- Email: [email protected]
Comparison of direct and indirect two-site binding enzyme immunoassay
Clinica
C&mica Acta, 122 (1982)
Eisevier Biomedical
1-9
Press
CGA 2119
Comparison of direct and indirect two-site binding enzyme immunoassay Baerbel Porstmann UDivisian
as*, T. Porstmann b, Raisa Seifert a, Hdga Meisel ’ and Elsa Nugel a
of Ciinicaf ~~oc~ern~str~, b Division o,f Clinical fmmunotogy and ’ Division of Virology,
Hum~ldt-Unjversit~
Berlin, department
ofMedicine
(Received
~Ch~r~t~~, Ziegelstrasse 5- 12, 1040 Berlin (GDR)
November
24th. 1981)
Summary
Rabbit IgG, directed against HBsAg, was purified by positive and by negative affinity chromatography and applied in horseradish peroxidase labelled as well as in unlabelled form in the direct and indirect two-site binding enzyme immunoassay (EIA), Comparing direct and indirect assay the latter is more sensitive and less conjugate consuming. In contrast to the indirect assay in which antibodies, purified by positive affinity chromato~aphy~ do not alter detection limit, a 4- to &fold higher sensitivity was achieved in the direct EIA in contrast to antibodies, purified by negative affinity chromatography. In the indirect EIA unlabelled second and labelled third antibodies were incubated successively as well as simultaneously. The latter procedure shortened the assay time but needed antibodies purified by positive affinity chromatography and a IO-fold higher conjugate concentration. Greatest sensitivity was obtained in the indirect EIA by the use of labelled second and labelled third antibodies (20-30 rig/l HBsAg).
Two-site binding assays with labelled antibodies are easily automated and, therefore, the most frequently employed type of EIA. For antigen determination the direct sandwich assay is mainly used in which the antigen specific second antibody is coupled to a marker enzyme. Although the application of isotope-labelled anti species antibodies as universal reactants has been described for immunoradiometric techniques of antigen and
* To whom correspondence
0009”8981/82/0000-0000/$02.75
should
be addressed.
Q Etsevier
Biomedical
Press
2
haptene quantification [l], and also enzyme-labelled anti species antibodies are commercially available, there are only few reports on the use of such conjugates in EIA for antigen determination [2,3]. These labelled immune reactants have the advantage of universal applicability for a variety of antigen determinations as for example in the indirect ELISA technique for detection of specific antibodies. We therefore compared the direct and the indirect two-site binding assay for hepatitis B surface antigen (HBsAg), using identical immune reactants and assay conditions, and discuss the advantages and disadvantages of the two methods.
Materials and methods Purification of HBsAg and production of anti HBs antibodies HBsAg was prepared as described by Nath et al. 141, further purified by ultracentrifugation and used to immunize guinea pigs and rabbits [5]. Antibodies were purified by the immunosorbent technique. preparation of conjugates Crude horseradish peroxidase (HRP) (AWD, Dresden, purity number (PN = absorbance 403 nm/absorbance conjugated with various antibody preparations by the method [7] at a molar ratio of 12 : 1 (HRP to IgG). For the antibodies were labelled:
GDR) was purified to a 280 nm) of 2.8 [6] and two-step glutaraldehyde direct assay the following
Rabbit IgG monos~eci~c for HBsAg, prepared by negative af~nity chromutogra~hy (the antibodies against normal serum proteins were removed on a macroglobuiin fraction of HBsAg-negative human serum, coupled to CNBr-Sepharose (Pharmacia Fine Chemicals, Uppsala, Sweden)). The content of specific anti HBs antibodies was 10% of the purified IgG fraction. Rabbit IgG monospecific for HBsAg, prepared by positive affinity chromatography (anti HBs containing y-globulin was chromatographed on purified insolubilized HBsAg and specific antibodies were eluted with potassium rhodanide (3 mol/l)). For the indirect assay sheep IgG monospecific for rabbit IgG was purified by positive affinity chromatography on insolubilized rabbit IgG, eluted with potassium rhodanide (3 mol/l) and then labelled with HRP [5]. The conjugates were purified by affinity chromatography on Concanavalin ASepharose (Pharmacia Fine Chemicals, Uppsala, Sweden) and gel filtration on Sephadex G-200 (Pharmacia Fine Chemicals, Uppsala, Sweden) [8,9]. Assays Polystyrene tubes coated with 0.2 ml centration 10 mg/l, phosphate buffered different types of
(VEB guinea for saline assay
Polyplast, Halberstadt, GDR) acting as solid phase were pig anti HBs y-globulin (first antibody), protein con20 h at room temperature, washed 3 times with 1 ml (PBS), dried and stored until use at -2O’C. For the a standard serum was diluted geometrically in PBS,
3
containing 0.5 mmol/l human serum albumin (HSA). Standards were recorded around 0.02264 f.cg/l HBsAg. In all experiments, HBsAg was incubated for 16 h at 2O’C; the labelled and unlabelled antibodies (rabbit anti HBs antibody as second, sheep anti rabbit IgG as third antibody) were incubated for 2 h at 37°C. Unbound reactants were removed by rinsing the tubes three times with 1 ml PBS. The substrate reaction, using as chromogen ABTS (Boehringer, Mannheim, FRG), was run for 30 min at 20°C and was stopped by adding azide [lo,1 11. The absorbances were registered automatically at 41.5 nm wavelength using a Spekol 20 analysis system (VEB Carl Zeiss Jena, GDR). An HBsAg negative pooled serum was used as a blank control and its absorbance subtracted from that of the HBsAg containing samples. The detection limit of HBsAg was taken to be the average absorbance of the blank plus 0.1 delta absorbance. For comparison, the assay was performed with Direct two-site binding assay. labelled rabbit anti HBs antibodies, purified by positive and negative affinity chromatography. The conjugates were adjusted with PBS, containing 0.5 mmol/l HSA to conjugate concentrations of 5, 10 and 20 mg/l. For examining the influence of the molar ratio of HRP to IgG as well as the effect of rising concentrations of uncoupled HRP in the conjugates on the position of the standard curve, the antibody containing fractions obtained by gel filtration were divided into four groups according to the sodium dodecyl sulphate polyacryl amide gel electrophoretic pattern, pooled, and adjusted to a concentration of 10 mg protein/l for the assay. assay was performed in three Indirect two-site binding assa,v. The indirect different ways. Successive incubation of the second and the labeiied third antibodies: the incubations were performed separately, with intervening washings. By varying the protein concentrations of both antibodies, the optimal concentration ratio was. determined. Simultaneous incubation of second and third antibodies: the antibodies were preincubated together at different concentration ratios either for 20 h at room temperature or for 2 h at 37°C in order to form immune complexes and subsequently transferred into the assay tubes. The preformed compIexes reacted for 2 h at 37°C with HBsAg, bound to the wall-fixed antibodies. In both of indirect assay, identical amounts of rabbit anti HBs IgG, purified by positive or negative affinity chromatography were used as the second antibodies. Successive incubation of Iabelied second and labelled third antibodies: HRP labelled rabbit anti HBs IgG, purified by positive affinity chromatography was used as the second antibody with which the labelled third antibodies reacted in the following step. Results Direct two-site binding assay Although antibody-enzyme
conjugates
of the IgG prepared
either by positive
or
4
Q25 05
10 20
LO
80 160 320
C0timtKuldHBsAgl~l1l
Fig. 1. Standard curves for HBsAg in the direct EIA. The HRP labelled second antibodies were anti HBs-IgG, purified by positive (closed symbols) or negative (open symbols) affinity chromatography. Conjugate concentrations: 5 mg/l (A, A), 10 mg/l (0.0) and 20 mg/l (0, n ); -.-.-.-.-.-, detection limit for HBsAg.
by negative affinity chromatography differ neither in their PN nor in specific enzyme activity, the former yielded steeper curves in the EIA and a 4-8 times lower detection limit for HBsAg of about 0.5-l pg/l (Fig. 1). The further subdivision of antibody-containing fractions after gel filtration into those with a molar ratio HRP to IgG of 2: 1 (Fig. 2, fraction 1) and those with a ratio of 1: 1 (Fig. 2, fraction 2-4) had no influence on the detection limit. However, the blank value of the assay increased systematically from fraction 1 to 4, due to rising amounts of free HRP in the conjugate, binding non specifically to the solid phase. Indirect
two-site binding assay
Successive incubation with second and labeled third antibodies: when using rabbit anti HBs, purified by positive or negative affinity chromatography as the unlabelled second antibodies, the detection limit was constantly 4 to 16 times lower (0.25-0.5 pg HBsAg/l) than in the direct assay (2-4 pg HBsAg/l) with the same antibodies, but labelled with enzyme Figs. 1 and 3). If unlabelled second antibodies, purified by positive affinity chromatography are employed in the indirect assay, the sensitivity cannot be further enhanced (Fig. 3). Simultaneous incubation of the second and the Iabelled third antibodies: the simultaneous incubation of the second and the labelled third antibodies approaches the sensitivity of the successive incubation procedure (0.25 pg HBsAg/l) only if anti Hbs IgG, purified by positive affinity chromatography is used as second antibody (Fig. 3). However, a 10 times greater amount of conjugate (10 mg/l) is necessary in comparison with the successive incubation (1 mg/l). Reducing the concentration of
5
Absorbance
WSnm
f 30 mm)
9
12” 11 ” 1.0 .’
I 0
09 ‘. 0.8
0I
07 .’
05 06 -’ .’
i 0
*I
0.4 _’ 0.3 .. 02 I.
’
/*
o,..-_.YL.-._. M’ 05
&i@! /
LJ _s&+-.10
20
40
Concentration
80
160 320 a.0
af HBsAg (pgll)
Absorbance
170 205 24c 275 310 w
380 415 150 &85 520 ml elutlon volume
Fig. 2. Standard curves for HBsAg in the direct EIA (closed symbols), above, using different conjugate fractions following gel filtration (fraction 1 to 4). below, labelled antibodies 10 mg/I, and standard curves for HBsAg in the indirect EIA with successive incubation (0), second antibodies 10 “g/l, labelled third antibodies 1mg/l. For the direct and indirect assay second antibodies purified by negative affinity chromatography were used; -, absorbance 280 nm; -- - - - -. absorbance 403 nm.
labelled third antibodies by one half raises the detection limit of HBsAg 8-fold (2 pg/l); reducing the second antibodies by one half only leads to a 4-fold increase (1 pg/l). When second antibodies, purified by negative affinity chromatography, were used in the simultaneous incubation, the detection limit of HBsAg increased 16 times to 4pg/l, but a 20-fold higher concentraiion of conjugate was necessary
6
(Fig. 4). A further enhancement was limited by increasing blank values. The sensitivity reached in this way corresponds to that of the direct two-site binding assay using the same, but enzyme-labelled, anti HBs antibodies (Fig. 3).
06 02
I--
Fig. 3. Standard curves for HBsAg in the indirect EIA (closed symbols) following simultaneous and successive incubation of unlabelled second and labelled third antibodies, in comparison to direct EIA (open symbols); the antigen specific second antibodies were purified by positive or negative affinity chromatography.
Symbol
Kind of purification
Cont. of second ab. (* labelled form)
0
pos. pos. neg. neg. neg. pos.
10 10 10 10 10 10
: n 0 A
aff. aff. aff. aff. aff. aff.
chrom. chrom. chrom. chrom. chrom. chrom.
mg/l mg/l mg/l mg/l mg/l mg/l
Cont. of third ab.
Technique of incub.
1 w/*
successive simultan. successive
10 mg/l
1 w/l * *
20 mg/l _ _
The temperature and the time course of the incubations with second and third antibodies for complex formation had no influence upon sensitivity of the assay. Successive incubation with labelied second and labelled third antibodies: when both the second and the third antibodies, prepared by positive affinity chromatography, were applied in the enzyme-labelled form in the successive incubation procedures, a significant increase in sensitivity resulted (Fig. 5). The detection limit of HBsAg in this double labelled EIA fell to 20-30 rig/l and the amount of second antibody could be reduced from 10 mg/l rabbit anti HBs IgG to 2 mg/l HRP conjugate, attaining&e sensitivity of the indirect EIA with unlabelled second antibodies (0.25 pg HBsAg/l).
7
Discussion According to our results, the separation of conjugates into higher and lower labelled antibodies has no effect on the sensitivity of the assay, but does contribute
025 05
10 20
LO
80 160 320
Cmcentrotim of H&@
t*r Blank mlues
l,q/ll
Fig. 4. Standard curves for HBsAg in the indirect EIA with simultaneous incubation and labelled third antibodies. The second antibodies were purified by positive negative affinity chromatography (open symbols).
Symbol
Cont.
w 0 A q 0 A
IO 5 IO IO 10 10
of second
mg/l mg/l mg/l mg/l mg/l mg/l
ab.
Cont.
of unlabelled second (closed symbols) or
of third ab.
10 mg/l 10 mg/l 5 20 15 IO
mg/l mg/l mg/l mg/l
to increasing blank values, due to the presence of free polymeric HRP in the low molecular mass fractions. High blank values prohibit an increase of conjugate concentration. Therefore, preceding the coupling reaction with activated HRP, the polymers should be removed by gel filtration [ 121. For the direct sandwich EIA it is necessary to purify the second antibodies by positive affinity chromatography in order to achieve high sensitivity. The indirect sandwich EIA for HBsAg generally allows an increase of sensitivity which cannot be further improved by the preparative isolation of an antigen specific second antibody. This offers a great advantage compared with the direct assay since the antigen to be determined is often very difficult to prepare in the desired amounts and purity for the immunosorbent technique. On the other hand, the preparation of species specific antibodies by positive affinity chromatography is much less problematic since it is easy to prepare large amounts of IgG for insolubilization from that species donating the second antibodies.
8
The approximately doubled sensitivity of the indirect assay (unlabelled second antibodies) over that of the most sensitive variation of the direct assay is caused by the fact that the second antibodies act as an amplification factor and fully retain their immunological reactivity, having circumvented purification by positive affinity
t GUl5IM3 QXi 012 025 050 100 200 LW corce~mtmof
d
Ellonkmlues
HBSAg f/_g/ll
Fig. 5. Standard curves for HBsAg in the indirect EIA with successive incubation using labelled antibodies (10 mg/l) and labelled third antibodies (1 mg/l), closed circles, and unlabelled antibodies (10 me/l) and labelled third antibodies (1 mg/l), open circles. As second antibodies, anti HBs-IgG were used, purified by positive affinity chromatography.
second second rabbit
chromatography and labelling. An extraordinary increase in sensitivity can be gained with the indirect technique when both the second and the third antibodies are employed in enzyme labelled form. The detection limit can thus be lowered 8-16 times compared with the optimal direct assay technique. This effect is, however, probably restricted to second antibodies with a low degree of labelling (in our case l-2 mol HRP/mol IgG) and low molecular mass marker enzymes as for example HRP or lysozyme, in order to prevent steric hindrance of the binding of the third antibodies. The indirect assay offers the additional advantage of an universal antigen independent applicability of the enzyme labelled anti species specific antibodies. One disadvantage is the necessity to raise antigen specific first and second antibodies in two different species; another, the longer duration of the indirect compared to the direct assay, due to a further incubation and washing step. By preincubation of second and third antibodies in parallel to the antigen incubation (simultaneous incubation technique) and subsequent transfer of the preformed immune complexes into the test tubes, the duration can be reduced to that of the direct assay. The sensitivity of the successive incubation procedure can be approached only by employing second antibodies, purified by positive affinity chromatography and by increasing the concentration of the labelled third antibodies 10 times. The necessity for such a high concentration of conjugate is caused by its ability to react with a much greater amount of second antibodies regardless of the
9
antigen binding by the latter. On the other hand in the successive technique the non reacting second antibodies have been removed by washing prior to the reaction with the third antibodies. it is clear that saturation occurs at higher conjugate concentrations and that, therefore, the simultaneous incubation procedure of the indirect two-site binding assay is uneconomical. The possibility of shortening assay time by simultaneous incubation of antigen and labelled second antibodies, using monoclonal antibodies with a different determinant specificity, together with a high sensitivity and economical antibody utilization, was recently demonstrated for a two-site binding assay of alphafetoprotein [ 131. References
10 11 12
13
Stafford JEH, KiBgallan W. Universal reagent immunosorbent assay (uri) for haptens. J Immunoi Methods 1980; 34: 339-343. Belanger L, Sylvestre C, Dufour D. Enzyme-linked immunoassay for alpha-fetoprotein by competitive and sandwich procedures. Clin Chim Acta 1973; 48: 15- 18. Porstmann T, Porstmann B, Schmechta H, Nugel E, Meisel H, Geserick G. Development of a double sandwich enzyme immunoassay for the quantitative analysis of HBsAg, Part I. Dtsch Gesundheitswes 1980; 35: 598-600. Nath N, Mazzur S, Ledman R, Fang CT. Purification of hepatitis B surface antigen using polyethylene glycol, pepsin and Tween 80. VOX Sang 1976; 31, suppl. 1: 84-89. Porstmann T, Porstmann B, Schmechta H, Nugel F, Meisel H, Geserick G. Development of a double sandwich enzyme immunoas~y for the quantitative analysis of HBsAg, Part 2. Dtsch Gesundheitswes 1980; 35: 1587-1593. Porstmann B, Porstmann T. Reinigung handels~bIicher Meerrettich-Peroxidase fur den Einsatz im Enzymimmunoassay. 2 Med Labor-Diagn 1979; 20: 87-95. Avrameas S, Ternynck T. Peroxidase labelled antibody and Fab conjugates with enhanced intracellular penetration. Immunochemistry 1971; 8: 1175-I 179. Porstmann T, Porstmann B. A simple and effective method for purification of crude horse radish peroxidase (HRP) and HRP-conjugated antibodies. 12th FEBS-Meeting, Dresden 1978. Abstract Nr. 0734. Porstmann T, Porstmann B. Purification of horse radish peroxidase conjugated antibodies by affinity chromatography. Comparison of glutaraldehyde and periodate conjugation. Acta Biol Med Germ 1979; 38: 1039-1054. Gallati H. Peroxidase aus Meerrettich: Kinetische Studien sowie Optimierung der Aktivitatsbestimmung mit den Substraten H,O, und ABTS. J Chn Chem Chn Biochem 1979; 17: 1-7. Persijn JP, Jonker KM. A terminating reagent for the peroxidase-Iabe~led enzyme immunoassay. J Clin Chem Clin B&hem 1978; 16: 531-532. Boorsma DM, Streefkerk JG. Some aspects of the preparation, analysis and use of peroxidase-antibody conjugates in immunochemistry. In: 24th Coil. Protides of the Biological Fluids. H. Peeters, ed., Oxford: Pergamon Press, 1976: 795-804. Uotila M, Ruoslathi E, Engvall E. Two-site sandwich enzyme immunoassay with monocional antibodies to human alpha-fetoprotein. J Immunol Methods 1981; 42: 1I-15.
C&mica Acta, 122 (1982)
Eisevier Biomedical
1-9
Press
CGA 2119
Comparison of direct and indirect two-site binding enzyme immunoassay Baerbel Porstmann UDivisian
as*, T. Porstmann b, Raisa Seifert a, Hdga Meisel ’ and Elsa Nugel a
of Ciinicaf ~~oc~ern~str~, b Division o,f Clinical fmmunotogy and ’ Division of Virology,
Hum~ldt-Unjversit~
Berlin, department
ofMedicine
(Received
~Ch~r~t~~, Ziegelstrasse 5- 12, 1040 Berlin (GDR)
November
24th. 1981)
Summary
Rabbit IgG, directed against HBsAg, was purified by positive and by negative affinity chromatography and applied in horseradish peroxidase labelled as well as in unlabelled form in the direct and indirect two-site binding enzyme immunoassay (EIA), Comparing direct and indirect assay the latter is more sensitive and less conjugate consuming. In contrast to the indirect assay in which antibodies, purified by positive affinity chromato~aphy~ do not alter detection limit, a 4- to &fold higher sensitivity was achieved in the direct EIA in contrast to antibodies, purified by negative affinity chromatography. In the indirect EIA unlabelled second and labelled third antibodies were incubated successively as well as simultaneously. The latter procedure shortened the assay time but needed antibodies purified by positive affinity chromatography and a IO-fold higher conjugate concentration. Greatest sensitivity was obtained in the indirect EIA by the use of labelled second and labelled third antibodies (20-30 rig/l HBsAg).
Two-site binding assays with labelled antibodies are easily automated and, therefore, the most frequently employed type of EIA. For antigen determination the direct sandwich assay is mainly used in which the antigen specific second antibody is coupled to a marker enzyme. Although the application of isotope-labelled anti species antibodies as universal reactants has been described for immunoradiometric techniques of antigen and
* To whom correspondence
0009”8981/82/0000-0000/$02.75
should
be addressed.
Q Etsevier
Biomedical
Press
2
haptene quantification [l], and also enzyme-labelled anti species antibodies are commercially available, there are only few reports on the use of such conjugates in EIA for antigen determination [2,3]. These labelled immune reactants have the advantage of universal applicability for a variety of antigen determinations as for example in the indirect ELISA technique for detection of specific antibodies. We therefore compared the direct and the indirect two-site binding assay for hepatitis B surface antigen (HBsAg), using identical immune reactants and assay conditions, and discuss the advantages and disadvantages of the two methods.
Materials and methods Purification of HBsAg and production of anti HBs antibodies HBsAg was prepared as described by Nath et al. 141, further purified by ultracentrifugation and used to immunize guinea pigs and rabbits [5]. Antibodies were purified by the immunosorbent technique. preparation of conjugates Crude horseradish peroxidase (HRP) (AWD, Dresden, purity number (PN = absorbance 403 nm/absorbance conjugated with various antibody preparations by the method [7] at a molar ratio of 12 : 1 (HRP to IgG). For the antibodies were labelled:
GDR) was purified to a 280 nm) of 2.8 [6] and two-step glutaraldehyde direct assay the following
Rabbit IgG monos~eci~c for HBsAg, prepared by negative af~nity chromutogra~hy (the antibodies against normal serum proteins were removed on a macroglobuiin fraction of HBsAg-negative human serum, coupled to CNBr-Sepharose (Pharmacia Fine Chemicals, Uppsala, Sweden)). The content of specific anti HBs antibodies was 10% of the purified IgG fraction. Rabbit IgG monospecific for HBsAg, prepared by positive affinity chromatography (anti HBs containing y-globulin was chromatographed on purified insolubilized HBsAg and specific antibodies were eluted with potassium rhodanide (3 mol/l)). For the indirect assay sheep IgG monospecific for rabbit IgG was purified by positive affinity chromatography on insolubilized rabbit IgG, eluted with potassium rhodanide (3 mol/l) and then labelled with HRP [5]. The conjugates were purified by affinity chromatography on Concanavalin ASepharose (Pharmacia Fine Chemicals, Uppsala, Sweden) and gel filtration on Sephadex G-200 (Pharmacia Fine Chemicals, Uppsala, Sweden) [8,9]. Assays Polystyrene tubes coated with 0.2 ml centration 10 mg/l, phosphate buffered different types of
(VEB guinea for saline assay
Polyplast, Halberstadt, GDR) acting as solid phase were pig anti HBs y-globulin (first antibody), protein con20 h at room temperature, washed 3 times with 1 ml (PBS), dried and stored until use at -2O’C. For the a standard serum was diluted geometrically in PBS,
3
containing 0.5 mmol/l human serum albumin (HSA). Standards were recorded around 0.02264 f.cg/l HBsAg. In all experiments, HBsAg was incubated for 16 h at 2O’C; the labelled and unlabelled antibodies (rabbit anti HBs antibody as second, sheep anti rabbit IgG as third antibody) were incubated for 2 h at 37°C. Unbound reactants were removed by rinsing the tubes three times with 1 ml PBS. The substrate reaction, using as chromogen ABTS (Boehringer, Mannheim, FRG), was run for 30 min at 20°C and was stopped by adding azide [lo,1 11. The absorbances were registered automatically at 41.5 nm wavelength using a Spekol 20 analysis system (VEB Carl Zeiss Jena, GDR). An HBsAg negative pooled serum was used as a blank control and its absorbance subtracted from that of the HBsAg containing samples. The detection limit of HBsAg was taken to be the average absorbance of the blank plus 0.1 delta absorbance. For comparison, the assay was performed with Direct two-site binding assay. labelled rabbit anti HBs antibodies, purified by positive and negative affinity chromatography. The conjugates were adjusted with PBS, containing 0.5 mmol/l HSA to conjugate concentrations of 5, 10 and 20 mg/l. For examining the influence of the molar ratio of HRP to IgG as well as the effect of rising concentrations of uncoupled HRP in the conjugates on the position of the standard curve, the antibody containing fractions obtained by gel filtration were divided into four groups according to the sodium dodecyl sulphate polyacryl amide gel electrophoretic pattern, pooled, and adjusted to a concentration of 10 mg protein/l for the assay. assay was performed in three Indirect two-site binding assa,v. The indirect different ways. Successive incubation of the second and the labeiied third antibodies: the incubations were performed separately, with intervening washings. By varying the protein concentrations of both antibodies, the optimal concentration ratio was. determined. Simultaneous incubation of second and third antibodies: the antibodies were preincubated together at different concentration ratios either for 20 h at room temperature or for 2 h at 37°C in order to form immune complexes and subsequently transferred into the assay tubes. The preformed compIexes reacted for 2 h at 37°C with HBsAg, bound to the wall-fixed antibodies. In both of indirect assay, identical amounts of rabbit anti HBs IgG, purified by positive or negative affinity chromatography were used as the second antibodies. Successive incubation of Iabelied second and labelled third antibodies: HRP labelled rabbit anti HBs IgG, purified by positive affinity chromatography was used as the second antibody with which the labelled third antibodies reacted in the following step. Results Direct two-site binding assay Although antibody-enzyme
conjugates
of the IgG prepared
either by positive
or
4
Q25 05
10 20
LO
80 160 320
C0timtKuldHBsAgl~l1l
Fig. 1. Standard curves for HBsAg in the direct EIA. The HRP labelled second antibodies were anti HBs-IgG, purified by positive (closed symbols) or negative (open symbols) affinity chromatography. Conjugate concentrations: 5 mg/l (A, A), 10 mg/l (0.0) and 20 mg/l (0, n ); -.-.-.-.-.-, detection limit for HBsAg.
by negative affinity chromatography differ neither in their PN nor in specific enzyme activity, the former yielded steeper curves in the EIA and a 4-8 times lower detection limit for HBsAg of about 0.5-l pg/l (Fig. 1). The further subdivision of antibody-containing fractions after gel filtration into those with a molar ratio HRP to IgG of 2: 1 (Fig. 2, fraction 1) and those with a ratio of 1: 1 (Fig. 2, fraction 2-4) had no influence on the detection limit. However, the blank value of the assay increased systematically from fraction 1 to 4, due to rising amounts of free HRP in the conjugate, binding non specifically to the solid phase. Indirect
two-site binding assay
Successive incubation with second and labeled third antibodies: when using rabbit anti HBs, purified by positive or negative affinity chromatography as the unlabelled second antibodies, the detection limit was constantly 4 to 16 times lower (0.25-0.5 pg HBsAg/l) than in the direct assay (2-4 pg HBsAg/l) with the same antibodies, but labelled with enzyme Figs. 1 and 3). If unlabelled second antibodies, purified by positive affinity chromatography are employed in the indirect assay, the sensitivity cannot be further enhanced (Fig. 3). Simultaneous incubation of the second and the Iabelled third antibodies: the simultaneous incubation of the second and the labelled third antibodies approaches the sensitivity of the successive incubation procedure (0.25 pg HBsAg/l) only if anti Hbs IgG, purified by positive affinity chromatography is used as second antibody (Fig. 3). However, a 10 times greater amount of conjugate (10 mg/l) is necessary in comparison with the successive incubation (1 mg/l). Reducing the concentration of
5
Absorbance
WSnm
f 30 mm)
9
12” 11 ” 1.0 .’
I 0
09 ‘. 0.8
0I
07 .’
05 06 -’ .’
i 0
*I
0.4 _’ 0.3 .. 02 I.
’
/*
o,..-_.YL.-._. M’ 05
&i@! /
LJ _s&+-.10
20
40
Concentration
80
160 320 a.0
af HBsAg (pgll)
Absorbance
170 205 24c 275 310 w
380 415 150 &85 520 ml elutlon volume
Fig. 2. Standard curves for HBsAg in the direct EIA (closed symbols), above, using different conjugate fractions following gel filtration (fraction 1 to 4). below, labelled antibodies 10 mg/I, and standard curves for HBsAg in the indirect EIA with successive incubation (0), second antibodies 10 “g/l, labelled third antibodies 1mg/l. For the direct and indirect assay second antibodies purified by negative affinity chromatography were used; -, absorbance 280 nm; -- - - - -. absorbance 403 nm.
labelled third antibodies by one half raises the detection limit of HBsAg 8-fold (2 pg/l); reducing the second antibodies by one half only leads to a 4-fold increase (1 pg/l). When second antibodies, purified by negative affinity chromatography, were used in the simultaneous incubation, the detection limit of HBsAg increased 16 times to 4pg/l, but a 20-fold higher concentraiion of conjugate was necessary
6
(Fig. 4). A further enhancement was limited by increasing blank values. The sensitivity reached in this way corresponds to that of the direct two-site binding assay using the same, but enzyme-labelled, anti HBs antibodies (Fig. 3).
06 02
I--
Fig. 3. Standard curves for HBsAg in the indirect EIA (closed symbols) following simultaneous and successive incubation of unlabelled second and labelled third antibodies, in comparison to direct EIA (open symbols); the antigen specific second antibodies were purified by positive or negative affinity chromatography.
Symbol
Kind of purification
Cont. of second ab. (* labelled form)
0
pos. pos. neg. neg. neg. pos.
10 10 10 10 10 10
: n 0 A
aff. aff. aff. aff. aff. aff.
chrom. chrom. chrom. chrom. chrom. chrom.
mg/l mg/l mg/l mg/l mg/l mg/l
Cont. of third ab.
Technique of incub.
1 w/*
successive simultan. successive
10 mg/l
1 w/l * *
20 mg/l _ _
The temperature and the time course of the incubations with second and third antibodies for complex formation had no influence upon sensitivity of the assay. Successive incubation with labelied second and labelled third antibodies: when both the second and the third antibodies, prepared by positive affinity chromatography, were applied in the enzyme-labelled form in the successive incubation procedures, a significant increase in sensitivity resulted (Fig. 5). The detection limit of HBsAg in this double labelled EIA fell to 20-30 rig/l and the amount of second antibody could be reduced from 10 mg/l rabbit anti HBs IgG to 2 mg/l HRP conjugate, attaining&e sensitivity of the indirect EIA with unlabelled second antibodies (0.25 pg HBsAg/l).
7
Discussion According to our results, the separation of conjugates into higher and lower labelled antibodies has no effect on the sensitivity of the assay, but does contribute
025 05
10 20
LO
80 160 320
Cmcentrotim of H&@
t*r Blank mlues
l,q/ll
Fig. 4. Standard curves for HBsAg in the indirect EIA with simultaneous incubation and labelled third antibodies. The second antibodies were purified by positive negative affinity chromatography (open symbols).
Symbol
Cont.
w 0 A q 0 A
IO 5 IO IO 10 10
of second
mg/l mg/l mg/l mg/l mg/l mg/l
ab.
Cont.
of unlabelled second (closed symbols) or
of third ab.
10 mg/l 10 mg/l 5 20 15 IO
mg/l mg/l mg/l mg/l
to increasing blank values, due to the presence of free polymeric HRP in the low molecular mass fractions. High blank values prohibit an increase of conjugate concentration. Therefore, preceding the coupling reaction with activated HRP, the polymers should be removed by gel filtration [ 121. For the direct sandwich EIA it is necessary to purify the second antibodies by positive affinity chromatography in order to achieve high sensitivity. The indirect sandwich EIA for HBsAg generally allows an increase of sensitivity which cannot be further improved by the preparative isolation of an antigen specific second antibody. This offers a great advantage compared with the direct assay since the antigen to be determined is often very difficult to prepare in the desired amounts and purity for the immunosorbent technique. On the other hand, the preparation of species specific antibodies by positive affinity chromatography is much less problematic since it is easy to prepare large amounts of IgG for insolubilization from that species donating the second antibodies.
8
The approximately doubled sensitivity of the indirect assay (unlabelled second antibodies) over that of the most sensitive variation of the direct assay is caused by the fact that the second antibodies act as an amplification factor and fully retain their immunological reactivity, having circumvented purification by positive affinity
t GUl5IM3 QXi 012 025 050 100 200 LW corce~mtmof
d
Ellonkmlues
HBSAg f/_g/ll
Fig. 5. Standard curves for HBsAg in the indirect EIA with successive incubation using labelled antibodies (10 mg/l) and labelled third antibodies (1 mg/l), closed circles, and unlabelled antibodies (10 me/l) and labelled third antibodies (1 mg/l), open circles. As second antibodies, anti HBs-IgG were used, purified by positive affinity chromatography.
second second rabbit
chromatography and labelling. An extraordinary increase in sensitivity can be gained with the indirect technique when both the second and the third antibodies are employed in enzyme labelled form. The detection limit can thus be lowered 8-16 times compared with the optimal direct assay technique. This effect is, however, probably restricted to second antibodies with a low degree of labelling (in our case l-2 mol HRP/mol IgG) and low molecular mass marker enzymes as for example HRP or lysozyme, in order to prevent steric hindrance of the binding of the third antibodies. The indirect assay offers the additional advantage of an universal antigen independent applicability of the enzyme labelled anti species specific antibodies. One disadvantage is the necessity to raise antigen specific first and second antibodies in two different species; another, the longer duration of the indirect compared to the direct assay, due to a further incubation and washing step. By preincubation of second and third antibodies in parallel to the antigen incubation (simultaneous incubation technique) and subsequent transfer of the preformed immune complexes into the test tubes, the duration can be reduced to that of the direct assay. The sensitivity of the successive incubation procedure can be approached only by employing second antibodies, purified by positive affinity chromatography and by increasing the concentration of the labelled third antibodies 10 times. The necessity for such a high concentration of conjugate is caused by its ability to react with a much greater amount of second antibodies regardless of the
9
antigen binding by the latter. On the other hand in the successive technique the non reacting second antibodies have been removed by washing prior to the reaction with the third antibodies. it is clear that saturation occurs at higher conjugate concentrations and that, therefore, the simultaneous incubation procedure of the indirect two-site binding assay is uneconomical. The possibility of shortening assay time by simultaneous incubation of antigen and labelled second antibodies, using monoclonal antibodies with a different determinant specificity, together with a high sensitivity and economical antibody utilization, was recently demonstrated for a two-site binding assay of alphafetoprotein [ 131. References
10 11 12
13
Stafford JEH, KiBgallan W. Universal reagent immunosorbent assay (uri) for haptens. J Immunoi Methods 1980; 34: 339-343. Belanger L, Sylvestre C, Dufour D. Enzyme-linked immunoassay for alpha-fetoprotein by competitive and sandwich procedures. Clin Chim Acta 1973; 48: 15- 18. Porstmann T, Porstmann B, Schmechta H, Nugel E, Meisel H, Geserick G. Development of a double sandwich enzyme immunoassay for the quantitative analysis of HBsAg, Part I. Dtsch Gesundheitswes 1980; 35: 598-600. Nath N, Mazzur S, Ledman R, Fang CT. Purification of hepatitis B surface antigen using polyethylene glycol, pepsin and Tween 80. VOX Sang 1976; 31, suppl. 1: 84-89. Porstmann T, Porstmann B, Schmechta H, Nugel F, Meisel H, Geserick G. Development of a double sandwich enzyme immunoas~y for the quantitative analysis of HBsAg, Part 2. Dtsch Gesundheitswes 1980; 35: 1587-1593. Porstmann B, Porstmann T. Reinigung handels~bIicher Meerrettich-Peroxidase fur den Einsatz im Enzymimmunoassay. 2 Med Labor-Diagn 1979; 20: 87-95. Avrameas S, Ternynck T. Peroxidase labelled antibody and Fab conjugates with enhanced intracellular penetration. Immunochemistry 1971; 8: 1175-I 179. Porstmann T, Porstmann B. A simple and effective method for purification of crude horse radish peroxidase (HRP) and HRP-conjugated antibodies. 12th FEBS-Meeting, Dresden 1978. Abstract Nr. 0734. Porstmann T, Porstmann B. Purification of horse radish peroxidase conjugated antibodies by affinity chromatography. Comparison of glutaraldehyde and periodate conjugation. Acta Biol Med Germ 1979; 38: 1039-1054. Gallati H. Peroxidase aus Meerrettich: Kinetische Studien sowie Optimierung der Aktivitatsbestimmung mit den Substraten H,O, und ABTS. J Chn Chem Chn Biochem 1979; 17: 1-7. Persijn JP, Jonker KM. A terminating reagent for the peroxidase-Iabe~led enzyme immunoassay. J Clin Chem Clin B&hem 1978; 16: 531-532. Boorsma DM, Streefkerk JG. Some aspects of the preparation, analysis and use of peroxidase-antibody conjugates in immunochemistry. In: 24th Coil. Protides of the Biological Fluids. H. Peeters, ed., Oxford: Pergamon Press, 1976: 795-804. Uotila M, Ruoslathi E, Engvall E. Two-site sandwich enzyme immunoassay with monocional antibodies to human alpha-fetoprotein. J Immunol Methods 1981; 42: 1I-15.