to be optimal at a certain concentration, above which coating was less efficient (Sorensen, 1986). Our results iridicated that in the optimized shaking system the Coating concentration was not as critical as previously has been observed with the static system. I The capture antibody could be coated at high dilutions which is an important consideration whe n using reagents which are in short supply or are~ expensive. The complex between one a~atibody and two epitopes ('divalent binding') is more stable than the complex between one antibody and one epitope ('univalent binding') (Vos, J987). The ratio between divalent and univalent bindings depends on the epitope density bound t0 the solid phase and on antibody coating concentration. In our system using the shaking incubator, the results indicated an optimal even coating of the plate with antibody promoting stable binding of the test sample. 'Edge effects' were eliminated in the shaking system because of the even heating across and from beneath the plate by a continual warm air flow. In the static incubator 'edge effects' were observed because outer wells of the solid phase reached 37°C faster than inner wells.

Conclusion We have established an ELISA system in our laboratory using a novel shaking incubator for the detection and measurement of monoclonal antibodies using short incubation times of 15 min at 37°C for all stages. Once the rapid shaking system had been established we were able to extend its use to other ELISA procedures for the measurement of h u m a n I g G and mouse IgM, I g G and IgA monoclonal antibodies. This ELISA is sensitive, reliable, consistent, quick and easy to perform. Its use m a y be invaluable in areas where large numbers of samples must be processed quickly and efficiently.

Acknowledgements The authors wish to thank Mr. A. Guest for the purification of the M A D - 2 used as a standard in the ELISA.

References Butler, J.E., Spradling, J.E., Suter, M., Dierks, S.E., Heyermann, H. and Peterman, J.H. (1986) The immunochemistry of sandwich ELISA I. The binding characteristics of immunoglobulins to monoclonal and polyclonal capture antibodies adsorbed on plastic and their detection by symmetrical and asymmetrical antibody-enzyme conjugates. Mol. Immunol. 23, 971. Fleming, J.O. and Pen, L.B. (1988) Measurement of the concentration of murine IgG monoclonal antibody in hybridoma supernatants and ascites in absolute units by sensitive and reliable enzyme linked immunosorbent assays (ELISA). J. Immunol. Methods 110, 11. Mason, D.W. and Williams, A.F. (1980) The kinetics of antibody binding to membrane antigens in solution and at the cell surface. Biochem. J. 187, 1. Mattila, P.S. (1985) Quantitation of antibody isotypes in solid-phase assays. Comparison of myeloma protein and monoisotypic antibody standards. J. Immunol. Methods 83, 43. Pruslin, F.H., Roman, R.C., Jones, J. and Winston, R. (1986) An ELISA for IgM titre of human serum. J. Immunol. Methods 94, 99. Sorensen, K. and Brodbeck, V. (1986) Assessment of coatingefficiency in ELISA plates by direct protein determination. J. Immunol. Methods 95, 291. Stenberg, M. and Nygren, H. (1988) Kinetics of antigen-antibody reactions at solid-liquid interfaces. J. Immunol. Methods 113, 3.

89 Stenberg, M., Werth6n, M., Theander, S. and Nygren, H. (1988) A diffusion limited reaction theory for a microtitre plate assay. J. Immunol. Methods 112, 23. Voller, A., BidweU, D.E. and Bartlett, A. (1979) The enzyme linked immunosorbent assay (ELISA). Dynatech Laboratories California.

Vos, J.G., Krajnc, E.I. and Beekhof, P. (1982) Use of the enzyme linked immunosorbent assay (ELISA) in immunotoxicity testing. Environ. Health Perspect. 43, 115. Vos, Q., Klasen, E.A. and Haaijman, J.J. (1987) The effect of divalent and univalent binding on antibody titration curves in solid-phase ELISA. J. Immunol. Methods 103, 47.

A fast and efficient method for quantification of monoclonal antibodies in an ELISA using a novel incubation system.

We have developed a sensitive, reliable, optimized ELISA to measure human IgM monoclonal antibodies using a novel shaking incubator system with short ...
508KB Sizes 0 Downloads 0 Views