Biologicals
(1990) 18, 123-126
Quantktation of Pertussis Toxin in an Enzyme Linked Immunosorbent Assay with Improved Specificity t Vaccine Department,
P. Ibsen** and I. Heron-jState Serum Institute, Aflillerivej 5, DK-2300 Copenhagen
S, Denmark
Abstract. The quantitation of pertussis toxin (PT) in two sandwich ELlSAs was tested for specificity. The detection of the captured PT was obtained by using either polyspecific rabbit anti Bordetella pertussis serum (RaBp-ELISA) or a monoclonal anti-PT antibody (McaPT-ELISA). No major differences in the estimation of PT in highly purified preparations were noted using either ELISA variants. In contrast, the quantitation of PT in crude extracts of B. perfussis cultures by the RaBp-ELISA was found to be over-estimated and showed greater variability when compared to the McaPT-ELISA. Comparison of the distribution of PT in the eluate fractions following partial purification by hydroxylapatite chromatography revealed that the results of the McaPT-ELISA were more specific as judged by SDS-PAGE analysis. Introduction Pertussis toxin (PT) is recognized as one of the major exotoxins ofBordetellapertussi and also as one of the major protective antigens of current and future vaccines. PT exerts a variety of biological activities some of which have formed the basis for PT quantitations over the years. These include assessment of the lymphocytosis promoting activity, islet activating activity and histamine sensitizing activity. For rapid and accurate quantitation of IT antigen capture enzyme-linked immunosorbent-assays (ELISA) is of great interest particularly for laboratories involved in production, purification and quality control of pertussis vaccines. We have developed monoclonal antibodies specific for PT and in the present paper such antibodies have been applied in two ELISA modi& cations in order to quantify unknown amounts of PT present in bacterial extracts and samples of PTenriched material derived during the purification process of PT. The concentration of PT is assessed relative to a highly purified PT reference preparation. Materials
and methods
Antibodies against PT Polyclonal rabbit anti-B. pertzmis (RaBp) antibodies were raised by repeated immunixations of the animals with conventional whole cell pertussis vaccine as described by Schou et al.’ * To whom correspondence KM-1056/90/020123+04
should be addressed.
$03.00/O
Murine monoclonal anti-PT antibodies (McaPT) were produced and characterized as described earlier by Schou et aZ.l B. pertussis extracts Cultures of B. pertussis strain 509 grown for 48 h in an 8-litre fermentor were concentrated by ammonium sulphate precipitation (50% saturation) and subsequently extracted using 15 ~I.M sodium phosphate 1 M NaCl pH 8 at 4°C. The extracts were concentrated by ammonium sulphate precipitation and extracted once more as described above and those with the highest PT-specific activity were pooled and stored at 4°C. Before performing the hydroxylapatite chromatography samples of pooled extract were centrifuged at 4000 g for 80 min. Purified P!l’ was a kind gift from Dr C. Capiau, Smith Kline-FUT, Belgium. Enzyme linked immunosorbent assay (ELISA) Two different ELISA systems were used in this study: RaBp sandwich ELISA. Polystyrene microtiter plates (Nunc, Roskilde, Denmark) were coated overnight at 4°C with 0.1 ml per well of purified monoclonal anti-PT antibody (O-5 pg/ml) in a carbonate buffer pH 9-6. The monoclonal antibody used most extensively is an IgGl subclass antibody (designated 20.6) which has toxin neutralizing ability in mice and which recognizes the 53 and 54 subunit of PT in immunoblotting.l In ELISA, McaPT 20.6 ~l99OTheInternationdAt3mciationofBiologicaIStandardization
124
P. Ibsen and I. Heron
reacts with the S2-S4 and S3-S4 dimers and the purified S4 subunit. Two-fold serial dilutions of test samples (0.1 ml/well) were made in dilution buffer consisting of phosphate buffered saline (PBS) 1% (v/v) Triton X-100,1% (w/v) bovine serum albumin pH 7.2 and incubated for 2 h at room temperature. Following this step, O-1 ml of rabbit anti-pertussis antiserum (150) was added and the mixtures were held at 20°C for another hour. Then 0.1 ml of horseradish peroxidase (HRP) conjugated swine anti-rabbit Ig (1:lOOO) (Dakopatts, Denmark) was added and incubation was continued for 1 h. Finally, 0-phenylene diamine was used as substrate and after 30 min the absorbance was read at 492 nm in an Immunoreader NJ 2000 (Tecnunc, Denmark). Between the individual steps the plates were washed in PBS pH 7.2, containing 1% (v/v) Triton x-100. All dilutions were made in dilution buffer. McaPTsandwich ELBA. Essentially, this assay was performed as described for the RaBP-assay with the exception of two steps: First, instead of using rabbit anti-pertussis antiserum a biotinylated monoclonal anti-PT antibody (designated 21.3 D 11) with a fine specificity differing from the coating antibody (0.3 pg/ml) was added. According to immunoblotting analysis McaPT 21.3 Dll reacts with the S2 and S3 subunits of PT. In ELISA the S&S4 and S&S4 dimers and the purified S2 subunit are recognized. The proposed difference in specificity between 20.6 and 21.3 Dll was further supported by the results of competition ELISA’ in which the ability of one McaPT to inhibit the binding of another to PT adsorbed to ELISA plates was studied. Second, HRP-conjugated streptavidin (Zymed Lab. I nc. U.S.A.) diluted 15000 was used instead of HRP-conjugated swine-anti rabbit Ig. Quantitation of PT in ELlSA A two-fold dilution of a proposed World Health Organization reference preparation of PT (JNIH-5) was used as antigen standard in all ELISA assays. Calculation of antigen concentration in unknown PT containing samples were carried out relative to the reference using parallel line assays.2 Hydroxylapatite chromatography Partial purification of PT was achieved by means of hydroxylapatite (HA) chromatography (HA Ultrogel, IBF Biotechnics, France) as described in the following: The HA-column (25 x 20 cm) was equilibrated in
O-01 M sodiumphosphate pH 8 followed by application of 20 ml of pooled extract which was diluted 1:lO in distilled water and adjusted to pH 8.7 using sodium hydroxide. Elution of PT was carried out using either O-5 M potassium phosphate buffer pH 6 (buffer A) or a linear gradient of 0.01-0.7 M potassium phosphate pH 6 (buffer B). Throughout, a flow rate of 50 ml/hour was used. Sodium dodecylsulphate-polyacrylamide gel electrophoresis (SDS-PAGE) The individual PT containing fractions from HA chromatography of B. pertussis extractions were analysed by SDS-PAGE as described previouslys and visualized by silver staining according to the method of Morrissey.4 Results One of the main differences between the results obtained in the two types of PT-ELISA described in this paper is illustated in Fig. 1 which shows the titration of purified PT. As is noted the curve obtained by using the McaPT-ELISA was much steeper and showed a higher maximum level of OD readings before the curve started to decline. Titration of crude PT-containing B. pertussis extracts gave similar results (data not shown). When a highly purified preparation of PT was estimated quantitatively relative to the standard,. only a minor difference in PT content was noted using either ELISA variants. The amount of PT in this preparation was found to be 72.3 t 5.7 and 80.1* 7.3 ,ug/ml (a SEMI according to the RaBp and McaPT ELISA, respectively.
I I I I I I I \q,
2
4
8
I6 32 64 128 Ddution
256
/ 512
, ,
2048 1024 4096
factor
Figure 1. Titration curves of highly purified F’T according to the RaBp-ELISA (O--O) (U).
and the McaFT-ELISA
Quantttation
of Pertussis toxin
125
Table 1. Distribution and recovery of PT (expressed in percentages) following partial purikation by hydroxylapatite chromatography (elution buffer A) as measured by RaBp- and McaPT-ELISA. ND = Not determined. All data are based on a single chromatographic run Experiment
Starting material Run-through/washings Eluate fraction 1 Eluate fraction 2 Eluate fraction 3 Eluate fraction 4 Eluate fraction 5 Recovery
A
RaBp
McaPT
RaBp
100.0 8.7 0.1 3.7 33.8 6.2 0.3 53.6
lOO*O 4.6 0.1 2.6 66.5 19.0 o-5 90.7
lOO*O 5.3 0.1 3.9 57.3 8.3 0.2 75-9
In contrast, measurement of PT in crude extracts using RaBp-ELISA seemed to be over-estimated and showed greater variability (48.8 f 10.2 ,uglmI, f SEM) when compared to the McaPT-ELISA (24.0 f 3.9 &ml, f SEM). In Table 1 the application of the two ELISA variants used for quantitation of PT following partial purification by hydroxylapatite chromatography is described. In this run PT contaminated with several other antigens (Fractions 3 and 4) was eluted using O-5 M potassium phosphate buffer pH 6. The recovery of PT according to the RaBp-ELISA varies between 32 and 76% compared to the recovery of 91-97% found by using the McaPT-ELISA. A higher degree of purity of PT was achieved by eluting PT with a linear gradient of 0.01-0.7 M potassium phosphate buffer pH 6. The recovery of PI’ in four of these eluate fractions as estimated by ELISA is listed in Table 2. According to the RaBp-ELISA the majority of PT was Table 2. Differences in the distribution of PT (expressed in percentages) with a higher degree of purity (elution buffer B) in the HA eluate fractions as measured by RaBp-and McaPT-ELISA. AII data are based on a single chromatographic run Experiment
Eluate, total Fraction A Fraction B Fraction C Fraction D
A
Experiment
RaBp
McaPT
RaBp
100.0 24.3 47.7 23.9 4.1
100.0 59.4 31.9 6.3 2.4
100.0 43.8 38.4 16.0 1.8
Experiment
B
B
Experiment
C
McaPT
RaBp
McaPT
100.0 3.1 0.1 3.0 65.4 23.7 0.4 96-9
100.0 N? .
100.0 Nk6 .
2.2 21.6 4.2 0.1 31.8
1.0 67.0 22.0 0.2 93.4
found in fractions A-C in varying amounts (l&48%) whereas the McaPT-ELISA detected the presence of most PT in fraction A (59~35%) and some in fraction B (2632%). When subjected to SDS-PAGE the distribution of PT in profiles of fractions A-C was found to be in agreement with the results of the McaPT-ELISA as judged by the intensity of staining (Fig. 2). The sensitivity of the two ELISA modifications was found to be similar and PT concentrations above 5 rig/ml were regularly detected and quantified.
sSh s3
s4. ss-
McaPT 100.0 65.3 25.8 6.5 2.4
A
B
C
D
Figure 2. SDS-PAGE profiles of HA e1uat.e fractions (elution buffer B) containing partially purified PT visualized by silver staining. The positions of the individual PT-subunits (Sl through S5) are indicated by bars. Lane 1: fraction A; Lane 2: fraction B; Lane 3: fraction C; Lane 4: fraction D.
126
P. Ibsen and I. Heron
Discussion PT in its toxic form can be quantitated in different biological assays among which the ability to induce leucocytosis histamine sensitization and islet activation are the most widely used. In vitro methods based on ELISA technology have major advantages and also allow quantitation of PT following detoxification, which is of interest for vaccine production. Estimation of PT by the McaPT-ELISA as described revealed at least two advantages compared to the RaBp-ELISA. First, it was found to be less variable from experiment to experiment and second, the study of recovery of PT following purification by chromatography .was found to indicate that this variant has a higher degree of specificity. The second point was based on the apparent distribution of Pl’ in the individual eluate fractions as seen in silverstained profiles of these fractions subjected to SDSPAGE. No major difference in PT content, however, was found when already purified preparations were assayed in the two ELISA variants. The different results obtained and the greater variability seen using the RaBp-ELISA may be due to several factors one of which is the presence of several other pertussis antigens in non-purified preparations which adhere to PT. These mixed antigenic clusters might permit the binding to various degrees of rabbit antibodies with irrelevant specificities. This problem would be particularly marked when PT is being determined, because of the known tendency of PT to bind to carbohydrate residues on glycoproteins. Comparison of two similar ELISA modifications to quantify filamentous haemagglutinin (FHA) revealed no major differences between a polyclonal and monoclonal system. The steeper slope of the dos*response curve of the McaPT-ELISA is also a factor of importance for the precision of this assay.
As the establishment of reproducible and specific assays for the estimation of antigen yields following purification procedures is an essential step in connection with the development of various biological products such as vaccines, the described ELISA modification with a greater specificity might prove to be useful for related purposes. At present, the quantitation using a similar approach to other pertussis antigens such as agglutinogen 2 and 3, 69 kD outer membrane protein and lipopolysacchardie by ELISA is studied. Regarding sensitivity, samples containing 5 ng PT/ml have been assessed reproducibly, which makes this ELISA method. useful for chasing contaminating PT present in other antigenic fractions of B. pertussis material intended for vaccine development. Acknowledgements We would like to thank Ingrid Balle for skillful technical assistance and Carine Capiau, Smith Kline RIT, Belgium, for the preparations of highly purified PT, dimers and subunits. References Schou C, AU-Jensen, M, Heron I. The interaction
between pertussis toxin and 10 monoclonal antibodies. Acta Path Microbial Immunol Stand Sect C, 1987; 95: 177-187. Simonsen 0, Schou C, Heron I. Modification of the ELISA for the estimation of tetanus antitoxin in human sera. J Biol Stand 1987; 15: 143-157. Ibsen P, M&er S, Heron I. Lipopolysaccharides in a traditional pertussis vaccine. J Biol Stand 1988; 16: 299-309. Morrissey JH. Silver stain for proteins in polyacrylamide gels: A modified procedure with enhanced uniform sensitivity. Anal Biochem 1981; 117: 307310.
Received for publication 23 November accepted 25 January 1990.
1989;
(1990) 18, 123-126
Quantktation of Pertussis Toxin in an Enzyme Linked Immunosorbent Assay with Improved Specificity t Vaccine Department,
P. Ibsen** and I. Heron-jState Serum Institute, Aflillerivej 5, DK-2300 Copenhagen
S, Denmark
Abstract. The quantitation of pertussis toxin (PT) in two sandwich ELlSAs was tested for specificity. The detection of the captured PT was obtained by using either polyspecific rabbit anti Bordetella pertussis serum (RaBp-ELISA) or a monoclonal anti-PT antibody (McaPT-ELISA). No major differences in the estimation of PT in highly purified preparations were noted using either ELISA variants. In contrast, the quantitation of PT in crude extracts of B. perfussis cultures by the RaBp-ELISA was found to be over-estimated and showed greater variability when compared to the McaPT-ELISA. Comparison of the distribution of PT in the eluate fractions following partial purification by hydroxylapatite chromatography revealed that the results of the McaPT-ELISA were more specific as judged by SDS-PAGE analysis. Introduction Pertussis toxin (PT) is recognized as one of the major exotoxins ofBordetellapertussi and also as one of the major protective antigens of current and future vaccines. PT exerts a variety of biological activities some of which have formed the basis for PT quantitations over the years. These include assessment of the lymphocytosis promoting activity, islet activating activity and histamine sensitizing activity. For rapid and accurate quantitation of IT antigen capture enzyme-linked immunosorbent-assays (ELISA) is of great interest particularly for laboratories involved in production, purification and quality control of pertussis vaccines. We have developed monoclonal antibodies specific for PT and in the present paper such antibodies have been applied in two ELISA modi& cations in order to quantify unknown amounts of PT present in bacterial extracts and samples of PTenriched material derived during the purification process of PT. The concentration of PT is assessed relative to a highly purified PT reference preparation. Materials
and methods
Antibodies against PT Polyclonal rabbit anti-B. pertzmis (RaBp) antibodies were raised by repeated immunixations of the animals with conventional whole cell pertussis vaccine as described by Schou et al.’ * To whom correspondence KM-1056/90/020123+04
should be addressed.
$03.00/O
Murine monoclonal anti-PT antibodies (McaPT) were produced and characterized as described earlier by Schou et aZ.l B. pertussis extracts Cultures of B. pertussis strain 509 grown for 48 h in an 8-litre fermentor were concentrated by ammonium sulphate precipitation (50% saturation) and subsequently extracted using 15 ~I.M sodium phosphate 1 M NaCl pH 8 at 4°C. The extracts were concentrated by ammonium sulphate precipitation and extracted once more as described above and those with the highest PT-specific activity were pooled and stored at 4°C. Before performing the hydroxylapatite chromatography samples of pooled extract were centrifuged at 4000 g for 80 min. Purified P!l’ was a kind gift from Dr C. Capiau, Smith Kline-FUT, Belgium. Enzyme linked immunosorbent assay (ELISA) Two different ELISA systems were used in this study: RaBp sandwich ELISA. Polystyrene microtiter plates (Nunc, Roskilde, Denmark) were coated overnight at 4°C with 0.1 ml per well of purified monoclonal anti-PT antibody (O-5 pg/ml) in a carbonate buffer pH 9-6. The monoclonal antibody used most extensively is an IgGl subclass antibody (designated 20.6) which has toxin neutralizing ability in mice and which recognizes the 53 and 54 subunit of PT in immunoblotting.l In ELISA, McaPT 20.6 ~l99OTheInternationdAt3mciationofBiologicaIStandardization
124
P. Ibsen and I. Heron
reacts with the S2-S4 and S3-S4 dimers and the purified S4 subunit. Two-fold serial dilutions of test samples (0.1 ml/well) were made in dilution buffer consisting of phosphate buffered saline (PBS) 1% (v/v) Triton X-100,1% (w/v) bovine serum albumin pH 7.2 and incubated for 2 h at room temperature. Following this step, O-1 ml of rabbit anti-pertussis antiserum (150) was added and the mixtures were held at 20°C for another hour. Then 0.1 ml of horseradish peroxidase (HRP) conjugated swine anti-rabbit Ig (1:lOOO) (Dakopatts, Denmark) was added and incubation was continued for 1 h. Finally, 0-phenylene diamine was used as substrate and after 30 min the absorbance was read at 492 nm in an Immunoreader NJ 2000 (Tecnunc, Denmark). Between the individual steps the plates were washed in PBS pH 7.2, containing 1% (v/v) Triton x-100. All dilutions were made in dilution buffer. McaPTsandwich ELBA. Essentially, this assay was performed as described for the RaBP-assay with the exception of two steps: First, instead of using rabbit anti-pertussis antiserum a biotinylated monoclonal anti-PT antibody (designated 21.3 D 11) with a fine specificity differing from the coating antibody (0.3 pg/ml) was added. According to immunoblotting analysis McaPT 21.3 Dll reacts with the S2 and S3 subunits of PT. In ELISA the S&S4 and S&S4 dimers and the purified S2 subunit are recognized. The proposed difference in specificity between 20.6 and 21.3 Dll was further supported by the results of competition ELISA’ in which the ability of one McaPT to inhibit the binding of another to PT adsorbed to ELISA plates was studied. Second, HRP-conjugated streptavidin (Zymed Lab. I nc. U.S.A.) diluted 15000 was used instead of HRP-conjugated swine-anti rabbit Ig. Quantitation of PT in ELlSA A two-fold dilution of a proposed World Health Organization reference preparation of PT (JNIH-5) was used as antigen standard in all ELISA assays. Calculation of antigen concentration in unknown PT containing samples were carried out relative to the reference using parallel line assays.2 Hydroxylapatite chromatography Partial purification of PT was achieved by means of hydroxylapatite (HA) chromatography (HA Ultrogel, IBF Biotechnics, France) as described in the following: The HA-column (25 x 20 cm) was equilibrated in
O-01 M sodiumphosphate pH 8 followed by application of 20 ml of pooled extract which was diluted 1:lO in distilled water and adjusted to pH 8.7 using sodium hydroxide. Elution of PT was carried out using either O-5 M potassium phosphate buffer pH 6 (buffer A) or a linear gradient of 0.01-0.7 M potassium phosphate pH 6 (buffer B). Throughout, a flow rate of 50 ml/hour was used. Sodium dodecylsulphate-polyacrylamide gel electrophoresis (SDS-PAGE) The individual PT containing fractions from HA chromatography of B. pertussis extractions were analysed by SDS-PAGE as described previouslys and visualized by silver staining according to the method of Morrissey.4 Results One of the main differences between the results obtained in the two types of PT-ELISA described in this paper is illustated in Fig. 1 which shows the titration of purified PT. As is noted the curve obtained by using the McaPT-ELISA was much steeper and showed a higher maximum level of OD readings before the curve started to decline. Titration of crude PT-containing B. pertussis extracts gave similar results (data not shown). When a highly purified preparation of PT was estimated quantitatively relative to the standard,. only a minor difference in PT content was noted using either ELISA variants. The amount of PT in this preparation was found to be 72.3 t 5.7 and 80.1* 7.3 ,ug/ml (a SEMI according to the RaBp and McaPT ELISA, respectively.
I I I I I I I \q,
2
4
8
I6 32 64 128 Ddution
256
/ 512
, ,
2048 1024 4096
factor
Figure 1. Titration curves of highly purified F’T according to the RaBp-ELISA (O--O) (U).
and the McaFT-ELISA
Quantttation
of Pertussis toxin
125
Table 1. Distribution and recovery of PT (expressed in percentages) following partial purikation by hydroxylapatite chromatography (elution buffer A) as measured by RaBp- and McaPT-ELISA. ND = Not determined. All data are based on a single chromatographic run Experiment
Starting material Run-through/washings Eluate fraction 1 Eluate fraction 2 Eluate fraction 3 Eluate fraction 4 Eluate fraction 5 Recovery
A
RaBp
McaPT
RaBp
100.0 8.7 0.1 3.7 33.8 6.2 0.3 53.6
lOO*O 4.6 0.1 2.6 66.5 19.0 o-5 90.7
lOO*O 5.3 0.1 3.9 57.3 8.3 0.2 75-9
In contrast, measurement of PT in crude extracts using RaBp-ELISA seemed to be over-estimated and showed greater variability (48.8 f 10.2 ,uglmI, f SEM) when compared to the McaPT-ELISA (24.0 f 3.9 &ml, f SEM). In Table 1 the application of the two ELISA variants used for quantitation of PT following partial purification by hydroxylapatite chromatography is described. In this run PT contaminated with several other antigens (Fractions 3 and 4) was eluted using O-5 M potassium phosphate buffer pH 6. The recovery of PT according to the RaBp-ELISA varies between 32 and 76% compared to the recovery of 91-97% found by using the McaPT-ELISA. A higher degree of purity of PT was achieved by eluting PT with a linear gradient of 0.01-0.7 M potassium phosphate buffer pH 6. The recovery of PI’ in four of these eluate fractions as estimated by ELISA is listed in Table 2. According to the RaBp-ELISA the majority of PT was Table 2. Differences in the distribution of PT (expressed in percentages) with a higher degree of purity (elution buffer B) in the HA eluate fractions as measured by RaBp-and McaPT-ELISA. AII data are based on a single chromatographic run Experiment
Eluate, total Fraction A Fraction B Fraction C Fraction D
A
Experiment
RaBp
McaPT
RaBp
100.0 24.3 47.7 23.9 4.1
100.0 59.4 31.9 6.3 2.4
100.0 43.8 38.4 16.0 1.8
Experiment
B
B
Experiment
C
McaPT
RaBp
McaPT
100.0 3.1 0.1 3.0 65.4 23.7 0.4 96-9
100.0 N? .
100.0 Nk6 .
2.2 21.6 4.2 0.1 31.8
1.0 67.0 22.0 0.2 93.4
found in fractions A-C in varying amounts (l&48%) whereas the McaPT-ELISA detected the presence of most PT in fraction A (59~35%) and some in fraction B (2632%). When subjected to SDS-PAGE the distribution of PT in profiles of fractions A-C was found to be in agreement with the results of the McaPT-ELISA as judged by the intensity of staining (Fig. 2). The sensitivity of the two ELISA modifications was found to be similar and PT concentrations above 5 rig/ml were regularly detected and quantified.
sSh s3
s4. ss-
McaPT 100.0 65.3 25.8 6.5 2.4
A
B
C
D
Figure 2. SDS-PAGE profiles of HA e1uat.e fractions (elution buffer B) containing partially purified PT visualized by silver staining. The positions of the individual PT-subunits (Sl through S5) are indicated by bars. Lane 1: fraction A; Lane 2: fraction B; Lane 3: fraction C; Lane 4: fraction D.
126
P. Ibsen and I. Heron
Discussion PT in its toxic form can be quantitated in different biological assays among which the ability to induce leucocytosis histamine sensitization and islet activation are the most widely used. In vitro methods based on ELISA technology have major advantages and also allow quantitation of PT following detoxification, which is of interest for vaccine production. Estimation of PT by the McaPT-ELISA as described revealed at least two advantages compared to the RaBp-ELISA. First, it was found to be less variable from experiment to experiment and second, the study of recovery of PT following purification by chromatography .was found to indicate that this variant has a higher degree of specificity. The second point was based on the apparent distribution of Pl’ in the individual eluate fractions as seen in silverstained profiles of these fractions subjected to SDSPAGE. No major difference in PT content, however, was found when already purified preparations were assayed in the two ELISA variants. The different results obtained and the greater variability seen using the RaBp-ELISA may be due to several factors one of which is the presence of several other pertussis antigens in non-purified preparations which adhere to PT. These mixed antigenic clusters might permit the binding to various degrees of rabbit antibodies with irrelevant specificities. This problem would be particularly marked when PT is being determined, because of the known tendency of PT to bind to carbohydrate residues on glycoproteins. Comparison of two similar ELISA modifications to quantify filamentous haemagglutinin (FHA) revealed no major differences between a polyclonal and monoclonal system. The steeper slope of the dos*response curve of the McaPT-ELISA is also a factor of importance for the precision of this assay.
As the establishment of reproducible and specific assays for the estimation of antigen yields following purification procedures is an essential step in connection with the development of various biological products such as vaccines, the described ELISA modification with a greater specificity might prove to be useful for related purposes. At present, the quantitation using a similar approach to other pertussis antigens such as agglutinogen 2 and 3, 69 kD outer membrane protein and lipopolysacchardie by ELISA is studied. Regarding sensitivity, samples containing 5 ng PT/ml have been assessed reproducibly, which makes this ELISA method. useful for chasing contaminating PT present in other antigenic fractions of B. pertussis material intended for vaccine development. Acknowledgements We would like to thank Ingrid Balle for skillful technical assistance and Carine Capiau, Smith Kline RIT, Belgium, for the preparations of highly purified PT, dimers and subunits. References Schou C, AU-Jensen, M, Heron I. The interaction
between pertussis toxin and 10 monoclonal antibodies. Acta Path Microbial Immunol Stand Sect C, 1987; 95: 177-187. Simonsen 0, Schou C, Heron I. Modification of the ELISA for the estimation of tetanus antitoxin in human sera. J Biol Stand 1987; 15: 143-157. Ibsen P, M&er S, Heron I. Lipopolysaccharides in a traditional pertussis vaccine. J Biol Stand 1988; 16: 299-309. Morrissey JH. Silver stain for proteins in polyacrylamide gels: A modified procedure with enhanced uniform sensitivity. Anal Biochem 1981; 117: 307310.
Received for publication 23 November accepted 25 January 1990.
1989;
