APMIS 100. 1027-1032, 1992
Enzyme=linkedimmunosorbent assay for determination of antibodies to Vibrio chlerae toxin-coregulated pili JACEK OSEK, GUNHILD JONSON, ANN-MAR1 SVENNERHOLM and JAN HOLMGREN Department of Medical Microbiology and Immunology, University of Gothenburg, Gothenburg, Sweden
Osek, J., Jonson, G., Svennerholm, A.-M. & Holmgren. J. Enzyme-linked immunosorbent assay for determination of antibodies to Vihrio clzolerae toxin-coregulated pili. APMIS 100: 1027-1032, 1992. An ELISA for determination of antibodies to F cholerae TCP was developed. Since purified TCP preparations contained detectable amounts of LPS (as shown by ELISA and immunoelectron microscopy with anti-LPS polyclonal serum), a capture ELISA was used. In this test the plate was coated with anti-TCP monoclonal antibody followed by incubation with TCP fimbriae. By this procedure no LPS bound to the solid phase as shown by the loss of reactivity with anti-LPS serum. The capture ELISA allowed sensitive and specific determination of TCP antibodies in sera of rabbits immunized with classical but not El Tor ! I rholerae strains. There was good agreement between results in the TCP ELISA and reactivity with the TcpA band in immunoblot analyses when antisera raised against classical and El Tor vibrios were studied. Key words: Vibrio cholercie; toxin-coregulated pili; monoclonal antibody; ELISA. Jacek Osek, University of Gothenburg, Department of Medical Microbiology and Immunology, Guldhedsgatan 10, S-413 46 Gothenburg, Sweden.
Vihriu ckolerar 01 strains, both of classical and El Tor biotypes (Inaba and Ogawa serotypes), cause disease mainly, if not exclusively, by producing cholera enterotoxin (CT). This toxin, despite well-defined molecular mechanisms, induces electrolyte and fluid secretion from the small intestine, leading to watery diarrhea, dehydration and, in severe cases, death ( 5 ) . However, colonization of the small intestine is an essential early step in the pathogenesis of cholera. I.: cholrme bacteria produce various factors which may facilitate penetration of the mucus and adherence to the epithelial cells: these include mucinase and other enzymes as well as putative adhesins, e.g. hemagglutinins (9). Of particular interest recently has been the toxin-coregulated pilus (TCP) described by Taylor et ul. (14), which has been shown to be important for colonization
of at least I.I cholerae 01 classical biotype in both humans and in mice (4, 14). Since TCP may also be a potential protective immunogen in K cholerae 0 1 and is therefore being considered for inclusion in future cholera vaccines (2, 12, 13), it is important to obtain a sensitive and specific immunoassay for measuring antibody responses to TCP, e.g. after infection or vaccination. We here report on our development of a capture ELISA using a monoclonal antibody (MAb) against TcpA, the 20.5 kDa pilin subunit of TCP and purified TCP fimbriae as solid phase antigen. We show that the capture ELISA can measure anti-TCP antibodies without any risk of false-positive reactions caused by antibodies to LPS. MATERIALS AND METHODS
Received April 27, 1992. Accepted September 3, 1992.
Bacterial strains and growth conditions Vibrio clzolerae 569Bt 165 strain, a conjugate between V cholerae 569B (classical Inaba) and the non-
1027
OSEK et ul.
01 serogroup vibrio 165, was used. This strain ex-
g for 20 min), the supernatant was carefully collected
presses the non-Ol polysaccharide antigen of strain 165 on the non-LPS antigenic background of 569B ( I ) . The strain was kindly supplied by Dr R. S. Attridge, Adelaide, Australia. Bacteria were grown using the AKI method (6) at 30°C for 16 h.
and the pellet was treated twice with sucrose (as above) yielding supernatants 2 and 3. The three supernatants were pooled and precipitated overnight with salts (6.5 mM Na,HP04, 1.5 mM KH,PO,, 136.9 mM NaCI, 2.7 mM KCI and 0.8 mM NaN,) without agitation. The precipitate was centrifuged first at 10,000 x g for 25 rnin and then again at 25,500 x g for 45 min. The second pellet was used and resuspended with sucrose (152 mM), stirred for 30 min, and centrifuged at 45,000 x g for 45 min. The final pellet was resuspended with sucrose ( I 0 ml volume) and examined for TCP by direct and immune transmission electron microscopy, and for TCP subunit TcpA by sodium dodecyl sulphate - polyacrylamide gel electrophoresis (SDS-PAGE) and immunoblotting, respectively (see below).
Monoclonal antibody ( M A b ) and antisera A monoclonal antibody (MAb) against TcpA raised in our laboratory was used (7). The MAb reacts with TCP-positive bacteria both in a whole cell ELISA and with TcpA in immunoblot analyses. Polyclonal antisera against TCP were produced by immunizing rabbits with a classical V cholerae strain JS1569 (pJS752-3), a derivative of the classical 569B, expressing a high concentration of TCP when grown by the AKI method as described previously (7). The immune sera were extensively absorbed with TCPnegative bacteria of the homologous serotype that had been grown on blood agar at 37°C. For specificity control purposes the antiserum with the highest anti-TCP titer was extensively absorbed with strain JSI 569(pJS752-3)grown under conditions which promoted TCP expression to eliminate anti-TCP antibody activity. Two monoclonal antibodies (MAbs) against lipopolysaccharide (LPS) produced in our laboratory were used. One of them was of IgGl and the other of IgM class; the anti-LPS ELISA titers in tissue culture fluids were 15,000 and 1,500, respectively. Polyclonal antisera against LPS were obtained by immunizing rabbits with the purified V cholerae LPS as described previously (8). Antisera against classical and El Tor biotypes were prepared as described previously (8). The bacteria used for immunization were grown either in vitro under TCP-expressing conditions, i.e. by the AKI method at 30°C or on CFA agar at 2 7 T , or in vivo in intestinal loops of cholera-infected rabbits. The immune sera were extensively absorbed with TCPnegative bacteria of the homologous serotype that had been grown on blood agar at 37°C. In some instances the antisera were further absorbed also with the homologous strains grown under TCP-promoting conditions. Purijication of TCP TCP was purified using the method of Hall et al. (1990) (3) with some modifications. The method is based on differences in aggregation state of TCP in saline and in sucrose. All steps were done at 4°C. Bacterial cells (1000 ml culture volume) were harvested by centrifugation (10,000 x g for 20 rnin), washed with phosphate-buffered saline (PBS, pH 7.2), and sheared by careful passing through an 18-gauge needle. After centrifugation at 10,000 x g for 20 min, the supernatant was discarded, and the pellet resuspended with sucrose (152 mM) to a 200 ml volume and stirred for 15 min. After centrifugation (10,000x
I028
SDS-PAGE and immunoblotting SDS-PAGE was performed using the Laemmli buffer system with a 17% separation gel; proteins were stained with Coomassie Brilliant Blue (10). For immunoblot analyses SDS-PAGE-separated proteins were transblotted to nitrocellulose paper and developed essentially as described (1 5). Immunoelectron microscopy ( I E M ) TCP was identified by using anti-TCP MAb or polyclonal anti-TCP serum; anti-LPS MAb and polyclonal anti-LPS serum were also tested for control purposes. A gold-conjugated anti-mouse IgG or protein A (AuroProbe, Amersham, U.K.) was used. The grids were stained with 1% ammonium molybdate and examined in a JEOL electron microscope (JEOL Ltd., Tokyo, Japan). TCP ELISA Polystyrene microtiter plates (Nunc, Roskilde, Denmark) were coated with purified TCP (1 pg/ml) diluted in PBS, 100 pl per well, at room temperature overnight. After washing with PBS, any remaining binding sites were blocked by incubation with 0.1% bovine serum albumin in PBS (BSA, Sigma, St. Louis, Mo, USA) for 30 min at 37°C. Anti-TCP and antiLPS MAbs and rabbit antisera against TCP and LPS were tested serially diluted in 0.1% BSA-PBS-Tween 20 (0.05%); the antibodies were incubated for 90 min at room temperature. Thereafter, the plates were washed with PBS-Tween, anti-mouse or anti-rabbit horseradish peroxidase-conjugated IgG was added as a secondary antibody (Jackson, West Grove, PA), and incubation continued for 60 min. Antibody titers were determined as the reciprocal dilutions giving an A,,, of 0.4 above background after reaction with the enzyme substrate for ca. 20 min. Capture TCP ELISA In initial experiments the optimal coating concentrations of MAb and purified TCP were determined
ELISA FOR ANTIBODIES TO Y CHOLERAE TCP
by checker board titrations. Based on these analyses, polystyrene microtiter plates (Dynatech, Chantilly, VA) were coated with anti-TCP MAb at a concentration of 0.6 pg IgGl /ml diluted in 0.05 M carbonate buffer, pH 9.6, 100 pI per well, at room temperature overnight. After washing the plates with PBS they were blocked with 1.0%1BSA-PBS. Purified TCP diluted in carbonate buffer to 10 pg/ml was added, 100 pl per well, and the plates were incubated for 90 min. Anti-TCP and other sera were then tested in serial dilutions on the plates as described above for the regular ELISA test.
RESULTS Analyses of purijlied TCP Immunoelectron microscopy studies of the purified TCP revealed the presence of TCP fimbriae specifically stained with anti-TCP MAb as well as rabbit polyclonal antiserum. Additionally, contaminating vesicle-like material reacting with anti-LPS polyclonal antiserum but not with LPS MAb was also seen. When the purified TCP was tested by SDSPAGE, a strong protein band corresponding to 20.5 kDa (TcpA subunit) and some weaker bands were detected (Fig. 1 A). Immunoblot analysis of the TCP material showed a single band at 20.5 kDa when tested with the anti-TCP MAb, and this and several other bands when rabbit polyclonal antiserum was used. When
Fig. 1. SDS-PAGE and immunoblot analyses of purified TCP. Lane I , Coomassie brilliant blue stained; lanes 2, 3, and 4 were developed with anti-TCP MAb, anti-TCP serum, and anti-LPS MAb, respectively; lanes R, low molecular weight references: 97.4, 66.2, 45.0, 31.0, 21.5, and 14.4 kDa, respectively.
anti-LPS MAbs were tested, no reaction with the purified TCP preparation was observed (Fig. 1 B, lane 4). ELISA and capture ELISA Using the purified TCP preparation as a solidphase bound antigen we found that both antiTCP rabbit polyclonal sera and anti-TCP MAb reacted strongly with the TCP antigen, while no reaction with either of the two anti-LPS MAbs was observed (Table 1). However, hyperimmune rabbit antisera against purified K cholerae LPS also reacted in substantial titers with the solidphase bound TCP antigen (titer 7,300 as compared with an anti-LPS ELISA titer of ca 25,000). These results confirmed the IEM examination and indicated that a direct ELISA for determination of antibodies to TCP after e.g. infection or vaccination cannot be used because of the possibility of false-positive results (serum antibody titers against LPS). We therefore tried to establish a specific capture ELISA. An anti-TCP MAb was used as a first coating followed by incubation with the TCP preparation. This capture ELISA allowed us to reduce the anti-LPS titer of the rabbit LPS antisera to an undetectable level (titer < 30) without any corresponding loss in reactivity of TCP antibodies, which were found in high titers in the TCP sera tested (Table 1). The specificity of the test was further assessed by the use of hyperimmune serum against TCP after extensive absorption with a K cholerae TABLE 1. Antibody titers of’ MAbs and antisera against TCP in direct ELISA and caature ELISA Antiserum or monoclonal Antibody titera antibody (MAb) ELISA Capture ELISA Anti-TCP MAb 4,500 NTh Anti-TCP serum 24,300 10,700 Anti-TCP serum absorbed with 810 < 30 strain JS1569 (TCP+) Anti-TCP serum absorbed 24,300 5,600 with strain N16961 (TCP-) Anti-LPS MAb
Enzyme=linkedimmunosorbent assay for determination of antibodies to Vibrio chlerae toxin-coregulated pili JACEK OSEK, GUNHILD JONSON, ANN-MAR1 SVENNERHOLM and JAN HOLMGREN Department of Medical Microbiology and Immunology, University of Gothenburg, Gothenburg, Sweden
Osek, J., Jonson, G., Svennerholm, A.-M. & Holmgren. J. Enzyme-linked immunosorbent assay for determination of antibodies to Vihrio clzolerae toxin-coregulated pili. APMIS 100: 1027-1032, 1992. An ELISA for determination of antibodies to F cholerae TCP was developed. Since purified TCP preparations contained detectable amounts of LPS (as shown by ELISA and immunoelectron microscopy with anti-LPS polyclonal serum), a capture ELISA was used. In this test the plate was coated with anti-TCP monoclonal antibody followed by incubation with TCP fimbriae. By this procedure no LPS bound to the solid phase as shown by the loss of reactivity with anti-LPS serum. The capture ELISA allowed sensitive and specific determination of TCP antibodies in sera of rabbits immunized with classical but not El Tor ! I rholerae strains. There was good agreement between results in the TCP ELISA and reactivity with the TcpA band in immunoblot analyses when antisera raised against classical and El Tor vibrios were studied. Key words: Vibrio cholercie; toxin-coregulated pili; monoclonal antibody; ELISA. Jacek Osek, University of Gothenburg, Department of Medical Microbiology and Immunology, Guldhedsgatan 10, S-413 46 Gothenburg, Sweden.
Vihriu ckolerar 01 strains, both of classical and El Tor biotypes (Inaba and Ogawa serotypes), cause disease mainly, if not exclusively, by producing cholera enterotoxin (CT). This toxin, despite well-defined molecular mechanisms, induces electrolyte and fluid secretion from the small intestine, leading to watery diarrhea, dehydration and, in severe cases, death ( 5 ) . However, colonization of the small intestine is an essential early step in the pathogenesis of cholera. I.: cholrme bacteria produce various factors which may facilitate penetration of the mucus and adherence to the epithelial cells: these include mucinase and other enzymes as well as putative adhesins, e.g. hemagglutinins (9). Of particular interest recently has been the toxin-coregulated pilus (TCP) described by Taylor et ul. (14), which has been shown to be important for colonization
of at least I.I cholerae 01 classical biotype in both humans and in mice (4, 14). Since TCP may also be a potential protective immunogen in K cholerae 0 1 and is therefore being considered for inclusion in future cholera vaccines (2, 12, 13), it is important to obtain a sensitive and specific immunoassay for measuring antibody responses to TCP, e.g. after infection or vaccination. We here report on our development of a capture ELISA using a monoclonal antibody (MAb) against TcpA, the 20.5 kDa pilin subunit of TCP and purified TCP fimbriae as solid phase antigen. We show that the capture ELISA can measure anti-TCP antibodies without any risk of false-positive reactions caused by antibodies to LPS. MATERIALS AND METHODS
Received April 27, 1992. Accepted September 3, 1992.
Bacterial strains and growth conditions Vibrio clzolerae 569Bt 165 strain, a conjugate between V cholerae 569B (classical Inaba) and the non-
1027
OSEK et ul.
01 serogroup vibrio 165, was used. This strain ex-
g for 20 min), the supernatant was carefully collected
presses the non-Ol polysaccharide antigen of strain 165 on the non-LPS antigenic background of 569B ( I ) . The strain was kindly supplied by Dr R. S. Attridge, Adelaide, Australia. Bacteria were grown using the AKI method (6) at 30°C for 16 h.
and the pellet was treated twice with sucrose (as above) yielding supernatants 2 and 3. The three supernatants were pooled and precipitated overnight with salts (6.5 mM Na,HP04, 1.5 mM KH,PO,, 136.9 mM NaCI, 2.7 mM KCI and 0.8 mM NaN,) without agitation. The precipitate was centrifuged first at 10,000 x g for 25 rnin and then again at 25,500 x g for 45 min. The second pellet was used and resuspended with sucrose (152 mM), stirred for 30 min, and centrifuged at 45,000 x g for 45 min. The final pellet was resuspended with sucrose ( I 0 ml volume) and examined for TCP by direct and immune transmission electron microscopy, and for TCP subunit TcpA by sodium dodecyl sulphate - polyacrylamide gel electrophoresis (SDS-PAGE) and immunoblotting, respectively (see below).
Monoclonal antibody ( M A b ) and antisera A monoclonal antibody (MAb) against TcpA raised in our laboratory was used (7). The MAb reacts with TCP-positive bacteria both in a whole cell ELISA and with TcpA in immunoblot analyses. Polyclonal antisera against TCP were produced by immunizing rabbits with a classical V cholerae strain JS1569 (pJS752-3), a derivative of the classical 569B, expressing a high concentration of TCP when grown by the AKI method as described previously (7). The immune sera were extensively absorbed with TCPnegative bacteria of the homologous serotype that had been grown on blood agar at 37°C. For specificity control purposes the antiserum with the highest anti-TCP titer was extensively absorbed with strain JSI 569(pJS752-3)grown under conditions which promoted TCP expression to eliminate anti-TCP antibody activity. Two monoclonal antibodies (MAbs) against lipopolysaccharide (LPS) produced in our laboratory were used. One of them was of IgGl and the other of IgM class; the anti-LPS ELISA titers in tissue culture fluids were 15,000 and 1,500, respectively. Polyclonal antisera against LPS were obtained by immunizing rabbits with the purified V cholerae LPS as described previously (8). Antisera against classical and El Tor biotypes were prepared as described previously (8). The bacteria used for immunization were grown either in vitro under TCP-expressing conditions, i.e. by the AKI method at 30°C or on CFA agar at 2 7 T , or in vivo in intestinal loops of cholera-infected rabbits. The immune sera were extensively absorbed with TCPnegative bacteria of the homologous serotype that had been grown on blood agar at 37°C. In some instances the antisera were further absorbed also with the homologous strains grown under TCP-promoting conditions. Purijication of TCP TCP was purified using the method of Hall et al. (1990) (3) with some modifications. The method is based on differences in aggregation state of TCP in saline and in sucrose. All steps were done at 4°C. Bacterial cells (1000 ml culture volume) were harvested by centrifugation (10,000 x g for 20 rnin), washed with phosphate-buffered saline (PBS, pH 7.2), and sheared by careful passing through an 18-gauge needle. After centrifugation at 10,000 x g for 20 min, the supernatant was discarded, and the pellet resuspended with sucrose (152 mM) to a 200 ml volume and stirred for 15 min. After centrifugation (10,000x
I028
SDS-PAGE and immunoblotting SDS-PAGE was performed using the Laemmli buffer system with a 17% separation gel; proteins were stained with Coomassie Brilliant Blue (10). For immunoblot analyses SDS-PAGE-separated proteins were transblotted to nitrocellulose paper and developed essentially as described (1 5). Immunoelectron microscopy ( I E M ) TCP was identified by using anti-TCP MAb or polyclonal anti-TCP serum; anti-LPS MAb and polyclonal anti-LPS serum were also tested for control purposes. A gold-conjugated anti-mouse IgG or protein A (AuroProbe, Amersham, U.K.) was used. The grids were stained with 1% ammonium molybdate and examined in a JEOL electron microscope (JEOL Ltd., Tokyo, Japan). TCP ELISA Polystyrene microtiter plates (Nunc, Roskilde, Denmark) were coated with purified TCP (1 pg/ml) diluted in PBS, 100 pl per well, at room temperature overnight. After washing with PBS, any remaining binding sites were blocked by incubation with 0.1% bovine serum albumin in PBS (BSA, Sigma, St. Louis, Mo, USA) for 30 min at 37°C. Anti-TCP and antiLPS MAbs and rabbit antisera against TCP and LPS were tested serially diluted in 0.1% BSA-PBS-Tween 20 (0.05%); the antibodies were incubated for 90 min at room temperature. Thereafter, the plates were washed with PBS-Tween, anti-mouse or anti-rabbit horseradish peroxidase-conjugated IgG was added as a secondary antibody (Jackson, West Grove, PA), and incubation continued for 60 min. Antibody titers were determined as the reciprocal dilutions giving an A,,, of 0.4 above background after reaction with the enzyme substrate for ca. 20 min. Capture TCP ELISA In initial experiments the optimal coating concentrations of MAb and purified TCP were determined
ELISA FOR ANTIBODIES TO Y CHOLERAE TCP
by checker board titrations. Based on these analyses, polystyrene microtiter plates (Dynatech, Chantilly, VA) were coated with anti-TCP MAb at a concentration of 0.6 pg IgGl /ml diluted in 0.05 M carbonate buffer, pH 9.6, 100 pI per well, at room temperature overnight. After washing the plates with PBS they were blocked with 1.0%1BSA-PBS. Purified TCP diluted in carbonate buffer to 10 pg/ml was added, 100 pl per well, and the plates were incubated for 90 min. Anti-TCP and other sera were then tested in serial dilutions on the plates as described above for the regular ELISA test.
RESULTS Analyses of purijlied TCP Immunoelectron microscopy studies of the purified TCP revealed the presence of TCP fimbriae specifically stained with anti-TCP MAb as well as rabbit polyclonal antiserum. Additionally, contaminating vesicle-like material reacting with anti-LPS polyclonal antiserum but not with LPS MAb was also seen. When the purified TCP was tested by SDSPAGE, a strong protein band corresponding to 20.5 kDa (TcpA subunit) and some weaker bands were detected (Fig. 1 A). Immunoblot analysis of the TCP material showed a single band at 20.5 kDa when tested with the anti-TCP MAb, and this and several other bands when rabbit polyclonal antiserum was used. When
Fig. 1. SDS-PAGE and immunoblot analyses of purified TCP. Lane I , Coomassie brilliant blue stained; lanes 2, 3, and 4 were developed with anti-TCP MAb, anti-TCP serum, and anti-LPS MAb, respectively; lanes R, low molecular weight references: 97.4, 66.2, 45.0, 31.0, 21.5, and 14.4 kDa, respectively.
anti-LPS MAbs were tested, no reaction with the purified TCP preparation was observed (Fig. 1 B, lane 4). ELISA and capture ELISA Using the purified TCP preparation as a solidphase bound antigen we found that both antiTCP rabbit polyclonal sera and anti-TCP MAb reacted strongly with the TCP antigen, while no reaction with either of the two anti-LPS MAbs was observed (Table 1). However, hyperimmune rabbit antisera against purified K cholerae LPS also reacted in substantial titers with the solidphase bound TCP antigen (titer 7,300 as compared with an anti-LPS ELISA titer of ca 25,000). These results confirmed the IEM examination and indicated that a direct ELISA for determination of antibodies to TCP after e.g. infection or vaccination cannot be used because of the possibility of false-positive results (serum antibody titers against LPS). We therefore tried to establish a specific capture ELISA. An anti-TCP MAb was used as a first coating followed by incubation with the TCP preparation. This capture ELISA allowed us to reduce the anti-LPS titer of the rabbit LPS antisera to an undetectable level (titer < 30) without any corresponding loss in reactivity of TCP antibodies, which were found in high titers in the TCP sera tested (Table 1). The specificity of the test was further assessed by the use of hyperimmune serum against TCP after extensive absorption with a K cholerae TABLE 1. Antibody titers of’ MAbs and antisera against TCP in direct ELISA and caature ELISA Antiserum or monoclonal Antibody titera antibody (MAb) ELISA Capture ELISA Anti-TCP MAb 4,500 NTh Anti-TCP serum 24,300 10,700 Anti-TCP serum absorbed with 810 < 30 strain JS1569 (TCP+) Anti-TCP serum absorbed 24,300 5,600 with strain N16961 (TCP-) Anti-LPS MAb
